e-v.-‘-..-.-__-‘ A 3333’! .m Q . . ' ._. -. . a ’ . « «.3 ‘_ THE ECONOMICS er UGANDA’S 2 _ _ .., 1 HEALTH SERVlCE SYSTEM * . . ~ . , MPUGAHGMS FOR HEALTH AND . -- ; :‘~ ’ _ ,__ Economc PLANNING , ~ ‘ 1 . . . ,3. ' n ' . “a. - V . ,A r . . V l V .r ' ' _ n r . -_H ' ' . ‘ Z. . ~ - . '. . ;' ' , , . ' ( . ' ' . . y , ~ , . for the Degree of Ph D UNIVERSITY Dissertation 0 ”"1 v . A ’0'“ 3 i. i, .iRY ,' ,._..~ _ _ h‘.1;i,‘:')til State ‘ UlllVCfslt)’ .‘m' I | I This is to certify that the I l thesis entitled l The Economics of Uganda's Health Service System: Implications for Health and Economic Planning presented by David Wallace Dunlop has been accepted towards fulfillment of the requirements for Ph.D. ‘degreein Economics. ajor professor 0-7639 , pens. - SPR'MEPHHY mam..- Q OK“ ~-~ 0! U6A»9M‘« a~z¢w 1;. l0! usarr. «we gnnwm 0‘.le Je; . .i '- t . Tuurpooo of [hit o! a. health “- ,,:.‘ A construct-d 1p . a .M ' a Inocatzon -<‘.-.. " mica flyttvm .mt, ‘ ifi L ' .Mco dyl‘tn-e '3“. mm or i'gh: {3‘ ,1 ~i r, ' ‘ ”W160 plfiz’iuil _~ v , rim: . DI available fihw? O .-.; of health g«y.. ‘en “I! with inf; ,. ,_ fi‘ ‘ z';l'\ 4 IHMCM It. pz'nfle'm .. ,~ ‘7: he \ 6,3,, .. a If 5‘ .Ak 'i 0 O V {coal every-run '.;.‘.- ’ N ' 2.31! ‘2“ “”1: 10'- 48:3! mad 2 “MR? ~>!‘.‘»." " .~r~.-‘.'..-,__\.' . ‘..-. ' ‘” “'1th and maternal nus. 31‘ rd twin. ‘WV cw‘ .- ‘I ' ~90 V . N,” m the «91¢:an setting of 1“,“:de “ .. . Jug? ":=?{[.lv‘. ‘ ‘!"‘ ehoanttlcel teen. in I-chnl nacho-doc. ‘7... ABSTRACT ms ECONOMICS OF UGANDA'S HEALTH SERVICE SYSTEM: IMPLICATIONS FOR HEALTH AND ECONOMIC PLANNING By David Wallace Dunlop The major purpose of this study is to analyze the economic implications of the health service system of Uganda. An analytical framework is constructed in order to (a) isolate the economic implica— tions of resource allocation decisions made in the several components of the health service system and (b) analyze the macro—economic impact of the health service system. The thesis focuses on the output of the health service system of Uganda, the differences in the alternative curative health service production processes and the economic implica- tions of the choices available to health and economic planners involved in the development of health services throughout the country. The analysis begins with information as to the economic context in which health services are provided in Uganda. The manpower, financial and institutional characteristics of the health service system are described. Analysis is made of the changing pattern of demand for curative health services, away from the predominance of infect— ions and parasitic diseases toward a demand which increasingly includes health problems related to malnutrition and maternal and child health. The focus of the study shifts from the empirical setting of Uganda's health service system to a major theoretical issue in medical economics. —~ 1.... .. , u.-. ... . . 7 u... --.. .., "‘n. .‘. "‘ I <- 7- P. . -\S n. - a. u.._ .5- “‘c ...' i- ~§ -' u. ‘ N‘ s David Wallace Dunlop analent COnceptualizations of the output of a curative health service system are examined and found lacking. An alternative conceptualization of output is developed, with particular attention to the non—homogeneous and qualitative nature of the output. A methodology is proposed for the development of an empirical measure of health service system output. A linear programing framework is then developed, incorporating the output conceptualization, in order to examine the relationships between resources utilized in the production of health services and the number of successfully treated persons. The framework developed, also accounts for the multi—product nature of Uganda's health facilities. In addition, a methodology is developed for examining the long-run effect of socio- economic and health variables on (a) the resources available for the production of health services, (b) the number of persons, with given age, sex and disease characteristics, demanding health services, and (c) the probability of successful treatment. An empirical comparative analysis utilizing the linear programming framework is made of three subsectors of Uganda's health service system: (1) government hospitals, (2) mission hospitals, and (3) government rural health facilities with inpatient services. Using an objective function derived from the government's stated health objectives, the analysis reveals that despite a common assumption that doctors, nurses and beds are in short supply, considerable excess capacity exists in those factors in both hospital sectors and, to some extent, in rural units as well. The most binding supply constraints appear in such specialized diagnostic resources as lab and radiographic technicians. At the macro level, the analysis focuses on the impact of the health service system on (a) the rate of population growth, (b) the David Wallace Dunlop rate of change in the age structure, (c) the balance of payments, (d) employment and its distribution, and (e) the extent to which the resources used in the delivery of health services were equitably dis— tributed. Findings related to the demographic variables were weak but generally consistent within the theory of demographic transition. Concluding remarks highlight the health and economic policy implications of the study, particularly with respect to Uganda's health manpower training strategy. Future research activities having health policy relevance in Uganda and similar countries are suggested. 7;. a“. ' - .t; or mum's nALTu SERVICE SYSTEM: 7. ‘roa mun AND ECONOMIC PLANNING 1 By David Wallace Dunlap I'QV‘X‘ ’ 1 A DISSSNTATION Submitted to i __ -., “‘3' 11mm State University ' » ~~ ._£u1fillnsnt of the requirements for the degree of _. v" - l I 1).! - ,' 3 i 1"- llu bean to germinate. in 1913 this “ .m has [issued in the intervening pcxiw' in its d-‘uelnpmu. , _ not have been ruwlelcd .111.ch use new: I ”wk“, Sm: not "1‘ pm! ,1" _ um“ gar“, ‘. .1 ..-2 . titties with stile: ':. i .- w. r m Ibo "scarf-x». .- . _’ - 'NVLded we -._r.?" ‘-i J”an this yr .~ ,.o¢aeatsc-.i. on. “lodging .u- .V : " ’ Copyright by ‘50 gratefully at! .{.~ .25., pwmmmm - ' ‘4 Dr. Carl El~LHJ . 1973 .. “.138 obtaining '10? . .mottiun ‘u. 'l e; -. :v ,7 - 17' __ individual out cull-.- ,. .w .. 133' W i m maxim hum lain It ‘. 01 Vtkting, :t-n, .--a -,,1,.-.,;-,.- -. . s We! in «am-orbs,- 11m. . ,~; -. M3! on cerm'u 34H- ' 'ixv . | ~ Jig. . \ ' .t .5. “AM trail“. butt».- n‘cs .: a " W. 8". Vile-3's.“ saunas-go or. UL ”l . - m n mm or. the “meltatlm L, ~ «mm a: distorted“ u mantis \ I ' ”1.1x Pm. .. . . ,i u , - ACKNOWLEDGMENTS 131968 the seeds of an idea began to germinate. In 1973 this thesis w to fruition. Much has passed in the intervening period and my have participated in its development. This project would not have been completed without the never ending assistance of my wife Kathy. She not only performed many editorial and secretarial responsibilities with skill and enthusiasm during the period of writing but was the "perfect" research assistant in Uganda during 1969/70. She also provided me with the moral support needed at all difficult times during this project, as well as through the earlier years of graduate education. Unfortunately the accolades at my command are eager in acknowledging her assistance. I would like to gratefully acknowledge the assistance given me by the mesbers of my dissertation committee, Dr. John Henderson, Chairman, Dr. Carl Liedholm and Dr. Carl Eicher. I particularly appreciate their collective support in obtaining field research assistance from the Mid- western Universities Consortium for International Activities. In addition I appreciate their individual and collective guidance during my years as a graduate student at Michigan State University. During the period of writing, two other colleagues provided par- ticularly insightful consents on certain parts of the dissertation. Dr. Jesse Mason was very helpful in commenting on ideas contained in Chapter Three and Chapter Four. Dr. Derek Byerlee, besides being an excellent golfer, paddleball player and friend, gave valuable assistance on Chapters Four and Five; his help and willingness to serve on the dissertation cor-ice», during the final period of writing the dissertation is gratefully iii Many other people assisted in the development of this document. First, I want to thank the people of Uganda who, through their Govern- ment, provided me with an opportunity to address some of the most important health service delivery problems facing the country. In particular the Ministry of Health, The Ministry of Planning and Economic Development, The Ministry of Regional Administrations, and Makerere University's Makerere Institute of Social Research assisted me in many ways during my field research phase in Uganda, 1969/70. Dr. Rwakihembo of the Ministry of Health deserves a special note of thanks for his interest and assistance. Also I want to acknowledge the assistance and cooperation of Drs. Manche, Kilkarni, Wallace, Masters and Beri as District Medical Officers in conducting various parts of the field research in West Mengo, East Mengo, Busoga and Ankole Districts. Many other Ugandans working in hospitals and rural health facilit- ies throughout the country helped me in many ways. Although I can not name them all, there are several who deserve special thanks: Mr. Murari, Medical Assistant Kinoni Health Center, Ankole District; Mr. Kaliisa, Ministry of Health Statistician; Mr. Kijambu, Assistant Hospital Secre— tary, East Mengo District; Mr. Rukuba Head Accountant, Ministry of Health; and Mr. Ziraba Medical Assistant, Namwendwa Health Center, P Busoga District. A special word of thanks is extended to Henry Nganwa who worked as an interviewer and field research assistant during the school term breaks. He not only conducted his research work with diligence and care, but also made many valuable comments which materially improved the research. He also helped make Uganda a completely enjoyable and lelorable experience. The field research assistance of Mr. Aloysius Bikwaisse is also very much appreciated. ”flit-i iv ' ‘1""5' I want to acknowledge also the early assistance provided by Dr. J. Gales, who had just completed an excellent survey of Uganda's Health services for W.H.O. prior to my arrival. His successor, Dr. Van der Hoeven also helped me while in Uganda. Dr. Glynn; W.H.0. Advisor, provided useful suggestions concerning the research and was helpful in acquiring governmental cooperation for the research effort. Three doctors, F. J. Bennett, D. Minkler and G. Saxton, who were teaching at Makerere Medical School in Kampala provided many insights about the medical system in Uganda. Not only did they act as inter— preters for me when I was unclear ab0ut the technical points of health care, but they also gave me useful critiques of many ideas found in the dissertation. Their dedication to improving the health of the people of Uganda impressed me and I feel as though I obtained an informal degree in preventive health by interacting with them. Dr. Michael Thuriaux, in charge of the Ankole Preschool Protection Programme, not only gave guidance in terms of his medical knowledge and a willingness to discuss and challenge my ideas, but also opened his home on numerous occasions to my wife and me while we conducted research in Ankole district. Dr. Roy Billington of the Protestant Mecical Bureau, Dr. D. Doyle of the Catholic Medical Bureau and Jane Hallway, (now Mrs. Jane Jones) a VLS.0. volunteer working on a study of the protestant mission hospitals with Dr. Billington, freely gave me information and a much improved understanding of the role of mission health facilities in Uganda. I would also like to acknowledge the fine cooperation I received from Dr. Haddel and Mr. Jerry Reiner of Kagando Mission Hospital and Dr. Krausand Dr. Mortensen of Ishaka Mission Hospital. Dr. Irving Gershenberg, Dr. Mark Haskell, Dr. John Dawson, and Dr. Bartell Jenson of the Economics Department at Makerere University provided a professional critique of my ideas while in a formative stage and their support is appreciated. I would also like to acknowledge the intellectual stimulation given by Dr. Maurice King, Dr. John Bryant and Dr. Martin Feldstein at various stages during the completion of the dissertation. The research project on which the dissertation is based would not have occurred without the generous financial field research support of the Midwestern Universities Consortium for International Activities. I would also like to thank the Departments of Economics and Agricultural Economics at Michigan State University for the generous support given in the form of graduate assistanceships, computer time, instructorships, and computer programming assistance. Special thanks for a job well done and many a potential headache avoided to Mrs. Judith Stephenson of the Agricultural Economics Computer Programming Section. Increasingly, persons with her expertise are the unsung heros of many research projects. The other unsung heros are the secretaries who have read the many crossovers, misspelled words and unintelligible equations and tables contained in the myriad drafts and redrafts of these pages. Mrs. Linda Owenby, Mrs. Graciela Abkin, and Mrs. Michelle Pratt deserve big bouquets for their respective efforts in this regard. I want to reiterate one final acknowledgment. Kathy, thank you. vi Chapter .k TABLE OF CONTENTS I. THE ROLE OF A HEALTH DELIVERY SYSTEM IN A LESS DEVELOPED COUNTRY . . . . . . . . . . . . . . (A) (B) (C) The Importance of Health in Government Priorities: Sub-Saharan Africa The Case of Uganda . . . . . . . . . . . (1) Economic Background . . . . . . . . . . (2) Ugandan Development Policy Related to Health. (3) Uganda's Health Status. . . . (4) The Health Service System in Uganda . . . . (5) Uganda's Financial Commitment to Health Services. . . . . . . . . . . . . Summary . . . . . . . . . . II. THE HEALTH SERVICE SYSTEM IN UGANDA . (A) (B) Preventive Health Services . . . . . . . . (1) Environmental Health. . . . . . . (2) Immunization Services . . . . . . (3) Ante-Natal Services . . . . . . . . . . (4) Young Child Services and Health Education . (5) Family Planning Services. . . . . . . . . . . Curative Health Services. . . . . (1) Health Service Facilities . . . . . . . . . . (a) Number of Facilities . . . . . . . . (b) Distribution of Facility Types . . . . . (c) Size of Facilities . . . . . . . . . . (2) Estimated Expenditures on Health Services . (3) Expenditures and Employment in Three Repre— sentative Government Health Facilities. (4) Demand for Health Services. . . . . . Page 12 15 18 21 21 21 22 25 26 28 30 30 3O 30 33 35 39 41 III. IV. (C) (a) Attendances. . . . . . . . . . . . . . . (b) Disease Mix. . . . . . . . . . . . . . . (i) Inpatient Treatment. . . . . . (ii) Outpatient Treatment . . . (iii) Comparison of Disease Mix Between Inpatient and Outpatient Treat— ment and Government and Mission Hospitals. . . . . . . . . . . . . (5) Employment. . . . . . . . . . . . . (a) Total Employment. . . . . . (b) Ministry of Health Establishment. (c) Registered Medical Manpower . . . . (d) Employment in Rural Health Facilities . Summary. . . . . . . . . . . . . . . . . . . . . THE DEVELOPMENT OF AN OUTPUT CONCEPT FOR ANALYSIS OF CURATIVE HEALTH SERVICES. . . . . . . . . . . . . (A) (B) (C) (D) (E) Review of Past Conceptual Development. . . . . . . Further Conceptual Development . . . . . . . . . Output Specification for the Health Services Firm. Specification of the Output of Each Treatment Process. . . . . . . . . . . . . . . The Issue of Output Homogeneity. . . . . . . . . . A THEORETICAL FRAMEWORK FOR ANALYZING THE HEALTH SERVICE SYSTEM OF UGANDA. . . . . . . . . . . . . . . (A) (B) (C) Statement of the Problem and 3 Consideration of Alternative Methodologies. . . . . . . . . . . . Production Process of the Health Service System. . Linear Programming Model . . . . . . . . . . . . . (1) Assumptions . . . . . . . . . . . . . . . . . (2) Constraints Related to the Production Process (3) Significance of the Technical Coefficient; of the Second Set of Elements of the Vector e o o e a o a o e o e e o o e e o e e e (4) Objective Function of the Health Service syst- O O O U I . . O O O O C I I I C I I C O (5) Specification of the Quality Objective. . . . viii 41 45 45 49 51 54 54 56 58 58 62 69 72 76 80 83 91 91 94 100 100 100 104 105 109 (D) (E) (a) Graphical Example of the Analysis. . . . (b) A Summary of the Quality Specification in the Objective Function. . . . . (6) Cost of Curative Health Services . Dynamic Factors Affecting the Model . . . . . . . (1) Factors Affecting the Vector of Initial Demanders. . . . . . . . . . . . (2) Factors Affecting the Rate of Successful Treatment. . . . . . . . . . . . . . . (3) Factors Affecting the Service Providing subset of the Input Vector . . . . . Summary . . . . . . . . . . . . . . . EMPIRICAL TESTING OF THE LINEAR PROGRAMMING MODEL OF THE HEALTH SERVICE SYSTEM IN UGANDA AND THE FACTORS AFFECTING THE OUTPUT AND RESOURCES AVAILABLE FOR DELIVERING HEALTH SERVICES IN THE FUTURE . (A) (3) Empirical Specification of the Model. . . . . (1) The Model. (2) The Variables. (3) The Data . . . . . . . . . . . . . . . . . . (4) Procedures Used to Determine the Value of the Elements of the Input Vector and Tech- nological Matrix . (a) The Input Vector . . . . . . . . (b) The Technological Matrix . . . . . (i) The Service-Providing Input Sub— matrices. . . . . (ii) The Service—Demanding Input Sub— matrices. . . . . . . . . . (c) The Objective Function . . . . . . . . . Presentation of Empirical Results of the Linear Programming Model . . . . . . . . . . . . . . . . (1) The Results. . . . . . . . . . . . . . . . . (2) Policy Implications. . . . . . . . . . . ix 112 117 119 120 123 125 126 129 136 136 136 139 141 142 143 146 146 148 152 152 152 161 (C) An Empirical Exploration into Factors Affecting the Output of the Health Service System and Re- sources Available for Delivering Health Services in the Future . . . . . . . . . . . . . . . . (1) Factors Affecting the Total Number of Persons Demanding Service. . . . . . . . (2) Changes in Disease Mix of the Initial Demanders. . . . . . . . . (3) Factors Affecting the Rate of Successful Treatment. . . . . . . . . . . . . . . . (4) Factors Affecting Selected Service Providing Inputs . . . . . . . . . . . . . . . . . . (a) Drugs. . . . . . . . . (b) Recurrent Budget . (D) Projected Solution of the Linear Programming Model for the Year 1980/1 . . . (1) Methods and Procedures Used to Project the Input Vector to 1980/81. . . . . (2) Linear Programming Results for 1980/81 . (E) Summary . THE MACRO-ECONOMIC EFFECTS OF UGANDA'S HEALTH SERVICE SYSTEM . . . . . . . . . . . . . . . . . . . . . . . . (A) Health and Demographic Change . (1) Health Effects on Population Growth. . . (2) Effects of Population Growth on Health . . . (3) Health and Population: The Ugandan Case . (a) Population Growth and Health Service Availability . . . . . . . . (b) Changes in the Age Structure Related to Health Service Availability. . . . . (B) Balance of Payments and Foreign Exchange Considerations. . . . . . . . . . . . . . . . (1) Imports of Drugs and Related Items . . . . (2) Imports of Building Materials and Other Capital Goods for Health Services. . . . . 163 164 169 172 175 176 177 179 180 183 189 197 197 198 198 200 201 212 218 218 220 (C) (D) (3) Non-Trade and Capital Account Considerations Health Services Effect on Employment. . . . . . . (1) An Interindustry Analysis of the Relationship Between Output and Employment. . . . (2) Secondary Employment Effects of the Health Services Industry. . . . . . . . . . . . . . (3) The Geographical Distribution of Health Ser— vice Employment in Rural Areas . . . . . . . Equity Considerations in the Distribution of Health Services . . . . . . . . . . . . . . . . (1) The Distribution of Expenditures on Health Services . . . . . . . . . (2) A Comparative Analysis of the Distribution of Health Resources. . . . . . . . . . . . VII. SUMMARY AND CONCLUSIONS. . . . . . . . . . . . . . . . APPENDICES I '1..; (A) (B) (C) (A) (B) (C) Conceptual Developments . . . . . . . . . . Policy Implications . . . . . . . . . . . . . . (1) Specific to the Health Service System. . . . (2) Macro-Economic Policy Implications . . Comments on an Unfinished Agenda. . . . . . . . Classification of Health Units in Uganda. Administrative Relationships in Uganda's Health Service System. . . . . . . . . . . A Further Elaboration of the Variables, Methods and Procedures Used in Chapter Five . . . . . . (1) The Data, Methods, and Procedures Used in Estimating the Elements of the Technological Matrix, the Service-Providing and Service- Demanding Input Vector . . . . . . . 224 228 228 229 232 239 239 242 256 257 258 258 262 263 267 270 275 _ —'_‘-'v (2) The Data and Procedures Used in Specifying the Factors Affecting the Output and Re— sources of the Health Service System in Uganda . . . . . . . . . . . . . . . . . . (3) Summary of Linear Programming Solution for Uganda Government Hospitals, Mission Hospitals and Government Rural Units for 1968/69 Where the Constraint of a Minimum of One— Half of Every Type of Initial Demander Must be Treated . (4) The Supporting Tables. . (D) Other Supporting Tables . (1) Chapter Two. . (2) (3) (E) Notes to Tables Included in the Text. (F) Classification of Diseases, Survey Forms, and Uganda Government Medical Forms . . . . . BIBLIOGRAPHY. Chapter Five . Chapter Six. 282 284 297 298 303 311 315 325 339 xii ‘ ‘v‘ n. .m- t... .vn. ..—\~— - ,‘-.- y ... 5 1 if ‘ ‘I .. .;‘ ¢ .‘ . . ‘§ 0 ‘O K . .- i? ‘ h 'A a . s: \- ‘K \— , ‘ ,- . .‘~‘L. \ ‘- s ..‘ .._ - LIST OF TABLES Table Page 1.1 Population Estimates and Proportion of Government Expenditure Allocated to Health in Selected African Countries . . . . . . . . . . . . . . . . . . . . . . . 3 1.2 Projected Capital Expenditures on Health in Selected African Countries During the Development Plan Period Indicated . . . . . . . . . . . . . . . . . . . . . . . 4 1.3 Uganda's Development Expenditures During the Second and Third Plans: 1966-1971 and 1971/72 - 1975/76. . . . 10 2.1 Percentage of Total Budgeted Establishment of the Ministry of Health to Hygiene and Sanitation. . . . . . 23 2.2 Number of People Receiving Various Types of Immuniza- tions, 1966/67. . . . . . . . . . . . . . . . . . . . 24 2.3 Ante-Natal (Pre-Natal) Services . . . . . . . . . . . . 26 2.4 Distribution of Governmental Health Facilities in Uganda. . . . . . . . . . . . . . . . . . . . . . . . . 31 2.5 Distribution of Voluntary Health Facilities in Uganda . 31 2.6 Number of Government and Voluntary Hospital Beds and Average Number of Beds per Hospital in Uganda . . . . . 34 2.7 Number of Government Rural Health Facilities and Average Number of Beds per Facility in Uganda . . . . . 34 2.8 Estimated Total Expenditure on Health Services in Uganda: 1959, 1963/64, 1968/69. . . . . . . . . . . . . 36 2.9 Distribution of 1968/69 Expenditures on Health Between Preventive and Curative Services. . . . . . . . 38 f | 2.10 Financial and Employment Structure of Three Representa— f tive Governmental Health Facilities for the Year 1968/ 69 or 1969. . . . . . . . . . . . . . . . . . . . . . . 40 Utilization of Governmental Health Facilities . . . . . 42 Attendance Data From a Selected Sample of Rural Units for the Time Period 1969 or 1969/7O . . . . . . . . . . 44 Distribution of Diseases Treated in Uganda on an In- patient Basis in Government Hospitals; Percentage of Total Cases by Disease Category in Selected Years . . . 46 xiii 2.14 2.15 5.6 5.7 Distribution of Diseases Treated in Uganda on an In— patient Basis in Mission Hospitals; Percentage of Total Cases by Disease Category in Selected Years . Distribution of Diseases Treated in Uganda on an Out— patient Basis in Government Hospitals; Percentage of Total Cases by Disease Category in Selected Years . Distribution of Diseases Treated in Uganda on an Out— patient Basis in Mission Hospitals; Percentage of Total Cases by Disease Category in Selected Years Comparison of Disease Distributions . . . . . . Employment in Uganda's Health Services Industry . Ministry of Health Establishment. . . . . . . . Medical Manpower Registered to Practice in Uganda . . Employment Structure of Four Types of Rural Health Facilities in Uganda: 1969/70 . . . . . . . . . . . Summarization of Output and Input Variable Specifica— tion. . . . . . . . . . . . . . . . . . . . . 1968/69 Service Providing Input Constraints for the Three Sectors of Uganda's Health Service System . . 1968/69 Initial Demanders Input Constraints for the Three Sectors of Uganda's Health Service System . . Estimates of the Rate of Shccessful Treatment, (aj , for the Outpatient Treatment Process. . . . . Estimates of the Rate of Successful Treatment, (3—), for the Inpatient Treatment Process . . . . Summary of Linear Programming Solution for Uganda Government Hospitals, Mission Hospitals and Government Rural Units for 1968/69 . . . . . . . . . . . . . . Proportion of Total Initial Demanders Comprised by Successfully Treated Cases, by Sector . . . . . . Relative Proportion of Slack for Each Service Providing Input by Sector of Uganda's Health Service System in 1968/69, Given the Linear Programming Solution. . Estimates of the Annual Rate of Change in the Distribu— tion of Diseases Treated on an Outpatient and Inpatient Basis in Government and Mission Hospitals . . . . . . xiv 48 50 52 53 55 57 59 60 138 144 145 150 151 153 154 160 171 5.10 Summary of Linear Programming Solutions for Uganda Government Hospitals Mission Hospitals and Government Rural Units for 1980/81 . . . . . . . . . . . . . . . . 184 5.11 Relative Proportion of Slack for Each Service Pro— viding Input by Sector of Uganda's Health Service System in 1980/81, Given the Linear Programming Solution. . . . . . . . . . . . . . . 186 6.1 Test of Differences Between Mean Population Growth Rates for Groups of Subcountries in Uganda, Grouped According to the Type of Health Services Available. . . 203 6.2 Test of Differences Between Mean Population Growth Rates for Groups of Subcountries in Uganda, Grouped According to the Type of Health Services Available and Adjusted for Migration and Undercounting. . . . . . . . 207 6.3 Results of O.L.S. Regression Analysis of the Relation- ship Between Population Growth and the Availability of Curative and Maternal Health Services in Uganda . . . . 213 6.4 Test of Differences in Mean Changes in Age Structures Between Subcounties in Uganda, Grouped According to Health Service Availability . . . . . . . . . . . . . . 214 6.5 Results of O.L.S. Regression Analysis of the Relation- ship Between Changes in the Age Structure and the Availability of Curative and Maternal Health Services in Uganda . . . . . . . . 215 6.6 Trends in Uganda's Importation of Drugs and Related Medical Supplies and Equipment. . . . . . . . . . . . . 219 6.7 Estimated Quantity of Imported Items for Health Facility Construction . . . . . . . . . . . . . . . . . 221 6.8 Estimated Proportion of Imported Construction Materials Used in the Construction of Health Facilities . . . . . 223 6.9 Health-Related Financial Transactions Affecting Uganda's Balance of Payments Position in 1969 . . . . . 225 6.10 Incremental Output Employment Ratio (IOER) and Elastic- ity of Employment With Respect to Output for Uganda, According to Industrial Sector. . . . . . . . . 230 Inter-Industry Linkages of the Health Service Industry and Estimates of the Secondary Employment Impact in 1968/69 . . . . . . . . . . . . . . . . . . 233 Proportion Health Service Employment is of Total Employment in Uganda by Districts, Towns, and Regions in 1968 e o a o o a a o o o o o a o e I 236 XV w———_,_ , , A.l A02 C.4 C's C.6 C.9 CI 10 C.ll Curative Services by Facility Type. . . . . . . . . Preventive and Other Health Services by Facility Type . A Reconciliation Between the Eleven Manpower Input Categories Used in Chapter Five and the Health Occupa— tional Titles Used in the Ugandan Health Service System Adjustments Made to the Availability of Service Pro- viding Inputs for Each Sector of Uganda's Health Service System. . . . . . . . The Allocation of Service Providing Inputs According to the Diagnostic Health Service System, 1968/69 . . . . Average Diagnostic Time for Each Major Disease Category Treated on an Outpatient Basis at Mpigi Health Center and Kajans; Sub- dispensary, West Mengo District, Uganda September and October 1970. . . . . Average Length of Inpatient Stay for Each Major Disease Category Treated on an Inpatient Basis in Each Sector of Uganda's Health Service System 1968/69 . . . . . . Proportion of Surgical Services Consumed in the Treat— ment of Each Major Disease Category in Each Sector of Uganda's Health Service System in 1968/69 . . . . . . . Proportion of Laboratory Services Consumed in the Treatment of Each Major Disease Category in Each Sector of Uganda's Health Service System 1968/69 . . . . . . . Proportion of X—Ray Services Consumed in the Treatment of Each Major Disease Category in Each Sector of Uganda's Health Service System 1968/69. . . . . . Average Cost of Drugs and Medical Supplies Consumed in the Treatment of an Initial Demander in Each Major Dis— ease Category and in Each Sector of Uganda' 5 Health Service System in 1968/69 . . . . . . . . Elements, a, of the Diagonal Submatrix for the Govern- ment Hospital Sector in Uganda in 1968/69 . . . . . Elements, a, of the Diagonal Submatrix for the Mission Hospital Sector in Uganda in 1968/69. . . . . . . . . Elements, a, of the Diagonal Submatrix for the Govern— ment Rural Unit Sector in Uganda in 1968/69 . . . . Summary of Linear Programming Solution for Uganda Government Hospitals, Mission Hospitals and Government Rural Units for 1968/69: A Minimum of One-Half of Each Type of Initial Demander Must Be Treated. . . . . . . xvi 268 269 284 285 286 287 288 289 290 291 292 293 294 295 296 D.3 D.‘ D.5 D.6 D08 D.9 D.lO D.11 Health Facilities in Uganda . . . . . . . . . . . . . . Selected Indices on Size and Structure of Uganda's Health Service System . . . . . . . . . . . . . . . . . Structure of Attendances at Uganda Government Health Facilities. . . . . . . . . . . . . . . . . . . . . . . Number of Attendances at Government Health Facilities . Technological Matrix for Uganda Government Hospitals 1968/69 . . . . . . . . . . . . . . . . . . . . Technological Matrix for Uganda Mission Hospitals 1968/ o e e o o u a o o a a o o o e a e Technological Matrix for Uganda Government Rural Units with Beds 1968/69 . . . . . . . . . . . . . . . . . . Results of the O.L.S. Regression Analysis of the Factors Affecting the Total Number of Initial Demand— ers . . . . . . . . . . . . . . . . . . . . . . . . . Results of the O.L.S. Regression Analysis of the Factors Affecting the Rate of Successful Treatment. . . Results of the O.L.S. Regression Analysis of the Factors Affecting the Availability of Drugs . . . . . Results of the O.L.S. Regression Analysis of the Factors Affecting the Ugandan Governments Recurrent Health Budget . . . . . . . . . . . . . . . . . . . . . Results of the O.L.S. Regression Analysis of the Demand for Employees in Uganda Using 1960 GDP Data. . Results of O.L.S. Regression Analysis of the Demand for Employees in Uganda Using 1966 GDP Series Data. . . Estimated Health Expenditures by District in Uganda: 1968/69 . . . . . . . . . . . . . . . . . . . . . International Classification of Diseases. . . . . . . 299 300 301 302 304 305 306 307 308 309 310 312 313 314 326 LIST OF FIGURES Figure 1.1 Uganda's Population Distribution: 1969 (Map). . . . 1.2 Distribution of Health Facilities in Uganda (Map) . . . 1.3 Uganda: Central Government's Expenditures on Health Expressed as 3 Percentage of Total Government Expenditures. . . . . . . . . . . . . . . . . . . . . . 1.4 Uganda: District Administration Government's Expendi- tures on Health (R&C) Expressed as a Percentage of Total District Government Expenditures (R&C). . . . . 4.1 Sub- -dispensary and Aid Post Outpatient Treatment Process . . . . . . . . . 4.2 Dispensary, Dispensary-Maternity Unit, Health Center Treatment Process . . . . . . . . . . . . . . . . . 4.3 Hospital Treatment Process. . . . . . . . . . . . . . 4.4 Relationship Between Output and Quality in the Product- ion of Health Services. . . . . . . . . . . 5.1 Conceptual Specification of the Linear Programming Model of Uganda's Three Sectors of the Health Service System. . . . . . . . . . . . . . . . . . . 6.1 Relationship Between Health Services and Population Growth in Uganda. . . . . . . . . . . . 6.2 Relationship Between Health Services and Population Growth in Uganda Adjusted for Migration and Under- counting. . . . . . . . . . . . . . . . . . . 6.3 Demographic Transition and the Approximate Location of the Northern and Eastern Regions of Uganda in That Transition in 1969.. . . . . . . . . . . . . . . . 6.4 The Distribution of Expenditures on Health Services: A Lorenz Curve Analysis . . . . . . . . . . . . . . . . . i I I The Distribution of Health Services Resources: A Lorenz Curve Analysis . . . . . . . . . . . . . . . . Page 13 16 17 95 96 98 114 137 204 208 240 243 Ol'rwles Ann smors “-41:10 to the “"l "a; care [or c m " aydwls are used in the Tables. a. I neighlm 1! Magnitude zero, or less than one-half the unit employed. Wrative uh Tt‘ PROLOGUE "It is to the common man that this analysis has relevance.... It is in his care both medically and otherwise, that the command to care for one's neighbor, the humanistic ideal to make the most of mankind, and the biological common sense not to let the infinite potential of our species lie wasted, all unite in one final and compelling imperative—-do all you can--ACT." Maurice King, Medical Care in Developing Countries London: Oxford University Press, 1966, Epilogue. CHAPTER ONE During the 1960's, many sub-Saharan African countries allocated a substantial portion of budget expenditures to health services, amounting to as much as ten percent of the total in some cases. In a number of the countries, expenditures on health comprised one of the five largest items in the national budget. The allocation of resources to health in several developing African countries has thus been considerable. In this thesis, analysis is made of the health service system of Uganda, with particular reference to (a) the microeconomic tradeoffs within the health service delivery system and (b) the contribution, at the macro level, of the health service system to the country. Within this context, the study focuses on the development of an analytical framework based on an improved conceptualization of the out— put and production process of the health service system. A second major focus is to be found in the empirical application of the conceptual and theoretical ideas developed. The first and second chapters of the theses provide general back- ground concerning Uganda, its health problems and its health service delivery system. A conceptual framework for the analysis is developed and presented in Chapters Three and Four. Empirical application of the conceptual and theoretical ideas presented is undertaken in Chapter Five. In Chapter Six, the discussion shifts from micro—level analysis of the health service system per se, to an analysis fo the important macro-inte actions between the health service system and other sectors of the economy. The final chapter summarizes the findings and proposes further related research activities which are seen as contributing to a more complete understanding of a complex social phenomenon. The Importance of Health in Government Priorities: Sub—Saharan Africa As mentioned above, many sub-Saharan African countries allocated between five to ten percent of total budget expenditures during the 1960's to health (Table 1.1). In the majority of cases, expenditure on health was one of the five largest items in the government budget, even in instances where the expenditures in health were only five percent. When the countries are classified according to population size, an interesting pattern emerges. In the larger countries, the proportion of total government expenditures allocated to health tended to be significant— ly lower (p <0.001) than in the medium and smaller countries.1 There are at least two plausible explanations for this difference. First, the larger countries tend to have larger military and internal security obligations than do smaller countries which tends to reallocate resources away from social services such as health services. Second, economics of scale undoubtedly exist in a health service delivery system. This fact tends to manifest itself through the development of less expensive facilities (and the use of paraprofessionals) throughout the country after the development of a large hospital complex in the capital city.2 There is a large inter-country variability in the proportion of pro— jected development expenditures on health (Table 1.2). In some countries, less than three percent of projected plan expenditures have been allocated . to health. In other countries, such as Sierra Leone, Uganda, and Ghana, a relatively high proportion of the total projected budget has been allooened to health. Although it is difficult to know the extent to 3 Table 1.1 Population Estimates and Proportion of Government Expenditure Allocated to Health in Selected African Countries (all data pertain to the mid-1960's) Estimated Population 2 Health Expenditures mm __ in Millions is of Total Govt . Egg . (countries: population > five million) Algeria 11.0 4.6 Cameroon 5 , 2 7 . 7 Ethiopia 22 . 7 6 . 3 Ghana 7 . 5 4 . 7 Kenya 9.4 4.8 Madagascar 6. 1 8. 8 N188?“ 57 . 5 5 . 4 South Africa 18.2 3. 7 Sudan 13.1 5.4 (1) Tanzania 11. 7 6. 3 Uganda 7. 8 6. 3 United Arab Republic 28. 7 4.1 (countries: population between one and five million) Burundi 3. 2 8 . 0 Central African Rep. 1. 3 7. 8 Chad 3. 3 9 . l Dahomey 2 . 3 12 . 7 Liberia 1 . 0 9 . 4 Malawi 3. 9 5 . 8 Mali 4 . 6 11 . 9 Mauritania 1. 0 6 . 7 Rwanda 3. 2 8. 0 Senegal 3. 5 8. 0 S. Rhodesia 4.1 5,4 Togo 1.7 9.0 Tunisia 4 . 2 10 . 4 Upper Volta 4.8 10.4 Zambia 3.6 4 . 7 (2) (countries: population < one million) Botswana 0-5 3.5 Comoro Islands 0.2 11.1 Congo (Brazzaville) 0. 9 7 . 4 Gabon 0.5 5,4 Gambia 0. 3 ‘ 8 . 1 Lesotho 0.9 9.6 Mauritius 0 . 8 9 . 9 Seychelles 0.1 12.5 Swaziland 0- 4 9 . 4 Sources: Demographic Yearbook (New York: United Nations, selected years); lthigpian Statistical Abstract, 1967/68 (Addis Ababa: Government Printer, 1968); Demographic Yearbook (New York: United Nations, selected years). (1) Intergovernment transfers are excluded. Central government expenditures only. 4 Table 1.2 Projected Capital Expenditures on Health in Selected African Countries During the Development Plan Period Indicated Projected health (1) Development Expenditures as a Z of Country Plan Period Total Planned Expenditures Cameroon 1961-1965 6.6 1966—1971 2 5 Chad 1966—1970 3.5 Cony)(Brazzaville) 1964—1968 2.4 Efluopia 1968-1973 1.4 Gabon 1966—1970 0 . 8 Ghana 1963—1970 7.0 1970-1971 6.0 Kenya 1966-1970 1.6 1970-1974 0.3 Nigeria 1962-1968 2.5 1970-1974 5.3 (2) Senegal 1965-1969 2.0 Sierra Leone 1963-1972 20.3 Smmfli Republic 1963-1967 3.0 mman 1961/62—1970/71 3.0 Tanzania 1964-1969 2.0 Uganda 1966—1971 12.0 1971/72-1975/76 5.7 Sources: Development Plans of individual countries. (1) Health is defined narrowly here and excludes expenditures on (a) water supplies, (b) housing, (c) social services, (d) various agricultural improvements, and (e) funds allocated separately ‘ ‘ for population control activities. (2 v governments. Includes total public capital investment by federal and state C which projected expenditures actually occur, it can be said that the projected commitments to health are relatively high in a number of sub— Saharan African countries, given that there are other competing claims 3 t1 resources. The Case of Uganda Economic Background Like most sub-saharan African countries Uganda's economy is based on agriculture. The production of coffee and cotton are the main cash crops and provide a substantial share of foreign exchange earnings. In addition, a major portion of the economically active population are engaged in food cash production. There are more than 1.3 million estimated small farmers in the country out of a total population in 1969 of 9.5 million. The spatial distribution of Uganda's population is shown on the accompanying figure (1.1). Since attaining independence in 1962, the government of Uganda has taken an active leadership role in the economic development process. Although recent political disruptions have left the country's short run economic picture in disarray, the government has systematically planned for the socio-economic development of the country. This planning effort as spelled out in three development plans since independence -- the first from 1962-66, the second from 1966-1971, and the third from 1971/72 - 1975/76 -- has laid the foundation for governmental encouragement of industrial development, mining, transportation, communications, electric power, education, and health. The rate of growth of GDP at factor cost has fluctuated over the years, depending upon the output of the world prices for cotton and coffee. Figure 1.1 Ugandan Population Distribution: 1969 _,... Qu'mu -_ \3 \ ' v_.. . ., .\_— ' ' vain .caov' \unmnnu- '1 . ,1 I?" o 9 _ ..,_\°‘.' _¢:~-rocAu9-"' jg.‘ , ,.. ',, '—FF4--— ‘I. u _ p-—~—.—--.—-— ——.—— -.—-__.—.— ——.— _1.. .. - -.. .' . - s i 3 I \ '\".RURI\L and URBAN E \ P O PU LAT I OH I_ fDISTI’HBUTIONHQSQ : ,-' _ _;_ 5000 I :- ""' smut. PERSON; I "1' POPULATION In TOWNS out 2000 ' INNAIITANTS \ Source: The Republic of Uganda, Uganda's Plan III, Third Five-Year Development Plan 1971/72 — 1975/76, (Entebbe: The Government Printer, 1972) From 1954 to 1965, the monetary economy grew at an average annual rate of 4.22 and from 1966 to 1970, the first four years of the second develop— ment plan, it grew at 4.81 per year.4 The rate of economic growth for the economy as a whole-including both the monetary and subsistance sectors during the last four years of the 1960's averaged 4.42 as com— pared to the projected target rate of growth for the economy during those years of 6.32 per year. During the period of the Third Plan, it is pro- jected that the monetary economy will grow at an annual rate of 5.6% from 1969 to 1976, with non-monetary activities (subsistence food product- ion) growing at an estimated rate of 3.6%, giving an overall growth rate of 5.01.5 As referred to earlier, Uganda's economy is heavily dominated by agriculture. Throughout the 1960's the agricultural sector comprised between 502 and 55% of total GDP (including both the monetary and sub— sistence sector), and the third development plan projects a similar figure through 1976-6 Agriculture is important to Uganda's development from at least three other perspectives. First, it has consistently accounted for about 85% of the exports Outside of the East African Common Market area. Second, it has provided employment and income for approx- imately 90! of the population. In addition, large farms have provided about 201 of all wage jobs in the monetary sector, making it the second largest sector for wage employment, after services. Finally, the agricultural sector is closely tied to industrial development in the country. Agricultural processing industries such as coffee curing, cotton ginning, textile production, and the processing of cooking oils, sugar, and tea are all ranked in the top seven industries in the country, as measured by their contribution to gross output.7 Thus, it is not "Flll'.’l"ffi"" 8 surprising to find that the overall rate of growth of the industrial sector depends on increasing agricultural output. Attempts have been made to diversify agricultural production away from the primary crops of coffee and cotton, toward tea, tobacco, and sugar production. During the Second Plan period, tea production expanded rapidly, primarily due to the successful introduction of small holder tea programs in the Western highland areas. During the Third Plan period, ,the tea program, as well as new projects in tobacco and sugar, are expected to result in rapid output increases.8 In addition, substantial efforts are to be made to increase food and livestock production. With its strong commitment to rural development during the third five-year plan, not only in terms of increasing total output, but also in terms of improving the distribution of social amenities and the standard living in rural areas, it appears that the government's expecta- tions of economic and social prospects for development is increasingly based on the realities facing the country. Ugandan Development Policy Related to Health The role of health in Uganda's development has been given high priority in recent years. This is reflected in the country's two recent planning strategies, Work for Progress (1966-71 and Plan III (1971—76). During the second development period (1966-1971) the development strategy "aimed to change the structure of the economy so as to lessen its depend— 9 The campaign to develop the economy ence on the existing export crops." had "three spearheads": (l) agricultural development; (2) industrializa- 10 , tion; and (3) expansion and improvement of education and health services. The government's concern for the third "spearhead" — the improvement of education and health services - was manifested during the 1966-71 plan A 9 by a combined expenditure of 380.6 million shillings, which comprised approximately 182 of all development expenditures during the period, with health receiving slightly more than half of the total, (191.3 million shs.) See Table 1.3.11 During the third five year planning period 1971/72 — 1975/75 govern— ment's concern for health has continued; although its priority, in terms of the proportion of the total development expenditures, has declined from about 9.12 to 5.72. The absolute expenditure is estimated to remain constant (183.5 million shillings), however if inflation is taken into consideration, this figure represents only 80% of the second plan's expenditure on health services. This decline in total expenditures can be explained by the construction of twenty—two 100 bed hospitals during the second development plan, while the third plan emphasises improvment of rural health facilities such as health centers and training more health workers. In addition, a substantial increase is projected for two preventive health programs: water supplies (to a level of 159.8 million shs.) and population control (a nominal 1.0 million shs. allocated from government funds). The development of rural areas clearly has high priority during Plan III. By its statements in the Plan, the government recognizes that (a) its resource endowment requires the development of rural areas and (b) rural living conditions, including health services, must be improved in order to increase agricultural production and to minimize the rate of rural-urban migration. Uganda's Health Status The health at Uganda's population has improved markedly over the last seventy years. In the early years of its protectorate status, the ‘ é I 1 I i l 10 Table 1.3 Uganda's Development Expenditures During the Second and Third Plans 1966-1971 and 1971/72 - 1975/76. Defense Public works, incl. housing Agriuclture incl. Animal industry Health Water Supplies Health, including water supplies Education Contributions to Govern- ment Corporations All other Development Expenditures Total Actual Expenditures Second Plan (1966-71) mill. shs. Z of total 243.5 11.5 355.3 16.8 321.3 15.2 191.3 9.1 84.7 _QLQ 276.0 iggi 189.3 9.0 339.3 16.1 384.8 18.2 2109.5 100.0 Estimated Expenditures Third Plan 1971/72—1975/76 mill. shs. Z of total 500.0 (1) 15.6 491.9 15.4 493.6 15.4 183.5 5.7 159.8 _§,g 343.3 10 7 235.3 7 4 624.2 19.5 572.9 17.9 3200.0 100.0 Source: Republic of Uganda, Uganda's Plan III, Third Five Year Development Plan, 1971/72-1975/76, Government Printer, Entebbe, 1972, Tables VIII-2, VIII-4, and VIII-5. (1) estimated minimum figure. No precise figures are provided in the plan but this figure may be arrived at by reviewing the text of the plan between pp. 116-120. L... _ _"‘I'IIaI'IIE5E-"—"""""""""""""""""""""""""'"__________'_____—___'_ ! s 11 country had a devastating epidemic of sleeping sickness, which contributed heavily to an absolute decline in population. Since then, however, sleeping sickness, as well as other infective and parasitic diseases such as smallpox, yaws, and meningitis have been largely controlled. The extent of this control was revealed in part in preliminary analyses of the 1969 census, which indicated a decline of 20% in infant mortality from about 150 per thousand to 120 per thousand over the previOus ten year period. At the same time overall life expectancy at birth is now well above 40 years, and is as high as 46 years in some areas.12 Infectious and parasitic diseases, such as malaria, tuberculosis, measles, gonorrhoea, and helminthic (worm) disease, are the major health problems today; such diseases comprise at least 25% of sickness episodes experienced by Ugandans. In addition, a number of respiratory illnesses are common, as are alimentary diseases (e.g., dental problems, gastro— enteritis, etc.), diseases of the skin (e.g. tropical ulcers), complica— tions of pregnancy, and miscellaneous infections, injuries and accidents. Subclinical malnutrition is also related to the contraction of infectious diseases, such as measles and upper respiratory infections. The major health concerns of the more developed countries of the world, such as heart disease, related circulatory problems, and cancer, do not posa a serious threat to the welfare of the majority of Ugandans. Although some diseases are prevalent throughout the country, there are important regional differences in the composition of disease dis- tribution.13 For example, in areas of the country above 5,000 feet in elevation (e.g., the southwestern part of the country) the incidence of malaria is quite low, as the survival of the mosquito vector is sub- stantially more difficult at higher elevations. Cultural patterns also ‘ 12 affect the distribution of disease; gonorrhoea, for example, comprises approximately 5% of all diseases treated at rural health facilities in the eastern region, whereas the figure in other regions is not above 3%, and in some areas is less than 0.5%. Economic differences between various areas of the country affect the distribution of disease, too; for example differences in diet, housing conditions, sanitation and waste disposal affect the incidence of malnutrition, tuberculosis, and helminthic (worm) diseases. In sum, the differences in disease dis- tribution play an important role in any analysis of the health service system, as well as in planning for the future development of the health system. The Health Service System in Uganda The curative health service system in Uganda is characterized by a number of different types of health facilities, as well as several administrative structures through which services are delivered (Figure 1.2). The government provides curative services, without charge, in hospitals, health centers, dispensaries, sub-dispensaries, maternity centers and aid posts. The Catholic and Protestant Church Medical Bureaus also provide curative health services through hospitals, sub- dispensaries, and maternity centers for a small fee. Curative health facilities are also operated by large commercial firms for employees and their families. The type of facility maintained by the firm is determined primarily by legal requirement: firms employing more than 1,000 persons must have a hospital, whereas smaller firms may either operate a dispensary or contract with a private physician for service as required. The Army and Prisons also offer curative health services to their specialized populations through dispensaries and, in the case ‘ us’r' 13 Figure 1.2 Distribution of Health Facilities in Uganda ’ \ l . \ ,0 ' x was; ‘ who». 3'! ...._...t':,!"."" 38.3; V" - .\ aumu I .. I \y- - meant-own. R Gevcmmcm Nsspvun V ‘ nmmsrl :‘:rr\;: "up fiféfimy Ga 3‘ 15 Dunnsury I Mum-.3, an: DIspcnsnvy Sub-U-spcnsu, AM Post Mussxon Um! PHVfllI Induny Hrsz” 0,: I Oo-onhlooo I .. D .3 n n 2 : tuna D! J Gel", VINO III-t Mullh Sun-u- fun" 0 ID Source: S. A. Hall and B. W. Langlands, eds., Uganda Atlas of Disease Distribution, Occasional Paper No. 12 (Kampala: Department of Preventive Medicine and Department of Geography, Makerere University College, 1968) 14 of the Army, a hospital. Finally, there are a number of private practitioners in the larger cities and towns who provide a range of curative services to those willing to pay. Government health facilities are integrated in such a way that an individual may be referred to a facility providing more intensive care or treatment than that offered by the facility originally attended. It is theoretically possible for an individual who initially attends a weekly outpatient clinic in a rural aid post to eventually receive treatment at Mulago Hospital in Kampala, the country's national referral, teaching, and research hospital. In addition, private physicians, mission facilities and other population-specific facilities may refer individuals to government facilities for certain specialized services. The most common referral relationship, however, exists between rural government health facilities and government district hospitals. Preventive health services in Uganda are usually provided by local governments - district administrations, municipalities and townships. Environmental health services such as sanitation, waste disposal, vector control, and clean water supplies are administered by special health manpower, headed by the health inspector. Other preventive services, such as ante-natal clinics, young child clinics, and immunizations, are usually delivered through weekly clinics held at local health facilities. The central government also supports an immunization team, which travels throughout the country and conducts daily immunization clinics, and in one district, the preventive services of static facilities are supplimented by a mobile health team, which brings immunization, young child, ante— natal, and health education services to 30 different locations in the area one day each month. ‘ 15 Uganda's Financial Commitment to Health Services Uganda has maintained a fairly large development commitment to health services for some time (Figure 1.3). From the mid—1930's to the present, the central government has consistent allocation a minimum of 6.52 of the total recurrent and capital budget to health services during years of minor capital improvements.14 In addition to this central government commitment, local government expenditure on health has increased substantially in recent years. Since 1947, the percentage of total District Administration expenditures allocated to health has risen from approximately 3.5% to nearly 20% (Figure 1.4). The first upward shift (1956/57) was related directly to the implementation of the so—called Frazer Report, one of whose main recommendations was the improvement of rural health services.15 The second major shift occurred near the time of Independence, October 9, 1962; the major cost increase at that time was due to shifts in power and political relationships between the central and various local governments.16 Finally, the launching of the second five-year plan in 1966/67 gave emphasis to the expansion of health services. This expansion occurred not only in hospital facilities, but also in rural health facilities, such as health centers and dispensaries. The combined expenditures of local govern- ments on health services increased from 22 million shs. in 1965 to 35 million shs. in 1970, in spite of a large decline in expenditures re— corded in Buganda district. As a share of total local government expenditures, health services increased during this period from 82 in 1966 to approximately 202 in 1970. Given the present policy of develop- ing at least one sub-dispensary grade health facility in every sub- .cmuns in the country as soon as possible, it would appear that spending '. s t,- A 16 Hm me we me me we mm wm mm \ON \wo we \qo \No 00 mm om am an mm am me we me me He mm um mm mm HmoH muse» o H wousuficsoaxm unmacuo>oo HmuOH N we smoucwouwm m we wmmwmunxm suamom do woususpsomxm m.usoasuo>oo Hmuusoo m ”spasm: m.H wuswwm a AHMuHQMU was usounsummv mmuSqucoexm .u>oo Hooch mo N m mm Ausoupsome m mousufipawmxm Loans: o n w a UH Aamuaaeo use unusuauomv mouauwvconxw .u>oo Hmuos mo N a an AHeuHmmu use usouusuamv nu mmusuaumomnm space: “H 2 nouauwm omnueoouom s 17 Figure 1. 4 Uganda: District Administration Government's Expenditures on Health (R & C) Expressed as a Percentage of Total District Government Expenditures (R & C) Percentage Figures HHHHP‘HHHHHNN o hare ea.» U10‘N m\o Old HNWJ—‘U‘O‘mGo Years 47 49 51 53 54 54/ 56/ 58/ 60/ 62/ 64/ 66/ 68/ 70/ 55 57 59 61 63 65 67 69 71 Note: R & C = Recurrent and Capital Budget Expenditures 18 on rural health services will expand rapidly in the foreseeable future, not only for capital expenditures, but also for increased recurrent requirements which result from high levels of capital spending in health.17 Finally, it is important to point out that approximately 75% to 80% of expenditures on health services in Uganda have been directed toward curative as opposed to preventive services. Given that such a large proportion of present resource commitment is allocated to curative ser— vices, it is important to analyze in some detail this sector of the health service system in Uganda. Summary This chapter has presented a brief overview of the rationale for an economic analysis of the health service system in a less developed country such as Uganda. The importance of research on the economic aspects of the health service delivery system in sub-Saharan African countries was emphasized. Basic socio-economic background to the case study of Uganda was presented. Its economic status and development policy as related to health were discussed. In addition, the status of the country's health, its health service system, and the governmental financial commitment to health were reviewed for their importance to the subsequent analysis. 19 Footnotes The results of the analysis are as follows: Mean Z Large countries, more than five million population (n=12) 5.68 Medium countries, population between one and five million (n—15) 8.49 Small countries, population less than one million (n=9) 8.54 Liberia can be cited as a good case in point of this latter develop- ment with the John Kennedy hospital complex being developed in Monrovia during the 1960's and only in the last year or so has rural health facilities been seriously considered. Unfortunately there is little comparative information available from developed countries or from other developing countries in other regions - Asia and Latin America - on capital expenditures in health. The World Health Organization in its Monthly Statistical Bulletin started in 1968 to annually publish figures on recurrent health expenditures by governments. However, there are no inter- national data available on either private recurrent health expendi— tures or capital expenditures irrespective of source private or public. The only international comparable data available on re— current and capital expenditure on health are found in Able—Smith, Brian, An International Study of Health Expenditure and its Relevance for Health Planning, Public Health Papers #32, World Health Organ— ization, Geneva, 1967. In Chapter 3, pp. 40-75, he systematically analyzes recurrent and capital health expenditure data collected in a special WHO survey conducted in 1963. From his sample of 29 countries he found a statistically significant positive relationship between per capita income and total recurrent health expenditures as a proportion of national income (pp. 44) but no statistically significant relationship was apparent between per capita income and total capital expenditures on health as a proportion of national income (inferred from data shown in Chapter 3, especially tables six and twenty; and discussion on pp. 67 & 68. See Republic of Uganda, Uganda's Plan 111: Third Five Year Develop- ment Plan 1971/72 - 1975/76 (Entebbe: Government Printer 1971), Table II - 1, p. 28; Republic of Uganda, Background to the Budget 1970/71 (Entebbe: Government Printer, 1970), p. 4; the Republic of Uganda, The Real Growth of the Economy of Uganda, 1954-1962 (Entebbe: Government Printer, 1964). Republic of Uganda, Uganda's Plan III, Table III — l, p. 42. Republic of Uganda, Uganda's Plan III, Table III - l, p. 42. Republic of Uganda, Background to the Budget, 1968/69 (Entebbe: Government Printer, 1968), p. 15. 10. ll. 12. 13. 14. 15. 16. 17. 20 See Republic of Uganda, Uganda's Plan III, Chapter 10, for the specific proposals developed for implementation during this period. Republic of Uganda, Second Five Year Development Plan, 1966-1971: Wbrk for Progress (Entebbe: Government Printer, 1966), paragraph 1.63, p. 14. Republic of Uganda, Work for Progress, paragraph 1.65, p. 15. If the separately listed expenditures on water supplies is included as a health expenditure the estimated total expenditure during the period was 465.3 million shs. or 22% of total plan expenditures. These figures also do not include the developmental expenditures incurred by local governments on health facilities. Republic of Uganda, Uganda's Plan III, p. 301. The single most important publication on th epidemiology of dis- eases in Uganda is S. A. Hall and B. W. Langlands, eds., Uganda Atlas of Disease Distribution, Occasional Paper no. 12 (Kampala: Department of Preventive Medicine and Department of Geography, Makerere University College, 1968). This is seen by noting the difference between the upper line, which aggregates both recurrent and capital expenditures on health, and the lower line, which shows the percent which recurrent health expenditures comprise of total government recurrent and capital expenditures. Uganda Protectorate Government, Report of the Committee to Examine Medical and Health Services in Uganda (Frazer Committee Report), (Entebbe: Government Printer, 1956). For an analysis of this period in the political history of Uganda, see David Apter, The Political Kindom of Buganda: A Study of Bureaucratic Nationalism (Princeton, New Jersey: Princeton Univer— sity Press, 1961). See Republic of Uganda, Work for Progress, paragraph 13:14; and Republic of Uganda, Uganda's Plan III, p. 122. It was announced in the third development plan that rural health facilities will by interated with the government hospitals, forming a single health service system to be administered by the Ministry of Health. This policy is a major shift toward a centralized health system. CHAPTER TWO In this chapter, the salient features of the health service system in Uganda are examined. The institutional, financial and manpower resources of the system are considered, particularly since they cour prise the primary mechanisms through which government policy can operate. The health services system has two sub—sectors: the preventive and the curative. These sectors are related to one another in that (a) both have‘a similar goal - improvement of the "level of health" in the population and (b) both are based on the assumption that the effectiveness of each in working toward this goal can reduce the de- mand for the others sector's services. Each sector is distinct, how- ever, in its approach to achieving the similar goal. Preventive services are directed primarily at reducing the causes of disease whereas curative services are related primarily to the treatment of persons already afflicted with disease. Preventive Health Services Environmental Health In Uganda, preventive health services include malaria vector control, water purification, sewage treatment, air pollution reduction, food cleanliness and meat inspection, building and land use standards, occupational health, and transportation safety. Within urban areas,1 such services are normally provided by the town's department of public health. Water supplies are ordinarily administered separately and occupational and transportation safety are administered by the central 21 22 government. In rural areas, environmental health comes under the jurisdiction of the district medical department; with the exception of market cleanlineSS, meat inspection, and improvement of water sources. Environmental health services provided in rural areas are limited however. Although financial and manpower matters related to the provision of all health services are discussed later, it is useful to note here the government's resource commitment to environmental health services. Approximately 34 million shs. were spent on preventive health services in 1968-69; of that sum, approximately 13.5 million shs. (6% of all expenditures onhealth) were spent on environmental health.2 Approx— imately 92 of the total Ministry of Health's budgeted positions from 1966/67 to 1970/71 were engaged in environmental preventive health services. (Table 2.1). Although a decline in establishment positions allocated to hygiene and sanitation is discernable over this period, an increasing number of hygiene and sanitation personnel were being hired directly by district and town health departments, indicating that the responsibility for such services is shifting-away from the central government to local governments. Immunization Services Immunization programs have focused on eradication of specific infective or parasitic diseases such as smallpox, yaws, polio, diptheria, tentanus, whooping cough, tuberculosis and, to a minor extent, measles. In recent years, both poliomyelitis and smallpox have been the focus of major immunization campaigns.3 As a result of the national immunization program, the incidence of several diseases has fi-I- 'll‘ - . I, 23 been reduced to a very low level, particularly in areas of the country where a high proportion (802-902) of total population has been im— munized. In certain districts of the country, (e.g., Ankole and Busoga) there are very few reported cases of yaws, polio, whooping cough, smallpox, diptheria and typhoid; this is a direct result of an immunization program which has been in effect since the early 1950's. It is believed that some progress is also being made in reducing the incidence of tuberculosis among young people through administration of BCG vaccinations, although there still exists a large reservoir 4 of infection among inadequately or non treated persons. Table 2.1 Percentage of Total Budgeted Establishment of the Ministry of Health to Hygiene and Sanitation. Year Total Establishment Percentage to Hygiene and Sanitation 1966/67 3,364 positions 11.0 1967/68 3,774 positions 10.9 1968/69 3,954 positions 8.3 1969/70 4,360 positions 7.5 1970/71 5,132 positions 6.9 Source: Draft Estimates of Recurrent Expenditures, (Entebbe: Uganda Government Printer), selected years. Immunization services are primarily provided by government non- hospital units. (Table 2.2). In 1966/67, for example, most immuniza- tions were provided either through specialized mobile clinics 24 Table 2.2 Number of PeOple Receiving Various Types of Immunizations, 1966/67 (1) Facility Type Govt. Health Non Govt. Immunization Units Excl. Z of Units Excl. Government Total Type Hospitals(2) Total Hospitals (3) Hospitals (4) (5) DPT Dose 1 172,312 92.3 4,194 10,631 187,137 DPT D086 2 77.337 90.7 2,356 5,522 35,215 DPT Dose 3 41,831 89.0 1,670 3,477 46,978 Smallpox 305,598 96.2 1,767 9,449 316,814 BCG 98,085 88.5 2,855 9,899 110,839 Polio Dose 1 502,137 95.0 7,918 17,823 527,878 Polio Dose 2 226,102 93.7 3,762 12,017 241,881 Polio Dose 3 143,128 93.3 2,727 7,720 153,575 (1) (2) (3) (4) (5) The figures represent the number of pe0p1e who received the particular type of vaccine or dose during 1966/67. It is likely that a number of persons receiving doses 2 or 3 of Polio or DPT also received the first dose during the year. A number of per- sons may also have received at least one other type of immuniza- tion during the year. The data were complied from Ministry of Health records and from Dr. J. Galea's "Inventory, Appraisal, and Assesment of the Basic Health Services of Uganda, Developments for a Malaria Eradication Programme", (Jinja, Uganda: Malaria Pre-eradication Programme, World Health Organization 1967), Appendix IVE. The figures pre- sended here represent approximately 902 of all vaccinations given, inasmuch as some health facilities failed to submit returns for 1966/67. Data taken from Gales Basic Health Services, Appendix IVJ 1. Data taken from Galea Basic Health Services, Appendix VF(i). Immunization se“vices were also provided at non-governmental hos- pitals, but the figures are not available. The number of immuniza- tions administered at such facilities, however, is so small relative to the total number provided by the other three types of facilities cited above, that the total figures given in the Table give a relatively accurate indication of the magnitude of the immunization services offered in the country. 