INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. U n iversity M icrofilm s International A Beil & H ow ell Inform ation C o m p a n y 3 0 0 North Z e e b R o a d . A nn Arbor. Ml 4 8 1 0 6 - 1 3 4 6 USA 3 1 3 /7 6 1 - 4 7 0 0 8 0 0 /5 2 1 -0 6 0 0 Order N u m ber 9208813 P rofession al activities o f w om en and m en physicians in M ichigan. (V olum es I and n) McNiven, Margaret Deirdre, Ph.D. Michigan State University, 1991 C o p y rig h t © 1991 by M cN iven, M a rg a re t D eird re. All rig h ts reserved. UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106 PROFESSIONAL ACTIVITIES OF WOMEN AND MEN PHYSICIANS IN MICHIGAN Volume I By Margaret Deirdre McNiven A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of College and University Administration ABSTRACT PROFESSIONAL ACTIVITIES OF WOMEN AND MEN PHYSICIANS IN MICHIGAN By Margaret Deirdre McNiven The purpose of this study was to describe the professional activities of all women and men physicians (MDs and DOs) in Michigan. Women have entered USA medical schools in annually increasing proportions since 1970, when the single digit enrollment quota was lifted by medical schools and their universities. Greater access to medical education has led to changes both in the number of women physicians in the USA, and in their medical careers. Gender differences were analyzed regarding specialty, hours worked in hospitals and offices, geographical distribution, and professional inactivity rates. Nation of training and professional degree differences within gender differences were examined. The study examined whether such differences were related to age or years since graduation. This study was based on a secondary analysis of an amalgamated data base drawn from three archived surveys, the 1986 Physician Survey by the former Michigan Office of Health and Medical Affairs, the 1986 American Medical Association Physician Masterfile, and the American Osteopathic Association Physician Masterfile. The sources of data were matched to each other by name, address, professional degree and specialty of the respondents. The amalgamated database described all professional activities of 15,770 Michigan licensed physicians, MDs and DOs who were in Michigan in 1986 and for whom there was an identical entry in either the AMA or AOA Physician Masterfiles. There were 1,966 women included, or 12.5% of the total. Full time, fully trained physicians were selected as the ultimate unit of measurement, due to the uneven gender distribution of part time physicians, graduate medical education trainees, and geographical location. The results were that women and men physicians in Michigan in 1986 had different professional activities, specialty, patient care hours per week, and geographic location, whether MD or DO. There were greater similarities between more recent medical graduates. There were fewer distinct differences for the latter in hours of patient care, specialty practiced and geographical location, and small or nonexistent within certain specialties and practice types. Differences were smaller between USA trained MDs and internationally trained MDs, than between MDs and DOs. Women were entering nonprimary specialties at high rates. Women and men had the same rate of activity and remained active in their career at the same rates. Perhaps actual changes in norms were occurring as more physicians are women toward a gender-independent pattern. Copyright by Margaret Deirdre McNiven 1991 DEDICATION This inquiry is dedicated to two special family educational role models and mentors. My late father, William D. McNiven, Esq., inspired me with his love of learning and discovery, and his professional model. Roger D. McNiven, my eldest cousin, encouraged me to follow him to the United States of America. He died in May 1988 when working on his dissertation. They gave me direction for lifelong learning. v ACKNOWLEDGEMENTS The College and University Administration program has been an important part of my work and life. Grandstaff directed this dissertation. Dr. Marvin J. Dr. Eldon Nonnamaker, my academic advisor suggested the topic of women in medicine to me, among his many other pieces of sound advice. My other committee members were equally crucial, Dr. Katherine E. White who initially suggested this doctoral program to me, and Dr. Louis Hekhuis who as Admission Coordinator saw fit to admit me to the program. Acknowledgement is made to the Office of Health and Medical Affairs formerly of the Michigan Department of Management and Budget, now in the Michigan Department of Public Health for the use of the 1986 Physician Survey Data, and special note is made of Kenneth J. Darga's endless cooperation and interest. The extensive guidance and personal effort in the management and analysis of data by Randall P. Fotiu, Ph.D., Statistician and System Analyst at Michigan State University's Computer Laboratory, is gratefully recognized. Acknowledgement and thanks is made to the American Medical Association (Ms. Gene Robak) and the American Osteopathic Association (Mr. Michael Wallis) for vi vii sharing data. I wish to credit the following publishers and authors for permission to reproduce copyrighted material: American Council on Education, American Medical Association, American Osteopathic Association, American Medical Women's Association, Anesthesiology, Annals of Internal Medicine, Canadian Medical Association, Hawaii Medical Association, Michigan Association of Governing Boards of State Universities, The New England Journal of Medicine, Praeger Publishers, Sage Publications, Social Problems, University of Chicago Press, University of Illinois Press, Yale University Press. Lasting thanks is to my spouse, Robert J. Henry, who kept the home fires burning. viii TABLE OF CONTENTS LIST......... OF T A B L E S ............................. LIST xiii OF F I G U R E S ................................. xvi IINTRODUCTION .......................................... The Context of Women inMedicine in the U S A ........... Role of Higher Education ............................. Purpose of the S t u d y ................................. Major E l e m e n t s ........................................ G e n d e r ...................................... A g e .......................................... Year of Graduation fromMedical School ... Professional Degree ......................... S p e c i a l t y .................................... Hours Worked ............................. ..................... Geographical Location Hospital Position ........................... T e a c h i n g ................................... Administration ............................. R e s e a r c h .................................... Professionally Inactive ..................... Proposed Relationships BetweenVariables ......... R a t i o n a l e .............................................. S u m m a r y .......................................... Research Questions ................................... H y p o t h e s e s ........................................ Generalizations From The Research ............... M e t h o d ................................................ Research Design ................................. P o p u l a t i o n ........................................ Data C o l l e c t i o n ................................. Office of Health and Medical Affairs . . . . American Medical Association................. American Osteopathic Association .......... Office of Health and Medical Affairs........ Confidentiality ............................. Final Data B a s e ........................................ Independent Variables........................ Dependent Variables.......................... Confounding Variables........................ M e a s u r e m e n t ..................................... Data A n a l y s i s ................................... Sources of E r r o r ................................. 1 1 5 8 11 11 11 11 12 16 16 17 19 21 22 22 23 25 28 28 29 29 34 36 36 37 38 39 39 40 41 41 41 42 43 43 43 43 44 ix Measurement Error ........................... Sampling Error ............................. S u m m a r y ................................................ 44 44 44 II REVIEW OF THE LITERATURE ON WOMEN IN USA MEDICINE The Medical Education Of USA Women Physicians ........ International Comparisons ....................... Increase in Enrollment of Women in USA Medical Schools Shortage of Physicians ........................... Encouragement of Women ........................... Removal of Formal Barriers ....................... Lowering of Informal Barriers ................... Sex D i s c r i m i n a t i o n ......................... Sexism in Medical Training ................. High C o s t ................................... Social and Learning Environment in Medical T r a i n i n g ............................... Performance in Medical Education .......... Lack of Role Models, Mentors and Sponsors . . Summary .............................................. Medical Specialties Of Women ......................... Access to Graduate Medical Education ................. Women Physicians in T r a i n i n g ..................... Mentor Relationships ............................. Specialty Distribution For Women Physicians .......... Interpretations of Women's Specialty Distributions . . Reasons Expressed by Women ....................... Stereotyped Roles ...................................... Patient Preferences ............................. Present Trends in Specialty Choices ............ Comments on Specialty Selection By Women ........ Professional Activities Of Women Physicians .......... Patient C a r e .......................................... Underemployment ................................. Private Practice ................................. salaried Positions ............................... Hospital-Based Practice ......................... Professional Productivity of Physicians ............ Hours Per Week and Weeks Per Year W o r k e d ........ Why Women Physicians Work Differently From Men . . Part-Time Practice ............................... I na c t i v i t y ........................................ Length of C a r e e r ................................. Convergence in Hours Worked by Women and Men . . . T e a c h i n g .............................................. Status of Women as Medical Faculty ............... Administration ........................................ R e s e a r c h .............................................. Geographic Distribution ............................... Definition of Rural and Metropolitan ............ Reasons for Distribution Patterns of Women Physicians . Age, Year Of Medical School G r a d u a t i o n ............... United States Versus International Medical Education . 46 46 49 49 49 50 50 51 51 53 53 54 55 56 57 57 58 58 60 60 61 61 63 66 67 67 68 70 70 70 71 72 72 72 73 75 75 76 76 77 78 79 80 80 81 81 82 83 X Contemporary Issues .................................... Narrowing Of The Gender Differences ................... Men C h a n g i n g ...................................... Women C h a n g i n g .................................... Specific Efforts to Create Change ..................... S u m m a r y ................................................ 85 85 86 87 88 88 III RESEARCH DESIGN AND METHODOLOGY ................. 91 92 Review of Research Methods of Prior Studies .......... Research Design ........................................ 94 P o p u l a t i o n ............................................ 96 Data Collection and Instruments ....................... 97 Data S o u r c e s ...................................... 97 Data A c q u i s i t i o n .................................. 99 Office of Health and Medical Affairs .. . . 99 American Medical Association ............... 99 American Osteopathic Association .......... 101 OHMA Survey D a t a .............................. 102 V a r i a b l e s ...........................................103 Independent Variables ....................... 105 Dependent Variables ......................... 106 Confounding Variables ....................... 106 Sources of E r r o r .................................... 106 Measurement Error ........................... 106 Sampling Error ............................. 107 M e a s u r e m e n t ........................................ 107 Nominal Measurement ......................... 107 Interval Measurement ....................... 108 Data A n a l y s i s .............................................108 H y p o t h e s e s ............................................... 109 S u m m a r y ................................................... 126 IV RESULTS OF THE A N A L Y S I S ............................ 128 Acquisition of the D a t a .................................. 128 Computer Tape Specification ..................... 129 office of Health and Medical Affairs Initial Data 129 American Medical Association Data ............... 130 Matching L&R and AMA Data.......................131 American Osteopathic Association Data .......... 132 Matching L&R and AOA Data...................... 133 Verification ...................................... 133 OHMA Survey D a t a .................................... 134 Final Data B a s e .................................... 134 The Study G r o u p ...........................................136 Distribution of the Variables of the Study Group . 137 Professional Degree and Gender ............. 137 Nation of T r a i n i n g ............................ 139 S p e c i a l t y ...................................... 140 Practice T y p e .................................. 140 Professional Activity Status ............... 142 Graduate Medical Education ................. 146 Testing of the H y p o t h e s e s ................................ 159 Subsequent Analyses .................................... 202 xi Practice T y p e ...................................... 207 Hospital Only Practice ..................... 207 Office Only P r a c t i c e ......................... 207 Mixed P r a c t i c e ................................208 Practice Type and Patient Care H o u r s ............... 208 Nation of Training and Specialty ................. 209 Length of C a r e e r .................................... 210 V : SUMMARY AND CONCLUSIONS..............................214 Summary of The S t u d y ................................214 Purpose of the S t u d y ................................215 Comments on Secondary Analysis ................... 217 Discussion of The F i n d i n g s ............................. 219 Tentative Interpretations ....................... 219 S u m m a r y ............................................ 219 Patient C a r e ........................................ 220 Hospital C a r e .................................. 220 Nonhospital or Office C a r e ................... 221 Forty or More Patient Care Hours Per Week . . 221 Less Than Forty Patient Care Hours Per Week . 222 Primary Care Physicians ..................... 223 Nonprimary Physicians ....................... 223 S p ecialties.................................... 224 Trends Over T i m e .................................... 224 Specialty and Geographical Location ........ 225 Specialty and Practice T y p e ................... 225 Specialty and Hospital Practice . . . . 225 Specialty and Office Practice ........ 226 Uncontrolled Variables ........................... 226 Relationship Between Findings and Other Resea r c h e r s .................................... 227 Increase in Enrollment of Women ............ 227 Lack of Role Models, Mentors and Sponsors . . 227 Women in S p e c i a l t i e s ......................... 228 Present Trends ................................... 230 Practice T y p e .................................. 230 Patient Hours Worked Per W e e k ................. 230 Part Time P r a c t i c e ....................... 232 I n a c t i v ity................................232 Length of C a r e e r ......................... 233 Convergence in Hours ................... 234 Geographic Distribution ..................... 234 Age, Year of Medical School Graduation . . . 235 USA and International M D s ..................... 235 Theoretical Implications ......................... 236 S pecialties.................................... 236 H o u r s .......................................... 237 Part Time P h y s i c i a n s ......................... 238 Practice T y p e .................................. 238 MD and D O ...................................... 239 Geographical Location ....................... 240 Recognized Limitations of This Study ............ 241 Conclusions Of The S t u d y ................................242 xii M y t h s .......................................... 243 Implications For Practice ............................. 244 245 Implications For Medical Education ................... Recommendations For Further Research ................. 246 Improvement Of This S t u d y ..................... 246 Methodological Issues ....................... 247 Beyond Research ........................................ 249 APPENDICES Appendix A T a b l e s .............................................. 251 Appendix B Michigan Department of Commerce Local Development Services Rural Development Office Classification of all Counties in the State of Michigan, 1990.......... 298 Appendix C Office of Health and Medical Affairs 1986 Physician Census Survey Questionnaire ..................... 305 Appendix D Office of Health and Medical Affairs Technical Notes on Physician Census Results, 3/20/90 306 Appendix E American Medical Association Record of Physicians' Professional Activities (PPA) ................... 311 Appendix F American Osteopathic Association 1989 Census of the Osteopathic Medical Profession ................... 315 Appendix G American Medical Association List of Self-Designated Practice Specialty Codes (85 Categories) ........ 318 Appendix H American Medical Association List ofPractice Specialty Codes (39 Categories) ..............................320 Appendix I American Medical Association List of Specialty Board Abbreviations (4 categories) ..................... 321 Appendix J Final Data B a s e .................................... 322 Appendix K Physician Profiles ............................... 327 Appendix L Characteristics of Specialties ................... 335 BIBLIOGRAPHY ............................................ 338 General References ............................... 372 xiii LIST OF TABLES Table 1 Women MDs in the USA, Selected Years, 1850-1986 . 4 Table 2 MDs in Hospital Based Patient Care in 1986 . . . . 20 Table 3 MDs in Nonpatient Activities, 1986 ............... 21 Table 4 Changing Practice of Women Physicians, 1969-1986 . 68 Table 5 Outcome of the Match Between L&R, AMA, AOA and OHMA R e c o r d s ............................................ 135 Table 6 Activity Status by Gender ........................ 143 Table 7 Activity and Hours Worked byGender .............. 154 Table 8 Primary Care as Principal Specialty by Gender . . 160 Table 9 Total Patient Hours for Specialty by Gender If Fully T r a i n e d ............................................ 168 Table 10 Activity Status of Women and Men, For Year of Birth 170 Table 11 Specialty, Degree and Nation of Training of Women P h y s i c i a n s .......................................... 194 Table 12 Nation of Training, Year of Graduation of MDs, by G e n d e r .......................................... 197 Table A-l Women In MD Medical Schools, 1894-1986 .......... 252 Table A-2 Women In DO Medical Schools, 1956-1986 .......... 254 Table A-3 MD Physicians' Activities In 1986 255 Table A-4 Age Distribution Of MDs In The USA, 1986 ........ 256 Table A-5 Specialty Of Patient Care MDs By Gender 1986 .. . 257 Table A-6 Specialty Of Women MDs In Patient Care, 1986 .. . 258 Table A-7 Specialty Of Women MDs In Practice And Training, 1986 .............................................. 260 xiv Table A-8 Specialty Of Women MDs In Training, 1986 ........ 261 Table A-9 Academic Rank Of Women (1978 And 1989) And Men ( 1 9 8 9 ) ............................................ 263 Table A-10 Gender And Age Of USA Medical School Faculty, 1988 264 Table A-ll Professional Degree, Nation Of Training By Gender 265 Table A-12 Principal Specialty By Gender ................... 266 Table A-13 Secondary Specialty By Gender ................... 268 Table A-14 Active Physician Practice By Gender............... 270 Table A-15 Gender Differences In Training Status Of Active P h y s i c i a n s .......................................... 271 Table A-16 Specialty Of Active Physicians In PracticeAnd Training By G e n d e r .......................................... 272 Table A-17 Population Based Location By Degree,Nation ofTraining And Gender Of Active Physicians ................. 273 Table A-18 Metropolitan Contiguity By Degree And Gender Of Active, Fully Trained Physicians ......................... 275 Table A-19 Metropolitan or Rural Location By Degree, Gender And Graduation Period Of Active Physicians .......... 276 Table A-20 Patient Care Hours For Full Time, Fully Trained Physicians By Gender ............................. 277 Table A-21 Years Of Birth And Graduation, ByG e n d e r ............ 279 Table A-22 Historical Specialties By Gender ................. 280 Table A-23 Specialty If Born After 1945, Or Graduated After 1970 .............................................. 281 Table A-24 Active Physicians' Patient Hours BySpecialty, Year Of Birth And G r a d u a t i o n ................................282 Table A-2 5 Physician Inactivity By Year Of Birth And Gender . 283 Table A-26 Physician Inactivity By Year Of Graduation . . . . 284 Table A-27 Location Of Active and Fully Trained Physicians By Specialty And G e n d e r ................................285 Table A-28 Location Of Fully Trained Women By Age, Year Of G r a d u a t i o n .......................................... 286 XV Table A-29 Specialty, Age, Years Since Graduation By Gender . 287 Table A-30 Percentage Of Graduates Of Each Decade In A Rural Location By Specialty And Gender Of Fully Trained Patient Care Physicians ......................... 288 Table A-31 Percentage Of All Active MDs And DOs, BySpecialty And Gender, Showing Those In A RuralLocation . . . . 289 Table A-32 Active Physicians By Specialty And Location, By G e n d e r ....................................... 290 Table A-33 Hours By Specialty And Gender Of Full Time, Fully 291 Trained Physicians, By Gender ................... Table A-34 Practice Type and Characteristics of Active Phy s i c i a n s ................................... 292 Table A-35 Hours For Practice Types by Specialty And Gender For Full Time, Fully Trained Physicians ........... 293 Table A-36 Hours By Specialty And Gender For Full Time, Fully Trained Mixed Practice Physicians .............. 294 Table A-37 Distribution Of Specialty By Gender And Nation of Training Of Active M D s ...................... 295 Table A-38 Activity After Graduation by Gender ............ 296 Table A-39 Hours Worked After Graduation By Gender ........ 297 Table B-l Classification of Michigan Counties as Metropolitan or R u r a l .............................................. 300 Table B-2 Adjacency of Michigan Rural Counties to Metropolitan Counties .......................................... 303 Table J-l Codebook for Individual Physician File .......... 323 Table J-2 Codebook for Physician Records By Location . . . . 325 xvi LIST OF FIGURES Figure 1. Women MDs in the USA, 1860 - 1986................. 3 Figure 2. Ages Of All Women And Men MDs In The USA, 1986. . 13 Figure 3. Women MDs and DOs as Percent of Graduating Class, 1914 - 1986 15 Figure 4. Nonfederal MD Practice Sites, 1986 ............... 18 Figure 5. Specialties of Women MDs, Practising and In Training, 1986 .............................................. 24 Figure 6. Specialties Of Women And Men MDs, 1986......... 27 Figure 7. Data A c q u i s i t i o n ............................... 38 Figure 8. Characteristics of the Study Group ............... 138 Figure 9. Part or Full Time Patient Care by Physician Gender 144 Figure 10. Activities and Location of the Study Group. . . . 145 Figure 11. Hours Per Week Of The Study G r o u p ............. 152 Figure 12. Part and Full Time Practice Hours, by Gender . . . 155 Figure 13. Year Of Graduation For The Study Group, By Gender 158 Figure 14. Profile of the Woman Physician ................. 161 Figure 15. Profile of Full Time Recent Graduates ........... 172 Figure 16. Specialty and Location by Nation of Training of Women Physicians ........................................ 195 Figure 17. Profile of Active Full Time M D s ................... 199 Figure 18. Geographical Location by Decade of Graduation . . 203 Figure 19. Length of Career By G e n d e r ..........................212 Figure 20. Hours Throughout Career By Gender ............... 213 XVI1 Figure 21. Profile of Active D O s ............................. 334 I INTRODUCTION The Context of Women in Medicine in the USA Over the last two decades women entered medical schools in the United States of America at rapidly increasing rates to realize an 850% increase (Appendix A contains Tables A-l and A-2, and refer to Figure 1 below). Women were 15.2% of physicians in the USA in 1986 (Table 1). Twenty six years earlier in 1960 women were only 6% of USA physicians, one of the lowest proportions in the world two decades ago. Due to the acknowledged single digit quota for accepting women into USA medical schools, in 1970 the Women's Equity Action League filed sex-discrimination suits against all USA medical schools (Levin, 1980). Later in the same year the Association of American Medical Colleges' (AAMC) 1970 resolution on equal opportunity lead to annually increasing proportions of women accepted into medical school (Levin, 1980; J. Wallace, 1969). Greater access to medical education has led to greater equity in the medical careers of women. Other pressures on universities with medical schools ensured maintenance and growth of enrollment of women in medical education, and their participation as faculty and leaders. Women had not been the only class of applicants whose numbers were apparently limited in medical schools, such as jews, catholics and African Americans (K. E. White, personal communication, 23 July, 1991), but women are the focus of this study. It is important to study the professional activities of the woman physician, USA or internationally trained, as she is contributing an increasing portion of medical services to the USA (Bluestone, 1990). By 2010, 30% of all physicians in the USA will be women (Kletke, Marder & Silberger, 1990). Their 1986 impact on the USA physician supply will double, and increase their visibility as a large and growing pool of talent within the medical profession. 