AN («mums OF 5mm: FACTORS ON THE 9mm seawca 95mm: AT EDWARD w. $FARRQW HOSPITAL. m LANSING, MICHEGAN The“: he HmDogroe of M. A. . ‘ MICHIGAN STATE UNIVERSITY ’ Frederic Boiten Plasman ‘ 1984 THESIS LIBRARY Michigan State University FREDERIC BOITEN PLASMAN ABSTRACT The purpose of this thesis was to analyze the affect of specific factors, the school calendar, weekends, holidays, temper- ature, and precipitation, on the patient service demand at Edward W. Sparrow Hospital in Lansing, Michigan. It was hypothesized that an understanding of these factors may better enable hospital administr— ation to achieve the most effective, economical and efficient utiliz— ation of the institution. A study of the patient census statistics for the year 1960 was undertaken. The patient service demand was studied for the following patient classifications: Total Patients Men Surgical Men Medical Women Surgical Women Medical Women Obstetrical Children Surgical Children Medical Children Nursery Births AAAAAAAAAA O \O 00 \7 OK)“ 1-\ ho M ——* D 0 C O 6 O O O U 0 vvvvvvvvvv ._L The census information for each separate interest area was obtained from the Edward W. Sparrow Hospital Official Census Book. The affect of the Lansing, Lansing Parochial, and East Lansing school calendars, weekends, and holidays on the census in each of the above patient classifications was determined using an analysis of variance statistical procedure. Men Surgical patient census was found to be higher during periods when school was in session. The other patient census classifications were not affected by the school calendar. Total Patients, Men Surgicals, Women Surgicals, and Children Surgicals were higher on weekdays than weekends. There were more births on Saturday and Sunday than on weekdays. The other patient census classifications showed no change over the seven day period. A census drop on holidays was found in the Total Patient, Men Surgical, Women Surgical, and Children Surgical classifications. The other patient census classifications showed no change over holidays. A linear regression statistical procedure was utilized in determining the affect of temperature variable. All other patient census classifications were independent of the temperature variable. None of the patient census classifications exhibited any effect by the amount of precipitation. AN ANALYSIS OF SPECIFIC FACTORS ON THE PATIENT SERVICE DEMAND AT EDWARD W. SPARROW HOSPITAL IN LANSING, MICHIGAN By Frederic Boiten Plasman A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Hotel, Restaurant, and Institutional Management 1964 TABLE OF CONTENTS ACKNOWLEDGMENT. . . . . . . . . . . . . . . . . . o . PREFACE . LIST OF TABLES. . . . . . . . . . . . . . . . Chapter I. INTRODUCTION . II. REVIEW OF THE LITERATURE . . . . . . o . Average Number of Persons Admitted to Hospitals Estimation of Hospital Bed Requirements Hospital Utilization in Michigan Calculation of Occupancy Factors Distinctive Patient Facilities Poisson Distribution in Patient Admissions Time—Series Statistics Effect of Empty Beds III. FACTORS INFLUENCING THE PATIENT'S ADMISSION TO THE HOSPITAL . In—school Versus Out-school Factor Seasonal Fluctuations Temperature Changes Weekends and Holidays IV. SELECTION OF DATA. V. STATISTICAL PROCEDURE. VI. FINDINGS . In—school Versus Out—school Patient Census Comparison Weekend Versus Weekday Patient Census Comparison Holiday Patient Census Analysis of Temperature and Precipitation Affects Page iv vi 11 18 20 22 VII. DISCUSSION. VIII. APPLICATION . IX. CONCLUSION. APPENDIX A. APPENDIX B. APPENDIX C. BIBLIOGRAPHY. Page 37 41 44 47 53 56 69 ACKNOWLEDGEMENTS The author wishes to express his sincere appreciation to the many people whose assistance and encouragement made this thesis possible. Especially, the author would like to mention Mr. Frank D. Borsenik, of the School of Hotel, Restaurant, and Institutional Management, Michigan State University, whose wise counseling and endless patience was of unestimable help. Appreciation is expressed to the other members of the Graduate Committee, Doctor Lendal H. Kotschevar and Mr. Douglas C. Keister. Those from Edward W. Sparrow Hospital who were of assistance were Mr. Donald H. Pound, past Director and the author's Residency Preceptor, Mr. Forrest K. Neumann, Director, as well as the many physicians, nurses and staff. A special note of gratitude is due to the Joseph Schlitz Brewing Company for providing financial support through scholarships granted by the School of Hotel, Restaurant and Institutional Manage— ment, and to Mr. Thomas Hain for his aid with the statistical work. And thank you my dear Susan for the wonderful times we had graphing census trends. East Lansing, Michigan January, 1964 iv PREFACE The purpose of this thesis was to analyze the affect of specific factors on the patient service demand at Edward W. Sparrow Hospital in Lansing, Michigan. It was hypothesized that an under— standing of these factors may better enable hospital administration to achieve the most effective, economical and efficient utilization of the institution. A study of the patient census statistics for the year 1960 was undertaken. LIST OF TABLES Table Page 1. Total Patients; In—school Versus Out-school Patient Census Comparison . . . . . . . . . . . . . . . . . . 22 2. Men Surgical; In—school Versus Out-school Patient Census Comparison . . . . . . . . . . . . . . . . . . 23 3. Men Medical; In—school Versus Out—school Patient Census Comparison . . . . . . . . . . . . . . . . . . 23 4. Women Surgical; In-school Versus Out—school Patient Census Comparison . . . . . . . . . . . . . . . . . . 23 5. Women Medical;.In-school Versus Out-school Patient Census Comparison . . . . . . . . . . . . . . . . . . 24 6. Women Obstetrical; In—school Versus Out—school Patient Census Comparison . . . . . . . . . . . . . . . . . . 24 7. Children Surgical; In-school Versus Out-school Patient Census Comparison . . . . . . . . . . . . . . . . . . 24 8. Children Medical; In—school Versus Out—school Patient Census Comparison . . . . . . . . . . . . . . . . . . 25 9. Children Nursery; In—school Versus Out—school Patient Census Comparison . . . . . . . . . . . . . . . . . . 25 10. Births; In—school Versus Out-school Patient Census Comparison . . .‘. . . . . . . . . . . . . . . 25 11. Summary Table; In-school Versus Out-school Patient Census Comparison . . . . . . . . . . . . . . . . . . 26 12. Total Patients; Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . . 27 13. Men Surgical; Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . . 27 14. Men Medical, Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . 27 15. Women Surgical; Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . . 28 vi Table Page 16. Women Medical; Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . . 28 17. Women Obstetrical; Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . . 28 18. Children Surgical; Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . 29 19. Children Medical; Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . . 29 20. Children Nursery; Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . . 29 21. Births; Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . . 3O 22. Summary Table; Weekend Versus Weekday Patient Census Comparison . . . . . . . . . . . . . . . . . . 3O 23. Total Patients; Holiday Patient Census. . . . . . . . . 31 24. Men Surgical; Holiday Patient Census. . . . . . . . . . 31 25. Men Medical; Holiday Patient Census . . . . . . . . . . 31 26. Women Surgical; Holiday Patient Census. . . . . . . . . 32 27. Women Medical; Holiday Patient Census . . . . . . . . . 32 28. Women Obstetrical; Holiday Patient Census . . . . . . . 32 29. Children Surgical; Holiday Patient Census . . . . . . . 33 30. Children Medical; Holiday Patient Census. . . . . . . . 33 31. Children Nursery; Holiday Patient Census. . . . . . . . 33 32. Births;Holiday Patient Census . . . . . . . . . . . . . 34 33. Summary Table; Holiday Patient Census . . . . . . . . . 34 34. The Analysis of Variance in Regression for Temperature . . . . . . . . . . . . . . . . . . . 36 35. The Analysis of Variance in Regression for Precipitation . . . . . . . . . . . . . . . . . . 36 36. Summary of Analyses . . . . . . . . . . . . . . . . . . 46 vii I. INTRODUCTION Illness of varying forms and degrees requires patients to be hOSpitalized. However, it has been suggested that the need for hospitalization is apparently more acute on certain days of the week than on others; it varies with the month of the year; by specific weather conditions; by the school calendar; and according to the community's employment situation. Since no proof exists which shows that neither illness nor accidents are reduced, for example, on week- ends, it is obvious that factors, as those mentioned above, other than the immediate physiological need, influence the patient's admission to the hospital. Therefore, to enable a realization and an understanding of the patient service demand, an analysis of patient census data and the comparison of hypothesized hospitalization-influencing factors was attempted. II. REVIEW OF THE LITERAURE It is desirable to present some preliminary considerations which will help to establish the significance of hospitalization and the magnitude of the patient service demand appreciated by hospitals. Approximately ten or eleven persons out of every 100 in any average population can be expected to be admitted for general hospital care 1 each year. The exact number of admissions and length of stay during any particular time period depends upon numerous and often nebulous 1Eli Ginzberg, A Pattern for Hospital Care; Final Report of the New York State Hospital Study (New York: Columbia University Press, 1949) p. 41 inter—related factors, such as, the size of the community, age- groupings, occupations of its inhabitants as well as a myriad of other factors.2 The estimation of hospital bed requirements has been developed into several basic statistical formulae which may serve as a guide. Stewart3 classifies these into three main groups: "1. Estimates based solely on population. It has been recommended that 4.5 or 5 general hospital beds should be provided per thousand population. 