25 conducted by the National Immunization Team or static rural health facilities--hea1th centers and dispensaries. Financial support for immunization services has come primarily from (a) the central government through the Ministry of Health, (b) international organizations such as W.H.0., UNESCO/UNICEF, and OXFAM (a British foundation) and (c) national foundations, such as the former A. M. Obote Foundation, which supported the polio immunization program. The exact amount of support for immunization services is not known, primarily because the extent of support provided by the Ministry of Health is not made clear. Ante-natal Services Ante-natal (pre-natal) services have been provided for some time in Uganda. Recently however, ante-natal clinic attendances have increased rapidly. This is due in part to the rapid growth in the number of health facilities providing such services. Presently, all government and mission hospitals, all governmental health centers and dispensary/maternity units, and most rural units Operated by mission groups conduct at least one ante-natal clinic per week. Where the supply of maternity beds and mid-wifery services is minimal, one of the major functions of the ante-natal clinic is to determine the re- lative probability of delivery complications. If a patient is deter— mined to be "at risk", she is advised to deliver at the health facility; if she is not "at risk", she may deliver at the facility, and is provided with simple sterile implements to help protect against infection in case she delivers at home. 26 The expansion of ante-natal services from the late 1940's to 1966-67 is documented in Table 2.3. Total attendances have increased at an annual rate of 4.0%; new attendances have increased even more rapidly - 5.72 per year, which is nearly double the estimated annual rate of increase of the population over the period. In 1950, only 452 of all pregnant women attended ante-natal clinics. By 1967, ap- proximately 65-702 of pregnant women attended an ante-natal clinic at least once prior to delivery.5 Table 2.3 Ante-natal (Pre-natal) Services Estimated (2) New Cases ‘Total New as a Z of Old Births Cases Total Cases Total ngr (in thous.)_ (in thousg) Births (in thOUSL) (in thous.) 1949 213.3 98.5 46.1 211.3 309.8 1952 229.1 113.5 49.6 163.7 277.2 1956 252.9 150.7 59.6 236.2 386.9 1960/61 320.6 156.6 48.8 235.6 392.1 1963/64 (1) 358.6 198.6 55.4 290.3 488.9 1965/66 386.4 276.1 71.5 465.5 741.6 1966/67 401.1 259.9 65.0 351.5 611.4 Source: Annual Reports or Statistical Reports from the Uganda Ministry of Health. Notes: (1) Figures for Buganda Region, although included, are incomplete. (2) A new case refers to the first visit a woman makes to .a particular clinic for a new pregnancy. For each new pregnancy, the woman is counted as a new case. Young Child Services and Health Education Young child and health education services are often provided simutaneously. Although an estimate of the exact nunber of attendances 27 at young child clinics is not available, there are probably as many attendances at such clinics as there are at ante-natal clinics.6 Two important services provided at young child clinics, in addition to health education and immunization, are (a) diagnosis and treatment of illnesses and (b) weighing and measuring of children in order to chart growth in relation to age.7 Young child services are commonly provided at a weekly clinic. Between one to five members of a health facility's staff operate the clinic; the most sophisticated piece of equipment in use is a scale constructed to withstand the rigors of active children. Virtually all mission health facilities and most government facilities, with the exception of aid posts operate young child clinics. The most intensive young child service program in Uganda, however, is not operated in static rural facilities. The Ankole Preschool Pro- tection Programme, which focusses on children under the age of five, operates 40 different clinics throughout Ankole District with two mobile teams of six staff members. Each team travels daily from its base of Operations to a new clinic site, returning once a month to that site. An average of 200 persons per day attend each clinic; the program thus has approximately 100,000 attendances per year.8 Health education has been used intensively in government health centers in an effort to spread preventive health measures throughout the country. Not only are special lectures and informal discussions an integral part of all clinics, but health education also comprises an important component of the center's daily curative services, par- ticularly in regard to diseases related to malnutrition or poor mudtation. In addition, health centers have a particular responsibility 28 for the health of the community adjacent to the center. The staff focuses on the health problems of its "defined area" and attempts to improve environmental health standards in that area through education and by example. Unfortunately, no systematic evaluation of this ap- proach has been conducted as yet. Family Planning Services Family planning services are not widespread in Uganda at the present time, although there has been an increase in such services since they were initiated in 1957 by the Uganda Family Planning Assoc- iation. 9 Although there are a number of doctors and midwives trained in contraceptive technology working in government and Protestant mission facilities, the number of persons receiving family planning services is not large, several thousand at most. Until December 1971, the government of Uganda had not developed a definite stand on the issues related to demographic change. Before the results of the 1969 census became known, the general governmental attitude toward family planning was somewhat negative. This lack of a formulated policy and rather negative attitude can be related to at least three factors: (a) the religious balance in the country (approx- imately 40% Catholic, 40% Protestant, 15% Muslim and 5% other), (b) the general attitude of competition between the three East African countries relative to indicators of growth and size, one such indicator being total population, and (c) the widely held belief that Uganda has excess land and needs more people in order to increase total product— ion. 'When the results of the 1969 census became known, however, the government established a "working party" to gather information and 29 develop recomendations concerning population matters. Recently, the government announced its positive attitude toward family planning activities. The general view among medical personnel in Uganda, at this time is that family planning services can be implemented only if integrated into a country-wide maternal and child health program. It is generally felt that only in this way can attitudinal problems toward family plan- ning services be dealt with effectively. A single objective clinic -- operating solely to provide contraceptives and other medical advice related to family health and size -- is generally not considered to be a viable institutional arrangement for long—run maximization of the use of family planning services in Uganda.10 The present resource commitment to family planning services in Uganda is difficult to determine. The International Planned Parent- hood Federation supports the activities of the National Family Plan- ning Association in Uganda. The level of this support in fiscal year 1970 was U. S. $81,200.11 The Rockefeller Foundation has supported a position in the Department of Obstetrics and Gynaecology at the Makerere Medical School in 1969-70, at U. S. $47,000, in order to further teach- ing and research related to (a) the medical aspects of contraception and (b) the clinical procedures necessary in the delivery of family planning services. Other international organizations have provided assistance of related to family planning; however, the amount of finan- cial aid involved is unclear. Finally the Ugandan Government indicated in its Third Five Year Plan 1971/72 to 1975/76 that it would allocate at least 1.0 million shs. to a population program during that period.12 30 In sum, it can be said that although family planning services are available in Uganda, and although attendance at the Uganda Family Planning Association clinic in Kampala (the largest in the country) has risen in recent years,13 the services have had no significant effect on fertility, birth rate, age structure, and infant and mater- nal mortality, because the attendances represent a very small pro- 14 portion of the total population "at risk" (i.e., women susceptible to becoming pregnant). Curative Health Services Health Service Facilities Number of Facilities The number of government health facilities (hospitals and rural units) rose from a total of 176 in 1950 to 342 in 1969-1970, (Table 2.4); the rate of increase over the last 10 years Of the period has been in excess of 6.0% per year. The number of health service fac- ilities has increased rapidly in the last decade primarily as a result Of central government emphasis on the improvement of health services during the second development planning period, 1966-1971.15 While governmental units have increased in number mission units have been declining, from a high Of 88 units in 1965-66 tO 71 in 1969- 70 (Table 2.5). This reduction in number Of facilities can be attri- buted primarily to (a) re-evaluation by the missionary organizations Of their role in providing health services and (b) the growing finan- cial difficulty Of Operating non-hospital health facilities. Distributional Mix Of Facility Types In 1950, approximately 15% Of all government facilities were 31 Table 2.4 Distribution Of Governmental Health Facilities in Uganda1 Total Hospital Rural Units Govt. Units % Of With Beds % Year (excl. Aid Posts) Total of Total (3) 1950 176 15.3 43.2 1955 183 12.0 62.3 1960 (2) 197 12.2 69.0 1965/66 257 8.9 59.1 1969/70 342 9.9 49.4 (1) The information presented in this table is derived from data compiled in Appendix D, Table D.1. (2) Figures for 1960 are for January - June only. In that year, there was a change in the reporting period from a calendar to a fiscal year. (3) Rural units with beds include health centers, dispensary/ Table 2.5 Distribution Of Voluntary Health Facilities in Uganda maternity units, dispensaries, and maternity units. 1 Total Hospital Rural Units Voluntary Units % Of With Beds % X£é£__ liexcl. Aid Posts) Total of Total 1953 50 18.0 N.A. 1955 50 18.0 N.A. 1960 (2) 73 21.9 53.4 1965/66 88 29.5 35.2 1969/70 71 36.6 42.3 (1) The information presented in this table is derived from data compiled in Appendix D, Table D.1. (2) Figures for 1960 are for January - June only. In that year there was a change in the reporting period from a calendar to fiscal year. N.A.: not available 32 hospitals; the percentage Of hospital units declined over the period however, to approximately 10% in 1969-1970 (Table 2.4). It is ex- pected that the latter percentage will remain constant over the next several years, for although the number of government hospitals is rising, the number of other types of health facilities is also expected to increase proportionally. The percentage Of rural units with beds (i.e., with the capacity for inpatient care) has shown a slightly different pattern over this period. They comprised approximately 45% of all facilities in 1950; rose to approximately 70% in 1960, but declined to less than 50% in 1969-70 (Table 2.4) 16 This drop in percentage of rural bedded units is due primarily to governmental policy to expand rural health ser- vices, such that daily outpatient service will be available in every gombolola (sub-county) in as short a time as possible. The genesis Of this focus lies at least in part in national political considera- tions and in the residual impact of the W.H.O. strategy to erradicate malaria and other infectious diseases (the strategy included, among other things, expansion of the basic health service system as widely and rapidly as possible.) 17 Within the category of bedded rural units, however, the number of health centers grew rapidly after 1958 (from 3 tO 40) relative to the number Of dispensaries and dispensary/maternity units.18 As a percentage Of the total number of bedded rural units, health centers increased from 6.6% Of all rural bedded units in 1960, to 23.6% of all such units in 1969-70. Within the mission sector, the importance of hospital facilities 33 relative to all other mission units has grown over the period 1953 to 1969-70. The number of hospitals nearly tripled between 1953 and 1962— 63; but has remained constant since then. It is unlikely that the number will increase; in fact, a reduction is likely primarily as a result Of (8) increasing financial pressure on Church organizations and (b) the growing number of government facilities. Political pressure from the government on the Churches to keep the facilities Open, however, has been great. Hospitals have increased in importance within the mission sector throughout the period, from 18% Of all facilities in 1953, to 37% in 1969-70. This increase has been due to (a) a relatively slow rate Of growth in the total number of mission facilities through 1966-67 and (b) an absolute drop in the number of non-hospital units between 1966- 67 and 1969-70 (Table 2.5). The maintenance Of a constant number of hospitals throughout the period 1960-1970 suggests that there is high priority placed by mission leaders on high quality, hospital care vis a vis preventive medicine and mass treatment. The fact that two new mission facilities, both hospitals, Opened between 1968 and 1970 re— inforces this unannounced priority. Size of Facilities In 1951, there were 3,392 beds in government and mission hospitals, 80% of which were in government units (Table 2.6). 19 From 1951 to 1965-66, the greatest growth in number of beds occurred in mission hospitals, primarily because the number of mission hospitals doubled during that time. The average number Of beds per mission hospital also increased during the period from 75 to 120. After 1965-66, however, 34 Table 2.6 Number of Government and Voluntary Hospital Beds and Average Number of Beds per Hospital in Uganda (1) Number No. Of Average NO. Average NO. Total # of Of Govt. Mission of Beds of Beds Hospital Hospital Hospital Government Voluntary Year Beds Beds Beds Hospital Hospital 1951 3392 2715 677 104.4 75.2 1955 3616 2932 684 133.3 76.0 1960 4826 3360 1466 140.0 91.6 1965/66 7282 4294 2988 186.7 144.9 1969/70 8792 5650 Est. 3142 Est. 166.2 120.8 (1) The information presented in this table is derived from data compiled in Appendix D, Table D.1. Table 2.7 Number of Government Rural Health Facilities and Average Number of Beds Per Facility in Uganda (1) Number Of Beds Including Average Number of Maternity Beds in Govt. Beds in Govt. Rural Year Rural Units Units With Beds 1951 1458 14.9 1955 1718 15.1 1960 2477 18.2 1965/66 3158 20.8 1968/69 3512 21.8 (1) The information presented in this table is derived from data compiled in..Appendix D, Table D.1. g. 35 the number of government hospitals increased more rapidly and by 1969/ 70, government hospital beds comprised 63% Of the total (8,792). The increase in government facility beds is likely to continue into the near future as a result of the government's expansion of the hospital system during the Second Development Plan, 1966-1971. The average number of beds per hospital will continue to decline as a result as it did from 182 in 1966/67 to 166 in 1969/70, because the new hospitals only contain 100 beds. The number Of beds in government rural units rose from 1458 in 1951 to 3512 in 1968-69, representing an annual rate of increase of 5.1% per year. The average size of individual units also increased over the period 1951-1968/69, from 15 beds to 22 (approximately) which represents an increase of 46% in total size (Table 2.7). The steady increase in size of these facilities has been due primarily to the increasing number of maternity beds in rural units. Estimated Expenditures on Health Services Uganda has allocated substantial resources for health services (Table 2.8).20 The magnitude of the commitment is indicated by the fact that the estimated total expenditure on health rose from approx- imately 85 million shs. in 1958/59 to 235 million shs. in 1968/69, an increase of 175% over the decade. By contrast, total government expenditures increased by 132% over the period. Also Of significance is the fact that health expenditures as a percentage of gross domestic product (monetary sector in current prices) increased from approximately 2.72 in 1958/59, to 4.72 in 1968/69. 21 36 Table 2.8 Estimated Total Expenditure on Health Services in Uganda: 1959, 1963/64, 1968/69 (1) 1959 1963/64 1968/69 Amount % Of Amount % of Amount % Of (in Mill. Total (in Mill. Total (in Mill. Total Source of Expenditure shs.) Expt. shs.) Expt. shs. Expt. (A) Government (I) Central Govt. (2) 52.47 62 .4 52.61 48.0 118 .41 50.1 (2) District Adminr istration Govt. 7.71 9.2 17.43 15.9 31.53 13.3 (3) Urban Authorities(3) 0.50E 0.6 1.02 0.9 2.72 1.2 (4) Municipalities 1.30E 1.5 2.80 2.6 14.00E-1 5f9 Less Inter-Govt. transfer payments for health services (2.91) (3.5) (7.05) (6.4) (9.26) (3.9) Total direct govt. expenditure (4) 59.07 70.2 66.81 60.9 157.40 66.6 (B) Mission Medical Services (5) Catholic 4.00E 47.00E 10.17 Protestant 1.50E 3.02E 3.24 Total Mission Expenditure 5.50 6.5 10.02 9.1 13.41 5.5 (C) Medical Training - Makerere 2.30E 2.7 3.00E 2.7 7.43 3.1 (D) Other Govt. Expendi- ture on Health(2)(7) 0.65E 0.8 1.00E 0.9 1.80 0.8 (E) UN Organizations (8) (1) WHO 0.42 0.84 4.61 (2) UNESCO/UNICEF ---- 0.97 ---- Total UN Expenditure 0.42 0.5 1.81 1.7 4.61 2.0 (F) Industrial Health - Private Firms (9) 2.00 2.4 2.53 2.3 3.82 1.6 (G) Other Voluntary & International Sources of Med. Service (10) 0.50 0.6 0.50 0.4 1.56 0.7 (H) East African Medical Research in Uganda(11) 1.88 2.2 1.82 1.7 2.86 1.2 (1) Private Consumption Expenditures on Health (12) 11.80 14.0 22.14 20.2 43.90 18.6 Total Expenditures 84.12 109.63 236.79 For notes on the derivation of the figures presented, See AppendixIL Table D.1. 3? The central government, through the Ministry of Health, is the primary source Of finance for health services, contributing approx- imately 50% Of the country's total expenditures on health. Most of the resources expended by the Ministry are used to operate an main- tain government hospitals throughout the country. District govern- ments, which have primary responsibility for rural health services, spend approximately 15% of the total sum spent of health in the country. The health expenditures of the four municipalities have increased in recent years from approximately 1.5% to over 6% of the country's total expenditure on health. This increase in expenditure by the municipalities has gone primarily to the creation Of urban health centers. Mission organizations account for a relatively small proportion of the country's total expenditure on health. Over the period 1958/59 to 1968/69, mission expenditure was less than 10% of the total spent on health in the country. This low figure may be due in part, however, to underreporting. The fact alone that there were 26 mission hospitals in 1968/69 containing (40% Of the total number of beds in the country) ‘makes the low figure all the more surprising. Total expenditure on private health services - private physicians, drugs, traditional healers - in Uganda is large. According to the estimate shown in Table 2.8 approximately 20% of the total expenditure on health occurs in the private sector of the health service industry, veven though government curative health services are free. The proportion of resources committed to curative vis a vis pre- ‘ventive health services are presented in Table 2.9, for the year 38 Table 2.9 Distribution of 1968/69 Expenditures on Health Between Preventive and Curative Services Nun-Direct Service 1968/69 Curative Preventive (Admin., Expenditure Health Health Research, Source Of Expenditure on Health(l) Services(1 Services(1) Training)(l-Z) _(A) Government (1) Central Govt.(3) 118.41 104.14 7.99 6.28 (2) Diet. Administra- tion Govt. (4) 31.53 24.13 7.17 0.23 (3) Urban Authorities 2.72 --—- 2.72 ---- (4),Municipalities(5), 14.00 3.52 10.48 ---- Less Intergovern- mental transfer payments for health serviggg» (9.26) (9.26) ---- ---- Total direct government expenditure 157.40 122.53 28.36 6.51 (B) Mission Medical Services (6) Catholic 10.17 9.70 ---— 0.47 Protestant 3.24 3.00 ---- 0.24 Total Mission Expenditure 13.41 12.70 ---- 0.71 (C) Medical Training - Mbkerere 7.43 ---- ---- 7.43 (D) Other Governmental Expenditure on Health 1.80 0.80 1.00 ---- (ES UN Organizations (1) WHO (7) 4.61 ---- 3.41 1.20 (2) UNESCO/UNICEF ---- —--- ---- —--- Total UN Expendi- ture 4.61 —--- 3.41 1.20 (F) Industrial Health- Private Firms 3.82 3.82 --—— -—-- (G) Other Voluntary and International Sources ‘ of Medical Service(8) 1.56 0.20 1.30 0.06 (H) East African Medical Research in Uganda 2.86 ---- ---— 2.86 (I) Private Consumption Expenditures of Health 43.90 43.90 —--- ---- Tbtal Expenditure 236.79 183.95 34.07 18.77 For notes on the derivation of the figures presented, see Appendix:E. 39 Although some of the figures shown are approximations, points can be made. Curative health services command a large >f total health resources. In 1968/69, approximately 77% of L was used to provide curative health services; 14% went to re health services and approximately 8% was used for training 1 manpower, medical research and administrative needs. Exr as made by the Ministry Of Health are even more heavily weight- I curative health services. Approximately 88% Of the Ministry's : and capital expenditures are allocated to curative health ‘ while 7% are allocated to preventive services and 5% is l to other services.22 The thrust of the analysis is evident; health services consume a large percentage of total resources : for health services in any given year. nditure and Employment in Three Representative Government Health Facilities Of the most important differences between (a) a large hospital, -bed hospital, and (c) a health center, is the level of re— xpenditure required to Operate the facility. The figures 2.10, using the health center as the unit of comparison, that for every shilling spent to Operate the health center ear, 12.6 shillings and 24.0 shillings are required to lOO—bed and a 250-300 bed hospital respectively. structure Of expenditures in the three representative fac- yields some further contrasts. Personnel costs comprised of total expenditures in the hospitals, whereas the figure ealth center was slightly over 50%. There were also notable 40 10 Financial and Employment Structure of Three Representative Governmental Health Facilities for the Year 1968/69 or 1969 (l) are Category Large Hospital Health Center 25-30 beds Hospital 250-300 beds 100 beds ((Expenditure category percentage of total est. expenditure) L Emoluments for 2d Labor 43.2 43.4 33.8 . Emoluments for .1ed Labor 18.9 19.9 18.9 ;Onnel Emoluments (2) 62.1 63.3 51.9 [Sundries 14.5 12.7 17.9 7 6.8 5.8 0.8 for Patients ployees (3) 2.4 1.6 14.0 ght, and Telephone 3.9 2.9 0.6 ion 10.3 13.7 8.6 3d Maintenance '7 (4) ---- ---- 6.1 :ges (less than) 0.01 ---- ---- :mated Annual Ltures 1,944,000 1,025,000 81,000 :Category (Percentage of total employment) :rsonnel (5) 38.5 35.4 28.6 EPersonnel 61.5 64.6 71.4 Eonnel (6) 260 147 14 L—— on the derivation of the figures presented, see Appendix E- 41 ces between both types of hospital on the one hand and the enter on the other in food, transport, and utilities. The expenditure by the health center for food is due to the fact generally does not provide food for inpatients. There are 0 utilities in rural health centers, although in some dis- larger units have electricity. In hospitals on the other are there is a wide range of equipment available, the use of ity is essential. The most plausible explanation for the :es in percentage of total expenditures for transport is 1sport represents a fixed cost; as the total cost of operat- 11th facility rises the amount used for transport services as a percentage of total expenditure. relative employment ratio between the health center and the : of hospital is shown in Table 2.10. For every one person in the health center, 10.5 persons are employed in the 100- -tal and 18.6 persons are employed in the 250-300 bed hospital. .On, the average expenditure per employee is greatest in the »spital,(7,500 shs.) declining in the smaller hospita1,(7,000 Rd lowest in the rural health center, (5,800 shs.). Demand for Curative Health Services SE e 2.11 presents data on the number of attendances at govern- th facilities over a twenty year period, 1950 to 1969/70. r of attendances increased during the first decade at an te of 6.3% and at an even more rapid rate of 7.7% over the 1f Of the period. When the rate of increase in total 42 2.11 Utilization of Governmental Health Facilities (1) (2) (3) (4) (5) (6) Estimated Total Total NO. Total No. Total NO. Estimated Total Number of New of Out- of Inpatient Average NO. Rnnflation of Cases Outpatient patient Re- Admissions of cases (000) (000) Cases attendances (000) per person (000) (000) per year 5,322.0 4,873.0 2,329.0 2,422.0 122.0 0.92 5,874.0 5,459.7 2,732.3 2,597.1 130.3 0.93 6,573.0 7,784.8 4,335.2 3,245.4 204.2 1.18 8,221.0 13,083.9 7,178.6 5,658.6 246.7 1.59 ),l9l.0 17,826.5 9,537.3 7,884.0 405.2 1.94 res shown for 1960 are estimated from the data reported for t six months of the year. In 1960, the Government changed alendar to a fiscal reporting year. pulation estimates were derived from estimated rates of popu- rowth between censuses. Republic of Uganda, Ministry Of Health, Annual Reports and :al Records, (Entebbe: Government Printer, selected years). 43 ldance is related to population growth, the figures take on :ional significance. Even though the population is increasing at the present natural of increase of 3.3% per year, the average number of attendances vernment facilities has increased from 0.92 attendances per lper year in 1951 to 1.94 attendances per person per year in 59. During this period, 65% Of all attendances were at rural facilities. This attendance figure is particularly significant e rural facilities received only 20% of government financial t for curative health services during the same period (Tables I 2.9). While health centers provide the majority of outpatient Lospitals provide the majority of inpatient services, with ad- s tO government hospitals comprising 44%,Of the total mission ls 18%, and rural units 38% in 1968/69.23 th inpatient and outpatient attendances at government units 3d rapidly between 1951 to 1968/69, by 365% over the entire Outpatient attendances, however, increased more rapidly than .tient attendance, 410% vis a vis 332%. The proportion of ces at all government units thus shifted toward outpatient services -- from 97.5% in 1951 to 97.7% in 1968/69. 1 presented in Table 2.12 indicate the extent to which patients .' the treatment process as outpatients at rural health facilit- a) admitted to inpatient care in the facility, (b) transferred strict hospital, or (c) die as inpatients in the rural health Of the 587,000 new attendances in the years indicated, an E 51.4 per thousand were admitted to inpatient care. in the :ility; and average of 3.4 per thousand were transferred to .... ...: ~ . u...l.ud. 44 .mvuoomu hufiafiomm Eouw vmuomaaou mama ANNoV Aaoo.Nv AHNH.OMV Nao.amm Hmuoa o.o~ q.m q.am 05\moma N.NH m.q H.~m mam.wa maoxn< mummammmfio mufimamnmmza omwmemfl w.HN H.N m.qm omo.HH maoxa< summamamaa meanness ca\msma m.HN H.N s.wm caomww maoxca “mucus spasms ascmsmam ca\mcmfl H.NN m.H H.Nm mas.- mfloxa< “mucus spasms access ca\aoma «.0N o.m ~.me oswqwm awamsm XNummaummHn ouHHmM oa\moma m.m~ m.~ m.ma Haodaa «woman summcmmmaa meammsm omwmoma ~.om m.a o.mq omn.m~ mwmmsm summaumwao usaaxH ca\msma 0.0H N.H w.mm ommdma «woman summamamaa assess omwmomfi s.qm c.~ s.sm use.os «woman summamamaa “sawsm Om\moma m.oH n.o H.mn qoq.nm mmomsm umuamo nuawmm msammoaamz 0: $2 0 . Z a . m m . 3 SW S mflomam $38 53% 983M oa\HomH m.- H.o m.ma moa.nm mwomsm umuamo :uHmmm muafimz ca\moma a.s~ o.H o.~m mq~4~m madman “muamu nuammm «seemzamz oaxmoma «.mH m.q H.om oHo.Ns \memmwm umuamu suammm muassas oa\msma o.oH m.~ o.~o mflmqu «woman umuamu auammm mxoxamsmz aoma u--- q.~ o.qm H~m+w omcmz .m [summammmfin mas>=m mesa o.HH o.m H.Nm Hom.ou amass .m wummamaman ougfimm mama o.mH m.H m.em mam.mm owns: .m umucoo nufimmm essays momH 5.0m H.m «.ma ooo.o~ ‘mmmuz .m saunas spasms wzxfiam Aoooa Hos mummy Aoooa “ma mummVAOOOH uma mamas maowmmfiav< wmoamvamuu< mmocmvamuu< mmoawvcmuu< vOHHmm mEHH ou 0>HumHOM 302 Cu 5.52 Cu 552 Hmuom. uUHHum 09 an msummn w>fiumamm w>HumHmm . an A ARV uHHHomm WHMHmCflwud 04.— 34. D 0 433,40 45 pital; and an average of 20.6 persons per thousand died while ving inpatient care in the rural health facility. The Ministry of Health analyzes attendances at government and >n hospitals by 14 major disease categories adapted from World 1 Organization classification standards.24 Analysis Of this 'ized information over time yields a good idea of (a) the dis- ion of diseases treated, on inpatient and outpatient bases, at n and government hospitals and (b) changes in the distribution ime. The following analysis of diseases treated in government ssion hospitals is based on the data presented in Tables 2.13 1 2.17. ant treatment process 1e distribution of diseases treated as inpatient cases in lent hospitals is presented in Table 2.13. Admissions for in- s; and parasitic diseases (malaria, measles, helminthiases, .losis, veneral disease, etc.), have dominated the entire period 5M58/69), but the percentage of the total comprised by infections ausitic diseases has declined by 50% from 34% of the total in 217% in 1968/69. The categories of normal delivery and com- )rus of pregnancy and puerperium both grew more rapidly than any rsrr the period. In combination, these two categories grew from ‘total admissions in 1952 to nearly 30% of admissions in 1968/ ruzrease of 140% over the period. This increase is due at Iiart to changes in traditional practices related to birth, 46 :2.13 Distribution of Diseases Treated in Uganda on an Inpatient Basis in Government Hospitals; Percentage of Total Cases by Disease Category in selected years. se Category 1952 1955 1958 1961/62 1964/65 1968/69 tive and Parasitic 33.8 28.2 21.3 19.6 15.8 16.8 rowths 1.3 1.7 2.0 2.2 2.9 1.8 gic, Metabolic 1 Blood 2.1 2.5 3 7 5.3 4.5 4.7 es of the Nervous tem and Sense Organs 3.2 3.7 3 7 3.1 4.0 2.7 :atory 0.7 1.0 1.1 1.4 1.9 1.7 fatory 9.9 10.3 9.5 9.7 10.4 10.6 itary 6.6 6.2 7 4 9.0 9.8 10.0 ::0rinary 2.9 4.8 5 5 5.6 5.2 3.9 hey and Puerprium 4.8 6.8 8 5 9.9 10.4 11.0 ryywithout plication 7.6 10.6 13.4 14.1 13.6 18.7 id Musculo- letal 8.6 6.2 5 2 5.1 5.5 4.6 ;;of the Born 1.2 1.0 0.9 0.8 1.2 0.9 ;;ned Diseases 7.3 7.4 8.2 4.7 3.8 1.7 ;; 10.0 9.7 9 7 9.5 11.0 10.8 100.0 100.0 100.0 100.0 100.0 100.0 txed from Republic of Uganda, Ministry of Health, Annual Report or igzal Records, (Entebbe: Government Printer, selected years). 47 eases in ante-natal preventive services, increases in the avail- .ty of maternity wards, and changes in demographic variables (e.g. mreased birth rate). Allergic, metabolic and blood diseases (e.g. malnutrition, anemia, rose from approximately 2.1% in 1952 to approximately 5% in the s. Alimentary cases (e.g. gastritis, hernias, etc.) increased nuously, from 6.6% in 1952 to 10.0% in 1968/69. Both skin and lo-skeletal diseases and ill-defined diseases declined markedly the period. The drop in ill-defined conditions may reflect (a) vements in diagnostic procedures and/or (b) increased availability >oratory services, such that a smaller percentage of all diseases 1 ill-defined; it may also reflect the pressure of demand, with :ain self-selection process Operating such that only more readily »sib1e conditions are treated today on an inpatient basis than : past, given the limited supply of inpatient treatment facilities. en; remained a fairly constant 10.0% of all cases admitted to EHIC treatment in government hospitals; an increasing proportion 'ury cases, however, were related to increased automobile traffic. rue distribution of mission hospital inpatient cases is shown in 2.214. As in government hospitals, infectious and parasitic Em; are most prevalent, comprising 33% in 1958 and 27% in 1968/69. icy-related cases, have seen no real trend in mission hospitals; 3, ‘however, comprise a large percentage (approximately 24%) of :ases treated over the entire period. leergic, metabolic, and blood diseases as well as alimentary 35:, have increased as a percentage of all inpatient cases over 48 2.14 Distribution of Diseases Treated in Uganda on an Inpatient Basis in Mission Hospitals; Percentage of Total Cases by Disease Category in selected years. 3e Category 1958 1961/62 1964/65 1968/69 :ive and Parasitic 33.2 33.9 26.8 26.9 rowths 1.6 1.9 2.4 1.7 ;ic, Metabolic L Blood 5.2 5.1 6.8 8.2 l.— es of the Nervous tem.and Sense Organs 2.2 2.4 2.2 2.8 atory 1.2 1.4 1.5 1.5 atory 10.3 10.4 9.5 10.6 tary 7.2 8.1 10.2 10.2 :Urinary 5.0 4.3 4.6 2.7 Hey and Puerprium 8.9 8.4 8.3 8.2 ;y without plication 13.2 14.2 17.3 15.1 idansculo-skeletal 4.1 3.2 3.3 3.7 ;; of the New Born 1.1 1.7 2.2 2.4 ;;ned Diseases 3.9 2.6 2.5 3.2 ;;7 2.9 2.4 2.5 2.8 100.0 100.0 100.0 100.0 Lted from Republic of Uganda, Ministry of Health, Annual Repgrt abstical Records, (Entebbe: Government Printer, selected years). 49 >eriod. Diseases of new born infants have increased rapidly as rcentage of total cases in mission hospitals, although they do ,onstitute a large percentage of total cases treated. There rs to have been a fairly significant drop. in the percentage of o-urinary cases treated in mission hospitals, from 5.0% in 1958 7% at the end of the period, although close examination of the as reveals that the drop in the last year may have been a random Injury cases, as a percentage of the total cases treated, were r stable over the period (2.5% to 2.9%) but comprise a much :r percentage of total cases treated than in government hospitals. ient Treatment 3 was true in the case of the distribution of inpatient cases, ive and parasitic diseases account for the largest percentage as treated on an outpatient basis in government hospitals 2.15). Again, however, this group of diseases has declined arcentage of total cases treated from 38% in 1952 to approx- . 30% throughout most of the 1960's. Other important shifts in ‘ucture of government hospital outpatient cases include (a) an e in allergic, metabolic, and blood cases (0.6% of the total to 1.3% in 1968/69), (b) a large increase in the number of system and sense organ diseases (from 1.8% to 6.1%); (c) an e in respiratory and alimentary diseases (respiratory: from > 17.2%; alimentary: from,12.7% to 15.5%); and (d) decreases skin and musculo-skeletal diseases and injuries as a per- of total cases treated (each, however, still comprises a 50 2.15 ‘Distribution of Diseases Treated in Uganda on an Outpatient Basis in Government Hospitals; Percentage of Total Cases by Disease Category in selected years. 3e Category 1952 1955 1958 1961/62 1964/65 1968/69 tive and Parasitic 38.4 37.7 33.4 28.7 28.7 31.8 rowths 0.1 0.1 0.1 0.1 0.1 0.1 gic, Metabolic [Blood 0.6 0.7 0.7 1.2 1.0 1.3 :es of the Nervous tem and Sense Organs 1.8 2.0 2.5 5.9 5.4 6.1 atory 0.1 0.2 0.2 0.2 0.2 0.2 atory 12.0 12.9 13.9 15.8 12.6 17.2 tary 12.7 16.8 14.4 14.8 15.3 15.5 r—Urinary 0.6 1.0 1.5 1.8 3.4 3.1 ncy and Puerprium 0.2 0.2 0.7 0.6 1.6 0.8 ;;Without plication ---- ---- ---- ---- ---- ---- I: Musculo-skeletal 14.4 11.3 12.2 14.2 13.0 10.5 g of New Born 0.6 0.2 0.2 0.4 0.7 0.3 Mined Diseases 4.5 3.8 7.3 5.6 7.7 3.5 : 13.0 13.0 12.9 10.8 10.5 9.5 I & (excl. s Innoculations) 100.0 100.0 100.0 100.0 100.0 100.0 bf tions and culations 17.3 16.1 12.1 17.5 12.4 10.5 1 (incl. Exams & :ulations) (1) 769.2 811.2 1200.0 1522.9 1781.5 3288.5 hf ted from Republic of Uganda, Ministry of Health, Annual Report or Lcal Records, (Entebbe: Government Printer, selected years). Total in thousands of recorded diagnoses. 51 In mission hospitals, the structure of diseases treated on an outpatient basis remained basically stable over the period 1958 to 1968/69, (Table 2.16). The decline in ill-defined conditions as a percentage of total cases, from 13.2% in 1958 to approximately 2.0% for the remainder of the period, is indicative of a change in disease reporting by mission units after 1958, the first year in which the government required reports from mission health facilities. Infective and parasitic diseases are most prevalent among the outpatient cases treated at mission hospitals, comprising approximately 40% of all cases. Other important disease groups treated in the mission hospitals include respiratory diseases, alimentary diseases, skin and musculo-skeletal diseases, the diseases of the nervous system and sense organs. The percentage of injury cases treated in mission hospitals is very small, and declined, in fact, from approximately 3.5% of all cases treated in the years prior to 1968/69 to less than 1.0% in that year. Comparison of Disease Mix Between (a) Inpatient and outpatient and (b) Government and Mission Hospitals Table 2.17 provides a comparison of the distribution of diseases between (a) government and mission facilities and (b) inpatient and outpatient treatment processes. The standard of comparison is the government hOSpital disease mix for both inpatients and outpatients; the mission hospital disease mix is compared to this standard. The disease mix structure for mission hospitals in both periods contained a larger proportion of infective and parasitic diseases; allergic, metabolic, and blood diseases; respiratory diseases and 52 Table 2.16 Distribution of Diseases Treated in Uganda on an Outpatient Basis in Mission Hospitals; Percentage of Total Cases by Disease Category in selected years. Disease Category 1958 1961/62 1965/65 1968/69 Infective and Parasitic 41.8 40.7 41.7 41.5 New Growths 0.6 0.4 0.6 0.3 Allergic, Metabolic and Blood 3.5 4.5 6.1 5.3 Diseases of the Nervous System & Sense Organs 1.6 5.3 5.4 6.2 Circulatory 0.5 0.6 0.6 0.8 Respiratory 9.5 13.1 11.1 13.4 Alimentary 9.8 13.8 12.8 13.1 Genito-Urinary 2.3 4.0 3.6 3.9 Pregnancy and Puerprium 6.1 1.8 2.1 1.9 Delivery without Complication ---— ---- ---- ---- Skin & Musculo-skeletal 6.8 9.2 8.9 9.6 Diseases of New Born 0.6 1.1 1.0 0.8 Ill-Defined Diseases 13.3 2.0 2.4 2.3 Injuries 3.6 3.5 3.7 0.9 Total (Excl. Exams & Innoculations) 100.0 100.0 100.0 100.0 Examinations and Innoculations 23.3 25.2 32.6 35.7 Total (Incl. Exams & Innoculations) (1) 166.6 164.5 497.6 640.6 Calculated from the Republic of Uganda, Ministry of Health, Annual Report or Statistical Records, (Entebbe: Government Printer, selected years). (1) Total in thousands of recorded diagnoses. 53 .oH.N was .mH.N .QH.~ .MH.N moanwa cw nmuoomonn mums Scum vo>fiumw mum newsman .auow owMuaoouom a“ ao>fim mum mounwwh "ouoz H.Ham o.wo o.MHH ~.mn m.m «.mm m.m~ m.mN wwqumsH H.mMH o.omH o.ma m.NNH n.qo n.mm N.me o.mq mummwmfln vocwwmnIHHH o.oom m.¢mH o.oom o.omq n.oo~ o.mmm n.00m N.NNH chem 3oz mnu mo mmmmmmfln m.mm w.qm w.mq o.~q q.Hm w.qo c.0w w.wn Hauwamxm 10H=om=2 was afixm III- III: In: :1... I}: III: A . om m .3 sofiumoflmaoo u:0£ua3 >HM>fiHoQ o.amq n.ooq o.mmma m.qa~a m.mmm o.oom m.qu m.qoa Esfiumumsm was hocmcwmum «.mo m.moa m.mNH 5.00m w.mNH N.NNN ~.mo o.oo \Nwmcfiusuouanoo $.55 o.mm m.qo q.Hm m.qw N.mm o.NOH m.nm mpmuamawa< H.mn «.mn o.Hm m.wo m.nn m.~m 0.00H c.moa muoumuflmwwm m.nwa m.m- 0.0mm 0.0mm o.ooq o.o0m N.wm H.mOH muoumaounao «.mc m.mq m.q< o.qu o.aoa a.mw n.moH m.mm mammuo madam can Ewummm m=o>umz wnu mo mmwwowfln n.¢ma m.maa m.aom o.w~m n.noq o.nnm m.an m.o¢H vooam new 023331 .3332 5.00m o.mn< o.oomH o.ooo~ o.oom o.ooq q.HuuomaH mo\mmmH mw\aomH mo\mooa mmma mo\womH ~o\HomH immxmoaa mmma mahe ammomfin manqumom doammwz mamuwawom ucofisuw>oo nomfiummfioo uamwumnuao aomfiumnaoo udmwumnaH unuaumnuso ou vuumqaoo uauauonaH unassum>oc ou noumeoo scammwz maneuopflnuman ommmmHn mo noufiuwnaoo NH.N wanna 54 diseases of the new born. Government hospitals, on the other hand, had a higher proportion of new growths; genito-urinary diseases; normal deliveries; skin and musculo-skeletal diseases; and injury cases. The inpatient disease mix in both government and mission hospitals in 1968/69 contained a greater proportion of cases in the following disease categories than in the outpatient case: new growths; allergic, metabolic, and blood diseases; circulatory diseases; diseases of preg— nancy and puerprium; diseases of the new born and injuries. The out- patient structure in both types of hospitals, however, contained a greater proportion of infective and parasitic diseases; nervous system and sense diseases; respiratory diseases; alimentary diseases; and skin and musculo- skeletal diseases. Employment Total Employment Trends in total employment in the health services industry are presented in Table 2.18. In 1951, approximately 7,500 persons were employed in both the public and private sectors of the health services industry. By 1968, total employment in the health services industry had risen to over 17,000 persons, representing an annual rate of in- crease of 4.9% over the period. The rate of increase in employment was greater than the increase in total recorded employment in the country over the same period; the health services industry increased its percentage of total recorded employment from 3.8% in 1958 to 6.0% in 1968. From 1958 to 1968/69 recorded employment in the government sector of the health service industry increased by 81% while it in- creased in the private sector by 133%. Since 1964, however, employment 55 Table 2.18 Employment in Uganda's Health Services Industry Total Medical Earnings Employment (in thousands) vaernment and Private Year Total Government Private (in million shs.) 1951 7. 48 1 1958 9.18 8.39 0.80 24.36 1959 9.87 8.80 1.07 27.93 1960 10.27 9.00 1.27 38.72 1961 10.07 8.61 1.46 31.28 1962 10.57 8.87 1.69 37.26 1963 10.55 8.95 1.61 37.44 1964 12.66 10.60 2.06 45.59 1965 13.42 11.50 1.91 45.38 1966 14.36 12.37 1.98 52.68 1967 15.29 13.42 1.86 61.14 1968 17.03 15.17 1.86 65.52 Sources: Data for 1951 taken from Uganda Protectorate, Report on the Enumberation of African Employees in Uganda, March 1951; East African Statistical Department, March 1952. Data for 1958-1968 provided by Ministry of Planning and Economic Development. * Unfortunately, the data regarding employment in the private sector of the health services industry may be suspect, due to inconsistencies in mission reports of employment allocation between their activities in health services and education. Personal come munication from the Government Statistican, January 27, 1970. 56 in the public sector has increased more rapidly than the private sector, the latter showing a decline since 1964. Earnings in both the public and private sectors of the health service system increased by 169% from 24.36 million shillings in 1958 to 65.52 million shs. in 1968. During the same period, total reported earnings for all sectors of the economy increased by 125%. The share of total earnings represented by the health services sector of the economy thus increased from 5.1% to 6.1% over the period. Average annual earnings per worker in the health services sector of the economy also increased during the period, from 2,653 shs. per year to 3,848 shs. per year--a 45% increase. This increase, however, was smaller than the increase re- corded for average earnings per worker in the total economy, which rose from 1,960 shs. per year to 3,800 shs. per year--a gain of 94%. There was therefore, an increase in earnings parity for all workers in the economy relative to health service employees. Ministry of Health Establishment During the period 1967/68 to 1970/71, the proportion of positions allocated to each employment category remained fairly constant, although the total number of positions increased by 36% (Table 2.19). Approx- imately 75% of all positions were allocated to direct curative medical services over the four year period. The allocation of positions to public health services has shown a slight decline over the period, from approximately 8.5% to 7.0% of all positions. Mulago Hospital and Butabika Mental Hospital are listed by the Ministry as separate categories. Mulago Hospital, the national teaching and referral hospital with approximately 1,000 beds, has consistently 57 .mmmum muoumuoema ofiusmmmuunuoaoso AUV new ”mmmum mofiHmeom new Hmofiuomomahmsn Anv “mmmum hamuunuo«mhna was amoewOHOfinmu .Houuaoo uouoo> .kuoumuonma Amv Hanna noumHH maoaufimom monaaoaw moow>umw wawuuoemnm mo showmuwo sea AmV .mamawhm new oofiumufiamm ou voumooaam mooaufimom mmvsaoofi nuamum oaanbm mo knowouwo may Amv .:mmmum Hmuuauu new u>HumHumfiafianmz new :umumaaaz mnu mo oofimwo: ou nmumooaam maofiufimou mum knowMumu doauwuumwnfiap< kuuaoo mnu ca novoaoaH Aav .mumoh nmuooaom Aumunapm unmaauo>ou "unnoucmV .ousuavaomwm ucmuuzomm mo nonmaauwm .mnamwp mo owansmmm ”mouaom r. Huam mmm mmm flee omH cow mmam H~\onH mmem hem NNm mmm boa Nae come ON\mme mmam Now sum mom had mmo «mam mo\momH HmHN ku own omm hm Ham «mum wo\noma Amomusa .muouoovv Amvmooa>uom Amvsuamom cowuwuumfiafiap< Hmufinwom Hmuwamom uaufinmfiapmuom use» amusemz guano waauuoaesm ufianam Aav Houuamo fineness mxaauuam amass: Hauoa uou nonwooaa< maoauwmom unmasmfianwumm mo wonabz usuanmaanwumm endow: no shamans: ma.~ «Home 58 taken approximately 16% of total Ministry positions. On the other hand, Butabika Hospital, the only mental hospital in the country and also with nearly 1,000 beds, has been allocated approximately 3.8% of all Ministry positions over the last four years. 'Rggistered Medical Manpower Table 2.20 provides information on the number of trained medical manpower of various types, registered in Uganda. Although the data imply that there has been an increase in the number of doctors and dentists in Uganda registration lists of medical manpower in East Africa are greatly inflated for the following factors: (a) registered expatriate personnel often leave the country after a short-term as- signment; (b) registrants die, but are not promptly removed from the lists; (c) some registrants practice medicine in one of the other East African countries, but not in Uganda; (d) some registrants leave the labor force or obtain alternative employment. 25 Unfortunately, the manpower registration figures also offer no insight into rural/urban differentials in availability of medical personnel.26 Employment in Rural Health Facilities Table 2.21 presents information in the employment structure of rural health facilities in two districts, Busoga and East Mengo. The units represent four types of facilities and the personnel in each type of facility are categorized into three groups: trained medical employees, untrained medical employees, and others. The average number of employees in the group of health centers is 19.2 persons per unit; the average in dispensaries is 12 persons, in sub-dispensaries 5.5 persons, and in maternity centers 9.7 persons. 59 Table 2.20 Medical Manpower Registered to Practice in Uganda Doctors Year Registered Licensed Total Dentists Midwives Nurses Pharmacists 1951 151 81 232 10 732 1952 167 82 249 9 755 1953 178 79 257 11 775 1954 223 53 276 13 836 1955 255 41 296 14 897 1956 292 53 345 15 980 27 1957 305 54 359 17 1023 34 1958 371 52 423 19 890(1) 174(2) 34 1959 422 54 476 19 907 301 40 1960 441 43 484 18 968 366 60 1960/61 476 52 528 18 1060 410 61 1961/62 479 73 552 28 1156 1354(3) 72 1962/63 504 80 584 22 1290 1557 84 1963164 538 113 651 28 1430 1748 95 1964/65 588 140 728 31 1565 2271 85 1965/66 642 171 813 39 1911 2682 61 1966/67 727 214 941 40 2199 3040 104 1967/68 797 181 978 42 2551 3277 116 1968/69 1969/70 Source: Republic of Uganda, Statistical Abstract, (Entebbe: Government Pr inter, selected years. ) (1) A new ordinance for the registration of midwives was iniciated in 1958. As a result, the series is discontinuous from that date. (2) State Registered Nurses only. (3) Includes State Registered Nurses, Enrolled Nurses, and Male Nurses. 60 These variations reflect both the range of service provided at each type of facility and the total demand for service. Table 2.21 Employment Structure of Four Types of Rural Health Facilities in Uganda: 1969/70 Average Employment Sample Facility Type Total Trained Untrained Other Size Health Centers 19.2 6.25 7.5 5.4 8 Dispensaries 12.0 2.2 5.8 4.0 12 Sub-dispensaries 5.5 1.1 1.4 3.0 8 Maternity Centers 9.7‘ 3.0 3.3 3.3 3 Information derived from facility records. The sub-diapensary provides daily outpatient clinic services only. Employees here are likely to be distributed as follows: 1 diagnostician, 1 or 2 persons providing treatment and 3 persons engaged in maintainance work. In a dispensary, both inpatient and outpatient services are provided, and speciality preventive clinics (e.g., ante-natal, young child, etc.) are usually offered. The in- creased range of services requires additional personnel; there is an average of one trained person above the staffing requirements of the sub-dispensary, an average of 4.4 more untrained medical personnel and one additional untrained employee providing other services. Maternity centers provide inpatient maternal care, ante—natal ser- Vices, and young child services. The three trained employees in the maternity center are midwives, and the three untrained medical employees assist in providing inpatient care. 61 The health center offers the full range of services provided by the other three types of facilities, and engages in a more in- tensive program of health education and public health program than in the other units. As a result, the health centers in the sample employed an average of 6.3 trained medical employees per unit, 7.5 untrained medical personnel and 5.4 employees engaged in other, non-medical activities. The implications of the varying personnel requirements of the different types of rural health facilities are as follows. First, since district governments are primarily responsible for building and staffing rural units, they should have information about different employment patterns exist- ing in different types of health facilities. Even the decision to upgrade a facility is likely to lead to a different pattern of de- mand upon the labor market; the upgrading of a sub-dispensary to a dispensary, for example, is likely to increase the demand for un- trained medical personnel, while the decision to upgrade a dispensary to a health center would increase the demand for trained medical personnel. Second, employment patterns amongst facility types imply differing financial obligations to districts when they decide whether to build or upgrade facilities. This factor is particularly true in the case of future recurrent cost obligations, particularly in View of the fact that personal emoluments in health centers and possibly in other units are likely to comprise at least 50% of total recurrent costs. Finally, employment mixes of rural health facilities have implications for the central government's training programs for 62 non-professional medical personnel. Not only must the central government analyze manpower requirements for hospitals, but needs also to consider the variability of demand by districts for trained medical personnel; the demand variability in turn is dependent upon the districts' investment mix in various types of facilities. Summary In this chapter, Uganda's health service system was examined. An overview of preventive health services was presented. The discussion centered on the extent to which environmental health, immunization, ante- natal services, child health services, health education, and family planning services are used by Ugandans. The curative health service system was then analyzed, focusing on four important aspects. First, the government and mission health facility structure was analyzed. Second the analysis centered on financial matters, with attention given to the rapid increase in cost of all health services and the large percentage of total health resources committed to curative service. Third, the structure of demand for curative health services was examined; the rise in attendance was noted and analysis was made of the shifts and differences in disease mix among government and mission hospitals. Finally, the pattern of employment in health was discussed. In this context, some of the manpower implications of possible expansion of the service systemnwas noted. 63 Footnotes Urban areas include the four municipalities - Kampala, Jinja, Mbale, and Masaka - as well as the 15 urban authorities and nine town boards. The terms municipalities, urban authorities, and town boards are used in Uganda in reference to the various levels of urban or city development. Municipalities are the largest units, usually have a relatively high percentage of non-African residents, and are very wealthy relative to other parts of the country. All of the municipalities are focal points of economic activity, governmental administration, and social and cultural amenities. They each have a population of more than 20,000 persons. Towns classified as urban authorities are the next most sophisticated urban areas. They are free to run their own affairs with minimal supervision from the Ministry of Regional Administrations. Some are nearly as large as the two smaller municipalities, Masaka and Mbale, but have not been chartered and are not nearly as wealthy as the municipalities. Most of these towns have less than 10,000 p0pulation, and serve as re- gional or district urban centers from which many government activities emanate. Town boards govern small urban areas which are closely controlled by the Ministry of Regional Administrar tions. Mbst of the town boards' personnel and virtually all financial support is determined (or seconded, in the case of personnel) by the Ministry. The towns governed by town boards are focal points for a part of a larger district, and many of them have large markets and perhaps two or three streets with permanent buildings; most have a population of 2,000 to 5,000. The figures do not include resources committed to water develop- ment and sewage treatment. To what extent are immunization services disseminated during any given year? The pOpulation of Uganda in 1966/67 was approxi— mately 8.5 million, increasing at approximately 3.5 percent per year. (See Steven R. Taber, "A First Look at the Provisional Results of the 1969 Uganda Census," a paper delivered at the Seminar on Population Growth and Economic Deve10pment, Nairobi, Kenya, December 15 - 22, 1969, (Nairobi: unpublished paper, 1969). This rate of increase implies an annual net increase of approximately 390,000 — a birth rate of 45 per 1,000. The dif- ference between 390,000 and 310,000 is comprised of approximately 120,000 deaths and net in migration of 40,000. See Taber, Provisional Results of the 1969 Uganda Census," p. 12.) Assuming that approximately 580,000 persons were vaccinated against Polio (Table 2.2)and that some underreporting did occur, approximately 7.0% of the total population was immunized during that year. However, if one takes into account (a) the dynamic aspects of population change, (b) the fact that the portion of the population most at risk consists of those people under 25, and (c) the fact that the largest proportion of immunization 64 services is provided to children under five years of age, perhaps 11% of the at-risk population were immunized against Polio, although the percentage of the total population imr munized is less than 6.5 percent. (There are approximately 5.1 million persons in the at risk population and approxi- mately 530,000 persons were innoculated against polio in 1966/67)- For a discussion of the problem of tuberculosis in Uganda and possible ways to control the disease, see R. H. Morrow, "Tub- erculosis Control in East Africa", (unpublished paper, Makerere Medical School, Kampala, Uganda, 1969) and R. H. Mbrrow, "TUb- erculosis", in Uganda Atlas of Disease Distribution, Occasional Paper no. 12, edited by S. A. Hall and B. W. Langlands (Kampala: Department of Preventive Medicine and Department of Geography: Makerere University College, 1968). The percentage figures have an upward bias for two reasons: (a) the percental computations were made comparing the total number of ante-natal cases in time to with the estimated number of births in time to; a more appropriate comparison would use new ante-natal visits in time to and the number of births in time t0+0.75 to reflect the lagged effect of birth; and (b) not all women with first visits to ante-natal clinics have live, viable births; some result in abortions or other problems lead- ing to fetal or new-born death. This statement is based on personal observation at numerous clinics and on a cursory analysis of several different facilit- ies' monthly statistical returns. Often the young child clinic and the ante-natal clinic is held on the same day. The latter is a means of detecting malnutrition, even though some of the obvious clinical symptoms may not yet be manifest. For more information on the Ankole Preschool Protection Program, see Malcom Moffat, Mbbile Young Child Clinics in Rural Uganda: A Report on the Ankole Preschool Protection Programme, 1967-69, (Kampala: Department of Pediatrics and Child Health, Makerere University College, 1970). Although the Uganda Family Planning Association was established in 1957, with a grant from the International Planned Parenthood Association, a full range of family planning services was not offered until 1962. The information compiled in this section results from many formal and informal discussions with numerous people involved or interested in population issues in Uganda. The points of view represented included medical, demographic, and socio- economic. Principal among these individuals were Dr. George Saxton, Dr. Donald Minkler, Dr. Keith Masters, 10. 11. 12. 13. 14. 15. 65 Dr. MiChel Thuriaux, Dr. Steven Taber, Mr. Kenneth Hill, and Dr. Richard Trussell. A good deal of further background in— formation on family planning services in Uganda and East Africa has been assembled as a result of the Ford Foundation sponsored Workshop on Needed Research in Family Planning and POpulation Growth in East Africa, Nairobi, July, 1970. See David Radel and Shelley Ross-Larson, eds., Proceedings of the WOrkshop on Needed Research in Eastern Africa on Family Plan- pipg; Medicaeresearchggnd Programme Trials, Nairobi, Kenya, 23—25 July 1970, Sponsored by the Ford Foundation, July 1971, (mimeoed document). No research has yet been conducted in Uganda or in East Africa on the relative effectiveness of any institutional arrangement. A research project on this problem may develop in Kenya as a result of that country's recent commitment to a mass family planning program. International Planned Parenthood Federation, Report to Donors; Programme Development and Financial Statements 1970-72, Inter- national Planned Parenthood Federation, London, England, Sept- ember 1971, page 190. See Republic of Uganda, Uganda's Plan III, page 119. Attendances at the clinic are approximately 300 persons per month; of that number, 60-70 percent are new attendances. See U. 8., Agency for International Development, Population Prgf 4gpam,Assistance; Aid to the Developing Countries by the United States, Other Nations and International and Private Agencies, (Washington, D. C.: Agency for International Development, 1969), page 137. This statement is based on the assumption that an attendance represents a new "protected" woman, which is obviously not the case for a variety of reasons. See the following on the opera- tions and problems involved in contraception method effective- ness: Robert G. Potter, "The Multiple Decrement Life Table as an Approach to Measurement of Use Effectiveness and Demographic Effectiveness of Contraception," a paper presented at the Con— ference of the International Union for Scientific Study of Popu— lation, Sydney, Australia, 1967, (unpublished paper, Sydney, Australia, 1967); George Saxton and M. C. Pike, "Followup Study of 921 WOmen Using IUCD's in Kampala," (unpublished paper, Kampala, Uganda, 1967); and "Family Planning Technology: Pre- sent Status and Limitations" in Population Control: Implications Trends, and Prospects, edited by Nafis Sadik ggflal. (Islamabad: Pakistan Family Planning Council, 1969). Refer to the 22 100-bed rural hospital scheme of the second development plan, WOrk for Prggress. 10 of these facilities had been opened by October 1970, and construction of most of 16. 17. 18. 19. 20. 21. 