2 3 5 .2% 8 0 ,0 0 0 # W o m e n MDs % a l l MDs 6 0 ,0 0 0 7 3% 4 0 ,0 0 0 6% 5% 4 4% 20,000 .4 . 4 % 4 6.%' 28% 0.86% 1 8 6 0 1870 1880 18 00 1 90 0 1010 1020 1030 1 0 4 0 19 5 0 10 6 0 1070 1080 1086 Figure 1 . Women MDs in the USA, 1860 - 1986. 4 The rapidly increasing enrollment of women in medical school (40% of the 1990 entering class) will take a generation to make an impact on the gender composition of the physician workforce in the USA (Figure 1 and Table 1), due to the existing large proportion of men physicians. Table 1 Women MDs in the USA. Selected Years. 1850-1986 Women Physicians Year Total Physicians Percent Women 1850 1860 1870 1880 1890 200 544 2,432 4,557 40,755 55,055 64,414 85,671 104,805 0.4 0.8 2.8 4.4 1900 1910 1920 1930 1940 7,387 9,015 7,219 6,825 7,708 132,002 151,132 144,977 153,803 165,989 5.6 6.0 5.0 4.4 4.6 1950 1960 1970 1980 1986 11,823 15,672 24,088 54,284 86,670 191,947 260,484 334,023 467,679 569,160 6.1 6.0 7.2 11.6 15.2 Note. Data 1850- 1960 adapted from Doctors Wanted: No Women Need Apply, p. 186, by Mary Roth Walsh, New Haven: Yale University Press, 1977, Copyright Yale University Press 1977. Adapted with permission. Data 1970-1986 adapted from Physician Characteristics and Distribution in the US. 1987 Edition, pp. 72-74, Copyright 1987 American Medical Association, 1987. Adapted with permission. Role of Higher Education USA medical schools and their parent universities had limited access of women by restrictive admissions quotas and procedures. Sex discrimination and unequal treatment of women in medical education restricted their professional opportunities. Enrollment changes were forced by external pressures on medical schools and universities, such as the 1970 class action suit, the AAMC resolution and federal antidiscrimination legislation and requirements. This study will measure the changes for women physicians and for all of medicine that followed the enrollment policy changes by USA medical schools in the 1970s. The problem to be investigated is the manner and the extent to which women physicians have participated differently in professional activities from men physicians (Table A-3; Bauder-Nishita, 1980; Cartwright, 1977b; Dykman & Stalnaker, 1957; Maheux, Dufort, Lambert & Berthiaume, 1988; Morantz-Sanchez, 1985; Rinke, 1981a, 1981b). They have been more likely to train in the primary care specialties than men (AMA, 1987; Custer & Dimon, 1987; Spieler, 1977; Wunderman, 1980). Women have previously worked fewer average hours per week than men (Bobula, 1980b; Cartwright, 1977b; Heins, Smock, Jacobs & Stein, 1976; Heins, Smock, L. Martindale, Jacobs & Stein, 1977; Lisoskie, 1986; Lopate, 1968; Materra, 1980; Phelps, 1968; Powers, Parmelle & Wiesenfelder, 1969; Silberger, Marder & Willke, 1987). Women physicians have more previously chosen a position in teaching (AAMC 1988a, 1988b, 1989; Bowman & D. I. Allen, 1985) or in a hospital (AMA 1987; Spieler, 1977) more than men. Women have been in other salaried positions more than have men (Bobula, 1980a, 1980b; E. D. Cohen & Korper, 1976a, 1976b; Custer & Dimon, 1987; Heins, Smock & L. Martindale, 1978; S. C. Martin, R. M. Arnold & Parker, 1988; Silberger, et al, 1987). Women physicians have been less likely than men to be in administration (Dickstein & Nadelson, 1986; Dickstein & Stephenson, 1987; W. Gross & Crovitz, 1975; Maheux, Dufort & Beland, 1988; S. C. Martin, et al, 1988; Rinke, 1981a). Women did less research (B. A. Levey, Gentile, Jolly, Beaty & G. S. Levev, 1990; Nadelson. 1983; Ramey, 1984). Women physicians have worked proportionately more in metropolitan areas when men have a more widespread geographical distribution (Bowman & M. L. Gross, 1986; Kutner & Brogan, 1980; Maheux, Dufort & Beland, 1988; S. C. Martin, et al, 1988; Vogel, 1985). Women have been more likely to be inactive professionally than men physicians (Ducker, 1986; Dykman & Stalnaker, 1957). Internationally trained women physicians practicing in the USA have been in lower status specialties and 7 professional activities than either internationally trained men physicians or than USA trained women physicians (Bowman & M. L. Gross, 1986; A. Goldblatt & P. B. Goldblatt, 1976; Rosenthal & Eaton, 1982; Worobey & Mick, 1987). There has been no conclusive evidence that USA racial or ethnic minority women physicians were in different specialties or professional activities from USA majority women (E. K. Adams & Bazzoli, 1986; Babbott, Baldwin, Killian, & S. 0. Weaver, 1989; Morgan, 1971; Swanson, Randlett, Haynes & Killian, 1989; Titus-Dillon & D.G. Johnson, 1989). However in the past two decades each of the above patterns of professional activity have been changing, for both men and women (Bergquist, Duchac, Schalin, Zastrow, Barr & Borowiecki, 1985; Weisman, Levine, Steinwachs & Chase, 1980). Women began entering a broader range of specialties in the 1970s and 1980s (E. K. Adams & Bazzoli, 1986; Geyman, 1980; Weisman, et al. 1980). Women physicians increased their work hours at the same time that men physicians reduced their hours (Curry, 1983; Freiman & Marder, 1984; Geyman, 1980; Weisman, et al, 1980; P. B. Williams, 1978). Women still work more frequently in academic medicine than do men (AAMC, 1988a, 1988b, 1989; Bickel, 1988; Hojat, Gonnella, Moses, & Veloski, 1987; S. C. Martin, et al, 1988). Women physicians were working more in administration (Rothstein in Baxter Foundation, 1989; Sullivan, 1974), and 8 now participate in research at the same rate as do men (AMA, 1987). Women students and residents indicated the same degree of preference for rural practice locations as did men (E. K. Adams & Bazzoli, 1986; R. H. Rosen, Heins & L. J. Martindale, 1981). However practising women physicians were still more likely to be in metropolitan areas than men (Bowman & M. L. Gross, 1985; D'Elia & I. Johnson, 1980; Kutner & Brogan, 1980; Ogle, R. C. Henry, Durda & Zivick, 1986). In the 1980s, for the first time in the USA, proportionately more women physicians were active professionally than were men (AMA, 1987; USDHHS, 1987). Internationally trained women physicians continued to participate in a narrow array of professional activities (A. Goldblatt & P. B. Goldblatt, 1976; Worobey & Mick, 1987). Purpose of the Study The purpose of this study is to describe the professional activities of all women and men physicians in Michigan. Both doctors of medicine, MDs, and doctors of osteopathy, DOs, were included. The study will compare gender differences in specialty practiced, hours worked, hospital, office, administration and research activities, graduate medical education status, geographical distribution, and professional inactivity rates. Nation of training and professional degree differences will be examined within gender differences. Younger physicians have graduated, trained and been active during the period of increasing opportunities for women in medical education after 1971 (E. K. Adams & Bazzoli, 1986; Bergquist, et al, 1985; Clavan & Robak, 1978; Geyman, 1980; Lisoskie, 1986; Lorber, 1984; McGrath & Zimet, 1977a, 1977b; Weisman, et al, 1980; P. B. Williams, 1978). Therefore the study will examine whether professional activities differ between younger (born after 1945) and older physicians, and between recent (after 1970) and early graduates. The increasing supply of USA women physicians during the past two decades reflected changes in women's roles in USA society, together with the impact of federal antidiscrimination legislation and voluntary efforts by medical schools. These latter forces interacted to shape the enrollment policy for women in medical schools. This study will measure the changes in physician activities in Michiaan since more women enrolled in USA medical schools after 1970. There will be confirmation value in the results for medical education on changes in enrollment policy and the subsequent impact on USA society. There will be predictive value in the intended and unintended outcomes of the 1970s1 enrollment policy changes for women. One major contribution of this study will be to determine for the first time the professional activities of all men and women physicians (MDs and DOs) in Michigan 10 (population of 9.2 million in 1986). Another major contribution will be inclusion of both MDs and DOs as the first such research on DO women physicians' professional activities. The opportunity for comparison between women in these two medical professions will be a further first study of women physicians in the USA. The research on the distribution of specialties will assist in predicting the future supply of each specialty group, based on the pattern of specialty activity by women physicians, and the trends of all physicians over time. The research on the geographical distribution of women and men physicians will provide a basis for planning by local and state bodies for their physician needs. The profile of the metropolitan and the rural physician in each specialty will aid recruitment of physicians. The research on the professional activities of women and men physicians will also facilitate physician recruitment by providing a profile of the physician who is in patient care, administration, research or teaching. The trends over time of specialty selection by women and men physicians will be a factual basis for planning for future recruitment of specific specialists. Recruiters will need to take into account the greater availability and specialty selection of women physicians and the reduction in supply of men physicians, and anticipate any differences in between women and men physicians in the same specialty or function. 11 The research on age and year of medical school graduation will indicate how each variable can predict professional activities of women and men physicians. Formulation of the Major Elements Gender The medical profession has been predominately male (Walsh, 1977a). The small proportion of women physicians has had professional activities different from men's. With the increasing proportion of women in medicine (AMA, 1987; Figure 1; Table 1), the professional activities of women physicians will increasingly influence the medical services offered in the USA. Age Figure 2 and Table A-4 show the age distribution of physicians in 1986, reflecting the recent growth of enrollment of women. The differences in activities between women and men physicians have been reducing in the last two decades. Analysis of the trends over time will compare the professional activities of younger (born after 1945, based on the age of entering medical students from when enrollment opened up to women in the early 1970s) and older (born before 1946) physicians. Year of Graduation from Medical School Due to the recency of women's medical school graduation (Tables A-l and A-2, and Figure 1) comparison between the professional activities of recent graduates (graduated after 12 1970, the year of the AAMC Assembly's resolution on equal opportunity) and those graduating before 1971 will detect trends. This may be a more sensitive index for change in professional activities than age alone, because older students have been more common in the past two decades (S. R. Kaplan, 1981, 1982). Retroactive studies have divided observations by graduation dates (L. M. Arnold et al, 1981; Beil et al, 1980; Bergquist et al, 1985; Cartwright, 1977a, 1977b; P. B. Williams, 1975). Professional Degree MDs are doctors of allopathic medicine. In the USA DOs are doctors of osteopathy, a second medical profession solely trained, recognized and licensed as physicians in the USA (American Osteopathic Association, 1989). 13 III Age Women L IMen ( 35 l........ 35-44 45-54 55-64 65 + Figure 2 . Ages Of All Women And Men MDs In The USA, 1986. 14 In 1989 in the USA, 95.8% of all active physicians were MDs and 4.2% were DOs (Council on Graduate Medical Education [COGME], 1989). In 1987 in Michigan, 17% of all licensed physicians were DOs and 83% were MDs (Michigan Department of Licensing and Regulation, 1987), with a greater impact than , in other parts of the USA. DOs were more likely to be in primary care (COGME, 1989) and to practice in rural locations more frequently than MDs (Denslow, Hosokawa, Campbell, C. R. Roberts, & Samuels, 1984). DO medical schools admitted women less frequently than did MD schools (Figure 3). In 1986 28% of total DO students were women (Table A-2) compared with 34% at MD medical schools (Table A-l). Thus the professional activities of women DOs may not demonstrate the same absolute change over time as of women MDs, but may illustrate similar trends. North American or International Medical School Graduates of international medical schools (including Canadian graduates) constituted 27% of all MD physicians in Michigan in 1985 (USDHHS, 1987). Historically women had greater access to medical education in most other nations. Therefore in the USA there were proportionately more women from international schools within the subtotal of internationally trained physicians, than there were women from USA medical schools within the subtotal of USA trained physicians (Bowman & M. L. Gross, 1986; Pennell & Renshaw, 1973) . 15 3 2 .3 % % -* & - a ll MD Gradu ate s % a ll DO Graduates 25% 68 . 4 .9 % 4 . 1% 2, 1914 2% 1924 1934 ' 1 944 1954 2.6 1964 1974 1986 Figure 3. Women MDs and DOs as Percent of Graduating Class, 1914 - 1986. 16 In the 1980s there was a diminishing number of international medical school graduates entering the graduate medical education in Michigan (Michigan Council on Graduate Medical Education [MCGME], 1987). Hence the professional activities of internationally trained women have influenced the aggregate description of all women physicians' professional activities in the USA. Specialty The medical specialty entered after completion of medical school influences many aspects of the subsequent professional activities of a physician (Davidson, 1979; McGrath & Zimet, 1977a). The specialty distribution of women has been narrow, predominately primary care specialties, namely family practice, general practice, internal medicine, obstetrics and gynecology, and pediatrics. Hours Worked Women physicians have historically worked fewer hours than men physicians (Dykman & Stalnaker, 1957). Bobula's (1980b) analysis of the hours worked by physicians in 1978 found lower hours per week and fewer weeks per year for women as did Woodward, M. L. Cohen and Ferrier (1990). However, men were shortening their hours concurrently (Freiman & Marder, 1984; M. J. Lanska et al, 1984; W. B. Schwartz et al, 1988). Average patient care hours for each specialty will disclose the proportional patient care contribution by women and men physicians, and MDs and DOs. 17 Geographical Location The population based model used by the Michigan Department of Commerce defines a rural county as one with no city greater than 50,000 persons (Appendix B, Table B-l). An alternative definition sorts counties by both population and contiguity to metropolitan areas (Table B-2). Women have been less likely to practice in rural areas than men physicians (Bowman & M. L. Gross, 1986; Maheux, Dufort & Beland, 1988; S. C. Martin, et al, 1988). Internationally trained MDs practice twice as often in metropolitan areas, than do USA trained MDs (Meija, Pizurki, & Royston, 1980). DOs on the other hand practice twice as often in rural areas than do USA trained MDs (Denslow, et al, 1984). A description of those women and men physicians in rural and metropolitan locations, further analyzed within each specialty and profession, will provide a basis for physician recruitment to rural areas. Patient Care Men physicians reported patient care more often (Table A-3) and with their longer work week one expects gender differences in the quantity of hours spent in patient care. 18 Hospital 22 % • Hospital Other k 36% Other 6% / 18%/ Office O ffic e 45% 72% MEN WOMEN Figure 4 . Nonfederal MD Practice Sites, 1986. 19 Hospital Position Women physicians were in full time hospital based care more than their representation as physicians (Tables 2 and A-3, Figure 4). The same direction of difference existed in 1970 (Spieler, 1977) when 35% of women physicians were in hospital based practice, but only 7.2% of all physicians were women. Gender differences in hospital only practice are expected. Women were proportionately less in fellowships in 1986 (22.2%) when they were 27.4% of all residents, when women have been over 30% of medical school graduates since the early 1980s (Table 2). It is still less likely that a woman resident will proceed to subspecialty fellowship training (Table 2), 12 to 1 for women, and 9 to 1 for men. Women have practiced less often than men in specialties that combine outpatient and inpatient care, such as the surgical specialties. Gender differences are expected in hours spent in inpatient care in conjunction with their outpatient care activities within the same specialty. 20 Table 2 MDs in Hospital Based Patient Care in 1986 Activity Women Physician % Women Men Physicians % Men Graduate Medical Education Residents 22,072 27.4 58,404 72.6 Fellows 22.2 6,433 77.8 Ratio 1,831 12:1 9:1 Staff Staff 8,809 19.0 Total 32,712 24.2 Note. 37,599 81.0 75.8 Data adapted from Physician Characteristics and Distribution in the US. 1987 Edition. American Medical Association. Copyright 1987 Adapted with permission. Table 3 illustrates physicians' activities when they report their professional activity as other than in patient care. 21 Table 3 MDs in Nonpatient Activities. 1986 Women % Total Activity % Activity Men % Total % Activity Administration 1.8 11.2 2.6 88.8 Teaching 1.6 17.7 1.3 82.3 Research 3.4 16.2 3.1 83.8 Other Activity 0.6 16.0 0.6 84.0 Inactive 7.4 13.7 8.4 86.3 Not Classified 3.2 20.1 2.3 89.9 15.2 National Total Note. 84.8 Data adapted from Physician Characteristics and Distribution in the US. 1987 Edition, p. 45, Copyright 1987 by American Medical Association. Adapted with permission. Teaching Women are an increasing percentage of new faculty (AAMC, 1988b). AAMC reported (1988a, 1989) that 15% of all USA women medical school graduates will at some time have a faculty appointment whereas only 10% of men will take an appointment. Bowman and D. I. Allen's (1985) considered that "this is consistent with their general trend toward choosing institutional settings for practice more frequently 22 than men,...(and)...it may also reflect societal permission to be a teacher" (p. 13). The rate of reporting teaching as a professional activity, and the number of hours and proportion of total professional activity reported in teaching will quantify any gender differences. Administration Women were underrepresented in administrative posts (Maheux, Dufort & Beland 1988; S. C. Martin, et al, 1988; Rinke, 1981a). Dickstein and Nadelson (1986) and Dickstein and Stephenson (1987) noted that women were also underused if employed in medical administration, as had Sullivan (1974) and Loutsch, Yandow, Semiltz and Berger (1982). The rate of reporting administration as a professional activity, and the number of hours and proportion of total professional activity reported in administration will document women and men physicians' contributions. Research Women have historically been less involved in research (B. A. Levey, et al, 1990), although (Table 3) recently they reported research activity in proportion to their numbers. They have less research funding (Nadelson, 1983), are less likely to be principal investigators (AAMC, 1991), have less research space and lower research productivity as well as being younger and lower in rank (Ramey, 1984) than men. Data on physicians doing research as a professional activity, and the hours and proportion of total professional activity they spend in research, will quantify any gender 23 differences in the quantity of research activity, within and between specialties. Professionally Inactive Some women physicians have had periods of professional inactivity, mostly temporary for child bearing (Ducker, 1986; Ferrier & Woodward, 1982; Geyman, 1980; Heins, Smock, Martindale, Jacobs & Stein, 1977; Lerner, 1981). The 1980s was the first decade in which proportionately more men physicians were inactive, than were women (Table 3). This may be partly attributable to the lower median age of women physicians (Table A-4). The rate that physicians report that they are professionally inactive will describe any differences in inactivity between men and women in Michigan in 1986. 24 U n s p e c i i l e d 8% O t h e r 19% O t h e r 20% S u r g i c a l 8% S u r g i c a l 5% P s y c h i a t r y il% P s y c h i a t r y 10% M e d i c a l 5% M e d i c a l 3% I n t e r n a l 24% ! n t e r n a l 18% O b / G y n 7% w u / G y 29 P e d i a t r i c s 17% 1 0 % P e d i a t r i c s 14% Fa m i l y / G e n 11% Women MDs In P r a c u c e n=59.5I0 n F ar m l y / G e n .0% W o m e n MDs in T r a m me P.=20 746 Figure 5. Specialties of Women MDs, Practising and In Training, 1986. 25 Proposed Relationships Between Variables Walsh (1977a) wrote that sex discrimination had prevented women entering medicine and restricted their advancement. One relationship to be explored is the interaction between gender and specialty choice and trends over time, as women have entered the primary care specialties and psychiatry proportionately more than have men (Tables A-5, A-6). The data may replicate Crovitz' (1980) findings that younger women physicians qualified in a broader range of specialties than older women (Figure 5, Table A-7). Using year of graduation from medical school as another independent variable may separate the effect of recency of medical education and the changes in practice differently from using age as the independent variable. Hours in patient care typically vary according to the medical specialty practiced (Phelps, 1968; Wheeler, Candib & M. Martin, 1990). Women may work fewer hours overall and less within each specialty, compared with men physicians. Working in a metropolitan area has been proportionately more likely for women than for men physicians and for specialists. Therefore the trend of women physicians to enter a broader range of specialties may sustain their pattern of metropolitan practice. Selected observations and descriptions will be made for particular events and attributes. Medical specialty has a major effect on a physician's subsequent professional 26 activities (McGrath and Zimet, 1977b). Some specialties have a tendency to a longer work week, such as surgical specialties, and others may be more readily practiced in rural areas, or in teaching or research. Women physicians's ultimate specialties differ from men's (Figure 6, Table A-5) . There will be predictable differences between women and men based on their specialties in their hours worked, their rate of entrance to teaching and research careers, administrative positions, and geographical location. Predictions include that women international graduate physicians will be more frequently in primary care specialties and in teaching and hospital positions, than USA trained women physicians. It is predicted that women DOs will practice more frequently in primary care specialties and rural sites than USA trained women MDs. 27 ! n c c ! i ve 7% I n a c i i ve 8% U n s p e c i t i e d 6% U n s p e c i l i e d 4% O t h e r 18% O t h e r 19% S u r g i c a l 5% M e d i c a l 3% S u r g i c a l 19% P s y c h i a i r y 9% O b / G y n 7% M e d i c a l 7% ni er nal P s y c h i a t r y 6% O b / G y n 5% 18% I n t e r n a l 16% P e d i a t r i c s 15% P e d i a t r i c s 5% F a m i l y / G e n 12% arr.t i v / G e n 10% Women Men Figure 6 . Specialties Of Women And Men MDs, 1986. 28 Rationale By 1986, 42% of women physicians had been born after 1942 (Figure 2, Table A-4). Younger women have different medical careers from older women physicians who were earlier medical school graduates. Women physicians have entered a narrower range of specialties than men, (Figure 6, Table A- 5) mostly in the primary care specialties (Table A-6). For example, when women were 15% of all physicians in 1986, they were 22.3% of all psychiatrists in 1986 (Table A-6), and 3.9% of all surgical specialists. Women in training in 1986 were in more specialties than those currently in practice in 1986 (Figure 5, Tables A-7 and A-8), yet still predominately in primary care. Summary Women physicians have worked fewer hours, entered teaching more often, and worked proportionately less in research than have men. Women physicians have been located in metropolitan areas proportionately more than have men physicians. Younger women physicians and those who were more recent graduates were working longer hours and entering more specialties than had older women physicians. Younger women physicians have been an increasing proportion of new faculty. Women physicians have been more common in other nations than in the USA. Therefore the women graduates of international medical schools are a greater proportion of 29 older (born before 1946) women MD physicians and earlier medical school graduates (before 1971). The women graduates of international medical schools are proportionately less among younger (born after 1945) women MD physicians and more recent (after 1970) medical school graduates. Because of the rise in the proportion of MD and DO USA trained women physicians (Tables A-l & A-2, Figure 1), they cluster in younger age groups (Table A-4). Women DOs are in primary care specialties more than USA trained women MDs. Internationally trained MD physicians were proportionately more in primary care specialties and teaching, than the USA trained MD physicians. Research Questions The research questions center on whether women physicians report differences from men physicians in hours worked, professional activities, geographical locations and t specialty choices. Do the latter observations vary with the specialty choice, professional degree, nation of training, age of physicians, or recency of graduation from medical school. Hypotheses Hypothesis l. Women physicians report specialties of family practice, internal medicine, pediatrics and psychiatry more than do men, as their principal specialty. 