2. Estimates based on birth and death statistics. This is based on the assumption that there is a constant relationship between the number of patients requiring treatment in a hospital. The average figure commonly used is 250 days of patient care for every death that occurs among hospital patients. 3. Hospital insurance morbidity statistics for the given area." Therefore, it can be expected that a certain portion of a given population will be hospitalized, requiring a particular number of beds, and accompaning ancillary and paramedical services and facilities. The first question that may be asked is, what amount of patient service demand is to be expected at any given hospital? Several authors have offered hypotheses for foreCasting or 2Thomas R. Ponton, "Factors to be Considered in Survey of Community Hospital Needs," Hospital Management, Vol. 62. No. 3. (September 1946) pp. 40—41 3C. G. Stewart, "The Estimation of Hospital Bed Requirements," Canadian Journal pf Public Health, Vol. 42 (July 1951) p. 283 3 / - predicting patient service demand. BaileyT has developed a formula for calculating the total demand for medical care at a given hospital in a given geographical area, accounting for the effect of another hospital in the area. Briefly his formula is: "Total Demand for Medical Care for 'A' Hospital = —§—E3 a+b Where, a = number of patients admitted to hospital 'A' b = number of patients admitted to hospital '8’ N = population of the geographical area in which hospital 'A' is located." The Michigan Department of Health,5 outlines a method for estimating need in a community. The average utilization of hospital acute beds is between 950 and 1,000 patient days per 1,000 population. The plan recommends 1,300 patient days be made available per 1,000 population to allow for population increases, hospital occupancy fluctuations, and varying degrees of utilization between different areas. “Of the 1,300 patient days of care, it is expected that 1,000 days will be rendered in community hospitals, 200 days in regional center hospitals, and 100 days in base area hospitals." The average number of beds required is determined by using census population figures, the 1,300 per 1,000 ratio, and dividing by 365 days. The plan utilized an “occupancy factor" to adjust for the fact that hospitals cannot operate at 100% occupancy. This factor is obtained by taking 2% times the square root of the average number of occupied beds for each hospital in the area. The resulting total number of beds needed 4Norman T. J. Bailey, “Statistics in Hospital Planning and Design," of Applied Statistics, ed. Donald G. Beech and Leonard H. C. Tippett. (10 Vol.; London: Oliver and Boyd Ltd, 1956) Vol. 5, p. 146-157 5"Michigan State Plan for Hospital and Medical Facilities Construction," (Michigan Department of Health. 1961) pp. 7—8 4 in a given area proposed by the plan is the sum of the average number of occupied beds needed plus the beds provided by the occupancy factor. Fifteen years previous to the Michigan Department of Health study, a similar one was sponsored by the W. K. Kellogg Foundation6 "Both statistical theory and study of individual Hospital data indicate that the extreme limits of occupied beds will not be greater or less than the average census plus or minus approximately four times the square root of the average daily census. That is to say, it is unlikely that the need for beds in the course of a year will exceed the average census by four times the square root of that average. Corre- spondingly, it is improbable that the minimum number of beds used will fall below the average census less four times the square root of that average." The occupancy factor in the Kellogg study was four times, rather than 2% times, the square root of the average daily census, as suggested by the Michigan Department of Health. The difference is considerable and can be accounted for by the advent of an "antibiotic 1 age". ( ) Just fifteen years ago communicable diseases could and 6Hospital Resources and Needs Commission 9n Hospital Care; Report 9: the Michigan Hospital Survey, (Battle Creek: W. K. Kellogg Foundation, 1946) p. 105 (1)As an example let us assume that community X has a popul— ation of 100,000 and has 400 acute patient care beds. 1,300 patient days per 1,000 population per year produces 130,000 patient days of care. The average number of beds required will be 130,000 patient days divided by 365 days, which equals 356 beds. The occupancy factor is 2% times the square root of the average number of occupied beds, or, 47 beds. The number of acute beds needed in community X is 356 beds plus 47 beds, or 403 beds. Using 4 times the square root of the average number of occupied beds would yield an occupancy factor of 76 beds or, an acute bed need in comunity X of 356 beds plus 76 beds or 432 beds. In each example community X is experiencing a shortage of acute beds, 3 and 32 beds, respectively. on would almost double any given hospital's daily admissions overnight thus the greater occupancy factor was needed. Bed needs may be predicted utilizing the "D. P. F. Concept",7 whereby, hospital beds are grouped into Distinctive Patient Facil- ities (D.P.F.). "A distinctive patient Facility may consist of one or many nursing units in a given hospital. Thus, three 15—bed pediatric wards in one hOSpital are equivalent to one 45-bed nursing unit, provided that any pediatric patient can equally well occupy any one of the 45 beds which is available. Such a unit is considered to be one 45-bed pediatric D.P.F.. If, however, 15 of the beds are set aside for infants and the rest for other children, then there are two D.P.F.'s. Three 15-bed pediatric units in three different hospitals in a community would be con- sidered as three different D.P.F.'s unless the choice of which unit a patient entered was based entirely on the availability of space. Beds which are unoccupied in a hospital are insurance against the risk of not having enough beds when the number of patients goes higher. The 'premiums' for this insurance are made up by the cost of having unoccupied beds and include uncompensated depreciation on the facilities, and the cost of staff who are partly idle while beds are unoccupied. The former is almost negligible, while the latter is substantial. The 'benefits' of this insurance result in preventing the increased disability of patients which may result from a shortage of beds because of (1.) delay in admission of those needing hospitalization, (2.) the necessity of placing a patient in a substitute or inadequate hospital facility, and (3.) the premagure discharge of a patient to make room for a new one. The provision of enough facilities to give adsolute protection for the largest conceivable patient load is not economically feasible because there is always the chance that some epidemic or other catastrophe 7Mark S. Blumberg,"'D. P. F. Concept' Helps Predict Bed Needs," Modern Hospital, Vol 97. No. 6. (December, 1961) p. 75 8ibid. will overload facilities that are more than ample for ordinary needs. The problem is rather one of deter- mining what chance of overloading in a given service can be tolerated. Unpublished studies of data from hospitals in several communities have indicated that daily (midnight) census figures on a D.P.F. are generally Poisson—distributed. The poisson distribution is a form of skewed bell-shaped curve in which the entire shape of the carve may be predicted when only the average is known. The daily census may be expected to be poisson- distributed when the occurrence of the condition requiring hospitalization is random, and only a small proportion of the eligible population falls sick at one time. Obstetrics is a good example, although induced labor on weekends may alter the distribution. Any service, when admissions are governed by convenience, such as elective surgery, with heavy admissions early each week, is probably not poisson-distributed." Poisson distributions have been applied utilizing the queuing 11 Possion distribution theory as they apply to the delivery suite. means that independant events occur at random and the probability of this occurrence in an increment time interval is small, the number of such events occurring in a fixed interval of time will follow a predictable pattern known as poisson distribution. The Delivery Suite, like the Emergency Room, is not subject to the usual control that can be used to predetermine the load of the Operating Room or the X-Ray Department. Emergencies do occur in the latter departments, but, they are far outweighed by the great bulk of scheduled, I 9ipig. p. 70 10ibid. p. 78 11John E. Thompson, Oscar Wade Avant, and Ellawyne D. Spiker, ”How Queuing Theory Works for the Hospital," The Modern Hospital, Vol. 94. No. 3, (March, 1960) p. 75. noncritical activity that can be shifted, postponed or, if necessary, canceled, without any serious consequences. The hospital "unnot schedule the arrival of the patient who requires the services of the Delivery Suite, nor is it possible to admit patients selectively so that their stays will fill a predictable length of time. Admissions to the Delivery Suite are random and independent, and the length of time they may stay varies a great deal. There are, however, still limits as to the probable number of patients who will require these facilities at any one time and, according to Thompson, et. al., these limits and the probability of any given number of patients being in the maternity suite at one time can be determined with reasonable accuracy, through application of queuing theory. The queuing theory, an extension of the law of probability which holds that the number of people in a facility at any one time will follow a certain pattern around the average number of people in the facility in a given pericd. The curve showing the frequency with which any given number of people will be observed in the