66 the remaining 12 units had been started by that time. The rapid expansion of the number of rural health facilities occurred after 1968/69, when this development project began. Although there is some evidence to suggest that there have been several changes in classification criteria over the period, such changes were adjusted for through the author's use of constant criteria throughout the period. See J. Galea, "Inventory, Appraisal and Assessment of the Basic Health Services of Uganda: Developments for a Malaria Eradication Programme," (Jinja, Uganda: Malaria Pre—eradication Programme, World Health Organization, 1967), and World Health Organization, PlanningTRural Health Services, Technical Report Series no. 215 (Geneva: World Health Organization, 1961). The health center in Uganda had its genesis in the Frazer Com— mittee Report (Uganda Protectorate, 1956). The establishment of such a rural facility, however, did not occur until three years later. The growth in the number of health centers had occurred in two ways: (1) upgrading of existing smaller rural units and (2) develOpment of a complete new facility. The former of the two is the more common way of creating a new health center. See Appendix D, Table IL].for the number of health centers in the country from 1958 on. The analysis does not include mental hospital beds. There has been one other recent attempt to estimate the total expenditure on health in Uganda; see F. J. Bennett and G. Saxton, "Utilization of the Social Sciences for Some Aspects of Health Planning in East Africa", a paper presented at the Conference on Africa in World Affairs: the Next Thirty Years, Makerere University, Kampala, Uganda, 1969, (unpublished paper, Kampala, Uganda, 1969). For a discussion of the prob- lems involved in estimating the total health expenditures in any country, see Brian Abel-Smith, Paying,For Health Services: a Study of the Costs and Sources of Finance in Six Countries, Public Health Papers No. 17, (Geneva: Worth Health Organization, 1963) and Dorothy P. Rice, Estimating the Cost of Illness, Health Economics Series, No. 6 (Washington, D. C.: U. S. De- partment of Health, Education and Welfare, 1966). The gross domestic product figures are taken from the revised series in Republic of Uganda, Background to the Budget, 1970/71 (Entebbe: Government Printer, 1970). The revised figures show an increase in total economic activity when compared to the old series; the percentage of total GDP comprised the health expen- ditures was approximately one percent larger when compared to the old series figures. Data for the new series were revised to 1961, so that the 1958/59 figure for GDP had to be estimated. In estimating, the percentage that the old series figure com- prised of the new series figure was calculated. From 1961 to 22. 23. 24. 25. 67 1968 (the overlapping period), the percentage consistently moved from 71.5 percent to 81.8 percent. This trend was ex— tended to the earlier years and was used to derive a new series gross domestic product estimate for 1958/59. These percentages must be viewed with some caution, as they are first approximations. No attempt has been made to isolate the extent to which certain hospital costs may be related to preventive services, such as ante-natal care, child welfare clinics, and immunizations. Even if such information were available, however, it would not significantly change the major thrust of this analysis. See Republic of Uganda, Ministry of Health, Medical Services Statistical Records, lst July 1968 to 30th June 1969 (Entebbe: Government Printer, 1969), Table 12, page 17. The changing pattern of in patient admissions between the different facility type during the period 1958-1966/67 is interesting to consider. The apportionment of admissions between the various types of facilities in 1958 was as follows: government hospitals, 47%; government rural units, 38%; and mission hospitals, 15%. This pattern shifted in 1962/63 as a result of a rapid rise in number of attendances at rural units and an increase in number of mission hospitals in 1962-63, 37% of total admissions were to government hospitals, 40% to government rural units, and 23% to mission hospitals. As more government hospital beds became available, however, and as mission hospitals slowed their expansion, the structure of inpatient admissions shifted again; in 1966/67, 44% of total inpatient cases were admitted to government hospitals, 30% to government rural units, and 26% to mission hospitals. See World Health Organization, International Classification of Diseases (Geneva: World Health Organization, 1955, and revisions, 1965). A11 diseases have been grouped by WHO into 17 different major classifications which (a) conform to basic life systems (circulatory, respiratory, alimentary, etc.), (b) are related to a specific major disease (e.g., cancer), or (c) are related to the causality of disease (e.g., parasites, accidents, etc.). Appendix F, Table F.1 indicates the correlation between W.H.O. classification of diseases, Uganda's classification, and the author's classification, which deviates in only one respect from Uganda's. The extent of the discrepancy between manpower registration and manpower actually practicing in some form of medical service in Uganda is indicated by Bennett's estimation in 1964 that there were 375 doctors practicing in Uganda which the registra- tion list for the same year shows 651 doctors. See F. J. Bennett; S. A. Hall; J. S. Lutwama; and E. R. Rado, "Medical Manpower in East Africa: Prospects and Problems," East African 26. 68 ‘Medical Journal, 42, 4 (April 1965), 149-161; and Mark Wheeler, "Medical Manpower in Kenya: A Projection and Some of Its Im- plications," East African Medical Journal, 46,2 (February 1969), 93-101. Bennett ggual., Medical Manpower, p. 151, estimates the urban ratio for Uganda at one doctor per 4,000 persons; the estimated rural ratio was one doctor per 26,000 persons. The ratios for Kenya and Tanzania were even more inequitable. CHAPTER THREE In this chapter, a relevant concept of output for health sector planning (particularly in regard to curative health services) is developed. This issue must be addressed, for without a conceptually sound theoretical formulation of the nature of the product of the health service system, a case study of the allocation of resources to sudh a system has little validity. The concept of output developed in this Chapter will be amenable to quantitative estimation so that changes in output may be determined.1 Perhaps the most desirable conceptualization of health services output would relate changes in inputs into the industry (such as (a) a larger number of doctors or more hospital facilities, (b) an in- crease in new preventive programs, or (c) changes in the larger socio- economic environment such as increased economic output, increased literacy, etc.) to changes in an overall index of the health of the society.2 Such an index would be desirable for any society engaged in planning its socio-economic development. Even if an Optimal theoretical measure were agreed upon, however, the statistical and data collection problems are overwhelming; even the most wealthy countries have not yet developed sudh an index. What are the alter- natives, then, that are both feasible from a statistical point of view and have theoretical support? In the field of health economics, several measures have been used as indicators of output. Measures relating pOpulation to various types of health resources (such as doctors and other personnel, as 69 70 well as facilities) have been used frequently either for health planning purposes or for international comparisons. In such con- texts, a "need" perspective has permeated the planning process and has considerably reduced the analytical usefulness of the exercise.3 These ratio measures have been utilized without reflection since they provide gross measures of the distribution of resource inputs and have no direct relation to any output measure. The use of such in- put measures as planning 30818 and, subsequently, the criteria for evaluation of implemented projects,not only assumes fixed prOportions between inputs and outputs but also indicates a lack of conceptual understanding of the nature of the problem.4 Deve10pment of an adequate conceptual framework requires clear answers to the following questions: What are the objectives of a health service system? What are the resource inputs into the system? What is the production process? What is the process producing? Before discussing other output measures that have been used in economic studies of health services, it is important to distinguish between several types of sectors which exist within the entire health services industry. Each sector has its own resource requirements, production processes, and outputs, which develop from the principle objectives of each sector. For purposes of the analysis presented here it is appropriate to distinguish between curative and preventive health services. In the case of preventive health services, an output concept may be derived rather readily from the rationale or objective for providing 71 the service in the first place. Given an effective immunization program, for example, reductions in disease-specific mortality and morbidity rates comprise a fairly precise measure of the output of that particular program. An economist, however, would add another dimension to that output concept: an estimate of the economic pro— duction gained (as a proxy for an increase in welfare) as a result of the reduction in mortality and morbidity rates.5 This gain is most commonly measured in terms of the present value of foregone income streams.6 In cases where a given public health project is designed to eradicate a wide—spread disease such as malaria, or to change significantly an important population or demographic variable such as the infant mortality or fertility rate, it may be more import- ant to analyze the program's effect on a society from a broader per- spective which would not only take into account the program's effect on specific objectives such as disease-specific incidence rates, but also the multiple social and economic effects resulting from a major change in the incidence of the disease. In such cases, it would be important not only to identify properly the primary "outputs" of a particular project, but also to identify how these outputs are related and how the resulting effects can be measured in terms of a single variable. In the case of economic analysis, the most common variable used to measure the impact of a multiplicity of economic effects is per capita GNP.7 In discussing curative health services, we are primarily inter- ested in the set of diagnostic and treatment services which are most 72 commonly provided in the hospital or health center facility. It is within the confines of such facilities that most, if not all, cura- tive services are provided.8 The question which is of primary interest here is the following: what does the curative health service system produce, and what is an appropriate output measure of this production? Review of Past Conceptual Development In previous studies of the economics of medical care where an indicator of output was essential to the analysis, a primary object- ive of the analysis has been a determination of whether or not economies of scale exist in the set of medical services provided in a hospital setting.9 In addition to such empirical studies, other analyses have focused on the concept of output from theoretical or planning perspectives.10 Upon reading these empirical, theoretical and planning studies, several features manifest themselves. They are important in understanding the type of output concept which has been developed for a curative health services firm or industry. First, a homogeneous measure of output has been used for two major reasons: (a) ease in analyzing the economics of a hospital within the framework of the microeconomic theory of the multi-product 11 firm, and (b) ease in applying statistical or econometric models to hospital data. The requirement of homogenity has led to the frequent utilization of one or more of the following indicators as a measure of the output of a hospital (particularly where the analysis has been limited to those services rendered to inpatients): (a) number of cases, (b) patient days, (c) bed days, or (d) patient weeks. 73 In analyses which have included outpatient services, common output measures used have included either (a) the number of patients, or (b) the number of patient visits. A second feature of the commonly used "output" measures is that most of them have only an obscure relationship to the set of individ— uals who are the recipients of service. In measures such as patient days or number of visits, the emphasis is placed on the extent of the resources used to provide service (i.e., inputs) rather than on the recipients of service (i.e., outputs). Such measures may have a useful function in the development of certain internal management control procedures, but in the context of a broader conceptual per- spective, the measures lack both comprehensiveness and adequate focus on the rationale for providing health services. Use of one of the above—mentioned indicators of output implies that one or more of the following issues will be given only cursory attention: (a) to whom are curative health services provided; (b) why are health services demanded in the first place; (c) if human beings are the recipients of health services, is it possible then to conceive of the situation where one individual may require a differ- ent set of services than another individual, depending upon disease, age and sex characteristics; and (d) is it possible to conceive of the possibility where, for a variety of reasons, individuals may respond differently to the same set of services provided? The use of measures such as patient days, bed days, patient weeks, and patient visits do not seriously consider that health services were produced for human beings. How can the situation be improved? 74 Martin Feldstein, in a study of the British National Health Service, provides an important beginning towards improving the situation by rejecting the need for a homogeneous output concept and uses an output mix concept, where the mix is specified accord- ing to the type of case treated.12 For practical reasons, Feldstein's specification of case type was defined as a result of the way in which British hospitals group patients according to major service department,or service rendered within the hospital, such as surgical, gynaecological, E, N & T, medical, etc. He suggests that specifying ouput mix according to the characteristics of disease type, age and sex may also be a theoretical improvement. Further Conceptual Development In discussing conceptual development, it is useful to explore a rather fundamental question, alluded to above: what is the raison d'etre of curative health services? Individuals and, increasingly, entire societies, place a rather high value on human life; when life is JeOpardized by illness, resources have been committed increasingly to restoring, to the extent possible, a prior state of well-being. Most resources so committed today are for the provision of curative health services. Such services thus do not exist for themselves, but rather for the restoration of individuals who have contracted an illness.13 Given that curative health services have a rationale only to the extent that human beings have been afflicted and will continue to be afflicted with disease, it is important to consider the possiblity that a logically sound measure of the output of curative health 75 services would have a direct relationship to the number of human beings who have comsumed such services. The following discussion focuses on an economic rationale for the provision of curative health services and reinforces the idea that the individual (the patient), or a set of individuals, comprises the essential basis of a relevant output concept. Additional information, related to the individual characteristics of those receiving service, as we shall see, can greatly improve the level of understanding derived from any study of health services, but the individual is primary. The economic rationale for health services can be presented best as follows. Let us assume that there is a set of individuals -- perhaps called a society -- which, for ease of understanding, is not expanding (i.e., we rule out population growth). This set com- prises the 1abor input for the production of all goods and services consumed by that society, assuming a closed economy. These individuals can be viewed as a stock of human capital, which presumably can be augmented by the application of education but which can also depreciate primarily as a result of aging and illness. One way to reduce the rate of depreciation is to interject a set of health services designed either to prevent certain illnesses, or to restore individuals ex- periencing a sickness episode, as much as possible, to their prior state of health. The latter restorative process is commonly pro- vided by curative health services. It provides a "repair" or main- tenance function for the continuance of a certain stock of human capital. Thus, in an analysis of the economic performance of the curative health services industry, an appropriate output measure 76 must be based on the number of individuals successfully maintained or repaired during a given period of time; this group, whose rate of depreciation has been reduced, comprises a portion of the human capital stock of the society. In addition to emphasizing the restored individual in the con- ceptualization of the output of a health services firm, it is also important to consider that variations in the quality of the services rendered during the treatment process can have an important impact upon the output of the health services firm. Understanding of the production process of curative health services thus is important in the development of an improved output concept.14 Output Specification for the Health Services Firm An analysis of the production process can be undertaken in the following way. Each day, a number of persons with perceived sickness episodes arrive at a health facility. They seek some diagnostic ser- vice in order to determine (a) the nature of the illness and (b) the appropriate treatment services for curing or reducing the discomfort (either physically or mentally), in order to return to their primary activities and responsibilities. The number of persons seeking and securing at least one such service during a specified period of time is represented by 8. Some persons, upon receipt of one or more diagnostic services, are considered to be very sick and, assuming the resources required are available, are admitted for further services provided on an inpatient basis. Such individuals are likely to consume a different set of curative services, or at least a different amount of any given 77 set of services, than those who are treated on an outpatient basis, primarily because the illness episode is more severe. In such cases it is useful to distinguish between the two groups initially demanding service, on the basis of treatment provided. Returning to an analysis of the entire curative treatment pro- cess without distinguishing between the treatment intensities, we note that there are other individuals who are referred to a different facility for various reasons, such as distance from home and the availability of special types of services which cannot be provided at the original facility due to lack of staff, treatment equipment, or drugs. These individuals may receive partial treatment as out- patients in the initial facility, but cannot be considered a com- pleted unit of output and, must be deducted from S. The potential for transfers from another facility to the given facility must also be recognized, but for convenience can be considered a part of the initial group of S individuals seeking service.15 There are still other persons who die as a result of their ill- ness, regardless of the curative services rendered. (Empirically speaking, virtually all persons who die at a health facility in Uganda have been admitted to the inpatient treatment process. The location of death, however, is not important for present conceptual considera- tions.) Given the fact that some persons die at the health facility while in receipt of some set of curative services, it is necessary to adjust potential output to account for this occurance. One final adjustment, proceeding from an analysis of the entire treatment process, is theoretically useful to explore. As noted in the discussion of the economic rationale for the provision of curative 78 health services, an important objective of such services is to restore the capacity of an individual to resume his major activity. Thus, one must investigate the extent to which persons seeking and receiv- ing medical services are able to resume their major activity subsequent to a normal recovery period. Those who are unable to resume major activities after such a period have consumed a certain set of health service resources, but cannot be viewed as economically viable output. Therefore, the cost of producing a "successfully treated" patient, e.g., one who can resume his major activity, must include the cost of those who did not benefit to the extent that they were able to 16 This unsuccessfully treated set of 17 resume their major activities. individuals must also be deducted from 8. By reducing the initial set of individuals by these groups, a net output concept for the health services firm (facility) is derived. Symbolically, (1) P = S - St - Sd - Su where S is defined as above, and P - output of a curative health services production unit (i.e., hospital, health center, or dispensary), St a the number of persons provided with a curative health service who are transferred to another health service facility for further treatment. Sd a the number of deaths of 3 Su 8 the number of persons who were treated but were unable to resume their major activity subsequent to treatment. It must be noted that the last set of individuals deducted from the initial group of persons seeking services implicitly involves a measure of the quality of medical care. The literature is lengthy 79 on this issue and the quality of medical care has been analyzed from many points of view and by several disciplines. Physicians, for example, have been interested in whether patients receive the appro— priate diagnostic and treatment services from appropriately trained staff. Nurses have been concerned with the quality of the care they provide to patients primarily on an inpatient basis. Pharmacists have been concerned with the rate of improper drug use as well as dosage. Administrators and public health officials have a general concern for the quality of the entire set of health services provided either within a given type of health firm or by the service system in general. Economists as well have considered medical care quality and changes in the quality as an important cause of relatively rapid increases in costs of medical care services.18 It is likely that each approach to the quality of care has a rationale of its own, depending upon the objectives of the investigator and his field of specialization. Whether any approach has developed an adequate empirical measure is another issue to be determined by the available evidence. From the perspective developed in this investigation, however, it is consistent to introduce a quality variable into the analysis which has, as its rationale, an output derivation. It seems logical for purposes of economic analysis to view the quality of any given curative health services treatment center as dependent upon the extent to which the individuals receiving services are rendered able to resume their normal major activities. In using such a measure of quality, it is recognized that persons may, for reasons other than the health services received, "get well" or remain ill. However, it 80 is theoretically possible to measure the frequency of such occurances and thus allow for that possibility by acknowledging the asymptotic nature of the measure. By using the method of follow-up surveys of service recipients, one can derive the necessary information required to estimate values for the quality variable defined. Instead of de- fining the variable in terms of absolute numbers of persons who have regained the ability to perform their previous major activities, the variable may also be expressed as a probability of occurance of suc- cessful treatment among the number of initial demanders S. This method of expression generally has greater usefulness in analytical models and it is incorporated in such a manner in subsequent chapters. Specificatiog_of the Output of Each Treatment Process As indicated earlier, it is useful to analyze the output of the two treatment processes separately. In many countries, as in Uganda, the two basic treatment processes of a health services firm have several differences which provide an important rationale for develop- ing a conceptual framework for determining the output of each treat— ment process separately. These differences induce (a) the relative seriousness of the illness of individuals treated, (b) the range of services available, and (c) differences in technical training of the manpower providing the services. The output of the total firm, thus, can be specified as some combination of outputs of the two primary treatment processes, sudh that, (2) P = 0P1 + 8P0 where P is defined as above, 81 Pi = output of the inpatient treatment process, P0 = output of the outpatient treatment process, and where a and B are non-specific weights which derive meaning when a clearly defined set of objectives for the health sector has been specified. (In a typical multi-product firm, the weights would assume the values of each product's price). A further analysis of the production process provides one with the insight necessary to specify the output of each method of treat- ment, P1 and Po' Let us assume that all demanders of service enter the system as outpatient cases where initial diagnostic services are provided. Given this perspective, it is convenient to specify the output of the outpatient treatment process. After the patient receives diagnostic services, a decision is made as to the appropriate method of treatment. For two groups of persons, subsequent outpatient treatment is considered inappropriate, particularly for those persons requiring inpatient care and those persons transferred to another facility for reasons mentioned above. These groups are deducted from the initial group of persons demand- ing services. Those treated on an outpatient basis but who are unable to resume their major activities subsequent to treatment must also be deducted. Thus, (3) Pas-si-sto-s O 110 where PO and S are defined as above, and Si - the number of persons provided with at least one curative health service on an inpatient basis at a given health service facility, Sto = the number of persons provided with a curative health service who are transferred, at some point in the out-patient treatment process, to another health service facility for further treatment, 82 Sun . the number of persons who were treated on an outpatient basis but were unable to resume their major activity subsequent to treatment. A similar derivation process can be undertaken to determine analytically an appropriate concept of output for the inpatient treatment process. The relationship between the outpatient and inpatient processes, as referred to above, takes on added signific- ance in the development of the inpatient output concept. Some authors have noted, in analyzing the health service delivery system in the U. S., that the demand for hospital services in that country is a derived demand -- derived from the additional service require- ments doctors deem essential for some of their patients in order that the individuals' major activities may be resumed.19 A parallel situation exists in the curative health service system of Uganda, where the demand for inpatient health services is a demand derived from the provision of certain diagnostic services which are provided primarily in the initial stages of the outpatient treatment process. Thus, those persons who were deducted from the output of the out- patient treatment process due to a determination that inpatient care was required become the total potential output of the inpatient treatment process. It is possible to determine -- in the same manner as for the outpatient treatment process -- the adjustments necessary to derive a measure of net output from the total potential output for the in- patient treatment process. In certain cases, it may be necessary to transfer patients to a more sophisticated treatment facility. (The most frequent cause of sudh transfers in Uganda are obstetrical or 83 gynaecological complications). Other persons who are admitted to the inpatient treatment process die, regardless of the curative services rendered. (Without great loss of empirical validity, it is assumed that all deaths occuring at a health facility do so during or sub- sequent to admission for inpatient care). Still other persons are unable to resume their normal activities after receiving treatment services and for that reason, must be deducted from the initial group of inpatient health service recipients. Symbolically, the derivation of a net output concept for the inpatient treatment process can be stated as follows: (4) P1 ' 51 " Sti " Sd " Sui where P1, Si, and Sd are defined as above, and Sti and Sui are the same as Sto and Suo’ except that they refer to the inpatient treatment process rather than the outpatient process. The Issue of Output Hompgeneity, To this point, it has not been made clear whether P is a homo- geneous or non-homogeneous variable. Given the specification of equa- tion (1) and the subsidiary treatment process equations (3) and (4), it has been implied that all variables are homogeneous. However, by excluding non-homogeneous characteristics from the analysis of the output composition of a health service firm it is difficult to under- stand why the cost of providing service may vary between otherwise similar health services firms. Let us explore a means through which these Characteristics, in accordance with Feldstein's concept of output mix, can be introduced into the analytic conceptualization of output of the treatment processes described above. 84 For conceptual purposes, three individual characteristics will improve the analytical usefulness of output: age, sex, and disease type. It is possible to specify a facility's output as a combination of the individual subsets of patients -- subsets which are defined by the three characteristics of interest. Let us begin by specify- ing a particular subset of a facility's total output. The specifica— tion of a subset of either or both treatment processes is analogous in its methodological development. A subset of a facility's output can be denoted as follows: Pj’ where the subscript j specifies the particular age, sex, and disease characteristics of that subset. Thus, j = (1,. . . ,n) which includes the j possible combinations of the three characteristics of interest. Total output can thus be specified as follows: P = ZijPj, where Y3 represents some non—specified weight of the import- ance or value of each P3 of the total output. In order to determine the subset outputs, i.e., each Pj, it is necessary to disaggregate the other terms in equation (1). For example, in order to determine the net output of the subset P15, which may represent the output of successfully treated males, aged 0-5 years, who suffer from.a respiratory disease, it is necessary to know what proportion of the other variables, 8, St’ Sd, and Su are males, aged 0-5 years, suffering from a respiratory disease. In a general- ized form, any output subset can be determined by (5) P1 " SJ " St: ’ 3d: ’ Suj The output subset for both treatment processes is similarly defined by disaggregating the variables in equations (3) and (4) for 85 the outpatient and inpatient treatment processes respectively. Thus, for any outpatient output subset Poj’ (5) P03 ' Sj - Sij - stoj - Suoj and for any inpatient output subset Pij’ <7) P13 ‘ $13 " 5:13 ‘ de ‘ Sui: With the deve10pment of an age, sex, and disease-specific out- put, it is useful to analyze health service facilities as multi- product firms whose total cost function is not only related to the total number of persons successfully treated (i.e., consistent with a homogeneous output concept), but more importantly, to the output mix, i.e., the proportion of total output represented by each output subset. By using this approach, one can analyze the impact which the following factors have on total cost: (1) different proportions of output as defined by the patient characteristics of that output; (2) differing proportions of output from the two treatment processes which may exist in the facility; and (3) different proportions of persons who seek health services -- assuming there is no change in the output mix in terms of treatment process proportion of total output and patient characteristics -- who are transferred, die, or are unsuccessfully treated. By using such an output conceptualiza- tion, it is also possible, from a planning perspective, to improve upon present analyses of the effect on cost of potential policy changes in the curative health services sector such as: (1) changes in emphasis as between health center and hospital development; (2) changes in government-voluntary institution relationships; (3) changes in staffing patterns; and (4) changes in the availability of (a) 86 supporting diagnostic services, such as laboratory and radiographic facilities, and (b) drugs. In the chapter which follows, the analytical framework necessary to focus on the effects of output and policy change on the cost of delivering curative health services is deve10ped. 87 Footnotes It is significant that Rothenberg suggested in 1962 that increased attention be given to improving upon the output concepts then in use when analyzing health service systems. To date,this research priority has generally received only cursory attention. See, Rothenberg, J., "Agenda for Research in the Economics of Health", in The Economics of Health and Medical Care, Proceedings of the Conference on the Economics of Health and Medical Care, May 10-12, 1962, (Ann Arbor, Michigan: The University of Michigan, 1964) p. 314. See Sullivan, D. F., Conceptual Problems in Developing an Index of Health, National Center for Health Statistics, Series 2, No. 17, Washington, D. C., 1966, for a discussion of the conceptualization problems involved in (and a practical second-best method for) developing such an index in one of the more economically developed countries of the world. The sc0pe of this paper does not include a discussion on concept- ualization problems related to the definition of health. The World Health Organization definition is admirable but unmanagable from a quantitative point of view. We concur with Wylie, M., "The Defini- tion and Measurement of Health and Disease", Public Health Ref pgggg, 85, 2 (February 1970), pp. 100-104, and Sullivan, D. F., pp, gi£., that a more limited definition of health which focuses on changes in mortality and morbidity, or on what may be termed minimization of sickness episodes, is a positive first step in deve10ping more adequate indices. See Boulding, K. E., "The Concept of Need for Health Services," Milbank Memorial Fund Quarterly, 44, 4, Part 2 (October 1962), See King, M., editor, Medical Care in Developing_Countries: A Primer on the Medicine of Poverty and a Symposium from Makerere, London, Oxford Univeristy Press, 1966, p. 1:10a. King describes the con— ceptual problem in the following way: it "is equivalent to a motor manufacturer trying to maximize not his output of vehicles, but the number of his workers and the size of his factory." It is unfor- tunate that a subsequent study of the development of health services in Malawi, he commits the mistake he earlier warned against by de- fining objectives in terms of input measures. See Feldstein, M., "Health Sector Planning in DevelOping Countries," Economica, New Series, 37, 146 (May 1970), pp. 139-163 and Weisbrod, B., Economics of Public Healph: Measuringgghe Economic Impact of Diseases, Philadelphia, University Press, 1968. In addition to mortality and morbidity (in terms of a loss in the ability to perform major activities such as work), Selma Mushkin identifies another way in which sickness can effect labor pro- ductivity. She describes it as debility -- the loss of productive 11). 88 capacity to perform a major activity even though one is still able to perform it. Unfortunately, the debility concept has not been defined operationally nor have indices been developed to measure its magnitude. See Mushkin, S. J., "Health as an Investment," Journal of Political Economy, 70, 5, Part 2 (October 1962), pp. 138-143. See Barlow, R., The Economic Effects of Malaria Eradication, Ann Arbor, Michigan School of Public Health, University of Michigan 1968, and Coale, A., and Hoover, E., Population Growth and Economic Development in Low Income Countries, Princeton, New Jersey, Princeton University Press, 1958, as examples of empirical studies using per capita income as the primary variable of analysis. I am aware that in some western countries, e.g., the U. S., a great number of diagnostic services are provided by doctors out- side the confines of such facilities, but for the purposes of the present analysis, it is easy to conceive of an institutional situation where all diagnostic services occur within the con- fines of the hospital or health center framework. Alternatively, one may view private doctors' offices as adjuncts to the rest of the curative service framework. For example, see Feldstein, M. 8., Economic Analysis for Health Service Efficiency: Econometric Studies of the British National Health Service, Amsterdam, North-Holland Publishing Co., 1967; Feldstein, P. J., An Empirical Investigation of the Marginal Cost of Hospital Services, Chicago, University of Chicago, 1961; Lave, J., and Lave, L., "Hospital Cost Functions," American Economic Review, 60, 3 (June 1970), pp. 379-396; and Ingbar, M. L., and Taylor, L. B., Hospital Costs in Massachusetts: An Econmetric Study, Cambridge, Massachusetts, Harvard University Press, 1968. See Abel-Smith, B., Paying for Health Services, A Study of the Costs and Sources of Finance in Six Countries, Public Health Papers No. 17, Geneva, World Health Organization, 1963; Feld- stein, M. S., Economic Analysis for Health Service Efficiency: Econometric Studies of the British National Health Service, Amsterdam, North-Holland Publishing Co., 1967; Feldstein, J., Research on the Demand for Health Services, reprinted from Milbank Memorial Fund Quarterly, 44, 3, Part 2 (July 1966), by the Health Services Research Study Section of the United States Public Health Services, 1966; King, M., 22, 215.; New- house, J. P., "Toward a Theory of Non-profit Institutions: An Economic Model of a Hospital," Amerigan Economic Review, 60, 1 (March 1970), pp. 64-75; Health Planning: Problems of Concept and Method, Washington, D. C., Pan American Health Organization, Pan American Sanitary Bureau, Regional Office of the WOrld Health Organization, April 1965 (scientific publication No. 111); and Statistics of Health Services and of their Activities, Thirteenth 'Report of the WHO Expert Committee on Health Sta- tistics, Geneva, World Health Organigation, 1969 (Technical Report Series No. 429). ll. 12. 13. 14. 15. 16. 17. 89 This is not meant to imply that all such applications are not useful for developing a better understanding of certain aspects of the economics of hospitals. Feldstein, M. 8., Economic Analysis for Health Service Efficiency: Econometric Studies of the British National Health Service, Am— sterdam, North-Holland Publishing Co., 1967. The output mix concept was introduced into the literature at this same time by P. D. Bonnet, "Increased Production and Better Utilization", 327 port of the NatiopalpConference on Medical Costs, June 27-28, 1967, U. S. Department of Health, Education and Welfare, Wash- ington, D. C., 1969. I am aware that some illnesses may not be thought of as "real", but demands are placed on the service system irrespective of the scientific reality of some conditions. See Donabedian, A., "Evaluating the Quality of Medical Care", in Shulberg, H. C.; Sheldon, A.; and Baker, F., eds., Program Eyaluation in the Health Fields, New York, Behavioral Publica- tions, 1959, pp. 214 and 215 for a similar point of view. The essence of Donabedian's view is as follows. "Greater neutrality and detachment are needed in studies of quality. More often one needs to ask, 'What goes on here?‘ rather than, 'What is wrong; and how can it be made better?‘ ....Emphasis must be shifted from preoccupation with evaluating quality to concentra- tion on understanding the medical care process itself". For certain analytical purposes, however, it may be useful to include them in the analysis as a specific group, to be included in a final net transfers figure. The major activities concept has been developed by the National Center for Health Statistics for the United States National Health Survey. It can be operationally defined for any society to conform to the social and cultural roles expected of each age and sex group. The concept does not specify a minimum level of role performance, nor, if the activity is basically an econ- omic one, does it specify a minimum productivity performance. Thus the concept as presently defined and as it is used in the present research, is related solely to a morbidity measure, rather than to a measure combining morbidity and debility. See footnote 6. This argument does not imply that certain persons, as a result of a poor initial prognosis or any other reason, should be denied health services; they are unfortunate, as determined by the technical capabilities of the given set of health services. However, if the age, sex and disease specific rate of occurance is known and monitored over time by staff involved in the de- livery of health services, perhaps incentives -- internal or otherwise -- would develop to initiate a reduction in such rates. 18. 19. 90 See Barker, K., Kimbrough, W., and Heller, W., A Study of Medicatiop Errors in a Hospital, university of Arkansas, University of Mississippi Press, 1968; Donabedian, op. cit., Feldstein, M. 8., Economic Analysis for Health Service Ef- ficiency: Econometric Studies of the British National Health Service, Amsterdam, North-Holland Publishing Co., 1967; Shapiro, 3., "End Result Measurements of Quality of Medical Care", Milbank Memorial Fund Quarterly, 45, 7 (1967); Sheps M. D., "Approaches to the Quality of Hospital Care", in Shulberg, H. C.; Sheldon A. and Baker, F., eds., 22°.21£-: pp. 286-303, and Thompson, J. D., Marquis, D. B., Woodward, R. L., and Yeomans, R. C., "End-Result Measurement of the Quality of Obstetrical Care in Two U. S. Air Force Hospitals", Medical Care, 6, 2 (March-April 1968), p. 131. On the issue of quality, see also, Statistics of Health Services and Their Activities, WHO Technical Report Series #429, pp. 29-31. See Feldstein, P. J., Research on the Demand for Health Services, reprinted from Milbank Memorial Fund Quarterly, 44, 3, Part 2 (July 1966), and Hixson, J. S., The Demand and Supplyyof Profes- sional Hospital Nurses: Intra—Hospital Resource Allocation, an unpublished Ph.D., dissertation, Michigan State University, East Lansing, Michigan, 1969. CHAPTER FOUR In Chapter Three, the production process of a health facility was discussed in order to focus on the concept of output for the health service industry; this chapter will explore the relationship between outputs, inputs, and services which exist in the health service firms of Uganda. At the present time little empirical or theoretical work has been conducted on these basic relationships to improve the under- standing of how this service system operates. Without this basic know- ledge, attempts at planning the future development of the health services - in light of the government's recently announced rural development strategy - or analyses of potential health policy changes will be conducted on an ad hgg_basis. Thus, the objectives for this chapter are (a) to develop a theo- retical framework which will enable health and economic planners to understand the relationships between the inputs and outputs of the curative health service system and (b) to develop the theoretical framework which will encourage health planners to consider changes in population growth, disease mix, and the rate of economic growth as they evaluate the short and long run effects of health policy. In the next chapter empirical investigations are conducted which use the method- ology developed in this chapter. Statement of the Problem and a Consideration of Alternative Methodologies In order to improve the understanding of the relationship between the output produced and the resources used in providing health services, it is useful to analyze the production processes which are 91 92 found in differing health service firms. An introduction to this analysis is found in the previous chapter and these relationships are formalized later in this chapter. Given that the production relationships existing in Uganda's health service system can be adequately described, the theoretical development turns to an exploration of potential performance indicators which can be used by society to determine the efficacy of alternative institutional mechanisms which provide health services. Assuming that certain economic and curative health policies have been and will be implemented in the country, the analysis then turns to exploring the output and cost implications of these policies. The analytical frame- work is specified such that the implications can be made explicit for a ten year planning horizon. The relevant question to raise at this point is, given the problem objectives formulated above, what are the potential methodological frameworks which can be employed and which is the most efficient? Al- though the question posed above can be formulated as a fairly standard production problem faced by the optimizing firm or industry, the fact that (a) the problem is complicated by the multi-product nature of health services firm and (b) the exercise has a raison détre in terms of improving the managerial and planning capacity of the Ministry of Health and other health sector decision makers, it is useful to ex- plore a methodology appropriate for the attainment of such objectives. Therefore, instead of employing the analytical framework based on usual theoretical and mathematical assumptions such as (1) continuous and differentiable and (2) convexity, which was done by Pfouts1 and 93 further explored by weil2 for a multi-product firm, the application of programming methodologies is more apprOpriate.3 In recent years a body of literature has deve10ped in which the problems of resource allocation in the health care sector has be- come subject to increasing analysis. The techniques of linear pro- gramming, operation, research, and systems analysis have been increas- ingly used to conduct such analyses." Although it is argued in Chapter Three that most analyses of health services conducted in the past have not adequately specified the output of curative health services, and as a result have not adequately specified the objective function of health services, the formulation of the resource allocation problem in a programming framework has assisted in developing a better under- standing of some of the production relationships existing in different types of organizations used in the provision of health services. By respecifying the objective function in a more appropriate manner for use in linear programming analyses of the health service system, it is envisioned that the problems of previous studies which have employed programming techniques can be overcome, thus enabling such techniques to be profitably employed. After an initial section of the chapter in which the formal pro- duction relationships found in Uganda's three main types of curative health service firms - the hospital, the health center, and the sub- dispensary - are schematically developed, the analysis turns to developing a linear programming framework for use in analyzing the curative health service system. In the first instance, the production activities and constraints are described, after which, a detailed 94 statement explores alternative objectives which may he used in deve10p- ing a maximand for the model. The model is then specified in such a way that it incorporates both the importance of quality of the health services in the objective function, as well as the cost of providing the services for purposes of managerial control and planning. Finally, the model is respecified in order to analyze some of the important factors affecting the health service system.over time. In this context, the use of such techniques as recursive and parametric linear program— ming can be employed as a "...useful tool for both descriptive and predictive inquiry". 5 Production Process of the Health Service System As noted in the previous chapter, the production of curative health.services in all facilities can be viewed as a process consist- ing of two phases: the diagnostic phase and the treatment phase. Resources (denoted vr, where r-(l,...,m)) such as medical personnel and supplies are allocated between these two phases of the production process. Within each phase of the production process, a set of services is provided to a certain proportion of the total number of persons demand- ing curative service. In the rural facilities (Figures 4.1 and 4.2), the only service provided in the diagnostic phase is the diagnostic consultation by trained auxiliary personnel (medical assistants). This diagnostic service has the following components: (a) discussion related to the patient's medical history; (b) notation of primary symptoms; (c) a decision to treat at the local facility; and if treatment SIS Sick Poopls Question: What Type of Medical Service? "Trsditionsl" or "Modern” Exit Diagnostic Phase Traditional Diagnostic a Phase vpk -——I Diagnosticisn Question: What frost-nut? Send to Transport . trust-Int Phase Figure 4.1: Another‘T' Services ' V, Facility k Exit: T, Treatment . Phase Notes: (1) 5 refers to the vector of persons having characteristics 3, and who seek at least one curative health service in the modern sector. I I I I I I I c. (2) T refers to the vsctor of persons ving characteristics 1 who are transferred to another hsslth uctl‘t’s Minor Surgery I Treatment rt (3) : refers to the vector of medical inputs - factors of production - ”Quired to produce the L1! curative service. (4) ‘3 refers to the vector of "success- Dt“! fully trusted" persons having Treatment characteristics 1. 4. rt o (5) D refers to the vector of persons having characteristics 3 who die while receiving services. (6) 03 refers to the vector of "unsuc- oghgg cessfully treated" persons having rrggcngng. characteristics 1. k-(3,...,l) Successful Unsuccessful Treatment Trsstnsnt 1 ‘ ‘“E3 ’1 C 2:13:01,“ 1.1D Sub-dispensary and Aid Post Outpatisnt Trust-ent Process 4. __L_____.________ mg— i l I I .-1I .__II 96 J Figure L2: Ij Dlupmmnry-Mntcrnity Unit, Health Louie! Treatment Process Predefined through the Dia- gnontic Service Phase r1 . 0.1) rt Diagnostic L Phase Send to , Another -——e Transport Services I Exit: 1, Facility Treat as Inpatient Innetient_ or ' tpati - Predefined Treat th:::t Outpatient Inpatient Treat-ent Treat-ent Phase Phase (Pig. 5 ) L r_. l Successful Unsuccessful * I P"“ V _u Treat-ent Treatment rk Drugs ___. I I I C I [nulls G1uzlds+$j)I I _|___.I ' I—“ Surgery ; I Food e Nursing led V“ H and Custodial I Service * I Other v‘k " Treat-ent I Services . Maternity V I Service I Note: A - the proportion of the 1th patient type entering the , . Outpatient Treatnent Phase. ~-—-.———.__.-r- if, Successful Unsuccessful Treat-ent Treat-ent Exit: (1-0 :1 lair: (1-" ¥ 97 is to occur, (d) a decision concerning the amount and type of treatment services.6 In both government and mission hospitals, the diagnostic phase is not only characterized by the diagnostic services performed by auxiliary (nonrprofessional) manpower, but also includes supporting services: professional diagnostic, laboratory, and radiographic ser- vices. As noted in Figure 4.3, all persons demanding curative service at a hospital first receive diagnostic services from trained auxiliary personnel. The auxiliary then must decide whether to use one or more supporting services. If the patient receives laboratory or radiographic services, he returns to the diagnostician who requested such services. Upon completion of the diagnosis, a decision is made whether to treat, and, if it is to treat, then the appropriate set of treatments is determined. In the treatment phase, services are designed to restore health or minimize the effect of the illness episode. The demands placed upon the various treatment services available at the health facility are determined by the diagnostician. As a result of training and practice, medical personne1--both professional and auxiliary--have deve10ped a standardized set of treatments for given diagnoses.7 In the Figures 4.1, 4.2, and 4.3, however, a specific set of treat- ment services for a given disease has not been specified; the range of the important treatment services is noted, and the direction of flow indicates that a person may receive any one or all of these services in a variety of sequences (depending upon decisions made by the diagnostician) while in the treatment phase. 98 essou one-useuk n~.«.u«ao eeu:h ocean-one unuauenon vwauueveum wosuoufi nae-aeouh donation an.¢ afiaidh ocean-ausc venuueueum ea seeooum announcer so» “due use ness:: use-seen» en» «a» no unease ssh .eael III I ' .«3 8.5a 3 In... as: 3 13:53 h owe-nuso no eeuwpusm use neon ee . queen“ . uuueoeuean decouaseueu- Is as» 3H: e an E» .e. «my ICOOOHQ “Chg: ‘CHgU‘ onenessuh how «can .. eeou>uem ouueonue . esuu>hem equeonueu. escapee» \ueuueuel nesuoc< uoauo so» genome-euoum Aboueuoned venuuuueuh on anew enact nauseous“: oz eeuu>uem eeunsheu an» 1 3308-3 5.x e uneuuwx=< H.u«hv sneak savanna ego-onus“. eeu.ou use: 38.. C, Juan 99 The diagnostician in hospitals and larger rural units has one additional decision to make; he must determine whether to treat the individual on an inpatient or outpatient basis. If he decides to admit the individual for inpatient treatment, two additional treat- ment services are provided: (a) a composite nursing, bed, and custodial service, and (b) food services. (In addition to these services, there is a separate maternity service). Upon receipt of the prescribed treatment services, the patient leaves the health service system. After passage of a normal recovery period, the output of the system can be determined: some individuals reapond to treatment and recover, while others, for a variety of reasons (discussed in Chapter Three) do not recover. Exit from the system for each firm type is shown in each figure; each type of exit is specified a8 some proportion of the total number entering the system. In each of the three figures, transportation to another health facility (necessitated by a diagnostician's decision not to treat an individual at the given facility) is shown as a supporting service.8 The availability of this particular service, however, is subject to Substantial variability throughout the country. All government hospitals have one or more ambulances which.are used to transport Patients from.one facility to another; in addition, each district has One or more ambulances, the number depending primarily on the dis- tricts"budget constraints. Where a district has more than one ambul- ance, some are usually stationed at specific rural units. Where government ambulance transport is not available or cannot be relied upon, use is made of private transport (busses or taxis) which operate throughout the country . 100 Linear Programmin39Model Assumptions Before analyzing Uganda's health service system in the context of a linear programming model, it is important to discuss two central assumptions - linearity and certainty - which underlie this method- ology. The assumption of linearity is particularly important, as it implies a production process characterized by constant returns to scale. It is assumed here that constant returns to scale are realized by each major firm type over the relevant range of production.9 The assumption of linearity also requires that constraint relationships be expressed in a linear manner (or at least that linear approximation procedures be used) and that there be some approximation of reality. The second assumption, certainty, requires that the values of parameters such as prices are known and do not vary. This assumption is important for determining a particular solution to a problem. The assumption can be relaxed, however, by using parametric programming or an other nonlinear programming technique to analyze consequences of changes in important policy parameters. Although no real world system completely conforms to the two assumptions of the linear model, it provides a convenient framework for understanding some of the im- portant production relationships of health services and highlights some of the policy issues which must be confronted in the years ahead. Constraints Related to the Production Process Over any time period, a certain number of persons demand services from a given health service firm i, i-(1,...,<9. As noted in Chapter Three, these persons, P, can be subdivided into subsets according to 101 age, sex and disease, denoted by the subscript j. (Each jth_subset of P is specified as pj.) Each firm can then be viewed as engaging in the production of j curative activities, j-(l,...,n), where the intensity of each activity is measured in terms of thousands of suc- cessfully treated persons who have received health services at that firm.10 In order to operate any set of health services for the production of 3 activities, each firm requires a set of inputs, r. In the short run, (e.g., a year) some inputs, such as trained medical manpower and buildings to house the firm's Operations, are limited and cannot be increased by increased expenditure (i.e., the short run supply function is inelastic). Other inputs, such as drugs and equipment, can be increased by increased expenditure to the point where the relevant budget constraint (central government, local government, or mission) limits the procurement of these inputs. Thus, for any given firm i in the health service system, there exists some maximum quantity of 98Ch input vri! r=(1,...,m) available for the production of a set of health services k, k-(l,...,|) leading to the successful treatment of persons demanding medical care. One subset of elements in the input vector is comprised of persons initially demanding service...,i.e., S where s-Zij. Thus, for 6 (vector of inputs with m elements), the first set of elements Vl-(l,...,(prn)), comprise the service providing inputs: i.e., medical personnel, drugs, equipment, etc. The second set of elements, Vz-g-((m-(n-l)),...,m), represents the number of each 3 type person demanding curative health care. For a given firm 1, some quantity of thelflgl input is required to 102 provide one unit of curative service k. This quantity is denoted mrk° In order to operate each.j activity at unit intensity (i.e., 1,000 persons with.the set of characteristics successfully treated) in firm 1, some quantity (defined in units of service) of service k is required. This quantity is specified ckj° Each ckj reflects present medical knowledge about the diagnostic and treatment services required for the successful treatment of each j activity. In the case of the multi-product health service firm, where each output may require a different definition of the term "one unit of service k" and where the specified service (as in the case of drugs) is an amalgamation of related services, it may be required that the concept of one unit of service for each k service be standardized in terms of one of the j activities. For example, in the case of diag- nostic service, one unit of service can be defined as the amount of a diagnostician's time required to diagnose a case of clinical malaria; in the case of a heterogeneous service such as drugs, however, there are at least two possible methods for dealing with the situation: (a) divide the heterogeneous service into homogeneous sub-sets, such as by detailed drug type and strength, or (b) homogenize the service by the use of prices and weights, and standardize the unit of service according to one of the 1 output activities. The standardized unit of service is specified operationally in the next chapter for each service. Finally, for any firm i, arj is the amount of input r required to operate activity j at unit intensity. The magnitude of each arj: is thus determined as follows: 103 (1) at: = 31. ma. cm In the production of the 1 activities, the health service firm can use any one or all of the r inputs to the point where it has exhausted the ‘maximum available quantity of that input. This relationship is formally expressed in the following way: (2) £3 arj Pj < Vr or, using matrix notation, A§<6 where 6 a vector of elements vr, P - vector of elements pj, and A = matrix of elements arj° For the second set of elements of vector $=§=Cm—(n-1),...,n0 which can also be more conveniently specified §=(1,...,n), the available quantity of each 1 type person must be exhausted in the production ‘process. This requirement formalizes the situation in Uganda in which all persons demanding medical care are provided with some ser- ‘Vice.11 Equation (2) can be rewritten to account for such a require- ment as follows : (2a) 23 arj Pj < Vr and where r=(l,...,(mrn)) and m>n, (2b) 2:j arj Pj ' Vr where r=((mr(n-1)),...,m). Equation (2b) can be simplified by splitting the input vector 6 :into its 2 component parts. It was shown above that the second set IOf elements in vector 6 can be respecified as g which is comprised of ‘D.e1ements. Therefore, it is possible to rewrite (2b) as follows: (2C) 23 arj pj - 81 where j-(1,...,n) and where the elements BJ‘Vr for every r'((‘“"(n'1)) s 0 ' ' am) 0 104 Significance of the Technical Coefficients arj of the Second Set of Elements of the 6 Vector The arj's in equation (2b) and (2c) are significant in that they trepresent a technical relationship between successfully treated persons .and the total number of persons initially demanding service. Using the formulation of (2c), each arj represents the inverse of the pr0por- ‘tion that each pj is of the corresponding SJ, such that 5;; .‘21_, r s ‘fihere j'(ls-°°sn) and where 53- = the rate of successful treatment.12 Those persons who exit from the system prior to (or without) being successfully treated, comprise a pr0portion of pj equal to arj -1. ‘This group of unsuccessfully treated persons for any activity j, l-arj, can be disaggregated into its component parts: (a) T3 - the number of persons transferred to another health services firm for treatment, (b) “j - the number of persons unsuccessfully treated,13 and (c) Dj — the number of persons who die in the process of treatment. Thus, (3) arj'l'a-l=3fiuifl)i. p3 Since the p:) are, in practice, not readily observable, it is useful to specify the arj's in terms of other variables. With a little algebraic manipulation, it can be seen that: (4) an - “i Tajijj)‘ The 81's, Tj's and Dj's are readily observable from data available at each health facility, and an estimate of U3 can be derived from patient follow up studies.14 The production constraints for the health service system as a Whole, i.e., all 1 firms or production units, can be written as in 105 equation (2)-15 The sunnnation of firms will take the following form:16 1 - firms (1,...,o), i - (1,...,h) where i so, then the production relationship shown in quadrant (1) would shift up to the locus denoted e1. Turning to the quadrant (3), a set of alternative relationships between the service providing inputs v1. and the initial service de- manders SJ is shown. For illustrative purposes, three relationships are included where LEE!) > (1131 > (:52. The ratio of the two . 