30 Hypothesis 2. of Women physicians report primary specialties family practice, general practice, internal medicine, obstetrics and gynecology, and pediatrics, more than do men, as their principal specialty. Hypothesis 3. Men physicians report more even distribution across specialties than do women. Hypothesis 4. Younger (born after 1945) men and women physicians have similar specialty distributions. Hypothesis 5. More recent medical school graduates (after 1970) men and women physicians have similar specialty distributions. Hypothesis 6. Women physicians report a lower mean total patient care hours worked per week than men. Hypothesis 7. Women physicians have a larger variance in the total hours worked per week than do men. Hypothesis 8. Women physicians report a lower mean total patient care hours worked per week than men, within the same specialties. Hypothesis 9. Women physicians have a larger variance in total patient care hours worked per week than do men, within the same specialties. Hypothesis 10. Younger (born after 1945) women and men physicians have the same mean total hours of patient care. Hypothesis 11. More recent medical school graduates (after 1970) women and men physicians have the same mean total hours of patient care. Hypothesis 12. Younger (born after 1945) physicians have the same mean total hoursof patient care, within the same specialties. Hypothesis 13. women and men More recent medical school graduates (after 1970) women and men physicians have the same mean total hours of patient care, within the same specialties. Hypothesis 14. Younger (born after 1945) women and men physicians have the same percentage reported as professionally inactive. Hypothesis 15. More recent medical school graduates (after 1970) women and men physicians have the same percentage reported as professionally inactive. Hypothesis 16. Older (born before 1946) women physicians have a higher percentage reported as professionally inactive than older (born before 1946) men. Hypothesis 17. Early women medical school graduates (before 1971) have a higher percentage reported as professionally inactive than early men medical school graduates. Hypothesis 18. Women physicians report patient care less often than do men. Hypothesis 19. Women physicians report inpatient hospital care as the sole professional activity more often than do men. Hypothesis 20. Women physicians report a lower mean of inpatient hospital care hours worked per week than men, when reporting both hospital and nonhospital practice. 32 Hypothesis 21. Women physicians report administration as a principal professional activity less than do men. Hypothesis 22. Women physicians report a lower mean total hours worked per week in administration than do men. Hypothesis 23. Women physicians report teaching as a professional activity more often than do men. Hypothesis 24. Women and men physicians report the same mean total hours worked per week teaching, when it is the sole professional activity reported. Hypothesis 25. Women physicians report research as a professional activity less than do men. Hypothesis 26. Women physicians report fewer total hours worked per week in research than do men. Hypothesis 27. Women and men physicians are in graduated medical education training at the same rate. Hypothesis 28. Women physicians are less likely to be training at the fellowship level than are men physicians. Hypothesis 29. Women physicians are less broadly distributed among the counties in Michigan than are men. Hypothesis 30. Women physicians are more likely to practice in metropolitan areas than are men. Hypothesis 31. Men physicians are more likely to practice in rural areas than are women. Hypothesis 32. Women primary care physicians are more likely to practice in rural areas than are women specialists. Hypothesis 33. Women primary care physicians are more likely to practice in metropolitan areas than are men primary care physicians. Hypothesis 34. Younger (born after 1945) women physician have a broader distribution between metropolitan counties, rural counties adjacent to a metropolitan county, and rural counties that are not adjacent to a metropolitan county, than older (born before 1946) women physicians. Hypothesis 35. More recent women medical school graduates (after 1970) have a broader distribution between metropolitan counties, rural counties adjacent to a metropolitan county, and rural counties that are not adjacent to a metropolitan county, than early (before 1971) women medical school graduates. Hypothesis 36. There are more USA trained women MD physicians among all USA trained MD physicians, than there are women DO physicians among all DO physicians. Hypothesis 37. DO women physicians are more likely to be in primary care than USA trained women MD physicians. Hypothesis 38. There are more internationally trained women among all women MD physicians, than internationally trained men among all men MD physicians. Hypothesis 39. Internationally trained women MD physicians are more likely to be in primary care than USA trained women MD physicians. Hypothesis 40. Women physicians are younger on the average than men physicians. 34 Hypothesis 41. Women physicians are more recent medical school graduates on the average than men physicians. Hypothesis 42. Women in nonprimary specialties are younger than women physicians in the primary specialties. Hypothesis 43. Women physicians are skewed to the younger age groups in specialty care, more than men in specialty care. Hypothesis 44. Women physicians in specialty care are skewed to the more recent medical school graduation years, more than men in specialty care. Hypothesis 45. Certain specialties are more likely to be reported by physicians whose principal activity is research, for both women and men physicians. Generalizations From The Research The generalizations from this research will be applicable to no less than 97% of physicians practicing in Michigan. There was a 98% response rate to the 1986 State of Michigan survey, and no less than 99% of these physicians had an AMA or AOA Physician Masterfile record. Formal surveying of tens of thousands of respondents is expensive and a 98% response rate would be exceedingly difficult and costly to achieve. Therefore, the secondary analysis of three valid and reliable surveys was the method selected to answer the research questions. One limiting condition of this study was that activities of two percent of physicians were unknown. 35 Either they did not respond to the survey in 1986 or follow up mailing in 1987, or they did not have a complete record in the Department of Licensing and Regulation. The study will be restricted to describing the professional activities of women and men physicians in Michigan only. Key determinants of women physicians' activities (mentoring experience, marital status, number of children, employment status, and academic rank) were not measured in any of the source surveys. Another limit of the study was that all source the data were self-reported, which may have reflected the trend of men to over-report their work hours, and women to underreport (Bauder-Nishita, 1980). The final limitation will be that there must be an AMA or AOA Physician Masterfile record that is accurate and complete for each physician to remain in this study. Loss of physicians from the study will occur by retaining only matched records. This source of error was accepted as the lesser error rather than inclusion of incomplete or inaccurate information. The information from the three source surveys combined to create the study database will be unique in the nation, and valuable for reliable answers to the questions posed. 36 Method Research Design The basis of this study will be the secondary analysis of an amalgamated database drawn from three archived 1986 physician surveys. Matching of the three sources of data to each other will be by name, suffix (if any, such as Jr.), county of license address, professional degree and specialty of the respondents. There are many advantages of secondary survey analysis (Kiecolt & Nathan, 1985), more fully discussed in chapter III. Each source has maintained a high level of reliability and validity and as each survey uses a full census count there has been no sampling error introduced in the original data sources. The research design combines elements of several survey designs. The primary design is a single measurement cross- sectional study. Temporal analysis is another component, by using trend analysis to investigate changes in physician activities, by age and by year of graduation from medical school. Comparative analysis will be a primary part of the design, through the comparison between women and men physicians; between USA educated and internationally educated MDs; between USA educated MDs and DOs; between primary and nonprimary specialists; between younger and older physicians; between early and recent medical school graduates; between office-only, hospital-only, and mixed office and hospital practice physicians; and between 37 metropolitan and rural physicians. Chapter III contains a full description of the database sources. Population The population is the full census of MDs or DOs licensed in Michigan who responded to the OHMA survey in 1986 or to 1987 follow up surveys, or for whom there was a Department of Licensing and Regulation (L&R) file, and for whom there was an entry in the AMA or AOA Physician Masterfiles (chapter III contains detailed discussion). The pool of subjects with an AMA or AOA file (99%) and a response to the OHMA survey (98%), minimally represent 97% of the physicians in Michigan in 1986. There have been few statewide studies of women physician activities reported. The only full census based on a survey and the largest (98% return rate) statewide survey published, was the 1978 California Board of Medicine Quality Assurance questionnaire, required for license renewal. Bauder-Nishita (1980) reported that approximately 44,000 nonfederal physicians responded, of which 8% were women, 91% were men, and one percent did not report gender. Californian women physicians in 1978 were more in cities, salaried, in primary care specialties, younger and worked fewer hours than men. Some studies have relied solely on reanalysis of AMA Physician Masterfile data (M. S. Stern, 1976). However, the OHMA survey included physician-supplied details of hours, practice types and geographical location of professional 38 activities. DOs. No statewide study reports inclusion of women Methods of prior studies are reviewed in chapter III. Data Collection The plan was to acquire the database in four steps after obtaining assurances of cooperation from the three sources (Figure 7). OHMA names matched matched AMA AOA data data OHMA < r\/o\/ W5iW.i. .^ J data Final Data Base Figure 7. Data Acquisition. 39 Office of Health and Medical Affairs. The Office of Health and Medical Affairs (OHMA) of the State of Michigan Department of Management and Budget conducted a survey of Michigan physicians' activities during 1986. The OHMA method was to include a brief questionnaire (Appendix C) with the state controlled substance license renewal application materials. The net return rate of 98% came after follow up surveys in 1987 and hand coding of data from L&R files (Appendix D). Instructions on the data collection card stated that the requested information was voluntary, and did not affect the renewal of the controlled substance license, and assured confidentiality and anonymity. Reliability of the OHMA study was not determined. As the State of Michigan had no other source of data, and felt that the individual physicians were the best source for individual activity data, there were no reliability studies conducted. The first step in the present study was that the chief OHMA researcher in Lansing Michigan would copy the names, addresses and professional degree of individual physicians onto a standard computer tape. American Medical Association. Second, the AMA would match these MD physicians by their names, suffixes, and county with its Physician Masterfile. For the matched physicians, AMA would add to 40 the tape their gender, year of birth, school and year of medical school graduation, and principal specialty. The sources and verification process of AMA Masterfile data assures its reliability. The AMA uses secondary sources and direct mail surveys to maintain the Physician Masterfile to describe all MD physicians in the USA (569,160 in 1986). A sample of the Physicians Professional Activities record is in Appendix E. The AMA assures confidentiality and anonymity. American Osteopathic Association. Third, the AOA would match all DOs on the OHMA standard computer tape, with its Physician Masterfile. To the matched DO records the AOA will add to the tape gender, year of birth, school and year of medical school graduation, and principal specialty. The AOA database is on all DOs in the country totalling 26,794 in 1986. The AOA uses verified public sources (mailing lists of state government licensing agencies, state osteopathic professional organizations, osteopathic specialty societies) and direct mail surveys (example in Appendix F) to maintain their files. assures confidentiality and anonymity. The AOA The sources and verification process assure the reliability of the AOA Census. 41 Office of Health and Medical Affairs. Fourth, the tape with AMA and AOA data added would return to the OHMA chief researcher and the OHMA 1986 survey responses (Appendix C) would be added for each matched physician. Confidentiality To preserve the confidentiality and anonymity of the survey responses after adding the AMA or AOA variables to the individual record, the name and address data will be deleted. License number will identify each physician, to enable analysis of parts of survey responses for each physician. After assuring this anonymity, the amalgamated data set on the standard tape will be returned to this researcher, entered as a disk file on MSU's IBM 3090 mainframe. SPSS-X (1990) software will permit statistical analysis. Final Data Base The composite data base will consist of the following classes of variables. Chapter III contains detailed discussion of classes of responses within each variable. 1 Professional degree. 2 Specialty. 3 Current professional activity status. 4 Patient care practice locations in office and hospital settings. 5 Hours worked per week in each professionalactivity. 6 Year of birth. 42 7 Gender. 8 Nation/ state of medical education, coded from school of medical education. 9 Year of graduation from medical school. Independent Variables. Gender will be treated as an independent variable with respect to specialty, hours worked, professional activity (including current training), geographical location, professional degree and nation of training. Specialty will be treated as an independent variable with respect to hours worked, professional activity, and geographical location. Nation of medical education will be treated as an independent variable with respect to specialty, and patient care in an inpatient setting as a class of professional activity. Professional degree for USA trained physicians will be treated as an independent variable with respect to principal specialty. Year of birth will be treated as an independent variable with respect to hours worked, distribution among specialties, principal specialty, and professional activity. Year of graduation from medical school will be treated as an independent variable with respect to hours worked, distribution among specialties, principal specialty, and professional activity. 43 Dependent Variables. Total hours worked per week will be treated as a dependent variable with respect to gender, year of birth, year of medical school graduation, and specialty. Professional activity will be treated as a dependent variable with respect to gender, specialty, geographical location, year of birth and year of medical school graduation. Geographical location will be treated as a dependent variable with respect to gender, specialty, nation of training, year of birth and year of medical school graduation. Confounding Variables. Two potentially confounding variables will be year of birth and year of medical school graduation. It appears that younger (born after 1945) and more recent graduates (after 1970) practice differently from older physicians and earlier graduates, for both men and women. Measurement The classes of variables were predominately nominal measures, with three classes of interval measures (year of birth, year of medical school graduation, and hours worked). Data Analysis There would be limits to the statistical analyses feasible and meaningful. The large size of the population studied, together with the absence of sampling induced errors, enhanced the power of the study. The exclusion of 44 records could occur if the matching of AMA or AOA Physician Masterfile to state records failed. The hypotheses were phrased according to the predictions of the literature. Sources of Error Measurement Error. Bias may have operated on the self-report of specialty or hours worked, but OHMA, AMA and AOA did not employ any methods to detect bias or faking. Cases were not eliminated, as this would effect the generalizability of the findings of the study. Sampling Error. The other source of error was the loss of any subjects (Spector, 1981) without a verifiable, accurate and complete AMA or AOA record for the requested variables. Loss was accepted as less of an error than inclusion of incomplete or inaccurate information on physicians. Summary The impetus for this study came from the rapid increase in the enrollment of women in USA medical schools since 1970. There is emerging data, that as well as being relatively more numerous, women physicians are practising differently from their historical professional activities. In addition, men are changing some aspects of their professional activities. The impact of the gender and age changes are studied for the full population of physicians in Michigan. 45 The secondary analysis of data from three surveys in Michigan in 1986 will provide the database for this study of the professional activities of women and men physicians in Michigan in 1986. The research design was primarily single measurement cross section, with temporal and comparative analysis. The study was generalizable to the entire population of physicians in Michigan. It would take 13 months to create the final database from state and national sources. Analysis was restricted to predominately descriptive statistics due to the primarily nominal scales in the final data set. II REVIEW OF THE LITERATURE ON WOMEN IN USA MEDICINE In the last two decades, the enrollment rate of women entering medical school in the United States of America increased over 850%. By 2010, 30% of all physicians in the USA will be women (Kletke, et al, 1990), including those graduated from an international medical school. Distinct differences in the number and proportion of women and men physicians exist due to a single digit enrollment quota for women existing at USA medical schools and their universities up to 1970. Distinct differences in professional activities also existed between women and men physicians had been observed (Dykman & Stalnaker, 1957; Maheux, Dufort, Lambert & Berthiaume, 1988; Morantz-Sanchez, 1985; Rinke, 1981a, 1981b). A review follows of the historical and theoretical reasons for the gender differences in specialty choice, professional activities and geographical location of physicians in the USA. The Medical Education Of USA Women Physicians Preparation for the medical profession is first by completion of medical school educational prerequisites. Graduate medical education commences with either or both of a one-year internship and 3 to 5 years of residency training in a specialty. Fellowship training to sub-specialize may take up to three years more (Berliner, 1987). Today, women make up the highest percentage of USA medical students, 46 47 residents, fellows and physicians in the history of women in USA medicine (Tables 1, A-l, A-2, A-4, and A-6, and Figures 1 and 3). Women had practiced medicine since ancient times (Marks & Beatty, 1972) , but there had been restrictions on women in medicine in the USA (Bowman & D. I. Allen, 1985; Campbell, 1973; Storrie, 1966). Women were excluded from formal medical education in the USA when the first school opened in 1765 (R. J. Abram, 1985). The first woman formally admitted to a USA medical school was Dr. Elizabeth Blackwell in 1847, and the first African-American women physician, Dr. Rebecca Lee, graduated in 1864 (Goodwin, 1985). By 1870 0.8% of the physicians in the USA were women (Bowman & D. I. Allen, 1985) illustrating the increase in women enrolled at both women's and coeducational medical colleges. Four factors interacted to reduce the number of women entering medicine earlier this century. Educational and professional reform raised the admission requirements for medical school. These higher pre-admission qualifications were difficult for women to obtain as they largely studied in less competitive all-women schools. Second, the higher standards subsequently effected a 58% reduction in the total number of USA medical schools from 162 in 1906 down to 69 in 1944 (Walsh, 1977a), so there were fewer medical school positions nationwide. Third, the number of schools especially for the woman student reduced. Some women's colleges merged with other 48 schools, some closed altogether due to an inability to meet the nationwide higher academic standards (Santilli, 1990), and other medical schools became coeducational. Fourth, there were few places assigned to women in the coeducational medical schools. It became increasingly rare for women to achieve admission to the dwindling places in medical school (Morantz, Pomerleau, & Fenichel, 1982). Early in the twentieth century, women physicians "watched the doors through which they had passed all but close behind them" (Moldow, 1987, Introduction, page 4). In the nineteenth century the USA had been leading the world in advancing women in medicine. By 1950 there were fewer women physicians in Boston than there had been in 1890 (Walsh, 1990). The percentage of women enrolled in USA medical schools appeared set at 5% throughout the first part of this century (Table A-l, Figures 1 and 3). The exception was a rise during each World War (Heins, 1985; Morantz, et al, 1982; Notman & Nadelson, 1973; Walsh, 1977b; Table A-l). Barondess (1981) summarized USA medical school enrollment of women (Table A-l) as doubling at MD schools in 29 years (1920-1949), but recently tripling in just 5 years (1970-1975). The rate of increase had slowed in the 1980s (D. I. Allen, 1989; A. G. Swanson, in E. Ginzberg, 1986). In thirty years (1956-1986) the osteopathic medical school enrollment of women increased from 43 to 1,857 total women students or 4300% (Table A-2, Figure 3). 