delivery suite is known as the poisson dise tribution and is predictable. The delivery suite application differs from the industrial application in that although the approximate service demand might be predictable as to the number of patients requiring the use of the facility, no waiting queue may ever exist and so a certain staffing complement must always be ready to provide the necessary service. Joseph P. Peters points out some of the uses of time series 12 statistics. The graphic presentation of the daily census or of the 12Joseph P. Peters, "Facts at your Fingertips - Part 2: Some Uses of Time Series Statistics in Hospital Administration," The Modern Hospital, Vol. 80 No. 2. (February 1953) p. 73. average daily census during a specific period of time over a period of several years, will show that there are high points and dips in any one calendar year, and when comparing from year to year, certain recurrences can be observed. These recurrences, or seasonal vari- ations, are important from planning and operational points of view. Because all hospital activities focus on the patient, an increase or decrease in the average daily census, for example, results in a corresponding change of over-all hospital activity. The number of meals prepared, the amount of expendable supplies and drugs issued, income earned from patients, and related happenings are affected accordingly. Time series data may reveal the recurring, wave-like increases and decreases of economic activity. These may occur with certain degree of regularity and are termed "cycles”. Trends of a steady increase of patients, as have been experienced by hospitals across the nation, will also appear in time series information. In addition, there are the irregular variations, disasters, epidemics, and episodic occurrences. "However, unless one has some understand- ing of the behavior patterns of prospective hospital patients, fore- casting future admissions or average census based solely on statis- tical trends can often be a seductive illusion...". "Hence, the administrator must be ever wary of over-generalizations even though they are based on well grounded statistics." "In short, time series statistics are an extremely useful tool if the administrator well realizes their capabilities and limitations."13 131bid. p. 76 A "reverse technique" utilized to analyze patient service demand is that of considering nonutilization of beds. As Ray E. 14 Brown points out, "When we remember that 'available' bed is defined as a bed fully staffed and ready for occupancy by a patient, and that payroll makes up 64% of the hospital's total operating costs, we can safely assign the major portion of the difference in cost per available bed and the cost per occupied bed to unutilized payroll. The above computations support the generalization often repeated by hospital administrators to the effect that avoidable costs represent considerably less than half the hospital's total costs. It also supports the axiom that 'the empty bed is the costly bed'. The effect of the empty beds on operating costs being so significant, the question immediately arises as to why hospitals have not done a better planning job so as to minimize unutilized beds." Morris London and Robert M. Sigmond 15 state that, “On a typical day in the United States last year (1960) one out of every four beds was empty in nonfederal general hospitals. This represented a daily total, on the average, of 150,00 empty beds." Some of the factors suggested by the suthors, preventing a hospital from operating at an occupancy of 100 per cent are seasonal, week—end and holiday drops in census, the need to segregate beds by service, pay status of patients and accommodation, necessity to hold beds open for peaks in demand,emergencies, disasters, house cleaning and maintenance, and so forth. The authors suggest that hospital administrators should be as conscious of the "Vacancy Rate" as the rate of occupancy, if not more so. 14Ray E. Brown, "The Nature of Hospital Cost," Hospitals, Vol. 30, (April 1, 1956) p. 39. 15Morris London and Robert M. Sigmond, "Are We Building Too Many Hospital Beds?" The Modern Hospital, Vol. 96, No. 1. (January 1961)p.59. 10 As Brown suggests16 "The average per cent of unoccupancy during a year multiplied by the hospital's bed complement equals the number of unused beds and indeed the amount of idle investment." The above brief explanation of the various ways of ascer— taining patient service demand indicates the importance and com- plexity of this matter. It must be noted, however, that not all of the above mentioned methods for estimating patient service demand will apply to one given hospital or community. With some hospitals or communities several methods may apply, but with others none of the methods may be suitable. 16Ray E. Brown, “Let the Public Control Through Planning," Hospitals, Vol. 33 (December 1, 1959) p. 35. III. FACTORS INFLUENCING THE PATIENT'S ADMISSION TO THE HOSPITAL When discussing the importance and complexity of ascertain- ing the patient service demand it can be noticed that in each theory statistical mechanisms were present, or recognition was given to the fact that the patient service demand fluctuated in one way or another and at one time or another. Let us now look at some of the factors mentioned previously and then attempt to analyze whether or not their effect on admissions could be shown at Edward W. Sparrow Hospital. In-school versus Out-school Factor It has been the observation of this author, from his experi- ences as an Admitting Office employee at Sparrow Hospital, that the scheduling of school vacation time produces a specific effect upon the number and type of admissions. Hospital personnelA in contact with children surgical patients will emphatically agree that during school vacation periods there is a considerable increase in child "T & A." (tonsil and adenoid) patients. Employees on the Eye, Ear, Nose and Throat Nursing Station at Sparrow Hospital, in the Laboratory, Surgery, Recovery Room, Admitting Office and even Dietary departments, all have found this to be the case. Seasonal Fluctuations Seasonal fluctuations are also believed to have a great effect upon hospital admissions. "It must be conceded that hospitals were subject to great seasonal fluctuations before the discovery ASee Appendix A p. 47 11 12 of sulfonamides and the antibiotics. During the 1930's one could expect a large influx of pneumonia and infectious bacterial diseases during the winter months which would raise the occupancy of the hospital to 100 per cent or more. During the summer months. there would be a drop in bacterial disease which would bring occupancy down to as low as 50 per cent. The situation was inevitable. Hospital people tried to cope with it as best they could. The public and the medical profession got a great deal of comfort from the empty beds standing by because epidemics were frequent at the beginning of this century. In fact there were entire hospitals for infectious diseases which were almost vacant during a large part of the year, but they were kept standing by. This situation no longer remains. The disease picture in hospitals has undergone a considerable change. The inevitability of seasonal change can no longer be accepted as it was 25 years ago. Despite this fact, fluctuations in seasonal occupancy sill seem to follow the pattern of 25 years ago. Annual studies conducted by Hospital Management magazine reveal an almost identical seasonal ocgupancy pattern year after year for general hospital." Louis Block18 concurs with the before-stated opinion regard- ing seasonal fluctuations stating that "...this has been explained by the increased prevalence of respiratory infections requiring hos- pitalization during the winter months, the decreased use of hospitals for children's services in the summer, and the drop in elective pro- cedures during the summer vacation season." In an article entitled "Hospitals Do Little to Leve113ccupancy Rates"19 Modern Hospitpl points out the results of a study whereby "...77 per cent of the hospitals surveyed reported seasonal fluctu- ations in occupancy, only about a third of the administrators were 17C. U. Letourneau and M. Ulveling, "Vacant Hospital Bed—-A Study of Occupancy," Hospital Management. Vol. 88. (October 1959);» 48. ) 18Louis Block, "Bed Occupancy," Hospital Topics, (October 1956 p. 42. 19"Hospitals Do Little to Level Occupancy Rates," Modern figspital Vol. 96. (July 1961) pp. 86-87. 13 doing anything to overcome them." This article cited vacations as the chief cause of seasonal fluctuations: "...two-thirds of the respondents attributed these seasonal fluctuations to patient or physician vacations, or both." In the "Small Hospital Questions" section of the Mpdggp Hospital magazine20 the panelists answer a question, pertaining to the low average occupancy -- less than 50 percent -- in a community having a seasonal industry, by stating that if surplus cannot be earned during the high occupancy season then “... the hospital must cut costs during the low occupancy period by closing floors, departments or even services, in order to achieve a reason- able balance of expense and revenue." To cope with the recurring problem of finding beds for the many extra patients each winter, the hospitals in the London, England, area implemented a warning system "...for reducing, in times of stress, the number of non-urgent admissions. This system depends on the proportion of patients, referred by practitioners, for whom admission is secured."21 Temperature Changes In conjunction with the season it seemed only logical that there may be a correlation of the effect of change in temperature, that is divergence in temperature from the normal, upon admissions. 20Jewell M. Thrasher, §£.al., "Small Hospital Questions." 15mg Hospital, Vol. 87. No. 6. (December 1956) p. 47. 21"Winter Admissions To The Hospital." L ncet, Vol. 2. (November 7, 1953) PP- 975-976. 