1 SJ 81 inputs v1. and s1 is constant along the entire ray. If, during any given time period t, where the level of vr is fixed at (vr)o, the 114 (1) (31 > 81) ('90 (vr)2 (vr)1 Vr - service producing inputs 2 45° line / “° ‘1 (2) “2 P1 / (given e o) 1”o O (sj)° (sJ)l (33)2 8:] - initial service demanders (3) A (83).) (81)]. (3:92 3 .1 Figure 4.6:. Relationship Between Output and Quality in the Production of Health Services 115 number of initial demanders of health services increases from value (81)o t0 (33)1, (shown in quadrant (3)), the ratio between the two inputs falls and one moves to a lower ray, say from do to «1. Given the production function relationship in quadrant (1) and the value of V: - (vr)o, output can be determined such that P=Po. Transfer- ring the value Po and (sj)o to quadrant (2), it is possible to specify the relationship between initial demanders and total output. Given the production function so, a relationship between output P and initial demanders 31 can be developed for each ray in quadrant (3). This relationship is developed in quadrant (2) with the «1 reflecting a constant ratio between vr and sj. The 45° line in quadrant (2) imposes an upper constraint on the feasible relationships due to the fact that output P cannot be greater than initial demanders sj. This constraint in turn imposes a constraint on the feasible rays in quadrant (3). The constraint ray in quadrant (3) is also dependent on the given production function. If, for example, the production function shifts up to e1 as a result Of improved medical technology, the maximum feasible ratio Vr/Sj must fall; for any other ratio of V1. and sj not subject to the constraint, the rays in quadrant (2) consistent with a given ratio of the two inputs in quadrant (3) would rise. Let us turn to the implications of this analytical framework in determining the firm's level of output and quality. Assume the budget- ary process has established a budget constraint for the firm at level (vr)o° If the number Of initial demanders at the start of period t is (83)1, the ratio of vr and 33 - ocl. Output level PO is determined by the production function and is transferred to quadrant (2). As noted 116 in quadrant (2), output 1?0 is less than the total number Of initial demanders (Bj)1. At (sj)1,IP is equal to the distance AF, which is O some proportion of BF (BF equals the number of initial demanders): the rate of successful treatment (the measure of quality), is thus equal to the ratio AF/BF. If during the course Of period t, there is an increase in the number of persons attending the facility, for example, to (sj)2 (and assuming no increase occurs in the availability of service providing resources (vr)o), the ratio of vr to sj must fall to rays: in quadrant (3), which indicates a spreading of the service- providing resources over a larger number of initial demanders. When this increase in Sj is reflected in quadrant (2), the relationships between output and initial demanders has shifted from.point A to point C. At point C, total output has remained constant, CG - AF, whereas the number of initial demanders has increased DC > BF, such that the rate of successful treatment a, has fallen; 81 > a2 where, 81: . AF .95; BE" and 32 DC Assuming no upward shift in the production function e improve- 09 ment in quality can occur only through an increase in the ratio of productive service inputs V1. to initial demanders 33. This can occur by (a) increasing vr, holding sj constant (for example, at level (sj)1), (b) reducing the level of initial demanders, for example, from.(sj)2 t0 (3311’ holding Vr constant at (yr)o; or (c) by some combination of changes in v or 33, such that «2 < a i where «2 a I' 117 1 ('v'l) and a 1 . (VI 83 2 2 33'2, which.is some ray from.the origin with a lepe greater than ‘2. If output maximization as well as quality maintenance is an Objective, then at least two of the possibilities discussed above are not feasible, assuming constant technology. Output maximization is realized only where vr increases from an initial point, say (vr)o, given the production function e0. Thus Cb) and any combination in (c) above, where Vr declines, are not feasible, if output is to be maximized. In the case where the number Of initial demanders 33 remain con— stant throughout a given period t (for example, at level (sj)2), how- ever, it can be shown that the two objectives of quality and output maximization are complementary. For example, assuming an initial bUdSGt constraint (Vr)o and production function e0, the initial level of output is'Po, such that point C in quadrant (2) represents the initial position with a rate of successful treatment a2 ..99, If DG additional resources enter the firm or health service system, such that V1. rises to (vr)1, output will rise to Po. Output P1 can also be attained with an increase in inputs vr from (vr)o to (Vr)2 < (Vr)1, with a concommitant increase in medical technology to e1. Regardless of the method of output increase from P0 to P1 the quality of the medical care also increases from.a2 -'%g to a3 -.§§, where 33 > cc, DG . such that a3 > a2, A Summary of the Quality Specification in the Objective Function The prOblem facing the curative health service system during any given time period t 18 to Max 2"5' %, subject to the constraints, A 118 e (a) AP O, pj > o, where A - technological matrix and, s A = a diagonal submatrix of the A matrix and is comprised of elements arj’ which are the inverse of the rate of successful treatment . For humanitarian and equity reasons, a minimum value restriction must be imposed on rates of successful treatment such that '§%3 > o, where r-((m-(n-l)),...,m). In the graphical analysis above, the quality variable - rate Of successful treatment - was discussed as an endogenous function of the ratio of the two parts of the input vector 6: (1) that part comprised of service - providing inputs vr, r < (m-n), and (2) that part come prised of initial demanders sj, r=((m—(n-1)),...,m). It was shown that quality is also dependent upon the level of medical technology e, under- lying the production function. Taking the later consideration into account, the functional relationship can be formally stated, such that within any given facility 1, the rate of successful treatment,'§%33 is a function 0f (g!) and e, such that l I 7 —— - v ( ) 311 a ( 8;), e)e In order to Obtain an aggregate measure of the quality of activity j for all 1 firms, the aij's can be summed and each weighted by the ith firm's proportion of total output of activity j, pij' Thus, (8) a - E 1’1 3 1 (1:31) ‘11- TO conclude, it is possible to determine the value of the output vector P for Uganda's health service system, given (a) the existing structure of health service firms, (b) a constant medical technology, 119 and (c) knowledge of the elements of the technological matrix A. Cost of Curative Health Services In any firm i, the total cost of production of the output vector P can be determined from the production relationship specified in the previous section. Thus, the total cost of the ith firm's output vector (h. may be specified as follows: (9) c - mg“) 1 "r vri r-l where Wr equals the price or wage of the rth input. (For r=((m-(n-1)), ...,m), wr - 0.) For the hospital subsector, (93) C1 .- I; (mEn) w v i=1 ral r r1. Similarly for the other two sub-sectors, ( C 3 (mrh) 9b = E w ) i 1-h+1 r-1 r vri for rural units with inpatient services and for rural units with no inpatient services. To estimate the total cost for any given activity j in firm i, the total cost of firm C1 must be allocated among the several activit- ies. The process of allocating costs, however, is facilitated when the elements (arj) Of the technological matrix A are known. Assuming for the moment that the elements are known, the total cost of operating the jth activity in firm i can be specified as follows:35 (In-n) C I Z 11 r-l arji wr Vri' (10) 120 Total cost of operating the jth activity in the system as a whole may be specified: (11) C (In-n) ‘ZC .2 X j 1 13 i r=l arji wr vri- It is important to note that the specification of the total cost 0f operating the 1th activity in equation (11) includes the costs of the persons who, for reasons mentioned above, are not treated success- fully. This is because every arj’ where r < (m—n), reflects the usage of inputs in operating activity j at unit intensity. Where the rate of successful treatment 3%; - é}, where r=((m—(n-1)),...,m), is less than 1. (#5 < 1), some service-providing resource Vt will be used in providing one or more curative services to those unsuccessfully treated. Dynamic Factors Affecting the Model An Objective of this research is to develop an appropriate method- ology for estimating future trends in total cost of the curative health service system. The methodology must incorporate the impact of factors such as (a) population growth, (b) changes in income, (c) changes in input prices, (d) changes in quality standards, (e) shifts in the health services facilities mix, (which includes both shifts in the mix of government units between hospitals and non-hospital units, as well as the mix of government units and mission units), (f) changes in the disease mix of initial demanders, (3) changes in the total number of health facilities and (h) increased usage of existing facilities. The impact of decisions byxthe Ministry of Health.have long-term affects, and once decisions are made relative to facility expansion, 121 health manpower training programs, relationships with voluntary units, and standards of care, the cost implications manifest themselves in subsequent periods. Only under conditions of extreme financial crisis (resulting, for example, from a major crop failure), would the govern- ment seriously cOnsider the possibility of decreasing its commitment to health service provision. In the static model develOped above, the primary endogeneous variables are (a) the output vector P and (b) total cost C. Given the output vector P, the elements of the technological matrix arj’ and the input price vector wr, the cost of the curative services can be determined. However, once dynamic considerations are introduced, it cannot be assumed that exogeneous variables, such as (a) service providing inputs (yr, r < (men)) and (b) initial demanders 81, remain constant. In addition, if the possibility exists for some substitut- ability between one or more of the v1. inputs as a result of relative input price changes or changes in medical technology, the elements of the technological matrix arj will change as a result of the impact of the change on the technological elements ‘tk and ckj' During each time period t, the problem can be reformulated as follows for the health service system as a.whole: e (12) Max 2 - P(t) subject to the congtraints A(t) P(t) 0; Pj(t) > o; and arj(t)=§r (t), where £=_l___q>o, arj(t) when r=((m-(n-l)),...,m) and l is a target rate of successful treatment. arJ(t) Given the solution of the problem for time t, the matrix of arj's J 122 and the vector 0f Vr can be used to determine the total cost at time t of each.sctivity j as well as the total cost for the entire cura- tive health.service system. Thus, the total cost of activity j, (mon) (13) 03m -r§1 arj(t)wr(t)vr(t) and total cost _ (In-n) (14) C(t) ZjCJCI) = 21 r51 arj(t) wrCt) Vr(t). If the planning horizon extends beyond the single period t, it is important to determine whether the desired objectives are to be attained by the end of the planning horizon or whether the time path of attainment is an important component of the objectives. In some circumstances the two perspectives may be synonymous with.respect 36 to the end result, but such.is not necessarily the case. This problem.is particularly critical in the development of a consistent health policy which takes into consideration the integration of pre- ventive and curative services. Given that a primary economic rationale for providing curative health services is to maintain the existing stock of human capital (see Chapter Three), it can be argued that a strategy which seeks to maximize output during each sub-period will maximize the stock of human capital over the entire period to a greater extent than will a strategy~whicthoes not necessarily require maximization in each.sub-period of a planning period. If one assumes that there are no material shifts in the age and sex specific incidence of disease episodes over the planning period, regardless of the shifts 37 which may occur in the distribution of diseases , a strategy which seeks to maximize output in each.sub-period will maintain the human 123 capital stock at a higher level than will alternative strategies. Thus, the objective function may~be specified such that (15> Max 5 Cr) ~31 3 (t). t where P (T) is the output vector over the entire planning period T, t-(l,...,T). Factors Affecting the Vector of Initial Demanders During a given time period t, it is assumed that the total number of initial demanders s-stj is a function of (a) the size of popula- tion, (b) the availability of health facilities, (c) income, (d) perceived quality of service, (e) price of service, and (f) the incidence of illness episodes. For purposes of projecting changes in the number of initial demanders over time, the functional relation- ship will be specified in terms of the rates of change in these vari- ables, however, for expository purposes, the variables are expressed in terms of absolute values. The relationship may be specified as follows: (16) S=S(O,Y,a,b,I,d) where S - the total number of initial demanders 0 - population size Y - GDP per capita (used as a proxy for per capita income) a - the rate of successful treatment (quality, given the disease mix of initial demanders) h - the price of.a set of curative health.services, I - the incidenCe of illness episodes, d - the average distance to health facilities as a measure of availability. 124 Over the usual planning period of five to ten years, it is reasonable to assume that the variables mentioned above with.the exception of (a) the rate of successful treatment (a), (b) the average distance to health.facilities (d), and, to a certain extent, (c) the price of a set of services (b), are exogeneous variables and thus are not af- fected by decisions related to the expansion or operation of the system. The variables b (the price of health services) and d (average distance) are determined by changes in the mix of facilities and d is additionally affected by policies related to the expansion of the en- tire service system. In the case of the price variable b, the shift in the mix of health services away from mission units (assuming no change in the price policy of mission units with their present financial requirements) will lead to a decline in the average price charged initial demanders. The availability of health facilities, as measured by the average distance to the nearest facility, changes as the number of facilities increases - assuming that the new facilities are not built in close proximity to existing ones. If the mix of governmental units shifts toward a relatively larger share of rural units (rather than hospitals), not only will there be an increase in availability, but also more rapid expansion (in terms of total numbers of units) may take place since the largest rural facility requires approximate1y~52 of the 38 initial capital cost of the one one-hundred bed hospital. A reasonable proxy variable for availability and average distance may be the average number of attendances per person which.is negatively correlated'with distance from the health facility.” relat 125 The relationships may be formalized as follows: (17) b - b $1) and . 23. g (18) d d (n. (G . can where H - total number of health facilities, (£5 = the ratio of mission units to all health facilities, GR *5? - the ratio of rural government units to all govern- ment facilities. In addition, the above discussion implies the following §_priori relationships: ng > o 3% "L1 d 3H __ 3 < o; < 0; an > o. (a) a (915-) 3 (9%) Factors Affecting the Rate of Successful Treatment Discussion earlier in this chapter and in Chapter Three provide insight into the nature of this quality variable and provide some rationale as to why it may vary over time. As noted above, at any given point in time t, the rate of successful treatment is defined as some function of (a) the ratio of service providing inputs V1. to service demanding inputs sj, and (b) the level of medical technology. By taking into account the non-homogeneity of output, additional vari- ables must be included in the analysis to explain the differences in the rate of successful treatment. It is thus hypothesized that this quality variable can be specified more precisely as a function of (a) differences in the ratio of inputs providing diagnostic ser- vices, (b) medical technology, (c) disease mix of initial demanders, and (d) differences in the ratio of service-providing inputs to initial 126 demanders. Symbolically, the relationship can be expressed in the following way: v * (19) a = a ((Tyi'r)’ 3’ S9 (:2?)), where a is defined as above, v QE'I). the ratio of inputs providing diagnostic services, where ra'rl, r e = medical technology, * S - the vector of initial demanders sj, and v ((_E5_,the ratio of service providing inputs to‘initial demanders. s It must be noted that the rate of successful treatment, is a function of both the service-providing and initial demander input constraint sub- sets. Thus, it is possible to analyze changes in the elements of the constraint vector in order to determine the impact of such changes (H: the rates of successful treatment. Factors Affecting the Service Providing Subset of the Input Vector In general it can be said that the service providing input vector is a function of each rth input, particularly (a) financial resources from (i) governmental allocations, (ii) individual demanders through fees paid for services consumed, and (iii) gifts from third parties, and (b) the supply of direct service-providing inputs such as medical man- power, drugs, equipment, and facilities. The first factor (which, for convenience, will be called the budget input), is controlled in large part by the central and local government's ability to generate revenue; this in turn is dependent on the over-all 127 economic performance of the country. Governmental medical policy on the issue of fees for services (i.e., no charge will be made for normal services rendered in governmental facilities) has placed a constraint on the level and rate of increase in fees charged by non-governmental health facilities. Gifts, made primarily by external church groups, can be considered exogenous in the present model, although political stability and general governmental attitude toward mission activities may be important factors in the external donor's decision to give. The second factor, the supply of individual inputs, is largely determined by decisions related to (a) manpower deve10pment, (b) salary levels, (c) the number of facilities, and (d) the availability of other inputs, particularly drugs and medical equipment. In the area of man- power development, there are two questions of importance: (a) What types of medical training programs are being and should be developed, and (b) how large should such programs be? In determining the salary levels for each type of health manpower, decision-makers must consider not only previous levels of pay, but also the potential mobility of each type of manpower, given the amount of training received and the skills which have been developed through experience. Levels of pay received in jobs with similar training and skill requirements, both in and out of government and the health field, must also be considered. In deter— mining the number of facilities in which curative health services will be provided, four factors are particularly important: (a) development plan strategy in the health field; (b) availability of financing (from internal as well as external sources, for both government and mission facility construction); (c) government attitude toward mission expansion; 128 and (d) relative emphasis on rural unit v. hospital construction. The availability of other inputs (non-labor, non-capital, e.g. drugs and other operating expenses) is largely determined by (a) total economic activity as measured by G.D.P., (b) the capacity to import, (c) the internal production capacity, and (d) the prices of such inputs. These relationships can be specified as follows: for the budgetary input B: (20) VB - vB (GB. MB. F. Q). where VB the budget input for curative heatlh services, GB - the government budget, MB the mission budget for operating curative health services, F = the quantity of fees collected for services rendered, and Q = the quantity of gifts received; for the manpower inputs L: (21) VL= vL(wL. JL), where v].3 the quantity of manpower inputs L, where L=(l,...,o) w];' the price (wage) of the manpower inputs, JL = the output of health manpower training programs operated by the government and missions; for capital inputs K:40 (22) vK = vK (EG, EM, GB, (ggll). (353), am), where GB and (g5) are defined above, and wk - the amount of capital inputs K, where K=(l,...,e), EG - the amount of external assistance to government, EM - the amount of external assistance to missions, DBH (fifi') - the proportion of the development plan's expenditures for health services, and 129 GAM a a dummy variable indicating whether the government has a positive attitude to missions in the delivery of health services; and for the non-labor, non-capital inputs N:41 (23) VN - vN (Y. CI. W , CAPD) where VN . the quantity of non-labor, non-capital inputs N, where N= (1: ° ' ° ,8) 9 Y a Total Monetary G.D.P., CI = the capacity to import, “N - the prices of the N inputs, and CAPD - the existence of domestic production capacity. To conclude, the service providing input vector 6r 13 a function of the four primary input components specified above. It is possible to write this relationship in the following way: 6r - v (VB, vL, v , v ), r K N where the variables are defined above. In Chapter Five, the postulated relationships between service-providing inputs and the factors affecting the supply of these inputs will be sub- jected to empirical verification through standard regression techniques. The significance of the estimated relationships will also be explored. Summary A theoretical framework, using linear programming concepts, was developed in this chapter for analysis of the relationship between the inputs and outputs of the curative health service system. The import- ance of the quality index (rate of successful treatment) and the multi— product nature of health service facilities were stressed and were 130 incorporated into the analytical framework. A further framework for analyzing the important factors affecting (a) service-providing inputs, (b) service demanders, and (c) rate of successful treatment over time was developed. In Chapter Five, the framework developed here is used in empirical analysis. 2. 5. 6. 131 Footnotes Pfouts, Ralph W., "The Theory of Cost and Production in the Multi- Product Firm", Econometrics, Vol. 29, No. 4, October 1961, pp. 650-658. Neil, Roman L. Jr., "Allocating Joint Costs", American Economic Review, Vol. 58, No. 5, Part 1, December 1968, pp. 1342-1345. See page 654, Pfouts op. cit. for a statement about the relative usefulness of each approach, particularly for decision making purposes. On this point also see Dorfman R., Samuelson P., Solow R. W., Linear Programming:and Economic Analysis, (New York: McGraw-Hill Book Co., Inc., 1954), and Dana 8., Industrial Production Models, (New York: Springer-Verlag Inc., 1966). Important examples of the use of such methodology-all of which are describing basically the same approach include: Feldstein, Martin 8., Economic Analysis for Health Service Efficiengy,(Amsterdam: North Holland Publishing Co., 1967) Chapter Six; Feldstein, Martin 8., "Health Sector Planning in Developing Countries," Economica, Vol. 37, No. 146, May 1970; and Taylor, Carl, E£.§l'a Functional Analysis of Health Needs and Services, A Report Compiled by Johns Hopkins Univer- sity, School of Hygiene and Public Health, Department of International Health, December 1970. See recent issues of Medical Care and Inquiry for related articles focusing on resource allocation in the health field in the United States. Day, Richard H., Recursive Programming and Production Response, (Amsterdam: North Holland Publishing Co., 1963) pp. 109. In some larger rural health centers (e.g., in Busoga District) a small laboratory service has been developed over the last five years. Thus, in addition to the diagnostic services described, some persons, whose condition has not been adequately diagnosed from historical and symptomatic information in conjunction with a physical examina- tion, may be given laboratory tests to improve diagnostic precision. Although this supporting diagnostic service is not yet widespread in rural health units, it could easily become part of a program whose policy objective is to improve the quality of medical care in rural areas by improving the rate of successful treatment. The book by J. R. Billinghurst, Trowell's Diagnosis and Treatment of Diseases in the Tropics, (London: Bailliere, Tindall and Cassell, 1968) is an excellent example of the extent to which standardized medical treatment procedures have been catalogued for a country such as Uganda. The National Formulary,gl966 (Entebbe: Government Printer, 1967) written for the Uganda government and Ministry of Health, provides standard dosages for a number of different drugs and diagnoses. The decision may also be made to refer a patient already receiving treatment services to another facility if it becomes apparent that more intensive care is required. This flow is not included on the diagrams primarily because it would tend to overly complicate them. 10. 11. 12. 13. 14. 15. 16. 132 Dano, 8., Industrial Production Models (New”York: Springer - Verlag, New‘York., Inc. l966)and Hixson, Jesse, "A.MOdel of Hospital Product- ion, Cost and Capacity Determination", Workshop Paper 7009, Econo- metrics Wbrkshop Papers, Michigan State University, East Lansing, Michigan, April 1971. As the analysis is expanded, the activity intensity is defined in terms of the entire set of firms comprising the health service system. Although every person demanding medical care receives some service, and in the process consumes a certain amount of service providing resources, the statement does not intend to imply that every person receives enough service for every one to be successfully treated as determined in a subsequent analysis, for everyone to obtain the services deemed medically appropriate at the time of initial demand. In terms of Equation (2b) specification, each ar shows the pro- portion that each pj is of the corresponding vr, such that arj "‘31, where r=((m-(n-l)),...,m). r This set of unsuccessfully treated persons theoretically can include those persons who sought medical help but leave prior to the receipt of service and yet recover. Such individuals reflect the fact that there is an opportunity cost of receiving "free" medical care. If one were to observe that this set were increasing over time, a plaus- ible inference could be made that the opportunity cost was rising over time. See Appendix I? for a description of the methodology employed in a patient follow-up study carried out in conjunction with this study. See also Taylor, Carl, gt al., Functional Analysis of Health Needs and Services, a report compiled by Johns Hopkins University School of Hygiene and Public Health, Department of International Health, December 1970, where the methodology of a patient follow-up study conducted in rural India is described. In that case, the vectors and matrices denoted in equation (2) will thus be a summation of the elements of the 1 firms. Thus, v for all firms will be a column vector of r elements, with each element vr - zivri' It may be required that the elements of the A matrix for 811 firms, 8:3, be weighted in some fashion to reflect the relative importance of larger firms in the health service system. Where this is the case the 2:method of calculation of each arj’ may be denoted 3‘ fOIIOVB at 1. where 91 is the weight given to the ith-firm.'s technical production relationship. Since a primary interest of this analysis centers on comparisons between several of the subsectors of firms in terms of input ratios, output mixes and total costs, it is important that the 1 firms 17. 18. 19. 20. 21. 22. 24. 25. 26. 133 within the system are grouped according to an apprOpriate taxonomy. The categorization develOped takes into account the administrative structure as it relates to the structure and financing of Uganda's health.service system and takes into account also the existence of inpatient treatment services, which.significantly raise the number of diagnostic and treatment service options and, pari passu, the cost. This kind of objective function has been used in analyzing educa- tional systems in less develOped countries. See Samuel Bowles, Plannigg;Educational_Systems for Economic Growth, Harvard Economic Studies, Vol. 113, (Cambridge, Mass.: Harvard University Press, 1969). See Martin Feldstein, "Health Sector Planning in Developing Countr- ies", Economica, Vol. 37, No. 146, May 1970. In the P.A.H.O. planning model, the objective function contains a single item, a measure of reduced mortality. This is much too simplistic in terms of the possible health objectives that people and societies may desire at any one time. See P.A.H.O., Health Planning: Problems of Concept and Methgd, Washington, D. C.: Pan American Health Organization, Pan American Sanitary Bureau, Regional Office of the World Health Organization, April, 1965, (Scientific Publications No. 111). Data of this type have been collected to some extent in India. See Taylor, Carl, Functional Analysis of Health Needs and Services, A Report Compiled by Johns Hopkins University, School of Hygiene and Public Health, Department of International Health, December, 1970. See Dunlap, David W., "Research on the Economics of Health Services in East Africa", Rural Africans, No. 13, Winter 1970, pp. 77-84. Republic of Uganda, Work for Progress: Uganda's Second Development Plan, 1966-1971, Government Printer, Entebbe, 1966, pp. 13; King, ‘Maurice, ed., Medical Care in DeveloPing Countries, Oxford University Press, London, 1966, pp. 2.6 and 2.7; personal interview with Dr. Semambo, Medical Superintendent of Mulagc>Hospital 1970; and Republic of Uganda, Uganda's Plan III: Third Five-Year Development Plan, 1971/72 - 1975/76, Government Printer, Entebbe, 1972, pp. 301 and 308. Republic of Uganda, Work for Progress, p. 18. Republic of Uganda, work for Prpgress, pp. 151 and Plan III, pp. 308. Republic of Uganda, Work for Progress, pp. 18, 51. Nathan Epenu, Uganda Argus,_25 February 1971. 27. 28. 29. 31. 32. 33. 35. 134 Sharpston, Michael - Ghana Harvard Advisory Service, no date, approx. 1967/68, mimeoed document. See work.f0r Progress, Statement 13.6, p. 15. The Ugandan govern- ment takes thefposition that bymaintaining a policy of free medical service, the country is medically and socially advanced. This long-standing positon has been questioned on occasion in the past, but has not been changed for political and public health reasons. See Frazer Committee Report, Medical and Health.Services in Uganda, Government Printer, Entebbe, 1956, and International Bank for Re- construction and DevElOpment, Economic Development of_yganda (Baltimore: Johns Hopkins Press, 1962). Feldstein, Martin 8., Economic Analysis for Health Service Efficienpy, (Ansterdam: North Holland Publishing Co. , 1967 ) Chapter Six. This procedure is followed for all government hospitals for most variable cost items such.as local employees, food, transport, and electricity. The problem of intergovernmental cooperation may not be as important in the health field today as it was before the change in policy'announced in Uganda's Third Development Plan 1971/72 - 1975/76 where the Minis- try of Health was placed in charge of all rural health facilities. See Republic of Uganda, (1972), Ugsnda's Plan III, op. cit., pp. 203, 204. Some public press reports indicate that a relatively effective regime has been developed treating cholera in Bangaladesh refugees. In the initial set of constraint conditions discussed above, the elements in vector A comprise a diagonal submatrix of the A.matrix. Included in the concept of medical technology as used in this con- text are organizational and administrative changes which affect the productivity of existing resources, as well as qualitative changes in one or more of the medical inputs. The methodological problems of estimating a will be examined in a subsequent chapter of this dissertation. Th2 problems involved are related to issues in the area of cost accounting. The conventions of time and space are used; sample data are also used to indicate the magnitude of certain stable relationships such as standard drug treatments (given the diagnosis) or the proportion of total costs which are administrative overhead costs. In Taylor, g£.'§l., Functional Analysis of Health Needs and Services, Report Compiled by Johns Hopkins University School of Hygiene and Public Health, Department of International Health, December, 1970, 493 pages, xeroxed, four health centers in India were studied and some tentative efforts were made to allocate the cost of operating a health center to the several activities in which it is engaged. See Chapter Five, Part II, especially pp. 351—412 of the study. 36. 37. 38. 39. 40. 41. 135 See Bowles, Samuel, Planning Education Systems for Economic Growth, (Cambridge Massachusetts: Harvard University Press, 1969). The assumption of no significant shifts in the incidence of disease episodes over the most normal planning periods of 5-10 years has considerable epidemiological validity except in situations where a major infectious disease - such as malaria - is erradicated or its incidence substantially reduced by the introduction of a series of preventive health measures, including the widespread use of prophylactic medicines. In the case of Uganda, it does not appear likely that any major breakthrough will occur to materially affect the incidence of illness episodes during the next five years, although steady progress is likely to be made to reduce the number of pe0p1e who have not been immunized. Given that a sizeable proportion of Uganda's population is impoverished, however, the overall incidence of illness episodes, particularly among the population over five years of age, is not likely to change materially in the near future. Figures taken from author's working papers. See Maurice King, ed., Medical Care in Developing Countries: A Primer on the Medicine of Poverpy, A Symposium from Makerere,(London, Oxford University Press, 1966);Haskell, Mark, "Medical Service in Masaka District", unpublished paper, New York University, 1971, 21 pages; Galea, J., "Assessment of Uganda's Basic Health Service System", World Health Organization, Malaria Erradication Program, Jinja, 1967, Mimeoed; and Taylor, Carl g5..sl., Functional Analysis of Health Needs_spd Servipss, A Report compiled by Johns Hopkins University School of Hygiene and Public Health, Department of International Health, December, 1970. Capital inputs include facilities, transportation equipment, and other medical equipment, such as x-ray machines, surgical room equipment, and a complement of beds and laboratory equipment. SuCh inputs include drugs. sundries, small equipment replacement, transport expenses, electricity and telephone. CHAPTER FIVE In this chapter, Uganda's health service system is empirically analyzed. The chapter is divided into four sections. The first section contains (1) an empirical specification of the analytical model developed in Chapter Four, (2) a discussion of the sources and methods used in obtaining data and (3) a description of the procedures used in estimating the value of the model's parameters. The second section of the chapter presents the linear program— ming results for the three major health service delivery systems in Uganda —— government hospitals, mission hospitals, and government rural health units with inpatient care. A comparative analysis of the findings for each sector follows, focussing on the similarities and differences in the outputs, resources, and costs of the three systems. The third section of the chapter discusses the empirical analysis of the factors affecting the output, resources and cost of health services over time. In section four, projections are made to 1980/81 -- the end of the fourth five-year planning period -- utilizing the linear programming model. The results of this analysis are presented in order to examine some of the long range implications of certain health policies recommended by the Third Plan. Empirical Specification of the Model The Model The linear programming model used in the empirical analysis is schematically presented in Figure 5.1 and Table 5.1. The system has three basic components. First, there are the input constraints, vr, depicted down the left-hand side of the figure. For each sector of the system (government hospitals, mission 136 137 as 325:5 ~33. n V . e. . . . v . . .. b. . . . .. H _ , ._ . ., nusasH .. AmuOH uooausasu soofiumauso usowusmsu usaaumauso , usuuuoaom unusuaauso mooa>uom usowuuasH no“; «use: Hausa masuaawom coswmaz mawufiamom noossuo>oo musauso - . _ . _ .souuhm oua>wom guano: us» we muouuom cough <._ m.wvoaw= mo Home: wswaauuwoum Luanda ecu mo soauooawauoam Hosumousoo . .H.m ouswwm 138 Table 5.1 Summarization of Output and Input Variable Specification A. Output Variable - (1) B. Service Providing — Disease Characteristics (j) Input Variables (r) (1) Infectious and Parasitic I. Manpower(2) (2) New Growths (1) Medical Officers (2) Medical or Nursing Assistants (3) Allergic, Metabolic and (3) Professional Nurses or Midwives Blood (4) Enrolled Nurses or Midwives (5) Trained Lab Staff (4) Nervous System and Sense (6) Trained X-Ray Staff (7) Other Trained Medical Staff (5) Circulatory (8) Other Trained Non-Medical Staff (9) Non-Trained Medical Staff (6) Respiratory (10) Other Non-Trained, Non-Medical Staff (7) Alimentary (11) Student Staff (8) Genito-Urinary II. Capital (9) Pregnancy and (12) Beds Puerperium III. Intermediate (10) Delivery without Complication (13) Drugs and Medical Supplies (14) Food (11) Skin and Musculo- (15) Vehicle Operation and Maintenance Skeletal (16) Electricity (17) Other Operating and Maintenance (12) New Born Expenses (13) Ill Defined IV. Service Demanding Inputs (14) Injuries Notes: (1) (2) (See the Output Variable Speci— fication for the 14 categories of service demanding inputs, numbered from (18,...,45), to reflect the two treatment processes. See Table F.1, Appendix F. for a precise enumeration as to how the output variable corresponds to the WHO Inter- national.classification of Diseases. In Appendix C, the specific job classifications are pre- sented which are included in the manpower input categories (5-10). 139 hospitals and government rural units with beds), there exists a separate set of input constraints. Second, there are the outputs, pj; these are depicted at the bottom of the figure. Finally, there is a technological ‘matrix, A(i,j), i-(l,...,45), j=(l,...,28), for each of the sectors. Each of these technological matrices can be divided into three sub- ‘matrices for analytical purposes. The upper left-hand sub—matrix of each A(i,j) matrix, i=(l,...,l7), j=(l,...l4), depicts the quantity of service- ‘providing input, 1, required per case of type, j, treated on an outpatient basis. The upper right-hand sub-matrix, i=(1,...,l7), j=(15,...,28), in- dicates the similar parameters for the inpatient treatment process. i=(l8,...,45), j=(1,...,28), is a diagonal matrix. The elements of this matrix are the proportion of each element in the final output vector P com-\ prised by initial demanders with type j characteristics. The inverse of the principal diagonal is the rate of successful treatment.1 The Variables The output classification system employed in the model is disaggregated solely on the basis of disease category for each of the two treatment processes —— inpatient and outpatient -— employed in each sector of the health service sys- tem. As there are 14 categories of disease for each treatment process, the output mix contains 28 categories. This classification system does not include reference to age and sex in the determination of output, but is used to conform to the data available from all three sectors of the health service system.2 The inputs used in the production of curative health services are divided into two categories: (a) the resources used to produce the curative health 140 services demanded and (b) the persons initially demanding service. The second set of elements in the input vector, the initial service demanders, is comprised of 28 elements corresponding to the vector of outputs described above. The first sub-set of inputs, however -— resources used to produce curative services -- is defined as follows. There are three main categories of service-providing inputs: (a) manpower, (b) capital, and (c) intermediate inputs such as drugs and transport (see Table 5.1). A fourth input, the budget, is not included in the linear programming model because it is reflected in the model by the specific amounts of all other inputs. Pre- sumably if one or more of the inputs constitutes a binding constraint on the ability of the system to offer service at some point in time, a budgetary reallocation between inputs may be undertaken to alleviate the constraint. Unfortunately, capital budgets for Uganda during the period 1935 to 1970 could not be disaggregated to allow the separation of capital inputs. It is assumed that beds can be used as a proxy for all capital inputs, particularly those consumed in the process of treating patients on an in— patient basis. Very few capital inputs, with the exception of a building and minor supplies,are used in outpatient treatment. It is assumed, thus, that capital inputs are perfect complements to beds and are consumed in the 'production of health services in fixed proportions with that variable, specified in terms of bed days. * In summary, the input vector V is defined as follows. there are eleven labor elements, one capital element, and five intermediate elements in the 141 vector. In addition, the set of initial demanders (disaggregated into 28 elements) is included in the input vector. The input vector thus has 45 elements: 17 service-providing inputs and 28 initial-demander inputs. The Data Data for the empirical analysis were collected in the following ways. First, a follow-up study of persons who had recently demanded service from a nearby health facility in Ankole District was conducted. The data obtained from this set of surveys were used to develop estimates of the rate of suc- cessful treatment, disaggregated according to case type.4 The second major source of data was the records of a selected number of health facilities in .Ankole, Busoga, East Mengo, Karamoja and West Mengo districts; information on the inputs used in producing health services in rural facilities and 'hospitals were gathered from these records. District-wide data were also gathered on (a) the number of persons (disaggregated according to case type) attending specific health facilities,5 (b) the rate of input usage dis- aggregated according to facility, and (c) drug dissemination and the use of transport facilities. Data gathered from individual facility and district records were supplemented by data gathered previously by Dr. J. Galea, W.H.O. Basic Health Services Project Director from 1965-1967. In addition, the central government Ministry of Health and the Catholic and Protestant Medical Bureaus provided additional information on the initial demanders and resources used in their respective hospitals.6 142 Finally, the above sources of information were supplemented by pub- lished data sources. The most important include (a) the government's Annual Statistical Abstract, (b) the government's Annual Report of the Public Accounts and Budget Estimates, and (c) the Ministry of Health's Annual Re- port and Annual Statistical Report. These sources were used primarily in determining the value of various socio-economic variables affecting the development of Uganda's health service system. Procedures Used to Determine the Value of the Elements of the Input Vector and Technological Matrix. As in any statistical analysis, there exists a hierarchy of methods ‘which can be used to estimate the value of desired parameters and variables. In the case of the linear programming framework, the most desirable method utilizes technologically determined values based on engineering studies of the production process being analyzed. Similarly, the most desirable in- formation for the input vector is exact data on input use. Where such information is unavailable, however, other estimation pro- cedures must be employed. In applying linear programming to health services, 'Martin Feldstein used regression analysis techniques to estimate the values of the elements in the technological matrix.7 Such methods are useful when the problem is formulated such that one is analyzing the "representative firm", as Feldstein was, and when one has the appropriate cost accounting data for all firms in the industry. In Uganda, such cost accounting data did not exist for any sector of the health service system in 1969. With the assistance of the Ministry of 143 Health,8 however, as well as that of mission medical bureaus and several District Medical Officers, budgets and cost-accounting data were obtained for a number of hospitals and rural facilities. This information was used in estimating the value of (a) the service-providing and service-demanding input vector and (b) the technological matrix. The Input Vector The elements of the input vector for each sector of the health service system are shown in Tables 5.2 and 5.3. Additional details concerning the estimation of the elements are presented in Appendix C, but several comments of a general nature are appropriate here. In the case of the service-providing inputs (Table 5.2), the following procedures were used to estimate the value of each element. Estimates of the total supply of each input were developed from the data sources described above. The total supply estimates were adjusted in two ways in order to obtain an estimate of the actual supply of each input available for direct health service provision to specific initial demanders. The first adjustment deducted the proportion of the input used in the production of administrative services. Although administration is a necessary service which assists in the provision of all health services, there is no justifiable criterion which can be used to allocate administrative services to specific outputs. The second adjustment deducted an estimated proportion of the input consumed in the production of preventive health services and delivered through specialized clinics. 144 Table 5.2 1968/69 Service Providing Input Constraints for the Three Sectors of Uganda's Health Service System Service Providing Inputs vr (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) Doctors Med. Assts./ Nursing Assts. Prof. Nurses/ Midwives Enrolled Nurses/ Midwives Trained Lab Staff Trained X-Ray Staff Other Trained Med. Staff Other Trained Non-Medical Staff Non-Trained Medical Staff Non-Trained Non-Medical Staff Students Beds Drugs/Medical Supplies Food Vehicle Operation and Maint. Electricity Other Operation and Maint. Exp. hrs. hrs. hrs. hrs. hrs. hrs. hrs. hrs. hrs. hrs. hrs. Bed Days shs. shs. shs. shs. shs. Government Hospitals 550,800 575,400 529,920 2,395,995 126,000 50,400 396,900 244,650 2,691,255 4,178,160 1,854,000 1,956,035 10,438,134 4,167,795 747,723 2,880,252 1,945,927 Mission Hospitals 126,921 6,300 292,608 404,130 105,000 23,100 13,965 7,875 895,775 845,502 770,000 1,111,060 1,977,712 734,373 393,242 656,442 978,549 Government Rural Health Units W/Beds 117,250 897,750 1,593,900 1,281,880 2,173,626 285,935 1,757,191 122,676 969,392 145 Table 5.3 1968/69 Initial Demanders Input Constraints for the Three Sectors of Uganda's Health Service System Service Demanding Inputs Government Mission Government Hospitals Hospitals Rural Health Units W/Beds Out Patient VI. = (18) I & P 1,856,893 357,826 2,524,020 (19) NC 4,969 1,245 153 (20) AMB 57,181 32,459 50,015 (21) NS 316,427 46,811 358,048 (22) Circ 5,422 3,157 136 (23) Resp. 1,080,744 125,270 1,670,109 (24) Alim 760,258 93,821 836,656 (25) GU 141,520 26,308 150,793 (26) Reg & Puer 30,838 10,720 65,725 (27) Del w/o ------------------ (28) S & MS 763,935 105,501 938,236 (29) NB 11,448 3,320 2,682 (30) Ill Def. 245,761 23,867 493,659 (31) Ins. 533,597 21,382 609,980 Inpatient (32) I & P 37,464 29,147 89,954 (33) NC 4,085 1,855 480 (34) AMB 10,467 8,876 1,570 (35) NS 6,107 2,971 480 (36) Circ 3,845 1,758 369 (37) Resp. 23,642 11,526 40,453 (38) Alim. 22,505 12,039 11,226 (39) GU 8,830 5,498 2,659 (40) Preg. & Puer. 25,017 8,888 21,872 (41) Del. w/o 41,911 16,359 57,995 (42) S &‘MS 10,379 3,980 7,976 (43) NB 1,979 2 ,558 129 (44) Ill. Def. 4,371 3.461 12,925 (45) Ins. 24,103 3,081 15,104 Total 0P 5,808,993 851,687 7,710,212 IP 224,702 111,997 263,192 OP & IP 6,033,695 963,684 7,973,404 146 In addition to the use of government and mission health facility records to estimate the service-demanding input vector, special analyses ‘were conducted in order to (l) disaggregate re-attendances by disease cate— gories and (2) classify outpatient and inpatient data from rural units accord— ing to the major disease categories used in hospitals. The resulting estimates are shown in Table 5.3. The data presented in Tables 5.2 and 5.3 make it clear that while government hospitals command the largest pr0portion of the service-providing inputs, government rural units receive the largest number of initial demanders, both on an inpatient and outpatient basis. Mission hospitals command the greatest amount of professional manpower time on a per-initial-demander basis, and, although not completely in evidence in Table 5.3, service a different mix of initial demanders than do the other two sectors.9 The Technological Matrix The Service-Providing Input Submatrices The values of each element found in the technological matrix linking the service-providing inputs, i-(1,...,l7), to the outputs, j=(l,...,28), ‘were estimated in the following way. (The three submatrices, one for each sector of the health service system, are presented in Appendix C.) First, the proportion of each input used in treating persons on an inpatient or outpatient basis was determined. Foroertain inputs, a further disaggregation 'was made specific to the production of surgical, radiographic and laboratory services. Given the disaggregation of the input vector, the quantity of each input used in the delivery of service to each disease category of initial 147 demander was estimated by using one of the following three criteria. The first allocation criterion, used for inputs into the delivery of outpatient services, was the proportion of total available diagnostic time used by disease category j. This allocation criterion was utilized not only for the diagnostician's time but for other inputs used in the outpatient treatment process as well, since most of the other manpower and non-drug/ medical supply inputs are factors of production complementary to the diagnos- tic factor and hence, are consumed in fixed proportions. The second allo- cation criterion, used for inpatient care, was average length of stay. This criterion reflects the fact that inpatient care is often not disease-specific in the intensity of resource use. The third allocation criterion was service- specific input use. It was used to allocate (a) manpower engaged in the production of laboratory, radiographic, and surgical services, (b) drugs and medical supplies, (c) electricity used by laboratory, radiography, and operating theatres, and (d) transportation costs. In the case of drugs and medical supplies, a detailed analysis was made of the quantities of each drug consumed in the treatment of each disease listed on the hospital outpatient medical form (MF75). The cost of drugs consumed in the treatment of each disease was calculated by using the Ministry of Health's drug price list. It was assumed (with the exception of injury cases) that medical supplies could be allocated in the same proportions as drugs. Although the distribution of diseases treated on an inpatient basis within each major disease classification is likely different than that treated 148 on an outpatient basis, it was assumed that such differences would not materially affect the resulting calculations. As a consequence, the average cost of treating each major disease category on an outpatient basis was adjusted to reflect the average length of inpatient care received.10 The Service—Demanding Input Submatrices As discussed earlier in this chapter, these three submatrices are diagonal, with the values of each element along the principal diagonal repre— senting the inverse of the rate of successful treatment. The following formula was used to estimate these elements for the outpatient sub-set of elements: (1) aopj = 3 - (T :jU + S ) , where j=(1,...,14), and j J j 191 opj = the inverse of the rate of successful treatment, Sj ' the number of initial demanders of the jth disease category, S = the number of initial demanders of the jth disease category ipj receiving inpatient care, Tj = the number of initial demanders of the jth disease category transferred to another health facility for further treatment, and U a the number of initial demanders of the jth disease category unsuccessfully treated. A similar formula was used to estimate the value of the inpatient sub- set of elements: S (2) aipj - 1P3 where j=(15 28) and S - T + U + D ’ "°" 1p: ( j 1 J) a. , S , and T , and U are defined as above but refer to the inpatient 191 193 j J 149 set of initial demanders, and Dj = the number of inpatients of the jth disease category who die in the health facility.11 Using the equations presented above, the inverse of the elements of the principal diagonal of the service-demanding input submatrices, i.e., the disease-specific rate of successful treatment for the outpatient and inpatient treatment processes of each sector, were calculated and are summarized in Tables 5.4 and 5.5. In both the outpatient and inpatient treatment processes, the rate of successful treatment for government rural units is lower than for either type of hospital (an average of 6.5% lower for all disease categories on an outpatient basis and 11.8% lower for inpatients). In the hospital sectors, government hospitals have a consistently higher rate of successful treatment for outpatients than do mission hospitals. However, mission hospitals have a higher rate of successful treatment for each disease group treated on an inpatient basis. These results may only reflect, however, the absence of evidence needed to estimate the number transferred or treated "unsuccessfully". Finally, at least 5% of all outpatient initial demanders (with the exception of the government hospital treatment of skin and musculo—skeletal diseases), and in many cases more than 102, are not treated in such a way that they can resume major activities. With the exception of mission hospitals in certain disease categories (pregnancy and puerperium, and skin and mus- culo-skeletal), this is true for inpatient care as well, although it appears that the overall rate of successful treatment for inpatient care is generally higher than that for outpatient care. 150 _ the Outpatient Treatment Process. Table 5.4 Estimates of the Rate of Successful Treatment, (ti), for a Government Government Mission Rural Units Disease Category Hospitals Hospitals W/Beds I&P 0.92 0.86 0.87 NG 0.40 0.40 0.22 AMB 0.75 0.69 0.80 NS 0.86 0.85 0.78 Circ 0.22 0.39 0.26 Resp 0.95 0.89 0.83 Alim 0.94 0.86 0.67 GU 0.89 0.75 0.95 Preg & Puer 0.58 0.57 0.94 Del w/o ---- ---- —--- S&MS 0.96 0.95 0.89 NB 0.73 0.55 0.92 111 Def 0.90 0.81 0.90 Inj 0.91 0.83 0.77 151 Table 5.5 Estimates of the Rate of Successful Treatment, (5;), for the Inpatient Treatment Process. Government Government Mission Rural Units Disease Category Hospitsls Hospitals W/BedS I&P 0.90 0.94 0.86 NG 0.82 0.89 0.66 AMB 0.83 0.95 0.70 NS 0.84 0.90 0.83 Circ 0.84 0.90 0.72 Resp 0.91 0.95 0.87 Alim 0.92 0.96 0.81 GU 0.94 0.97 0.81 Preg & Puer 0.97 0.98 0.79 Del w/o 1.00 1.00 1.00 S&MS 0.96 0.98 0.77 NB 0.79 0.82 0.83 Ill Def 0.87 0.95 0.72 Inj 'o.93 0.97 0.64 152 The Objective Function As Martin Feldstein discusses,12 there are a number of objective functions that can be specified with some justification when linear programming is used in the analysis of health services. The analysis of Uganda's stated objectives for curative health services undertaken in Chapter Four indicates that the government has placed a high priority on the provision of quality health care. The analysis further indicates that the government could, in pursuing this objective, maximize the number of successfully treated patients as well. In the empirical analysis discussed in the section which follows, an objective function consistent with the Ugandan government's quality and quantity objectives for its curative health services was used. The objective function uses the rate of successful treatment as its set of weights.13 (See Tables 5.4 and 5.5 for the values of these weights for each sector and treat- :nent process.) In addition experimentation was conducted with a second objec— tive function which differed from the first as follows: the second one possessed an additional constraint that at least one-half of all initial demanders in each disease category were provided with service. By adding this constraint to the objective function, all disease types from each treatment process were included in the linear programming solution. The results of this analysis are presented in Appendix C. Presentation of Empirical Results of the Linear Programming Model The Results The linear programming solutions for each sector of Uganda's health searvice system are presented in Table 5.6. The table describes the number ()f' successfully treated cases by disease category in each health sector, II Optimum Number of Case 153 Outputs Preg ‘ gbjegtlve 101:; To. . unc ion n is Puer Del w/o S 5 HS ND 111 Def Inj Equals Demanders Goverment Hospitals'l' 0'00 1 00 0‘00 0'00 0'00 0-00 4.553.032 Current Values of h5'017 “'911 10,379 1 976 ‘ 7 Initial Demanders ' '3 1 23’103 6'033'695 0.98 1.00 0.00 0.00 0. . Mission Hospitals"' 95 0 0° 61"7‘9 Current Values of 8,888 16,359 3,930 2.558 461 lnltlnl Demanders 3' 3'081 963'6" 0. 0 . . . . . Government Rural Units 0 1 0° 0 0° 0 0° 0 00 0 00 5.192.066 Current Values of'lzl 572 57 995 7 976 129 12 925 15 10‘ 1 973 ‘0‘ Initial Demanders ' " ' ' ' : : Shadow Prices of Case Initial Demanders Preg ‘ Puer Del w/o S h M5 N8 Ill Def Inj Government Hospitals (0) Shadow Prices(' 0‘00 0'62 0'00 0-00 0.00 0.00 (b) Cost of son-om °-“ ‘ 0-79 0-35 0.89 0.00 Mission Hospitals (a) Shadow "luau 0.02 0.11 0.00 0.00 0.33 0.00 (b) Cost of Non-Opt ' ‘ 0-39 0-29 - 0.11 Government Rural Unit: (a) Shadow Prices'4 0'00 1'00 0'00 0'00 0-00 0-00 0.41 - 1.12 0.27 0.58 2.70 (b) Cost of Hon-Op1 Shadow Prices and Sla4 Inputs'z) Government Hospitals (0) Shadow Prices' (1:) Black Quantit 11 (c) Current Values Mission Hospitals (I) Shadow Prices (b) Slack Quantltl ' (c) Current Values Government Rural Unit (I) Shadow Prices (b) Slack Quentlti (c) Current Values ) The optimum number of successfully treated cases is expressed as a proportion of the current value figure which is the number of 1n1t1al demanders. ) Inputs numbered l-ll are measured in hours; input 12 is measured in bed days: and inputs 13-17 are measured in shs. The current value figures under the optimum number of cases treated and slack quantitles of service—prov1d1ng inputs represent the two parts of the input vector, the serv1ce-domandlng and service-providing inputs respectively. disease character1stics were to demand service. The shadow prices of initial demanders show the amount by which the objective function could increase if an addltional initial demander with the given ) The cost of adding a non—optimal case to the solution shows the amount by which the objective function would decline if one of those types of cases were treated as opposed to those cases which are treated. ) The shadow price of the service-providing inputs shows the amount by which the number of successfully treated persons would increase if one additional unit of that input were made available. 154 when (a) the Objective function is specified in terms of the rate of success- ful treatment (RST) and (b) no constraint is placed on the minimum number for each type of initial demander included in the solution. This number is pre— sented in the table as a proportion of the total number of initial demanders. For example, 94 percent of all persons treated at government hospitals in 1968/9 on an outpatient basis for alimentary diseases (760,258) were success- fully treated. Similarly, 85 percent (of 93,821) and 54 percent (of 846,656) persons so afflicted were successfully treated at mission hospitals and government rural units respectively. The total number of successfully treated in each sector is shown on the far right-hand side of the table, under the heading "objective function equals." The proportion of total initial demanders successfully treated across sectors is presented in Table 5.7 below. TABLE 5.7 Proportion of Total Initial Demanders Comprised by Successfully Treated Cases, By Sector Sector Total Number of Initial Demanders Government Hospitals 6,033,695 Mission Hospitals 963,684 Government Rural Units 7 ,973 ,404 All Sectors 14,970,404 (Note: iFigures for this table were derived from Table 5.6) Total Number of Successfully Treated 4,663,032 614,749 5,192,066 10,469,847 Percentage Successfully Treated 77.3 63.8 65.1 69.9 155 The figures suggest that a larger pr0portion of initial demanders at govern- ment hospitals are included in the number of successfully treated in the optimal solution. Although the disease-specific RST's are generally higher in mission hospitals than in either government sector, the overall proportion of total initial demanders comprised by the successfully treated is lowest in the mission sector, 63.8 percent. This finding is due primarily to the fact that the proportion of initial demanders treated on an inpatient basis at mission hospitals is greater than is true for either government sector. Inpatient cases comprise 11.6 percent of total mission hospital cases, 3.7 percent of total government hospital cases, and 3.3 percent of rural unit inpatient cases. Since inpatient treatment requires a larger set of re- sources than does outpatient care, mission hospitals will necessarily treat a smaller proportion of initial demanders if they must be treated on an inpatient basis. The overall proportion of successfully treated cases in government rural units is lower than in government hospitals because the disease-specific RST is generally lower in rural units compared to hospitals. In regard to the output case mix, a similar pattern emerges across all three sectors of the health service system. First, with the exception of infectious and parasitic diseases (I&P), respiratory conditions (Resp.), and uncomplicated deliveries (Del. w/o), virtually all output is treated on an outpatient basis. It is significant that these three disease categories, when treated on an inpatient basis, require little, if any surgical service and have the lowest average length of stay for all disease categories.14 156 Second, allergic, metabolic and blood (AMB), does not appear in any sector's outpatient disease distribution solution. In addition, accident and injury cases (Inj.) appear in two of the three sectors (mission and government hospitals), but are significantly below the potential rate of successful treatment, and circulatory diseases (Circ.) only appears in the mission hospitals solution. The disease types mentioned above do not appear (or appear only mar- ginally) in the solutions for at least two reasons. First, most of the excluded case types are relatively large users of scarce resources. This fact is seen in each sector's technological matrix (See Appendix C), com— paring the disease-specific 'resource use' elements found in the tables. For example, the AMB disease category has a very high rate of use of scarce trained laboratory staff time, which is a binding constraint in all three sectors. (The constraint is measured by the input's shadow price in the third section of Table 5.6.) Second, most of the excluded case types have a lower rate of successful treatment than do those included in the solution. (See Table 5.4 for the disease-specific RST.) A lower RST tends to exclude a category from high priority in the Optimal solution since the weight attached to the objective function is low. Two sets of figures for each sector —- the shadow prices of case/disease types appearing in the LP solution and the cost of forcing non-optimal disease types into the solutionls -- are presented in the second section of Table 5.6. The shadow prices of disease types suggest a demand constraint imposed upon the optimal solution for each sector, as each sector could increase its output 157 of successfully treated cases of more initial demanders had illnesses fall- ing within particular categories. For example, in the case of government rural units, if one more person with a respiratory illness, rather than an alimentary disease, had presented himself at a health center, the objective function for government rural units would have increased by 0.60, which would be the increase in the number of successfully treated persons.16 The shadow price figures can be used to obtain a relative ranking of the disease category which is the most binding in the sense of increasing the value of the 'maximized objective function. Among disease groups which have the highest shadow prices and manifest the most binding demand constraints on the out- patient treatment process of all three sectors are respiratory (Resp.), nervous system and sense (NSS), genito-urinary (GU), skin and musculo-skeletal (S&MS), and new born (NB) cases. Only uncomplicated delivery (Del. w/o) and respira- tory (Resp.) cases appear consistently on the inpatient side; uncomplicated deliveries have shadow price values of 0.62, 0.41, and 1.00, and respiratory cases have shadow price values of 0.02, 0.11, and 0.31, for government hospi- tals and government rural units respectively. The second set of numbers presented in this section, the cost of forcing non-optimal case types into the solution, indicates the size of the trade-off ‘which is imposed on each sector when it treats one more case of a disease which does not appear in the Optimal solution. For example, the cost of treating an AMB case in a mission hospital on an outpatient basis means that the value of the objective function will decline by 7.63 successfully treated ‘persons. The cost figures presented in Table 5.6 indicate that most of the 158 non-Optimal disease types are treated on an inpatient basis, although the highest cost disease category is AMB, particularly when treated on an outpatient basis in the two hospital sectors. If, over time, each sector of the health service system operates under an implicit policy of treating persons on a "first come, first served" basis, and the entire spectrum of diseases (as indicated by the current value figures for each sector's initial demanders, contained in the first section of Table 5.6) appears in the queue for treatment, the cost figures for non-optimal case types is high. The third section of Table 5.6 presents the shadow prices and result- ing slack quantities of service-providing inputs implied by the linear programming solution for each sector of the health service system. The shadow price figures first indicate that the most binding supply constraints in the delivery of health services are trained laboratory and radiographic staff and other trained, non—medical staff (primarily ambulance drivers), all of which are manpower inputs. The high shadow price for medical and nursing assistants in mission hospitals is explained by the fact that there were only three such employees in mission hospitals at the time of the research. The actual shadow price figures show that if, for example, one more hour of trained laboratory staff time becomes available in government hospital's objective function would rise by 5.51 successfully treated persons. These shadow price figures thus can be interpreted as a first approximation of the marginal value of an hour of overtime for those already employed or the marginal value of the first few hours worked by a new employee. 