49 International Comparisons In 1966, the USA had proportionately fewer women enrolled in medical school than any other developed country (Beshiri, 1969; Bowers, 1966; Hellstedt, 1967; Holstrom, 1967; Parrish, 1971; Pirami, 1967; Yamakazi, 1967) or most of the developing countries (Beshiri, 1969; Gomez, 1967). The only countries with lower medical school enrollment of women in 1966 were Spain and Myanmar (formerly Burma). Strong support for more women in medicine came from around the world (Bauknecht, 1967; Crosse, 1967; Day, 1982; Hellstedt, 1967; G. Henry, 1967; Schindler-Baumann, 1967). Increase in Enrollment of Women in USA Medical Schools Most of the impetus for change for women came from outside medicine (Barondess, 1981). Shortage of Physicians Temporary shortages of physicians and of men medical school applicants in the USA occurred during the two World Wars. Medical schools then admitted more women, either by increasing the women's portion of a coeducational medical school, or by eliminating a men-only tradition (Morantz, et al, 1982; Nadelson, 1983; Walsh, 1977b). Postgraduate internships also opened to women during World War II because hospitals were understaffed. Another shortage of physicians occurred in the 1960s due to a rapidly expanding population with no increase in the physician supply. Women were actively sought for 50 medicine (D. G. Johnson & Hutchins, 1966). Battle's 1966 editorial ("Wanted: more cheap, docile women doctors") decried the perception of women physicians as more malleable and as easier to recruit than men (Bunker & Pool, 1971). Encouragement of Women There were two interacting factors in addition to the physician shortage that enabled more women to enter medical school. Societal attitudes increasingly encouraged women into a broader range of career options (Geyman, 1980). More women had the academic preparation for medical studies after the 1958 National Defence Act promoted scientific training (D. G. Johnson & Hutchins, 1966). Further, the American Medical Women's Association [AMWA] with the Women's Bureau of the Department of Labor, sponsored a conference in 1968 to encourage women to enter medicine (Beshiri, 1969; Lopate, 1968; Norris, 1969). Explicit encouragement of women to enter medicine was a critical factor in a woman choice of medicine as a career (Nemir, 1978), and encouragement has continued (J. Harris, 1984; Romm, 1982; Walsh, 1976). Removal of Formal Barriers In 1970 the Women's Equity Action League filed sexdiscrimination suits against all USA medical schools (Levin, 1980). In 1970 the AAMC Assembly's resolution on equal opportunity lead to annually increasing proportions of women accepted into medical school (Levin, 1980; J. Wallace, 1969). The Federal Government prohibition of sex discrimination and affirmative action requirements were also 51 influential on medical schools. The USA Government requires compliance with these laws to receive federal funds, such as student financial aid, research grants, and special program funds (Boris & Honey, 1988; Dube, 1973a; Finseth, 1978; D. G. Johnson, 1983; Spieler, 1977). Lowering of Informal Barriers There had been a greater acceptance of women physicians within the medical profession (Crovitz, 1980; Lovelace, 1985; Spiro, 1975) and by patients (J. W. Adams, 1977; Baum, 1980). Sex Discrimination. Because medicine had been a men-oriented and a mencontrolled profession (AMWA, 1968; J. L. Weaver & Garrett, 1978) it was implicit that men were suitable but that women were not suitable for a medical career (Hammond, 1981; Zimet & Held, 1975). Men physicians may not fully recognize women physicians' experiences of discrimination (Crovitz, 1980; Grant, 1988; W. Gross & Crovitz, 1975; Raymond, 1982? H. M. Wallace, 1980; Walsh, 1977a). A. Lake (1967) illustrated discrimination with a succinct example of the double bind on women physicians; "An occasional woman complains because a woman doctor is a spinster ('How can she understand a mother's problems?'), another because she's married ('She's probably got half her mind on her husband and children')" (p. 403). Discrimination that favored men physicians by employers, by patients (J. W. Adams, 1977; Kasteler & Hulme, 52 1980), by society (Kehrer, 1976), and by the profession (J. W. Adams, 1977; Kehrer, 1976) tainted the professional experiences of women physicians. Pre-Admission Discrimination. Counsellors advised eligible women students against applying, or re-applying to medical school (Bowers, 1967; W. Gross & Crovitz, 1975; A. Lake, 1967; Morgan, 1971; Radius, Becker, D. R. Smith and Katasky, 1979; M. A. Thomas, 1970), as did women's families (P. A. Williams, 1971). D. G. Johnson (1983) found that many more women students did not apply to medical school after sitting the Medical College Admission Test [MCAT] or attending pre­ medical school programs, than men students (Florentine, 1987), as though directed away from application to medical school, and from medicine as a career. Discrimination in Medical School Admissions Process. Higher education permitted this sex discrimination, through the medical schools' acceptance systems based on a single-digit quota for women medical students (H. I. Kaplan, 1970; Morantz, et al, 1982; Morgan, 1971; Tables A-l & A-2). Between 1930 and 1966 the number of women applying to medical school had increased 300% but their acceptance rate actually fell, while men's application rate increased only 29%. Marquart, Franco and Carroll (1990) reported that distinctions still exist along gender lines in admission interview questions. 53 Wilson (1981) emphasized "...the important factor in increasing female population of medical students is the commitment of the admissions committee members, be they male or female, to admit and ultimately enroll female medical students" (p. 70). Sexism in Medical Training Overt sexism encountered by women medical trainees included humor that denigrated women, instructors' negative attitudes, and less attention given to women (Baum, 1980; Clark & Rieker, 1986; Sclabassi, 1984; Surawicz, Norton & Abies, 1973). 1988) Subtle and passive discrimination (Waxman, included women being overlooked, ignored or not treated as an equal, and the use of male phrases, roles and dominance in medical education texts and classes (Cox & L. A. Lewis, 1981; I. K. Smith, Lancaster & Fleming, 1984). The AMWA Ad Hoc Committee on Gender Equity (Bernstein, 1989) documented sexism in training. High Cost Chapman (1969) and S. R. Kaplan (1981, 1982) found that the high cost of medical education was a deterrent to women, especially for older applicants. Morgan (1971) noted a prevalent family preference to educate a son before a daughter, so women owed more on graduation (Titus-Dillon & D. G. Johnson, 1989) . The observed inequities in pay and promotion for women physicians, may also have deterred women from entering medicine (Chapman, 1969; I. K. Smith, et al, 1984; Tamburrino, Franco, Evans & Seidman, 1985). 54 Social and Learning Environment in Medical Training Minority Status of Women. As a minority, women often felt ignored, isolated, resented and harassed in medical education and practice (Bowers, 1968; Clark and Rieker, 1986; Roos, Gaumont & Colwill, 1977). Bonar, J. A. Watson and Koester (1982) noted that "the norms of 'feminine' behavior deviated from the behavior considered desirable in physicians" (p. 300) (M. F. Myers, 1985; Notman & Nadelson, 1973). Women were not viewed as competent physicians (S. L. Brown & Klein, 1982), or suitable as leaders (Costanza, 1989). Leserman (1981) inferred that small numbers of women may change the prevailing medical model, as "minority status often promotes liberal views" (p. 35). Sources of Stress. Authors studying women's sense of isolation in medical training, found the following sources: their minority status; the stereotyped identification of the physician as an aggressive, competitive, scientifically astute man; and finally due to the low representation of women role models and mentors in the medical training environment (Blumberg, Flaherty & Morrison, 1984; Bowman & D. I. Allen, 1985; Clark & Rieker, 1986; Coombs & Hovanessian, 1988; Hoferek & Sarnowski, 1981; C. L. Janus, S. S. Janus, Price & Adler, 1983; S. R. Kaplan, 1981, 1982; I. K. Smith, et al, 1984) Specific medical school offices for women have facilitated the success of women medical students, as 55 sources of support, and redress when necessary (AAMC, 1990a; Finseth, 1978). Potter (1983) and M. J. Gray and Ackerman (1978) described hostility, jokes, innuendo and threatening warnings against women. Baum (1980) and Hafferty (1986) attributed some of the hostility against women physicians to financial resentment by men. Furthermore, there were few women physician role models and mentors to counter-balance such negativism toward women physicians, or to encourage women medical trainees in such an environment. Performance in Medical Education Crovitz (1980) reported that since 1970, the drop-out rates were the same for women and men in medical school (C. B. Thomas, 1976). Also, the stop-out rates from medical training and practice for women physicians, primarily for childbirth, were briefer than previously believed, limited to weeks or a few months (Heins, 1985; Heins, Smock, Jacobs, & Stein, 1976; Heins, Smock, & Martindale, 1978; Heins, Smock, Martindale, Jacobs, & Stein, 1977; Heins, Smock, Martindale, Stein, & Jacobs, 1977). Harward, Lyons, Porter and Hunter (1981) found that the performance of men and women from the first year of medical school through the first year of postgraduate study and board examinations was virtually equal (Calkins, L. M. Arnold & Willoughby, 1987; Norcini, Fletcher, Quimby & J. A. Shea, 1985). There may have been cultural expectations that 56 women would perform less well than men (L. M. Arnold, Willoughby, Calkins & Jensen, 1981; Heilman & Krara, 1983). Fiorentine (1987) concluded that "women who persist and enter a medical career tend to be more committed than their male counterparts" (p. 1136) . Lack of Role Models. Mentors and Sponsors Leserman (1981) emphasized that "women are to some extent pioneers in a male profession" (p. 38) with different obstacles for women students than for men students (LloydGreen, 1967). The availability and effectiveness of active mentors and sponsors to help in training and practice was and still is an important for all physicians (Dean N. E. Gary, in L. Gross, 1990; Lorber, 1981, 1986). Women have benefitted less from the protege, mentor and sponsor system in the academic and hospital community. One reason was due to a lack of women physicians to provide these functions (Gruppen & D. R. Brown, 1981; Hsu, 1986; s. R. Kaplan, 1982; McNamara, 1985; Nadelson, 1983; Nadelson & Notman, 1972; Roeske & K. Lake, 1977; I. K. Smith, et al, 1984; Weisman, 1984). Lorber (1981) and Reiman (1989) noted the impact of age, as younger women integrated better into the medical community, to be, or to find, a mentor or sponsor. 57 Summary Women were 15.2% of all USA MD practitioners of medicine nationwide in 1986 (Table 1), and 16% of all nonfederal physicians in Michigan (AMA, 1987). These figures parallel the increased participation by USA women in the workforce in last two decades. In 1987 (Michigan Department of Labor, personal communication, September 17, 1987), women were 40.6% of the total labor force in Michigan. In 1986 in the USA, 34% of MD medical students and 32% of all MD medical school graduates in 1986 were women (Table A-l). The comparable figures for the osteopathic profession were 2*8% and 25% respectively (Table A-2, Figure 3). The DO impact was small on the national pool of women physicians, as Dos were only 4.2% of all active physicians in the USA in 1986 (COGME, 1989). However in 1985 Dos were 17% of physicians in Michigan (USDHHS, 1987) so the impact was greater in Michigan. While the total physician workforce grew 70% between 1970 and 1986, the total of women physicians increased nearly 241.2% (AMA, 1987; Tables A-l, A-2, A-4). Medical Specialties Of Women "Women physicians gravitate to certain specialties, earn less than men, work fewer hours, are less likely to be board-certified, are more likely to be found in salaried positions, and are more apt than men to interrupt their careers." (Cartwright, 1977b, p. 316). 58 Access to Graduate Medical Education The specialties selected by women physicians (McGrath & Zimet, 1977b) "determines many aspects of the life style, role, and contribution of the future physician" (p. 361). Medical specialization had increased since first accepted in 1869 (from Transactions American Medical Association. Vol 20 (1869), pp. 28-29, in G. Rosen, 1972; K. L. White, 1964). The rise in the number and percentage of women in undergraduate medical education preceded increasing gains in the post-graduate training programs (Lorber & Ecker, 1988; L. M. Lowenstein, 1971). Women were in all specialties in 1986 (Tables A-5, A-6 and A-7), although less in the specialties of surgery and its sub-specialties (AMA, 1987). Swanson, et al, (1989) noted that a slightly higher percentage of 1989 graduating women than men were not accepted into any residency. Women and minorities had pursued their graduate medical education into subspecialty fellowships at a lower rate than do majority men (Kohrman, Lyttle, Andersen, & G. S. Levey, 1989; Table A-5; A. P. Williams, Domnick-Pierre, Vayda, Stevenson, & Burke, 1990). Women Physicians in Training Women in post-graduate training in 1986 (Tables A-7 and A-8, and Figure 5) were proportionately over represented in the primary care specialties (family and general practice, internal medicine, obstetrics and gynecology, pediatrics) and psychiatry and child psychiatry. The only race difference noted (Bowman, 1986) was that African-American 59 women were over represented in family practice, and other women were over represented in neonatal and perinatal care. Women residents were under represented in all the surgical specialties (American College of Surgeons, 1988; Dalessandri, 1988; Ramos & Feiner, 1989; Wright, 1967), and those trainee surgeons were mostly younger women (AMA, 1987). Women were underrepresented in the cardiovascular and gastroenterology specialties. Women advanced to fellowship level training at a lower rate (8%) than men (10%, Table 2). Barondess (1981) emphasized that USA medicine needed to be more flexible and open to women (Bowers, 1966, 1967; H. S. Kaplan, 1962; O'Connell & Beighton, 1979; Parkhouse, 1989; J. Shapiro, 1981, 1982; M. A. Thomas, 1971). Part- time residencies were one such flexibility, first permitted in 1964 (American Medical Association, 1967; J. Wallace, 1969; Weinberg, 1969). Federal Legislation (Section 709 of the 1976 Health Professions Education Act) promoted shared primary care residencies by funding them (E. Shapiro & S. Driscoll, 1977). All specialties except aerospace medicine and colorectal surgery offered a reduced schedule option (Baucom-Copeland, Copeland, & Perry, 1983; Coulter, 1972; Freeman & Waickman, 1988; H. I. Kaplan, H. S. Kaplan & Freedman, 1964; H. I, Kaplan, 1972; H. S. Kaplan, 1962; E. Shapiro & S. Driscoll, 1977; E. C. Shapiro, Lane, S. G. Driscoll & Books, 1980). 60 Mentor Relationships Lorber (1975, 1981, 1982, 1984, 1986, & 1987) emphasized the importance of sponsorship, or identification and help by established physicians, and the negative effect of a lack of mentors and sponsors (Ducker, 1988; Grant, 1988; S. R. Kaplan, 1982; Lyacki, Josef & Chapin, 1983; Ochberg, Barton & West, 1989; Sirridge, 1985; H. M. Wallace, 1980). "Sufficient women role models, then, for the next generation of women in medicine may create the impact that we all wish could have happened yesterday" (Fagin, 1984, p. 90) . Specialty Distribution For Women Physicians The seven specialties practiced most by women were the same in 1986 as in 1967, as though some constant operated in specialization by women physicians (Lopate, 1968; Powers, et al, 1969; Spieler, 1977). Primary care specialties were the major niche for women physicians in 1986 (Table A-5, A-6 and A-7), namely 16% in internal medicine (Golden, 1989), 6.4% in pediatrics (Coffin & Babbott, 1989; Golden, et al, 1989; Nadelson & Notman, 1972), 12.1% in general and family practice (Black, Schmittling, & T. L. Stern, 1980; Campos-Outcalt & Senf, 1989; Ellsbury, Schneeweiss, Montano, Gordon & Kuykendall, 1987; Ogle, et al, 1986), and 5.5% in obstetrics and gynecology (Golden, 1980; Gough, 1975; Spellacy, 1982). 61 When including women residents (Custer & Dimon, 1987; Table A-7, A-8, Figures 5 and 6), 56% of women practiced in 1986 in the primary care specialties. The specialties in which women were practicing in 1986 no less than in proportion to their numbers in the workforce (15%) included: anesthesiology (Calmes, 1985), child psychiatry, dermatology (Ducker, 1978), diagnostic radiology (Becker-Manheimer, 1967; Ducker, 1978), forensic pathology (Lorber, 1984), general preventative medicine, neurology, pathology, pediatric allergy, pediatric cardiology, physical medicine and rehabilitation, psychiatry (Nadelson & Notman, 1972), public health, and radiological oncology. Interpretations of Women's Specialty Distributions The data on women physicians from the AMA masterfile reflected the cumulative effect of the past choices by and prospects for women in medicine (Wunderman, 1980). A similar pattern of limited choices by women was in other health occupations (Butter, Carpenter, Kay & Simmons, 1985) . Reasons Expressed bv Women The woman physician had been viewed as either a passive party, or as an active agent in her specialty choice process. Passive The most restrictive specialty selection process had been posited by Bodel and Short (1972) and M. Cohen, Woodward and Ferrier (1988). They proposed that women were 62 conditioned to accept limited places and roles in social and medical institutions (H. M. Wallace, 1980; WestlingWikstrand, Monk & C. B. Thomas, 1970). Social and cultural factors defined medicine as men dominated and controlled (Ehrenreich, 1975; Kritzer & Zimet, 1967; S. C. Martin, et al, 1988), so women had been excluded as exceptions (Anonymous, 1986; Barthauer, 1989; E. D. Cohen & Korper, 1976a, 1976b; Ducker, 1978; M. B. Harris & Conley-Muth, 1981), or inhibited (Lyacki, Josef and Chapin, 1983). Hayes (1981) and Lorber (1984, 1987) wrote that women took the culturally approved choices in the light of their reduced options (Dickstein, 1990), thus reinforcing the perceived lack of ambition for women physicians. A similar point of view was that women were being tracked (Elliot, 1981; Lorber, 1984) or directed (Nadelson & Notman, 1972; Nadelson, et al, 1981; Notman & Nadelson, 1973; Reiman, 1980) toward certain fields or settings (Psathas, 1968) . These internal and external pressures were not necessarily logical (Quadango, 1976) nor most appropriate for their talents and interests (M. J. Gray & Ackerman, 1978; A. Lake, 1967). Men faculty, regardless of their own field, recommended specialties differently for women than men students (Ducker, 1978). Dralle, Daum, Heim and R. C. Elston (1987) found that medical students described characteristics of "most" and "male" physicians alike, but the "female" physician was dissimilar to each of the former. Matteson and S. V. Smith (1977) found that 63 women were less likely to be in a residency that was their real preference, but men were much more likely to be so. Day (1982) identified women physicians dissatisfied with their present career, countering the conclusion that women chose these specialties and practice arrangements. Davidson (1979) and Rinke (1981b) likewise concluded that choices by women physicians were constrained by limited options. Lorber (1984) considered that women physicians were viewed and treated as a "reserve army", achieving neither full training nor full participation in medicine. Active An alternate view was that women were active selectors of their specialties (Eccles, 1987), according to the value of the specialty and the likelihood of success. Bourne and Wikler (1978) described women as "in collusion", as women settled for "comfort zones" within the structure and organization of men-defined medicine (Shore, 1984). Women reported intellectual challenge, talent, patient population and emotional challenge as their reasons for their specialty choice (Bergquist, et al, 1985; Schermerhorn, Colliver, Verhulst, & Schmidt, 1986; Zimny & Shelton, 1982). Stereotyped Roles Career opportunities for women physicians had been restricted (Perun & Bielby, 1981) by stereotypes of female physicians as exceptions, just for support roles (Clark & Rieker, 1986). Double-binds of women as bad doctors were an 64 extreme sex stereotype (Freedman, H. I. Kaplan, fit Sadock, 1980; A. Lake, 1967; Potter, 1983). Stereotypical views of women in medicine were published by Kosa and Coker (1965), Holtan (1969), Barclay (1973), and recounted by M. J. Gray and Ackerman (1978) and Davidson (1979). Quadagno (1976), C. Eisenberg (1983), Potter (1983), Kris (1985) and W. C. F. (1986) refuted the stereotypes' validity. Dr. C. Eisenberg (1981) criticized the methodological assumptions of "the (sex) 'Difference' argument - whether it is offered as a rationalization for the status quo or as a justification for change - is a two-edged sword" (p. 45). Specialty Selection bv Family Responsibilities. Women's sex role and traditional gender role expectations as parent and spouse were considered dictators of career decisions in medicine (Bulkley, 1981; Cartwright, 1987; Clark & Rieker, 1986; M. Cohen, et al, 1988; Elkin, 1971; J. D. Gray, 1983; Jussim & Muller, 1975; S. R. Kaplan, 1981; Lorber, 1982; S. C. Martin, et al, 1988; MorantzSanchez, 1985; P. A. Williams, 1971). Lorber (1984) protested these assumptions as circular. First, medicine expected less of women in medicine based on restrictive sex role assumptions (Alumnae, 1897, in Abir-Am & Outram, 1987; Bergquist, et al, 1985; Bonar, et al, 1982; Davidson, 1979; Ducker, 1988; Moore-West, Lucero, Christy & Kaufman, 1982) . Second, medicine "tracked" (Elliot, 1981) women into lower status careers, and third, ultimately concluded that women made these choices based on their sex 65 and gender roles. Angell (1981, 1982) and Brodkin, Shrier and Buxton (1982) encouraged medicine to take into account family needs for all physicians. Specialty Selection bv Personality Factors. Data were inconclusive whether personality factors contributed to a physician's selection of specialty (Beil, Sisk & Miller, 1980; Bodel & Short, 1972; Cartwright, 1972a, 1972b, 1977a; L. Eisenberg, 1981; Graves & C. B. Thomas, 1985; McGrath & Zimet (1977a, 1977b); Notman & Nadelson, 1973; Sclabassi, 1984). Specialty Selection bv Status Factors. Aguirre, Wolinsky, Nierderauer, Keith, and Fann (1989) found that the status of health professions was tied to gender (man was higher status) and race (white was higher). A. Goldblatt and P. B. Goldblatt (1976) uncovered a hierarchy of physician status, with USA trained men were at the top, USA trained women in the middle, and graduates of international medical schools were in the lowest status specialties (Davidson, 1979; Rosenthal & Eaton, 1982; Worobey & Mick, 1987). Men medical students assumed that women were content in the lower levels of the profession, justifying imposition of restrictions on women's careers (W. Gross & Crovitz, 1975). Bowman and M. L. Gross (1986) wrote that "the specialties that women have chosen tend to be those with lower prestige and lower income....It may also be that certain specialties carry less prestige partly because they 66 have high numbers of women" (p. 515). Schermerhorn, et al (1986) found that women rated status as unimportant in their choice of specialty. Specialty Selection bv Income Potential. Women earned less than men in the same specialties, after equalizing the hours worked (Avery, 1981; Bergquist, et al, 1985; Bobula, 1980a, 1980b; Bowman & M. L. Gross, 1986; Hendrickson, 1975; Lisoskie, 1986; Mitchell, 1984; Ogle, et al, 1986; Ramos & Feiner, 1989; Rinke, 1981a). Women physicians received a lower payment for the same medical procedure as men even with personal, educational and experiential factors matched (J. W. Adams, 1977), indicating sex discrimination. Schermerhorn, et al (1986) showed that USA-trained women physicians in Illinois rated income as unimportant in choice of specialty. Possibly the encouragement of women into lower income specialties and practice patterns (Custer & Dimon, 1987) may have contributed to women's lower income. Patient Preferences Patients held competency stereotypes of men physicians as more, and women as less competent (Ackerman-Ross and Sochat, 1980; Engleman, 1974; Fennema, Meyer & Owen, 1990; Haar, Halitsky & Strieker, 1975; Weyrauch, Boiko & Alvin, 1990). Engleman (1974) found that 75% of patients preferred a man doctor (M. B. Harris & Conley-Muth, 1983; Kasteler & Hulme, 1980). Women physicians had a different patient base 67 than men physicians for the same specialty (Custer & Dimon, 1987). Patient prejudice had been reducing as women physicians were receiving more referrals (Ackerman-Ross & Sochat, 1980; J. W. Adams, 1977; Antonelli, 1986; Baum, 1980; Lorber, 1984; D. Myers, 1986). Women residents were rated better in care by both male and female patients, than men residents (Linn, Cope, & Leake, 1984). Present Trends in Specialty Choices Women physicians have continued to enter primary care specialties more frequently than men (AAMC, 1988; Babbott, et al, 1989; Lieu, Schroeder, & Altman, 1989). E. K. Adams and Bazzoli (1986) noted convergence of men and women physicians' specialties (Ferrier & Woodward, 1982; Kutner & Brogan, 1980; Maheux, Dufort & Beland, 1988). et al, R. H. Rosen, (1981) found no gender differences in Michigan's MD medical students' preferences for either specialty or geographical location to practice. Comments on Specialty Selection Bv Women The maldistribution of women physicians among specialties reflected both passive and active choices. M. A. Elston (1977) and C. Eisenberg (1981) asked if these interpretations were post hoc rationalization of the existing sex-specialty differences, or a defensive justification of the sex distinct forces operating in medical specialty distribution and career opportunities. 1966, Dr. Marion Fay emphasized the importance of accurate In 68 data on women physicians' careers to disprove negative stereotypes, and to promote recognition of women physicians' achievements. Professional Activities Of Women Physicians Graves and C. B. Thomas (1985) isolated three career patterns of women physicians, dubbed academic (teaching, practice and research with a university affiliation), clinical (medical practice) and circumscribed (limited achievements). Table 4 Changing Practice of Women Physicians. 