14 The thinking here being, for example, that when the temperature rises above the expected normal for the summer months, the increased effort required to breathe. coupled with unaccustomed exercise or illadvised strain, precipitates a heart failure because of known or unknown cardio-vascular trouble or as a further complication of some other physiopathological disorder.B Weedends and Holidays How do weekends and holidays affect occupancy? Since there is no evidence that serious illness i.e., hospitable illness, regularly declines over weekends and holidays, why should there be a decline in the census during these periods and for that matter what group of hospital admission, i.e., what specific patient service demand classification is responsible for the decline? The Hospital Council of Western Pennsylvania22 conducted a study of this occupancy pattern in 14 hospitals. They concluded that "...the decline during these periods reduces total occupancy by only a few percentage point." They state that personal considerations of patients and their families, established work patterns of physicians and various hospital routines have a definite effect on the timing of hospital admissions and dis- charges. They found that the peak census day was on Monday and Tuesday BSee Appendix B. p. 53 22Morris London and Robert M. Sigmond, "How Weekends and Holidays Affect Occupancy," ng.flpg§:phflp§pipgl, Vol. 97. No. 2 (August 1961) pp. 79-82 15 and that the average daily census declined slightly Wednesday through Friday, and reached the lowest level on Saturday. Census rose again on Sunday because of the large number of admissions which character- istically occur on that day, but it was still below the average for the week as heavy weekend discharges continued. They point out that the fluctuation of the medical-surgical census by day of week was almost identical with the over—all pattern. Pediatric census had the widest fluctuation by day of week, and maternity having the narrowest range of census fluctuation by day of week, although discharges from maternity were heaviest on Sunday, which appears to reflect a variety of social and economic as well as professional considerations. It would seem that the hospitals have elective discharges as well as elective admissions. Regarding the effect of Thanksgiving, Christmas and New Year's on the census it was shown that a 15 percent, 40 percent and 18 per cent below normal occupany, respectfully, was experienced.23 "The holiday decline in census was not limited to the day of the holiday, but began one or more days in advance and continued for a number of days afterward. The depressant effect which holiday periods had on census was much less among hospitals with relatively high occupancy than among those with relatively low occupancy." Also it was noted that "Declines in census during holiday periods were much greater than the declines over weekends and had a greater effect on over all occupancy." 23ibid. \l' 16 Ray E. Brown24 pointed out that "...alnost universally the variation is downward by about 15 per cent over weekends as compared with other days of the week. This tendency of the patient to observe a five—day week is almost equalled by a similar tendency to observe holidays and vacation months. The average occupancy nationally drops more than 16 per cent during the heavy vacation month of August as compared with February." At present it is acknowledged that occupancy fluctuation occurs at least when studied nationally or when the data of several hOSpitals is combined. There is apparently some advanced thinking regarding this eccentric use of facilities. M. A. Simpson25 stated, "...obviously some patients will need inpatient care at the weekend, but. would it be beyond the powers of clever organizers to arrange for others -- I would suggest as many as two-thirds to spend the weekend at home? Will the day come when private cars and ambulances arrive at the hospital to take patients home at 5 P.M. on Fridays and return them at 9 A.M. on Mondays?" One hospitalz6 is aware of this week—end census slump and has implemented what they call a "Week-end 'Resort' Special" program, 24Ray E. Brown, "Let the Public Control Thru Planning," Hospital . Vol. 33. (December 1, 1959) p. 37 25M. A. Simpson, "Monday-to—Friday Wards," Lancet, Vol. 1. (April 30, 1960) p. 977 26Mark Berke, "Week—end 'Resort' Special," Hospitals, Vol. 28. (September, 1954) P. 75 17 which offers short-term treatment on the hotel plan. "It is especially designed to accommodate busy businessmen, housewives who profer leaving home when their husband is there to take care of the children, working people who might lose a day's salary during the week and persons from out-of-town who wish to make use of the hospital's extensive therapeutic and diagnostic services. Patients requiring short-term hOSpitalization for surgery or therapeutic and diagnostic work enter this program at their physician's request, but the unusual feature is that the patients are free to come and go as they please, leave the hospital for dinner or a movie or to attend any other business or social engage- ments during their stay." This is indeed one imaginative way of marketing hospital services to create a patient service demand during the times when business activity is needed to cover expenses. IV. SELECTION OF DATA Census information for each separate interest area studied was obtained from the Edward W. Sparrow Hospital Official Census Book. The midnight patient census for each day of the year 1960 was compiled on a master form for the following patient classifications: (1.) Total Patients - (2.) Men Patients a. Total b. Surgical c. Medical (3.) Women Patients a. Total b. Surgical 0. Medical d. Obstetrical (4.) Children Patients a. Total b. Surgical c. Medical d. Newborn (Nursery) (5.) Births The master form was arranged by month. Special entries were included to indicate weekday (Monday, Tuesday, Wednesday, Thursday, and Friday) versus weekends (Saturday and Sunday), holidays (New Years, Memorial Day, Independence Day, Labor Day, Thanksgiving, and Christmas), the variation of temperature from the normal, the amount of precipitation, and the days when school was either in session or not. Temperature variance from normal and precipitation readings were obtained from the United States Department of Commerce, Weather Bureau's "Local Climatological Data." The official temperature and CSee Appendix C. p. 56 18 19 precipitation readings by the Bureau were made at the Capital City Airport, Lansing, Michigan. The school calendars were obtained from the Lansing Board of Education, the East Lansing Board of Education, and the Lansing Parochial School Board of Education. The calendars were compared and found to be almost identical for the school year of 1960, with the exception of occasional Holy Days in the Parochial System. For the statistical study the calendar of the Lansing School Syetem was used. 20 V. STATISTICAL PROCEDURE When analyzing the data of the different interest areas, two statistical procedures were utilized. The first was analysis of variance, single variable of classification. This procedure "...con— cerns a comparison of the means of the ... populations, and the parts of the sample variance are analyzed for this purpose."27 The analysis of variance is based on the fact that "...if means of subgroups are greatly different the variance of the combined groups is much larger than the variance of the separate groups." In the single variable of classification all individuals, in this study patients, were classified into exactly one of two populations. For example, in the In—school versus Out-school interest area the various patient populations were segregated into groups defined by the variable of school being either in session or not. The hypothesis was that there is no difference in the means of the two populations. The F Statistic was at the 5 per cent level of significance. The second statistical tool utilized was that of linear regression. In the temperature and precipitation interest areas the variation in patient census was studied and compared with particular changes in the temperature from normal, and the amount of precipit- ation. They are then the "regression of patient census on temperature effect" and, "regression of patient census on precipitation". In the application of the linear regression theory, the mean of the dependent 27Wilfrid J. Dixon and Frank J. Massey Jr., Introduction t9 Statistical Analysis, 2ed Edition, (New York: McGraw-Hill Book Company 1957) pp. 139-140 ‘ 21 measurement, (the patient census) and the independent measurement (the temperature effect) were calculated. The variance of the means of the patient census was calculated as were the regression coefficients and the "t Statistic". The critical region was established at the 5 per cent level of significance. VI. FINDINGS In—school Versus Out-school Patient Census Comparison The first interest area analyzed was the In—school versus Out—school patient census comparison. Each patient census popul- ation, e.g., Total Patients, Men Surgical, Men Medical, Women Surgical, Women Medical, Women Obstetrical, Children Surgical, Children Medical, Children Nursery, and Births was analyzed using an analysis of variance procedure and a 5 per cent level of significance. The hypothesis was that the mean of the patient census during In—school periods equaled the mean of the patient census during Out-school periods. The results, Tables I through XI, indicate, with one exception, that the hypothesis was true and therefore, the school schedule had no effect upon patient admissions. The one exception to this finding was that Men Surgical patients and their census apparently varied with the school calendar. TABLE I. Total Patients Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 997.56 1 997.56 1.5958 Within 9376.44 15 625.09 . Total 10374.00 16 F RATIO STATISTIC F(1,15) F(.95) = 4.54 22 Men Surgical 23 TABLE II. Sum of the Degrees of Means Variance Squared Freedom Square F Ratio Means 134.67 1 134.67 4.5450 Within 444.39 15 29.63 Total 579.06 16 F RATIO STATISTIC F (1,15) F(.95) = 4.54 TABLE III. Men.Medical Sum of the Degrees of Means Variance Squared Freedom Square F Ratio Means 4.72 1 4.72 0.0515 Within 1374.22 15 91.61 Total 1378.94 16 F RATIO STATISTIC F(1,15) F(.95) = 4.54 TABLE IV. Women Surgical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 164.71 1 164.71 4.0124 Within 615.76 15 41.05 Total 780.47 16 F RATIO STATISTIC F(1,15) F(.95) = 4.54 Women Medical 24 TABLE V. Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 179.71 1 179.71 1.9722 Within 1366.76 15 91.12 Total 1546.47 16 F RATIO STATISTIC F(1,15) F(.95) = 4.54 TABLE VI. Women Obstetrical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 23.33 1 23.33 1.5870 Within 220.43 15 14.70 Total 243.76 16 F RATIO STATISTIC F(1,15) F(.95) = 4.54 TABLE VII. Children Surgical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio .Means 6.47 1 6.47 0.6393 Within 151.77 15 10.12 Total 158.24 16 F RATIO STATISTIC F(1,15) F(.95) = 4.54 25 TABLE VIII. Children Medical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 5.23 1 5.23 1.2482 Within 62.89 15 4.19 Total 68.12 16 F RATIO STATISTIC F(1,15) F(.95) = 4.54 TABLE IX. Children Nursery Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 84.19 1 84.19 3.1043 Within 406.87 15 27.12 Total 491.06 16 F RATIO STATISTIC F(1,15) F(.95) = 4.54 TABLE X. Births Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means .33 1 .33 0.1000 Within 49.56 15 3.30 Total 49.89 16 F RATIO STATISTIC F(1,15) F(.95) = 4.54 26 TABLE XI. Summary Table Hypothesis Interpretation Patient Populations Accept/Reject of Comparison Total Patients Accept No Effect Men Surgicals Reject Higher In-school Men Medical Accept No Effect Women Surgical Accept No Effect Women Medical Accept No Effect Women Obstetrical Accept No Effect Children Surgical Accept No Effect Children Medical Accept No Effect Children Nursery Accept No Effect Births Accept No Effect Weekend Versus Weekday Patient Census Comparison The Weekend versus Weekday Patient Census comparison was the second interest area analyzed. As in the before mentioned interest area, the same patient census populations were considered. The hypo— thesis was that the distribution of the patient census during the weekend equaled the patient census during the weekdays. The calcul- ation results, Tables XII. through XXII. Indicated that for half of the patient census populations the hypothesis was accepted: that the Men Medical, Women Medical, Women Obstetrical, Children Medical, and Children Nursery census population distributions were statistically the same over any seven day period. The other five patient popul- ations did have census variations from the weekend to the weekdays. The first four of these, Total Patients, Men Surgical, Women Surgical, and Children Surgical populations, were found to be higher on Monday through Friday. Conversely, Births occurred more often on the Saturday and Sunday weekend than on the weekdays. 27 TABLE XII. Total Patients Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 8376.92 1 8376.92 6.9924 Within 59899.77 50 1198.00 Total 68276.69 51 F RATIO STATISTIC F(1,15) F(.95) = 4.54 TABLE XIII. Men Surgical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 228.50 1 288.50 14.7705 Within 773.49 50 15.47 Total 1001.99 51 F RATIOS STATISTIC F(1,15) F(.95) = 4.54 TABLE XIV. Men Medical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 102.48 1 102.48 0.6852 Within 7477.27 50 149.55 Total 7579.75 51 F RATIO STATISTIC F(1,15) F(.95) = 4.54 28 TABLE XV. Women Surgical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 415.56 1 415.56 9.4103 Within 2207.89 50 44.16 Total 2623.44 51 F RATIO STATISTIC F(1,5o) F(.95) = 4.04 TABLE XVI. Women Medical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 56.08 1 56.08 0.2919 Within 9606.15 50 192.12 Total 9662.23 51 F RATIO STATISTIC F(1,50) F(.95) = 4.04 TABLE XVII. Women Obstetrical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 30.77 1 30.77 1.9111 Within 804.92 50 16.10 Total 835.69 51 F RATIO STATISTIC F(1,5o) F(.95) = 4.04 Children Surgical 29 TABLE XVIII. Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 880.71 1 880.71 57.1518 Within 770.29 50 15.41 Total 1651.00 51 F RATIO STATISTIC F(1,5o) F(.95) = 4.04 Children Medical TABLE XIX. Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 8.48 1 8.48 1.1276 Within 376.19 50 7.52 Total 384. 67 51- F RATIO STATISTIC F(1,50) F(.95) = 4.04 Children Nursery TABLE XX. Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 4.33 1 4.33 0.2052 Within 1055.12 50 21.10 Total 1059.44 51 F RATIO STATISTIC F(1,5o) F(.95) = 4.04 3O TABLE XXI. Births Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 37.25 1 37.25 8.3333 Within 223.68 50 4.47 Total 260.92 51 F RATIO STATISTIC F(1,50) F(.95) = 4.04 TABLE XXII. Summary Table Hypothesis Interpretation Patient Populations Accept/Reject of Comparison Total Patients Reject Higher-Weekdays Men Surgical Reject Higher-Weekdays Men Medical Accept No Effect Women Surgical Reject Higher-Weekdays Women Medical Accept No Effect Women Obstetrical Accept No Effect Children Surgical Reject Higher-Weekdays Children Medical Accept No Effect Children Nursery Accept No Effect Births Reject Higher-Weekends Holiday Patient Census Holiday Patient Census was the third interest area considered. The hypothesis was that the mean of the patient census during holidays equaled the mean of the patient census during the month in which the holiday occurred. The calculation results, Tables XXIII. through XXXIII., indicated that in four of the ten patient census populations this hypothesis was rejected. Total Patients, Men Surgical, Women Surgical, and Children Surgical populations did experience a census drop on holidays as compared with their census during the holiday month. 31 TABLE XXIII. Total Patients Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 6075.10 1 6075.10 5.2485 Within 15047.30 13 1157.48 Total 21122.40 14 F RATIO STATISTIC F(1,13) F(.95) = 4.67 TABLE XXIV. Men Surgical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 221.14 1 221.14 7.7159 Within 372.59 13 28.66 Total 593.73 14 F RATIO STATISTIC F(1,13) F(.95) = 4.67 TABLE XXV. Men Medical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 3.34 1 3.34 0.0256 Within 1689.60 13 129.97 Total 1692.94 14 F RATIO STATISTIC F(1,13) F(.95) = 4.67 32 TABLE XXVI. Women Surgical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 1081.20 1 1081.20 21.3718 Within 657.73 13 50.59 Total 1738.93 14 F RATIO STATISTIC F(1,13) F(.95) = 4.67 TABLE XXVII. Women Medical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 100.81 1 100.81 0.5297 Within 2473.59 13 190.28 Total 2574.40 14 F RATIO STATISTIC F(1,13) F(.95) = 4.67 TABLE XXVIII. Women Obstetrical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means ’69.14 1 69.14 3.2736 Within 274.59 13 21.12 Total 343.73 14 F RATIO STATISTIC F(1,13) F(.95) = 4.67 33 TABLE XXIX. Children Surgical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 378.16 1 378.16 5.8556 Within 839.59 13 64.58 Total 1217.75 14 F RATIO STATISTIC F(1,13) F(.95) = 4.67 TABLE XXX. Children Medical Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 7.62 1 7.62 1.4216 Within 69.71 13 5.36 Total 77.33 14 F RATIO STATISTIC F(1,13) F(.95) = 4.67 TABLE XXXI. Children Nursery Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 13.89 1 13.89 0.8070 Within 223.71 13 17.21 Total 237.60 14 F RATIO STATISTIC F(1,13) F(.95) = 4.67 34 TABLE XXXII. Births Sum of the Degrees of Means Variance Squares Freedom Square F Ratio Means 18.30 1 18.30 1.7562 Within 135.43 13 10.42 Total 153.73 14 F RATIO STATISTIC F(1,13) F(.95) = 4.67 TABLE XXXIII. Summary Table Hypothesis Interpretation Patient Populations Accept/Reject or Comparison Total Patients Reject Holiday Drop Men Surgical Reject Holiday Drop Men Medical Accept No Effect Women Surgical Reject Holiday Drop Women Medical Accept No Effect Women Obstetrical Accept No Effect Children Surgical Reject Holiday Drop Children Medical Accept No Effect Children Nursery Accept No Effect Births Accept No Effect Analysis of Temperature and Precipitation Affects The analysis of temperature and precipitation affect on the patient census populations was the fourth interest area analyzed. The hypothesis was that the patient census population was independent of the temperature. The calculations, Table XXXIV., indicated that all patient census populations, except Men Surgical and Births were 35 independent of the temperature variations. It appeared statistically, that the Men Surgical population and Births were dependent upon the temperature variable. Regarding the patient census populations with respect to preticipation, Table XXXV., all populations were found to be independent of the latter variable. TABLE XXXIV. 36 The Analysis of Variance in Regression for Temperature Patient Populations N X S B Sx N-1 B t(Byx=0) yx yx yx Total Patients 26 331.38 33.49 —0.0696 222.62 -0.941 -0.463 Men Surgical 26 44.46 5.73 —0.0720 222.62 —0.496 —2.797 29 Men Medical 26 56.19 12.53 +0.0086 222.62 +0.031 +0.153 Women Surgical 26 48.19 5.97 -0.0469 222.62 -O.337 -1.750 Women Medical 26 60.92 13.57 +0.0155 222.62 +0.059 +0.254 Women Obstetrical 26 41.38 4.14 +0.0170 222.62 +0.184 +0.914 Children Surgical 26 20.08 3.60 —0.0082 222.62 -0.104 -0.508 Children Medical 26 17.15 2.30 —0.0142 222.62 -0.271 -1.372 Children Nursery 26 44.23 4.05 +0.0230 222.62 +0.250 +1.263 Births 26 8.58 0.82 +0.0215 222.62 +0.118 +5.840 3O t STATISTIC DISTRIBUTION t(2,24) t(.95) = I 2.06 TABLE XXXV. The Analysis of Variance in Regression for Precipitation Patient Po ulations N X S B Sx N—1 B t(B —O) p yx yx yx yx_ Total Patients 26 331.38 33.29 +7.4754 3.1923 +0.145 +0.717 Men Surgical 26 44.46 6.56 —1.1368 3.1923 —0.113 —0.553 Men Medical 26 56.19 12.48 +1.8367 3.1923 +0.095 +0.470 Women Surgical 26 48.19 5.18 +1.7753 3.1923 +0.577 +1.095 Women Medical 26 60.92 13.42 +3.2702 3.1923 +0.157 +0.778 Women Obstetrical 26 41.38 4.21 -0.3453 3.1923 -0.053 —0.262 Children Surgical 26 20.08 3.61 -O.1454 3.1923 -0.026 -O.128 Children Medical 26 17.15 2.35 +0.7130 3.1923 +0.194 +0.970 Children Nursery 26 44.23 4.17 -0.5367 3.1923 -0.084 -0.411 Births 26 8.58 0.82 +0.0773 3.1923 +0.061 +0.300 t STATISTIC DISTRIBUTION t(2,24) t(.95) = i 28 t :33 (2x24? Elan (n—2) S yx 29 Y = 42.255 + (-.O720) x This is the relationship of the temperature effect. 