159 Second, the shadow price figures indicate that at a time when many countries are concerned about the "need" for more doctors, other high level professional medical manpower, and beds, all of these inputs are in a position of considerable slack in each sector of the health service system in which they are used. Table 5.8 (derived from figures presented in the third section of Table 5.6), for example, presents the results of an analysis of the proportion of slack contained in the total supply of any given input. The results suggest that at least 66 percent of the available supply of doctor's time in government hospitals is not required in that sector's solution. Similarly, between 73 and 86 percent of the available supply of other "important" inputs - medical and nursing assistants, professional and enrolled level nurses, midwives, and beds in government hospitals are not required in the solution. Other important findings concerning slack supplies of inputs are con- tained in Table 5.8. First, the data suggest that both government and 'mission hospitals have the largest proportion of slack inputs, given the ‘nature of the linear programming solution.17 Second, the inputs least utilized in every sector (i.e., those which have the greatest pr0portion of slack) are the ones about which the government and the medical profession 'have been most concerned. Even in government rural units, the least utilized input is beds. Third, although some of the government rural unit inputs manifest a certain proportion of slack (as mentioned above), many of them were used almost completely in the optimal solution. Vehicle operation and maintenance, 160 Table 5.8 Relative Proportion of Slack for Each Service-Providing Input By Sector of Uganda's Health Service System in 1968/69, Given the Linear Programming Solution. Objective Function: Rate of Successful Treatment. Inputs Government Mission Government Hospitals Hospitals Rural Units Doctors 0.66 0.45 --—— Medical and Nursing Assts. 0.76 O 0.39 Professional Nurses/Midwives 0.73 0.49 ---- Enrolled Nurses/Midwives 0.73 0.67 0.26 Trained Lab Staff 0 O O Trained X-Ray Staff 0 O ---- Other Trained Medical Staff 0.49 0.52 0.12 Other Trained NonéMedical Staff 0 0.18 0 Other Non-Trained Medical Staff 0.23 0.52 0.11 Other Non-Trained Non- Medical Staff 0.38 0.50 0.03 Students 0.70 0.68 ---- Beds 0.86 0.78 0.56 Drugs, etc. 0.57 0.59 0.01 Food 0.70 0.68 0.37 Vehicle Operation and Maintenance 0.32 0.55 0.01 Electricity 0.56 0.48 0.31 Other Operation and Maintenance 0.57 0.58 0.22 Note: The figures found in this table were derived from data presented in the third section of table 5.6. 161 as well as drugs and medical supplies, are particularly noteworthy in this regard. This situation was not the case for any input in either hospital sector. Policy Implications Several health policy implications can be drawn from the findings of the linear programming solutions. First, the information suggests that over-emphasis has been and continues to be placed on the expansion of (a) hospitals vis-a—vis rural facilities and (b) high level manpower vis—a-vis technical personnel, whose specialties tend to facilitate improvement in accuracy of diagnosis. Results presented in Table 5.6 and 5.7 indicate that the rural health service system provides services to more initial demanders than do the two hospital sectors combined. It also provides inpatient care to a larger number of service-demanders than do government hospitals (263,192 compared to 224,702) and it performs approximately 50 percent of all normal deliveries (57,895 in rural units compared to 58,270 in hospitals). At the same time, its resource endowment is severely limited. Given these findings, in addition to the fact that beds represent the service-providing input with the greatest proportion of slack (Table 5.8), little economic justification is available for expanding the number of hospital facilities. Similarly, although training facilities for doctors, medical assistants, and nurses and midwives have been expanded in recent years, the most binding service-providing input constraint on the provision of high quality health services appears to be the under-supply of technical personnel, such as laboratory and radiographic technicians and skilled ambulance drivers. 162 Uganda's Third Deve10pment Plan notes the importance of training more laboratory and radiographic technicians,1'8 but the short supply of ambulance drivers, given present disease-specific transfer rates, has not been seriously addressed as yet. The data also suggest that more attention should be given to ways of improving the utilization of potentially slack inputs in each sector. The following set of possibilities is by no means exhaustive, but indicates the type of alternatives available. First, work within the curative treatment processes could be reorganized in order to (a) expand the supply of service- providing inputs for use in outpatient work or (b) improve the diagnostic process for patients of either treatment process by retraining some employees to work as laboratory or radiographic technicians. Second, service-providing inputs having considerable slack in hospitals (e.g., doctors, professional nurses and midwives) could be transferred to rural health facilities where such highly trained manpower currently does not exist. This reallocation is but one possible means by which the differential in the disease-specific rates of successful treatment between government rural units and hospitals could be reduced. (Complementary inputs such as drugs could also be included in any reallocation effort.) Finally, slack inputs could be used in the production of preventive health services to be consumed (a) at the health facility (e.g., young child clinics, to minimize AMB cases, which are not included in the solution), or‘ (b) on a community basis (e.g., improved sanitation facilities or water supplies). It is heartening to note that the Third Development Plan 163 significantly increases the budgetary allocation for water supplies, as opposed to curative health services per se.19 Perhaps manpower retraining schemes are needed to shift slack manpower resources into health-improving activities other than curative care. In addition, given the apparent demand by consumers for some form of health service, government pharmacies perhaps could dispense drugs at cost or at subsidized prices, as an alternative to the use of health facilities for such purposes; released resources could well be more productively employed in other sectors of the economy. If any of the above health policy implications are to be examined or explored, it is important that the Ministry of Health (a) continue to im- prove its statistics and statistical procedures and (b) expand its capacity to analyse the health service system by using techniques develOped in this research, expanding on them over time in order to answer other questions. Of particular importance would be the use of parametric programming techniques as a basis for exploring the extent to which a reallocation of inputs will both relieve currently binding constraints as well as introduce new ones. An Empirical Exploration into Factors Affecting;the Output of the Health Service System and Resources Available for Delivering;Health Services in the Future. The discussion which follows presents results of linear regression analy- sis of the factors affecting the output and inputs of Uganda's health service delivery system. This analysis is based on ideas developed in Chapter Four._ ‘It is undertaken in order to project trends implied in present curative health sservice policies so that decision makers can better evaluate the longer run :resource and output implications of present actions. 164 The analysis focuses on the factors affecting the total number of persons demanding service. It then turns to a discussion of changes that have occurred in the disease mix of initial demanders; these factors are of importance in understanding both the epidomiological transition which is under way in Uganda and the effects which that transition will have on health resources and outputs.20 An analysis of the factors affecting the rate of successful treatment is subsequently conducted. Finally the results of an analysis of the factors affecting the supply of two important inputs used in the production of health services, drugs and the recurrent budget, are presented. The results discussed in this section are incorporated into the final section of the chapter. Some of the estimated relationships found in this section are used in estimating the values of the elements of the input-vector for 1980/81. A further discussion of the important findings of that section is deferred until then. Factors Affecting the Total Number of Persons Demanding Service In Chapter Four, the functional relationship between the number of initial demanders and the variables affecting that number is specified. In a sense, the relationship specified is similar to a demand equation, where price, income and other factors peculiar to the market for curative health services are included. In Chapter Four, it was suggested that the relationship could be specified as follows: 165 (3) S = S(O,Y,a,b,I,d), where S = the total number of initial demanders, O - population size, Y 3 GDP per capita, a - the rate of successful treatment, b = price of curative health services, I = the incidence of illness episodes, and d = the average distance to health facilities. In addition, it was indicated that price and distance could be specified as follows: (4) b = b (3'1) and (5) d = d (H, (§%)), where H = total number of health facilities fig) = ratio of mission health facilities to all health facilities, and (9%) B the ratio of rural government facilities(non—hospital) to all government facilities. Empirical analysis, however, requires several changes in these equations. First, although there are time series data available for most variables, there .are no data available for the incidence of illness episodes and only limited «cross-sectional data are available on the rate of successful treatment; as a :result, the statistical analysis is likely to include misspecification error. Such specification error may result in biased and inconsistent estimates of ' tflne parameters of the variables included, however, only if the variables 21 (Excluded are correlated, positively or negatively, with those that are included. 166 In this explanatory analysis, it is difficult to determine whether correlation does exist between the two sets of variables. It may be argued that the in- cidence of illness episodes per person per year is negatively related to income, however, little evidence is available to support the hypothesis. For purposes of this analysis, it is assumed that correlation between the variables does not exist and, thus, that the parameter estimates do not contain any systematic bias. Another problem arises in estimating the effects of the price paid for the curative services consumed in mission health facilities, and the distance variable, d, which was used as a proxy for the opportunity cost of obtaining a "free" set of services provided by the government. Since each of the two variables are only partial indicators of the total price paid by the demander of a set of curative health services, the econometric analysis used several alternative specifications utilizing different combinations of these variables. As a first approximation, the analysis assumes that a linear functional form can be used and that the basic assumptions for using ordinary least squares (0. L. S.) techniques hold. The basic econometric equation estimated is presented below with the statistical results:-22 (6) S 8 22345.86 + 2.48(O) - 339.85(Y) - 80.88(%) + 43.49(H) - 397.97(§%) (O.68)*** (226.32) (57.82) (17.05)*** (174.66) (Linear Functional Form) ink n = 20 2 R2 =- .989, E = .984, d.w. = 2.01; 167 (7) s = 2.05 + l.6l(O) - 0.69(Y) + 0.19%) + 1.5301) - 3.8%-9%) (0.42)*** (O.3l)** (0.11) * (0.44)*** (1.38) (Linear in Logarithms Functional Form) n = 20 -2 -2 R = .991, R = .988, d.w. = 2.17; and * = significant at < 0.10, ** = significant at < 0.05, and *** = significant at < 0.01. The results suggest several important findings. Most important is the negative relationship between per capita income and the number of initial demanders. Results of the log equation indicate that the income elasticity of demand is negative and inelastic (—0.69). Reasons for this finding are difficult to discern but the following tentative solution is corraborated by other results. First, the data used to estimate the number of initial demanders were taken from the government sector of the health service system only. As a consequence, the sign of the income coefficient may reflect the fact that as incomes rise in Uganda, people demand a higher quality package of health services, i.e., a higher rate of successful treatment, which may be available either in mission hospitals or from urban-based private physicians. The data in Table 5.5 indicate that there is a generally higher rate of successful treatment for inpatient care in mission hospitals. The suggestion that people demand a higher quality package of health services as incomes rise also helps to explain the positive relationship found between the number of initial demanders and the ratio of mission hos- pitals to total health facilities, (see for example, the results of the linear 168 in logarithms equation (26) in Table 0.8, Appendix D. Of interest is the fact that this positive relationship occurs even though it might be expected that the existence of a positive money price in mission hospitals would retard the number of persons seeking such service. Since the rate of successful treatment is generally higher in mission hospitals, a person who requires care and has enough money will rationally seek service at mission hospitals rather than government hospitals. Another result of the analysis indicates, as expected, the positive effect of population growth on the number of initial demanders. The more interesting findings, however, lie in the relationships between the total number of initial demanders and (a) the total number of health facilities (H), and (b) the ratio of rural to total government facilities (2%). Total number of health facilities was positively related to the number of initial demanders, and the ratio of rural units to total government facilities was negatively related. These findings are not inconsistent with expectations concerning the opportunity cost of time and the way in which perceptions of quality would affect the consumption of health services.23 It is particular- ly interesting to note, however, that the number of demanders is highly responsive to changes in the number of facilities with the log equation, suggesting an estimate of the elasticity of initial demanders with respect to facilities greater than 2. In the case of the negative relationship between the ratio,-§%, and initial demanders, the findings suggest that although rural unit facilities . may be closer (thereby reducing the opportunity cost of time in consuming the 169 service) the perceived difference in quality between rural units and hospi— tals is sufficiently great to offset the reduced opportunity cost of con— suming. The data in Table 5.4 and 5.5 suggest that consumers are rational in their discrimination between the services and outcomes obtained in rural health units as opposed to either hospital sector. Since the signs of estimated coefficients for the variables'total health facilities' and 'ratio of rural units to total government units' are different, an interaction occurs between quality and price when the total number of health facilities increases due to an increase in the number of rural units. The net impact depends on the actual values of each variable, H and 93 but they do tend to offset one G’ another. Changes in Disease Mix of the Initial Demanders Although a complete analysis of the reasons for changes in the disease mix of initial demanders would encompass a study in itself, relating a number of socio—economic as well as medical and cultural factors, the results of an initial analysis of disease mix changes in hospital inpatient and out— patient categories are presented in this section. Annual data were used on the number of persons treated in government and mission hospitals on an inpatient and outpatient basis by 14 major disease classifications. The data for government hospitals extends from 1952 to 1968/69'118 observations), and the series for missions extends from 1958 to 1968/69 (12 observations).24 170 The analysis estimated the average annual rate of change for each disease group as a proportion of the total number of cases treated in a government or mission facility by inpatient or outpatient treatment pro— cess. Formally, the estimation equation was specified as follows: (8) Y -a +b X+ ewhere J J J Y - proportion of total disease treated in disease category j, J (1,...,14), (..I. I X a year, disturbance term, and M II a1 and b3 - are the estimated parameters. The results of this analysis are presented in Table 5.9. The most important finding, which corraborates information on disease mix change presented in Chapter Two (see Tables 2.13 - 2.16), is the sig— nificant decline in the proportion of infectious and parasitic diseases treated in hospitals on either an inpatient or outpatient basis. In govern- ment hospitals, for example, the annual rate of decline in the proportion of infectious and parasitic diseases was about 0.4% per year for outpatients and 0.8% per year for inpatients. This finding is coupled with a significant rise in the proportion of pregnancy and normal delivery cases treated on an inpatient basis in government hospitals. Also worthy of note are the positive coefficients for allergic, metabolic and blood diseases and alimentary diseases; within those categories many specific cases are related either to poor nutrition or to gastro-enteritis affecting mothers and children. Given that these four Table 5.9 , Estimates of the Annual Rate of Change in the Distribution 171 of Diseases Treated on an Outpatient and Inpatient Basis in Government and Mission Hospitals. Disease Category Government Hospitals Outpatient Inpatient 18 observations Mission Hospitals Outpatient Inpatient 12 observations Infectious & Parasitic New Growths Allergic Metabolic & Blood Nervous System and Sense Circulatory Respiratory Alimentary Genito—Urinary Pregnancy & Puerprium Delivery w/o Complication Skin & Muscule-Skeletal New Born Ill Defined Injuries -0.409** (0.151) 0.003 (0.002) 0.048*** (0.010) 0.273*** (0.055) 0.008** (0.004) 0.253*** (0.059) 0.093** (0.040) 0.121*** (0.016) 0.047*** (0.009) —0.oi1 (0.042) (0.006) 0.017 (0.063) -0.197*** (0.040) -0.822*** (0.147) 0.049*** (0.015) 0.197*** (0.026) -0.019 (0.019) 0.043* (0.023) ’ 0.050 (0.058) 0.252*** (0.023) 0.029 (0.035) 0.310*** :(0.063) " o.531*** (0.060) -0.181*** (0.046) (0.010) -0.400*** (0.051) (0.037) (0.121) —0.030*** (0.006) 0.095** (0.033) 0.183** (0.075) 0.003 (0.004) 0.083 (0.113) 0.233* (0.113) 0.004 (0.034) -0.126 (0.079) -q 0.090* (0.042) 0.024* (0.011) (0.192) -O.102* (0.055) -0.822*** (0.183) —0.035 (0.033) 0.424*** (0.059) 0.032 (0.023) 0.021 (0.012) (0.085) 0.247*** (0.069) -0.093* (0.050) £0.026 (0.110) 0.382 (0.393) -0.036 (0.034) 0.009 (0.046) -0.032 (0.050) -0.012 (0.027) NOTE: Significant at < 0.10 * Significant at < 0.05 ** Significant at < 0.01 *** The Figures presented in this table are the estimated bj's. 172 disease categories comprise 17.6% of the total outpatients treated and 44.4% of the inpatients treated in government hospitals in 1968/69,25 and given that the three latter disease categories are increasing as a proportion of all cases, it can be said that maternal and child health is rapidly becoming the most significant health problem facing Uganda today. Factors Affecting the Rate of Successful Treatment It was hypothesized in Chapter Four that the rate of successful treat- ment is a function of (a) differences in the composition of diagnostic v service inputs (—£,)26 * vr demanders S, and (d) the ratio of service-providing inputs to initial * demanders, (g). The medical technology variable could not be included in the S empirical analysis, as there were no data available to estimate its mag- , medical technology, (c) the disease mix of initial nitude. In addition, insufficient data were available on components of the rate of successful treatment for each health facility included in the sample. As a result, the dependent variable used in the analysis is a hybrid, 2. For the outpatient treatment process, it is the proportion of the total number initially demanding service, S, which is comprised of the number transferred to other health facilities or treated on an inpatient basis. For the inpatient treatment process, it is the proportion of those treated who are either trans- ferred to another facility or who die. Basically, 3 can be interpreted as a rate of unsuccessful treatment. Use of this dependent variable, 2, rather than a, makes more difficult the interpretation of the estimated coefficients; it is clear however, that the signs of the estimated relationships between the dependent and independent variables would change if a were to be substituted * 27 for a in the equation. 173 Three dummy variables were incorporated in the analysis in order to determine whether any institutional factors affect the rate of successful treatment. The first such variable was 'hospitals or non-hospital facilities'. The second was 'rural units with or without inpatient care', and the third distinguished between government and mission hospitals. The empirical analysis used cross-sectional data, as there was no time series information available on the rate of successful treatment.28 The econometric relationship estimated used ordinary least squares statistical techniques and it was assumed that a linear functional form was appropriate. The most interesting findings were obtained when the data for hospitals and rural units were combined to analyze the factors affecting the outpatient rate of successful treatment. The results are summarized below in a presen- tation of the estimated parameters obtained for one equation.29 * v (9) a = 299.46 + 338(35.) - 2.37(I&P) - 3.37(AMB) - 2.20015) - 2.36(Resp) r (2.27)* (0.67)*** (0.99)*** (0.74)*** (0.66)*** -2.33 (Alim) - 1.42(GU) - 3.00(Preg & Puer) - 2.58(S&MS) - 2.42(Ill Def) (0.69)*** (0.76)* (0.97)*** (0.66)*** (0.76)*** * -2.75 (Inj) + 0.08(g) + 4.38(Hosp) + 5.96(GRB) + 9.33(GOVMI), * (0.75)*** (0.07)8 (2.60)*** (1.68)*** (3,00)*** n a 20 R2 =- .81, £2 = .75, a * vr V where a, (3-3), and (x) are defined as above, I S I&P through Inj = the proportion of the total number of initial demanders comprised by that particular disease type, 174 Hosp = 'hospital, non-hospital' dummy variable, where 1 represents a hospital facility, GRB = 'rural unit with or without inpatient care' dummy variable, where 1 represents a rural unit with in- patient care, and GOVMI = 'government or mission health facility' dummy variable, where 1 represents mission units. Several important results emerge from the data. First, the positive relationship between the rate of unsuccessful treatment and the index of diagnosticians, b1, tends to suggest, given the construction of the index, that the more the training held by the primary diagnosticians in a given facility, the lower the rate of unsuccessful treatment. Second, the findings for the relationship between (a) the rate of unsuccessful treatment and (b) the ratio of total expenditures to initial demanders, b16 (a resource avail- ability variable), are inconclusive because the estimated parameter is not significant. The results suggest additional research, to disaggregate the resources variable, 3, in order to determine if there are particular factors, other than the diagnostician, which affect the rate of successful treatment. The three institutional dummy variables were very significant. The (Hosp) dummy variable suggests that there is a significantly higher rate of unsuccessful treatment in hospitals than in non—hospital facilities. This finding is primarily reflective of the fact that a higher proportion of outpatients treated at hospitals subsequently become inpatients. The same phenomenon manifests itself in the case of rural units (see the coefficient for the dummy variable GRB), where units with beds have a higher rate of unsuccessful treatment than do those without beds.30 The coefficient for 175 the third dummy variable (GOVMI) indicates that the rate of unsuccessful treatment in mission hospitals is significantly higher than in government hospitals, again reflecting a higher rate of inpatient treatment on the part of mission hospitals. The negative signs of the coefficients for case mix variables all suggest that the higher the proportion of a particular disease type in the" total disease mix, the lower is the rate of unsuccessful treatment for that disease. The results are somewhat puzzling in that all disease types intro- duced into the analysis have a positive effect on the rate of successful treatment. It may be that the case types not included in the analysis31 create a negative impact on the rate of successful treatment for outpatients. Even though the signs for all disease-type coefficients are negative, how- ever, their relative magnitude suggests that certain disease types have a greater impact on the rate of successful treatment than do others. For example, persons with diseases in the categories of AMB, Preg and Puer, and Inj tend to have a greater positive impact on the rate of successful treat- ment than do others. Factors Affecting Selected Service-Providing Inputs In Chapter Four, the service-providing input vector was disaggregated into four sub-sets: manpower, capital, other material inputs, and budget or financial inputs. The data were insufficient for econometric analysis of factors affecting the supply of manpower. Estimates of this input for 1980/81 projection, however, were based on projected training school output, adjusted 176 for retirements and turnover; for untrained personnel, it was assumed that the labor supply function is perfectly elastic at the existing minimum wage. An expected result was confirmed in an analysis of the capital input: the government's developmental priorities constitute the most important factor affecting the supply of capital available to the health service system. The present government's building and construction plans in the field of health were used as a first approximation of the future supply of capital to health, and these plans were utilized in the 1980/81 projections. More extensive analyses of two inputs -- imported drugs and recurrent budget -- were undertaken in order to better determine the factors affecting their supply. The results of these analyses were then used to estimate the value of several inputs in the service—providing input vector for the linear programming solution of the health service system in 1980/81. 2.1189 The availability of drugs is assumed to be related to the general economic well-being of the country as measured in three ways: (a) total monetary GDP (b) capacity to expand imports easily, as measured by the difference between total exports and total imports, and (c) the total level of imports. In addition, it is also expected that the number of initial demanders for health services may have an impact on the authorities' decision to purchase more drugs abroad. As a consequence, the statistical analysis included, in addition to the other factors mentioned, the number of initial demanders (the variable) and the variable lagged by one year, to reflect the lag in obtaining information. 177 Two examples of the statistical results of the analysis are presented below, along with the equations used in the estimation:32 (10) vn1 = -4.26 + 2.00Y -0.lOIM (Functional Form, Linear in (0.29)***(0.21) Logarithms) n = 16, 2 -—2 R = .96, R = .95, d.w. = 1.64, and (11) vn1 = -4.16 + 1.80 + O'OSS(t—l) (Functional Form, Linear in (0.42)***(0.22) Logarithms) n = 16, 2 2 R = .96, R = .95, d.w. = 1.49, where v = total drug imports from outside East Africa, Y - total monetary GDP, IM = total imports, and S(t-l) = total number of initial demanders lagged by one year. The results show clearly that total monetary GDP is the most important factor affecting the importation of drugs. The income elasticity to import (and demand) drugs appears to be highly elastic (1.8-2.0); this finding is consistent with other estimates of the income elasticity of demand for health services.33 Neither total imports nor the capacity to import (total exports minus total imports) were found to be statistically significant factors. The initial demander variable (an indicator of a demand-induced supply res- ponse) is also statistically insignificant. Recurrent Budget The budget, although not a direct service-providing input, provides money necessary for the acquisition of all other inputs. It is comprised of Several important factors: (a) the government health budget (central and 10681) (b) mission grants, (c) fees, and (d) gifts. Although grants and fees are generally increasing, the most important single item in the financia* 178 support of the entire health service system is the government's annual appro- priation. The empirical analysis, therefore, focuses on the factors which affect the government's recurrent health budget: (a) total monetary GDP (Y), (b) the pr0portion of recurrent expenditures comprised by recurrent health (GRHE GRE comprised by government rural units (9%), and (d) the government's capital (GRHE GRE record the effect of shifting governmental priorities (a) between health and expenditures ), (c) the proportion of total government health facilities budget for health. The two variables, and (9%), are incorporated to other sectors and (b) within the health sector. The capital budget for health is included to determine the effect of present capital priorities on recurrent expenditures; in addition, it is lagged by one and two years in order that the effect of past capital decisions on present recurrent costs may be analyzed. When either lagged form of the variable appears in the equation the non—lagged form is not included. It is assumed that the econometric analysis can be performed by specify— ing the functional relationship in a linear form and the OLS techniques apply. The results of two estimated equations are presented below as examples of the statistical results.34 (12) VGB(t) - -l8.23 + 1.7000 + 1.17(G'G'§:) + snug-g) (Linear in Logarithms (0.14)*** (0.20)*** (2.42)** Functional Form) n 8 20, 2 R =- .98, E2 - .98, d.w. = 1.58; 179 GRHE (13) VGB(t) = -5.77 + 1.96(Y) + 0.76(—————GRE) + 0.09 (GBC(t-—1)), (Linear in ‘°-15’*** (0...... 9032* $323332: 11 = 20, Form) R - .98, R2 = .98, d.w. = 2.02, where, VGB(t) - the recurrent government budget for health at time t, Y = total monetary GDP, GRHE C-Eii) = the proportion of government recurrent expenditures spent on health, GR C-E) = the proportion of total health facilities comprised by government rural health facilities, and GBC(t-l) = the government's capital budget for health lagged by one year. All results are significant. The coefficients for the monetary GDP variable are highly significant and the log equation values suggest a highly elastic income elasticity of demand for health services (around 2.0). Results for other variables suggest that they also significantly affect the size of the recurrent health budget. The fact that the capital budget variable is significant tends to indicate that capital expenditures have a lagged effect on recurrent health expenditures.35 Projected Solution of the Linear Programming;Model for the Year 1980/81 In order to obtain a better idea of the potential response of the health service system to future demands placed upon it by the growing papulation of the country, the input vector - both service-demanding and service-providing —- was projected to 1980/81 (the end of the fourth five-year planning period). The projections were subjected to the same linear programming exercise and the results of the projections are compared to the results for the year 1968/69. 180 Methods and Procedures Used to Project the Input Vector to 1980/81 In order to conduct the linear programming exercise for 1980/81, it was necessary to project the input vector from 1968/69 to that date. It was assumed that no significant changes in medical technology would occur during that period and thus, that the values of the technological matrix would re- main constant. It was also assumed that the basic methods of organizing resources and delivering health services within each sector would not change.36 The projection effort was conducted in five parts. First the disease mix from 1968/69 to 1980/81 was projected on the basis of information con- tained in Table 5.9. Where negative values occurred, a minimum value was assumed to exist. This was the situation, for example, in the case of the inpatient, ill defined disease category in government hospitals, where a minimum figure of 0.5% was used.37 It was assumed that the distribution of diseases for rural units would change at one half the average rate estimated for government and mission hospitals.38 The second projection concerned the total number of initial demanders for each sector. These estimates were developed in two ways. For mission hospitals, it was assumed that the total number of initial demanders would increase at the projected rate of population growth (3.6% per year), since no information exists on past trends. This estimate might have been adjusted by taking into consideration the rather high income elasticities of demand for health services from mission hospitals inferred from the analysis in the previous section. It was felt, however, that two factors mitigated against this adjustment: (a) the number of mission hospitals in Uganda will likely decline during the period, and (b) fees are charged for mission services. 181 An estimate of total initial demanders for government hospitals and rural units, was obtained by equations 4 and 8, Table D.8 , Appendix D. The estimates derived from these two equations were 36,915,000 and 38,016,000 persons respectively; for reasons of conservatism, the figure used was the lower of the two estimates. This figure was then disaggregated between hospitals and rural units on the basis of past trends in the percentage of initial demanders in each sector. The remaining three parts of the projection required estimates of each service-providing input for the three sectors of the health service system. For mission hospitals, a number of assumptions were made on the basis of available evidence. First, it was assumed that although no new facilities would be built, some additions might be made to existing facilities, such that available beds would rise to about 4,000 by 1980/81.39 It was assumed that the supply of doctor time would not increase over the period for two reasons: (a) the number of hospitals is not likely to increase, and (b) doctors tend to be in short supply throughout the world and the opportunity cost of being a missionary doctor is rising rapidly. The supply of trained nursing and midwifery staff is expected to expand over the period, primarily as expatriates are replaced and Ugandans trained in mission nurse/midwifery schools are retained. The remaining trained manpower categories are expected to expand to an average of one or two persons in each category of trained ' staff throughout the entire group of mission hospitals. The supply of student input was estimated on the basis of present levels of training and future plans for expansion described in the Third Five Year Plan.40 Finally, the remaining inputs were projected on the basis of population growth, adjusting that growth 182 upwards to reflect the high income elasticity of drug imports and recurrent health budgets reported in the last section. Turning to the two government sectors, the projections of the trained ‘manpower staff input categories were based on information presented in Third Five Year Development Plan.41 The total supply was then disaggregated between the hospitals and rural units on the basis of each sector's propor- tion of total available supply in 1968/69, and adjusted to reflect the rela- tively greater emphasis which has been given to rural health facilities since the beginning of the Third Plan. Other manpower input categories were pro- jected on the basis of personnel-to-bed ratios existing in 1968/69. The supply of students was estimated from information as to present and future trends in training programs. The supply of beds for each government sector was estimated on the basis of the short and long—run building plans described in the Third Plan, (i.e., an increase in the supply of beds to 2.0 per thousand in the early 1980's).42 Finally, estimates for other inputs used in hospitals were obtained by combin- ing estimated income elasticities for imported drugs and recurrent expenditures with income trends as described above. These figures were used, in turn, to estimate total recurrent expenditures for the hospital sector, and then, on the basis of 1968/69 figures for the proportion of total recurrent expenditures comprised by each input category, an estimate for 1980/81 was obtained. Esti- mates of other inputs used in the rural health facility sector, were based on the following: (a) a factor to reflect the increase in the average size of each rural unit, (b) a factor to reflect the increasing number of units, and 183 (c) a factor to reflect the changing distribution of types of rural units (an increasing proportion are health centers rather than dispensaries).43 Linear Programming Results for 1980/81 The linear programming results for 1980/81 are presented in Table 5.10. The results suggest the evolution of the health service system during the period 1968/69 to 1980/81, given the projected input vector for that date. These results are also indicative not only of future trends in health services but also of the scope of potential problems which may not be presently visible. First, the disease mix for the two government sectors in the 1980/81 solutions contains virtually no inpatient cases. This result differs con- siderably from the 1968/69 solution, where infectious and parasitic (I&P), nervous system and sense (NS), respiratory (Resp), and alimentary (Alim) cases appeared in at least one of the two government sector's solutions. (See Table 5.6). With respect to the mission hospitals, the most significant changes between the two solutions are the following: (a) outpatient injury cases are included to a much greater extent in the 1980/81 solution than they were in 1968/69 and (b) a larger proportion of inpatient disease types are represented in the 1980/81 solutions. (This increase is manifested in the fact that NS, Alim and NB disease categories are included in 1980/81 but not in 1968/69. At the same time, the GU category drops out of the solution in 1980/81.) The primary reason for this change in case mix -- and the resulting cost figures attached to the non-optimal case types -- is that the number of outpatient cases will likely increase significantly from 1968/69 to 1980/81, 1 II 111 Optimum Number of Cas1 Outputs 184 Objective Total No. Government Hospitals' Current Values of Initial Demanders Mission Hospitals'l) Current Values of Initial Demanders Government Rural Unit? Current Values of Initial Demanders Shadow Prices of Case Initial Demanders Government Hospitals (a) Shadow Prices (b) Cost of Hon-Op! Mission Hospitals (a) Shadow Prices" (b) Cost of lon-Op( Government Rural Unit (a) Shadow Prices (b) Cost of Non-Opi Shadow Prices and Slaq Inputs'z) Government Hospitals (a) shadow Prices (b) Slack Quantiti( (c) Current Values Mission Hospitals A (a) Shadow Prices' (b) Slack Quantitii (c) Current Values Government Rural Unit (a) Shadow Prices' (b) Slack Quantitid (c) Current Values (4} )reg.‘ Function Initial Puer Del w/o s a BS NC 111 Def Inj Equals De ders 0.00 0.00 0.00 0.00 0.00 0.00 10,960,762 2,516 122,159 11,936 3,246 2,497 50,041 13,108,200 0.60 1.00 0.00 0.02 0.95 0.00 1,045,231 3,141 33,050 5,366 4,111 4,661 4,409 1.480.000 0.00 1.00 0.00 0.00 0.00 0.00 19,300,561 6,607 99,156 9,542 120 14.434 22,492 24,908,700 reg.‘ Puer Del w/o s 5 N5 N3 111 Def In) 0.00 0.00 0.00 0.00 0.00 0.00 2.00 0.05 3.96 2.74 4.23 3.97 0 00 0.55 0.00 0.02 0.51 0.00 - - 0.27 - - 0.92 0.00 1.00 0.00 0.00 0.00 0.00 0.36 - 1.06 0.24 0.55 2.62 The optimum number of successfully treated cases is expressed as a proportion of the current value figure which is the number of initial demanders. ' Inputs numbered 1-11 are measured in hours) input 12 is measured in bed days; and inputs 13-17 are measured in shs. ) The current value figures under the optimum number of cases treated and slack quantities of service-providing inputs represent the two parts of the input vector. the service-demanding and service-providing inputs respectively. ) The shadow prices of initial demanders show the amount by which the objective function could increase if an additional initial demander with the given disease characteristics were to demand service. The cost of adding a non-optimal case to the solution shows the amount by which the objective function would decline it one of those types of cases were treated as opposed to those cases which are treated. + The shadow price of the service-providing inputs shows the amount by which the number of successfully treated persons would increase if one additional unit of that input were made available. I II Optimum Number of Casi Outputs 184 Objective Total No. Government Hospitals‘ Current Values of‘ Initial Demanders Mission Hospitals‘l) Current Values of Initial Demanders Government Rural Unit: Current Values of Initial Demanders Shadow Prices of Case Initial Demanders Government Hospitals (4 (a) Shadow Prices (b) Cost of Non-Opd Mission Hospitals (a) Shadow Prices (b) Cost of Non—Op! Government Rural Uniti (a) Shadow i’rices‘4 (b) Cost of Non-Opt Shadow Prices and Slaq Inputs‘z, Government hospitals (a) Shadow Prices( (b) Slack Ouantiti‘ (c) Current Values Mission Hospitals (a) Shadow Prices‘ (b) Slack Quantitit (c) Current Values Government Rural Unit! A (a) Shadow Prices (b) Slack Ouantitid (c) Current Values‘ <1 reg.‘ Function Initial Puer Del w/o S a MS NC Ill Def Inj Equals De ders 0.00 0.00 0.00 0.00 0.00 0.00 10,968,762 2,516 l22,159 11.936 3,246 2,497 50,841 13,108,200 0.60 1.00 0.00 0.82 0.95 0.00 1,045,231 3,141 33,058 5,366 4,111 4,661 4,489 1,480,000 0.00 1.00 0.00 0.00 0.00 0.00 19.300.561 6,607 99,156 9,542 120 14,434 22,492 24,908,700 reg.‘ Puer Del w/o s a MS NB Ill Def In) 0.00 0.00 0.00 0.00 0.00 0.00 2.08 0.05 3.96 2.74 4.23 3.97 0.00 0.55 0.00 0.02 0.51 0.00 - - 0.27 - — 0.92 0.00 1.00 0.00 0.00 0.00 0.00 0.38 - 1.08 0.24 0.55 2.62 The optimum number of successfully treated cases is expressed as a proportion of the current value figure which is the number of initial demanders. Inputs numbered 1-11 are measured in hours: input 12 is measured in bed days; and inputs 13-17 are measured in shs. The current value figures under the optimum number of cases treated and slack quantities of service-providing inputs represent the two parts of the input vector, the service-demanding and service-providing inputs respectively. The shadow prices of initial demanders show the amount by which the objective function could increase if an additional initial demander with the given disease characteristics were to demand service. The cost of adding a non-optimal case to the solution shows the amount by which the objective function would decline if one of those types of cases were treated as opposed to those cases which are treated. The shadow price of the service-providing inputs shows the amount by which the number of successfully treated persons would increase if one additional unit of that input were made available. 185 such that each sector in the health service system, particularly government hospitals and rural facilities, will be able to maximize the total number of successfully treated cases by treating a greater number of the less costly outpatient cases, rather than by treating any cases on an inpatient basis. The change in the disease mix of initial demanders will also lead to a sig— nificant increase in the proportion of successfully treated cases from the total number of initial demanders. In 1968/69, the proportion of success- fully treated cases in all sectors combined was approximately 69.9%. In 1980/81, the projected rate will be approximately 79.3%. Comparative analysis of the cost of forcing non-optimal case types into the linear programming solution between 1968/69 and 1980/81 suggests that a relatively constant pattern exists over the entire period (compare Tables 5.6 and 5.10). The only significant changes between 1968/69 and 1980/81 appear in the relative magnitudes of the costs. Inpatient costs will increase significantly over the period, particularly in government hospitals. Perhaps more important from a policy point of view, however, are the changes in the shadow prices of the constraints imposed by the service- providing input variables. The analysis presented in Table 5.11 concerning the relative proportion of slack for each service-providing input provides additional information as to the nature of these constraints, assuming that the estimated service—providing input vector for 1980/81 reasonably reflects the situation at that point in time. Assuming that the government's health Inanpower training programs are implemented, the constraints which existed throughout the health service system in 1968/69 in trained laboratory staff 186 Table 5.11 Relative Proportion of Slack for Each Service-Providing Input by Sector of Uganda's Health Service System in 1980/81, Given the Linear Programming Solution. Objective Function: Rate of Successful Treatment Inputs Government Mission Government Hospitals Hospitals Rural Units Doctors 0.83 0 -_-_ Idedical and Nursing Assts. 0.80 0.81 0.36 Professional Nurses/Midwives 0.98. 0.28 ---- Enrolled Nurses/Midwives 1.00 0.77 0.72 Trained Lab Staff 0.84 0 0.86 Trained X-Ray Staff 0.38 0.31 ---— Other Trained Medical Staff 0.89 0.62 0.72 Other Trained Non-Medical Staff 0 0 0 Other Non-Trained Medical Staff 0.34 0.19 0.29 Other Non-Trained Non-Medical Staff 0.55 0.21 0.17 Students 1.00 0.24 ---- Beds 1.00 0.54 0.83 Drugs, Etc. 0.78 0.73 0.20 Food 1.00 0.72 0.77 Vehicle Operation and Maintenance 0.46 0.69 0.14 Electricity 0.74 0.68 0.64 Other Operation and Maintenance 0.87 0.68 0.53 Note: The figures found in this table were derived from data presented in the third section of Table 5.10. 187 will have been largely erased by 1980/81 except in mission hospitals. The input variable which will remain the most binding constraint in 1980/81 (a variable which was not addressed in the third development plan) is the manpower category of other trained, non-medical staff, whose major component is ambulance drivers. (See the shadow prices for this category, Table 5.10.) When the small average annual rate of 3.4 transfers per thousand (the average rate in 1968/69 for a sample of 19 rural health facilities, Table 2.12), is applied to a base of approximately 25 million attendees at rural units, 13 ‘million at government hospitals and 1.5 million at mission hospitals, the total number of persons transferred becomes quite large. Even if the average transfer rate were to decline significantly, there would still be an absolutely large number of transfers. Further, given that the government is increasing the number of ambulances in rural areaséé and is spending large sums of money to improve roads throughout the country, it is difficult to conceive that the transfer rate will decline. It is also interesting to note that while the input complementary to drivers -- vehicle operation and maintenance -- is slack in the linear programming model, its proportion of slack is relatively small in all three sectors in both 1968/69 and 1980/81 (see Tables 5.8 and 5.11). In terms of training programs, the government should implement a program, not just to train ambulance drivers, but also to render basic first aid and emergency medical procedures. Perhaps some of the less intensively used manpower cadres, indicated as slack inputs in the linear programming solutions, could be trained as ambulance drivers. 188 Turning to an analysis of the relative degree of input slack, the information presented in Table 5.11 suggests that, given the linear program- ming solutions for each sector in 1980/81, the proportion of slack through- out the system will have generally increased from 1968/69 to 1980/81. This situation is particularly true in the case of trained manpower and in the government hospitals. Government hospitals have such a high proportion of inputs as slack due to the large proportion of all inputs used in the in- patient treatment process, whereas the linear programming solution suggests that little if any inpatient care is warranted. At the same time, while ‘not containing as much slack, inputs in rural units also will experience an increase in slack over the period. What are the policy implications of the above analysis? First, assum- ing that (a) the objective function used conforms to the priorities of Uganda, and (b) the estimated 1980/81 input vector conforms to reality in an approx- imate way, a considerable expenditure will have been made by the government for the training and provision of staff for very costly health maintenance activities, i.e., curative inpatient care. Perhaps spillover benefits will accrue to the entire population as a result of having many medically trained persons living throughout the country, but those benefits would have to be quite large in order to compensate for the income disparities which will likely result from increased opportunity for trained persons to engage in private practice.45 Secondly, it would appear that although rural health units will receive a larger proportion of available inputs through 1980/81, additional resource reallocation between government hospitals and rural health units is warranted. 189 A particularly important area for the beginning of such resource reallocation is in drugs. Although the drug input is never a binding constraint in any of the linear programming solutions, its pr0portion of slack in the rural units is very low. If the rural units were disaggregated, it is highly likely that many would experience a binding drug and medical supply constraint. The personal stories of medical assistants recounting how they often run out of monthly allotments in two weeks, if not before, provide graphic illus- trations of the general point. Third, no more units with beds appear to be needed since the slack proportions for government hospitals and rural units increases during the period. The only exception to this statement lies in maternity beds for normal deliveries and minor complications. Finally, given the large and increasing degree of slack in each sector, the input intensity in the health services provided each person demanding service might be increased. There appears to be slack diagnostic time generally available, both among doctors and medical and nursing assist- ants, as well as among the nursing staff generally. The nursing staff could be used to a greater extent in the outpatient department or, as was earlier suggested, in the provision of preventive health care. Such reallocations could well yield a higher quality of care and the rate of successful treatment presumably would manifest that fact. Summary In this chapter, the linear programming model for analyzing Uganda's health service system was empirically specified. The procedures, methods, and assumptions used for estimation of the input vector, the technological 190 matrix, and the objective function -- when specified in terms of the rate of successful treatment -- were described. The model was empirically tested using 1968/69 data from the three main sectors of Uganda's health service system (government hospitals, mission hospitals, and government rural health units with beds for inpatient care). The implications of the results were discussed. An exploratory empirical analysis of the factors affecting the vector of initial demanders, the rate of successful treatment, and several components of the service-providing inputs was conducted. Finally, projec- tions and analyses of the health service system in 1980/81 were made, utiliz- ing the linear programming model developed for and tested on 1968/69 data. This analysis was conducted by projecting both subsets of the input vector from 1968/69 values to 1980/81. The health policy implications of the analysis were presented. 191 Footnotes The linear programming model could have been constructed to conform more realistically to the production process discussed in the first section of Chapter 4 by incorporating into one vector the inpatient and outpatient initial demanders and then having two lower submatrices, each 14 x 14, containing a principal diagonal whose elements have the same significance described above. The scope of the s technological matrix would have been i=(l,...,31), j=(l,...,28). This procedure has not been followed for reasons of convenience, and the substance of the analysis is not affected. Output data for government and mission hospitals are summarized in monthly reports according to sex and disease. Government rural units, however, summarize data on a monthly basis according to a disease classification different than that used by hospitals, and data on age and sex are included only in summarized form, without disaggregation on a disease-specific basis. The disease classification scheme em- ployed by rural units has been adjusted to conform to that employed in government and mission hospital units through detailed analysis of the daily medical record books for several rural units. See copies of MP 74, 75, and 77 in Appendix F. Martin Feldstein, in his example of the use of linear programming in health services planning, used the budget as a variable. His use of the budget, however, was due to his desire to keep the analysis sugges- tive of the applicability of the methodology. He further implied that it would be desirable to disaggregate the input vector, particularly to include drugs and medical supplies, in order to provide a greater degree of realism in the planning problem. See Martin Feldstein, Economic Analysis for Health Service Efficiency (Amsterdam: North— Holland Publishing Company, 1967), Chapter 6. A copy of the survey instrument used is contained in Appendix F. The data on initial demanders were randomly audited by going back to original sources in rural facilities. See J. Galea, "Inventory, Appraisal, and Assessment of the Basic Health Services of Uganda, Developments for a Malaria Eradication Programme," (Jinja, Uganda: Malaria Pre-Eradication Programme, World Health Organi— zation, 1967). See also Hallway, Jane, A Survey of Church Related Hos- pitals in the Anglican Province of Uganda, Rwanda and Burundi, (Kampala:. Provincial Medical Board, 1972) and Survey of all our Catholic Medical Units, (Kampala: Catholic Medical Bureau, 1969). 10. ll. 12. 13. 14. 15. 192 See Martin Feldstein, Economic Analysis for Health Service Efficiency, pp. 171-175, for a discussion of the methodology he employed. The Ministry of Health was beginning to develop a complete hospital- specific budgeting system in 1969. See Tables 2.13, 2.14, 2.15, and 2.16 for an analysis of the differences in the disease distribution between government and mission hospitals. In Table C.9, Appendix C, the calculations made to obtain an estimate of the average quantity of drugs and medical supplies used by disease category on an outpatient and inpatient basis are shown. A better method of estimating the cost of drugs and medical supplies on an inpatient basis would be to examine a sample of inpatient case records. At one point, the researcher began such an analysis but ran into great difficulty deciphering handwriting and could not obtain necessary assist- ance from medical staff. At some point this research should be conducted. One additional complicating factor, not taken into consideration in the calculations but of which the researcher is aware, is the problem of theft. It was well known in certain markets that individuals could ob— tain drugs for a fee, particularly antibiotics such as penicillin. Exactly how to take this problem into consideration is unclear, but one must be aware of its existence when interpreting the results. Basically, the data used to estimate the disease-specific rates of successful treatment for each of the three sectors were obtained from the following sources: (a) published statistics of the Ministry of Health, (b) unpublished records of a sample of health facilities from all three sectors of the health service system, and (c) a patient follow—up survey used to estimate the number of unsuccessfully treated initial demanders. Martin Feldstein, Economic Analysis for Health Service Efficiency, pp. 175-178. See Chapter 4 for a discussion of the rationale for using the rate of successful treatment as the appropriate set of weights. It is possible to discern the differential resources requirements among disease groups by looking at the inpatient portion of the technological matrix for each sector. For example, differences in the average length of stay may be discerned by looking at the twelfth row, beds, in Tables D.5, D.6, D.7, Appendix D. The shadow price and cost figures are to be interpreted as estimates of marginal prices or costs rather than average price or cost since an extra marginal change in any one or more or the inputs would yield a solution different than that indicated. 16. l7. l8. 19. 20. 21. 22. 23. 24. 25. 26. 193 Since each initial demander is weighted by his rate of successful treatment, (less than one, except for inpatient uncomplicated deliv- eries), the values of the shadow prices for each disease type is constrained to values between zero and one. In Table c.13,Appendix C, an alternative linear programming solution is presented. The principal difference between the alternative and the one presented in the text is as follows: the alternative requires that a minimum proportion of each case type, on an inpatient and out- patient basis be treated before the optimizing process occurs for the remaining service demanders and service-providing inputs. Although the nature of the solution in terms of the Optimal disease mix and the values of particular shadow prices, costs of forcing non-optimal case types into the solution and the quantity of slack of each input are different, a similar pattern is observed with respect to the proportion- ality of slack in the "important" service-providing inputs such as doctors, nurses and beds. For example, in government hospitals the proportion of doctor's slack time is 36.8 percent, for professional nurses and midwives 52.0 percent, and for beds 73.9 percent. Republic of Uganda, Uganda's Plan III, para. 17.74, p. 322. Republic of Uganda, Uganda's Plan III, p. 120. See Abdel R. Omran, "The Epidemiologic Transition: A Theory of the Epidemiology of Population Change," Milbank Memorial Fund Quarterly, 49, 4, Part 1, (October 1972), pp. 508-538. See Jan Kmenta, Elements of Econometrics, (New York: The Macmillan Co., 1971), pp. 392-394. See Table D.8, Appendix D, for the statistical results of other estimated equations. As the total number of health facilities increases (assuming that there is some reasonable concern for equitable spatial and population distri- bution), the average distance required for travel to obtain health ser- vices is reduced, and the amount of time spent in the consumption process falls; the cost of consuming such services and is thereby reduced, leading an expansion of demanders. It was assumed that the one-half year observation for January to June 1960 was a valid observation, when the reporting period for the Ministry of Health changed from a calendar to a fiscal year in June 1960. Data to calculate these figures are found in Tables 2.13 and 2.15. The corresponding mission hospital figures are 20.3% and 41.7%, respectively. v The index of diagnostic service inputs, G-Er), is described more fully in section 4, Appendix C. Briefly, r however, it is constructed such that it can vary between one and five,with a value of one indicating that all diagnostic services are provided by doctors and a value of five meaning that all diagnostic services are provided by untrained dressers. 27. 28. 29. 30. 194 The statement regarding the changing signs between the desired and estimated coefficients can be demonstrated for the simple case as follows. The desired relationship to be estimates is: a a b1 x1. However, the data available enable the estimation of the following relationship only: * * a = ble, where a = l — a. The question is: given an estimate for b , what is b ? By simple algebraic manipulation, it can be shown that b = - b . 1 ‘1 2 As long as 1 x b > b < 0. Where there are more than one 2 :1’1 independent variables in the equation the relationship between b and the estimated b becomes more complex; however, its general form is of the following nature: b1 --l1, 0 - b2, where Q is a set of terms greater x than one. In the case of two independent variables in each equation: a = b x + b x and l l 2 2 l — a - b3):1 + b4x2, Q = x2 (b3 + b4) x1 ° In these more complicated instances it becomes even more evident that the sign will change given the additional impact of the Q term. Cross-sectional data were obtained from all government rural health facilities in Busoga, Ankole, and East Mengo districts (n-39), from 7 government hospitals and 18 mission hospitals. See Appendixl) for additional statistical results. The results presented are disaggregated between hospitals and rural health facilities for out- patient and show hospital inpatient results as well. Information on the number of unsuccessfully treated patients, as deter- mined by patient follow-up studies, is not contained in the value of the dependent variable. If it were, the signs and the magnitude of the estimated coefficients would likely change significantly. 