1969-1986 Practice Active All Active MDs in Primary Care Hospital Staff Note. 1969 1986 84.3% 92.6% 6.7% 15.2% 46.3% 53.3% 16.9% 19.0% Data for 1969 adapted from Renshaw and Pennell (1971), "Distribution of Women Physicians, 1969", The Woman Physician. .26(4) , pp. 187-195. permission. Adapted with Data for 1986 adapted from Physician Characteristics and Distribution in the US. 1987 69 Edition, pp. 45, 19, 20, 25. Medical Association, 1987. Copyright 1987 American Adapted with permission. Renshaw and Pennell (19714) and Pennell and Renshaw (1972, 1973) reported from AMA data that in 1971 83.5% of women were active and 7.1% of active physicians were women. Seventeen years later (Tables 4 and A-3) 92.6% of women physicians were active professionally, an 11% increase, and the percentage of active physicians who were women more than doubled by 1986 (15.2%). There have been few statewide studies reported in the literature. The only full census study and the largest statewide study published, was the 1978 California Board of Medicine Quality Assurance questionnaire (Bauder-Nishita, 1980). A Connecticut (Callan & Klipstein, 1981) survey, with a 44.2% total response of women physicians, found similar patterns of professional activities. In North Carolina (M. S. Stern, 1976) "female and male physicians tend to have different types of medical careers" (p. 1012), based on AMA data. Quebec women general practitioners were more likely to be in a salaried, office practice in a metropolitan setting than men (Maheux, Dufort, Lambert & Berthiaume, 1988). Most recently Dickstein (1990) surveyed every woman physician who had either graduated from a Kentucky medical school or who had ever practiced medicine in Kentucky (35% return rate). discrimination in their careers. These women reported gender 70 Patient Care Table A-3 illustrated that women were slightly less represented in federal medical services. In the nonfederal sector women were proportionately more likely to be in administration, teaching or research, and less in office based patient care (Table A-3 and Figure 4; AMA Center for Health Policy Research, 1988; Spieler, 1977). Underemployment "This is the real inefficiency of women in medicine: talented physicians reluctantly spending their entire professional lives on the fringes" (Angell, 1981, p. 1162; Bodel & Short, 1972; Kroser, 1987; Woodside, 1975). Women may not have been practicing in their specialty, or were practicing in a limited fashion, and rarely in authority (Lorber, 1981) or leadership positions (Blanchette, 1986), as though invisible (Shore, 1984). Women physicians have felt underutilized in many countries (Baumgartner, 1966; Kirchner, Kossoff & Pickens, 1984; Lloyd-Green, 1966) for primarily social factors (Crosse, 1967; Farrell, Witte, Holguin, & Lopez, 1979; Keyserling, 1969; D. Myers, 1986; Westling-Wikstrand, et al, 1970). M. J. Williams (1969) opined that "the profession has failed to evolve" (p. 862), and recommended reforms to utilize women physicians better. Private Practice In spite of encouragement for women to enter solo practice (Coryllos, 1984), Bobula (1980b) found in the 1970s that men physicians were more likely to work in traditional 71 solo, fee for service practice or in a group practice setting, than women. However, all physicians were more commonly in group practices in the 1980s, and especially women as men had infrequently included women physicians in medical groups (AAMC, 1986). Salaried Positions Silberger, et al, (1987) found women physicians less likely to be self-employed (52.4%) than men (77.2%) (Cartwright, 1977b; E. D. Cohen and Korper, 1976a, 1976b; Heins, et al, 1978; S. C. Martin, et al, 1988; Morrow, 1973). Women physicians have disproportionately been employees, up to twice as likely as men (E. K. Adams & Bazzoli, 1986; Custer & Dimon, 1987). Phelps (1968) assumed that salaried positions were appealing to women physicians because the fixed hours accommodated sex-roles, as had E. P. Brown (1962), Ruben (1972), Davidson (1979), and Maheux, Dufort and Beland (1988). However, R. W. Schwartz, Jarecky, Strodel, Haley, Young, and Griffen (1989) and Griffen and R. W. Schwartz (1990) found that residents of both genders were choosing their specialty based on a controllable time commitment. Bennett and Nickerson (1990a) suggested another reason that women accepted salaried medical positions was that women perceive barriers to other medical practice prospects. 72 Hospital-Based Practice In Table A-3, one anomaly in the work pattern of the woman physician was the high percentage in hospital-based practice (Davidson, 1979; Heins, et al, 1978; Spieler, 1977) . Professional Productivity of Physicians Hours Per Week and Weeks Per Year Worked Women and men physicians' work hours were a measure of their contribution to medicine (Shank, 1988). The gender difference had narrowed since Dykman and Stalnaker's (1957) historical report. Hours per week and weeks per year were lower for women, even for a full-time continuous appointment (Bobula, 1980a, 1980b). This recurring observation led to recent conclusions that with more women in medicine, overall medical productivity will reduce in the USA. The previously predicted physician surplus (Fulop, 1986a, 1986b; GMENAC, 1981) may be absorbed by the year 2000, due both to more women who worked a shorter work week, and men practising fewer hours than the previous average (M. J. Lanska, D. J. Lanska, & Rimm, 1984; W. B. Schwartz, Sloan, & Mendelson, 1988) . Silberger, et al (1987) reported from 1986 AMA data that the number of weeks practiced per year were almost the same (47.5 for men, and 47.6 for women). Women spent fewer hours per week in medical and administrative duties combined (52.6 for women and 58.2 for men). Men saw more patients 73 per week and per hour (99 for women and 120 for men). Dennis, et al (1990) confirmed the latter results, suggesting that traditional family roles continue to have an impact on both male and female physician behaviors. Hojat, et al (1990) excluded data from part-time men and women, and the hours worked by full-time physicians remained statistically significantly different for gender. The hours that a woman anticipated working was a factor in choice of both her specialty and sub-specialty (Davidson, 1979; Ferrier & Woodward, 1982). Patients Per Hour and Per Week. Curry's (1983) Canadian Maritime women physicians worked the same length of week, and over 45 years of age, saw more patients in their shorter work week than did men physicians. Using another productivity index (Silberger, et al, 1987), USA women physicians served fewer patients per hour and per week in 1986 (99 patients per week by women and 120 by men). In 1988, Mitchell, Schurman and Cromwell found that the average office visit had increased one minute over 7 years as there were more women physicians in the USA, who spent more time with each patient. British women and men general practitioners provided equal quantities of patient care (Hooper, 1989). Why Women Physicians Work Differently From Men Phelps' (1968) survey of women physicians gave three reasons why women work less professionally than men: child bearing absences; the types of medicine they practiced; and 74 their frequent choice of salaried positions with set hours (Silberger, et al, 1987). Fourth, men were more often self- employed with economic incentives to work more hours (Freiman & Marder, 1984). Lisoskie gave a fifth reason as women's non-paid family job, or "dual commitments to medicine and family" (Antonelli, 1986, p. 428). Cartwright (1977b) added a sixth dimension, that the hours worked by women physicians reflected their career continuity. A seventh reason women physicians worked "less" was methodological due to inclusion of those who work part time or were inactive, reducing the mean. Hours Worked bv Women if Married. With Children. Family commitments were assumed as why women physicians were professionally inactive (Dykman & Stalnaker, 1957), and remained the accepted expectation and conclusion regarding all women physicians (L. Eisenberg, 1971a, 1971b). Fett (1976) confirmed that women physicians worked shorter hours per week if married, shorter again with one child, and shorter still with three or more children. There was less post-graduate training for women physicians if married with children. The same findings were presented by Shaw and D. Shapiro (1987) from the National Labor Survey of Young Women and other investigators (C. S. Shapiro, Stibler, Zelkovic, & Mausner, 1968; Sinai, Weavil & Camp, 1988; Uhlenberg & Cooney, 1990; Vaschak, 1970, 1972; WestlingWikstrand, et al, 1990; Wheeler, et al, 1990; and Woodward, et al, 1990). 75 Part-Time Practice During physician shortage eras (Antoine, 1966; S. M. H. Roberts, 1967), part-time practice by women was acceptable and appreciated. Hojat, et al (1987) found that 15% of women and 2% of men worked part time. M. A. Elston (1977) noted an ambiguous attitude to part-time medical work due to double standards. Part-time was viewed negatively for women as a failure to utilize fully their training, and positively for men as a sign of professional and financial success (Elliot, 1981; Offenbach, 1982). Only Rosenlund and Oski's (1967) productivity findings took into account that women physicians practice with efficiency and resourcefulness in a shorter work week. Inactivity Another productivity index was the rate of inactivity which in the late 1980s for the first time in documented USA medical history was more for men than women (Table 3). There were only 7.4% of MD women inactive in 1985 (AMA, 1987), a 62% reduction from 12.3% in 1970, and 8.0% in 1978. There were 8.4% of MD men inactive in 1985, a 20% increase from 5.9% in 1978. For international comparison, M. A. Elston (1977) found that one-third of both men and women physicians trained in the UK were not practicing in the UK. Men were more likely to emigrate permanently from the UK, whereas women reentered the UK health system after career interruptions. 76 Ducker (1986) found that if women had career interruptions, they were brief, measured in months only for child birth and family responsibilities (Ferrier & Woodward, 1982; Geyman, 1980; Heins, et al, 1976; Heins, Smock, L. Martindale, Jacobs & Stein, 1977; Heins, Smock, L. Martindale, stein and Jacobs (1977); Lerner, 1981). Cartwright (1977b) found that "continuous” women physicians worked longer hours than those who had taken time away from training. Length of Career Women planned to work longer in their lives than men physicians. Women's lifetime productivity may equal men's (Batchelor, 1990; Heins, Smock, L. Martindale, Jacobs and Stein, 1977; Morgan, 1971; J. H. Watson, 1977). Convergence in Hours Worked by Women and Men Bowman and M. L. Gross (1986) wrote "There is evidence that change in the relative productivity of male and female physicians is occurring ...(and) that the differences in productivity between men and women physicians have decreased.... Interestingly, most of the convergence in productivity appeared to be caused by male physicians' working less rather than by women physicians working more" (p. 517). Curry (1983), Freiman and Marder (1984), Mitchell (1984), Hooper (1989) and Woodward, et al, (1990) determined that both men and women, especially younger physicians, were reducing their work hours, closer to that of the general population. The increase in salaried physicians, both men and women, was the major factor sustaining the trend, along 77 with sharing after-hours coverage in group practice settings. In the future, women and men medical students anticipated working similar hours of work (Kutner & Brogan, 1980), or slightly fewer for women (Bergquist, et al, 1985; R. H. Rosen, et al, 1981). Teaching "Although women students constitute 52 percent of all students today, women faculty comprise only 27 percent of total faculty and only 10 percent of all full professors. Women constitute only 10 percent of all college and university professors." (ACE Commission on Women in Higher Education, 1987, p. 6). Likewise in USA medical education in 1986 women constituted 34% of all medical students, and 19% of all medical school faculty, of whom 56% or 10% of the total faculty were women physicians (AAMC, 1988c). AAMC (1989) reported that 15% of all USA women medical school graduates will at some time have a faculty appointment in a medical school, compared with 10% of men. Teaching was 50% more likely for women physicians (AAMC, 1988a). Teaching was recommended to women physicians as easy to enter (Beshiri, 1969), as socially acceptable for women (Bowman & D. I. Allen, 1985), and advantageous to society (C. Eisenberg, 1981). Nevertheless, in 1988 women were only 7% of full professors (medical or other health degree combined) or "a 78 deficit of professors and a surfeit of instructors among women” (Abstract, Wilkinson & Linde, 1986). Status of Women as Medical Faculty The underrepresention of women physicians in academic leadership and in administrative posts (Rinke, 1981a) had perpetuated "negative stereotypes about the capabilities and contributions of women" (Michigan Association of Governing Boards of State Universities, 1987, p. 2). Women physician faculty clustered in low ranks as shown in Table A-9 (AAMC, 1988b; C. Eisenberg, 1981; Lorber (1984); S. C. Martin, et al, 1988; Shore, 1984; Spieler, 1977; Whiting & Bickel, 1990) . Nickerson, Bennett, Estes and S. Shea (1990) concluded that women were advancing at faster rates than previously noted, but at half the rate of men (Dial, Bickel & Lewicki, 1989; C. Eisenberg, 1989; Table A-8). Further, more women than men were in clinical training tracks, less prestigious or secure than tenure track positions (Schaller, 1990). Women physician faculty had been underused (Wallis, Gilder & Thaler, 1981). (1989) Levinson, Tolle and C. E. Lewis "studied the system's 'survivors', and we did not evaluate the attitudes of women who chose to leave academic careers" (p. 1516) . Underuse issues were discussed by Witte, Arem & Holguin (1976), Davidson (1979), Farrell, et al (1979), Scully (1982), Nadelson (1983), Bowman and M. L. Gross (1986), Englund (1987), Wilson (1987) and AAMC (1988, 1989). AMWA developed a nationwide talent bank of women 79 physicians (Bernstein, 1989), including academic physicians (Scadron, 1980). The number of women Deans of USA medical schools (127 MD and 15 DO) reached a historical maximum of three in 1988 (L. Gross, 1990). All were at MD schools, two being international graduates. In 1991 there were none. Age may have been a factor in the faculty rank attained by women physicians (AAMC, 1988c; Table A-10) as the cohort of younger women (Figure 2 and Table A-4) were recent medical school graduates new to academia. Flexible working conditions had been recommended, including child care provision (Bennett and Nickerson, 1990b). Affirmative action at higher faculty rank was encouraged to improve the status of women in medical education (Shore, 1984; C. Eisenberg, 1989; Dean N. E. Gary in L. Gross, 1990; Levinson, Tolle and C. E. Lewis, 1990; Palac, Lee and F. Collins, 1990; Stiles, 1990) . Administration Women physician administrators had been recognized as a valuable and essential part of health care administration (Rothstein, in Baxter Foundation, 1989), but women were underrepresented in administrative posts (S. L. Brown & Klein 1982; Costanza, 1989; Dickstein & Nadelson, 1986; Dickstein & Stephenson, 1987; W. Gross & Crovitz, 1975; Loutsch, et al, 1982; Maheux, Dufort & Beland, 1988; S. C. Martin, et al, 1988; Rinke, 1981a; Sullivan, 1974). 80 Research Recently women had achieved parity in quantity of time spent in research with men (Table 3). A circular problem developed where the low level appointments of women at universities reduced the status of their scientific endeavors and grant applications (B. A. Levey, et al, 1990; Ramey, 1984). Then there was a lower rate of promotion and sponsorship of women scientists as they had fewer research projects, grants and publications. Just as women entered research in greater numbers in the 1980s, federal research funds reduced, and there was still a low representation of women on peer review panels for research funds (Ramey, 1984). Nadelson (1983) reported that the grants awarded to women physicians were fewer, and lower in sum and status of sources. Research had not contributed to improved prospects for women in medical research and teaching (Fleischer & Abney, 1990). Geographic Distribution Metropolitan Location of Women Physicians Bowman and M. L. Gross (1986) reported that "physicians generally were not well distributed in relation to the population; however, male physicians are better distributed than female physicians, since women are more likely to locate their practice in physician-rich urban areas" (p. 515) . 81 Definition of Rural and Metropolitan The Michigan Department of Commerce's (1990) population-based model (Appendix B, Table B-l) defined rural counties as that with no town equal to or larger than 50,000 persons. A second model (Table B-2) defined rural counties both by their population and contiguity to metropolitan counties. Around the world, physicians tend to congregate in cities (Are there too many doctors, 1986). The distribution of health workers is a process objective toward the World Health Organization goal of "health for all by the year 2000" (Fulop, 1986a). Women physicians were more likely than men to practice in a metropolitan setting (Maheux, Dufort & Beland, 1988; S. C. Martin, et al, 1988; Woodward, 1986). Vogel (1985) outright recommended that two-doctor couples practice in a large city, each in a group practice, for professional support. However, in 1984, 50% of osteopathic physicians were practising in a rural county (Denslow, et al, 1984). Reasons for Distribution Patterns of Women Physicians E. D. Cohen and Korper (1976a, 1976b) and Lefevre and Colwill (1983) found physicians' choices of practice location could be predicted primarily from the location of their medical schools and graduate medical education, and their specialties. L. Eisenberg (1981) summarized the asymmetrical effect on women's location in a two career family, in that the 82 men's job comes first. Moore-West, et al (1982) surveyed women physicians in rural New Mexico (26 respondents). Married women had chosen rural.sites more often based on family reasons and their husband's preference, while single women physicians chose more often for professional and personal reasons. The choice of location to practice tended to be more mutual for physician couples under 40 (Lorber, 1982). Ogle, et al (1986) found that spousal issues were important to women family physicians in selecting their practice location, but not to men family physicians. Kutner and Brogan (1980) found too that women medical students, anticipating spousal problems whether single or married, were not choosing rural locations. Women physicians had been working more frequently in rural areas in recent decades. D'Elia and I. Johnson (1980) studied the distribution of women physicians in non­ metropolitan Illinois, finding a 156% increase in women in the area. Nationally, women and men medical students' and residents' preferred geographical location was comparable (E. K. Adams & Bazzoli, 1986; R. H. Rosen, et al, 1981). Age, Year Of Medical School Graduation Men and women physicians who were younger, and who were more recent medical school graduates were choosing similar specialties (E. K. Adams and Bazzoli, 1986; Ferrier & Woodward, 1982; Kutner & Brogan, 1980; Maheux, Dufort & 83 Beland, 1988; R. H. Rosen, et al, 1981). Kutner and Brogan (1980), R. H. Rosen, et al (1981), E. K. Adams and Bazzoli (1986) , and Reiman (1989) found that recent medical school graduates had similar plans for rural or metropolitan location, as had been projected by M. S. Stern (1976). The year of medical school graduation accurately identified the number of years that a physician had been available for practice. It was more sensitive than age, as older students had been admitted to USA medical schools after 1970. Quantitative changes were the increasing number of women and racial and ethnic minorities admitted to medical school. Qualitative changes were the active efforts to reduce inequities in the opportunities for women and minorities in medical education. Differences between USA and international medical graduates may show in the specialty and practice patterns of women and men graduating before the changes of 1970 being implemented in the USA. United States Versus International Medical Education Worldwide physician shortages had led to an 253% increase in the number of medical schools between 1950 and 1986 (Fulop, 1986b) among all 166 World Health Organization member nations, producing many more doctors (Are there too many doctors, 1986). The USA benefitted from importing internationally trained doctors. 84 Bowman and M. L. Gross (1986) noted that in 1981 women graduates of international medical schools were proportionately more common in the USA. In 1970 (Mejia, et al, 1980), women were 15% of all internationally trained physicians practicing in the USA, when USA trained women were 6% of all USA trained physicians. Nationwide in 1986, 24% of MDs in patient care were graduates of international medical schools, of whom 87% were originally citizens of the country in which they trained. In Michigan in 1987, 27% of MDs were graduates of international medical schools (AMA, 1989). Specialty practices of graduates of international medical schools, have been similar to USA trained women physicians. Internationally trained physicians were employed in institutionalized settings even more frequently than USA trained women. Geographical location of internationally trained physicians in the USA had been over twice as often in a metropolitan setting (Mejia, et al, 1980) as USA trained physicians. Women graduates of international medical schools were in lower status specialties than men graduates, and in different specialties than USA trained women. They lacked career mobility from their lowest level point in status (A. Goldblatt & P. B. Goldblatt, 1976; Mick & Worobey, 1984; Worobey & Mick, 1987). The role and status of USA trained women physicians may alter (Kletke, Schleiter & Tarlov, 1987; Rosenthal & Eaton, 1982) in response to a reduction in 85 the number of international physicians arriving in the USA in the 1990s (Fulop, 1986a). Contemporary Issues Shore (1984) exhorted medicine to learn from history, as women's participation in medicine had most strongly been influenced by external forces. Women had been encouraged to create an improved environment for advancement of other women (Fagin, 1984; Friedman, 1988; Morantz-Sanchez, 1985). Narrowing Of The Gender Differences Men and women had approximately the same results on medical education tests (Roessler, et al, 1975) as traditional gender based test differences were declining (E. K. Adams & Bazzoli, 1986). There were still wide differences in the current physician workforce (Bowman & M. L. Gross, 1986). Sex differences in medical practice were diminishing rapidly (Bergquist, et al, 1985; Table 4; Weisman, et al, 1980). Actual changes in norms were occurring as more women become physicians (Beck, 1988; Bluestone, 1990). The average patient visit time for all physicians lengthened (Mitchell, et al, 1988). Women physicians' hours were increasing and men physicians' hours were decreasing (Curry, 1983; Freiman & Marder, 1984; Geyman, 1980; Mitchell, 1984; Weisman, et al, 1980; P. B. Williams, 1978). 86 Both women and men students planned to work a shorter work week than the national physician average (R. H. Rosen, et al, 1981). Walsh (1990) argued that as the unique or historical barriers for women in medicine reduce, all medical trainees will benefit from an improved environment. As increasing numbers of men and women physicians were salaried, the average number of patients seen per week will drop (M. R. Schwartz in E. Ginzberg, 1986). The perception that women were different physicians was persisting even as women and men practiced more similarly (S. C. Martin, et al, 1988). Most measurable differences were narrow, including less prejudice against women physicians (Kasteler & Hulme, 1980). Men Changing Dr. C. Eisenberg (1981) defined the goal for medicine was for men to become more like women, with human values, not men or women values. Male medical students have developed more sympathy for women medical students (Leserman, 1980). their work week. Men were reducing More men physicians wanted to be more family oriented (Bennett and Nickerson, 1990a; Kinder, 1985; Lisoskie, 1986; Maheux, Dufort, Lambert & Berthiaume, 1988; Nadelson, et al, 1981; P. B. Williams, 1978), reflecting earlier advice that men physicians participate more in family life (Bodel & Short, 1972; Bourne & Wikler, 1978; Stohl, 1990). 87 Women Changing McGrath and Zimet (1977a, 1977b) noted that women's personality profile was becoming closer to the masculine profile over time. Women's career aspirations were changing (Clavan & Robak, 1978; Nadelson, et al, 1981; Weisman, et al, 1980). E. K. Adams and Bazzoli (1986) and Burnley and Burkett (1986) measured women's movement into underrepresented specialties, as Szkop-Frankiel (1967) had encouraged 19 years earlier. Geyman (1980) found less stereotypical practice pattern by women physicians. There was greater continuity of practice by women physicians (AAMC, 1988c; M. Cohen, et al, 1988; Lorber, 1984; Schermerhorn, et al, 1986). Curry (1983) and Freiman and Marder (1984) recorded the increase in women physicians' work week. Donoghue (1988) found faculty salaries comparable for women and men. Women residents were more likely to enter their selected residency (AAMC, 1990b) and join specialty societies at the same rate as men (D. I. Allen, 1990; Pearse, 1990). Nevertheless, N. B. Schwartz (1973), Fagin (1984) and Lorber (1984) recommended that women physicians organize, to counteract barriers and to become active role models, mentors, and sponsors for other women in medicine (Hilberman, Gispert & Harper, 1976). 88 Specific Efforts to Create Change Swanson, et al (1989) recommended support for women such as collegial encouragement and governmental affirmative action, and changes in tenure processes to establish more permanent women faculty. Women have been endorsed in graduate medical education as residency directors (Freeman & Waickman, 1988; Tucker & Margo, 1987). Ackerman-Ross and Sochat (1980) suggested altering the images of medicine and of women physicians, to reduce patients' stereotypes and prejudices against women in medicine. Recommendations to change the structure, organization and financing of medical work were advanced by Bourne and Wikler (1978), D'Elia and I. Johnson (1980), Barondess (1981), Day (1982), and Lorber (1984). Summary The impeti for the increase in the number and proportion of women entering medical education in the USA in the past two decades have been largely external of the medical profession and the medical education continuum, such as market demands, social changes and legislative requirements. While women had entered USA medical schools in greater proportions, women physicians had not entered the specialties or the status and leadership positions in a similarly increasing proportion. 