30 = 8.9374 + ( .0215) x This is the relationship of the temperature effect. VII. DISCUSSION The primary purpose for undertaking an analysis of the In— school versus Out—school Patient Census comparison was to ascertain the effect of the school calendar on the Children admissions, although the effect upon the other patient populations was also of importance. Table II indicates that the Men Surgical patient pop— ulation does vary with the school calendar. It is curious to note that the Men Surgical Census rises when school is in session. This is a fact of importance in this study. The majority of surgical admissions are elective, that is to say, the patient and his doctor arbitrarily set a time for admission with the hospital when the operation can be scheduled. Having this in mind a possible reason why Men Surgical patients more often choose the school-in—seSsion time of the year for operations might be so that they are not incapacitated during their children's school vacation time when most families schedule their trips and visits. Even though calculations did not reveal any effect of the school calendar upon Children admissions, there is, nevertheless, a definite impression of this correlation upon the minds and working arrangements of many hospital employees. It is worth while to review the data used for this statistical analysis. The Sparrow Hospital Official Census data utilized in this interest area segregates Children admissions into three classifications: Children Surgical, Children Medical, and Children Nursery. These classifications range in ages from the nursery patient, i.e., the newborn, to the pediatric 37 38 patient up to the age of 12. The official census data does not ' distinguish the Children patients by age groupings. In the results of this study it was the author's opinion that the effect of the school calendar upon the admission of school-age children was being disguised by the admissions of children below school age. It can only be suggested that there is an "informal practice" on the part of the parents and doctors to "reserve" the vacation periods for school-children admissions and the school-in- session periods for the younger children. The effect Of this "in- formal practice" was to approximately balance out the admissions over a year's time. The Weekend versus Weekday Patient Census comparison revealed that there was no particular patient census fluctuation over a seven day period for non-elective admissions: Men Medical. WOmen Medical, Women Obstetrical, Children Medical, and Children Nursery. There was a difference, however, in census on weekends for the Total Patients, Men Surgical, Women Surgical, in particular the Children Surgical classification, and in the number of Births. In the Review of the 31 32 33 P Literature, several articles ointed out that sociological customs and psychological attitudes of our population gear our living to a Monday through Friday routine, and so when arranging for the 31London and Sigmond, Op. Cit. 328impson, Qp. 9i.- 33Berke, 0p. 91;. 39 surgical admission there seems to be a tendency to approach this as one does his occupation and wish to allocate only weekday time for the surgery. Women Obstetrical is the non-elective patient clas— sification that statistically does not fluctuate over a seven day period. In other words, the number of obstetrical admissions on Tuesday or Wednesday are statistically equal to those on Saturday or Sunday. In comparison with the Women Obstetrics results, Births were more frequent on weekends than on weekdays, even though the "expecting female" was equally likely to be admitted to the Maternity Department any day of the week. One possible explanation might be that the Thursday and Friday Obstetrical admissions were spending more time in the Labor Rooms and that parturition doesn't occur until Saturday or Sunday. Another possibility might be, as one author suggested,34 that there are more cases of induced labor and delivery over the weekend. The Holiday patient census comparison revealed that there was a Total Patient census holiday drop. Accounting for this, there was a holiday census drop in the Men Surgical, a significant one in Women Surgical, and in Children Surgical. Once again, the patient and his doctor schedule the elective admission so as to not interfere with the patient's and doctors holiday. In particular, the Women Surgical census dropped, possibly because the female in the home takes on additional culinary responsibilities for the festive holiday occasions, e.g., the Christmas dinner, and the July 4th picnic. 34Blumberg, Op. Cit. 40 The fourth interest area, the effect of temperature vari- ations, indicated that statistically only the Men Surgical census and Births are dependent upon the divergence of temperature from normal. It might be suggested that the unseasonableness of weather was an important factor in the male's decision—making process. May- be the male in our civilization is "out—door oriented" and there— fore if the weather is unusually hot or cold this brings about his decision to be or not to be admitted for his elective surgery. That Births are effected by abnormal variations in the weather may not be news to the "spinners" of old wives' tales, but there doesn't seem to be any obvious explanation for the situation. When considering the analysis of precipitation effect there wasn't any indication that precipitation or the variance in barometric pressure had any effect upon service demand in spite of the extreme amount of attention placed on, for example, the number of hear attacks brought on by over-exertion when shoveling snow. VIII. APPLICATION Based upon the findings of this study regarding the effect of the school calendar, the hospital may wish to designate additional beds on a nursing station for men surgical cases during the school year. Also the Surgery Department may have to make allotments for that portion of the schedule for male-type opeations during this period of time. The weekend versus weekday analysis points out the variance in service demand created by elective surgery. Having this infor- mation certain personnel staffing patterns could be created whereby, the Admitting Office, Department of Surgery, Recovery Room, Central Supply, the Special Diet Kitchen, and of course the nursing stations would be able to reduce their staff on weekends. It might be to the hospital's advantage to implement the distinctive patient facility plan whereby part or all of a male or female surgical nursing station could be closed down over the weekend or a skelton crew be utilized on specific shifts. It might be possible to designate specific bed complements or nursing stations for surgical operations where the total length of hospital stay was known, (e.g., a surgical operation with a 5—day stay, 4—day stay, 3-day stay, etc.) and thereby selec- tively program elective surgical admissions based upon length of hospitalization. This might enable specific stations to operate at a high per cent of occupancy during the weekdays and enable the staff to have weekends off. Of course this would result in a savings by the hospital on payroll expense, which amounts to approximately 70 per cent of the cost of operation. 0n the other hand, knowledge of 41 42 these census fluctuations by administration could result in imple— mentation of selective patient assignments: the available beds on the weekends could be marketed, as was suggested by the hospital having a "Weekend 'Resort' Special", and the Hospital would not only avoid a loss, but more actively fulfill one of the purposes of a hospital's existence--that of being a place for the promotion of good health and the prevention of disease, in this example by the creation of a practice of preventive medicine. It was noted that more Births occur on weekends. This should dictate the staffing pattern of the Labor and Delivery Rooms and possibly the Premature and Term Nurseries, i.e., have additional nursing personnel on Saturday and Sunday. The holiday census drop in the elective admission classific— ations again would indicate that personnel staffing patterns and the organization of nursing stations should be flexible to these census changes. It is obvious that census drops on weekends and holidays have a direct and most noticeble affect on the demands on nursing personnel and the utilization of beds, but also there are extensions of this "vacancy" into the functions of the Laboratory, Dietary, Radiology, Central Supply, and Pharmacy departments. Depending upon the seri- ousness of the medical patients and those surgical cases remaining hospitalized over the weekend and holiday, there can be a reduction in the number of Laboratory and Radiological procedures that must be done, the number of dietitians and dietary personnel to plan and produce the special surgical diets, and in the number of special solutions and equipment packs to be cleaned, made up, sterilized and 43 distributed by Central Supply. Likewise, the unseasonableness of the temperature and its affect on Men Surgical admissions and Births also calls for direct planning on the part of administration to make the necessary staffing and equipment arrangements for these situations. The importance of being aware of the patient service demand and understanding the variations or fluctuations that occur extend into almost every aspect of the hospital operation. That the patient is admitted and occupies a hospital bed or the reverse that the patient admission does not occur, has far reaching consequences, for a hospital bed is not "just a bed". It is also the myrid ofarmillary and paramedical services supporting the patient while he is receiving nursing care in his hospital bed. IX. CONCLUSION The patient service demand at Edward W. Sparrow Hospital, Lansing, Michigan, was studied for the following patient classific— ations: Total Patients Men Surgical Men Medical Women Surgical Women Medical Women Obstetrical Children Surgical Children Medical Children Nursery Births O O vvvvvvvvvv __\ OxOOOQOUlbwm—s The census information for each separate interest area was obtained from the Edward W. Sparrow Hospital Official Census Book. The effect of the school calendar, weekends, and holidays on the census in each of the above patient classifications was determined using an analysis of variance statistical procedure. (1.) Men Surgical Patient census was found to be higher during periods when school is in session. The other patient census clas- sifications are not effected by the school calendar. (2.) Total Patients, Men Surgicals, Women Surgicals, and Children Surgicals are higher on weekdays than weekends. There are more Births on week- days than weekends. There are more Births on Saturday and Sunday than on weekdays. The other patient census classifications showed no change over the seven day period. (3.) A census drop on holidays was found in the Total Patients, Men Surgical, Women Surgical, and Child- ren Surgical classifications. The other patient census classific— ations showed no change over holidays. 