31. 32. 33. 34. 35. 36. 37. 195 The exclusion is due either to lack of information -- i.e., the disease comprises a very small proportion of total outpatient cases seen - or to the fact that normal deliveries are not conducted on an outpatient basis. Other statistical results are presented in Table D.10, Appendix D. See Benton Massell and Judith Heyer, "Household Expenditure Analysis in Nairobi: A Statistical Analysis of Consumer Behavior," Economic Development and Cultural Change, (January 1969), pp. 212-234. Their estimates of the income elasticity of demand for health services varied between 1.07 to 1.42. Other statistical results are presented in Table D.1l, Appendix D. The government appears to be cognizant of this important relationship. See Republic of Uganda, Uganda's Plan III, pp. 121-124. It is unfortunate that economic methods of analysis cannot adequately handle major changes in the socio - politico - economic system, such as have occurred in Uganda since 1971. A recent issue of Africa Reports indicates that one-third to one-half of all the doctors in Uganda prior to 1971 have left the country -- including nationals. In addition, many other health workers have left. I suspect the situation is parti- cularly acute in mission hospitals. The lack of trained health man- power has manifested itself in at least two ways: Africa Report mentions that the government is attempting to recruit health workers from North African countries, and Mr. Roy Innis of the Congress of Racial Equality has been recruiting Black Americans with health skills to work in Ugan- da. In addition there is evidence that the medical manpower training programs, (including both the medical school at Makerere and other training programs) have been seriously curtailed. Exactly how these shocks will work themselves out over a ten to fifteen year period is difficult to predict; the analysis conducted in this last section, however, is suggestive of the situation as it might have existed if the health service system had developed according to the plans and priorities developed in the 1970/71 period and spelled out in Uganda's Third Five Year Development Plan, 1971/72 - 1975/76, Chapter 1, and 17. Figures were used which indicated a rather precipitious rate of decline in the pr0portion of total cases comprised by infectious and parasitic diseases -- 0.4% per year for outpatient and 0.8% per year for inpatient. More research attention should be given to these figures in future pro— jections, since it is unlikely that the etiology in Uganda will change to the extent that such a rate of decline will continue. 38. 39. 40. 41. 42. 43. 44. 45. 196 This assumption is based primarily on pragmatism, as little alternative information is available. See Republic of Uganda, Uganda's Plan III, para. 17.50, p. 315. See Republic of Uganda, Uganda's Plan III, para. 17.62, p. 319. See Republic of Uganda, Uggnda's Plan III, Table 17-1, and pp. 315-324. See Republic of Uganda, Uganda's Plan III, para. 17.26, 36, 39, 40 and 43. pp. 308-313. The resulting figure used to project 1968/69 figures to 1980/81 was 4.0. This figure was derived in the following way: (1) The average size health facility, in terms of beds, would increase from 15.8 beds in 1968/69 to 25.0 beds in 1980/81, an increase of 58.2%. (2) The total number of units would increase by 97% over the period, from 169 in 1968/69 to 352 in 1980/81. (3) The average increase in cost due to the general improve- ment of the rural units from dispensaries to health centers is 30%. This figure is based on 1968/69 average cost data from a sample of rural health facilities in Busoga, East Mengo and Ankole Districts. The final figure was determined by multiplying these three figures together: 1.58 x 1.97 x 1.30 - 4.0. See Republic of Uganda, Uganda's Plan III, para. 17.32, p. 310. See W. Lee Hansen and Burton A. Weisbrod, "The Distribution of Costs and Direct Benefits of Public Higher Education: The Case of California," Journal of Human Resources, 4,2 (Spring 1969), pp. 176-191. CHAPTER SIX In this chapter, the analysis focuses on the macro-economic effects of Uganda's health service system. To the extent that data are available, the relationships between Uganda's development ob— jectives and the provision and development of the health service system are examined. The analysis focuses on the relationships between health services and (a) population growth and demographic change, (b) the balance of payments, (c) employment, (d) other industries, and (e) developmental equity.1 Health and Demographic Change The literature on health and demographic change has emphasized two relationships. Some analyses have examined the effects of avail- ability of health services on the rate of population growth and demo- graphic change. Others have discussed the effect of population growth on both the general level of health in the population and on a coun— try's health services. Most of the literature, however, is specula- tive in nature, offering very little empirical information. This problem is particularly acute when long run effects are discussed, for virtually all of the studies make inferences related to the impact of improved health standards on demographic variables. Approximate quantitative estimates of these relationships, and the relative importance of these relationships vis-a-vis other socioeconomic changes in the population, have not been developed. 197 198 Health Effects on P0pulation Growth The literature suggests that disease eradication tends to increase the rate of population growth by causing a decline in the death rate and/or an increase in fertility.2 Several studies have also indicated that the reduction in morbidity rates brought about by the eradication of disease lends to an increase in human capital.3 It is also maintained that the improvement of health services, par— ticularly the addition of both maternal and child health services and family planning services, leads to demographic change. It is assumed in this regard that the rate of population growth over time will decline as families perceive an increase in the probability that their children will live to adulthood.4 Health services are also said to affect demographic change when the availability of the services is improved, but family planning services are not provided concomitantly. Improvement of the health of females and reduction of high infant mortality rates are both assumed to result in an increase in population‘growth.5 It is also maintained, however, that an increase in nutritional standards will result in a decline in infant mortality as malnutrition is often complicated with other severe illnesses. This decline in infant mortality, it is suggested, will lead in the long run to a decline in desired family size, and thereby a reduction in fertility and decline in the rate of population growth.6 Effects of Population Growth on Health There has been some discussion in the literature of the effects of population growth both on health standards and on the demand for 199 health services. First, it has been suggested that rapid p0pulation growth leads to a reduction in health standards. Standards such as hospital-bed-to-population ratios and doctor-to-population ratios are said to decline as a result of population increase and the assumption is made that the health of the population will also decline. As indicated in Chapter 3,however, the use of medical resource-to-pOpu- lation ratios as a measure of health output and health standards is spurious. Second, it is maintained that rapid population growth will have an impact on government decisions with respect to budget allocations. It is suggested that although health services are likely to receive increased budgetary allocations, standards of health will not in- crease because rapid population growth itself retards the service system's ability to keep up with the increasing service demand.7 After suggesting that the relationship between population growth and health standards is negative, most authors explore alternative cost- minimizing policies, such as increasing paramedical staff,8 or ex- panding the training of both professional and paraprofessional staff. Population is also said to affect health in another way. As the age structure of the population changes, the composition of demand for health services shifts. There is a change in the incidence of certain diseases, which affects the disease mix within a pOpulation.9 When the rate of population growth increases, a larger proportion of the population is found in younger age brackets; an increasing pro- portion of the demand for health services thus results from maladies afflicting young children. As the rate of population growth falls 200 and the age structure of the population becomes older, an increasing pr0portion of the disease mix is composed of circulatory and heart related problems and cancer.10 Demographic changes also affect health and the health service system through the process of rural to urban migration. Health ser- vices in urban areas tend to be delivered in hospitals, while health centers or other less sophisticated facilities predominate in rural areas. As peOple migrate to urban areas, the cost of providing health services increases because hospitals are based on more expensive organizational structures.11 Finally, the impact of a shorter birth interval on health has 12 been analyzed in the literature. A shorter birth interval increases both the illness rate per family, and the rate of illness per family member per unit of time. It also tends to decrease the nutritional status of the entire family which may have the further effect of changing the distribution of diseases contracted. Health and Population: The Ugandan Case The purpose of this section is to present information which highlights some of the implications discussed above relating health and population change. Two studies were undertaken in this regard: the first examined the availability of health services in relation to the rate of population growth in Uganda's sub-counties from 1959 to 1969, and the second examined the extent to which the age structure has become younger over that period in relation to health service availability. 201 Population Growth and Health Service Availability_ In order to test the relationship between the availability of health services and the rate of population growth, two types of data were required: (a) population census data for at least two periods, and (b) the type and location of every health facility in both census years.13 Once gathered, the data were used to conduct three different tests on two relationships between health services and population growth. The first test was made to determine whether there was a significant difference in the mean annual rate of population growth over the decade in those subcounties (called divisions or gombololas) with no health facilities as compared to those gombololas in which there was a health facility. Health facilities were further classified into two groups: (a) those with maternity services (including antenatal and young child clinics) and (b) those without such services. The specific taxonomy used is as follows: (1) No health facility in 1959 or 1969; (2) No health facility in 1959, but a health facility in 1969 without maternity services; (3) No health facility in 1959, but a health facility in 1969 with maternity services; (4) A health facility in 1959 without maternity services, and a health facility in 1969 without maternity services; (5) A health facility in 1959 without maternity services, and a health facility in 1969 with maternity services; (6) A health facility in 1959 with maternity services, but no facility in 1969; and (7) A health facility in 1959 with maternity services, and a health facility in 1969 with maternity services. 202 The test was conducted to determine the effect of the avail- ability of health services on two demographic variables, births and deaths, which, in interacting, constitute the basic mechanism of population growth. In order to focus solely on the effect of changes in birth and death rates, the 599 gombololas in the country 14 were divided into a rural and urban groups. The urban gombololas were dropped from the analysis due to the likelihood of high levels of net immigration, which would bias the test results.15 The resulting sample of 467 "rural" gombololas were analyzed and the results are summarized in Table 6.1 and Figure 6.1. As seen in Table 6.1, few of the mean pepulation growth rates are statistically different from one another, primarily because the standard devations are quite large. Where there are significantly different means, no consistent pattern emerges. However, Figure 6.1 reveals several interesting trends which, although not statistically significant, do point to the possibility of a linkage between popula- tion growth and the availability of health services.16 Figure 6.1 illustrates the data found in Table 6.1, eliminating health facility categories 3 and 6 due to a lack of observations. The subcounties, grouped according to health facility classification, re- presenta.continuum of services: (a) no services, (b) recently intro- duced curative health services, (c) curative health services available for at least 10 years, and (d) both curative and maternal and child health services available for at least 10 years. In Figure 6.1, part (A), we see that in all cases - in the country as a whole as well as in each region - the rate of population growth tends to be greater in gombololas with no health.facilities, than it is in those where both 2(13 sowuuannon mo mum» amass. onus on» a“ wau«>uom .uaz an“) ac CH ...H .IIIQQUNBQM .umx as“; muaaauau nuance so as .u.= oz olmoua>uom .uax saw: an a“ auaaauam euaaum oua>uom .umx no“: so an .m.=unoo«>umm .uwx o\u on :« xuaaauam zuaaom uuoa>uum .umx o\a me a« .m.x nooa>uwm .uax o\3 an a« suaaau-m guano: nuou>uom .ua: sun: «0 an suaaauam asaaom--on c« >ufiafiuam suaoom oz mouu>uom .umx o\> me as xuaaauam euaaum--ma a“ zuamuu-m sud-um oz we no on a“ nuaaaoam cog-u: oz samba huaawuau guano: mo «squamouau mo noquasmaqu nma Aoaq.~0 ch.n o~ maso.nv aqn.« a Aomn.a0 coo.n n Aaoa.nv Nom.m an Amam.~v ono.¢ an ans: “Au «AN Haw 00H Acan.~v nn~.c Ad Aseo.uv «no.< an Aa«~.H0 cod.“ 4 Aoqn.~v oma.o Ga Assn.~0 can.e as euuoz ecu Assa.av ~mq.~ «a A000.0v 000.0 a Aomm.ov N50.~ 0H Amca.~o ona.n o Assn.av mom.~ as “Nao.ov man.~ an unflm «An axw «Ad 00 An-.~v owo.n ma A000.cv 000.0 H Ammo.«v neo.H n Aooo.«0 000.0 H “sno.~0 oaa.< an Aeso.n. an~.e «a ovsamsn sodas“ use» a noun: H0>0H 0~.0 us canoe .uuqa unnuwuucwum aaaduauouusuw msoaun>uomao «muck .aoauo«>ov vuavoauo osu monocuch vuwzu on“ new .nuaouu hopes: vacuum 05H .mcoaua>houno uo amass: may usuao«vsu «duo sumo cu yonssc u-uwu 03H ---- .u-- a.“ u--- ---- “An nun- use nae «An ¢A~ ~Ac as No on em dos use “aka.” A~o~.av Aasa.o0 Aeoa.av .s~0.~v Acmo.~v n~o.~ man.“ om~.~ omm.m coo.n onn.n 0H NH ca n“ «a on Aooo.o0 nooo.o0 “coo.~v 004.3 ooa.o u--- u..- u.-- cos.~ a H N Anqs.ov Aoos.av Amoa.~0 Aomn.~v Aomo.~0 Aoom.~v ooa.~ “as." can.” oom.n oao.. aao.n n o a «a an no ---- xono.ov Aona.o. “as".av Anna.nv Aooa.uv .... Ono.” emu." nae.“ “on.o ~a~.¢ a N a n ea lano.~v Aano.av Ammn.av AmNH.n0 Ano~.~0 Awom.~v man.~ nam.~ oa~.~ osn.q oos.q Nom.n ma - a“ oN an egg Aa-.aV Aonu.av Aooo.~. Amo~.~0 Aao~.~v namn.~v anx.~ soa.~ non.n cam.n as... ”no.n nu ou on an an con Hosfl coq-aon con-ao~ oo~-ao~ ooan u.ou seas sun-non scan-aaaom annoy asses canaaua>< acousuom guano: no «Aha can 0» usuvuoou< voaaouo .avaau: nu nowuaaouaam no naaouu now noun: gusouu coauaasaom ado: nuoauon noocououuwa uo away ”.0 m4n4 xx 2>4 xxx 7>4 xxx Eastern Region 56 2.548 (0.822) 40 2.605 (0.890) 5 2.300 (0.998) 17 2.512 (0.766) 4.400 (0.000) 22 2.236 (0.801) 141 2>7 x l>7 x at 0.15 level at 0.10 level at 0.05 level Test of Differences Between Mean Population Growth Rates For Groups of Subcounties in Uganda, Grouped According to The Type of Health Services Available and Adjusted for Northern Region 34 3.165 (0.983) 6 3.667 (0.761) 2 4.000 (0.600) 18 3.406 (0.643) 3.544 (0.813) 69 Western Region 29 2.890 (1.387) 31 2.368 (1.336) 4 3.375 (0.928) 6 2.500 (0.392) 7>4 x 1>2 x 208 Figure 6.2 Relationship Between Health Services and Population Growth in Uganda Adjusted for Migration and Undercounting Rate of Population Growth 52.[ 4 ll/z"“~\_ Northern Region 3 1) Buganda Region Eastern Region ' ' ” Total Country \ / on" Western Region 2 .. \ / 1 P 0 (1) Health Facility Type Data from Table 6.2 209 good proxy for two variables which certainly affect population growth: (a) population density and (b) the extent and duration of other socio- economic forces which,over time, affect desired family size. There are at least two socio-economic forces which are important in the case of Uganda: (1) education, particularly of females,20 and (2) participa— tion in the monetary economy, through cash crop agriculture as well as non-agricultural activities. In the Eastern Region both education and economic participation have been prevalent over a longer period of time in the Northern region.21 In addition, the Eastern region has a much higher population density than does the North.22 For these reasons it is possible that the relationship between population growth and health services shown in Figure 6.2 is a manifestation of different periods of the demographic transition within the country. The North, having more recently been drawn into socio-economic development and experienc- ing little population pressure, has increased in rate of population growth. In the East, however, both socio-economic development and population pressure have been at work for some time and it appears that health services may now be contributing not only to a declining death rate but also to a declining trend in fertility, as indicated by crude birth rates.23 These suggestions are presented in schematic form in Figure 6.3, which presents a "classical" demographic transition and the points at which the Northern and Eastern Regions are now to be found. A final test was conducted on the available data in a determina- tion of the relationship between health service availability and the rate of population growth. Gombolola observations were grouped into 210 Amman: 6H voHMfioon uozw mummy axon I‘l/ . _ _ _ . _ _ _ _ _ _ . _ _ _ _ _ _ . _ _ soauwmsoua. Fauna Tull. onu mo amend woumowmsm _ . _ . _ . _- L n.5uuoz can «0 _ swoon woumowwom v$umoonh non moomuom a“ moon 3 cowufimamua a we 0m comma can cause guano moma aw coauamsuua yank aw «comm: mo msowwom nuoummm vow suonuuoz can we soauooon oumaaxouee< onu was soauamowus cannouwoaoo n.0 muswwm 211 counties; again the urban counties were excluded from the analysis. A category of partially urbanized counties was created and this category was analyzed separately.24 There were two primary reasons for conducting the analysis on a county basis. First, the problem of migration bias (short rural— rural movements) is minimized, since many moves do not cross county boundaries. Second, it was possible to develop two additional crude measures of health service availability: (a) health facility beds per thousand population as of 1969, and (b) maternity service beds per thousand population as of 1969. These two measures are continuous in nature, making it possible to focus on the functional relationship between population growth and health service. The hypothesis tested assumes that the rate of population growth is affected by the availability of both curative and maternal health services (represented by measures (a) and (b) above). It was assumed that the functional relationships were linear and could be estimated according to O.L.S. assumptions. Econometrically, the relationships estimated were: x18 = a + blgxlg + e, ‘ a + b20X20 + E, and N H on I X18 = a + b21x19 + b22x20 + r , where X13 = average annual rate of population growth from 1959-1969, X19 = total health service beds per thousand persons in 1969, >4 M II total maternal health service beds per thousand persons in 1969, 212 a, bl9’ b20’ b21, and b22 = estimated parameters, and a = the disturbance term. These relationships were estimated for the sub-sample of partially urbanized counties as well as for all rural counties and the rural counties in each of the four regions. The results are summarized in Table 6.3. The most significant finding of this analysis is that even though (a) not all the estimated parameters are statistically significant and (b) the proportion of variation between the variables explained by the model is generally low (due largely to an incomplete model specifica- tion of the factors affecting population growth), 21 of the 24 health service parameters estimated had a negative sign. This tends to suggest that the relationship between the availability of health services and the rate of population growth may likely be negative. This finding tends to corroborate the tentative findings of the two tests discussed above. Chapggs in the Age Structure Related to Health Service Availability Two tests were conducted to determine the effect of health service availability on the age structure of the population of Uganda. The first test analyzed differences between mean changes in the ratio of "old" persons to "young" persons over the decade 1959 to 1969 in rural gombololas. The results are presented in Table 6.4. The second test analyzed the average annual rate of change in the ratio of old persons to young persons over the decade 1959 to 1969 in rural counties of the country.25 The results are presented in Table 6.5. 12]”3 000.0 ~00.0 000.0 «No.0 ~0n.0 400.0 2.2h 00.N H0.~ 00.H nm.H Nu.~ m~.N 00.N an.~ -.H 00." on.“ 0«.~ 00.0 n¢.~ 00.~ en.“ m~.« 0H.N 030‘ sno.ou 0-.0 050.0: 0n~.0n 000.0 na~.0u 500.0: n0u.0 0H0.0u «00.0: 000.0 0m:.0| 0m0.0n n0~.0 «~0.0n 050.0: «00.0 «00.0: I“ 000.0 ”05.0 000.0 000.0 «00.0 0h..0 000.0 «00.0 000.0 000.0 000.0 000.0 H00.0 000.0 ax~.0 000.0 000.0 000.0 «00.0 000.0 000.0 0N0.0 000.0 «00.0 000.0 ~0~.0 0~0.0 000.0 000.0 000.0 N 00:00: 00 000w>u0m guano: anon-00: va- 0>wu¢u=u no auauqnnawa>< 0:0 0:0 0 0~00° guano -0000000 nouns .000 “~0~.~0 «550.00 an0o.nv A000.~0 Awnn.av nsn~.0v a-0.00 “000.00 Ann0.00 Amua.nv A000.00 Amqm.n0 own no .000H ca ago-000 vac-:onu 000 0000 000>000 Ana-0: “onuuuqs «0000 I 00M ~0n.0| H0~.nn 000.0: 0~H.0 0NN.HI -0.Nu «0H.0I Non.0ml ~00.00 000.0: -0.~I 040.~| 0AA 000 000a cw 0000000 00000020 000 0000 00ua>hoa guano: ”0060 I 0a” mn0.0 0N0.0 000.0 000.0 qu.0 000.0 050.0 «no.0 000.0 ”no.0 «no.0 «00.0 us 00:00 Ao00.00 a0~0.~0 annn.00 Aa00.~0 a~00.00 1000.00 Annn.00 “000.00 an0~.00 An00.00 Anan.00 Aa00.00 000 .0 -«uuauam 00000 .000 50H.0 0M0.HI 000.0: 0NN.OI ~0n.0l www.0n ~a0.0 00~.nu 004.0: ~0n.0| 000.0: 0H0.0I 0H0 0000.0v 0000.0v 0000.0v 000.0 0000.0v 500.0 0000.0v 0000.0v 0000.0v 0000.0v nooo.ov «00.0 0000.0v 0000.0v 0000.0v 0000.0v 0000.0v 000.0 0 00 00:00 luuuuuum noun» .vum Am00.00 Aa0m.00 “000.00 “000.00 “Hu0.00 An~0.~0 .mhn.00 ANHN.«0 Anon.00 Rana.00 Aoan.00 Aqna.~0 Anae.00 Aan~.~0 Am0~.00 An00.00 Ao~0.00 Annn.av 0 «o £03000 saga-asaou 0003000 mazano«uoaou 0;» mo namaaac< couna0pwom .m.q.0 uo nuanmum 0.0 ugn< .ua>000 000.0: 00 ma0utouu< 0003000 000000 00 00000300030 0003000 0000003000 00< :« 0000000 000: 00 00000000000 «0 0008 0.0 ugm00003u 00 000H000H00>< 0:0 000 n000.ou 0000.0 HOHo.o n~00.0 0000.0 0000.0 «H00.01 0000.0: 0000.0 n000.0| 0000.0 nH~0.0. 00A .000H 00 0000000 00000000 000 0000 000>000 00H000 H0000005 H0000 I oux 000 000.0 NNH.0 0Hm.0 000.0 oHH.0 000.0 000.0 «00.0 000.0 000.0 H~o.0 nHH.0 0H0 00 .0000 00000 .000 .000H 00 0000000 00000000 000 000A 0000>000 :0H000 H0000 I 00n00.00 An~H0.0V 00000.00 A0m00.0v H0000.00 H0000.00 Hn~00.00 00000.00 Ah~0o.00 H0000.0v A-00.00 00000.00 0H0 nmoo.c 0000.0- 0000.0: 0000.0- ~000.0- n0m0.0- «Hoo.0 0000.00 «000.0: 0000.0! HH00.0I 00H0.0t 0H0 000.0 000.0 ~00.0 n000.0v H00.0 0N0.0 H00.0 000.0 ~00.0 0000.0V 000.0 nmm.0 nn0.0 0H0.0 n000.0v 0000.0v Hoo.0 000.0 0 00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 00000.00 Q .0000 00000..000 5000.0! ~mH0.0 0H0.0 o~u0.0 0000.0 00~0.0 0000.0- 00H0.0 00H0.0 0000.0 0000.0 0000.0 0000.0 noH0.0 00H0.0 nnN0.0 HOH0.0 0n~0.0 I ‘ 003003000 00< 0:0 00 0000000 0000000 000000000H00 0:0 00 0000000< 0000000000 .0.0.0 00 00H3000 0.0 mJn .muuomom ovmua Hmscs< «whom» m20fium> .uomuum0< Hmowumfiumum .ucmscHo>ou mvcmw: "mmouaom mofium< ummm .oosmwfl>m wafiumuoomuuoo m mmvfl>oum mzcmx Scum meow coauosvoum HmwuumavdH .HHmEm mum> mum? mafia umsu ou uoflum muaooem ozu umAu ummwwom vases vaouu mnu swoonuam .00oa ou uowum mmcmx aoum manomaw Hmoausmomaumna 0am mayo «0 uaouxo wnu ou mm maomfifim>mcs mum mama .mzaom Boom mums wouuomsa m5mua one 00 Add haamsuufi> 0.0a nw.~ 00.0 om.¢~ o.h¢NH H.5H mn.~ «0.0 Hw.0m 0.0NHH 0.0a oq.m 0H.0 mm.qm N.wwHH N.0 mH.N mu.a 05.0H N.0Nm 0.0 mm.H H0.o oo.NH m.~m0 mm.H AHV <2 «N.OH c.000 H0.H AHV <2 m0.H 0.0HH wuuoaaH SmuH woumamm 0cm mono Hmuou mo mum mEouH woumamm 0am wane cmowum< ummm coauuoaoum muuooEH dance 00 mum muuoasH woumaom 0cm mwsun cowuuoaoum .mnm soaaawz mofiuuasoo amowum< ummm scum meouH woumamm 0am mwsum mo muuonaH .mnm aowaawz moaum< ummm mvflmuso aoum maouH woumamm was mwdun mo muuoaaH .m:m coaaaaz muuomaH Hmuoa uaoaafiavm 0cm mowamaam Hmofiwoz woumamm 0cm mmaun mo aOfiumuHoasH m.m0cmwa ca mvcmue 0.0 mqm¢H Amy :0 "mmuoz m0mH m0aa 00mH ¢0mH O0ma 00¢H 0qma umow vmuooaom 220 nearly tripled.27 Other items comprise larger proportions of total imports, but health service imports comprise a rapidly growing pro- portion of the total. Finally, from the perspective of the long run economic integration of East Africa,the increase in the pr0portion of drugs and related imports from Kenya is welcome; over the 1960's, this proportion has grown from 5.1 percent to 20.2 percent. Imports of Building Materials and Other Capital Goods for Health Facilities Not only does the health service system require the importation of a large proportion of drugs, medical supplies and equipment, but it also imports a substantial proportion of the items for transportation and for the construction of health facilities (Table 6.7). As may be recalled (Chapter One, Figure 1.1), expenditures by the central government on health have varied over the decade, primarily as a result of capital expenditure lumpiness. Since 1964, however, health facilities construction activity has increased more than five-fold and the proportion of total imports comprised by health facility construct- ion materials has increased from 0.2 percent to 1.0 percent. The proportion of health facility construction imports within total construction material imports is illustrated in Table 6.8. Even during periods of relatively minor health facility construction as in 1966, a minimum of 6 percent of all imported construction materials went into health facilities; during periods of major health facilitv construction (1969/1970), as much as 15 percent of imported construction materials were so used. Aside from construction materials, the other most important imported capital items in health services are vehicles (ambulances, etc.). 221 .uouuan amm.mm_auuwuuuoau on» o» moauuum mousuwuaonxo huwawumm guano: Huuou «a non on ommu 0am I uouumw .JMMZNM .Hmoauuuoam ago on monsoon nousuwvcoaxo huaawuum guano: Hanan 00 Rod I Amy ammo "coauqsammm ucmuowmfip a no woman some .mcoquSuam N mafiamxm o3 mowonpnm m>fiumuu nasaaw you .uo>msom .muomavaw cofiuusuumcoo any 00 nouomm comm cu monsoon mounuavcmdxm muflafiuaw guano; Huuou wo coauuoqoum ums3 nooaoas ma uH .uouoom wcwucflmm 0am HMUfiuuouHm.wcfipE=Hm osu now No.00 vow noboom cofiuoauumnoo 0cm wafipfluam ozu now NN.0H was ouswaw may .0mumasoamo ma Auv>o lauauv moamm Hmuou mo mucmmmunwu mamwuoums vouuomsa umnu aofiuuoaoun «Lu A000H zumzcmn .onnouam .ucosaoam>o0 xuacassoo 0am maaaswam mo shaman“: .coama>wn mowumaumum Amwma .aowuosuumcoo_mmm mcfipfiwsm "pawuusvoum Hmfluumsch.Mm.Nmmwmwv .unonou mnu 00 A00 HH was any HH magma xavcoaq< :H .huumspcw =o«uoauumcoo 05m wawwflwsn mg» no >m>u=w m ovwa unusaum>ow cmvcmwp onu <00H cm ANV .muuonxouou 0cm muuonsa omfivcmnouoa vocamuou Haney mmvsaoau Aav "muuoz 0H.H «5.0 0.0H «.0 H.- 0.0a 0.00 0.~oo huwaaomm Sodom: ouauwpcmaxm woumswumm cowuuoaoum pmumsfiumm we ouauwvamaxm aowuuoqoum woumsaumw cowuuauumaou huaafiomm nuawmm pow msmun wouuomeH mo huauamso woumEHumm 5.0 wands 0 222 .oamasum .uamasuo>oo ovum»: .Hmuoaoo uoufics< Sofia mHHm co m00a|m00a mums» you mucmscuo>oo sowumuumfiaaa0< uofiuuman mo musosmuaum Hmaoamcam wouava< “mummm msowum> .uomuumn< Hmoaumwuoum .ucoaauo>ow avauws "moouaow Hmaowua00< .mamwuouma wouuomaa wow wow: mousuapaoaxo coauoauumsoo zuwawumw guano: mo acquuoaoum woumaaumm «an .Aan no Ammv cesaou nonufio ca ousmfim onu saws .sofiuozuumaoo kuaawomm guano: so maoammwa 0cm unmasuo>om mo musufipamaxm woumsaumo .ANV sasaoo ca mouawfiw osu magmamfiuasa An woumanoaoo mum mmuawwm one .mcowunasmmm nuon so“: o>wa ou waaaafia an H .cowumswxoumaw umuam a m< .maamuocmm coauosuumdou ca pump was conu mamfiuouma 00 was usouowmwv adamwuswumosm a wufisvou uo: moop cowuosuumaoo mowuwafiumu guano: Any 0am 000H|000H mvmoov onu um>o moauaawomm anamon mo acauoauumsoo can aw vow: mamauouma mo muwaanmawm>m map a“ owqmno Hmfiuswumnsm a soon was muonu Amv "maowumasuamo o>onm mnu ca voaflmaa mum mdoaamESmmm Hmaowuwpvm 039 A.aseouv 5.0 maps? 0» mauoz A3 A3 223 .ONma weapon mouauwvamaxo ucoaauo>om voumaaumo Scum vmumaauao mum; mmuswam ohm” .Aav 0am Adv nuouuasammw mo scammauwav a you .5.0 manna .N ouo: mom .2 a .q-HH magma .Nmoa .onpouam .uouswum uamacuo>oo .oa\mamfi--waamfl .amam sauemoaw>mp ummruw>fim qunH .HHH swam m.m0cmw0 .mvamws wo uwaosmom ona aouw cmxmu ma puma o.aa m.oH n~.mfl. ma.a n.~o m.mH n.m m.qa ~.a m.ea o.m o.e a.“ m.e a.m~ s.» ~.n a.» m.n «.ma m.m «.n m.n «.n 0.00 “p0 Amy «A90 aoauaaamm< Nagy soauaasom< .mnm nowafia: manomaH wauoumz .mgm ao«HH«Z.meauoumx Hmamauoumz aowuosuumaou Hmuoa aowuosuumdoo huuaaoah aoHuusuumcoo mo ma muuoasH nuawom mo muuoasH woumswumm mo muuooau soauosuumcoo xufiaflumm . nuamom aowuuomoum .mowuwawomm suamum mo coauusuumaoo usu aw 00m: nauwuouaz soauusuumaou vouuoasH mo nowuuomoum voume«uum 0.0 mqmdfi Amy va aflv “mmuoz 050a 000H 000a n0ma 000d you» 224 Financial accounts of both government and missions indicate that less than 500,000 shillings are being spent each year on new vehicles. In recent years, this may be true; however, figures on new vehicle registrations and data found in Galea's survey of Uganda's health facilities in 1966 and 1967 indicate that the health service industry spent, at a minimum, 700,000 shillings in 1965 and as much as 1,700,000 shillings in 1964 and 1967 on transport costs.28 If these estimates of imported capital goods are added to construction materials imports, the average annual proportion of total imports comprised by capital inputs for health services during the 1960's was approximately 0.7 percent. Non-Trade and Capital Account Considerations Although it is difficult to quantify the size of all public and private non-trade and capital balance payment flows attributable to health services, it is useful to review the sources of such flows, especially because Uganda's balance of payments position has deter- iorated markedly in the last several years.29 Important flows at- tributable to the health sector are (a) private remittances of earnings by non-Ugandans; 30 (b) public transfers, (e.g., membership payments and other fees to international health organizations); and (c) net capital flows from abroad, public and private, for health facility construction. Table 6.9 presents data showing the approximate effect on these factors on the balance of payments position of the country. The overall effect of health-related financial transactions on the balance of payments position in 1969 was positive, amounting to a net inflow of approximately 9 million shs. Closer analysis however, 225 TABLE 6.9 Health-Related Financial Transactions Affecting Uganda's Balance of Payments Position in 1969. Transfers mill. shs. mil]. shs. (1) (2) (3) (4) (5) Notes (1) (2) (3) Private Income Transfers (a) Non-African(2) (3) -0.32 (1) (b) Non-Ugandan African -l.59 -1.9l Public Transfers (a) to international health related organizations 4 -0.35 Capital Account, Long Term (a) Private-Mission Related<5> (6) +2.70 (b) Public-Health Service Construction (1) inflow of loans and grants +9.81 (ii) outflow of interest and repayment of loans -1.34 +11.l7 Net Effect on Balance of Payments +8.91 Total Balance of Payments of Uganda 1969(7) +70.00 a (-) sign indicates a financial outflow from the country, whereas a (+) sign indicates an inflow. It is estimated that 5% of total wages paid to non-African personnel in 1969 were remitted abroad. It is estimated that 50% of total wages paid to Non—Ugandan Africans were remitted abroad. Total wages paid to Non-Ugandan Africans were estimated on the basis of information found in The Republic of Uganda, Enumeration g§_Employees, (1968 and 1969), Statistics Division, Ministry of Planning and Economic Development on employment in health services. (a) Estimated Total Employment in Health Services 1969 17,700 226 Notes to Table 6.9 (contd.) (4) (S) (6) (7) (b) (c) (d) (e) (f) Estimated proportion of Non-Ugandan Africans employed in Health Services 1969. Based on Health and Education or Place of Birth 5.6% Estimated number of Non-Ugandan Africans employed in Health Services 1969 1,000 Estimated proportion of Non-Ugandan Africans employed in Government Health Services. Based on employment data between Private and Public employment in Health Services 802 (1) Average annual wage for Africans in Government Health Services in 1969: 3505 shs. (11) Average annual wage for Africans in Private Health Services in 1969: 1867 shs. Estimated total wages paid to Non-Ugandan African'employees. (1) Government Health Services 2,804,000 shs. (ii) Private Health Services 373,400 shs. Total Wages 3,177,400 shs. The Republic of Uganda, Medical Services Statistical Records 1968/1969, Ministry of Health, Entebbe, Appendix IV, pp. 55-56. Financial records of Catholic and Protestant Mission Health Facilities, provided by the respective Medical Bureaus for 1968/1969. The Republic of Uganda, The Public Accounts of the Republic of Uganda for the Year Ended 30th June, 1969, Government Printer, Entebbe, 1970, pp. 32-33, 81-82. The Republic of Uganda, Background £g_the Budget, 1970-71, Statistics Division, Ministry of Planning and Economic Develop- ment, June 1970, Table 8.1, p. 50. 227 reveals that the position is not as positive as it appears. Estimations of the net remittance abroad of employees' earnings place the figure at a minimum of 2 million shs. Available evidence, however, indicates that the amount remitted abroad is declining for at least 2 reasons: (a) the rapid decline in the number of European expatriates, who are being replaced by trained Ugandans or less expensive non-EurOpean ex- patriates, and (b) the reduction in the number of non-Ugandan Africans working in the country. In addition, the period of large capital in- flows from abroad for mission health facility construction appears to be at an end, as a result of governmental decisions (3) to expand the government health service system and (b) to discourage further expan— sion of mission health services. Although recurrent external support is thus likely to decline, support from abroad to assist in the opera- tions of present facilities will undoubtedly continue; this support of Operating costs amounts to approximately 1.2 million shs. at the present time. The net effect, from a balance of payments perspective, is that the positive inflow related to health services should decline in the future to a level of approximately 0.75 million shs. In the public capital account, a positive inflow of approximately 8.5 million shs. was recorded in 1969. This trend is unlikely to continue however, due to (a) decline in health facility construction 31 activities; (b) a decline in the use of foreign financing methods, particularly contractor finance, in future health services expansion 32 and (c) an increase in the level of interest and loan activities; repayment commitments, particularly as a result of the large rural hospital project of the second development plan.33 Interest and loan 228 repayment commitments have risen, due to the increase in outstanding contractor-financed projects, to a minimum outflow of 7.0 million shs. in 1970 through 1972, and 6.6 million through 1975. This substantial repayment commitment will tend to exacerbate Uganda's balance of pay- ments position during the early and middle 1970's. Health Services Effect on Employment In 1969, approximately 18,000 persons were employed in the health services industry -- approximately six percent of the total number employed in the wage sector of the economy.34 Since employment in this sector nearly doubled over the decade 1959-1969, while total employment in the country has been relatively stagnant, and since health service facilities have been expanded in recent times, it is useful to analyze the present and future impact of the health service system on employment and the related equity objectives of the process of economic deve10pment.35 The analysis will focus on three issues: (a) a comparison of the health service industry with other industries, in terms of output and employment relationships; (b) the important secondary employment effects of the health services industry, partic~ ularly in the case of the construction industry; and (c) the distribu- tion of employment in health services throughout the country. An Interindustry Analysis of the Relationship Between Output and Employment The analysis here focuses on the position of the health service industry relative to other sectors of the Ugandan economy, in terms of the relationship between output and employment. Two methodologies have been employed in determining the relationship for each industry: 229 (a) the incremental output employment ratio (IOER) and (b) estimation of the elasticity of employment with respect to output. In the IOER method, the percentage change in value added for each industry over the period 1963-1969 is divided by the percentage change in the number employed. The results of these calculations are presented in Table 6 6.10.3 In addition to the IOER calculations, a simple labor demand model has been developed in order to derive estimates of the elasticity 37 These elasticity estimates of employment with respect to output. are also presented in Table 6.10. Results of the first analysis indicate that the health service industry has a lower IOER than the economy as a whole, which implies 38 3 larger than average employment impact during the 1960's as a re- sult of increases in value added. Services in general, in fact, have values below the economy average, indicating a larger employment impact in those industries due to increases in value added. Employment elas— ticity estimates tend to corroborate the finding, for the service industries' employment elasticities tend to be larger than most other industries'. 39 Secondary Employment Effects of the Health Services Industry Although an input-output transaction matrix does not exist for the Ugandan economy as yet,"0 it is clear that health services demand goods and services from a number of other industries in the country. In doing so, the health service sector expands the derived demand for labor in those other sectors. Some of the most important linkages between the health services industry and other industries are summarized for 1968/69 in Table 6.11; included are estimates of the secondary employment 230 «00.0 00.0 «ee0H.H «ee0H.H HN.H 0H.H heocoom mufiucm «««00.0 «se00.H IIII IIII ~0.0 III mow>uom nuamom e*«qm.0 aee~m.0 IIII IIII 00.0 III mofi>uom soflumuovm «a0H.H «««HH.H «««0H.H «««0~.H H0.0 00.0 moofi>uom .omfiz IIII IIII «eewm.a «ee00.a III 00.0 uaosouo>ou Hmooa woo.o «00.0 «a00.o ««00.o em.m oa.m cofiumoweassoo 0am uuodmemua 0H.0I 00.0I «*00.H «00.H 00.0 HN.H mouoesoo «««q~.H «««0~.H «ee0n.0 «e«H0.0 00.0 0H.H eowuosuumcoo «««~0.H «eee0.a «e0m.0 «ee0c.0 00.0 00.0 waquauomwsqmz .omwz 00.0 00.0 «00.0 «a00.0 00.0 HH.H .0oum 0000 no ouauommsamz 00.0I 00.0I «ee0~.HI «eea0.HI 00.0 00.0 wdahhumso 0cm wows“: ma.o 0~.o mm.o om.o IIII IIII wagons: .meaemam .muumouom mH.o mH.o ««0~.o «som.o 0w.~ 00.N ousuasuaum< sees 00<> oe<> oe<> Hausa; 0000. .000M ammmmmmm woq mmocfiq mod uaamso ou uommmmm mansoa manson nufi3 newshoamsm 000mmmm0a mo suaofiummam emumaaumm uouoom HmfiuumsvaH ou mcavuooo< .susowo you uaouso ou uuommom £003 unmemoaasm mo muwoaummam 0am AmmoHv owumm uaoahoaasm uaauso HousmamuuaH 0H.0 mum¢a 231 n.0ma .m .n00H .mmmum 0ufimum>wa0 soumoauum .0mmum0 3mz .maaaoaou mo m>wuommmumm umsommwz .umumfin .xoonomw aoumv muouomm mmauwaau: madman Haaoauusuumaou .wmz .maflcaz m.o Housmaa mmumum causes muouomm HmuzuaaoauwmIaoc HH< 0.0 m0I000H mumwaam muouumm HmuauaaowuwmIooa ~04 H.m m0I000H womaom muouomm HmuauaoofiuwmIaoa 00¢ 0.0 00I000H mfi>mamowsw muouomm HausuaaofiuwmIeoa HH< 0.~ m0I000H mmawmmaaazm muouomm amusuaaoauwmIaoa HH< 0.0 m0I000H HmmHmH muouomm HmuauazofiuwmInoc HH< 0.0 00Im00H oofim ouumam muouomm Hmuauazoauwmlooa Had 0.0 m0Im00H omamn muouowm Hae H.H moumoaa «gamma mmmmmmm. mmmm. mummy “mummmm. .Boamn 0muammmum m0 mmauucsoo umnuo mom m.mm0H mo maaamm 4 .mmfiuuosoo 0m00Hm>m0 mmmH umnuo you m.mm0H 0mumaaumm £003 0mummaou 30H muwav m0 mvsmws now aaosm mmoH msH A00 .0000 HHHQ< .uomamon>m0 hufiaasaoo 0mm wawaqmam mo muumaafiz .sowmfi>«0 mowumfiumum .N00HI000H .qumwm:mm.0soaoom mnu.Mm.nu30u0 Hmmm may .ucmsdum>oo mvamwo "000a mash .uamaaoam>m0 oHaoaoom 0am mafiaamam mo muumwcfiz .cowma>ao mowumaumum .HNI000H .mmmmmm.mmm.mm.vasoummwmm .mvamwa we madnoemm "mummm 0muomamm .muomuuma< Hmoaumaumum .mvamws mo meansmmm "mmouzom A00 .maoaumsum ummaaa men you 0m00m maam> 0am unmahonmam mo mmaam> same man um vmumasoamo mma usmuoo ou uomammu £003 uomShquam mo muwuwummam 0mumawumm mna Amy .000HIH00H vowumq may now mumv mawmuaou mmaumm men .muommumnu ”H00H aw 0muumum .00<> .mmwumm 0m00m msam> magma uamumaoo 000a may .000HI000H vowumn um>o mump 0mm: .00<> .mmuumm 0m00m m=Hm> mofiua uomumaoo 000a mnu wafim: mmfiufiofiummam vmumafiumm Amy .uamshoanam 0am unmuao ammsumn awnmcoaumamu umumamuma woumaaumm msu mo moamuamaawwm mnu mumowvnw .noaumsvm ummswfl msu aoum 0mumasoamo mmumsaumm huaowummam msu mvammn wowummnmm coaumuo: unmofimwswfiw mna Ho>oa Ho.o we» as “smegmaamfim ««« Hu>ma mo.o use an uamuwmflawam «« Ho>wa oH.o one an uquoamaamfim « Aav 00.0 manna so mmuoz 232 effects of the health service industry. In l968/69,approximately 3,800 persons were employed in other industries as a secondary effect of the demands for goods and services by the health services industry. This constituted approximately 1.3 percent of the average total employment in the country. The most important secondary employment impact was in the construction industry, where at least 3,100 additional jobs existed as a result of health facility construction activities. That number comprised approximately 7.5 percent of the total number of persons employed in construction in June 1969. In addition, estimates based on Table 6.11 indicate that for every additional 19,200 shs. spent by the health service system for goods and services in other industries, the demand for labor in those industries is increased by one employee.41 The Geographical Distribution of Health Service Employment in Rural Areas Table 6.12 presents information, by district, concerning the pro- portion of total wage employment comprised by health services employ- ment. Of immediate importance is the fact that people are employed in the health services industry in every district and town with one ex- ception.42 Thus, while a completely equitable distribution of health services has by no means been accomplished, 43 there are at least some services available in every district and major town. Although employment in the health services industry does exist in all areas of the country, it comprises a significantly larger pro- portion of total employment in the towns than in the districts - 9.0 percent as opposed to 4.3 percent. Even adjusting the figures to 233 N m H 000.000 0000 000 <2 <2 000 0000 00 000 huum=0a0 mo0hmwm nu0mmm mo uquEH unmaho0nsm mumvaoomm 0muma0umm unmaho0msm mwmz 0muoa 00 m0 ucm8000050 mumvsoomm 0muma0umm :00uuoaoum uamsho0nsm mwmz 0muoa 0muma0umm 00.00 Ahm0 .om0z <2 00.0 m=o0umo0a=saoo «z «z uuoamamua 0000.0 00.00 Anvmoumsaoo 0000.0 00.00 A00:00uoshumaoo 0000.0 00.0 A0050003030 0000.0 00.0 A0vmu=u0so0uw< umho0nam umm .msm a00000a Nmmmmmmw. .w:m s00000a A0vhuum=0a0 000mm mm000asm mu0>umm :u0mmm umm mun0momm scum mun0mom0 0Hum50¢0 00\0000 s0 uom080 ucmsho0msm 0Hm0coom0 msu mo mmuma0umm 0am huunsva0 mo0>hm0 £u0mmm mnu mo mmwmxa00 huumawaHIumuQH 00.0 m00 g H c: 0 COO O 3 f: u on: 0 1:0 no u H Q) «I: <1: U H u a H H N In 'H (USP-i a: ..fi a UV <0 0 H Government ional Refe District H Centers a te a Di Dis erni d Posts Natl. Immunization Team District Health tor 1 Health Team Ankole P.P.P. Milita Health Services Private Missi ust Mission Matcrni a ter itals Hos itals Dispensary Units ssion Dis nsa Mission Sub-Dis Mission Maternit Mission Aid Post ...(5313111. fists r Urban Private Physicians Ofllres d - maybe * - the nervice is offered by Protestant units only I - health education is conducted within the defined area APPENDIX B APPENDIX B Administrative Relationships in Uganda's Health Service System Until very recently, at least three types of administrative relationships were important to understanding the operations of the health service delivery system and its prospects for future development: (1) the relationship between the several government ministries which have jurisdiction over various aspects of the health care system; (2) the relationship between the central government and various local units of government; and (3) the relationships between government at all levels and the private sector of the health service system, which includes mission medical bureaus, large firms, and private physicians and pharmacies. The administrative relationships of importance are found in three areas of policy: (a) medical standards, as related to personnel, care, and operating methods; (b) financial support policy; and (c) deve10p- mental policies related to future expansion of the service delivery system. Administrative relationships relative to these policy matters have their genesis in historical, political events, in the development of the role of governmental organizations, particularly since Independ- ence, and in the recurring financial problems faced by private health organizations. Central Government, Inter-Ministerial Relationships Until the recently announced Plan III, two ministries in the central government had to correlate their activities relative to health services. The Ministry of Health had (and continues to have) major responsibility for establishing broad medical policy in such areas as minimum qualifica- tions for several types of medical personnel, the development of pharma— cological policy, etc. The Ministry develops the medical, administrative, and financial policies related to the operation of all government hospitals. It is responsible for all medical education in the country, with the exception of the education of medical doctors, which is administered by the University; in conjunction with this responsibility, there are Ministry— run schools attached to the larger hospitals in the country to train 1 The information presented in thislkppendix was gathered in an in- formal way and there is no one source which can be cited for authenticity of the information presented. The author is responsible for any mistakes which may exist in facts or interpretation. 2 At the end of thiSJAppendix the implications of the recently an- nounced administrative policy changes described in Uganda's Plan III _p,.gi£., pp. 306 and 307, paragraphs 17.20 - 17.24, which strengthen the powers of the central government's Ministry of Health vis a vis those of local governments, are examined. 270 271 professional nurses and midwives, auxiliary level nurses and midwives, medical assistants, professional and auxiliary radiography and laboratory personnel, and public health personnel. The Ministry establishes medical policy for all government rural health facilities, through its appointed district medical officers. It also deve10ps the broad national public health strategy and, where appropriate, develops national legislation re— lated to health matters for consideration by Parliament. Finally, the Ministry of Health is involved in planning for the improvement of the health of the people of Uganda. In this capacity, it has assisted in planning for the expansion of health facilities, as well as expansion of the supply of other necessary inputs such as personnel, drugs, and equip- ment. This latter responsibility is undertaken in cooperation with the Ministry of Planning and Economic Development. The other Ministry, the Ministry of Regional Administrations, has had control over the activities of each local government's department of health. This Ministry approves pr0posed recurrent and capital budgets for each district and city and determines the taxing capacity of each. This Ministry also determines the criteria for dispersing central govern— ment funds to cover recurrent and capital budgetary requirements of the local governments. It has been essential that the Ministry of Health work in close cooperation with the Ministry of Regional Administrations in order to attain the medical objectives which the Ministry of Health may have developed for the country; this has been especially important given the rapid expansion of hospitals and rural health centers during the second five year development planning period, 1966-1971. Decisions about timing of new construction, facility type, financing, and location of facilities have been made initially at the local level, but are reviewed by the Ministry of Regional Administrations. The policies of this Ministry, therefore, have been vital to the way in which the health service system operates, particularly in rural areas where approximately 90-952 of the population live. The individual directly responsible for the daily operation and expenditures of health facilities in all districts, however, remains the district medical officer (DMD), who is a direct appointee of the Ministry of Health. ‘Many trained medical staff (medical assistants and other more highly trained auxilliary staff) in the districts are seconded to the districts from the Ministry of Health. A large percentage of all drugs and equipment (approximately 952) used in the districts have been purchased from the Ministry of Health. Patients who cannot be adequately cared for in local rural health facilities are transferred to the district government hospital, which is run by the Ministry of Health. The Operation of the health service system, thus, has been subject to the coordination between the two Ministries. Central and Urban Government Relationships The relationship between the largest municipalities (Kampala, Jinja, Mbale and Masaka) and the Ministry of Health is not as strong as the 272 Ministry's relations with the districts, as the municipalities hire their own personnel and coordinate their activities with the Ministry of Health primarily on such items as national.immunization campaigns. In addition, a large percentage of the municipalities' purchases of drugs and equipment are made from local suppliers, rather than from the Ministry of Health's central stores. The municipalities' finances are reviewed by the Ministry of Regional Administrations, but because they are largely able to finance their recurrent and capital expenditures on their own, their activities are not subject to the same degree of scrutiny as are the districts. Urban authorities and smaller towns remain tied more directly to the central government because of their inability to finance their own ser- vices completely. The Ministry of Regional Administrations is involved in their budgetary matters and in arranging the financing of their capital projects. The Ministry of Health is also involved with the urban author- ities and towns, because the disctrict medical officer is often the town's medical officer of health as well. The towns' most direct link to the Ministry of Health lies in the administration of public health policy, inasmuch as nearly all health services provided by the smaller towns are related to public health. Central Government and Private Health Services The private health sector interacts with the government in several ways. Private physicians are related through licensing which is adminis- tered by the Ministry of Health. At present, there are about 250 doctors in private practice in the country, nearly all of whom are located in the 10 largest cities and towns. Chemists are also licensed by the Ministry of Health, and must receive a special license in order to dis- pense chemical formulas appearing on the government's list of poisons. This licensing is the only direct link between the government and chemists. However, the Ministry of Health purchases a large percentage of its drugs, equipment and stores from local chemists and other vendors; this relation- ship is strictly contractual and is usually negotiated on an annual basis. The labor laws of Uganda state that firms which employ a certain number of workers must provide a minimum set of health services by qualified medical personnel. If a firm employs more than 1,000 persons, it must maintain a full range of hospital facilities. Such private hospitals are operated by three industrial firms at the present time: Kilembe mines, Mahdvani and Co. Ltd., and the Uganda Sugar Factory Ltd. These hospitals and the less complete curative services maintained by smaller firms are monitored by the district medical officer and must conform to the medical standards of the Ministry of Health. Also, industrial health and hygiene is monitored by the central government's Ministry of Labor and its personnel investigate plant safety standards and monitor general standards of hygiene. The government and mission health services have had a long relation- ship. Mission hospitals received central government financial assistance at least as early as 1932, and this support continues to the present time. Today the government provides grants to approve nurse and midwifery training schools operated by mission hospitals. In addition, the Ministry 273 of Health provides recurrent grants to the mission hospitals on the basis of the number of doctors and registered nurses on the staff. Mission health facilities, however, have been experiencing increased financial difficulties. The mission medical bureaus contend that govern— ment grants are much too small, but the amount of the grants likely reflects the government's general set of objectives, which does not lend much support to mission health facilities. The district administra— tion governments have also provided some financial support to mission health facilities in the past. Besides the financial relationship between the government and mission health facilities, the government controls mission activities by a form of accreditation. Each mission hospital must abide by medical standards established by the Ministry of Health in order to obtain the financial support provided by the central government. Every doctor working in a mission facility must also be licensed by the Ministry of Health. Central Government and Makerere Medical School Makerere Medical School has primary jurisdiction over medical education of doctors in Uganda. The School is related to the central government in several important ways, however. Since 1968, medical students have been required to agree formally to serve in government hospitals for two years after completing school. In addition, the Medical school must cooperate with the Ministry of Health in using Mulago Hospital, the largest hospital in Uganda, in the training of its students. Because it is a part of the national university, the Medical school receives a large proportion of its recurrent operating funds from the central government's Ministry of Education. The Medical School also receives funds from the WOrld Health Organization and other international foundations, but the central government, through the Ministry of Health reviews the appropriateness of the particular research or teaching pro- ject to be funded by international organizations. New methods of delivering medical care services have been developed or adapted by various departments of the Medical School, but the rate of innovation diffusion to date has been slow. Recently however, there appears to be more receptivity to change and cooperation between the Medical School and the Ministry of Health. The cause for optimism is related to the government's willingness to help fund the following pro- grams: (a) the Kasangati Teaching Health Center (previously funded by the W. Mengo District Administration); (b) the Ankole Preschool Pro- tection Program (previously funded by the Oxfam Foundation and the Ankole District Administration); (c) the Mbbile Maternal and Child Health Clinic program in the rural areas near Kampala (previously supported by the W. Mengo District Administration); and (d) the rural 1:331:11 service system improvement project operated by medical students and professors in the W. Mengo District (previously supported by the District). Plan III's Announced Administrative Policy Change for Health Services In Uganda's Plan III, 1971/72 - 1975/76, the government announced a 274 major health administration policy change. The major facet of this change strengthens the powers of the central government Ministry of Health vis a vis and local authorities, primarily the district administrations. The essence of the policy change is as follows: "The most difficult problems of coordination have arisen in connection with the health activities of local authorities. In the past, each local authority developed its health service with- out any regard to the activities of other local authorities. Also there has been very inadequate coordination between the activities of the local authorities, on the one hand, and those of the Government, on the other. This has created problems especially in relating to the provision (by Government) of staff for local authorities' health establishments. With a view of alleviation these problems,iit has been decided to transfer the responsibility for setting up and administering;health centres and all other rural medical units completely away from district administrations to the Ministry of Health. The district admin- istrations will thus cease to have any direct responsibilities in the field of health. It is anticipated that the net effect on government recurrent expenditure of the take—over of rural medical units will be neutral, as block grants to local author- ities will no longer cover the Operation of these units". (Paragraph 16.22, pp. 306 & 307) It is clear, thus, that a major shift in the organization of Uganda's health service system has been announced. Such a change has precedent in Uganda, for in 1966, the central government established its control in the field of education services. The announced shift in health administration policy was not as inclusive as was the case in education, however, since mission health facilities can still maintain a considerable degree of financial, administrative and medical control over their operations. The primary change is in the administration and financial control of rural health services; such a change may be viewed as a step by the central government to further support the interest of people living in rural areas and is an important institutional and administrative development. APPENDIX C APPENDIX C A Further Elaboration of the Variables, Methods, and Procedures used in Chapter Five The Data, Methods and Procedures Used to Determine the Value of the Elements of the Input Vector and Technological Matrix_ This section of Appendix C provides supplementary information on data sources, as well as methods and procedures used to estimate the value of (a) the service-providing and service-demanding inputs and (b) the elements of the technological matrices, including the elements of the diagonal submatrices. Values for the second part of the input vector - the service de- manders by major disease classification for government and mission hospitals - were developed from data appearing in Ministry of Health annual statistical records.1 The distribution of new cases by major disease classification for inpatient and outpatient cases was obtained directly from that source. Reattendances (persons returning for treatment of a new illness episode during the year) by disease category were determined by an allocation procedure based on analysis of a sample of reattendances at one govern~ ment hospital (Mbarara) and two government rural health facilities (Namwendwa, Busoga District, and Kinoni, Ankole District). The distri- bution of diseases according to major disease categories was derived by adding the new case attendance data to the estimated distribution of reattendances for the hospital sectors. Additional adjustments were necessary for rural health facilities in order to classify new cases in the major disease categories used by hospitals. Sample information was obtained from the daily record books of three rural health facilities (Kinoni, Ankole district; Buikwe, East Mengo district; and Busesa, Busoga districts) on the disease distribu- tion of the "other disease" category used by the rural health facilities. These data, and the above-mentioned information on reattendances, were used in constructing the distribution of diseases treated on an out- patient basis in rural facilities. The disease distribution of rural facility inpatients was also developed by analysis of a sample of inpatient records (Kinoni, Ankole district; Namwendwa, Nsinze and Busesa, Busoga district; and Buikwe, East Mengo district). This was supplemented by an analysis of rural maternity center returns at the same locations in order to incorporate diseases of pregnancy and puerperium and delivery-without-complication into the analysis. n—_ 1 Republic of Uganda, Ministry of Health, Medical Services Statistical Beggggg, lst July 1968 to 30th June 1969, (Entebbe: Government Printer, 1969). 275 276 Several procedures were used to estimate the service-providing element. For each sector, an accounting of the total available person- nel, disaggregated into the eleven categories, was developed. (See Table C.l for an analysis of how specific occupational categories were combined into eleven categories used in this research.) These data were procured from several sources: (a) annual district budgets (for rural facilities in districts which were not visited by the researcher), (b) health facility survey information (for hospitals and rural units in districts visited by the researcher), (c) the publications of the Catholic Medical Bureau and the Protestant Medical Bureau,2 and (d) Ministry of Health pay records. Data for other inputs (beds, drugs, food, etc.) were obtained from budget and expenditure records. For the manpower inputs, two adjustments were made to obtain the value used in the input constraint vector. The first adjustment de- ducted the amount of time engaged in administrative duties. This adjustment is particularly important in the high level medical manpower categories, where a large proportion of time is spent in "running the operation" rather than in direct service provision. Information on this adjustment was obtained from interviews of personnel in each sector of the health service system. (Although it may be possible to reorganize the health service delivery system in such a way as to conserve par- ticularly scarce manpower resources for delivery of health services, the existance of the present organizational constraints was deemed sufficiently important to warrant inclusion in the calculations.) A second adjustment deducted from several manpower categories the amount of time spent in specialized clinics which were not providing direct service to the vector of initial demanders. Such specialized clinics include many preventive services discussed in Chapter Two. Estimates of the amount of time spent in such activities were developed from interview information. It is important to note that the manpower inputs are expressed in man hours. The estimated number of man-hours for each category of worker was developed on the basis of interviews and observation of normal working schedules. The author is aware that not all doctors work the average, and the case of the all-night operation is well known. The calculations are based on estimated averages and it is obvious that the system is elastic enough to attend to "emergency" situations. It was assumed that other service-providing inputs were consumed in the process of delivering services to the vector of initial demanders. Although some of these inputs are consumed in administrative duties and/or the provision of specialized clinics, the amounts were assumed to be minimal (except for drugs and supplies, and in that case an ad- justment was made for specialized clinics). 