89 The more recent women medical graduates have broader distribution among the specialties and increased their hours worked and taken shorter breaks for child-bearing, and reduced their rate of professional inactivity. They have shown a greater inclination to practice in less metropolitan areas. At the same time, men physicians have shortened their work week, a sign of the change occurring among younger physicians. Women have influenced medical practice. They provide a more interactive relationship with the patient (Friedman, 1988) that had contributed to the lengthening of the average physician-patient interaction (Mitchell, et al, 1988). There were fewer malpractice suits against women physicians (Holder, 1979). They have contributed to the quality of health care (Maheux, Dufort, Lambert & Levesque, 1989). The rationale for this study is that as more women entered medicine in the USA, because women and men physicians have practiced differently, then the characteristics and quantities of medical services available to any one state may change. It is proposed to study the practice patterns of all physicians in a state, analyzing their specialty, geographical locations, hours and type of professional activities by gender, age, year of graduation from medical school, and nation of training. The research will be based on a secondary analysis of the surveys on professional activities completed by a total of 98% of the licensed physicians located in Michigan in 90 1986. This was a virtual census which included DO physicians, who have not been reported in any literature on women in medicine. The results will provide a clear picture of current physician professional activities, and projection of future trends, for planning for the needs for medical service, teaching, research and administration, and the needs of women in medicine. Ill RESEARCH DESIGN AND METHODOLOGY Secondary analysis of an amalgamated database drawn from three archived surveys was the basis for this study. The sources of data were matched by name, suffix, county of address in 1986 and professional degree. A single record was created for each physician for whom there was a match between the data sources. There are many advantages of secondary survey analysis (Kiecolt & Nathan, 1985), notably with the time and money saved in data collection. The results are comparable with those who have used the same instruments. The three data archives for this study were already encoded in machine readable data, enabling the research to be done independently. The technical expertise of each of the three source surveys' staffs was available to the researcher. The major limitation of secondary survey analysis is that the prior surveys may not have collected data precisely as a subsequent researcher would wish to operationalize a concept. The AMA database represented all MDs in Michigan on December 31, 1986. This date excluded those who responded to OHMA's survey while practising in Michigan during 1986, but had left the state prior to the last day of the year. The technical limitations of secondary analysis are the possibility of errors in sampling or coding of data, or insufficient documentation. None of these possibilities 91 92 interfered with the three data sets used for this study. Each was a full census count, without sampling occurring, and documentation and ongoing collection for each source maintained a high level of reliability and validity. Review of Research Methods of Prior Studies The historical baseline survey of women physicians' activities in the USA was by Dykman and Stalnaker (1957), on women physicians graduating from USA medical schools between 1925 and 1940. The next national study of physician activities and gender differences in the USA was a cooperative effort (Powers et al, 1969). It was based on a 1965 survey by the AMA, AAMC and AMWA of medical school graduates of the years 1931, 1936, 1941, 1946, 1951, 1956. It included all women (77% response) and a sample of men (75% response). Several national studies have been conducted on gender and physician activities solely through secondary analysis. AMA data were the basis for most (J. W. Adams, 1977; Bobula, 1980a, 1980b; Custer & Dimon, 1987; Freiman & Marder, 1984; Hendrickson, 1975; Kehrer, 1976; Kletke et al, 1990; Langwell, 1982; Pennell & Renshaw, 1972; Renshaw & Pennell, 1971, 1973; Silberger et al, 1987; Wunderman, 1980). Mitchell (1984) re-analyzed Health Care Financing Administration (HCFA) detailing hours and income data for gender and specialty differences. 93 With a national focus, E. K. Adams and Bazzoli (1986) re-analyzed the AMA 1983 Survey of Resident Physicians for gender and race differences in specialty choice. Similarly, A. Goldblatt and P. B. Goldblatt (1976) did an 8-year follow up on all physicians who had been interns and residents in 1963. Hadley (1977) introduced a two-source secondary analysis when utilizing both the AAMC longitudinal study of the 1960 pool of USA MD graduates, and their 1972 AMA physician profile for analyzing practice activities. Lorber and Ecker (1983) also used the above AAMC longitudinal database to follow up the 1960 graduates in 1976. There are a few multi-state studies of physician activities and gender differences. Curry (1983) studied all physicians in three Maritime Canadian provinces in 1979, and Golden (1989) surveyed honors graduates of 11 medical schools in several states. Statewide surveys have had return rates ranging from 44% (Callan & Klipstein, 1981) to 55% (Schermerhorn et al, 1986). The highest return reported was Bauder-Nishita's 1978 California study with a 98% return rate. Other statewide studies relied on re-analyzing AMA data (M. S. Stern, 1976). As a subset of state studies, some have selected a specific geographical region within a state to study physician characteristics. Such studies included physicians in non-metropolitan Illinois (D'Elia & I. Johnson, 1980), in 94 rural New Mexico (Moore-West et al, 1982), and in Detroit, Michigan (Heins, 1985; Heins, Smock, Jacobs, & Stein, 1976; Heins, Smock, & Martindale, 1978; Heins, Smock, Martindale, Jacobs, & Stein, 1977; Heins, Smock, Martindale, Stein, & Jacobs, 1977). Another state-focussed approach has been to first select a variable to define the population to be studied. Examples are studies of all specialists in Quebec (Maheux et al, 1989), all general practitioners in Quebec (Maheux et al, 1988) or all female physicians in Connecticut (E. D. Cohen & Korper, 1976a, 1976b). The most prevalent research design has been a survey of graduates of one or several medical schools (Cartwright, 1977a, 1977b, 1987; Graves et al, 1985; Harwood et al, 1981; P. B. Williams, 1978) or a particular residency or style of residency (Ellsbury et al, 1987; Ogle et al, 1986) . The results have limited generalizability but have been a primary source of information on gender differences and professional activities. Research Design The research design combined elements of several survey designs. The primary design was a single-measurement cross- sectional study. The purpose was to determine whether in 1986 in Michigan physicians' hours worked, professional activities reported, and geographical location varied with their specialty choice, gender, nation of medical training 95 or professional degree. More specifically, did the women physicians in Michigan report in 1986 different hours worked, professional activities, geographical locations and specialty choices than did men physicians. Did women DO physicians in Michigan in 1986 report different hours worked, professional activities, geographical locations and specialty choices than did USA trained women MD physicians. Did women MD physicians in Michigan who were international medical school graduates report in 1986 different hours worked, professional activities, geographical locations and specialty choices than did USA trained women MD physicians. Cross-sectional design is useful in field research as it is simple to take all measurements at one time, and easy to conduct in the field, by one survey card (Spector, 1981). The major shortcoming of the cross-sectional design is that it only measures relationships between variables. The research design included temporal analysis. Trend analysis technique mimics a simple time series analysis without making multiple measurements over time, by investigating change in the level or distribution of variables over time. Disaggregation of these trends can show changes in physicians' professional activities in Michigan. Separating observations of physicians by their age or by year of medical school graduation created two such time measures. Comparative analysis was also a primary part of the design, first with the comparison between women and men 96 physicians. Other comparisons were between USA educated and internationally educated MDs, and between USA educated MDs and DOs. Further comparisons were between primary and nonprimary specialists, younger and older physicians, early and recent medical school graduates, type of practice, and metropolitan versus rural location of practice. The purpose of comparative analysis was to determine whether observed differences between women and men physicians were universal, or whether other factors influenced them. Such factors included nation and year of medical school graduation, professional degree, specialty practiced, age and geographical location. The advantages of the three source data archives were that the same individuals responded to the items in the same year and state (OHMA, and either AMA or AOA surveys). Each was a complete census, of physicians in Michigan by OHMA, of all MDs in Michigan by AMA, and of all DOs in Michigan by AOA. Population The entire available population of physicians in Michigan in 1986 was the subject pool. No sampling occurred in any of the three source surveys, each designed to describe the entire population of physicians in Michigan by OHMA, all MDs by AMA, and all DOs by AOA. of the source surveys were voluntary. Responses to each Endorsement, 97 promotion and supervision by the professional societies fostered participation by physicians. The research was to be conducted on the full population of physicians licensed as MDs or DOs in Michigan in 1986 and with a controlled substance license in 1986, or for whom there was a 1986 record in the Department of Licensing and Regulation, and for each of which there was a verifiable entry in the AMA or AOA Physician Masterfiles. AMA Physician Masterfile has verified records on 99% of all MD physicians in the country, and the AOA Physician Masterfile has verified records on over 99% of all DO physicians in the country. Therefore the pool of subjects who have responded to the OHMA survey (98%) and have an AMA file or an AOA file would minimally represent 97% of the physicians in Michigan in 1986. The generalizability of the findings was limited to this available population, namely MDs and DOs licensed to practice in Michigan. Data Collection and Instruments Data Sources The data sources for this research were previously collected data from three existing questionnaire instruments. As described in the first chapter, the Office of Health and Medical Affairs of the State Michigan Department of Management and Budget conducted a survey of Michigan physicians' activities during 1986. A brief questionnaire 98 (Appendix C) was enclosed with the physicians' controlled substance license annual renewal application materials sent by the Department of Licensing and Regulation to over 90% of all physicians in Michigan. OHMA subsequently sent questionnaires to other MDs and DOs with active or recently expired professional licensees who had mailing addresses in Michigan or in adjacent portions of contiguous states. OHMA did not directly survey physicians with educational limited licenses, but used data from Department of Licensing and Regulation records. Follow up mailings were sent to physicians with mailing addresses in, or contiguous to Michigan who did not submit survey cards or who submitted incomplete cards in response to the initial mailing. The data collection card informed physicians that the requested information on their professional activities was voluntary, and did not affect the renewal of their controlled substance license. Further, all physicians were assured of confidentiality, as OHMA planned to report only anonymous data. Confidentiality and anonymity continued in this secondary analysis, as no survey responses were identifiable for an individual physician. Physicians with questions could call OHMA through a toll free telephone number. The purpose of the OHMA survey was for quantifying physician resources in the state for state planning. The 99 rate of return of the 1986 questionnaire, after 1986 and 1987 follow up, was 98% (technical notes in Appendix D ) . The results of the OHMA survey (Darga, 1988) were that professional activity was known for approximately 95% of Michigan's active physicians, and one or more practice locations for 93%. Patient care hours were known for 90% and specialty known for 93% of the active physicians who were not in graduate medical education. The validity of the OHMA study was based on the content validity model. The reliability of the OHMA study continued this study, as the secondary analysis was on the raw data. Data Acquisition The data was to be acquired in four steps. Office of Health and Medical Affairs First, the chief Office of Health and Medical Affairs (OHMA) researcher would copy the 1986 names, addresses, professional degree, license number, type and status for all physicians in Michigan in 1986 to a standard magnetic tape. Specialty designation was requested, but as it was data from the survey, rather than from public record, the commitment to confidentiality overrode the request for that data element. Specialty was added at the last step as survey data. American Medical Association Second, from its Physician Masterfile the American Medical Association (AMA) would match the OHMA names on the OHMA tape from its archived 1986 Michigan Physician 100 Masterfile. The requested individual data would be copied to the physician records. The specialty code (Appendix G's 85 categories) was added. The OHMA survey response for specialty would be the data used for actual analysis. The AMA uses secondary sources and direct mail surveys to maintain the Physician Masterfile to describe all MD physicians in the USA (569,160 in 1986) and many DOs (G. Robak, personal communication, 26 September, 1990). See Appendix E for items and method of scoring. The AMA Physician Masterfile is the most reliable, and has been the major national data source on USA physician activities since established in 1906. Assurance of the reliability of AMA Masterfile data was by its sources and verification process. As a historical file, AMA verified the 1986 data before being archived. The reliability continued in this study as the raw data was analyzed, without using any of the transformed variables developed by AMA. The Physician Masterfile is updated annually, and reports published from the databank for a variety of public and private uses (AMA 1987a, 1987b, 1988; USDHHS, 1987). The AMA ensured the consistency across the decades in the collection and verification process, the identical instrumentation and categorization, and checked for clerical errors before the data being released. 101 American Osteopathic Association Third, following the match by the AMA, the names and addresses of the DO physicians listed on the OHMA tape were to be matched by the American Osteopathic Association to their Physician Masterfile records (see Appendices H and I). For each matched DO the fields of gender, year of birth, school of medical education graduation, and self-designated principal specialty were to be added to the standard computer tape commenced by OHMA. The specialty code in the OHMA survey database was used in the analysis. The AOA uses secondary sources and direct mail surveys to describe all osteopathic physicians in the USA (26,794 in 1986) . See Appendix F for the items and method of scoring. The AOA Physician Masterfile has been the most reliable, and is the only national data source on USA DO physician activities since 1958. The first formal survey was in 1986, followed by a three year follow up survey in 1989, which the AOA considers the best available and most current source of practice detail on osteopathic physicians. Reliability of AOA Masterfile data was assured by its sources and verification process. The reliability continued in this study as the raw data was analyzed, without using any of the transformed variables developed by AOA. The Physician Masterfile is updated annually, and reports published from the databank for a variety of public and private uses (AOA, 1987, 1989). AOA ensured consistency in the collection and verification process, the identical 10.2 instrumentation and categorization, and checked for clerical errors before the data being released. OHMA Survey Data Fourth, after adding AMA and AOA data to the OHMA tape, this OHMA original tape was to be returned to OHMA by this researcher, and the matched physician name data set and the same physicians* OHMA survey responses combined by the chief OHMA researcher. To preserve the confidentiality and anonymity of the survey responses, the name and address data were to be deleted after AMA and AOA data had been added to the OHMA physician records. To enable analysis of parts of survey responses for each physician, a random number identification was to be assigned to each physician. An archived file would preserve the match between the individual license and the random identification numbers, for later follow up on OHMA's longitudinal database. After assuring this anonymity, the amalgamated data set was to be returned to this researcher on the standard tape. Physicians whose name and address were not accurately and completely matched with AMA or AOA sources, or for whom there was no OHMA survey response would be excluded from the study group. A disk file containing the data was to be established on the IBM 3090 computer at Michigan State University, and statistical analysis conducted by SPSS-X (1990). 103 Variables The composite data base consisted of the following variables, and classes of responses within variable: 1 Professional degree, either MD or DO. 2 Specialty, as the response to OHMA's survey question as to what was physicians "principal" (major) specialty, next their secondary, and any third specialty. Principal specialty (ascribed) and secondary specialty (reported) were the selected data elements. If the principal specialty was missing, a specialty had been assigned by OHMA based on at least one of the following: the physician's report of residency or fellowship; specialty information included in the physician's practice address; national databank records such as from the AMA for MDs and the AOA for DOs; and 7% were left blank as truly missing data when no data existed (Darga, 1988). Therefore the principal specialty variable used in this study contained reported and ascribed data. If the secondary specialty was missing, it was treated as "none", so no secondary specialties were ascribed. Therefore the secondary specialty variable contained only reported data. OHMA collapsed both principal and secondary specialty fields subsequently into AMA's 85 self-reported specialty categories. Expert advice from faculty of the Michigan State University College of Human Medicine assisted this conversion (K. J. Darga, personal communication, January 24, 1991). See Appendix G in which the 85 categories, collapsed in Appendix H into 39 categories, and again in Appendix I into 4 broad (abbreviated) categories of specialty code. Current professional activity status, as self-reported in the OHMA survey, whether undergoing internship, residency or fellowship training in Michigan, treating patients in Michigan, or working as a physician in Michigan, but no time spent in patient care, or professionally inactive or retired. Patient care practice locations, as self-reported in the OHMA survey. Zip code of each location was subsequently coded into the 83 Michigan counties, and ultimately collapsed into metropolitan or rural, using the Michigan Department of Commerce population based classification of counties (Appendix B ) . Hours worked per week, as self-reported in the OHMA survey. Hours worked per week for each patient care practice activity and location was reported, and summed into a total of hours worked per week for each physician. Hours in patient care, both hospital inpatient care and other patient care, or another activity were all reported. Examples of nonpatient care given in the survey instructions were "office, outpatient clinic, nursing home, etc.", but the class of the site was not reported individually. 105 6 Year ofbirth, as reported by the AMA 7 Gender, 8 Nation or the AOA. as reported by the AMA or the AOA. or USA state ofmedical school, as reported by the AMA or the AOA. 9 Year of medical school graduation, as reported bythe AMA or AOA. Independent Variables Gender was treated as an independent variable with respect to specialty, hours worked, professional activity (including current training), geographical location, professional degree and nation of training. Specialty was treated as an independent variable with respect to hours worked, professional activity, and geographical location. Nation of medical education was treated as an independent variable with respect to specialty, and with respect to patient care in an inpatient setting as professional activity. Professional degree for USA trained physicians was treated as an independent variable with respect to principal specialty. Year of birth was treated as an independent variable with respect to hours worked, principal specialty, and professional activity. Year of medical school graduation was treated as an independent variable with respect to hours worked, principal specialty, and professional activity. 106 Dependent Variables Total hours worked per week was treated as a dependent variable with respect to gender, year of birth, year of medical school graduation and specialty. Professional activity was treated as a dependent variable with respect to gender, specialty, geographical location, year of birth and year of medical school graduation. Geographical location was treated as a dependent variable with respect to gender, specialty, nation of training, year of birth and year of medical school graduation Confounding Variables There were two potentially confounding variables, year of birth and year of medical school graduation. It appears that younger physicians and more recent medical school graduates practice differently from older physicians and earlier graduates, for both men and women. Sources of Error Measurement Error Bias may have operated on the self-report of specialty or hours worked. Physicians may have reported more or less professional hours based on their perceptions of what might have been desirable or of interest to the State of Michigan. Bauder-Nishita (1980) wrote that women physicians under reported their hours of work and men physicians over reported hours of work. So systematic bias may have 107 occurred for gender in opposite directions in the hours reported worked by physicians. OHMA did not employ any methods to detect bias or faking. Cases were not eliminated as this would affect the generalizability of the findings of the study. Sampling Error The other source of error was the loss of any subjects (Spector, 1981) without a verifiable, accurate and complete AMA or AOA record for the requested variables. Loss was less of an error than inclusion of incomplete or inaccurate information on physicians. The extent of the error from loss could not be estimated with any statistical reliability (Jacob, 1984). One of Jacob's recommendations was to round the numbers to avoid the false impression of accuracy. It could be justifiable to do small roundings due to the large number of responses and the high reliability and validity of the three data bases combined for this study. There was no estimate of the magnitude of this error. Measurement Nominal Measurement The following classes of variables are nominal: gender, professional degree, principal and secondary specialty, nation or USA state of medical education, professional activity, and geographical location. 108 Interval Measurement The following classes of variables are interval: year of birth, year of medical school graduation, and hours worked. Data Analysis The levels of measurement are predominately nominal variables with several interval scaled variables. There were limits to the statistical analyses using SPSS-X (1990) that would be feasible and meaningful on the amalgamated database used for this study. The large size of the population studied, together with the absence of sampling induced errors, enhanced the power of the study. The exclusion of records occurred because the AMA or AOA Physician Masterfile matching to state records failed, or state survey responses were not available. The phrasing of the hypotheses was a special focus. As acquisition of scientific knowledge is a cumulative process (Henkel, 1976), the hypotheses were phrased according to the predictions of the literature. The directional or a priori hypothesis also circumvented the uninformative results from rejecting or accepting the null hypothesis. Meehl (1976, in Henkel, 1976) indicated "the usual procedure for corroborating theories - by refuting null hypotheses - has a high probability of corroborating the theory, even if the theory is false" (p. 84). 