44 45 A linear regression statistical procedure was utilized in determining the effect of temperature and precipitation on the census. (4.) None of the patient census classifications exhibited any effect by the amount of precipitation. The effects of the above mentioned factors on the various patient classifications are summarized on Table XXXVI, Summary of Analyses. 46 pommmm oz poommm poommm oz poommm poommm oz unppzm poommm oz poommm oz poommm oz poomzm oz poozmm oz zpomzsz copuazno poommm oz poommm oz poommm oz poommm oz poozwm oz Heoficoz concafino pommmm oz poomzm oz poommm poommm poommm oz amazmzsm concazzo poommm oz poommm oz voommm oz poomzm oz poommm oz HmOfizpopmno coaoz pommmm oz poommm oz poowmm oz poommm oz poommm oz amazes: :maoz poozzz oz poozzm oz poozzm poozzz poozzz oz zeozmssm sosoz poommm oz poommm oz poommm oz poommm oz poommm oz Hmozcoz cox poommm oz poommm poommm poommm oz poommm Hwowmzsm cox poommm oz poommm oz voommm pommmm .poomzm oz mpeozpmm Hepoa cozpmpfidzomzm ozdpmpomsoe heUHHoz zealxoozllccmnxomz HoonomIPSOIIHoozomICH mommawc< myopowm mcoflpmofiMzmmmHo pcoflpwm momzfiesd mo mhsEEdm H>NNN mqm¢a APPENDIX A LETTERS FROM HOSPITAL PERSONNEL REGARDING IN-SCHOOL—-OUT-SCHOOL EFFECTS ON THE PATIENT CENSUS 47 November 29, 1963 Dear Mr. Plasman: On your recent question concerning the relationship of school vacations to the work load in the Laboratory, I have found that they are directly related. Especially during the Easter or Spring vacation and Christmas holidays, our total work load in the Hematology and Urinalysis sections are increased, as all admitted patients receive a routine CBC and urinalysis. This increase is primarily due to an increase in short- term surgery patients during school vacations. Also, other laboratory areas have increased procedure totals, such as, histology, as they process all tissues removed in surgery. So from my experience in the Laboratory, dating back to 1954, I have observed that school vacations increases the total work of the Laboratory. Yours sincerely, Ann Spencer, M. T. (A S. C. P. ) Hematology Section Chief Edward W. Sparrow Hospital 48 DIETARY DEPARTMENT A sharp increase in the children's trays for the T. & A. ward is definite and very marked at the beginning of all school vacations and continues in this manner throughout the vacation period. A normal number of trays ordered for this area is again resumed at the end of a vacation period. All trays going to this area are designated by age so an appropriate tray can be sent to the child followed by the post- operative tonsil regime after T. & A. surgery. Preparation is always made in advance of the immediate start of a vacation period to accomodate this increased load on the Dietary Department. Doris Cox, A.D.A. Executive Dietitian 49 EDWARD W. SPARROW HOSPITAL INTER-DEPARTMENTAL CORRESPONDENCE T0: Mr. Plasman FROM: Elisabeth Munter, R.N. SUBJECT: Operating Room Schedule During School Vacation DATE: November 29, 1963 The increase of number of children in the hospital during winter and spring school vacation is evident on the Operating Room schedule. This is a convenient time to have elective surgery done without having the children miss classes. The greatest increase is in ear, nose, and throat surgery. Ordinarily we have only one ENT Operating Room which accomodates eight to ten tonsillectomies per day on an average day. Far in advance of the winter and spring school vacations we receive requests from the ENT staff of surgeons to have a second operating room for tonsillectomies during the vacation period. With adjustments in operating room schedules and staff in operating room, recovery room and on the nursing units, we are able to provide this service. {4244—42 22% fl/ EM:ac 5O December 17, 1963 Dear Mr. Plasman: In the last few years we know that the census in the E.E.N.T. Depart- ment is much larger during vacation time; that is, Christmas and Spring Vacation especially. This is especially so with the pediatric age group. A second operating room is in operation during the periods to accom- modate the increase. Many adults are scheduled at this time also to avoid having to take sick leave from work. Teachers and students often elect this time to have surgery done. We plan a full staff to be available at these periods to handle the situation. Sincerely, W W, flea/7' Eleanor Purdy, R.N. Head Nurse, 1-Main 51 July 15, 1963 F. B. Plasman, Administration. Dear Mr. Plasman: The Admitting Office at Sparrow Hospital has been a very interesting position for me during the last fourteen and one-half years. The hospital has it's busy seasons the same as any other business. .At Christmas, Easter and summer vacation time, we are *set up to open two operating rooms plus the necessary bed section, so that children may have throat, eye or other elective surgery. Approximately twenty to twenty-five children per day are admitted through this period. It is very gratifying to know that the hospital has these facilities available so that children need not miss school classes during the year. This hospital averages eight to ten children per day as emergencies with either medical or surgical diagnosis on a year round basis. The accommodations are improving each year to help patients have elective surgery when it is convenient for them. Very truly yours, Admitting Supervisor 52 APPENDIX B LETTERS FROM PHYSICIANS REGARDING TEMPERATURE VARIATIONS ON PATIENT CENSUS 53 ROBERT M. STOW, M.D. 2909 EAST GRAND RIVER LANSING 12. MICHIGAN PHONE IV 9.6596 INTERNAL MEDICINE JANUARY 20, 1964 MR. F. B. PLASMAN EDWARD W. SPARROW HOSPITAL 1215 EAST MICHIGAN AVENUE LANSING. MICHIGAN DEAR MR. PLASMAN: THIS Is THE LONG OVERDUE LETTER I PROMISED TO WRITE YOU REGARDING SUDDEN CHANGES IN WEATHER AFFECTING HOSPITAL ADMISSIONS. OF COURSE. CERTAIN TEMPERATURE CHANGES GREATLY EFFECT THE TRAUMA AD— MISSION DEPENDING ON WALKING AND DRIVING CONDITIONS PLUS EXPOSURE TO VARIOUS SPORTS SUCH AS: SKIING. SWIMMING. SKATING ETECTERA. THESE ACTIVITIES CHANGE THE TYPE OF INJURIES APPRECIABLY. THESE I BELIEVE ARE SELF EVIDENT. IN ADDITION TO THE ABOVE THERE ARE THOSE CHANGES IN TEMPERATURE. SUCH AS A DROP, AS A CAUSE OF INCREASED ANGINA PECTORIS AND EVEN FRANK MYOCARDIAL INFARCTIONS. WHEREAS. A PATIENT MAY BE ABLE TO EXERT IN WARM TEMPERATURES WITHOUT PAIN: THE SAME EXERTION IN COLD WEATHER IS LIKELY TO PRECIPITATE ANGINA OR A CORONARY THROMBOSIS DUE TO SPASM OF THE VESSELS SECONDARY TO THE COLD AIR IN THE BRONCHI. ANOTHER DIFFICULTY IS THE PRECIPITATION OF CONGESTIVE HEART FAILURE DURING HOT WEATHER BECAUSE OF PATIENTS' INCREASED INTAKE OF SODIUM CHLORIDE IN THE FORM OF SALT TABLETS AROUND VARIOUS PLANT DRINKING FOUNTAINS OR INCREASED SOFT DRINK CONSUMPTION. FREQUENTLY THESE WILL PRECIPITATE THE FIRST EPISODE OF ACUTE PULMONARY EDEMA. I HOPE THIS IS THE INFORMATION YOU NEEDED AND IF FURTHER EXAMPLES ARE NEEDED PLEASE DON 'T HESITATE TO REQUEST IT. SINCERELY YOURS. ROBERT M. STOW. M.D. RMS/WA 54 EDWARD w. SPARROW HOSPITAL DEPARTMENT OF PATHOLOGY February 24, 1964 Mr. F. B. Plasman Edward W. Sparrow Hospital 1215 E. Michigan Avenue Lansing Michigan Dear Mr. Plasman: The issue of effective changes of temperature or of temperature per se on patient admission is a moot one. Temperature Changes appear to be not only inductive or condusive to the deve10p- ment of acute infectious diseases, such as, the common cold, bronchitis etcetera. But the temperature Change is also a significant factor in myocardial infarction or angina pectoris by inducing broncho—spasm due to the cold air of winter. Again,0ver exertion in winter time by certain patients appears to be a significant percipitating factor in the-deve10pment of infectmous diseases, respiratory difficulties, such as, bronchitis or asthma, but the cold air in Combination with exertion are involved in the deve10pment of myocardial infarction. Obviously seasons and temperatures Control or significantly influence a variety of activities, such as, the winter sports of skiing, ice skating, versus boating, swimming, golf, base- ball, etcetera. Each of these sports or activities may have "epidemic" of injuries. Frequently these are relatively spec- ific for the sports. Numerous other examples could be developed. I should be happy to go into these additional details with you at any time you wish. TOM, FM" 9 ohn F. Dunkel, M.D. ethologist 55 APPENDIX C MONTHLY MASTER FORMS 56 OmN ONO . u O NO (in hOQ. N n—n.— R1.N Om!— OvN.m O—O.N Own.— m—h.n NCO. '— XX XX O— n1 On Om NC Oh— 5O Nv OO— On» XX XX 01 ON as. ”Q 01 NO n. N: OOH On XX XX XX —O. a... O! ON 0' NO n‘ ma On OO— OO 11 O: Nhn an WOO ON a... .3 no Ow NO NO NO p OO or. O: ON 2. c... O— O' NN 0' h! OO— Oh —' —: hN XX BO. O— OO ON NN OO O' OO OQ OO— Oh —N— can ON XX O— NN On n1 O! Oh— NO ht Q: Own ON XX XX XX XX 3. No. n | — I = O— 'O ON ON O— . u — — OO MO On OO —O Oh : an OO— OO— QO Nh O? 1' O: O: 8O tin (N ON Dm JOOIUW U—JMDQ UZ—WZj MIL. k0 t(OZNl—> h—O GUT—07: Z— DUIDW0 DMZ—ZKUFMD m4. .02-WZj Z— UFUD m( DummwleU MESH-(mUn—ZUL. N XX no. all on N.