2 See Survey of All Our Catholic Medical Units (Kampala: Catholic Medical Bureau, 1969), and Jane Hallway, A Survey of Church Related Hospitals in the Anglican Province of Uganda, Rwanda and Burundi (Kampala: Provincial Medical Board, 1972). 277 Given these two adjustments, net inputs were determined. In Table C.2 the figures used to adjust the gross manpower inputs to the net inputs are shown. It may be noted that high level manpower time in rural health facilities is consumes primarily in administrative duties or in specialized clinic assistance. Table C.3 presents the amount of each service-providing input allocated to each disease-specific initial demander. According to the three basic criteria specified in Chapter Five (diagnostic time for inputs used in the outpatient treatment process, length of stay tor inputs used in the inpatient treatment process, and service-Specific information, where data exist for appropriate use of this criterion), these calculations were necessary in order to estimate the value of each element in each sector's technological matrix. The figures are disaggregated according to sector and treatment process. Table C.4 presents data which show the average diagnostic time for each disease type. The data were obtained from two health facilities in West Mengo district, Mpigi Health Center and Kajansi Sub-dispensary. The data on diagnostic time were collected by direct observation of the diagnostic process. Timing began when the patient commenced inter- action with the primary diagnostician and ended after treatment was prescribed. This period included time required for patient examination. The total number of observations at the two locations was 450. In Table C.5, data are presented on the average length of inpatient stay at government hospitals, mission hospitals and rural health units. This summarized information was obtained from inpatient record books at a sample of facilities from each sector. The names of the facilities and the number of observations from each facility is shown below. Government Hospitals Mission Hospitals Rural Units Jinja 1362 Ishaka 534 Kinoni HC 582 Iganga 9790 Kagando 214 Kinoni MC 599 Mbarara 619 Namwendwa HC 712 Kawolo 346 Namwendwa MC 490 Bombo 1158 Buikwe HC 440 Busesa MC 370 Nsinze MC 486 Total 13275 Total 748 Total 3679 With respect to the service-specific allocation criteria, data in Tables C.6, C.7 and 0.8 show the proportion of surgical services, labora- tory services, and Xeray services, consumed by each disease category in each sector. The data for these tables were obtained from the departmental records of a sample of health facilities in Uganda. 278 A detailed analysis was made of the quantities of each drug consumed in the treatment of each disease listed on the hospital out- patient medical form (MF 75; see Appendix F). Data for this analysis came from the outpatient record books of four rural facilities (Kinoni Health Center, Ankole district; Busesa Dispensary Maternity Unit, Busoga district; and Mpigi Health Center and Kajansi Sub-dispen- sary, East Mengo District) and were supplemented by (a) information on the average number of treatments provided to each major disease classification and (b) two books on the diagnosis and treatment of most diseases found in Uganda. The cost of drugs used in various treat- ments was calculated from the Ministry of Health's drug price list- The average cost of drugs per major disease classification was calculated for each sector by summing up the total cost of drugs used to treat each disease in the category and dividing it over the total number of persons treated in the category. The pr0portion of the total cost of drugs and medical supplies constituted by medical supplies alone was then used as a factor (742) to increase the average cost of drugs consumed for each major disease category; for the category, injuries and accidents, however, drug cost was doubled to reflect the greater usage of medical supplies. The figures used in estimating the disease-specific rate of use of drugs and medical supplies for each sector and treatment process are presented in Table C.9. In order to estimate the value of the elements in the diagonal submatrices for each sector of Uganda's health service system, data were gathered on (a) the disease-specific transfer rate; (b) the disease- specific death rate; and the disease-specific rate of unsuccessful treatment. Transfer rate data were obtained from the primary records of a sample of health facilities in each sector of the health service system. Data on the number of inpatient transfers were obtained from the following government hospitals: Mbarara, Bombo, Iganga, Kawolo, and Jinja. For rural units, the data were obtained from Kinoni Health Center, Ankole district; Nsinze and Namwendwa Health Centers and Busesa Dispensary Maternity Unit, Busoga district; and Buikwe Health Center, East Mengo district. In the case of outpatient transfers from rural health units, the data used to estimate the disease-specific rates of transfer were obtained from monthly attendance data of 19 rural health units. (See Table 2.12 for the list of the units and the data obtained). The average transfer rate for all 19 units was then used in prorating the inpatient disease - Specific rates obtained. In the case of government hospitals, it was assumed that the rate of 3Data on the average number of treatments provided to each disease category was obtained in the patient followup survey in Ankole district and supplemented by data obtained in an analysis of diagnostic time requirements in Mpigi Health Center and Kijansi sub-dispensary in East Mengo district. The books are Republic of Uganda, Uganda National Form- ulagy, (Entebbe: Government Printer, 1967), and J. R. Billinghurst, Trowell's Diagnosis and Treatment of Diseases in the Tropics, (London: Bailliere, Tindall and Cassell, 1968). 279 transfer was one-half the disease-specific inpatient transfer rate. There was no information available on transfers from mission hospitals. However, there were few (a) inter-sectoral transfers between mission and government hospitals and (b) intra—sectoral transfers between mission hospitals. Also, most mission hospitals did not offer ambulance ser- vices. For these reasons, it was assumed that no outpatients were transferred from mission facilities and that inpatients were trans- ferred at a constant rate of 1 per 1000 patients. Information on the number of deaths by disease category for the two hospital sectors was obtained from the Annual Statistical Report of the Ministry of Health. Information on the number of deaths for the rural health units was estimated from inpatient records of the five rural units listed above. Finally, several sources of data were used to estimate the disease- specific rate of unsuccessful treatment for each sector of the health service system. For the inpatients of each sector, information on the final disposition of the case was obtained from inpatient records. Dis- charge dispositions included "well", "improved", "transferred", "died", or other, including "run away" or "on request". It was assumed that cases discharged in as run-aways or on—request were unsuccessfully treated, at least to the point of discharge. Others discharged as "improved" or even "well" may not have been successfully treated in the sense of being able to return to prior major activity; it was assumed for these calculations, however, that discharge notations could be taken at face value. Outpatient figures were estimated in two ways. For the rural health units, rates of successful treatment were estimated from infor- mation obtained on a patient follow-up study conducted in Ankole dis— trict. (See Appendix F for a copy of the survey form used in the follow-up study.) No comparable data were obtained on hospital out- patient services. As a consequence, it was assumed that hospital disease-specific outpatient rates of unsuccessful treatment were 50% greater than the estimated inpatient rate. The elements of the diagonal submatrices for each sector of the health service system were estimated on the basis of the above des- cribed data. The disease-specific rates of transfer, death, and unsuccessful treatment for each sector and treatment process are sum— marized in Tables C.lO, C.ll and C.12 in the third section of this appendix. 280 The Data and Procedures Used in Specifying_the Factors Affecting the Output and Resources of the Health Service System in Uganda The following discussion provides supplementary information on a selected number of the variables used in the empirical analysis presented in Chapter Five. Price of Curative Health Services, b. Government health facilities do not charge fees for services except for private and semi-private inpatient rooms. Mission health facilities, however, charge fees for all services; there is generally a consistent fee structure within the health facilities operated by a particular mission organization, but not across mission organizations. It is assumed that change in prices charged in health facilities is function- ally related to the change in the ratio of mission health facilities in the country (g). The major thrust of this assumption is that the relative importance of fee-charging health facilities is declining in the country. The Average Distance to Health Facilities, d. Although the location of each health facility in Uganda is known, the distance to the nearest facility from every point in the country has never been systematically analyzed. It is assumed that the average number of attendances at a health facility per person per year can be used as a reasonable proxy variable, since it has been demonstrated that distance and the number of attendances per person are highly negatively correlated. This proxy measure is available for the period 1949-1969/70, incorporating 20 observations. Total Number of Health Facilities, H. Data on the number of health facilities are available from the 'Ministry of Health's annual reports. This series began in 1900 and has continued to the present. In some cases, data contained in the Ministry's reports were corrected, where delays in reporting the opening of health facilities (particularly in rural areas and by missions) could be docu- mented by mission medical board reports and district medical records. A detailed annual series including 22 observations from 1949 to 1969/70 ‘was compiled. Ratio of Mission Facilities to Total Health Facilities, (g). Data used to develop this series were found in the Ministry of Health's Annual Reports. The series, composed of 22 observations, was compiled for the period 1949-1969. Ratio of Government Rural Units to Total Government Health Facilities, (9%), The data used to compile this series were taken from the Ministry of Health's Annual Reports. The series of 22 observations covered the years from 1949 to 1969/70. 281 Input Mix Providing Diagnostic Services, (;£1). r Although every resource used in the production of diagnostic ser— vices is of potential importance to an analysis of the effect of changes in the input mix on the rate of successful treatment, the most important input resource is the primary diagnostian. The measure used to indicate Changes in the resource mix is the ratio of doctors to non-doctors engaged in diagnostic work. Doctors who are primary specialists, con- sultants, or administrators are excluded. Similarly, medical assistants who are working in anesthetics or some other non-diagnostic role are excluded. The index is constructed such that it can vary from 1.00 to 5.00. A value of 1.00 means that all persons providing primary diagnostic services in a health facility are doctors, whereas a value of 5.00 means that all services are provided by untrained personnel, such as dressers or ward maids. A non-integer value such as 2.50 indicates that the primary diagnosticians at a particular health facility are not homogene- ous. The data were gathered from interviews at rural health facilities, unpublished district administration medical and personnel records, un- published medical and personnel records of the Ministry of Health, and unpublished documents of the Catholic and Protestant Medical Bureaus; these data were used to construct the index for 64 health facilities for the year 1968/69. * Ratio of Service Providing Inputs to Initial Demanders, (X). S An aggregated measure of this ratio can be developed by (a) quantify— ing all service-providing inputs in monetary terms and (b) measuring initial demanders in the aggregate as the total number of initial attend- ences at a particular type of health facility. Aggregate attendance data were available for the years 1948-1969/70. Ministry of Health and district administration recurrent expenditure data were used to develop an aggregated measure for government hospitals and rural health units for the period 1948-1969/70, incorporating 23 observations. Capacity to Import, C. Two measures of this variable are possible: (a) the change in total exports, and (b) the change in the size of the visible trade balance. fmhe first measure provides a good indication as to the upper limit on total.iaports since, in most less developed countries, it is not realistic to contemplate the continued possibility of financing imports from the combination of net inflows of factor income and capital loans or grants. The second measure provides a better indication of the potential to increase total imports, since it is a net figure - exports less imports. In addition, this figure provides an indication of the country's long run import preferences, preferences which tend to be relatively constant in the short run. Although Uganda has substantially curtailed certain types of imports through taxes and quotas, the trend of the visible trade balance provides a good indication as to the extent to which the inqxxrtation of health-service-related commodities can be expanded. Data on the visible trade balance were available for the period 1948 to 1969/70. 282 The other variables are self—explanatory or are discussed in the body of the text. Tables D.8 - D.1l present additional results of the regression analysis using the variables described above. Summary of Linear Programming Solution for Uganda Government Hospitals, Mission Hospitals, and Government Rural Units for 1968/69, Where a Minimum of One-Half of Every Type of Initial Demander Must be Treated In Table C.13, an alternative linear programming solution for Uganda's health service system in 1968/69 is presented. The differences between this solution and the one discussed in the text are due to the constraint imposed on the objective function, which requires that a minimum of one-half of each type of initial demander, both inpatient and outpatient, must be treated. As a consequence, all disease types are included in each sector's output solution, whereas the linear programming solution presented in the text incorporated only those disease types which had the highest rates of successful treatment and the lowest use of resources. Only after treating one-half of all initial demanders from each disease type, could the present linear programming problem use the remaining resources to maximize output. The case types having shadow prices (see Section II of Table C.13) indicate the disease types which were treated after one-half of each disease-specific initial demander was treated. In the case of the government hospitals, most of the case types having shadow prices were treated on an outpatient basis, whereas for missions, the opposite is the case, with the inpatient treatment process predominating. A significant similarity between this set of solutions for each sector and the one presented in Chapter Five is that the service-providing input constraints remain the same in each sector as indicated by the service-providing input shadow prices presented in Section III of Table C.13. Although the actual shadow prices may differ in magnitude between the two solutions, the policy implications discussed in Chapter Five, ‘particularly with respect to health manpower training, are further sup- ported by the results shown in Table C.13. Finally, it is important to mention that the imposition of the constraint on the linear programming solution as presented in Table C.13 :makes.the cost considerable in terms of the total number of successfully treated. A comparison of the objective function figures in Table 5.6 .mnd C.13 makes evident the magnitude of the tradeoff facing Uganda between (a) maximizing the number of successfully treated persons and (b) allowing all diseases to enter the solution, with at least 502 of all initial demanders receiving treatment. Number Successfully Treated No Constraints Constraints on Difference on Objective ftn. Obj. ftn. Government Hospitals 4,663,032 3,880,243 782,789 Mission Hospitals 614,749 480,416 134,333 Government Rural Units 5,192,066 4,235,417 956,649 Total 10,469,847 8,596,076 1,873,771 283 Given this tradeoff, it is clearly a political decision which must determine the course of health policy: should it focus on the objectives of maximizing output and quality (as defined in this study), or should it make curative health care accessible to all Ugandans, regardless of the illness contracted? (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) 284 APPENDIX C Table C.l A Reconciliation Between the Eleven Manpower Input Categories Used in Chapter Five and the Health Occupational — Titles Used in the Ugandan Health Service System Manpower Categories Doctors Medical and Nursing Assistants Professional Nurses/ Midwives Enrolled Nurses/ Midwives Trained Lab Staff Trained X-Ray Staff Other Trained, Medical Staff Other Trained, Non- Medical Staff Other Non—Trained, Medical Staff Other Non-Trained, Non- Medical Staff Students (1) (2) (3) (4) (5) (6) (7) (8) (9) .(10) (11) Included Occupational Titles All Medical Officers, including Specialists, Surgeons, Residents, and Medical Super- intendents Medical Assistants and Nursing Assistants including Senior designations Uganda Registered Nurses and/or Midwives, Nursing Sisters, Sister Tutors, and Health Visitors Uganda Enrolled Level Nurses and/or Midwives Laboratory Technicians and Assistant Laboratory Technicians Radiographers and Assistant Radiographers Pharmacists, Assistant Health Visitors, Theater Attendants, Dental Technicians, Storemen, Blood Donor Attendant, Dispenser, Orthopaedic Assistants All Clerical Office Staff (including Hospital Secretary, Assistant Hospital Secretary and Clerk), Ambulance Drivers, Clinic Writers, Statistical Clerks, Headman, Seamstress, Domestic Assistant, Laboratory Attendant, Carpenter and Mechanic Ward Assistants (including Ward Maids and Nursing Assistants) and Dressers Sweepers, Porters, Manual Laborers, Cooks, Orderlies (Pharmacy, Laboratory, X-Ray), Office Messengers, Watchmen, Peelers and Dhobies. All Medical, Medical Assistant, and Nursing School Students. 2£MS .uaHux on. o n n a II I... a n a II II a n o I «on one a n Hon noH H non one non one ans non H sun non H as ooHuouoeo uoauo .HH nHo.H~H nn nn nan.H~H . «ne.ono nu nu Hoe.ono «nH.ooo.~ nn nu ~n~.ooo.~ ..a. nuHoHuuuoHu .oH . . nn nn . . . nn nn HcH.nnn nna.aoa nu un nun.n a ..a. .uaHuz on. HaH an“ H HnH ens H Hou nan o ooHunuooo .HoHaoo .nH nnn.no~ nn nn nnn.no~ nan.onn nu un nan.onn noa.noH.o nn nu naa.anH.e ..a. ooon .oH . . soon H .naH H n o oao.onn.~ ~HH.HHn.H n un Noo.Hoo.~ enH.ono.oH n nn oon.noo.oH on. ooHH o n u . . . . unuo HuuHooxxuoouo .nH n n II II n n n . II .II n II II - ooo Hou H ooo Ho~ H ooo HHH H ooo HHH H nno one H nno nno H ooo noon .«H nnnnnn un nn unnunn ooo.o: nn un Soot o8.ono.H I l ooo.ono.H .85 . Bunsen .HH . n n n n It n n n a II- n n . uwam quail“; ooo non H n n oon «He H «on nno oH ooH Hno H ooH oHH o oa ooH «NH n .HH oooHuuanooz .oH . nn . . nn . . . nu . . . Hanan HuoHooz one son n ooo non nna nno n ooo Hon nnN Hon H n ooo «no H on: ooaHnuanooz .n . nn nn . . nn . . nn . . ..ua «noon HuuHouznooz onH aHH onN HHH new a on one on one so" on onN «Nu H ouaHona noeuo .o n n H II n n II a .09 Hmfium ova oHH Ho nN n onn noH non HH n ooH «H ooo non oH ooo Hoo a ouaHaua guano .H unnunn un uu unnunn ooH.n~ nu nu ooH.n~ ooo.on uu nu oon.on .oue nuoun nouns oooHaua .n oon.oH nu nn oon.oH ooo.noH nn nu ooo.noH ooo.o~H nn nn ooo.o~H ..ua mason noH oouHous .n a a a n I-. a a Q I A a s as; noH mam om n ooo ann onH oo« oo. nu. non non N n ooH -n N on: \uoauoz ooHHouom .o o oH on oon.on ooo.~n~ un oH H.non oun.n~n nu o . ..ua nopnaon on H ooo «on \nouuoz .noun .n oo~.oon nH n ooo.ano oon.o un nn oon.o ooo.nnn nn nn ooe.nan ..ua ..u-.< noH-non \s.u.: 0‘! ON o un ooH Hoa.no HHo.nHH nn «N o-.~oH oon.onn nu nH oon.onn ..ua nuouuoo .H unsouH ooz ooHaHHo goo a assauH ouoouH Hoe ooHaHHo ao< a“ noose «HoeaH Hoe ooHaHHo moo a .uoouH oooaH Hsnuoam a noose Henooem u Hensonm u scone undouboum ooneuow suds: auasom anus“ nauseous: noH-sax ndsunnoom use-snuboo soumnm ounouom euasom s.sossms mo uouuom noon sou musesu nuavwboum soapuom we muaflaosaas>< emu as one: sunssuosnu< «.0 sunny U Nunlflmmd 286 3 on . a . .10! 305' sons-usua- l.» on .8: EH3 :5 a. not...» .5 van! 3 focus. 3 8.3-H can A33»... worsen-In: «.3. II: A: vs- 3HH 9. H333 ve- swsuv .3 5n: A3 5.3-sofa n8 Eons... .3 39305 .35.- .83337 3300;- ouEs. 2F ".30: «$.39 03.52 35.3" 03.0: v3.8. 31: 30.30; .332; n33: .1. .3 .ui a do . . . 3...”: . . . . 38...: .55 Rain 31: 3. .n. a... n. 3.: H: ..o H3 2. a «nu 8. H 3.8.... H.386: 5.3.; HHo...H ..n. 53:83 . :15; 32%.: 333:. «34% 3:...3 52...: 3...: 33.3: =5...an .1. ...H... . ..o 3.3.. . “Ho-n.“ ”.03 EO‘N "finoxh “kchaO. “zuham. a... z . o~o.nh~.n 30.0: 2.§ofl angled Alv§.n~n go‘s .00.... gun-fives niv§.xncfl d2.‘.n fl:.§.£ .3. 369m .Q!\§s . 8..H.H.H 8..H.~.H 80.3.. .853 nno..n..H no.2; .F. o... a... . 3.9: 80.0: c8333 28.3...” .31 33an . Sign; mgr—3 C.;: «3.3. Auzno£= 333.»: 3...: 31.3.. 8.63.» o3...3.« ...us ..a .3 son .5. so: . Oflhnhg gog. as“? “Eh-“g as! a'adCH “and...“ 8ho§ed “aflods ..H‘ “.3” .3 .hH DO: . 3n.:H Assam: :0; =0; ’03 3.0.3 onnnn.~= A631: «3.3 is. I! influenza" . 31H. oH..Hn 8H... 3.5 3.5 no... 8...: onto: 34... ...a .3. 1......— ...so . 5.3 03.: 80.3 80.3 .31 and.” sun oceans.— . O n . oon.oH oon.oH ooo.noH . . So no. So .3 8o..HH ...n :3. .3 v.3... . 8st one..." nnn.n 8H 8. 33... H. on H! 3.88.. 33%": 5.21" .5. .8588..- 33... . §.u¢~ 3a.»: «3.3 89.9: 03.8. a... .5 nos!!!- Joum . 3...: 8...: 8...... 8n. 8.; non: 8...: 38.... 8.8. ..a. .3... 38.2.3: . «no.5 “353.3 03.: “0.3 98.63 ASH-«Haas gamma :1: .9:— Bouuoo . I; . a; ...H 5. ...o I... H38. ...H n... ...o In. 39: H38 53...... IFS-5 u. 6.15 8.86323 1... 33...... 3...... 5 .o 5’3 5 3.8.9.2: 38...... 8.385 2. 59.83 8.88.83 .538... 8.2.. 3.3.8 it I. 8 .82.... 3.8.3. 2!... 3.8.»... 8.33.. 1. 3 J35... .3333. all. 2:83.. .8328 I. 8 .5381 188:. all. ...- fl... 3.... Hula 33. 8.. 8...... 33....- null-.8 003$: .‘uosn 00E: Judo. nausea-8 05 cu 059.83 s!‘.— 03:90.: SE3 .0 83.00.34 IF n.U Candy. 0 Mass r-NHCGON. 287 APPENDIX C Table C.4 Average Diagnostic Time for Each Major Disease Category Treated on an Outpatient Basis at Mpigi Health Center and Kajansi Sub—dispensary, West Mengo District, Uganda, September and October 1970. Disease Average Diagnostic Category No. of Cases Time (in minutes) I&P 144 2.00 NC 1 0.67 AMB S 2.05 . NS 25 1.51 Circ -- -- Resp 76 1.91 Alim 38 1.92 GU 21 1.87 Preg & Puer -- -- Del w/o -- -- S&MS 45 1.18 NB - -- Ill Def 27 2.37 Inj 68 1.04 TOTAL 450 1.74 Disease Category 18? NG AMB NS Circ Resp Alim GU Preg & Puer Del w/o S&MS NB Ill Def Inj TOTAL 288 APPENDIX C Table C.5 Average Length of Inpatient Stay for Each Major Disease Category Treated on an Inpatient Basis in each Sector of Uganda's Health Service System 1968/69. Government Hospitals (days) .3.12 10.87 5.17 6.26 8.43 3.49 3.57 4.51 3.76 1.65 6.70 4.69 5.06 6.33 4.03 Mission Hospitals (daYS) 6.78 13.82 9.03 8.71 18.92 5.99 6.93 6.41 6.66 4.03 9.46 8.69 4.09 7.56 7.17 Rural Units with Beds (dayS) 3.72 2.50 3.82 3.75 5.29 4.07 2.98 3.58 2.00 2.17 4.70 2.75 2.82 6.12 3.38 289 APPENDIX C Table C.6 Proportion of Surgical Services Consumed in the Treatment of Each Major Disease Category in Each Sector of Uganda's Health Service System 1968/69. Disease Government Mission Category Hospitals Hospitals 2 Z I&P --- 0.1 NG 2.7 4.5 AMB 0.2 —-- NS 0.3 0.8 Circ -- --- Resp 0.3 0.2 Alim 25.5 15.8 GU 26.5 17.6 Preg & Puer 6.9 21.5 Del w/o --- --- S&MS 27.3 8.0 NB --- 0.2 111 Def --- 0.1 Inj 10.3 31.1 Note: (1) Surgical services reported here are consumed on an inpatient basis. (2) There are only minor surgical services provided in rural units. The records of rural units indicated that such services were infrequent. When a person required surgery, he was transferred to a government hospital. Disease Category I&P NG AMB NS Circ Resp Alim GU Preg & Puer Del'w/o 88MB NB Ill Def Inj Note: 290 APPENDIX C Table C.7 Proportion of Laboratory Services Consumed in the Treatment of Each Major Disease Category in Each Sector of Uganda's Health Service System 1968/69. Government Mission Rural Units Hospitals Hospitals ‘with Beds (days) (days) (days) 66.44 79.90 99.40 0 o 01 O a 4 6 ..- 29.72 11.45 0.60 0.41 --- --- 1 a 33 00 01 --- --- 0.19 --- --- 00 O6 -..- 2 o 08 O a 01 --- --- 1.49 --- --- 6.42 --- Laboratory services are generally consumed in the process of diagnosis -- an outpatient activity. As a result the disease categories refer to the outpatient treatment process. 291 APPENDIX C Table C.8 Proportion of X—Ray Services Consumed in the Treatment of Each Major Disease Category in Each Sector of Uganda's Health Service System 1968/69. Disease Government Mission Category Hospitals Hospitals 2 Z I&P 39.66 61.87 NG 0.40 0.27 AMB --- --- NS 0.84 0.53 Circ 4.40 2.80 Resp 0.25 0.16 Alim 4.96 3.15 GU 0.23 0.14 Preg_& Puer 0.45 0.29 Del w/o —-- --- S&MS 7.51 4.77 NB --- --~ I11 Def --— —-— Inj 41.06 26.04 Note: X—Ray services are generally consumed on outpatient basis, prior to inpatient admission. As a result, it is assumed that such services are consumed on an outpatient basis. 292 APPENDIX C Table C.9 Average Cost of Drugs and Medical Supplies Consumed in the Treatment of an Initial Demander in Each Major Disease Category and in Each Sector of Uganda's Health Service System in 1968/69. Government Hospitals Mission Hospitals Rural Units with Beds Ave. Ave. Ave. Ave. Ave. Ave. Ave. Ave. Ave. Drug Drug Drug Drug Drug Drug Drug Drug Drug Cost & Med. 8 Med. Cost 6 Med. 5 Med. Cost & Med. & Med. 09(1) Supplies Sup- or Supplies Sup- 0? Supplies supplies Disease Cost 0P plies (3) Cost 0? plies Cost OP Cost IP(3) Categories Cost IP Cost IP(3) I 8 P 0.54 0.95 27.81 0.71 1.23 11.12 0.24 0.41 3.95 NG 0.03 0.05 9.65 0.03 0.05 9.65 0.02 0.03 2.41 AMB 3.57 3.70‘2) 17.33 1.12 1.16(2) 4.31 0.10 0.18 1.79 NS 0.15 0.26 16.82 0.13 0.23 2.56 0.09 0.15 1.41 Circ 0.36 0.63 51.48 0.37 0.65 14.03 0.22 0.38 4.67 Resp 0.28 0.48 20.78 0.27 0.47 3.89 0.16 0.28 2.90 Alim 0.10 0.17 7.64 0.16 0.28 2.57 0.12 0.21 1.74 GU 0.28 0.48 18.08 0.30 0.52 4.48 0.17 0.30 2.80 Preg 5 Puer 0.01 0.02 10.39 0.01 0.02 6.00 0.01 0.01 2.00 Del w/o ---f ---- 2.24 ---- ~--- 2.50 ——-- ---- 1.00 S&MS 0.38 0.67 45.91 0.37 0.65 7.60 0.22 0.39 4.41 NB 0.11 0.19 10.22 0.09 0.15 1.67 .0.06 0.10 0.83 Ill Def 0.45 0.79 43.76 0.45 0.79 4.93 0.27 0.47 3.79 Inj 0.34 0.68 38.95 0.33 0.67 11.40 0.20 0.35 4.77 Note: (1) All figures are in Ugandan Shillings. (2) The average drug cost of this disease category is so high due to the continued treatment regime required by a limited number of diabetes patients who comprise a proportion of this disease category. (3) The estimated inpatient average cost figure was derived by using the average outpatient cost of drugs and medical supplies and adjusting it according to two criteria: (a) a severity factor and (b) length of inpatient stay. The severity factor assumed that the average consumption of drugs and medical supplies would double during the first two days of inpatient care. After that period, it was assumed that the average outpatient consumption of drugs and medical supplies would approximate the average daily cost of drugs and medical supplies consumed on an inpatient basis. Elements, a, of the Diagonal Submatrix for the Government Hospital Sector in Uganda in 1968/69. 293 APPENDIX C Table 0.10 Service 8 Demanding Estimate j Rate of Rate of Rate of 8 Inputs of a (1968/69) TJ UJ Dj ipj Outpatient I 6 P 1.08 1,856,893 0.66 4.95 --- 37,464 NG 2.50(E) 4,819 2.17 13.04 --- 4,085 AMB 1.34 57,181 1.46 5.47 --- 10,467 NS 1.17 316,427 0.70 11.54 --- 6,107 Circ 4.60 5,422 0.70 6.66 ---- 3,845 Resp 1.05 1,086,744 0.50 2.25 -—- 23,642 Alim 1.06 760,258 0.50 2.51 ---- 22,505 GU 1.12 141,520 0.54 4.06 ---- 8,830 Preg & Puer 1.72 30,838 0.70 1.98 --- 25,017 Del w/o --- --—- ---- —--- ---- 41,911 8 a MS 1.04 763,935 0.70 1.92 --- 10,379 NB 1.38 11,448 0.70 9.38 -—- 1,976 111 Def 1.11 245,761 2.03 6.09 --- 4,371 Inj 1.10 533,597 0.96 3.58 --- 24,103 Inpatient I & P 1.11 ----- 1.32 3.30 5.07 37,464 NG 1.23 ---- 4.35 8.69 5.41 4,085 AMB 1.20 ---- 2.92 3.65 10.10 10,467 NS 1.19 ------ 1.40 7.69 7.00 6,107 Circ 1.19 ------ 1.40 4.44 10.24 3,845 Resp 1.10 ----- 1.00 1.50 6.95 23,642 Alim 1.09 ----- 1.00 1.67 5.56 22,505 GU 1.06 ------ 1.09 2.71 1.78 8,830 Preg & Puer 1.04 ------ 1.40 1.32 0.76 25,017 Del w/o 1.00 ------ -—-— ---- ---- 41,911 8 8 MS 1.04 ------ 1.40 1.28 0.86 10,379 NB 1.27 ----- 1.40 6.25 13.36 1,976 111 Def 1.15 ----- 4.07 4.06 5.21 4,371 Inj 1.07 -—--- 1.91 2.39 2.54 24,103 Notes: (1) (E) - an assumed figure 31 (2)3 - _ _ opj Sj sj(Tj+Uj) sipj (3)11ipj - SIPJ sip: Sip: (Tiwiwi) 294 APPENDIX C Table c.11 Elements, a, of the Diagonal Submatrix for the Mission Hospital Sector in Uganda in 1968/69. Service 8 Demanding Estimate j Rate of Rate of Rate of 8 Inputs of a (1968/69) '1‘j Uj D.1 ipj Outpatient I & P 1.16 170,837 0.00 6.00 --- 29,147 NG 2.50(E) 1,203 0.00 9.60 ---- 1,855 AMB 1.46 22,138 0.00 4.05 --- 8,876 NS 1.18 25,586 0.00 9.00 -- 2,971 Circ 2.54 3,061 0.00 4.95 ---- 1,758 Resp 1.12 55,345 0.00 1.65 --- 11,526 Alim 1.17 53,873 0.00 1.50 —-- 12,039 GU 1.32 16,228 0.00 3.75 --- 5,498 Preg & Puer 1.75 7,605 0.00 1.50 -- 8,888 Del w/o ---- -- --- ---- --- 16,359 8 6 MS 1.06 39,783 0.00 1.50 --- 3,980 NB 1.83 3,228 0.00 6.90 --- 2,558 Ill Def 1.23 9,313 0.00 4.50 --- 3,461 Inj 1.21 10,655 0.00 2.70 --- 3,081 Inpatient I 8 P 1.06 --- 0.10 4.00 1.66 29,147 NG 1.12 --- 0.10 6.40 4.15 1,855 AMB 1.06 --- 0.10 2.70 2.44 8,876 NS 1 11 --- 0.10 6.00 3.43 2,971 Circ 1.11 ~-- 0.10 3.30 6.59. 1,758 Resp 1.05 --- 0.10 1.10 3.58 11,526 Alim 1.05 -- 0.10 1.00 3.25 12,039 GU 1.03 --- 0.10 2.50 0.60 5,498 Preg & Puer 1 02 -- 0.10 1.00 0.68 8,888 Del w/o 1.00 --- --- --- —--- 16,359 S 6 MS 1.02 --- 0.10 1.00 0.62 3,980 NB 1.22 --- 0.10 4.60 13.40 2.558 111 Def 1.05 —-- 0.10 3.00 1.47 3,461 Inj 1.04' --- 0.10 1.80 1.55 3,081 Notes: (1)(E) - an assumed figure. 3 (2)8 - J 995 - - SJ SJ(TJ+UJ) 5191 5191 (T (3)8 a: . ipj , sipj sipj +U +D ) J J J 295 APPENDIX C Table C.12 Elements, a, of the Diagonal Submatrix for the Government Rural Unit Sector in Uganda in 1968/69. Service Demanding Estimate S Rate of Rate of Rate of S1 Inputs of a J T U D 91 J J J Outpatient I 8 P 1.15 2,524,020 2.19 7.00 ---- 89,954 NC 4.62(E) 153 10.75 --—- -- 480 AMB 1.25 50,015 6.61 10.00 --- 1.570 NS 1.27 358,048 3.40 18.00 ---- 480 Circ 3.86(E) 136 3.40 ---- ---- 369 Resp 1.20 1,670,109 1.80 12.50 --- 40,453 Alim 1.49 836,656 6.47 25.00 ---- 11,226 GU 1.05 150,793 3.40 0.00 --- 2,659 Preg & Puer 1.06 65,725 3.91 0.00 --- 1,330 Del w/o —-- -—-- ---- ---- ---- 57,995 8 6 MS 1.12 938,236 5.86 4.00 -- 7,976 NB 1.09 2,682 3.40 0.00 —-- 129 Ill Def 1.11 493,659 4.42 2.60 ---- 12,925 Inj 1.30 609,980 10.32 10.00 ---- 15,104 Inpatient I 6 P 1.16 --- 5.10 4.76 4.03 89,954 NG 1.52 ---- 25.00 6.21 2.94 480 AMB 1.43 --- 15.38 11.54 2.94 1,570 NS 1.21 --—- 7.91 6.21 2.94 480 Circ 1.38 --- 7.91 16.67 2.94 369 Resp 1.15 --- 4.18 5.82 2.99 40,453 Alim 1.23 ---- 15.05 2.69 1.08 11,226 GU 1.24 ---- 7.91 6.82 4.55 2.659 Preg & Puer 1.27 -- 9.09 9.09 2.94 1,330 Del w/o 1.00 --- -- --- --- 57,995 8 6 MS 1.29 --- 13.64 7.57 1.52 7,976 NB 1.21 --- 7.91 6.21 2.94 129 111 Def 1.38 ---- 10.28 15.42 1.87 12,925 Inj 1.57 ---- 24.00 9.20 2.94 15,104 Notes: (1) (E) - an assumed figure (mOPJ - Si 51 SJ(TJ+UJ) sip: (3):!ipj - $1211 sipj sipjcrjijj) I II 296 Optimum Number of Cases 7: Outputs : Objective Total No. [.6 . Function initial 1: Del w/o s s as NB 111 De! In) Equals Demander, Government Hospitals(l) In 1.00 0.48 0.40 0.43 0.47 3,880,243 Current Values of(3) Initial Demanders 17 41,911 10,379 l,976 4,371 24,103 6,033,695 Hilsion Hospital-(1) 90 1.00 0.19 0.32 0.95 0.97 490,416 Current Values oft3) Initial Demanders 38 16,359 3,980 2,558 3,461 3,081 963.684 Government Rural Units(l)99 1.00 0.39 0.41 0.36 0.32 4,235,417 Current Values cf‘l) Initial Demanders bz 57,995 7,976 129 12,925 15,104 7,973.40! Shadow Prices of Case Typi Initial Demanders 3;‘ Del w/o s E as up 111 Def Inj Government Hospitals [ (a) Shadow Prices 00 0.60 0.00 0.00 o oo 0.00 (b) Cost of Nontoptimal7 - 0.88 0.12 o 98 o as I Mission Hospitals ' (a) Shadow Price! 26 0.55 0.00 0.03 0.52 0.16 (b) Cost of Non-Optimal - 0.05 - _ _ Government Rural Units in) Shadov Price! 00 1.00 0.00 0.00 0.00 0.00 (b) Cost of Non-Optimak9 - 3.05 l 42 2.02 6.07 Shadow Prices and Slack 04 Inputs‘Z) Government Hospitals (a) Shadow Prices‘s) (b) Slack Quantities (C) Current Values(3) Mission Hospitals (a) Shadow Prices(6) (b) Slack Quantities (c) Current Values(3) Government Rural Units (a) Shadow Prices(6) (b) Slack Quantities (c) Current Values(3) {The optimum number of successfully treated cases is expressed as a proportion Iof the current value figure which is the number of initial demanders. Ilnputa numbered 1-11 are measured in hours; input 12 is measured in bed days: iand inputs 13-17 are measured in shs. I The current value figures under the optimum number of cases treated and slack quantities of serv1ce—prov1dlnq inputs represent the two parts of the input vector, the service-demanding and serv1ce-providing inputs respectively. 1 gThe shadow prices of initial demanders show the amount by which the objective 'function could increase if an additional initial demander with the given disease characteristics were to demand serv1ce. The cost of adding a non-optimal case to the solution shows the amount by which the objective function would decline if one of those types of cases were treated as opposed to those cases which are treated. The shadow price of the service~providing inputs shows the amount by which the nLnber of successfully treated persons would increase if one additional unit of that input were made available. APPENDIX D APPENDIX D Other Supporting Tables 297 D01 D.2 D.3 D.4 298 Supporting Tables to Chapter Two Health Facilities in Uganda Selected Indices on Size and Structure of Uganda's Health Service System Structure of Attendances at Uganda Government Health Facilities Number of Attendances at Government Health Facilities 299 ..auouan n.0unuz can..uz on“ can sauna: no stunncn: ..nuaana gun: sonpuuuan can gong-uncanuou nunouuos noun nonununo ouca nuuounn no. ouunoou no. «age "unnunn-u ..c.=.a. .ocou .oau.oc= .ce>ue: ugc :.> ..c .a :59: ...ouju no unuuuvos unounonn .quauqu ..o.=.uv mumma: no ..unpuom nun-o: on... on“ no nan-..u.n< on. u-.ua. u gouaosuu .non.u .n “nouooon-no\eonn ..vuoouu uuouu-nu-um nounphom sauna: «o bun-«an: neoxncau smacks» guano: no ska-«an: osu uo_ummmummlmummm« unsanuOu on on. ou-v ecu you nuousom any .oonuq-coaonv van unnunn-on Alu< vac acunum ovanuuo nouawnu any .uuuaa anonouau van nono< .anuuu< coutuon cool on couuoanu-nv 0:» onus: nuao sync: caunuu< ovauucu nouaunu Ac. undue-.41uooou .ou-u» acnuuoauu an oucasu cu can .naoo ooau wean I anannin now can Iona I M .ooon .un .009 we no nouuuanuou «o goo». 03¢ ou bum-u Iuov ogu .nanuuauoz Aunucano> van .upom uo henna: osu you uaooxo I as .momu .un uoaauuoa uo an aonuunuuuu no soon. on» on name» nuuv as» ..uauunoo; huoucauo> van acoe=u0>om no uoaa=c ago now unnunn I c .Quqv u~na~na>¢ Ibuu coo-Inuuu I m undue: non u eon uu n nu nnn n uo o nn on unun nuqun «anon nu en uon men n nan uunu an nn un nnnn anon usnn nu nn non nunn unnu nnnu uuos nu nu uon u nan nu . on nu noun - on an no nu nusn nnon onnc nu cu non n nu . nn un ”nun . nu «u on nu unuu menu anus nu nu emu u nnnu onus usuu nuoe cu uu omu n nu n nu on nonu n uo nu nu nu unns ousu cue. nu uu can u nu u nu n. nnuu n no «u nn uu nun» essu unnn nu uu umn n nnnu unnn nnnu nnan ou nu one n ou n nu on nnnu u so en nn on «onn «nan ouqn ou su mno u on on un nusu e nn un on n cums oonn oonn on cu pnn u menu nuns ousu cann on .u one u nu u nu n. unnu u un en on n onn. nsqu noun on nu one u ou u on u. uuuu o no n un nnuc noun uuon nu nu ouq u . nc annu ua nn one. unen oeou a nu one u no u no ouun nu on ouon «no unnu n uu uun u nu n nn nunu un un noqn «no nuuu n uu uun n non u u. nunn nu on ouun one uonu a uu uun n non u un nnnu no un uqcn unu, nuuu a nu uun n can u un anon no nn unnn nun , nuuu n nu uun u on u nn unun .. on nnuu n nu on uu non ou nnnu nu ..ncu a nu nnnn uu non nu ovum .9: .uo> .u>ou .u.» .upr .nop .utoo .no> .uooo .upoo .no> .upoo .no» .usoo .upou nouoh .no> .»ou .uoh .«ob .pau A...ou. ...ou vu< .un::.uqx .unna .u-z .n.na.aam .10. .u.: gonna-aoa-na guns: .a-z nuouauu .vu- nuuuaoos .n-uns-o- .n.o= unusa: .uaunnunam nauuanuau nuqnaaaana gun-o: can: n-uow ave-u: 6« coauuauuoh sundos u.a canon a Nunzmmm< usxOsmu cuxooma cc ao\woo~ o mo\noau ~o\owon oo\moo~ noxcoau coxnooa nox~oo~ ~o\~oa~ nexooau cow”. amun anon hood anon nnau «nod nnou Nnom «non Onau anon scan «non «dad 3OC) n.uq o.nn nu s.on n.nu nn u.nn n.nu no u.uq n.nu um o.n¢ n.nn nn n.0e o.ou uu «.nn n.nu nu u.en u.uu uu n.nn n.ou us o.nn on o.nu on c.nn on nanny nauoa no u no a .non sun: n-uuuaan anus nuuau .0d0flHfiUIfi .HO> HQUOH «.mc mod m m o.nn Hon c.0u m.oo «mu m. «.mn and m n.mo oea 0.0H m.oo and H.HH n.no «nu .N.NH o.mc one N.NH ~.oe and «.HH n.9o mun o.~H m.~c «HA o.- m.~o ONH «.mn o.co can Q.NH n.nm om c.cu N.nc on n.nH «.He an 0.0H nouoa Huuoa no N we N «cocoa nuua .nnom aunaa nauau lauuhm oon>uom nun-on o.ovnlun mo ouauoauum can ounm no ooonuan vouoouow .uuom nauuuonuu an undone ou 03v .muno .oomn doom I muuacuh now can nouaunm ‘ .uunaa unannounx no. .oouuo-uumuna .ounua manau0u¢x\muonaomnun .quuaou sun-cu ovauoau uuoa nun: gang: nouns cc .qvonup on nonunnnoom sun-u: .n.a «upon an nonamaou ou-v aouu vo>nuov on sunny saga an wound-«um nonunauouan one c «an mmu nnN nnN wow mmH omH 5mm HON NmH nan sud can onfi own man Auuooa vnq .Huuov uununnnoum .u>ou Manon N.o- c.0uu <.no~ n.~mH n.0ma o.w~H o.mhd m.moH n.mn~ n.noH c.~nH c.5HH c.5Hu N.AHH 0.90H o.mm H.wo Aooalanmuv ovum .uunn .upoo no navan m.om~ m.ma~ ~.QHN n.mm~ N.omH H.omn s.mnH c.onH n.~ea m.oeH n.mna ~.o .u>oo anon .noo: we wanna N.a IHAIH a Nanumh< a.o~u c.hHH m.< m.snn d.wmu n.0md <.NmH m.~wa m. <3-(«w-nun~o\o c>25c>c>c>c>c3 OI comma acouuaaau Haunqmom oa nuaao uaonunnan and: Adana «0 oaunm Hauoa Haunm H0 0 O O O‘HHOO‘NHO ‘OOO‘GNOOO G60 0 C O O. O O HNHHNHNH HI-C ”NQOMB‘O \Dhafla‘HNO HHHHNNN oomwo Haunmuom ou «undo was: Huuoa no ouuqx s.~om n.no~ n.0oa «.mmd m.uwH m.owH «.5nd c.~¢H n.hca c.num o.muu o.H°H m.sm o.coa o.~m m.mn uaonuaauno a unonuumaH ooaavaouu< noun: Hanan Hanan mo Hanan w.mmn o.~nn «.nun o.en~ o.mmH c.0ma m.m- o.~cH m.coH H.~mH w.ced o.s¢H c.moa m.nm «.mm c.oon N.~HH c.mo uauuuqmuso a uconunnaH coucovcouu< .maom nanny uo seven .uoom aunuuoauu an unnunn on one .muno .moma econ n muqaaon now one noununu any .vovsnoan no: on. can .uuum and» .HHGD HH< mcowmmnamd noonuaasu m.on~ o.nom .o.nnu «.mon u.non n.mn~ u.omn o.cnu n.unn n.m¢~ o.mmn o.oou o.cnn n.own c.mmn A.mon n.onn c.umn o.m~n n.5nn ~.uon n.5nn .<.z n.mm m.wm o.mm n.mm n.mm o.oon o.ocn n.5mn o.wm .<.z m.mun .<.z m.n .uuuuu 3oz nuns: HH< uuaonuaauno no «oven nouns an «oven AHV monunnnoum sunuow uauaauu>oo «was»: an nooaovoouu< mo ousuoauum n.n ouauk n Nunzumm< n.nuuu oua «use o.ooH . nnnu nuv ou\nonn no\oonn mo\nomn nc\nonn no\nenn no\sonn sounonn no\uenn uo\nonn no\onnn any onnn nnnn nnnn unnn onnn nnnn «nun nnnn unnn nnnn onnn unnu nnnu your 302 .a-«onuanunum e.gunoo: no hug-«en: on» no nuance «guano: on» Ibuu non-A30uno nausunu «anonuuoas on. cough Aouv .ooununu swoon-anon: louu noun-«and In henna: cash Amy ...usuau noun-«an. one an vovauocu we. noun-nuns on. comma: nuonuuoz sea uou unnunn» any .vovauuoa «on on. caucus nuunuuox «nu uou couaunu an. .oomu no «no: u-unu can an canon» «o uoaaaa vovuouou can manugaov an van-uauuau .ouauuu ecu-Inna. ad on such nov .un wanna .on .m .unmu .ucoauunmun nuance: one we gnomes unacc< as” an chose ouamuu vaunlnuuo ca.nn one» “av .mnn .A .dwwwqmml..u .u .nxccnnosa cum .~nmn .un punaouua «o as aunnuun>~ unnunn use an «ugh new .uuom onsu hOw aoucmvcuuuaou ovauucn ooH. nunwu an: new 930:. nouawnu any .nnu .a .nosnu .unuum uunauo>uca auouxo ”concouv .u .no> qwuuaum uaucouoo gangsta on“ no nu>u=m unnmnuwoaua ..u .x .«sucauuas 00m .zaOnu-uanom m>nuaz: uauou vouaauuuu you .eumn .un noun: uo no oupauu~>n ousunu ecu an anew Anv .MMMMHMM1mmmmmfl guano: mo auunucnx aouu o~a~uu¢>¢ couumuuounn Iouu vounoaou one: .vuuounvau onus: uaouxu aunv ~n< .nuuaunu «o acuucsou on can unnunn» au-uOu fiasco no: ans nunuouueam .uvcu-aosu an coma» 0». nouamnu uu< any announuru an: I <2 use» annuuonuu an sac-=9 on one .5uao coma econ u unwound new a». chose nousmnm I c ”nouoz .umm cu.u o.oun.o n..o.unu . o.soo.ou cu\oonuuec «n.u o.n9u.n u.nos u.cnu n.ouu o.snn.u u.nuo.n o.ouu.u n.unn.n o.ooo.o n.uss.n n.suw.uu n.nou.nn u.unn.o no\nonn un.n o.nnn.u n.non n.oun s.onn u.onn.o n.nun.¢ u.uun.u o.n¢e.n s.cus.n o.nno.n u.uuo.nu u.uuo.n u.snu.n oo\nonu om.u o.nnn.n s.n0n n.4un u.uun n.nnn.o n.oos.. n.nnu.u n.nns.u o.Oss.n n.sno.n n.0un.nu n.son.n n.unn.n uo\oonn an.u o.nuu.n n.osu u.uOu o.nnu o.nno.n n.4cn.n u.nom.u o.nun.u o.ouu.q o.nnn.u o.uno.nu u.uon.o u.une.n ss\nonu o.oun.u nousonu nu.n o.ono.u n.nuu n.nnu ..non n.9un.n n.nsn.u o.uon.n n.nnn.s o.nnn.n s.unn.u u.nun.n s.nno.o n.nuo.n sounoon nu.n o.unn.u u.sou n.onn u.uun eco.cno.n n.uon.uncn.unu a.noq.n e.non.n o.ono.u n.nnn.n u.neu.o o.uun.n no\uonu on.n o.umo.u n.onu o.son n.unu u..un.n u.noo.u o.nuu.n u.unn.n q.Oun.n n.nns.u u.unw.n u.smo.o o.nnu.n uo\nonu nn H o.nuw.n n.nnu o.nn u.Onn o.onn.n u.omo.u o.oun.n o.nnc.n n.onu.n n.0en.u n.oon.n u.nno.o o.uso.u uo\oenu use»... o.nun.o u.uOu n.ns u.nn u.uuo.n u.uoo.n o.nun o.nou.u o.noq.u o.nnu s.uon.n u.ncn.u n.nen.n econ. su.n o.onn.o o.nun u.nu u.uOu n.uuu.n n.unu.u n.nmo.n o.osn.n n.onn.u u.nnu.u n.nsu.u o.nnu.e n.nwa.u nnnu n.n o.nun.o n.snn ¢.an ..on n.uou.n u.ouu.u a.Ouo.n u.¢oo.n u.eee.u o.cou.n o.unu.n a.n0u.e u.uon.u anon no.n o.uuu.o n.onn u.uo u.uu u.uno.n u.uun.n a.uuu.n u.qnn.n n.nOu.u u.nuo.n n.unn.n n.nno.c u.uoe.u unn— un.o o.uuo.o u.oqn ..nn u.oo n.9oo.u u.onu.n u.unn n.non.u o.nun.u u.sno o.oqn.n n.nnu.n n.nun.u onnn no.0 o.cnn.n n.onn o.un u.uu u.unn.u c..ou.u o.unn n.unu.u u.uun.u u.uu» n.one.n u.nno.n n.nonun nnau c.nnu.n n.nun u.on n.nn .<.z .<.z u.uun.u c.nso.u u.noo snau oo.o c.unn.n u.uuu o.n. n.uu o.nnn.u n.nuo.n n.nnu n.oon.u. n.ono.n n.nuo u.¢ua.¢ n.nnu.n u.unn.n nnou un.o o.nns.n n.uun n.no o.nn o.oun.u u.nnn.u n.nnn u.uun.u u.unn.u o.eeu n.oou.c u.unu.n u.noo.n unou un.o o.uun.n o.uuu nuco.cn o.nn o.uu..u o.ono.u u.uuu o.oun.u n.nnn.u n.nnu o.nun.c.. n.nnu.n u..no.n unon no.o o.uau.n u.uu .<.= u.uu n.noo.u o.ono.u u.unn n.non.u n.¢on.n u.unu n.noo.e o.oou.n o.nnu.u anau no.0 o.muo.n n.nnn n.nu o.nu u.uno.u n.0cn.n n.nnn n.neu.u o.un..n u.uun u.uua.n ..onn.n n.nun.n noon nn.o nnsn.non.n n.¢u o.n u.uu .<.z .<.z .<.z n.nno.n~...non.n n.unu.u o.neo.n n.oon.n n..nn.n unnn no.0 .w.n.noo.u .<.z .<.z .<.z .<.z .¢.z .<.z u.uu A, .<.z .<.= u.oo nunu .unxoo-uoa non condo cannon uo acnaanuon nouoH.: «quad .aooa «ouch .= «quad .auoa nouoH .a nouns .noo: Hauoa .9 Anna: .nooa no on .o>< vounlnuuu u- nouu-uaaom «ouch snow-nalv nag en . guacno.oO cacaao~.ov ccaaun~.9v and NO.N no. no. CO0.0 anh.O ano.~ an.nl On nu> O04 nu ¢¢Ah¢0.00 cac~¢o~.0v c¢caom~.ov .04 no.d 00. OO. OOH.O ~n0.0 an.d nc.nl ON nu> and on cu~n~c.~v cannoo~.0v uccaona.ov nag on.H DO. 09. OhH.o Asd.a OON.H "u.uu! ON nu> OOH ON OCCAAO~.OO «CCAOO0.0V non Oh.d no. so. nn0.0 nun.“ «0.0! ON nub .04 on a0-.¢v cucaeoa.6v ¢«cAnoo.oO oo.a no. no. ann.O Hen.c an0.0 On.do0 ON .09 OH accaaoa.0v ctaanuo.0v «Ccanco.cv AO.A 00. 0O. cO0.0 nonoo nNO.n «M0.0 Nn.nnu ON .99 s .A~c~.0v «canc~.nv oceanoo.ov nu.d no. so. O~0.0 «on.n . «M0.0 NO.OOI ON .0. o 9. «gnaw kuduih And-OI o n .38 I m m «38. .o 3 £1.- ouao» and» co oouuuuuu-u no unoam 2: 18... .8 .38: .33: .38.. .31.- H... 9.31:. u.uuan no. access quay: uaiuusuqu gnu 6.: 13.5 328 «338 .38 we .38 .9 £388. 53...» ...- nnaouuouah .3.0 an «a .ube .utou .ubou aouuuononm noduuoaouu aquou d I uaouaoaon .ru .325 51... uou unnunn ago-auopou uaouusuau oAu «aquuuuu< cacao-u gnu no nuns~¢a< sou-oou-ou .m.a.o «o nuns..- u~.n nus-p a 535...? 311 Supporting Tables to Chapter Six D.12 Results of O.L.S. Regression Analysis of the Demand for Employees in Uganda Using 1960 GDP Series Data D.13 Results of O.L.S. Regression Analysis of the Demand for Employees in Uganda Using 1966 GDP Series Data D.14 Estimated Health Expenditures by District in Uganda: 1968/69 312 ~0m.~ 000.~ ~0¢._ ~00.0 000.0 0—m.— p0~.— Nvm.0 00n.0 «cm.0 me.— «on u FNO. 0mm. mnw.- 050. 000. 00v.- moo. ntn. 000. 000. 0N0. 3.00<> L ~0~.n 0mm.n nm—.m omm.n awn.m v~—.n .o.~ ooo 050.~ moo.— 0pm. ~o~.0 Nap.— -s.0 m——.0 0mm.0 00m.0 < ”K., m_i.o < om._ oaa.o a ¢_.F ssm.o o m_._ a...c a ,o.o om..o a mo.. .ai.o a ¢~._ mo~.o a .o._ ~m_.o < an.~ _o..o < mo._ ooo.c ‘ -.~ .Nm.o xv Nu Ibo; disequousm om~.0 «no.0 0~0.0 -m.0 5mm.0 0—0.0 ~v~.0 oo~.o ~0v.0 nn—.O 50m.0 .oo. asamo.o. .m_. Apooo.ov cos. Amamo.o. am.— Am¢o~.o0 _m_. .Nma..o. _mm. .moao.o. mNC. A.mao.o. ~o~. Aoomo.o. 000. A0v0~.00 _op. .mms..ov ~m~. Amvo.o0 050m.0- ooo—.0. mvn0.o- spam.0- -~N.Oa on—0.o- 000—.0 nvO—.0- OOQ0.0 o~n~.Oo nm0.0- : Na mooo.v .Ne.~.o. mooo.n h~o~,.o0 «ooo.o aseip.o0 G's. Aomm~.o0 ace. “mowo.c. moon. Aaoop.o0 mm.c awcmp.o0 Nmo. iono~.o0 coo. .ooon.ov ~o~. .a¢a~.o. n~0. Am——.00 owe—.p ~0—.p 0K0. ~0.0n~ muoc.— 055.0 mmo. v~.—e 0m-.— mc—.~ s—0.- wa.o~ 0mm0.0 nwm.— «50. No.0 va0.— mm0.— 0mm. mn.o pomu.o vso.— vmn.- ~m.nn Nwmn.0 nmn.0 ans. ~m.0p oo~o.0 ~0—.— mnn. 00.0 ~0m~.—u 0.0.0- _no. ~0.m ~0~m.0 Om—.. mww. mm.n vo~.0 0nn.— mvo. o~.0¢ oo¢> . .n . x os<>t ( .\ I 00.00~p ~m.qnm~ mm.—mm— 0c.¢mm~ Pm.v~v~ 00.00v— -.0~mp 0~.000— m_.~mnp 0—.vmv— o—.—oo -.0—— —0.~— c_.~ mv.0 mp.op 0v.m Mr.— 0v.~ pv.~ m—.—m com, ououo>a ecu ous.-ounce coauaauauov o~n¢uu¢> can No ecu can hope ocua och Rev OH.o oanlh 0 «N.— 0m0.0 < m0.— mno.0 0 mp.p nmm.0 0 F~.F «ov.0 0 00.0 mm».0 0 ~¢.— 000.0 0 nn.~ —on.0 0 n0.~ 05F.o < co.— ~0n.0 < m0.~ ~v0.0 ( 0~.~ mom.0 xv NW uhuusvsw ow cacaoHaEm we hogan: I u vu>o~nau you now-3 I 3 .ousua ooaa a“ .uaaaasg. no cove. usau> . oo<> Ame cu uuuoe 00¢ nounsou coon was acauouuufiuoao aouoa uo oaaeuov you A~0 uncuuu>uunao can: adv “nous: 000. Foo. ¢o~.o Respo.oe also.o- i.emm.o0 .om~.~ .Fo.mm sauces” or.ECi huu.>uom zupoo: noov>som eovuouavm ~¢_. mooc.v mou.>tom asa.o appoo.ov ~_oo.o- iamwm.oe mn_m.n oem.o “sooc.Fpuuwez .io. mooo.v omo.o Aacoo.o. _~oo.o- Ammo..pv oc_a.v_ _o~.~ “cascto>oo _.uco mB. owo. :33:ch mam.o Amoco.ov ..oo.o- Amman.o0 aoaa.o ~o~.o vco “toamcote wm~. cos. 50¢.o A__oo.o. NFoo.o- imoon.oe om~o.o o_~.~ ourgse°u Koo. mooc.v cm~.o ao~oo.ov noooo.o- AONoa.ov mmnn.m cao.o =o_uu=tun=oo pno. 000. . ocvsauueeacor m_m.o imnooo.o. o_oo.c Ancpm.o. naam.P wmn.o naoo:.__oUm.x mam. Nee. “nuances coed Fm~.o ANoooo.o0 .mooo.o Ao~vm._. n_~m.. sea.” .0 meson..==.: moo. m00. oc_»uua:0 Na..o Ammooo.ov o.oc.o Aco-.ov .N...~- nnm.m ac. ac.=.z mvp. n- oewucax so..o ieooo.o. oooo.o- i_wmn.o0 mne..o mom." o=.;u_u “noted PFn. “No. n~s.o ,noo.00 moo.o- .mpp.o0 aw~.o uvq.on «rau.=u,to< a s ~¢ ~n wm< - stunavc_ < .< 1\ luau HiccuuunSL anon“; nuoo nouuom mac coed cad-0 Ivan-0 cu oooboaalu new muslin one no sausaoa< ecu-nouns: .m.u.o no snagged NA.O canch .O Kunilhhd 313 —m0. pr. msm. 000.0 3.00(» mwn.~ ~0—.F —n.p «no.— 000.0 ~P0.— sac.— mom.— omo. 005.0 mmm.0 ooo.— 00n.n nmm.n N—m.n nom.n 00¢.m ~m~.n ~0v.n onm.n —0_.n —om.n chm.— nnm.~ N—n.~ 000.~ 000.— 00n.~ man.— 050.— was.— w¢p.n .3 N0. 00. ms. Nu. n0. 0 ~—. No. no. 0 0—. < ~—. mm. 0:. no — <¢o.o — 0.0.0 — oqo.0 — ~v0.0 n 0vm.0 — 000.0 m opo.0 ~ nnw.0 — 000.0 N «00.0 N 000.0 — nn0.0 nnm.0 poo.0 000.0 F00.0 mm0.0 n~0.0 ~00.0 m00.0 vn—.0 ~0¢.0 0m—.0 cwn.0 one: sown-a ~00 ecou ua«-0 ovals: ad cookouolfl you uni-in on» «o naosnon< sou-nouns: .m.g.o no suds-cu va. Amon.ov was. Aooon.ov owv. iqmm~.c0 ONN. Amcm~.o. «as. .~o~n.°0 cue. Aon-.o. was. ap~o~.o. owe. ace—m.o0 “mo. A¢_F~.o0 n... ..Nc..o0 o-. Ammo_.o0 ad unonqu _mn.0- 00~—.0o F~v~.0- c00n.0- ~000.0 —vm¢.0o mm0~.0I 0m0¢.0 000v.0 swcn.o- vn~—.0 ~o ‘ no_. iahm.o. _o¢. inmm_.o0 ooo. ao~_~.o. v.0. Aonem.o0 sac. a..mN.o0 so». .~__a.o. ~oo. ..~m~.00 coo. chmn.ov mom. Annvn.00 _o~. Assam.c0 moi. isms..o. .—.. in~._.ov 0——.~ 0000.0 00'0.0 «mm—.— -00.0 ape—.0: Ov¢~.— mo—o.~ ~om—.0 anv0.0- pmmp.0 m0~—.0 u¢<> —n is mmm.0u n0—.0- 00m.0u Nnm.0- o~o.0 0-.— ~_0.0 ~—~.0- O—o.0- moo.0 oon.— n00.0 «no. 500. amp. and Ian—OH O Kan—mum: 0~.vm~ 00.n— mn.—~ no.0v 0n.0— ~—.—n 0m.0~ no.0. ~0.v0<~ 10.m0mm mm.mmnn n—.mm~n 0F.00~n —~.~m0~ «m.mom~ n0._m—~ v.m0¢~ n.mam— n~.00~— 0.0v0m 0m.co c¢.n0— 05.m0m —p.p—N —~.mo~ mo.v0 mm.nm~ -.m~ cc.o~ -.vo ——.m—vp some caduo>o ogu ausoaouniu abuunnwuuou can-«hop can no ecu can uo>o «cad ask 0 cc." 0N0.0 < om.p poa.0 < ~m.— vns.0 < —~._ 000.0 0 ~—.n 0—0.0 0 sn.— 000.0 0 00.~ 5.0.0 < 00.— p~0.0 0 00.0 000.0 < pn.~ 00~.0 < n~.~ 000.0 < vm.— v00.0 z: un uocaueouuu yo —o>u~ xm oocnuaouua yo po>op mm «is. ¢n_. mm~.o immo.o0 o~0.o- ammo.o. «mo.o oa~.°~ VNm. «ooo.w -a.o ip_oc.ov ”ooo.o- .hmao.ov an,~.o Noo.. _co. ~om.o Anemo.o0 omm_.o -m.~ Dem. moo. m_m.c awmoo.o0 .Noo.o- im_no.o. own—.0 on».— ..N. _mo. ..~.o Amoco.ov oooo.o- A¢_Pc.o. auwo.o «om.o m~n. can. mop.o Amoco.oe oooo.o- .mnoo.=0 meoo.o- _oo.n_ use. _cc. mam.o hovoo.ov omoo.o- im~ho.o0 PVN..o mom.~ ooo. Koo. noa.o iw~oo.c0 "moo.o- .omno.o0 na...o ohm.— ~¢m. was. .__.c iv~cc.o0 mpoo.o Ao~.,.ov mooo.o mso.. coo. mNN. nhc.o Ascoc.o. ~_oo.o A~.¢c.c0 ammo.o. cos.» «on. o_o.a ma..oinmooo.o. «coco. - Ammwo.c. .n_o.o _o~.v -~. can. no~.a .omco.a. ”moo.o .mcoo.°0 m.oo.o ~oa.mn x oe<> a a . w a. . luau nqaowuuaau and; Aav shunsvau a“ nouaoaolo mo hogan: I m cohoHnau you now-3 I 3 noouua coma ad nuauauunn nu vovuu oaan> I oo<> Anv 0H.o Qua-H ou oouon ooo usuuao- no.0 van neuu-ouuwooeo devo- uo nauauov you ANV adv "mayo: .8“ HIEOODO OUE Lao—ucana co.»~—onsoo ouaa u>.uuouz - o: uuohaauu covua—ouuou qua. o>_u_moa . o: uuonomum cc—uapustou case 0: - o: uaouoLum sup-o: «ou'>som covu-uaum “cotton . on E accessosow Paood cowuauvcasiou can «Loamcugh cutaneou co.auatuacou ocvtauoopacox .umr: nauseous wood eo ogauuoeacox u:.autoso ac. m=.:_z ocvucsz .oc_;m.u .atuuotod «gnu—au_so< haumavc_ .31u4 «0.5 00.0— 00.50— v0.0mp mm.~— ~m.0 o..m Nm.0 0~.5 58 ..2 . .u_a.o\.gm 0—.m 0v.~ 0P.m <0.— 0v.~ uu_0 mm0.mo~.o ..¢.c.m.¢ acm.~m oo..omn moo._mn eim._5m m._._om mmo..¢o cmm.~mv.~ 050.00 ~05.M50 m.m.v00 000.m0v 500.00w 000.000.— 0N0.05m voq.eo 000.5N0 omm.5om m5m.0om oa¢.cmv.~ 000.000 000.0v0 00v.m_m ram.~00 ~m0.0mm.~ 000— «cameo; mp0.c<0.—5 000.00~.0—— -..o~5.~ awo.¢mm.on mom.~mm.c ~Pu.woo.. ~._..~... mma..-.. 0—~.mo5.5— n00.000 05¢.0Q0.M ch.mov.v 0~0.0m0.0 m00.00p.~ 00¢.0m5.0— 000.000.. mm0.N—5 cmc.v-.v m00.m0¢.0 m05.005.v p0m.0mm.0~ 00~.500.~ 0.0._0m.0 ~m—._~0.n .mu.mm~.v 000.00~.5— ouo>.gd vco .u>oo .auO» acouuouosa “00.nom.. 5¢..m...n 0vc.00~.— 005.0mm 005.—5~ 5—0.50— p0m.055 .000.-~ 000.—0— me0.m5n .00.~vv ~00.~vv ...E.amo: —~_.000.— .N..aoo.P mom.—5 ~0m.50 p00.0v— 005.00~ ~0—.~v~ 00¢.——— 000.0m 000.~0 «50.000 050.00~ ~F5.5~ 00v.m_~ 0N5.¢0 0~n.00 00n.000 n~—.50 .00.00 05m.0_~ 000.0NN 000.0—m ..a..o a ..u: u..o...u 500.~00.0 -0.~c—.0 0—0.0~0.n 0q~.omv -0.00 50m.m—0 500.~m0.— 000.0em 000.000 505.mm0.— 000.000.— wm—.~0~ ~0v.0mm 000.5N0 0.0.50o._ «mm.oMn mac._m~ m_s.mmm.. 5.2:a0 aa:«-«:m savanna a“ on: nounuuu aa< 0—5.500.n0 0.0.00~.c I 50—.0mv.v~ ~00.~nm.em om—.~5—.0O— 050.000.0—I -0.055.5 n00.vmn.0n 00~.00~.o 0_~.—cm.m _5~.mm0.n an.o¢0.m 000.500.0— m00.000 00¢.0m0.~ 0vm.~00.m ~5_.000.m mmo.—00.— "~0.5F5.n5 005.505.n —v~.000 5—0.0¢0.n 0m~.0—m.0 050.00—.v «no.o_0.0p m...ov¢.~ cm..ocm.m mm...sn.n ~¢m.¢.5.m mam.m...m. «Pauwamo: ucoecto>00 u..ozoau pouch 000.000.n 00—.0~0.0 .)\. ac O5~ 0m0.—0~ ~05.00— 050.500 050.000 5m_.- «N5.05F 00~.005 5m~.v5~ ~00.m~p 000.000 000.05N 0m0.m- 5m0.n05.5 m0m.n00.— 000.000 005.0NN 0N5.55~ 0—0.000 .u.c== 0cm nova—u oo\ooo~ u.uu-u: a. gouty-«a an nouauweaoauu sundae e.g.-«gun c«.O OHAQH 0 xnnzmum< 50—.00¢.c~ ~mo.m¢¢._ m.m.voo.. cmm.m~m.. oss.m~o.. c~...om.m ~¢—.wan mm~.n~c.P ~0q.m-.p ~__.q5~._ m.5.~.o.P ows.omm.m 000.0~_.~ 5mm..0~ mm0.pm0.p —0~.000.— «c—.m05.~ 000.~—0.5 m0~.—00.5 ~0—.05m.— 00~.0N0.— 055.000., n00.050.0 .=.su< uuptuuvo .u>o0 F-tucou 000.000.00 -¢.omm.e nom.nmn.m~ .wp.5on.~ Nom.vam.~ am~.m¢m._ ~ms.~m~.~ com.w-.m c~¢.amc ss¢..~n.. «mm.q0m.~ nom.moq.~ 5s~.m_m .55.e«m.5 mm~.00n.— «00.mm0 0cm.pr5.— ovm.0——.v ~m0.55m.— c00.000.0 moo.ww5._ meo.m~_.n .«m.¢_m.. ev¢.~5m._ 5e5.no..m ...u_a.oz "0002 ..c.fi . ..x .puxm Pogoh «page» ohcvn .53 80%: 3m ago» o_oxc< .Nou.x on»: _u~x 0.02 .3 00cc; _Po;o< whosotmx gyro: emu» venom wumxam :m_cam omomam uuau unnunn: «some: 00cm: .3 coco: .0 nucuuam APPENDIX E Notes to Tables Included in the Text E.I Notes for Table 2.8 E.II Notes to Table 2.9 E.III Notes to Table 2.10 E.IV Notes to Table 6.11 315 316 APPENDIX E.I Notes for Table 2.8 E - Estimated on the basis of expenditures of recent periods. E-l - Estimated on the basis of partial figures. Kampala and Mbale figures for the year 1969. (1) Excluding expenditures on environmental health services such as water supplies and sewer systems. All figures include recurrent and capital expenditures. (2) The figure for central government expenditure on health does not in- clude health expenditures made by the Army, Prisons, vocational rehabilita- tion schools, of the Department of Labor on occupational hygiene. An estimate of the health expenditures of these four units is difficult, at best. A figure for the Army and Prisons can be estimated from the number of beds in facilities they Operate, assuming that the average cost per bed in other facilities is similar; this figure, in the case of the Army, can be corraborated by a constant percentage estimate of total expenses for purchases of drugs from the Ministry of Health (this would be a minimum estimate, because the Army very likely purchases some drugs and equipment from private chemists). A minimum estimate of the Department of Labor's expenditures on health can be made from establishment available in the annual estimates. Figures for operating rehabilitation programs and schools for the handicapped are available in the annual government estimates. The estimated recurrent expenditure for health made by these different programs and departments is shown below for the year 1968/69 (in thousands of shillings): Tl) Army One hospital, 35 beds 220.00 Six sub—dispensaries (estimated minimum),_assuming 0 beds 200.00 420.00 (2) Prisons 10 units, with 141 beds_jl966/67) 380.00 380.00 (3)'Ministry of Labor - Occupational Hygiene Section Eight persons 120.00 Other expenses 40.00 ‘T' 160.00 (4) Ministry of Culture and Community Deve10pment - Vocational Rehabilitation 752.00 752.00 (5) Schools for Handicapped Children 87.00 87.00 Total 1,799.00 317 (3) All Urban Authority expenditures are for preventive health services and are related primarily to environmental health services, such as malaria control, public conveniences, rodent and pest control, Abattoir cleanliness, and refuse collection. (4) Data for government expenditures are derived from various issues of (a) Uganda's Statistical Abstract, (b) Government Estimates of Develop— ment and Recurrent Expenditures, (c) District and Municipal Estimates and Financial Statements. The latter two sources (b and c) were used for estimating the most recent year's expenditures. (5) These figures do not include the cost to mission organizations of high level expatriate personnel or gifts of drugs and equipment. (6) Estimated from recent figures and corraborated with data available in Health and the Developing wogld by John Bryant (Ithaca: Cornell University Press, 1969), p. 267. (7) The estimated figures for 1958/59 and 1963/64 are speculative and were made solely on the basis of (a) whether the service existed and (b) in the case of Army and Prisons, some idea as to the number of beds in the facilit- ies. (8) Data taken from W.H.O. official records, financial reports for various years, and from UNICEF (Regional Office), Progress Report 1969 (Kampala, 1970). Expenditure data were converted from U.S. dollars tofifiganda shill- ings at the official rate of exchange: U.S. $1 = 7.14 U. shs. (9) Expenditures by industrial firms on health services were estimated on the basis of an estimated average recurrent cost per bed per year for each of the three years. This estimate, thus, is solely for medical care ser- vices and does not include the cost incurred by the firms for preventive services and plant safety. In 1968/69, the average recurrent cost per bed in a lOO-bed government hospital was approximately 10,000 shs. Corresponding figures for Protest- and and Catholic mission hospitals were 5,150 shs. and 3,090 shs., respect- ively. The mission figures, however, do not include all of the costs of expatriate skilled manpower, particularly in the case of the Catholic units, or the value of donated drugs and equipment. Using these figures as a guide, and considering that very large costs are not allowed for by mission facilities at present, a figure of 8,000 shs. per bed per year was arrived at and used as an estimated cost figure for industrial firm medical care services. The number of industrial firm beds for the three years is found in Annual Reports of the Ministry of Health. The average cost per bed figure was adjusted further for rises in the price of resources. From 1961 to 1969, retail prices in Kampala for all income classes rose 442 (see Table U0 1(a), (b), and (c), Statistical Abstract 1969). In terms of increases in the private medical sector, wages rose 49% from 1959 to 1968, and 271 from 1963 to 1968. Using these figures as a basis, it was assumed that the average recurrent cost per bed rose by 40% from 1958/59 to 1968/69, and 252 from 1963/64 to 1968/69. The figures, thus, are derived as follows: 318 1958/59 350 beds x 5,700 shs. per bed 1963/64 395 beds x 6,400 shs. per bed 1968/69 477 beds x 8,000 shs. per bed 1,995,000 shs. 2,528,000 shs. 3,816,000 shs. (10) An estimate of the expenditures on specific health projects funded by internal to international sources is very difficult to obtain. Some of these organizations assist in medical research, others fund special pro— jects (such as the Ankole Preschool Protection Programe or the Uganda Foundation for the Blind/U.S. Peace Corps National Campaign Against Trachoma), and others assist various Ministry of Health Programs (such as the National Immunication Team or food supplements for small children). For the earliest two periods, 1950/59 and 1963/64, an estimate is not attempted, but the fact that a certain level of expenditure did occur at these times is indicated by a token 0.5 million shs. For the latter period, a minimum estimate will be made from available data. The expenditures for some programs or organizations are not available to the author at this time, so that this minimum estimate is subject to upward revision at a future time. The following list of organizations and the specific program funded (if any), and the level of expenditure by the organization is indicated below: Oxfam Ankole PPP; medical research 255,800 shs. (PPP) Uganda Foundation for the Blind Trachoma campaign 119,182 Uganda Red Cross Save the Children Fund Swedish Red Cross Refugee Health services 70,000 (min) U.S. A.I.D. U.S. Peace Corps Trachoma campaign 1,000,000 ($7,000/ vol. for 20 vols.) British V.S.O. Protestant Medical Bur. study 15,000 Nuffield Foundation ' Research Obote Foundation Nutrition, polio immunization 100,000 (min) Estimated minimum total 1,559,982 This list is not exhaustive. It does indicate, however, that such sources spend fairly large sums on various services. Additional research in this area is indicated. Private Family Planning Services will be ins cluded in any revision made. (11) Data is taken from Table UM.15 of the Uganda Statistical Abstract for the years 1962, 1967, 1968, and 1969. The assumption was made that one- third of the medical research and health expenditures of the East African Common Services Organization was spent in Uganda. (12) Private consumption expenditures on health were estimated as follows. Over the ten years since 1958, the Uganda Government has taken periodic expenditure surveys of unskilled African workers (and coffee growers in the Buganda Region) in various cities through the country. One expenditure 319 item classified for the survey was medicine. According to the data collected on total and medicine expenditure for 26 classifications of workers (classified by income and social variables, such as tribe), medical expenditures comprised 1.54% of the total expenditures by workers. It is recognized that other socio—economic groups (either EurOpeans, Asians, or Africans with a higher or lower - i.e., sub- sistence - income) may spend a greater or smaller percentage of their income on medicine and other health items, but this estimate is taken to be a first approximation. Given this figure (1.542) the total private expenditure on health for the year 1961 was estimated by multiplying it by the estimated total paid employment compensation (wages) figure in gross domestic product estimates by source of income. It is likely that a certain proportion of the rent and profit component of national income is also spent on health, but there is no evidence available on which to base an estimate. To make estimates for other years - before and after 1961 — the following equation was used: d1 - p + y n1, where d a the rate of increase in demand for commodity i (i-health), = the rate of increase in income (total paid employment compensation was used as a proxy for this item, with awareness of potential problems of shifts in the dis- tribution in income or in the proxy variable, and the problem of failing to allow for price changes. n1 a income elasticity of demand for commodity i. p = rate of change in total population. As an estimate of population changes, 1959 and 1969 census figures were used. (1959 figures are available from published sources and the preliminary figures for 1969 are available from the Ministry of Planning and Economic Deve10pment). During the decade of the 1960's, the rate of natural increase of the population (excluding net in-migration during the period) was 3.3% per year. For an estimate of the income elasticity of demand for health service expenditures, a value estimated from recent Kenyan data was used: see Massell, Benton and Heyer, Judith, "Household Expenditure in Nairobi: A Statistical Analysis of Consumer Behavior," Economic Deve10pment and Cultural Changg, Jan. 1969, pp. 212-234. Massell and Heyer estimated five values for the income elasticity of demand for health expenditures, using several different functional forms, and the values varied from 1.07 to 1.42, with three of the values around 1.20 - 1.25. A simple average of the five estimates (1.22) was used for estimation purposes. Because the definition of health services (from drug purchases to private medical care to the services rendered by traditional healers) varied in scope (primarily in the narrow direction) in the several studies used to obtain estimated values of the variables required, and because the proxy variable used for income does not allow for expenditures in health out of profits (surpluses) and rents, the estimate is a minimal estimate. 