109 Covariation between any two 2-point nominally scaled variables was first tested by Pearson's chi square. When variables were collapsed to two points, it was not to overcome any asymmetry within the variables (H. T. Reynolds), but to study specific subgroups of categories of a variable. An example was physicians' year of medical school graduation, whether after 1970 or before 1971. When one of the variables was intervally scaled, such as total hours reported worked per week, statistical testing was by analysis of variance. This tested for the significance of effects, and provided an estimation of the variation in each effect (Iversen & Norpoth, 1976). The t- test was the preferred test of significance when homoscedasticity could not be assumed such with year of medical school graduation. Women were known to be skewed to recent graduation (Tables 1 and 2). The F-test was the preferred test of significance when homoscedasticity could be assumed (Andrews, et al, 1981; Henkel, 1976). Hypotheses Hypothesis 1. Women physicians report specialties of family medicine, internal medicine, pediatrics and psychiatry more than do men, as their principal specialty. First these four specialties (psychiatry including child psychiatry) will be identified from the AMA list of 85 categories of specialty (Appendix G ) . All 81 others will become a new category of "other" to create a five point 110 scale. Covariation of specialty with gender will be tested by Pearson's chi square. Hypothesis 2 . Women physicians report primary specialties of family practice, general practice, internal medicine, obstetrics and gynecology, and pediatrics more than do men, as their principal specialty. These five primary care specialties will be identified from the AMA list of 85 categories of specialty (Appendix G) and the remainder classified as "other". Covariation of primary care specialty and gender will be tested by Pearson's chi square. Hypothesis 3 . Men physicians report more even distribution across specialties than do women. With gender as one variable and specialty as the other (a nominal scale of 39 categories, Appendix H ) , covariation of gender and specialty will be tested by Pearson's chi square. Hypothesis 4 . Younger (born after 1945) men and women physicians have similar specialty distributions. Physicians will be divided by age into younger (born after 1945) and older (born before 1946). This year was chosen as the critical point at which to disaggregate the physicians, as physicians born after 1945 would have been admitted to medical school or postgraduate training, during Ill or after the events in 1971 that lead to the increased enrollment of women in medical school (Allison, 1984). Similar age based disaggregations have been applied by E. D. Cohen and Korper's (1976a) survey of 687 women Connecticut physicians. They grouped the respondents into their year of finishing medical school, as before 1945, from 1945 to 1959, and from 1960 to 1971, to have comparable sized groups. In addition, these epochs were meaningful for women in medical education. D'Elia and I. Johnson's (1980) secondary analysis of Illinois physician directories selected 1950, 1965, 1972 and 1979 as key years for changes for women in medicine. The first two resemble this study's current older age variable, and the second two are like the younger group. The assumption was that these younger physicians were influenced differently from those physicians trained before the increased enrollment of women in medical school. The AMA 4 broad (abbreviated) categories of specialty (Appendix I) will be the other variable. Covariation of specialty with age will be by Pearson's chi square. Hypothesis 5 . More recent medical school graduates (after 1970) men and women physicians have similar specialty distributions. Physicians will be divided into recent (after 1970) or early medical school graduates (before 1971) as explained 112 above. Those graduating after 1970 trained in a changing medical education environment for both women and men. The AMA 4 broad (abbreviated) categories of specialty (Appendix I) will be the other variable. Covariation of specialty and year of graduation will be by Pearson's chi square. Hypothesis 6 . Women physicians report a lower mean total patient care hours worked per week than men. Hypothesis 7 . Women physicians have a larger variance in the total hours worked per week than do men. Gender is the independent variable and total hours per week is the dependent variable. Gender differences in total hours worked per week will be tested using the t-test. Hypothesis 8 . Women physicians report a lower mean total patient care hours worked per week than men, within the same specialties. Hypothesis 9 . Women physicians have a larger variance in total patient care hours worked per week than do men, within the same specialties. For each of the 4 AMA broad specialty categories (Appendix I), with total hours per week as the dependent variable, analysis of variance will test the means and equality of the variances of the dependent variable for the two categories of the independent variable (women and men). 113 The t-test is the test of significance, as homoscedasticity cannot be assumed. Hypothesis 10. Younger (born after 1945) women and men physicians have the same mean total hours of patient care. Younger physicians (born after 1945) will be selected to test this hypothesis. gender, The independent variable is and the dependent variable is total hours worked per week per physician. The effect of gender and age can be considered additive for USA physicians, based on the recent increase in the number and proportion of women physicians. Therefore the mean and modal age of women physicians may be less (Table A-4) than that of men physicians. The statistical analysis for the effect of gender on the total hours worked per week per younger physician will be by the t-test. Hypothesis 11. More recent medical school graduates (after 1970) women and men physicians have the same mean total hours of patient care. Recent (after 1970) medical school graduates (before 1971) will be selected to test this hypothesis. The independent variable is gender and the dependent variable is total hours worked per week per physician. The effect of gender and year of graduation from medical school can also be considered additive for USA physicians, due to the recent increase in the numbers and proportion of women. The mean 114 and modal years postgraduate of women physicians may be less (Table A-l and A-2) than that of men physicians. The statistical analysis for the effect of gender on the total hours worked per week per recent medical school graduate will be by the t-test. Hypothesis 12. Younger (born after 1945) women and men physicians have the same mean total hours of patient care, within the same specialties. Younger (born after 1945) physicians will be selected to test this hypothesis. gender. The independent variable is The dependent variable is the intervally scaled total hours worked per week per physician in each of AMA's 4 broad specialty categories (Appendix I). The statistical analysis for the effect of gender on the total hours worked per week per younger physician for each of AMA's 4 broad categories of specialty, will be by the t-test. Hypothesis 13. More recent medical school graduates (after 1970) women and men physicians have the same mean total hours of patient care, within the same specialties. Physicians graduated from medical school after 1971 will be selected to test this hypothesis. variable is gender. The independent The dependent variable is the intervally scaled total hours worked per week per physician in each of AMA's 4 broad specialty categories (Appendix I). The statistical analysis for the effect of gender on the 115 total hours worked per week per recent medical school graduated for each of AMA's 4 broad categories of specialty, will be by the t-test. Hypothesis 14. Younger (born after 1945) women and men physicians have the same percentage reported as professionally inactive. All those physicians in the younger age group (born after 1945) will be selected to test this hypothesis. One variable is gender and the other variable is professional activity. Covariation of reported professional inactivity and gender for younger physicians will be tested by Pearson's chi square. Hypothesis 15. More recent medical school graduates (after 1970) women and men physicians have the same percentage reported as professionally inactive. All those physicians graduated from medical school after 1970 will be selected to test this hypothesis. One variable is gender and the other variable is professional activity. Covariation of reported professional inactivity with gender for recent medical school graduates will be tested by Pearson's chi square. Hypothesis 16. Older (born before 1946) women physicians have a higher percentage reported as professionally inactive than older (born before 1946) men. 116 All those physicians in the older category will be selected to test this hypothesis. One variable is gender and the other variable is professional inactivity. Covariation of reported professional inactivity with gender for older physicians will be tested by Pearson's chi square. Hypothesis 17. Earlier women medical school graduates (before 1971) have a higher percentage reported as professionally inactive than earlier men medical school graduates. All physicians graduated from medical school before 1971 will be selected to test this hypothesis. One variable is gender and the other variable is professional inactivity. Covariation of professional inactivity with gender for early medical school graduates will be tested by Pearson's chi square. Hypothesis 18. Women physicians report patient care less often than do men physicians. One variable is gender and the other variable is the single category of patient care on the nominal scale of professional activity. Covariation of patient care with gender will be measured by Pearson's chi square. Hypothesis 19. Women physicians report inpatient hospital care as the sole professional activity more often than do men. 117 One variable is gender and the other variable is the single category of inpatient hospital care as the practice type, without any nonhospital professional activity on the nominal scale of professional activity. Covariation of inpatient hospital care with gender will be tested by Pearson's chi square. Hypothesis 20. Women physicians report a lower mean of inpatient hospital care hours worked per week than men, when reporting both hospital and nonhospital practice. Physicians reporting both hospital and nonhospital practice types will be selected for this hypothesis. One variable is gender and the other variable is the total number of inpatient hours reported worked per week. The relationship between gender and inpatient care can be treated as additive. This is based on the two factors within women physicians' practice, that they are less likely to be in a specialty that includes inpatient practice, unless inpatient hospital practice is their sole activity. In addition, women physicians tend to a shorter work week than men. The t-test will test gender differences in total hours per week reported in inpatient hospital care, when the physician practices both in a hospital and elsewhere. Hypothesis 21. Women physicians report administration as a principal professional activity less than do men. 118 One variable is gender and the other variable is the category of administration reported on the nominal scale of principal specialty (reported or ascribed). Covariation of administration with gender will be tested by Pearson's chi square. Hypothesis 22. Women physicians report a lower mean total hours worked per week in administration than do men. Physicians reporting hours in administration were selected to test this hypothesis. One variable was gender and the other variable was the total number of hours reported worked per week. The relationship between gender and administration can be treated as additive, based on two factors within women physicians' practice. They have been less likely to be in a specialty that was commonly recruited for administration, and that women physicians tended to work to a shorter number of hours per week than men. The t-test will test gender differences in total hours per week reported in administration. Hypothesis 23. Women physicians reportteaching as a professional activity more often than do men. One variable is gender and the other variable is the number of physicians reporting teaching as their principal specialty or secondary specialty. Covariation of teaching with gender will be tested by Pearson's chi square. 119 Hypothesis 24. Women and men physicians report the same mean total hours worked per week teaching, when it is the sole professional activity reported. Physicians reporting teaching as their principal specialty will be selected to test this hypothesis. One variable is gender and the other variable is the total number of hours reported teaching per week. Therefore the t-test will test gender differences in total hours per week reported in teaching. Hypothesis 2 5 . Women physicians report research as a professional activity less than do men. One variable is gender and the other variable is physicians reporting research activity as their principal or secondary specialty. Covariation of research with gender will be tested by Pearson's chi square. Hypothesis 26. Women physicians report fewer total hours worked per week in research than do men. Physicians reporting research as principal or secondary specialty will be selected to test this hypothesis. One variable is gender and the other is the total number of hours reported worked in research per week. The relationship between gender and research care can be treated as additive, based on two factors within women physicians' practice. They have been less likely to be in research and historically have worked fewer hours per week 120 than men. Therefore the t-test will test differences in total hours per week reported in research. Hypothesis 27. Women and men physicians are in graduate medical education training at the same rate. One variable is gender and the other variable is the single category of current training reported as professional activity. Covariation of current training with gender will be tested by Pearson's chi square. Hypothesis 28. Women are less likely to be training at the fellowship level than are men physicians. One variable is gender and the other variable is the single category of fellowship on the category of current training as professional activity. Covariation of fellowship training with gender will be tested by Pearson's chi square. Hypothesis 29. Women physicians are less broadly distributed among the counties in Michigan than men. One variable is gender and the other variable is county location. Covariation of county with gender will be tested by Pearson's chi square. Hypothesis 30. Women physicians are more likely to practice in metropolitan areas than are men. 121 Hypothesis 31. Men physicians are more likely to practice in rural areas than are women. The counties in Michigan have been classified in Appendix B as metropolitan, rural adjacent to metropolitan, or rural not adjacent to metropolitan counties ("metropolitan, rural adjacent, or rural"). variable and gender is the other variable. This is one Covariation of gender and location will be tested using Pearson's chi square. Hypothesis 32. Women primary care physicians are more likely to practice in rural areas than are women specialists. Selecting all women physicians, one variable is the classification of principal specialty as primary or nonprimary. The other variable is the scale of metropolitan, rural adjacent, or rural county location (Appendix B ) . Covariation of specialty and location for women physicians will be tested using Pearson's chi square. Hypothesis 33. Women primary care physicians are more likely to practice in metropolitan areas than are men primary care physicians. Selecting all primary care physicians, gender is one variable and the other variable is the metropolitan category of county location. Covariation of gender and location for 122 primary care physicians will be tested using Pearson's chi square. Hypothesis 34. Younger (born after 1945) women physician have a broader distribution between metropolitan counties, rural counties adjacent to a metropolitan county, and rural counties that are not adjacent to a metropolitan county, than older (born before 1946) women physicians. Selecting all women physicians, one variable is the scale of younger (born after 1945) and older (born before 1946) physicians. The other variable is the classification of counties into metropolitan, rural adjacent, and rural counties (Appendix B ) . The covariation between geographical location and age of women physicians will be tested by Pearson's chi square. Hypothesis 35. More recent medical school graduates (after 1970) women physician have a broader distribution between metropolitan counties, rural counties adjacent to a metropolitan county, and rural counties that are not adjacent to a metropolitan county, than early women medical school graduates (before 1971). Selecting all women physicians, one variable is the scale of those graduated from medical school before 1971 or after 1970. The other variable is the classification of metropolitan, rural adjacent and rural counties (Appendix B). The covariation between geographical location of women 123 physicians and their year of medical school graduation will be tested by Pearson's chi square. Hypothesis 36. There are more USA trained women MD physicians among all USA trained MD physicians, than there are women DO physicians among all DO physicians. Proportionately more women have been admitted to allopathic medical schools than to osteopathic medical schools in the USA (AMA, 1987; AOA, 1987). As DOs are only trained as physicians in the USA, only USA trained MDs will be selected to test this hypothesis. variable. Gender is the other Covariation of gender with professional degree will be tested by Pearson's chi square. Hypothesis 37. DO women physicians are more likely to be in primary care than USA trained women MD physicians. DOs are more likely to be in primary care, and the specialty opportunities for women in osteopathy have expanded medicine. less rapidly than for women in allopathic All USA trained women physicians are to be selected with one variable being professional degree (MD or DO). The other variable is the classification of specialties as either primary care or nonprimary. Covariation of specialty with professional degree for women will be tested by Pearson's chi square. 124 Hypothesis 38. There are more internationally trained women among all women MD physicians, than internationally trained men among all men MD physicians. Only MDs will be selected to test this hypothesis because DOs trained in other countries are not trained as physicians. One variable is gender and the other variable will be nation of training, recoded to USA, Canada or international. As Michigan is a border state with Canada, there may be more Canadian trained physicians in Michigan than in other USA states. The covariation of gender with nation of training will be tested by Pearson's chi square. Hypothesis 39. Internationally trained women MD physicians are more likely to be in primary care than USA trained women MD physicians. After selecting all women MD physicians, one variable is the nation of training, USA or internationally trained MD physicians. The other variable is the classification of principal specialty into primary or nonprimary specialties. The covariation of specialty with nation of training of women physicians will be tested by Pearson's chi square. Hypothesis 40 . Women physicians are younger on the average than men physicians. Gender and year of birth were tested for differences by the t-test. 125 Hypothesis 41. Women physicians are more recent medical school graduates on the average than men physicians. Gender and year of medical school graduation were tested for differences by the t-test. Hypothesis 42. Women in nonprimary specialties are younger than women physicians in the primary specialties. Selecting all women physicians, the variable describing principal specialty as primary or nonprimary, and year of birth, were tested for differences by the t-test. Hypothesis 43. Women physicians are skewed to the younger age groups in specialty care, more than are men in specialty care. All specialty care (nonprimary) physicians will be selected from the classification of principal specialty as primary or nonprimary. Differences in the mean year of birth for gender will be tested with the t-test. Hypothesis 44 . Women physicians in specialty care are skewed to the more recent medical school graduation years, more than men in specialty care. All specialty care (nonprimary) physicians will be selected from the classification of principal specialty as primary or nonprimary. Differences in the mean year of medical school graduation by gender will be tested with the t-test. 126 Hypothesis 45. Certain specialties are more likely to be reported by physicians whose principal activity is research, for both men and women. An exploratory hypothesis, this will provide descriptive information on the specialties of women and men in research, and any gender differences. All physicians reporting research activity on the nominal scales of principal and secondary specialty will be selected. The specialty (from AMA's 85 categories, in Appendix G) and percentage of each gender of the physicians in each specialty will be determined. Summary The impetus for this study came from the rapid increase in the enrollment of women in USA medical schools since 1970. There is emerging data showing that as well as being relatively more numerous, women physicians are practising differently from their historical professional activities. In addition, men are changing some aspects of their professional activities. The impact of the gender and practice changes are studied for the full population of physicians in Michigan, both MDs and DOs. The secondary analysis of data from three statewide surveys in Michigan in 1986 will provide the database for this study of the professional activities of women and men physicians, generalizable to the entire population of physicians in Michigan in 1986. The research design was 127 primarily single measurement cross section, with temporal and comparative analysis. It would take 13 months to create the final database from state and national sources. Analysis was restricted to predominately descriptive statistics by the primarily nominal scales in the final data set. IV RESULTS OF THE ANALYSIS Acquisition of the Data It took six steps over a period of thirteen months to establish the data base, commencing in April 1990. Each institution's manager of data services gave verbal permission to release data, namely from the Office of Health and Medical Affairs (OHMA), the American Medical Association (AMA), and the American Osteopathic Association (AOA). During the data acquisition process, the Executive Branch of the State of Michigan changed in the 1990 election. Early in 1991, the new Governor eliminated the Office of Health and Medical Affairs with some of its functions transferred to the Michigan Department of Public Health, including the annual physician surveys. Closure of the Department of Licensing and Regulation was announced, with its functions, including the Board of Medicine and the Board of Osteopathic Medicine to be transferred to the Michigan Department of Commerce. Within the latter, the Rural Development Office of the Local Development Services (Appendix B) was slated for closure. Access to key individuals in these state offices was increasingly marked by delays. 128 129 Computer Tape Specification A standard computer tape setup was requested to conform with the recommendations of the Michigan State University (MSU) Computer Center Statistician and system analyst for the following format: EBCDIC, 6250 bpi, 9 Track, no label, and a fixed block format. The OHMA, AMA and AOA researchers specified the logical record length and the blocking factor when returning their respective data tapes. Office of Health and Medical Affairs Initial Data First, this researcher gave a new blank computer tape to the chief researcher at bOHMA in Lansing. Data was copied on 28,395 physicians (24,355 MDs and 4,040 DOs), originally from the Michigan Department of Licensing and Regulation (L&R), under which the Board of Medicine and the Board of Osteopathic Medicine have operated. Note the high proportion of DOs licensed in Michigan (14.2%), the highest in the USA (AOA, 1987; USDHHS, 1987). They were the same records used by OHMA for their 1986 Physician Survey. The variables were: license number, license status in 1986, 1987 license type, their name and address, including county, in aDr. Randall P. Fotiu, Ph.D., Statistician/System Analyst, Computer Laboratory, 409C Computer Center, Michigan State University, East Lansing, MI 48824-1042. (517) 353-1800. bMr. Kenneth J. Darga, M.A., M.S.W., Demographer, Demographic Research and Statistics, Department of Management and Budget, Lansing, Michigan 48909. (517) 373-9654. Formerly, Special Studies and Analysis, Office of Health and Medical Affairs, Department of Management and Budget. 130 both 1987 and 1989, and the date of their original Michigan license. OHMA separated MD and DO records, and within the professions, physicians were separated by their recorded state of residence to make four files. However, 4.7% of the out of state MDs had a Michigan county ascribed to them, and conversely 2.5% of the instate MDs had an out of state county ascribed to them. This probably reflected a previous address change but an unchanged county code at the time of the release of the data by L&R to OHMA (K. J. Darga, personal communication, January 18, 1991). There was a total of 21,040 instate MDs in the two files. Because a 1986 AMA instate record may have existed corresponding to the physician's Michigan county code in the L&R files, all records were retained to attempt a match with AMA data. There was a total of 2,746 instate DOs in the two L&R files. Because an AOA record may have existed for every DO, all records were retained to attempt a match with AOA data. American Medical Association Data It was intended that the tape would next be forwarded to aAMA in Chicago, and the requested study variables would be matched for MD physicians. However the AMA advised that it would take less time to reproduce an archived data tape aMs. Gene Robak, Director, Department of Physician Data Services, American Medical Association, 515 North State Street, Chicago, Illinois 60610. (312) 464-5181. 