- n— OO OQ 'h OO— OO b! On— Nnn NN XX XX —O. .F m I N I — O— N? mm m— m— n— n— Nh ”O 11 in NO —O O? 04.. —h— —O— OO Oh a? pm h: (N— OOO OOn pN ON XX no. a + NO On OO— O! nN— Nov XX v... O— Oh O- 2. hO O1 00 OO Oi ON— NOO XX XX XX XX xx xx .r ... n... ... an. .I .... n... .1. n... : Q— h w N O? am On in NO uN ON —N —N .N N— O— —N on N Oh Oh Oh Vb NO ho. OM NO On QM Oh Oh Oh R Oh on —O OO —O Oh Oh— Oh— 05— ..h— OO— OO OO Oh —O OO on N. N! O‘ O. O: nN— uN— hN— NO— hhn Ohm whn Nhn NOO h— Op n— (a O— oom— .>m_>DU 5.011.. OWZ_<.PMO (F(D WDWZUU JJ< XX XX 3‘ m— QN OO O1 Oh ON pm— OO NQ OO— man u F30 Z— ? JOOIUm WOZMv—MU; W>KNWKDZ I—(U—OU—Z JW JOOIUm 01..le UZ—WZd‘I— MIL. k0 t(OZMJm DNZ—S—CWPMD m< .07:ij 7: MPMD W< DmmmmKlXN MEDF(le$—NF 100m MDmZMU J(_O_h_k0 J> DE<>>DM ZOKk DMZ—(PDQ m<:mmm.._ man M —.PW JOOIUW U—Jmal UZ—WZS NIP LO C(OZUJ(U NIL. 20¢."— MJDOMIUW JOOIUW 1 EMF<>> LO mNIUZ_ Z- ONESM> .m.3 MIL- >m ONZ—S—m—UFMD m< .Uz_m2(l_ Z— ”F(O KjDU—Fm—(l F(IF 10.n— DMIm—JD(FWU MIDI—.(KULEUL: 2(U: DMPUNEXU NIP $.02."- ZO—L.<—>UD m< DMWWUEEXU MKDF> DE<>>DU 201k DMZ—(F00 (F(O MDWZMU JJ< — F30 Z— JOOIum MOZUXNM; W>KUMCDZ J(U_OM_2 J(U—U¢3W ...—(FOL. ZUEOJ—Io ...—(U—KFMFm no J(U—DM_)_ .....(U-Um—Dm JW JOOIUW U—Jmal 97:ij WI...- h—O C(OZNJ(U “WI-F ZOKk NJDDNIUN JOOIUm 1 NHL-(2’ .10 WUIUZ_ Z— OMIDM(M: .mDZ—X JJ< k0 ZO—F(P—l—UUCL O Dm DMZ-ZEWFUD m< .07:ij Z- MFUD m< DWWMUKn—XU UKDF>0E¢> DC<>>DM SOIL DMZ-(FOO (Flea WDWZUU JJ< — F30 17: JOOIum WOZMXUM; msrleO—JO—l— O 207—.(LL1-UUKQ N WCDFENI ZUI—I WIFE—m >EMW¢DZ ...—(”=sz J(U—@K3m JW JOOIUW U—Jmal UZ—WZS NIP to Em DUZ—S—KMPND m< .OZ—MZj Z— ”Pg EjDU—Fm(l F(IF [0h— DmIm—JflNO m< DMWWUKQXM MIDI—t(mulzwur N room MDWZWU J<_U_LLO I—(P—lmOI 3011(n—m .>> OK(>>DM 20ml DMZ—(F00 mummaz mam N. c. o. m. o. h. o. m. N .N NN N «N MN 3 ON 9 a. .N o. N N o. .N n. m. N m. h. «N .N J(U_ou.2 I moo Na 8 .. n. mN .... an «N .. o. m. «N n. ..N .N 3 v. o. u. o. N 3 3 m. m. «N NN 0N 2. mm NN 446.05.. m 6 Rod om mo 3 3 mo. no. no. 3 3 mu 3 .m 8 No 3 om no on 3 2 we no no 3 5 Nu on 2. 3 «a on 4.2.0... zu¢0J_IU NvN. . pm me we on .0 on N.. or. 2. an on ... km N... 9. cc Nv 3 3 N9 mm mm an on .n on 5N on 3” on on :20. mhmkm m0 8. . N 3 mm an B. an 8 3 no no on 3 no R 2. an an 8 R R No mm mu 2. 2. NN. .m on No 3 .o 2. ... <0 . a u s. 8.... . Nm 3. 2. an on an an 3 or. .n .m 3 B an we nu pm on .n 2. Na 2. .... mm me mm mm on em N» on J<05¢Dm 23... Nm. n... mm. on. an. 2.. 2.. 2.. mm. on. mm. no. .5. N2 mm. an. N... B. Nu. «a. No. mm. cm. .2 Nm. 8. on. No. Nb. 8. mm. J<...O... 2520?) 35.. 3. mm mm on an no 3 3 no ow mm 3 3 3 N5 mm on mm on be me 3. 3 .... Nm mm on mm. mm .o mm JW Joozom U—JDE UZ—WZj NIL. ....0 ¢> h—O mNIUZ— 7: OUC3W(U1‘ .mDZ—X JJ( ...—0 ZO—L.<.P—n.—UUK& O ....(Mmam EMT—F(m; .W.3 NIL. >fl DMZ—ZCUFMO m< .va—WZj 7: MF> OK(>>DM 20ml DWI—(F00 Nh.(n— MDWZNU JJ< — .50 N z. 4001 Um mozmxmm; 95.5.40: O ZO~F(.P—l — ONCE N UKDFKNKME SUI—l MIL-02m >¢mm¢32 J(U-DM$— J(U-UKDm I_(.PO.P ZUEOJ—Io J(U—K.FN.—.m DO J(0-DU—2 I—(U-UKDW J<.F0..—. INS—O; I—(O—OU—Z J(U—UKDW I—(Lunuh. ZN: — mL-ZM -.FW JOOIUW 07...le UZ—WZj HIP h—O C(OZUJ(U NIP SON—n.- MJDOMIUW Joo—LUM 1 EUP<>> k0 MMIUZ— Z— DMKDW(U2 .mOZ_V. JJ< k0 ZO—P(P—l—UU¢& O ....(mmDm EMT—F(w; .m.3 NIP \fm OUZ—ZEMFMD m< .07:ij 2— ”Fa m—(JDU—Lumzxn— F(IL- COL OMIm—J0(PWU “ESL-(CMl—ZUF ZNO m( OMWWUKIXM Mm—DFEMESMF N XX. Op. 0' 1— O1 O— —N O1 N1 11— ON OO NO NON XX XX XX XX .... O O —I —-l NI O1 Oh 11 hO— NO hO OO NON —N N— h1 O— OO N1 «O— O1 hO OOO ON O— O1 O1 NO OO— OO hO Oh NON O1 N— Oh O1 GO ON— hO Oh 1hN O— XX XX XX XX NO. O O I O I : O— 11 O1 Ow N— N— O OO OO N1 O1 hO OO h1 O1 ON— ON— OO —O ON OO OO —O hON OON h— O— XX XX XX XX XX XX OT OO Op ON Oh OO O1 OO— NO OO Oh OhN Op 1— hN NO O1 O1 OO OO— NO h1 Oh OON 1— ON. NI OO ON ..N OO hO O1 OO— O1 O1 OO OON O— 82 .52. O1 — N ON OO —1 O1 h1 ON— OO hOO N— O1 —N ON OO — 1 O1 O1 OOw —1 11 OO hOO XX O1 O— N— NO hO O1 OO— O1 NON O— XX XX XX O O — I O I O O— O1 h1 : ON hN ON .O NO- OO O1 O1 _.1 OO O1 ON» NO— hO —O —1 —O Oh NO OON O— O O O X X XX O OO . O I O I O h hO 01 O— N— hO OO hO OO NO hO OO N1 O1 h1 ON— ON— OO hO OO hO Oh 1h OON OON h O ZOOM ODMZNU J<_U_kk0 J(F_n_mOI 3°C¢> 01(2/0“ 20ml DMZ—(PDQ (F(O mDmZNU JJ< xx xx xx xx xx Xe. .ax .xx 0 o «a. an. ma. ml NT ml «I ...I .. n h q a. an «a on on a. o. w. m. w. o. v. on NN a. o. no om No .5 .o. on .n aN an m. an on NN .n «a on NN «a an Nu m.. nm .N NN. «m. on n» oN mN Nn N1 5N on h» an on No on no on NwN qu .mN nmN no» a v n N . F30 1 Z— JOOInvm WOZUXWN; OsrKmmw—DZ J(0—OM—2 I—(O—UEDM JW JOOIUW U—JflDn— UZ—OZj NIL. k0 E> .W.D MIL. >0 OMZ—S—EMFMO m< .UZ—WZS Z— UFw0 m< OUWWUIIXM HKDlemmleL. N XX .1. O1 — N ON NO —1 OO 11 ON— h1 ON Nh hON NN XX XX 8. 3. O1 ON 1— hh OO O1 OO —N— O1 ON OO OON —N XX XX OO. .1. O1 —N 1— Oh OO O1 —1 ON— ON nO OON ON XX XX XX XX XX XX XX XX O O O + O .f O —— O1 —O O— O— h— O— Oh 1O N1 O1 O1 NO . O1 11 —O— ..1— ON OO NO 1O OO Oh hON OON O— O— XX XX O O NO. hO. O O I O I h I O I p I O1 O1 .1 OO OO O— 1— — p h— O— —N O— ON ON —N OO hh hh Oh Oh N1 OO OO OO 1O O1 O1 N1 hO NO O1 N1 O1 —O 1O —O— hN— O— — —O— OO— OO hO OO OO 1O hO 1O hN ON ON Oh — h OO 1O OO OON OhN OON O1N OON h— O_ my 1— O— OOOF .FODOD< hO O— h— OO OO OO O1 1: 1O OO 1O h1N N— xx hO O— ON Oh —1 O1 ON— —O OO OO NON '— XX 1O O— ON 1h OO 11 hO 1—— —1 OO Oh 1ON O— XX —O. O— —1 O. N— Nh OO O1 OO O: OO OO OO OhN XX XX XX T... OO O— 1— OO OO O1 OO O: O1 OO —O OON XX 3. Nb. hO O— OO OO OO — 1 Ow— 11 ON Oh N1N XX N1 = N— OO O1 OO N1 ON— 01 ON OO OON XX 11 1— O— 1h 11 hO NO OO— O1 OO Oh th XX No. —O N— O— h1 h1 OO 11— O1 OO —O OOO XX ON— OO O— ON h1 N1 OO 11— O1 O1 O—O XX v... —O O— O— OO O1 O1 O1— O1 O1 OO O—O Zoom MDWZMO 171—0—th J(..F_n. WOI gomm<¢m .>> OK<>>OM 20th. DMZ—(PDQ (...-(D WDWZUU JJ< XX OO 1— O— hO 1O OO OO O1— O1 OO Oh O—O p F30 1 Z— JUDY—0m MOZUXUM; m><0 _ .... 0: O ZO—L-(F-n— —UU In N WCDFmmmmnz 4 .3025: 6 Jm JOOIUW ‘16—].le UZ—mzj HIP .u—O K> .m.D MI... \rm DUZ—S—KMFMD m< 392—.ij Z— UPNO m< Dummwmlxu ”ESL-(m—Ul—ZNF N 100m WDMZUU J<_U_..._LO I—(F-lmO—n— gomml‘lm .>> Own—(ZION 20ml.— ONZ—(Fmo (F(O MDWZUU JJ< — xx xx xx xx xx xx xx xx xx xx xx F30 xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx 3 z. Jooxom xx xx xx xx xx xx xx xx mozmxmmg xx 955on ... 8. o o o ... o .r ... o ... m..— o ... o o o no. r... o o o o o o o ... .r o ... n zo_._.<._._n:ommn._ :I ~+ ml n+ N... 3... u; ~+ a... ol .... ol ol a... «I .I ml .7 ml 3| ml ...? 2+ 2+ 5+ 3+ or. M... 3+ NI. N mm:...<¢un2m._u omu o— m o. a u— n 3 o m n a ~— ~— 3 v o m o 3 2 ~— 2 n N. o. m m m o h mIPEm as; 3 me S cm hm um 3 me 3n mm on .n we 3 S «3 me me 3. m3 m3 .3 3. me we 3 N3 3 m3 3 >mwmmaz «on on 2 o. o n. 2 o. 2 o. 2 2 2 mm 2 S n. m. 2 2 N. w. 2 MN Nu 2 _N .N 3 .u m. .33—ous— % n3 2 a. S... «a o. 3 Z n. aw n~ mu 3.. n h o. m. a. 2 m t 2 o. S 2 2 a e - 3 ca J> 8n.— mn m3 3. an cc 3 n3 2. we 3 on 3. m3 3 3 m3 F. «m on an 2.. 3 3 «3 me on «3 3. we 3 44535 .2 3Q. on 2. .3 cm on 3 R on m3 m3 3 we on on 9. 3 3 n3 9. 3 on on «3 m3 mm mm nu an on ma JM JOCIUW.U_JNDQ UZ—mzj NIL. k0 m—(DZUI—(U MIL. zomk UJDOUIUW JOOIUM Q mm.—.()> n.o mWIUZ- Z— DMKDW(U—Z .MOZ—v. JJ( ...—0 law—F(.P—l—UNCI n D> .m.3 MIL: >m DMZ—ZEUFUD m< .UZ—MZS Z- ”F(D Ejaouhfl(l FEB. KO... DWIM—Jmlrrmw UCDFéUn—ZUP 21‘“: OUPUUEXM NIP SOIL ZO-F(_>MO m< OUWWMKlXM MIDI—.(KMl—ZNF N 300m WDWZUU J<—U—h—k0 J<..F—lm01 30”.!(lm .>> DC<>>DM 201k DMZ—(F00 (P(D_JOI n— n ZO—L-(Lu—l—UUm—l Mal. N UCDFEUn—zul—l m mIFK—m nv \fGNWKDZ h— J(U—Dw2 n— I—(U—UEDM nh J(.FO.P ZMKDJ—IU ow J(U_¢.PU.PWDO om J(U-OU: n! I—(U—UCDW Nv— Jm I—OOIUfi U—JmDn— GZ—WZj MIL- .u—O C(DZMJ(U NIL- :OKE MI—DDNIUW Jooiom 1 EMP(>> .n—O mMIUZ— Z— DMKDM> .m.3 NI...- >m DMZ—ZKUPUD m< .OZ—mzj Z— ”F(D C(JDU—P¢(l F(IP CO... OMIm-JD(FWN WED-F(EMn—EWL. 2(M—2 OWL-UNIX” NIP 20ml.— ZO-F(—>MD W( DummMm—EXM UCDF> DMZ-£50” 20¢."— OUZ—(Luflo (F(D mDmZNU JJ< — xx xx xx xx xx P30 xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx 3 2. Joozom xx xx xx xx xx xx xx xx mozuxuu; xx 9530.40: ..—.. GO. ON. L. O O O O OO. o O O :o O no. on. .F O O O L. mm. O—. O no. —0. no. —O. NO. no. a ZO—F(.F-l—Uflr—m OI ”I O; 0!? RI. 0.... 0... on- at: m}. a... n.o: —.f .31 a... as. of 0+ OI bl ml. 1' nl :l 0| (I N... p... NI NI NMKDFEUEZUI—I hON 1— h o h an N— O— N— n h O 0 up Nu m on o o on : 1— O— O o h 0 o h C m WIFK—m paw; nn Nn «n Nm 0m QC MG mv Me. 03. h! h! 0‘ On 3 no. an an on NO in n! O! n? 0' Oi mv u! N! 01 >¢UG¢DZ 7 ma! 1 n— O— h— n— MN N— N— 3. O— N— n— O. a O— O an N— N— v— N. O— O— O— h— on O— Q— Q— m— J(U—OM2 6 Nan O— n— V: n— O. ON on 3 on n— up up n— rcN ON NN ON on n— O— —N “N ON MN h— h— —N —N ON pm I—(U—UCDM NO<.N no on «O NO be NO— ph mh mm mh Nh MN on on On Oh CO om hm mm hm DO n0 8 1h Mb 3 on on an J(.PO.F ZUKOJ—IU .hN.— Nn Nn Oi Oi a! fit mm 13 NM —1 N3 OQ N... on on NC at an on pm O! 01 an on in mm a! on hm mm J(U—KL-M.Pm $0 up. — Om pa h? N“ 3 an NO on N... «O N» On 10 O1 no 00 '0 no 00 OO NO O0 no No no On a in N» am JM JOOIUN U—JDDL UZ-WZS NIL. k0 IEZUJ(U NIL. SCI."— UJDOUIUW JOOIUM Q IUL..(>> ...-O MUIUZ- Z— DNIDfifl DUZ—ZIWFUO m< .52—WZS Z- “F(D I530—L-IUO m< DummuIn-XM MIDFIUn—SNP N Icon mDMZUU J(-0—LLO ...(P—l m0... >>OII>OM ZOIK DMZ-(FOO (F(D ”DOING l—I—( — xx xx xx xx xx xx xx xx xx xx xx xx xx ......0 xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx 32. Joozom xx xx xx xx xx xx xx xx xx uozuxmmg xx 953.40... ... ... no. ... ... .o. o ... .r oo. .o. ... .r oo. ... oo. oo. ... ... ... ... .o. ... ... ... .o. .o. o o o o ozo....(........ommn_ o+ o... .... ol ol ol o... 21 nol ool n1 nl nl .... nl nl 3... o... n1 3: o.l ol 3.. o1.l o1. o... n... n... ol nl oumahsnuno‘uh. ono o 3. n o o n o o. o. .. 3. o n o 3 o. o n n o. 3 o o 3. oo o n o n n. o mxk¢.m no3.. o3 o3 o3 oo .3 o3 33 o3 3n .n n3 no no on no o3 3 3 3n on .n on on no nn 3 n3 o3 on on on >cuumaz no3 oo o. n. o. o. n. 3. o. 3. n. o. n. n. oo o. n. n. 3. 3. n. o. oo oo .. o. o. o. o o. 3. .. J