320 APPENDIX E. II Notes to Table 2.9 (l) Expenditures are reported in millions of shillings. (2) Non-direct services includes administrative services, manpower training services, and research services related to delivery of curative or preventive health services. (2) The estimation procedure used to allocate Ministry of Health ex- penditures involved (a) analysis of each line item in the Ministry's recurrent and capital budget for 1968/69 and (b) deletion of those items which were related to preventive or other health services. The estimated expenditure on curative services is thus a residual estimate. Those line items which included expenditures for all three categories of ser- vice - such as personal emoluments, transportation, office, and miscellane— ous expenses - were segregated according to the percentage of total Ministry employees directly employed in delivering that category of service. A further check was made of the relative distribution of per- sonnel according to pay scales to determine whether there were significant differences between the three groups. There did not appear to be any major differences, so that for a first approximation these line items were allocated according to each service type's percentage of total em? ployment. The percentages for 1968/69 were as follows: 8.25% for pre- ventive services, 9.25% for other (non-direct) services, and the remaining 82.50% to curative services. With further analysis of expenditures at the hospital level (which will be undertaken in the near future), a reallocation of the percentages could occur. Such a reallocation would increase the percentage of the total manpower conmitment engaged in preventive services. At this time, the magnitude of this shift cannot be predicted. Therefore, the figures for preventive services should be viewed as minimum estimates and correspondingly the curative estimate is likely to be a maximum figure. Within a year or so, budget estimates for all Ministries in the central government will be presented on a program basis with line items within each program. This budgetary procedure will facilitate the analysis undertaken in Table 2.9. (4) For a first approximation, the expenditures of two districts (Ankole and Busoga) for 1969 were analyzed by line item, including personal emoluments, and the percentage estimates of the two districts' recurrent health budgets used in the delivery of the three types of service were obtained in this way. The results were as follows: 12 to other (non-direct) services, 312 to preventive services, and 68% to curative services. These percentages were applied to the total 1969 recurrent health ex- penditures (estimates in the case of ten districts) for all 17 districts in Uganda in order to obtain the estimates presented in the Table. It was assumed that all capital expenditures were related to expanding the 321 curative service system. This assumption will subsequently be checked and some shifting is likely to occur (particularly in relation to new health center construction), but the amount of shifting is unlikely to have a major impact on the conclusions to be drawn from the data pre- sented in Table 2.9- (5) Expenditures on curative services were estimated directly from the municipalities’financial reports. The remainder of the municipalities' health expenditures were related to preventive services, primarily of a societal nature. (See Appendix A. for explanation of the distinction between individual and societal preventive health services.) (6) The mission medical services were allocated between curative and other health services on the basis of 1968/69 recorded expenditures. The other health expenditures were for medical manpower training programs (nurses and midwives) operated by mission hospitals. On more thorough analysis of mission hospital data, expenditures on preventive services will be included, thereby decreasing the total curative expenditure. (7) The allocation of W.H.O. expenditures was made on the basis of an analysis of the estimated 1969 program budget for Uganda, wherein each W.H.O. program or project is listed separately. In addition, an estimate was made of the total expenditures by the Office of the W.H.O. Representa- tive to Uganda, and this figure was included in the Other (Non-direct) Services category. (8) The analysis was made on the basis of figures shown in Note 10, for Table 2.8, Appendix E.I. Undoubtedly, the figures may change upon receipt of additional information. ' 322 APPENDIX E.III Notes to Table 2.10 (l) The period of time is dependent upon the reporting period used by the various units. Hospital data is for 1968/69; other facility data is for 1969. (2) Personnel emoulments do not include the cost of fringe benefits such as social security, retirement, and workman's compensation. (3) The estimated expenditure for the Health Center does not include the percentage of the total expenditure used to operate one or two weekly Aid Posts, which involves transporting of staff and some supplies from the Health Center to the site of the Aid Post. (4) Figures for maintenance and repair costs are not available due to the accounting proceedures used by the Department of Public Works, which has responsibility for these services. This figure is therefore not included in the total estimated expenditure for hospitals. (5) Skilled personnel refers to persons who have received some formal training (medical or public health) necessary to fill the position they hold. (6) Total employment figure for the large hospital does not include the 40 or more students in the training schools adjacent to the hospital who work in the hospital as a part of their training. 323 APPENDIX E . IV Notes to Table 6.11 (1) Not all industries are included in the list. It is assumed that those not included do not have significant purchases made from them by the health service industry. This assumption is based on an analysis of financial statements of the respective governmental admission sectors of the health services industry. (2) Source: an analysis of the 1968/69 financial statements from the appropriate jurisdictions responsible for health services in the central, districts and municipal governments and the Protestant and Catholic mission organizations. (3) Estimates are rounded to the nearest ten employees. (4) The estimated secondary employment impact in Agriculture was made as follows. I took the monetary sectors estimated value, added in agricul- ture (for 1968 and 1969) per Table 1.2 in the Republic of Uganda, Background to the Budget 1970—71, Statistics Division, Ministry of Planning and Econ- omic Development, Entebbe, 1970, and made an assumption that it represented 80% of total receipts by that sector. I divided each year's figure by two to get an estimate of receipts for each half of the 1968/69 fiscal year. These estimates were added together and were divided by the estimated number of total employees in agriculture for the fiscal year 1968/69 (by averaging the 1968 and 1969 figures in the same manner as described above), per the Republic of Uganda's Enumeration of Employees, June 1963 and 1969. Est. Receipts Est. Receipts Est. Total per Employee Ratio ingAgriculture (pill. shag) Employees (thousands) (mill. shs) 2092.9 52.85 0.0396 (5) The receipts-per—employee ratio was estimated from data in Republic of Uganda, Survey of Industrial Production, 1967, Statistics Division, Ministry of Planning and Economic Development, Entebbe, 1969. Table 3.118, p. 83. (6) Receipts-per-employee ratio was estimated from data in Uganda Govern- ment, Survey of Indistrial Production: Building and anstruction, 1964, Statistics Division, Ministry of Planning and Community Deve10pment, 1964. Appendix Table 1. Labor productivity data in construction was analyzed to determine the extent to which the figures should be adjusted to reflect the 1968/69 situration. The annual rate of increase in productivity over that period was about 0.52 (7) The industry receipts from health services is estimated as follows: 324 mill. shs. mill. shs. (a) Drugs and Equipment 19.10 Less Direct Imports 2.47 16.63 (b) Misc. Commerce Purchases 0.28 (c) Transport, Petroleum and Related Purchases 4.72 TOTAL 21.63 The receipts-per-employee ratio was estimated from data in The Republic of Uganda, Census of Distribution 1966, Statistics Division, Ministry of Planning and Economic Deve10pment, 1967, Table A, IV, p. 35, and Table BIII,p.45). (8) The receipts-per—employee ratio was estimated as follows. It was assumed that the average labor cost per employee in miscellaneous services in 1968/69 was 6,000 shs. It was then assumed that labor cost constituted 502 of total receipts in miscellaneous services. On the basis of these two assumptions the receipts-per-employee ratio was estimated. Data on average cash wages in Misc. Services in 1968 and 1969 was approximately 4100 and 4500 shs., respectively. It is assumed that labor costs such as pensions, workmans compensation, social security and non-wage benefits comprise the difference. See Republic of Uganda, Enumeration of Employees, 1968 and 1969, Appendix Table XX. APPENDIX F APPENDIX F International Classification of Diseases Successful Treatment Followup Survey Form Uganda Government Medical Forms MF 74 MF 75 MF 77 325 326 APPENDIX F Table F. 1 International Classification of Diseases W.H.O. Major Disease Uganda Government Inpt. Author's Major Classification 5 Outpt. Major Disease Disease Classification Classification [1) Infective 5 Parasitic (l) Infective 5 Parasitic (l) Infective 5 I Diseases Diseases [ Parasitic 2 Neoplasms (2) New Growths (NeoplasmS)U2) New Growths (3) Allergic, Metabolic, 5 U3) Allergic, Meta- 5 Metabolic Diseases Blood Diseases F bolic 5 Blood [3) Endocrine, Nutritional, 4) Diseases of Blood 5 Blood—forminggopgans (as #3 above) (as #3 above) 5) Mental Disorders (6) Diseases of Nervous System 5 Sense Organs Diseases of Nervous 1 System 5 Sense Org. (4) [4) Diseases of the Nervous Sys. 5 Sense Organs Digestive System Diseasep; (7)4Diseases of the (5) Circulatory Diseases (5) Circulatory Circulatory System ' 8) Diseases of the (6) Respiratory Diseases [6) Respiratory Respiratory System (9) Diseases of the (7) Alimentary (Digestive) [7) Alimentary (10) Diseases of the Genito—Urinary»System (8) Genito-Urinary Diseases {8) Genito-Urinary (ll) Complications of Pregnancy, Childbirth and the Puerperium (97 Diseases of Pregnancy 5 Puerperium r9) Diseases of Pregnancy 5 Puerperium { ? 10)Delivery (Child birth) without complication (12) Diseases of the Skin and Subcutaneous Tissue (l3) Diseases of the Muscula- skeletal System 5 Con- nectingyTissue (10) Skin and Musculo- skeletal Diseases (included in #10) {11) Skin and Musculo-- skeletal (14) Congenital Anomalies K15) Certain Causes of Perinatal Morbidity 5 Mortality b (11) Diseases of the New Born (included in #11) [12) Diseases of the New Born (included in #12) (16) Symptoms of 111- defined Conditions (12) Ill-defined Diseases [13) 111-defined (17) Acclhvnts, Poisoninga, 5 Violence (13) Injuries (including _ppisoning) [14) Injuries Note: The table indicates the correlation between W.H.O. classification of diseases into major groups, the Uganda government's classification, and the author's classiflcntion (which deviates in only one respect from Uganda's classification). 327 Successful Treatment Followup Survey Form (A)(l) Health Facilty Name I l f for i Tl| L: l 1 4. 2 4 5 6 7 (A)(2) Patient Name 3 (A)(3) Patient Number (from record book) [ [I 'y i l i 8 9 10 11 12 13 (A)(4) Date of original attendance (A)(S) Place of residence (address): (a) Village (b) Gombolola (c) Other address information (how to find house, etc.[__ (A)(6) Distance in miles between health facility and place of residence Give number of miles 14 15 (A)(7) Method of travel between residence and health facility: (1) walking 16 (2) bicycle (3) taxi (4) bus (5) private car (6) motor cycle (7) other (A)(8) Diagnosis | I l l7 l8 (A)(9) Age: (1) infant (2) child (1-6 years) I I (3) school age 19 (4) adult (A)(10) Sex: (1) Male (2) Female L__l 20 s * * s s a * * s * * * * * *- (B)(l) Does live here? (patient's name) I 1 (1) yes 21 (2) no (a) if the answer is NO, ask "Where does this person live?" 328 (1) Village (ii) Gombolola (iii) Other address information (b) if the answer is YES, continue to question (B)(2) (B)(2) Let me see your medical chit (MF-S) E::] (1) yes 22 (2) no (If he/she gives you the MF-5, answer the questions from the information recorded on the MF—S, asking the person about it when necessary). (If he/she does not give you the MF-5, answer the following questions the best you can by talking with the person and asking him these questions). (a) Diagnosis or complaint (b) How many treatments were prescribed? Give number (c) Did you receive all of the prescribed treat- ments? (1) yes (2) no WU “$213 If the answer is NO, ask, "How many treatments did you actually receive?" Record the number here (d) Name of diagnostician 5E] 3D (1) Doctor (2) Medical Assistant (3) Dresser (4) Nursing Assistant (5) Other (specify) (e) How many days has it been since you first visited the health facility for treatment of this illness of com- plaint? Give number of days 29 30 Record the date shown on the MF-S for this sickness: (B)(3) What was your job before you became sick and went to the health facility? | i 31 (B)(4) 329 (l) farmer or herdsman (2) porter (3) clerk (4) private business/ taxi driver (5) professional: teacher, lawyer, accountant, doctor, nurse (6) other paid employment: service station worker, cook, factory employee (7) housewife (8) school (9) unemployed (if school leaver age or older) (10) pre-school age Other: please specify Are you doing the same job or work now? (If a child, is he/she back in school now, playing [::] normally, etc.?) 32 (1) yes (2) no (a) If YES, when did you begin your work or job here: Determine the number of days that he did not work or do his job from the date recorded above and the date of original attendance. Write down the number of days that he did pp£_work. .4"r'1 33 34 (b) If NO, why aren't you doing the same job or work now? (1) not well yet (2) no job or work available now [::1 (3) changed work or job 35 (4) on vacation or leave (5) other: please specify (c) If the answer was "Not well yet", go on to question (B)(S). If the answer was (2), (3), (4), or (5), then ask: When did you: find out that there is no work available? change your job or work? go on vacation or leave? begin the "other" reason specified in (5) above? Give date here Determine the number of days between the date given and the date of original attendance at the health facility. Record that number here [:1::] 36 37 330 (B)(S) Did you go to any other health facility or person (for example, see list below) for treatment of your sickness after going to the health facility? (1) Yes C] (2) No 38 (If NO, the interview is finished.) (If YES, answer these next questions.) (a) Did you go to another health facility? (b) (C) (d) (e) (1) Yes (2) No 39 40 Give name of facility or place here: Did you go to an Asian doctor or take traditional medicine? (1) Asian doctor (2) Traditional medicine (3) Neither SD Did you visit any other kind of health facility (e.g., Mobile health clinic)? L::] (1) Yes 42 (2) No Give name of facility here: How far in miles did you travel to this person or place? Give the number of miles :::I:] l 43 44 Did you pay something for your treatment from this person or place? (1) Yes [:3 45 (2) No (i) if yes, what was the form of the payment? (1) an animal (what kind?) I g (2) other gift (what kind?) 46 (3) money (ii)if you paid money, how much money did you pay? Write number of shillings here [ 47 48 49 What was the sickness or condition for which you were treated by this person or health facility? [::[:] Write diagnosis here 50 51 After you received your treatment from this person or place did you feel better? (1) Yes [::] (2) No 52 331 (f) After you received treatment from this person or place, did you go back to your work or job or to a new job or work? (1) Yes 53 (2) No (i) If yes, when did you go back to work or to your new work? Write date here Note to interviewer: (C)(1) Who did you talk to for this information? (1) Patient (2) Adult relative or friend i | (3) Child relative or friend 54 Interviewer's name Today's date 332 THE RE'UBLIC OF UGANDA-MINISTRY OF HEALTH "J- 7‘ nae-I RETURN or DISEASES (IN-PATIENTS) HOSPITAL "" ' C Imps-s For the month of I9...... “ m D ADMISSIONS DEATHS No. Md. Funk Total Male Funk Total INFECTIVI AND rAmITIc msmu with-008 u l unnunn-.- ol Raw-wry DWI-I DIO A.) 7 3. o! \1emfim m1 Liam-n) \dvvus SM 0" A.) T U. or ,“7flufi'l, Pcn‘m all Muslim Cindi 011.8!) LO 'l'abutuluua 0! m and Iona fin—on IL: Tubcrv'uulaz-E on... {onus 020 L6 Wm::J—S::hha .____ an A1 Early 99...... (I and H) .2. LI Tuba Dru..- — .33 5.9 COM Pun] ru- 0! Insane ‘ ' 03.2.02.) A10. Carine Vascular 57pm runs)» 4 III. an other smug. 030.0." A.Ila Com-xxx“: Ink-eons Cato-Um 01) All. Comma]h.!:1melhn moums a I I. 0;:JJJITITQQ. m (or. ——— m [.11 Tv;aitd Fever «1.94: In: Pu;r;p:¢s on! ma- M lat-anu- on A“ Chin W AIS ”natal-aloe (L'nd Aw Fm) .03 A?“ fistula" D\:cfif:ry o“ IL I» Anna-tn: ::I:d-nl 9mm“. cum-1) 007.0“ A!“ Other Umpoohed Omen .53 A.” $1110! Fever 051 A." Stnctml Son Tm 052 Al. Emma‘s .9 A39 Schema and m 035 L2! Orphan". 036 L1) WhoomM-aoush on L2) Mcmnmc: [aim 05' L24 Plant. .60 L25 1.49m" “I A26 Toumu 062 L17 Anthrax “0 A.II Anus Pohomnhus on A.” Anus Infamous mumm- sumo w Lu: cfiuu Panorama!» and [Whit 0&4 AJIC Smaller)! \anoh Mam: “4 “IO Smalls»: \‘mala Manor “5 m 310.“!!! e" A.” You... hm j— m “4 Idea-ou- Hep-hm W AJS RabIn IN A)“ Laue-borne Epamhc T795:- ‘9; A)“ Flu-borne lithium: Tyvhus 104 A!“ Tether“: Tvnhm :07 A)“ Hawaii-1;}... fi”. Ind-10' AJQ Olhfl Returns-d Unease. "0 L37. VI"! Nolan-(8.1.) III am um um". (or) u: an. rm mum M75551.) "Jan. I". In A 171 0.27.;‘EJEZ3‘JIAIuw us an. murmm 52:: m I;- “fiaa Sehnloulmsjum 1....“ I}: I... ";.7.:..T.......'.... I..'.4.....I IA...“ -.7 H n...”..‘.I‘I;;.;.’ - In A on. MM;;'-::o_ ”1;?- ” A in} '"F..’....“.‘..“:.’...T.Z...I 3:?"- .. "'11:“ "I I.:..'....’...: ..I" I 4.4.: 0...... "”73; ”Tu: "MIC." ”.331." “IS. 7 2 ii “We; LIN“: 02.0—- Alld 1 ous.-da.": Y m‘ “7 u. '7‘..:.'.:..:““ .' I 1):») an: Gum's—worm 31.01%.an an: on... ILL-am!!!- J an A u. LmKJQJI Vsnu‘u'm (:1!!!) on Lu; QMJIQ-Exfi; Vanna-3 on. an L4): (Mm and t:;~:vi_.5:'cmm Irma- 10!“. PA“ I 333 THE REPUBLIC OF UGANDA—MINISTRY OF HEALTH H.,. 7‘ m Page 3 OAe. RETURN OF DISEASES (I N-PATIENTS) ................................. HOSPITAL a leap-n For the month 0! ' . -. l9..,... “- Lin 0! ADMISSION: oeA'rus No. lIele Female Total Male 7m Total Inna-m- m PM W —T on AOL! Food Panama. “mum. Selmlle lam A.IJ) m ‘410 Relapsing I’m — .71 AA)! Leptoeomoue (Wed'e Dues-e) TT .7: AA): Y.-. m LOH! Rubella m AT). ous...-”- *— ul A41; Hm z..." 0” AJM - Mumps ”S A.‘ II Ted’vflu I” A‘l- Lush-nun-" In A0). Twp-m or A CT). I‘mnafephnuue ITumI I)! A Up TSTTIM. T——- N.Oi.050—IJI Air All other lube-nu and anme Ozone. . 5E“ CRO‘VI'HS ‘40-!“ 4L“ Holman! lenm 0! “out and Plasma I” A05 M Ila-hem Newman. of Omaha”;- ISI A 00 .\Ie.' lenem Stools-m- ol Slmh III. I” A47 .\Ie.‘..T-nam Neoplasm Te! TTnce-une I“ A a MslITM Novel-am d Rectum [.1 A19 Swans!" Ker alum—o! Lama I‘LIU A.;!) Mahmm Scout-m d Tm bee limbs! and lawn. (not m” no A“ Hahn-M Seoul-um of Drum In LS! .‘KInmenl Neoplasm 0! Con u I. trn‘ 171-110 A.” SILIf-wanl Scot-Java at on." pom or (Inn. 377 A, so Shhqnw .“.),:Mf;\—OI Pvmxue Ice-1, m -1 A 5:. Mm“... .xnp'u...“ ol 5... 0! Ln no.3. I90. m A; 3: 5mm... 5mm»... c! 5.... «I... m... I," no ASST Mot-mm xmsxumou... us, "1 A5» TM: .snam xéazz «3... Bone and Cam" 1'... III-I A 574! MTtlnrnem Nee: mum of Lwcr (Pnrrm) I?! '0 A $75 Mum-M Nous—'eTem ol OVA-’1 ‘70-. A37: Shun-rum Neoplasm o! Fen.» N.O.S. III-I” A $7.! “alien-M Stools-m of our" meenned Sues zoo ‘ A as team. and Noah-emu too. 20). 20! A.“ LW-Dhmuconu and other Neop'mn u! Lyvnphllic and “Maudie System I” A 600 Born; T53...“ of Bren: ZIQ L606 Clams Fubromyome II. L60! lem'n \qwlnm of Oven ".03.?10-230 L601 Other Heme-n and Curt. .ma Nmpl ume ALLERGIC \lETADO'JC AND BLOOD DISEASES 350,1” AGI Non-wear Genre 131 L62 Thromumu mm at m bout Go-m 3.. A 6] D 'abflel MeUIKTn ”0 AA” afl‘men‘ ”I L643T PIN-[n w A“. TsTcum 3“”. AMI! Kenna-notice 3.3-1“ A an TT0\her Dehum! Sui to 2.0 A 654 Perva'I'rTersTe:dT;T0;vTH»p-¢nhmmac An wvm‘u ”I A 650 ltonTEhTfl-Txn—cy AnaeT;:IIv;\Thromut) TTTm -e AM. 5...... ( 3.1.7.3; T 'T'iiIT RT {8;} Tch;TL..M..I..IT Mm... T 30! .T TA 66: Avhm - T T T_TT TTTT IN:: 'TTA-ITM TI’urouu en ITIwhzTHun-mvh-qu- (rm! mm H. O a no M T770: TELL—Thu ,1. T;:Ir:;lTnTOT.\-IIIT:I .I.. '....I film-J nun-o. TTTTT TTT TTTTT TTT Tum A‘fl IIT or ~19“-.an «vs II \: A~I§TFTN3I owns“ “355 ITQ'T MTG-T .3.‘..i;.:.' T T- T .TT TTT JIO. IIITIT‘T TIT—7;! TV": h.- .. -m. «Tor-331:3}: I «Now TIIITTT TTATnTT ' 3:2... I... m... 'T T TTJ-II In TTATIIITT TTu In I .......m... aim... “T7". —)TO;; A 7T|c:: \lumu I- c In: To II. TrTTTIeTmTuT _ TTT SCOT-l A suT T \I.§........ .I..;T..Tr::.uo..u. Joe A 7:: TOTLLKI... mm. (m... \Inun-vvntcel A :I mI TaAz) us A u \IITII. IIITATm—IZJITST:;‘~T.T ” Tom. has 1 334 THE IE'UBLIC OF UCANDA-HINISTRY OF HEALTH “J. W III... RETURN or DISEASES (IN-PATIENTS) HOSPITAL "" ’ er loop-- I _ . For the much of ' - - I9...... a. La D ADMISSIONS DEATHS Ne. Hele Female Tull Mde Female 1’.“ . Dunes ee Nam-I Sum u. Sal. WM "T 1!: An spa..." 31m A.“ MW Due-ea of Eye (aunt Tush—- L410 3” L75 Cow *— 3.7 A7 01mm ““3"-” An 0m. “an. no MM:- 3” AJIe (rm- Euem wan-m A?“ ow: Due-nee end cm...“ a «I. Eu “.05.,”4” A?“ Other Due“- end Contact! 01 the E” This. III-m And An aha Dunne- .nII. Nervou- 5...... 1' CIRCULATORY DISEASES rm A.” Rheum-LI: Fever {IO-4M A00 Chrome Rheumam Hem Dnen. 020-421 A.“ Area—althou: end Deccan-five Hun Dam ‘N‘ A.IZI Acute end Saba-me {knead [WM 7 OJI A.» EhJomrdeu I'Ibmen N.O.S. ‘30-43‘ All: 0M1 Um d Her! «0443 Ll) "\Wflcm Inna Hem Dive-u M1 A.“ “Wu-loan I-«Ihout Mame“ e! Hun ‘50-4“ “5 on!“ d Am 660-400 A.“ UNI" Due-ee- cl (.muhxon Swen .-I ”SPIRATORY DISEASES £70475 A." Anne L'm Ila-potion [alum IT 080-4” A.“ Inflm «0 A.” label PROM ’ 0‘! A.” From hm 092.69) A." I’M-h! Afnnal other and cum“ We 5” 4.92 Anne 3mm 50], 502 A.) Bron. hm- Chmmc end L'muhfied SIO L” Hm», o! Tannin AM MM 3". ”I A.” Lmrvmu Md Abteee of Lung ”9'0 A.“ Pleuney no..." Efluucn S" -l. 31,-! L99 Plum! with am when! «union of T3. 52) , L97. ”hum-e no.3 III-m Am ATI am Ila-m on... a ' ALIVEN‘TMY DISEASES ”0 Lfle Dem-I CAM SIN-SJ! A.,” A'Jolhev DIouueelTeuhend Gum 3‘0 A.” L'Ieev 0! W “I L:” TL'Icet of W TT'T 5“ AIM Gum ISO-SS) A.IO! A:;)endIam 560 L10). He mu of Am Cavity “than." Chan-m “I A.IOJO Herm- ol AMI-mun Camry snh Obsuueu‘oa ’70 '0 A. [01: In! uuuecepuon 3”" A.IOJJ \‘oIIUIuI N.O.S. 37° A.IOJe 01M? IFJG‘IINI om without "erni. SH '0 5.10“ (Intro L;I—:fIIIO And Calms (Au e man to 2 run) 37! '1 AIM (Intro I Menm and Col-(Is (Age 2 yen. and over) ’73 AIM C‘flmi: FmeI-me end L'keruue CoIIuI III A.IOS Carbon" 0! In" —_.310 III A.IOO CNI'TIIMT-Tejn—Il-T I'coIet—tPIu-I NIITIITSITO-T “I TmTT TTOTlher [)Iwe-eed lhgnuu Sun-m— TTT‘- T T T -T' GI .\'ITU I III -.AIII IIISIAacs 5‘” A.II‘I Anne Nerhmie MI- 3" A um ( Mu.“ uh] I‘ nape IfieJ .\cwmm ..— T'- T ATTIC; ...rT—;CI_I':. PVCIMF'PIVIC:;;bol LIITfi IN. “one of K6 In" (any! 1 I- A I) A ”GI-TOO; AIII I (—'eI:IIIII-I'I II-Iul\:ll:_-T T T “T”...- T. on A In II "III—INTJIT hum. iii-0‘“ [III TIMI-nu '3' "In“ (eeoepl Stool-pm.) T..- m “ ‘ 77.1" 12:537. .3 a...“ j. A OI) A II“ TIIydmuv e I I ON A! He ITIZ-wJen o! “en-(Mica on A IIU TxTITu-I... Ier'ule T“ .—.:.I; -I T—ATIMITT VapneI-Tnlul' T- 505 sou-en AI w “II-2.35;. 4......- oc (Mao-I'M», Snlem ’ Tam. has 1 .. 33S THE IE'UILIC OF UGANDHINISTRY OF HEALTH "J. N a... RETURN or DISEASES (IN-PATIENTS) HOSPITAL "" ‘ For the month of - - I’m... ADMISSIONS bums “ Ih m N. MeIe Fade Tote] I“ Tel-ll DISEASES 0' PREGNANCY A.\'D FUENTEIIUM MI A.III Sepue e! Pam Chum M [New “VIA-“1 '1 A.IIOe I‘Ic-«Iempen Tee-lie “I ') Al I“ Echmvuc 1 guanine EELOII,“ " LI I“ OIF:;;1 am“ el W end ”menu:- 070 LI l7e Human fine 0! Pregnancy end Cit-Hm (Into-one” 07L 671 L117) [hem-artiste o! Prom, end CIIJJDI'Hh (pet-mu” 050 Ad I. Abomm '1“ Seven ot Teeeeu‘e —— OS! All. Atom cub sex-II “I A.IZOO Envy“ Pam-«1 070 All“ Abnomul Lehuu: III-e to UM O74 A.Ine Abnormel Leheul due to Mela—elm .77 A.IZOJ Ruptured L'lenu mew-e” A.IZOe 00m Comphcel'm of PIM' CMdL-Inh end Wm .00 L120] Denver-y “thou: Camping-n. /-' SKIN AND MUSCL'LO-SKXLETAL msusu mm A!!! [Memo-I 0! Sim end 5'..th Tieeue —_13I>-7u A.II: AnImu. end Spur-dune m7), A.IJ) 5I WUNL'WMW 7’0 A.IIO Oueomve'nre end Fade-(Ill \ 731. 165-7" LII! Aliylrmn end Around theneIe Shel-“I Defmfi- (alcove AJOI 7I$ All“ Ukn cl Le; no.3. ’00-?“ AI!“ AI! other Dneeeee el Sb! 1” All“ Pies-linens. nl-7“ A.IIU All HM 0m d BMW 5!“!!- 151 L117 5m DIM: end Mom“ 7“ All. Cancun! \Ielfoemeuom d Gin-Idem" Syn-II m 730-?” All? OIIIee Con; cI-mel Meltmooe. DISEASES OF N£W«IOIN 760.76I A.IN MI lniunce 1e: A.IJI POI-Mal AID”)!!! It“ Aidan-i. 1“ A.IJZe Danton e! Sew-Dom (Ia-dc 0 mil) 7“ , A.mo 09mm"... Neon-conu- mag-flee A.IJZe Other Inlccum ‘ L 770 A.IJJ "laminae Dueeee el Sew-m 771,772 A.IM All other Defined Diane. e! let" Idem 710-770 A.IJIe lmmfify 77) Al)“ III-defined Dieeee. el Led, ldwv ILL-DEFINED DISEASES ”0 A.Ils Sendity Inhout m e! Fuzhou] ‘ m‘ A.IJ1e Prune of Cub-M Origin ”3 A.I.WO Mm WNW need ‘0! MM medial m N.O.8. 7%”! A I17: All other “1.9.6.4 Cm e! Mortmh'ty ' INIL'RIES M AN I)! frames of $1.14!! ees—em AMI)! I'm at Spine end Trunk "H” ANJeO Fncnue o.’ Limhe 830-039 ANJOI Dnlocenon without Freda" M044. AVJOJ Sprain- enJ Stream of loam end Mgeceee Mantle lib-ISO ANJO) Heed Iniurv («not Free-woe) Ito-MO Asuu lmemel My}; Um. Abdomen end rum rap—’01:. . Wfiéuum‘rmm: \\ u.uu. ___ ”53}; "- ‘ «To; " s...m‘.3'.‘..7.;, Ill-.536. end ( MN“. mm mm MM ”.53.?” ‘I§.F"‘*E..... .. .445.14;;...;.".:.“...:.::.‘:..;a.: "‘ “ 4 9;;0_Hm -mu Duffie end Scelde ____ WI. -7; HO CIIeue el Pee-tune __ gent-97'. .... A‘V I-m- “me; [jennm-eI I'"«. of Inland ('euIee .596” ...- "”"--‘N ' Tout. Pane 4 ' '“ TILT E.“ I . __ 2 Tarn; Hue I . . 1 TM“. Peal I GRAND TOTAL .__._ TO’I‘AL IN-I'ATII‘. \‘T DAYS not. men " Net Duke-tee Spec“ ". I.e. N 0.3 use-m neene ell othe dunno Include.) between three nee-hen In the lumen-eel Cleeeifimien wheel-“60M lie-e MID-O. " D I Josefina-nee evened III eny IIIIe eke-hue. 336 TH! REPUBLIC OF UGANDA—MINISTRY OF HEALTH I'L'. u up I ”I. . om RETURN OF DISEASES (OUT-PATIENTS) AT (UNIT) a III-p- For ch. month OI ' I, ou- No. mm as” , . . Mala Fund. Teal ~ .. INFECTIOUS AND PAMSITIC DISEASES III-N I “Ow-rusty Tubcnu'm OIHI, 3 Other TM“ m 1' SW-o M! O Gonorrhea: It‘duhnx Ophltuhnu 8mm. m 77) 036-039 5 01h" Vent-I Drum .03 6 8mm" Dim..." M 7 Amt»: Dynasty 0” O “whom“: 050 9 “Mal-mt- 057,10. [0 Maj-Imnt'u («madam Tuberculwo. m 2) I. ll Lean-y on 12 Ankh“: 071 I3 MM PM m 14 Yam "0 1! Ann. Relaunch"- fll IO Aunt Pohomychuu Lu: Eden! “t I? SIR-Upon (\‘Anola Moon “3 I. Snuflpou (\‘anoln .\I no!) “5 l9 Khaki N6 20 lulu-II: on 2| Cinch-u.uu j N. 22 "not: lat" | M 2) Mama. I 093 26 Them. I "0 35 Mal-no ET. I“ 20 Mama 0 T. n: 27 Mum» :‘r. k I“. M" 2! Four- not uI-tmuu Iwcafied III 29 Twain-mun! . I” '0 .. )0 Savanna-nu \‘uual m -l 3| Schumann. [nun-ml _ I26 32 Tapeworm (Tuniun) I27 13 Onchotnunu I39 )4 Anlflmrwnuc (Hmh'onn) I”'. 35 Ate-uni. (Round-am) 1’0'3 )6 Gumvm I30 37 0th" Hoknivnhuc Dncun III )8 Tuna I” J“ Sabin 8.0.3.0164)! ‘0 01M Infective and Pmu'm: Din-u NEW’ ORO\\THS “@203 CI Mahmm Secular!“ “MIUJWI‘ Inbound filo-2J9 42 Benita and Och" Stephens ALLERGIC METABOLIC AND BLOOD DISEASES “I .3 AIM. 260 ‘4 Duke". 1‘6 ’0 4! Ku «Mortar 2.0-2“ do Vital-hm Deficient: Sula I'M-29) 47 Am ”.71?“ am- m u on... MIut-c Mash-k; and mm 0me “M . _. I)|§i-Ts;~t'n.v—§riv«uw SVH'EM AN!) sassr (mo/um , m3}. 0'! .‘Irnul Ih‘mukn -m V. w m so MAJ-v --').iu no SI min} In"... M Nun»... 3”...“ 11.7731 .77 ~;:;‘;TM.;;-Icl;hu h-I; I v... N-uha. m .‘M "13.3” ":3" Tin‘JJK-‘m: ' " ‘ a - ‘ ' ..- 1:“...;1;-- — “:4" WITi;:u:—v‘-I‘Fu . “ __......_____ "'— ”#ETRFWHK'IRIAMA ...- 7W- 5:.— ' Fir-Mom... ; --_- “n"— 080- “I IA -« I an}: on when" Down“ RESPIRATORY DIfiF'J‘IIS I‘m-0O! !7 'W- ;‘3577. 57 A .. j..- OVIw-v Ila-c." 0' Renault- n \y-lcm Ton“ PMJ I 337 THE REPUBLIC OF UGANDA—MINISTRY OF HCALTH NJ. 1! ha 3 “I . u... RETURN or causes (OUT-PATIENTS) AT (UNIT) 0W.- ”0 53%!” 3.0 ”I ”3'3 5“ N.O.S 5.0-fl: 1‘03 3“ 47! SW-SN 720-717 715 5.0.5. 6"" #10 710-7" 7“ 7.3 TIC-776 7.53.0.7” 3.000-no ‘ N. loo-III N. tub-NV .\'. 8.0.8. USO-Wu) M795 Y. 00 Y. N! Y.” Y. o! VJ) Forth. monthd ' . I’m... ALINIENTAIY DISEASES Dam Cum 0!th than" cl Tooth and Gut-II "um. I Gunu Enhhln (.u. 0 inch and 0"!) Joundu: Cums-us ul Luv" 05': lhuun oI Lu!" and No th Other inn»!- ol Mom-r1 Syum CENITO-L’RINARV DIS‘ASE‘ Mphntfl Ilvdoonlo Crud-«fl "'IUU'. Othcr Uncut! 0! (.M~L'mry S\ «m Danna oI Fran-nut. “mic and human!» SKIS AND ML’SCI'LO-SKELITAL DISEASES Animu- oml Rheum-m Tnmnl L‘kn Infmnmo of *hn and Sat-tum Tit-val Olhu Dnum cl Nanak-Stolen! Drum DIfiEASES OF NEW-IOIN barman o! New-hm (Ag. and" 0 nuts) Ophlhlhnfimlmfll Inn-n many 00.! Muffin-om and Dunn of Inhnn INIL’IIu I‘nuum nu Unload“ 3mm. Duff. .04 Q0“. Mum-n1 All och" Input-s and Wound. ILL-DEFINED DISEASES Ill-defined (M and Conductor. IXAHINATIONS AND INOCI'LATION'S Ame-Vin! C‘hiIJ “'clfm 00in Eiumnam lmdmc Adamant Snuflno' \‘uunovm Proofiylann I'm!“ Tom. PAC: : Tom. PAL! I Tom. III-A1 hummus GRAND TOTAL .. ".05. M “No! Min $9.0M ". I... 3.0.3. GIG-Injun- all «he: dun-u incl-dd m thew W a the Imps-I Clan-Mum: I. it won! m IN. J not «hon-u- wifid in my line darken. ._._. .,_..._..__._= m: 338 MI. 77 PCs-1‘0! .. HIM OF HEALTH NAME CF unit. . . . . . 'DISTNIC REHCAT ‘- ’ Ti—E "(2.3-In” uF .............................. l9 ...... I. New Cases (m o “0 . . Hale ............... Female .............. Total ............ . ..... New Cases (five -..d under) . . Male ............... Female ............... Total .................. Reattendances . . . . Hale ............... Female ......... ‘ ..... Total .................. Admissions .. .. .. Hale ............... Female ............... Total Transferred to Hospital . . . . Hale ............... Female ............... Total .................. Deaths (six and over) . .. Male ............... Female............... Total .................. Deaths (five and under. excluding those given in section 6) . . Male ............... Female ............... Total .................. AID POSTS: Number ................................................ Number of Visits . ......................... , ............ Number of new patients .............................. Re-attendancea ............................................. 2. RETURN OF ALL NEW CASES. BOTH IN AND OUT PATIENTS (excluding Aid Posts): Clinical Malaria with Fever Malnutrition (five and under) ........................... Upper Respiratory lnlettlon ........................ Burns and Scalds .......................................... Pneumonia ................................................ Fractures ................................................... Tuberculosis 0! Lung ............................... Wounds ..................................................... Early Syphills ............................... . ............. Measles ........................ . ................. ..... Early Yaw: ................................................ Whooplng Cough .......................................... Late Syphilis or Yaw: ................................. Chicken Pox .. .............................................. Gonorrhoea ................... . ......................... Stables ...................................................... Leprosy ................. . ................................. Stricture of Urcter ....................................... Diarrhoea (over five) .................................... Hernla (inguinal or femoral) .............................. Diarrhoea (five and under) ................. . ......... Other Diseases ........... . ................................. Intestinal Worm Disease .................................................................. ......................... Conjunctivitls and Trachoma ........................................... . ........................ . ..................... ‘ Other Eye Diseases ...................................................................................................... Disthargin; Ear .................................................... . ................................................ Traplcal Ulcer ....................... . ......................... 3. SMALLPOX VACCINATION: Six and over ....................... five and undtr ........................... Total .............................. 4. OTHER PREVENTIVE INOCULATIONS (Whoopmg cough. T.A.B.. etc): 5. MINOR OPERATIONS: Dental ExtracUons ......... , ...... .. ....................... Other operations . 6. HATI‘RNITY: Antenatal new cases .............................. .. Re attendances ...................... . ...................... Deliveries ................................. . ............. Maternity tascs sent to Hospital . Maternal Deaths ........................ Dtath of Babies ......... . .............. Stillbirth: 7. CHILD WELFARE—CHILDREN'S CLlNIC: New Cases ........................................... Ike-attendances ............................................ [9.1.0. BIBLIOGRAPHY Bibliography The Economics of Health and Medical Care. Proceedings of the Conference on the Economics of Health and Medical Care, Ann Arbor, Michigan, May 10-12, 1962. Ann Arbor, Michigan: University of Michigan, 1964. International Financial News Survey: 24, 21 (May 31, 1972). Mathematical Model Building in Economics and Industry. Collected Papers of a Conference organized by CEIR, held in London, July 4-6, 1967. London: Charles Griffen and Co., Ltd., 1968. Report of the National Conference on Medical Costs, Washington, D.C., June 27-28, 1967. Washington, D.C.: U.S. Department of Health, Education and Welfare, 1969. Report of the Second Commonwealth Medical Conference, held in Kampala, Uganda, 1968, presented to Parliament by the Minister for Overseas Development. London: H.M.S.0., 1968. Abel-Smith, Brian, "Economics and Health Services." In Health Economics, a report on a seminar convened by the Regional Office for Europe of the World Health Organization, Moscow, June 25 to July 5, 1968. Copenhagen: Regional Office for Europe, World Health Organization, 1969. Abel-Smith, Brian, "Health Policies, Adjustments and Economic Development," in The Role of Social Security in Economic Development, Research Report no. 27. Washington, D.C.: Office of Research and Statistics, Social Security Administration, U.S. Department of Health, Education and Welfare, 1969. Abel-Smith, Brian, Paying for Health Services: A Study of the Costs and Sources of Finance in Six Countries. Public Health Papers no. 17. Geneva: World Health Organization, 1963. .Anderson, W.H. Locke, "Production Scheduling, Intermediate Goods and Labor Productivity," American Economic Review, LX, 1 (March 1970), 153-162. Apter, David, The Political Kingdom of Buganda: A Study of Bureaucratic Nationalism, Princeton, New Jersey: Princeton University Press,l96l. .Arriaga, Eduardo E., Mortality Decline and Its Demographic Effects in Latin America. Population Monograph Series. Berkeley, California: Institute of International Studies, University of California, 1970. 339 340 Bacharach, Michael, Biproportional Matrices and Input-Output Change, London: Cambridge University Press, 1970. Barker, Kenneth; Kimbrough, Wilson; and Heller, William, A Study of Medication Errors in A Hospital, Arkansas: University of Arkansas, University of Mississippi Press, 1968. Barlow, Robin, The Economic Effects of Malaria Eradication, Research Series no. 15. Ann Arbor, Michigan: Bureau of Public Health Economics, School of Public Health, University of Michigan, 1968. Barlow, Robin, Planning Public Health Expenditures with Special Reference to Morocco, Center for Research on Economic Development, Discussion Paper no. 27, Ann Arbor, Michigan: University of Michigan, 1972. Baryaruha, Azarias, "Factors Affecting Industrial Employment: A Study of the Ugandan Experience, 1954-1964." East African Institute of Social Research Occasional Paper no. 1. Kampala: East African Institute of Social Research Makerere University College, 1967. Bennett, F.J.; Hall, S.A.; Lufwama, J.S.; and Redo, E.R., "Medical Manpower in East Africa: Prospects and Problems," East African Medical Journal, 42, 4 (April 1965) 149-161. Bennett, F.J, and Saxton, G., "Utilization of the Social Sciences for Some Aspects of Health Planning in East Africa," a paper presented at the Con- ference on Africa in World Affairs: The Next Thirty Years, Makerere University, Kampala, Uganda, 1969. Unpublished paper, Kampala, Uganda, 1969. Benyouseff, Amor, "Health, Population Dynamics and Development;" paper presented at the African Regional Population Conference, Accra, Ghana, December 9—18, 1971. Unpublished paper, Accra, Ghana, 1971. Billinghurst, J.R., Trowell's Diagnosis and Treatment of Diseases in the Tropics London: Bailliere, Tindall and Cassell, 1968. Bohm, Peter, "An Approach to the Problem of Estimating the Demand for Public Goods", Swedish Journal of Economics, 73, 1 (March 1971), 55-66. Bonnet, Philip D., "Increased Production and Better Utilization", in Report of the National Conference on Medical Costs, Washington, D. C., June 27—28, 1967. Washington, D.C.: U.S. Department of Health, Education and Welfare, 1969. IBoulding, Kenneth E., "The Concept of Need for Health Services," Milbank Memorial Fund Quarterly, 44, 4, part 2 (October 1966) 202-224. 341 Bowles, Samuel, Planning Education Systems for Economic Growth. Harvard Economic Studies, no. 133. Cambridge, Mass.: Harvard University Press, 1969. Bowman, Edward H. and Fetter, Robert B., Analysis for Production and Operations Iganagement, Homewood, 111.: Richard Irvin, Inc., 1967. Bryant, John, Health and the Developipg World, Ithaca, New York: Cornell University Press, 1969. Caldwell, J.C., "Fertility Differentials as Evidence of Incipient Fertility Decline in a Developing Country: The Case of Ghana", Population Studies. 21, 1 (July 1967), 5-21. Carlsson, Robert; Robinson, James; and Ryan, Michael J., ”An Optimization Mode] for a 'Non-Profit Agency'", Western Economic Journal 9, 1 (March 1971) 78—87 0 Chaing, C.L., An Index of Healthl_Mathematica1 Models, series 2, no. 5. Washington, D.C.: National Center for Health Statistics, 1965. Chenery, Hollis B., "Process and Production Functions from Engineering Data", In Studies in the Structure of the American Economy, edited by W. Leontief. New York: Oxford University Press, 1953. Coale, Ansley J. and Demeny, Paul, Regional Model Life Tables and Stable Populations. Princeton, New Jersey: Princeton University Press, 1966. (Scale, Ansley and Hoover, Edgar, Population Growth and Economic Development in Low Income Countries. Princeton, New Jersey: Princeton University Press, 1958. (Zolbourne, Michael, Planning for Health, London: Oxford University Press, 1963. (Sorsa, Leslie Jr., and Oakley, Deborah, "Consequences of Population Growth for Health Services in Less Developed Countries: An Initial Appraisal", in Rapid Population Growth: Consequences and Policy Implications, National Academy of Sciences Study Committee, Roger Revelle, Chairman. Baltimore: John Hopkins Press, 1971. (Summings, E.C., "The Development of Health Services in Sierra Leone," in The Demographic Transition in Tropical Africap Proceedings of an Expert Group Meeting, Paris, l7-l9 November 1970. Paris: Development Center, Organization for Economic Cooperation and Development, 1971. IDano, 8., Industrial Production Models, New York: Springer-Verlag, Inc., 1966. 342 Day, Richard H., Recursive Programming and Production Response. Amsterdam: North-Holland Publishing Co., 1963. Donabedian, Avedis, "Evaluating the Quality of Medical Care,” in Program Evaluation in the Health Fields, edited by Herbert C. Shulberg, et. al. New York: Behavioral Publications, 1959. Dorfman, Robert; Samuelson, Paul; and Solow, Robert, Linear Progpamming in Economic Analysis. New York: McGraw Hill Book Co., Inc., 1958. Dow, Thomas E., Jr., "Fertility and Family Planning in Sierra Leone," Studies in Family Planning, II, 8 (August 1971) 153-166. Dunlop, David W., "Research on the Economics of Health Services in East Africa" in Rural Africana, no. 13 (Winter 1971), 77-84. East Africa, Annual Trade Report, 1945. Nairobi: East African Statistical Dept., 1946. Economic Commission for Africa, "Population Growth and Social and Economic Development in Africa: A Review and Discussion of Country—Case Studies." Paper presented at the African Regional Population Conference, Accra, Ghana, December 9-18, 1971. Unpublished paper, Accra, Ghana, 1971. Eicher, Carl; Zalla, Thomas; Kocher, James; and Winch, Fred, Employment Generation in African Agriculture. Institute of International Agricul- ture Research Report no. 9. East Lansing, Michigan: Institute of Inter— national Agriculture, Michigan State University, 1970. Enearnacion, J., "Some Implications of Lexicographic Utility in Development Planning," in Philippine Economic Journal, 9, 2 (1970) 231-240. Fein, Rashi, "Health Programs and Economic Development," in The Economics of Health and Medical Care, Proceedings of the Conference on the Economics of Health and Medical Care, Ann Arbor, Michigan, May 10—12, 1962. Ann Arbor, Michigan: University of Michigan, 1964. iFeldstein, Martin, "Cost-Benefit Analysis and Health Program Planning in Developing Countries," a paper presented at the Second Conference on the Economics of Health, Baltimore, Md., 1969. Baltimore: mimeograph, 1969. Feldstein, Martin 3., Economic Analysis for Health Service Efficiency: Econome— tric Studies of the British National Health Service. Amsterdam: North— Holland Publishing Co., 1967. 343 Feldstein, Martin 8., "Health Sector Planning in Developing Countries," Economicai new series 37, 146 (May 1970) 139-163. Feldstein, Paul J., An Empirical Investigation of the Marginal Cost of Hospital Services, Chicago: University of Chicago Press, 1961. Feldstein, Paul J., "Research on the Demand for Health Services," Milbank Memorial Fund Quarterly XLIV, 3, part 2 (July 1966) 128-165. Fisher, F.M., "Cost Functions and Budgets," in The Analysis of Public Output, edited by J. Margolis. New York: Columbia University Press, 1970. Fox, Karl A. and Thorbecke, Erik, "Specification of Structure and Data Requirements in Economic Policy Models," in Quantitative Planning of Economic Polioy, edited by Bert G. Hickman. Washington, D.C.: The Brookings Institute, 1965. Frank, Charles R., Jr., "Urban Employment and Economic Growth in Africa," Oxford Economic Papers 20, 2 (July 1968) 250-274. (Salea, J., "Inventory, Appraisal and Assessment of the Basic Health Services of Uganda: Developments for a Malaria Eradication Programme." Jinja, Uganda: Malaria Pre-Eradication Programme, World Health Organization, 1967. (Sales, J., An Inventory of Government Medical Units in Uganda, Entebbe: Ministry of Health, 1968. (Salenson, Walter and Pyatt, Graham, The Quality of Labour and Economic Develop: ment in Certain Countries. Studies and Reports New Series no. 68. Geneva: International Labour Organization, 1964. (Eass, Saul 1., Linear Programming Methods and Applications. New York: McGraw— Hill Book Co., Inc., 1964. (3ershenberg, Irving and Haskell, Mark, "The Distribution of Medical Services in Uganda." Social Science and Medicine, 6, 3 (June 1972), 353-372. (Sish, Oscar, "Health Planning in Developing Countries," The Journal of Development Studies; 6, 4 (July 1970), 67-76. (Soldberger, Arthur S., Econometric Theory. New York: John Wiley and Sons, Inc., 1964. (Soode, Richard, "Comment" on "Health Programs and Economic Development" by Rashi Fein in The Economics of Health and Medical Care, Proceedings of the Conference on the Economics of Health and Medical Care, Ann Arbor, Michigan, May 10-12, 1962. Ann Arbor, Michigan: University of Michigan 1964. 344 Hall, S.A., and Langlands, B.W., eds., Uganda Atlas of Disease Distribution. Occasional Paper no. 12. Kampala: Department of Preventive Medicine and Department of Geography, Makerere University College, 1968. Hall, Thomas L., Health Manpower in Peru: A Case Study in Planning. Baltimore: Johns Hopkins Press, 1969. Harris, John R., and Todaro, Michael P., "Wages, Industrial Employment and Labour Productivity: The Kenyan Experience." Eastern Africa Economic Review, I, 1 (June 1969), 29-46. Haskell, Mark, "Medical Services in Masaka District," unpublished paper, New York University, 1971. Heisel, Donald, "Attitudes and Practice of Contraception in Kenya," Demography, V, 2 (1968), 632-641. Henin, Roudshi A., "The Applicability of the Theory of Demographic Transition to African Countries," in The Demographic Transition in Tropical Africa, Proceedings of an Expert Group Meeting, Paris, 17-19 November 1970. Paris: Development Center, Organization for Economic Cooperation and Development, 1971. Henin, Roudshi A., "Population Growth and Economic Development: The Sudan- A Case Study." Paper presented at the African Regional Population Conference, Accra, Ghana, December 9-18, 1971. Unpublished paper, Accra, Ghana: 1971. Hickman, Bert 6., ed., Quantitative Planning of Economic Policy. A conference of the Social Science Research Council Committee on Economic Stability. Washington, D.C.: The Brookings Institute, 1965. Hillier, Frederick S. and Lieberman, Gerald J., Introduction to Operations Research. San Francisco: Holden-Day, Inc., 1967. Hixson, Jesse S., The Demand and Supply of Professional Hospital Nurses: Intra-Hogpital Resource Allocation. Unpublished Ph.D. dissertation, Michigan State University, 1969. Hixson, Jesse S., "A Model of Hospital Production, Cost, and Capacity Determination." Econometrics Workshop Paper 7009. East Lansing, Michigan: Michigan State University, 1971. Hallway, Jane, A Survgy of the Church Related Hospitals in the Anglican Province of Uganda, Rwanda and Burundi. Kampala: Provincial Medical Board, Church of Uganda, Rwanda and Burundi, 1972. 345 Huang, David 8., Introduction to the Use of Mathematics in Economic Analysis. New York: John Wiley and Sons, Inc., 1966. Idusogie, E.C., "Relation of Papulation and Nutritional Health Problems in African Communities," paper presented at the African Regional Population Conference, Accra, Ghana, December 9-18, 1971. Unpublished paper, Accra, Ghana, 1971. Ingbar, Mary Lee and Taylor, Lester D., Hospital Costs in Massachusetts: An Econometric Study. Cambridge, Mass.: Harvard University Press, 1968. International Bank for Reconstruction and Development. Economic Development of Uganda. Baltimore: Johns Hopkins Press, 1962. International Labour Organization, Employment Polioy in Africa, Part I: Problems and Policies. Report to the Third African Regional Conference, Accra, Ghana, December, 1969. Geneva: International Labor Organization, 1969. International Planned Parenthood Federation. Report to Donors: Programme Development and Financial Statements 1970-72. London: International Planned Parenthood Federation, 1971. Kaser, Michael C., "Observations on the Likely Economic Efficiency of the Malaria Eradication Program." In The Economic Effects of Malaria Eradication, Research Series no. 15, by Robin Barlow. Ann Arbor: Michigan: Bureau of Public Health Economics, School of Public Health, University of Michigan, 1968. King, Maurice, "The Development of Health Services in Malawi, 1970-1985." Blantyre, Malawi: mimeograph, 1969. King, Maurice, ed., Medical Care in Developing Countries: A Primer on the Medicine of Poverty. A symposium from Makerere University College, Kampala, Uganda. London: Oxford University Press, 1966. IKirk, Dudley, "A New Demographic Transition." In Rapid Population Growth: Consequences and Policy Implications, National Academy of Sciences Study Committee, Roger Revelle, Chairman, Baltimore: Johns Hopkins Press, 1971. iKuczyaski, R.R., Demoggaphic Survey of the British Colonial Empire, vol. 2; London: Oxford University Press, 1949. ZLapham, Robert J., and Mauldin, W. Parker, "National Family Planning Programs: Review and Evaluation." Studies in Family Planning, 3, 3 (March 1972), 29-52. 346 iLave, J., and Lave, L., "Hospital Cost Functions.‘ American Economic Review, 60, 3 (June 1970), 379-396. ILeontief, W., ed., Studies in the Structure of the American Economy. New York: Oxford University Press, 1953. iLipton, Michael, rapporteur. Population Growth: The Impact of Advances in Agriculture and Medicine. Report of a Conference at Ditchley Park, January, 1969. Enstone, England: The Ditchley Foundation, 1969. Iiaclean, Una, Magical Medicine: A Nigerian Case Study. London: Penguin Press, 1971. biahoney, J.A., "The Demographic Effects of Improvements in the Health Services in Gambia," in The Demographic Transition in Trapical Africa, Proceedings of an Expert Group Meeting, Paris, 17-19 November 1970. Paris: Development Center, Organization for Economic Cooperation and Development, 1971. lialinvaud, E., "A Planning Approach to the Public Goods Problem." Swedish Journal of Economics 73, 1 OMarch 1971), 96—112. ‘Manuwa, Samuel, "The Principles and Methodology of Planning the Deve10pment of National Health Programs in Underdeveloped Countries." West African Medical Journal (April 1961), 69-86. iMargolis, Julius, ed., The Analysis of Public Output. New York: Columbia University Press, 1970. iMassell, Benton and Heyer, Judith, "Household Expenditure in Nairobi: A Statistical Analysis of Consumer Behavior." Economic Development and Cultural Change 17, 2 (January 1969), 212-234. iMeyer, Robert A., Jr., "Private Costs of Using Public Goods," in Southern Economic Journal, 37, 4 (April 1971), 479-488. ‘befat, Malcolm, Mobile Young Child Clinics in Rural Uganda: A Report on the Ankole Preschool Protection Programme, 1967-69. Kampala, Dept. of Pediatrics and Child Health, Makerere University College, 1970. IMorrow, R.R., "Tuberculosis," in.gganda Atlas of Disease Distribution, Occasional Paper no. 12, edited by S.A. Hall and B.W. Langlands. Kampala: Dept. of Preventive Medicine and Dept. of Geography, Makerere University College, 1968. iMorrow, R.R., "Tuberculosis Control in East Africa." Unpublished paper, Mulago Medical School, Kampala, Uganda, 1969. 347 Ifiushkin, Selma J., "Health as an Investment." Journal of Political Economy 70, 5 supplement (October 1962), 129-157. 'Myrdal, Gunnar, Asian Drama: An Inquiry into the Poverty of Nations. New York: Twentieth Century Fund, 1968. National Academy of Sciences Study Committee, Roger Revelle, Chairman. Rapid Population Growth: Consequences and Policy Implications, vols. 1 and 2 Baltimore: Johns Hopkins Press, 1971. N'Doye, Thianar, "Infant Mortality and Nutritional Problems," in The Demographic Transition in Tropical Africa, Proceedings of an Expert Group Meeting, Paris: 17-19 November 1970. Paris: Development Center, Organization for Economic Cooperation and Development, 1971. Newhouse, Joseph P., "Toward a Theory of Non-Profit Institutions: An Economic Model of a Hospital." American Economic Review, 60, 1(March 1970), 64-75. Newman, Peter, Malaria Eradication and Population Growth, with Special Reference to Ceylon and British Guiana. Research series no. 10. Ann Arbor, Mich.: Bureau of Public Health Economics, School of Public Health, University of Michigan, 1965. 'Norman, John C., ed., Medicine in the Ghetto. New York: Appleton-Century- Crofts, 1969. ()kediji, Francis 0., "Socio—economic and Demographic Aspects of Nigeria's Second National Development Plan, 1970-74." Paper presented at the African Regional Population Conference, Accra, Ghana, December 9-18, 1971. Unpublished paper, Accra, Ghana, 1971. Chnran, Abdel R., "The Epidemiologic Transition: A Theory of the Epidemiology of Population Change."__Milbank Memorial Fund Quarterly, 49, 4, Part 1 (October 1972), 509-538. ()rganization for Economic C00peration and Development, The Demographic Transi- tion in Tropical AfricaJ Proceedings of an Expert Group Meeting, Paris, 17-19 November 1970. Paris: Deve10pment Center, Organization for Economic Cooperation and Deve10pment 1971. lian American Health Organization, Health Planning: Problems of Concept and Method. Scientific Publication no. 111. Washington, D.C.: Pan American Health Organization, Regional Office of the World Health Organization, 1965. 348 Papandreau, Andreas G., Introduction to Macroeconomic Models. Training Seminar Series 5. Athens: Center of Planning and Economic Research, 1965. Perlman, Mark, "Some Economic Aspects of Public Health Programs in Under— deveIOped Areas." In Economics of Health and Medical Care, Proceedings of the Conference on the Economics of Health and Medical Care, Ann Arbor, Michigan, May 10—12, 1962. Ann Arbor, Michigan: University of Michigan, 1964. Pfouts, Ralph W., "The Theory of Cost and Production in the Multiproduct Firm." Econometrica, 29, 4 (October 1961), 650-658. Pool, D.1., "Ghana: A Survey on Fertility Attitudes Towards Family Limita- tion." Studies in Family Planning. no. 25 (December 1967), 10-15. Potter, Robert G., "The Multiple Decrement Life Table as an Approach to Measurement of Use Effectiveness and Demographic Effectiveness of Contraception." A paper presented at the Conference of the International Union for Scientific Study of Population, Sydney, Australia, 1967. Unpublished paper, Sydney, Australia, 1967. Prais, S.J. and Houthakker, H.S., The Analysis of Family Budgets. Cambridge: Cambridge University Press, 1955. President's Science Advisory Committee, The World Food Problem, The White House, Washington, D.C.: U.S. Government Printing Office, May 1967. Republic of Uganda, Draft Estimates of Recurrent Expenditures, various years. Entebbe: Government Printer, various years. Republic of Uganda, Estimates of Recurrent Expenditure, various years. Entebbe: Government Printer, various years. Republic of Uganda, The Public Accounts of the Republic of Uganda for the Year Ended 30th June 1969. Entebbe: Government Printer, 1970. Republic of Uganda, Uganda's Plan III: Third Five Year Development Plan 1971/72-1975/76. Entebbe: Government Printer, 1971. Republic of Uganda,_gganda National Formulary, 1966, Entebbe: Government Printer, 1967. Republic of Uganda, Department of Lands and Surveys, Atlas of Uganda, 2nd ed., Entebbe: Government Printer, 1969. 349 Republic of Uganda, Government Statistician, Personal Communication, 27 January 1970. Republic of Uganda, Ministry of Health, Annual Report, for years 1958-1970. Entebbe: Government Printer, years 1958-1970. Republic of Uganda, Ministry of Planning and Economic Development. Back- ground to the Budget. Entebbe: Government Printer, various years. Republic of Uganda, Ministry of Health, Medical Services Statistical Records, lst July 1968 to 30th June 1969. Entebbe: Government Printer, 1969. Republic of Uganda, Ministry of Health, Statistical Report, various years. Entebbe: Government Printer, various years. Republic of Uganda, Ministry of Planning and Economic Development. Census of Distribution, 1966. Entebbe: Government Printer, 1967. Republic of Uganda, Ministry of Planning and Economic Development. Enumera- tion of Employees for various years. Entebbe: Government Printer, various years. Republic of Uganda, Ministry of Planning and Economic Development, Report on the 1969 Population Census, vol. 1, The Population of Administrative Areas. Entebbe: Government Printer, 1971. Republic of Uganda, Ministry of Planning and Community Development. The Real Growth of the Economy of Uganda, 1954-1962. Entebbe: Government Printer, 1964. Republic of Uganda, Ministry of Planning and Economic Development. Second Five Year Planpy1966/67-1970/7l1ySupplemental Projects. Entebbe: Government Printer, 1969. Republic of Uganda, Ministry of Planning and Economic Development. Second Five Year Development Plan, 1966-1971: Work for Progress. Entebbe: Government Printer, 1966. Republic of Uganda, Ministry of Planning and Economic Development. Statis- tical Abstract, various years. Entebbe: Government Printer, various years. Republic of Uganda, Ministry of Planning and Community Development. Survey of Industrial Production: Buildingyand Construction, 1964. Entebbe: Government Printer, 1964. . 350 Republic of Zambia. Development Division Office of the Vice-President. Zambian Manpower, Lusaka: Government Printer, 1969. Rice, Dorothy P., Estimating the Cost of Illness, Health Economics Series no. 6. Washington, D.C.: U.S. Department of Health, Education and Welfare, 1966. Rothenberg, Jerome, "Agenda for Research in the Economics of Health." In The Economics of Health and Medical Care; Proceedings of the Conference on the Economics of Health and Medical Care, Ann Arbor, Michigan, May 10-12, 1962. Ann Arbor, Michigan: University of Michigan, 1964. Ruderman, Peter A., "Lessons from Latin American Experience," in Economic Benefits from Public Health Services: Objectives, Methods and Examples of Measurement. Washington, D.C.: Public Health Service, U.S. Dept. of Health, Education and Welfare, 1964. Ruprecht, Theodore and Wahren, Carl, Population Programs and Economic and Social Developnent. Paris: Development Center, Organization for Economic Cooperation and Development, 1970. Sadik, Nafis, et. al.,eds., Population Control: Implications, Trends and Prospects. Islamabad: Pakistan Family Planning Council, 1969. Saxton, George A., "General Fertility and Infant Mortality," in Uganda Atlas of Disease Distribution, edited by S.A. Hall and B.W. Langlands. Kampala, Uganda: Department of Preventive Medicine and Department of Geography, Makerere University College, 1968. Saxton, George and Pike, M.C., "Follow-up Study of 921 Women Using IUCD's in Kampala." Unpublished paper, Kampala, Uganda, 1967. Schulberg, Herbert C.; Sheldon, Alan; and Baker, Frank, eds., Program Evalua— tion in the Health Fields. New York: Behavioral Publications, 1959. Shapiro, 8., "End Result Measurements of Quality of Medical Care." Milbank Memorial Fund Quarterly XLV, 2 (April 1967). part 1, 7-30. Sheps, Mindel D., "Approaches to the Quality of Hospital Care." In Program Evaluation in the Health Fields, edited by Herbert Schulberg, et. a1., New York: Behavioral Publications, 1959. Stone, Richard, Measurement of Consumers Expenditure and Behaviour in the United Kingdom, 1920-1938. Cambridge: Cambridge University Press, 1954. Strassman, W. Paul, "The Construction Sector in Economic Development." Scottish Journal of Political Economy, 17, 3 (November 1970), 391-409. 351 Suits, Daniel B., The Theory and Application of Econometric Models. Training Seminar Series 3. Athens: Center of Planning and Economic Research, 1963. Sullivan, Daniel F., Conceptual Problems in Developing an Index of Health. Series 2, no. 17. Washington, D.C.: Public Health Service, U.S. Department of Health, Education and Welfare, 1966. Tabborah, Ried B., "Toward a Theory of Demographic Development." Economic Development and Cultural Change, 19, 2 (January 1971), 257-276. Taber, Steven R., "A First Look at the Provisional Results of the 1969 Uganda Census," a paper delivered at the Seminar on Population Growth and Economic Development, Nairobi, Kenya, December 15—22, 1969. Nairobi, Kenya, Unpublished paper, 1969. Taylor, Carl E., "Functional Analysis of Health Needs and Services." A report compiled by Johns Hopkins University School of Hygiene and Public Health. Baltimore: Department of International Health, School of Hygiene and Public Health, Johns Hopkins University, 1970. Taylor, Carl E., "Health and Population." Forelgn Affairs, 43, 3 (April 1965). 475-486. Thompson, John D.; Marquis, Don B.; Woodward, Robert L.; and Yeomans, Richard C., "End Result Measurements of the Quality of Obstetrical Care in Two U.S. Air Force Hospitals." Medical Care, 6, 2 (March-April 1968), 123—143. Thurow, Lester D., "The Income Distribution as a Pure Public Good." Quarterly Journal of Economics, 85, 2 (May 1971), 327-336. Titmuss, Richard; Abel-Smith, Brian; MacDonald; Williams, Arthur; and Wood, Christopher, The Health Services of Tanganyika: A Report to the Government. London: Putman Medical Publishing Co., Ltd., 1964. Todaro, Michael P., "An Analysis of Industrialization, Employment and Unem— ployment in Less Developed Countries," Yale Economic Essays, 8, 2 (1968). Todaro, Michael P., "Education and Rural-Urban Migration: Theoretical Constructs and Empirhui Evidence from Kenya," a paper presented at the Conference on Urban Unemployment in Africa, University of Sussex, September 1971. Unpublished paper, Sussex, England, 1971. Uganda Catholic Medical Bureau, Survey of All Our Catholic Medical Units, Kampala: Uganda Catholic Medical Bureau, 1970. 352 Uganda Protectorate, Report of the Committee to Examine Medical and Health Services in Uganda (Frazer Committee Report). Entebbe: Government Printer, 1956. Uganda Protectorate, Report on the Enumeration of African Employees in Uganda, March 1951. Entebbe: East African Statistical Department, 1952. Uganda Protectorate, Uganda General African Census, 1959, Age Sex Analyses, vol. 1, Population by Sex and Age Gronp_for Protectorate Provinces, Districts, Countries, Divisions and Parishes. Entebbe: Government Printer, 1960. Uganda Protectorate, Ministry of Economic Affairs, Uganda Census, 1959, African Population. Entebbe: Government Printer, 1962. Uganda, Protectorate, Medical Department. Annual Medical and Sanitarleeport, 1932. Entebbe: Government Printer, 1933. Uganda Protectorate, Ministry of Health. Annual Report, for various years. Entebbe: Government Printer, various years. Uganda Protectorate, Ministry of Health, Statistical Reportl for various years. Entebbe: Government Printer, various years. United Nations, Demographic Yearbook, selected years. New York: United Nations, selected years. United Nations. Department of Economic and Social Affairs. MeasureslgPoliciesl and Programmes Affecting_Fertilityl_with Particular Reference to National Family PlanninglProgrammes. Population Studies no. 51. New York; United Nations, 1972. United Nations, Department of Economic and Social Affairs. Methods of Estimating Basic Demographic Measures from Incomplete Data. Population Studies no. 42. New York: United Nations, 1967. United Nations International Childrens Educational Fund, Progress Report 1969 Kampala: UNICEF Regional Office, 1970. United Nations Research Institute for Social Development, The Level of Living Index. Report no. 4. Geneva: United Nations Research Institute for .Social Development, 1966. 'United Nations Research Institute for Social Development. Social and Economic Factors in Development: Introductory Considerations on their Meaningl Measurement and Interdependence. Report no. 3. Geneva: United Nations Research Institute for Social Development, 1966. 353 U.S. Agency for International Development, Population Program Assistance: Aid to Developing Countries by the United States, Other Nations, and International and Private Agencies. Washington, D.C.: Agency for International Development, 1969. U.S. Department of Health, Education and Welfare, Economic benefits from Public Health Services: Objectives, Methods and Examples of Measurement. Washington, D.C.: Public Health Service, U.S. Department of Health, Education and Welfare, 1964. U.S. Department of Health, Education and Welfare, International Classification of Diseases, Adapted, vol. 2, Alphabetic Index. Washington, D.C.: U.S. Department of Health, Education and Welfare, 1962. __g VanderHooven, J., "An Inventory of Government Medical Units in Uganda, 1969." Jinja: Basic Health Services Project, World Health Organization, 1970. Veitch, M.D., "The Implications of Population Growth on the Health Services: An Economist's View, Zambia." Paper presented at the Seminar on Population Growth and Economic Development, Nairobi, Kenya, December 14-22, 1969. Unpublished paper, Nairobi, Kenya, 1969. Watts, Ronald E., "The Impact of Demographic Trends on Rural Development and Welfare," Rural Africana, no. 14 (Spring 1971), 11-18. Weil, Roman L., Jr., "Allocating Joint Costs," American Economic Review, 58, 5 part 1 (December 1968), 1342-1345. Weisbrod, Burton A., Economics of Public Health: Measuring the Economic Impact of Diseases. Philadelphia: University of Pennsylvania Press, 1968. Wheeler, Mark, "Medical Manpower in Kenya: A Projection and Some of Its Implications." East African Medical Journal, 46, 2, (February 1969), 93-101. Winslow, C.E.A., The Cost of Sickness and the Price of Health, Monograph series no. 7. Geneva: World Health Organization, 1951. World Health Organization, International Classification of Diseases; Geneva: World Health Organization, 1955, and revisions, 1965. World Health Organization, Measurements of Levels of Health, technical report series no. 137. Geneva: World Health Organization, 1957. World Health Organization, National Health Planning in Developing Countries, Report of a W.H.O. Expert Committee. Technical Report series no. 350. Geneva: World Health Organization, 1967. 354 World Health Organization, Official Records, No. 2. Geneva: World Health Organization, June 1948. World Health Organization, Planninngural Health Services. Technical Report Series no. 215. Geneva: World Health Organization, 1961. World Health Organization, Statistics of Health Services and Their Activities. Thirteenth Report of the W.H.O. Expert Committee on Health Statistics. Technical Report Series no. 429. Geneva: World Health Organization, 1969. Wray, Joe D., "Population Pressures on Families: Family Size and Child Spacing." In Rapid Population Growth: Conseguences and Policy Implications, National Academy of Sciences Study Committee, Roger Revelle, Chairman. Baltimore: Johns Hopkins Press, 1971. Wylie, Charles M., "The Definition and Measurement of Health and Disease.” Public Health Reports, 85, 2 (February 1970), 100-104. Zeuthen, F., Economic Theory and Method, Cambridge, Mass.: Harvard University Press, 1955. Zschock, Dieter, Magpower Perspective of Columbia, Princeton, New Jersey: Industrial Relations Section, Princeton University, 1967. Zubkoff, Michael and Dunlop, David W., "Consumer Behavior in Preventive Health Services;" paper prepared for Roundtable on Consumer Incentives in Health. Public Services Laboratory, Georgetown University, Washington, D.C., June 8-9, 1973, unpublished paper, Washington, D.C., 1973.