131 of MD physicians who were in Michigan in 1986, and include in it the requested variables for this researcher to perforin the matching. The first tape received from AMA did not meet specifications, and their next, usable tape came within one week of the return of the first tape. The AMA 1986 file of MDs contained 17,789 physicians who had been in Michigan on December 31, 1986, of whom 16% were women. Matching L&R and AMA Data. The L&R and AMA files were matched by name and address as the third step. The key variables for matching were the first 20 characters of the last name, first name initial, second name initial, any name suffix, and county of residence in 1986. The two files of MDs were concatenated to create one MD file. Each of the L&R and AMA files were reconstructed in the last name field to delete any spaces within names. Finally each of the L&R and AMA files were sorted alphabetically and read into system files and archived. The order of the sort was by last name, first initial, second initial, suffix if junior (as few in the AMA files were senior), and Michigan county of license. L&R records with out of Michigan counties were excluded, as only instate records came from the AMA, leaving 21,040 instate L&R MD records. There were two names with duplicate, indistinguishable records in each file that were excluded from matching. 132 Merger of the AMA and L&R files produced a total of 12,214 records that matched exactly with the AMA files, and 5,575 AMA unmatched records and 8,826 L&R unmatched records (Table 5). Matching by hand could not increase the computer matched total as the criteria for matching the two files still applied. Verification of the match was by reviewing the first fifty records in the combined and source files. All 26,613 records (L&R-AMA matched name data set, L&R-only, and AMA-only) were written to a disk file. American Osteopathic Association Data Fourth, the aAOA matched L&R records by name between the OHMA provided data tape with their AOA Physician Masterfile for 4,009 of the 4,040 L&R physicians. This process included hand matching subsequent to electronic matching (M. Wallis, personal communication, January 16, 1991). After generating a second readable tape AOA provided the requested variables: gender, year of birth, medical school year of medical school graduation, and self­ identified specialty. AOA did not have a historical file archived for 1986, but had commenced in 1989 to retain historical files. Therefore, at this researcher's request, OHMA had entered both the 1989 and 1986 L&R addresses for the DOs in Michigan. AOA then could match first from L&R to their aMr. Michael Wallis, M.P.H., Director, Department of Physician Data Services, American Osteopathic Association, 212 East Ohio Street, Chicago, Illinois 60611. (312) 280-5806. 133 archived 1989 Michigan file, and match the remainder of the L&R DOs from their national file. The AOA generated file of DOs contained 4,009 physician records of whom 12.6% were women and 87.4% were men. Matching L&R and AOA Data. As the fifth step, the entries were matched by name and address, between the two concatenated L&R DO files and the AOA file. Each of the L&R and AOA files were reconstructed as for the MDs. The order of the sort was by last name, first initial, second initial, and suffix if junior, but without county code as it was reported from different years. The 3,570 records that matched electronically were written to a disk file on the MSU IBM 3090, as part of the matched name data set. Matching by hand could not increase the total as the same criteria for matching the two files still applied. The rate of matching DOs was higher than for MDs due to the AOA data having first commenced with the OHMA-supplied files as originally planned. Only matched records could potentially became part of the final study group. All records were retained (3,570 L&R-AOA matched name data set, 470 L&R-only, and 439 AOA-only) in disk files. Verification Verification of the match was by reviewing the first one hundred records in each of the files. Careful checking of the matched name data set was conducted for consistency across all variables. Subtotals were reviewed within each 134 variable, across each original data set (L&R, AMA, AOA), and with the final matched name data set. All variables had been matched and saved consistently. OHMA Survey Data The sixth and final step was to return the matched name data set on tape for the addition of 1986 survey data by OHMA for each physician. The variables to be copied from the OHMA file for each physician record were specified. The tape of 15,770 matched physicians, now only identified by their license number, was hand delivered to the OHMA chief researcher for addition of their survey responses. Retention of the license number to be the identification for each record was requested, but rejected by OHMA in preference to generation of a random number identifier for each physician. The license number would have permitted anonymous verification of records, and longitudinal follow up study of individuals using it as a permanent identifier. OHMA archived the match between each license number and random identifier with the Michigan Department of Management and Budget to permit retrieval of the license number for matching 1986 responses with subsequent state physician surveys. Final Data Base The final data base is outlined in Tables J-l and J-2. The geographical location data was separated for each physician for each separate location of that physician, creating multiple records for each physician in Table J-2. 135 Table 5 Outcome of the Match Between L&Rf AMA. AOA and OHMA Records Data Source Total Records # Records Matched % Records Matched % Loss MDs All L&R 24 355 12 214 50 % 50 % L&R Instate 21 040 12 214 58 % 42 % AMA Instate 17 789 12 214 69 % 31 % 12 200 99.9 % 0.1 % OHMA Matched DOs All L&R 4 040 3 570 88 % 12 % L&R Instate 2 746 2 653 97 % 3 % All AOA 4 009 3 570 89 % 11 % 3 570 100 % 0 OHMA Matched Total Total L&R 28 395 15 770 56 % 44 % Total Instate 23 786 14 867 63 % 37 % Note; L&R is the Michigan Department of Licensing and Regulation; AMA is the American Medical Association; AOA is the American Osteopathic Association; OHMA is the (former) Office of Health and Medical Affairs of the Michigan Department of Management and Budget. 136 The Study Group The study group constituted 15,770 physicians for all of whom L&R and AMA (12,200) or AOA (3,570) records had matched, and an OHMA survey response was obtained (Table 5). The term study group is the reference term in this research for the physicians whose survey responses were the basis for the analysis and discussion. The loss of physicians from the full list of 23,786 physicians licensed and practising in Michigan in 1986 (21,040 MDs and 2,746 DOs) equalled 34%. OHMA provided a 1986 Physician Survey response for 99.9% of the MDs and 100% of the DOs in the matched name data set, and then deleted physician identifying data. The final tape containing the completed data base was copied to a disk file on MSU's IBM 3090 mainframe. A complete archive of key files remains stored at Michigan State University for follow up research. Data was provided by OHMA in two files (Appendix J ) . One file contained unique data for each physician in a single record. All physician-unique data was combined to create an enlarged first file, which was the final Table J1. The second file was all the data for each practice location, up to eight records per physician, with the same unique identification attached to each file (Table J-2). One conceptual problem encountered by this splitting of data by location, was that the individual physician was the unit of analysis for this study. The resolution was to select 137 the "principal location" defined as that practice location for each physician at which the highest number of hours was reported. Questions regarding geographical location were analyzed only using the so-defined principal location for each physician. Distribution of the Variables of the Study Group Professional Degree and Gender See Table A-ll and Figure 8. There were 15,770 physicians in the study, of which 12.5% were women and 87.5% were men, and 77% were MDs and 23% were DOs. This distribution compares with the 1987 state percentages that 84% of MDs were men and 16% of MDs were women in Michigan in 1985, and that 83% of practising physicians in Michigan were MDs and 17% were DOs (USDHHS, 1987), and that 14% of the physicians licensed by Michigan in 1986 were DOs. As already noted, DOs were matched with AOA data at a higher rate than for MDs, contributing to their slight over­ representation in this study group. Nationwide DOs were 4.2% of active physicians (COGME, 1989). Among the DOs, 11% were women and 89% were men, and while there were no published national figures for comparison, the AOA tape included 13% women, and so some were lost in the match. Among the MDs, 13% were women and 87% were men, comparing with the national figures of 15% and 85%, respectively (AMA, 1987; Table 1), and the AMA tape of 16% women and 84% men. Among women 81% were MDs, and 19% 138 were DOs. The study group was considered representative of the medical workforce in Michigan in 1986. Degree MDs 77% DOs 23% Gen d e r Women 12.5% Men 87.5% N a t i o n of T r a i n i n g USA 81% International 19% Specialty Primary 40% Nonprim ary 60% L ocation Metropolitan 90% Figure 8 . Rural 10% Characteristics of the Study Group 139 Nation of Training See Table A-ll and Figure 8. Eighty one percent of the physicians were USA trained, including DOs, and 19% were internationally trained MDs. Of the total MDs in the study group 74.6% were trained in medical schools in the USA (including Puerto Rico), 11% being women and 89% men. Of the MDs 25.4% were trained in international schools, of whom 18% were women and 82% were men, a significant difference for gender in the nation of training (p=.00). This was comparable with the 1985 figure that 27% of all MD physicians in Michigan trained in an international school (USDHHS, 1987) along with other studies (D'Elia & I. Johnson, 1980; A. Goldblatt & P. B. Goldblatt, 1976; Mejia et al, 1980). Thirty five percent of all women MDs in the study group were international graduates, virtually unchanged from 1971 when 36% of all women MDs were international graduates (Pennell & Renshaw, 1973). The five most frequent nations of training for MDs outside of the USA were India (5.9%), the Philippines (3.0%), Korea (1.6%), Iran (0.9%) and Pakistan (0.8%). Women and men who trained in international schools were distributed similarly among the nations. Rank ordered from the highest, women were more likely to have trained in India, the Philippines, Korea, with Pakistan and West Germany as equal fourth. Men were more likely to have trained in India, the Philippines, Korea, Iran, with Mexico and Pakistan as equal fifth. 140 Specialty The most frequent principal specialty reported by women and men physicians in Michigan in 1986 (Table A-12) were family and general and family practice (20.3%), internal medicine (13.1%), general surgery (5.8%), obstetrics and gynecology (5.4%), and psychiatry (4.9%). This was summarized in Figure 8 and Table 8 as 40.2% in primary care specialties and 59.8% in nonprimary specialties. Women were in most specialties, and reported most commonly (Table A-12) family and general practice, internal medicine, pediatrics, psychiatry and obstetrics and gynecology, totalling 42.8% in primary care. This was comparable with the 1986 national distribution of 40.8% of women MDs in primary care (Tables A-5, A-6 and A-7). Men reported most commonly as their principal specialties (Table A-12) general and family practice, internal medicine, general surgery, obstetrics and gynecology, and psychiatry. Of all men, 39.9% were in primary care comparable with 37.6% of MDs in primary care in 1986 nationwide (Table A-5). Secondary specialties of women and men were reported by 16.2% of the study group (Table A-13), or by 12.9% of women and 16.4% of the men (p=.00). Practice Type See Table A-14 and Figure 8 for the practice types reported by active physicians. For this analysis, any physician reporting zero patient care hours was excluded, 141 both those professionally active but not in patient care, and those completely inactive. of 2 practice locations. Physicians reported a median Among those in patient care, hospital-only practice was reported by 31%, practice solely outside of a hospital by 10%, while 59% of active physicians reported a mixture of practice in and out of hospitals. Physicians who were active in Michigan but did no patient care were 1.9% of the total (Table 6). Full time physicians reported primarily (68%) practice in and outside of hospitals. Practice types have been the most studies factor for gender differences among physicians (Custer & Dimon, 1987; Ellsbury et al, 1987; (Heins, 1985; Heins, Smock, Jacobs, & Stein, 1976; Heins, Smock, & Martindale, 1978; Heins, Smock, Martindale, Jacobs, & Stein, 1977; Heins, Smock, Martindale, Stein, & Jacobs, 1977; Maheux et al, 1988, 1989). Women physicians reported some differences in practice types from men (see Table A-14). Women physicians reported a median of l practice location with a range from 1 to 6, and 8 (none with 7) locations. Men reported a median of 2 practice locations, and had the full range of 1 to 8 locations. Among practice types, 44% of women reported only hospital practice significantly more than 28% of men, (p=.02). Women with a nonhospital practice, 12%, were not significantly different from 10% of men, (p=.33). Forty- four percent of women practiced both in and out of hospitals, significantly less than 62% of men (p=.00). See 142 Table A-3, Davidson, 1979, and Heins et al, 1978 for evidence that this was both the historic pattern, and the expectation of practice for women physicians. Full time women practiced mainly (57%) in and out of hospitals, and part time women reported more hospital only practice (62%). Nationwide in 1986, 35% of all nonfederal women physicians reported a hospital practice (Table 3), compared with 19.2% of all nonfederal men physicians reporting a hospital-only practice. Active women in Michigan who did no patient care were no less than men. Office-only practice tended to be a high proportion of part time physicians, but not significantly different for women and men (p=.45). This was re-analyzed to exclude women and men who were in graduate medical education in Michigan in teaching hospitals (Table A-14). A disproportionate share of women physicians were in training at any one time due to their relatively recent entry into medical education. Exclusion of these trainees captured the practice profile of fully trained physicians. The outcome after exclusion of physicians in training was that both full time women and men reported less hospital only practice. Professional Activity Status The most frequent physician activity for both women and men, was patient care, 83.5% of the total, broken down to 72.3% full time and 11.2% part time practice (Table 6 and Figure 9). Those who were active, but spent no time in 143 patient care were 1.9% of the total reporting activity status, and 10.4% were professionally inactive or retired. Table 6 Activity Status bv Gender Professional Activity Women Men Total aPatient Care 82.9 83.6 83.5 bFull Time 66.4 73.1 72.3 bPart Time 16.5 10.5 11.2 (n = 2) (n = 1) * 1.8 1.9 1.9 10.1 10.4 10.4 4.6 4.2 4.2 Total n 1 966 13 804 15 770 % 12.5 87.5 100 Locum Tenens Active (no patient care) Clnactive or retired Other ap = .38, patient care not different for women and men. bp=.00, significantly more women were part time, and more men were full time. cp=.70, inactivity not significantly different for women and men. Activity status of women was most frequently in patient care, which was not significantly different from men (p=.38). Significantly fewer women reported full time 144 patient care and more worked part time (p=.10), than did men (Figure 9). E H Part Time □ Full Time Women 12.6 % Men 87.4% Figure 9 . Part or Full Time Patient Care by Physician Gender 145 10% Inactive ! 6% Nonpatient Office Hospital Office + Hospital Ill 8% 26% 1 49% Practice Type Figure 10. 7% 3% Rural Adjacent Rural 90% Metro­ politan Location Activities and Location of the Study Group. 146 OHMA's definition of inactive physicians was broad (Appendix D) , including those temporary and permanently inactive and retired, and those who had left Michigan. Women were professionally inactive or retired at the same rate as men (p=.70). This compares with Tables 3 and 4, showing women MDs' national inactivity level in 1986 as 7.4% and men's level as 8.4%. Women have brief periods of inactivity, mostly for child bearing and rearing (Callan & Klipstein, 1981; Ducker, 1986; Ferrier & Woodward, 1982; Heins, Smock, Martindale, Jacob & Stein, 1977). The lower mean age of the women in this study group, 38.8 years, was significantly less (p=.00) than the mean of 46.2 years for men. The most frequent activity reported by men was patient care (Table 6, 83.6%), of whom 87.4% were full time and 12.6% part time (Figure 9). Graduate Medical Education The training status of women and men cams from either physicians' survey response if fully licensed, or was gathered from the educational limited license data, the first practice year license in 1986 from the Department of Licensing and Regulation (Table A-15). Hence none of the first year MD residents and the DO interns were surveyed directly, but were included in the limited license category. Of all physicians in the study group 15.3% were in graduate medical education training programs in 1986. 147 Women were disproportionately more in graduate medical education, 26.1% of all women, 1.9 times higher than the 13.8% rate of men. Another summary figure was that 41% of all active women physicians were in training, 2.3 times higher than the 18% of all active men physicians that were training in 1986. Of these women in training, 29.4% were known to be in second or subsequent years of residency and 6.4% were in fellowships. Nationally, of the MD women who were in training in 1986 (Table 2) 8% were in fellowship programs. Historically women have progressed to fellowships less than men (Kohrman et al, 1989; A. P. Williams et al, 1990). Men in fellowships were 9.8% of the total of men in training. Nationally, of the MD men who were in training in 1986 (Table 2) 10% were in fellowship programs. Women in second year residencies or fellowships were more often in part time programs (9%) than men (5%), significant at the 10% level (p=.10). As noted, women were almost twice as likely as men to be still in graduate medical education. Hence the decision was made to separate physicians in training from fully trained physicians for some of the research questions. Women MDs and DOs in training in 1986 were 41% more in more specialties than women physicians in practice in 1986 (p=.00). Women MDs in training were significantly more in nonprimary specialties (p=.00) than those already in practice (Tables A-5, A-6, A-7, A-8 and A-16). 148 Men in training in 1986 were also more in nonprimary specialties, 26% more than those in practice in 1986 in Michigan, (81% to 64%) as seen in Table A-16, and previously in Tables A-5. Women were closing the specialty gap. Physicians who were in graduate medical education were almost exclusively (99%) in metropolitan areas (footnote, Table A-16). The dominance of women in graduate medical education could have skewed their representation in metropolitan areas. For analyses of geographical location trainees would be excluded to test the ultimate location of physicians, after completion of graduate medical education. Geographical Location Principal physician practice location was defined as the zip code of their highest hour entry on the survey card, recoded to county (Appendix B). This was also the first entry for 80% of the physicians, confirming the OHMA intention and survey instructions that the first entry be the principal location. See Table A-17 for the geographical location (population based) of women and men by their degree. Only active physicians were included in the analysis, as the focus of the study was professional activities. Metropolitan settings were the most (90.4%) common (Fulop, 1986a; Maheux, Dufort & Beland, 1988; Maheux, Dufort, Lambert & Berthiaume, 1988; S. C. Martin, et al, 1988; Woodward, 1986). 149 Women physicians' principal practice locations were even more commonly, 94%, in metropolitan settings (p=.00) than men (90%). Internationally trained physicians' principal practice locations were significantly more in metropolitan settings, 93.4%, than USA MDs, 89.8%, (p=.00). DOs were significantly more in rural areas than MDs, 88.6% to 90.8% (p=.00). Women have been practising more in rural areas than they had historically (Bowman & Gross, 1986; D'Elia & Johnson, 1980; Ellsbury et al, 1987; MooreWest et al, 1982). Another definition of geographical location was to classify counties according to both their population and their contiguity to metropolitan counties (Appendix B ) . They were coded in Table C-2 as metropolitan, rural adjacent to metropolitan, or rural not adjacent to metropolitan counties ("metropolitan, rural adjacent, or rural"). See Figure 10 and Table A-18. Physicians' locations have been described using the population based model in Tables A-17 and C-l. Given the preponderance of metropolitan location for all physicians and even more so for women physicians, the second (Tables A18 and C-2) classification may have been more sensitive to dispersion of physicians. It may provide information for recruitment and planning strategies for rural and rural adjacent to metropolitan counties, especially if contiguity to a metropolitan area was an inducement for a physician to locate in a community. 150 Women physicians' principal practice locations were distributed more evenly between the rural adjacent and rural counties (Table A-18) than the population based model (Table-17) illustrated. Table A-19 simplified the description of location and the discussion to metropolitan or rural (using the population based model). Analysis by years since medical school graduation found no significant difference in the location of early or more recent graduates (p=.00). Location patterns appeared to be not changing. Recent DO graduates (after 1970) were more in metropolitan areas than earlier DO graduates, but there was no significant difference between women and men (see Table A-19). However 91% of all women DOs were recent (after 1970) graduates, as their enrollment increases have lagged behind MDs (Table A2), so the early/recent division of graduation years was less meaningful or utilitarian for describing changes among DO w om en n h v s i n i a n-s A. ~------------ -------- Hours Worked Physicians reported patient care hours. For all analyses of hours, those reporting none were excluded, in keeping with the methodological concerns in chapter II. Further, as some physicians had reported more than 168 hours per week when adding up all their sites, values over 80 were excluded as outliers. This was selected as the criterion because 80 hours per week was the recommended maximum for residents by the Accreditation Council on Graduate Medical 151 Education (1990, 1991). Eighty hours was verified as the most common policy maximum for residencies operating in Michigan in 1990 (Michigan Association of Medical Education, 1991). See Table A-20 for the hours of patient care reported by women and men and Table 7 for all work hour categories. Physicians practiced an average of 52.6 total hours in patient care per week in all their practice locations. Research has reported varying degrees of dissimilarity of hours worked by women and men. Silberger et al (1987), Curry (1983) and Heins' group (1976, 1977, 1978, 1985) found that hours were almost the same. Bobula (1980a, 1980b), Mitchell (1984) and Callan and Klipstein (1981) found that hours were less for women. Cartwright (1977) found that women physicians' hours varied according to their degree of continuous professional activity. Freiman and Marder (1984) noted that between 1970 and 1980 the male week was shortening faster than the female week was lengthening, but that both processes were occurring. Women physicians reported (Table 7) an average of 44 total patient care hours worked per week, 9.6% and 4.7 hours, less than men (p=.00) with 48.7 hours (Bowman & M. L. Gross, 1986; Freiman & Marder, 1984; M. J. Lanska, 1984; W. B. Schwartz et al, 1988). 152 Women ( I Men 44h A l l Patient Care 48.7h F u l l T i m e Care 59 Oh F ull Time T raining 50.8h Figure 11. Hours Per Week Of The Study Group 153 A subsequent analysis separated part and full time physicians (Hojat et al, 1990) and excluded those in graduate medical education (Figure 11, Table A-20). Full time women's mean hours in each activity increased, to 6.6% and 3.5 hours less than men's (49.4 to 52.9, respectively). The impact of part time work as a temporary work status for women was analyzed in research questions examining gender differences in hours worked. When physicians in training were separated, as well as examining full and part time physicians in training, women and men in full time training worked almost the same hours (two-tailed significance of p=.18, Table 7). This was another indicator of the gap closing in hours worked for gender. Trainees were in a salaried situation and do not control their schedules to the same extent as independent physicians. However, all active full time women physicians were only 5.4% and 2.9 hours less than men, 50.4 to 53.3 hours respectively (Table 7). This illustrated the trend of women and men to work similar work weeks (Figure 11). 154 Table 7 Activity and Hours Worked bv Gender Hours Women Total Men All Full Time All Full Time All Full Time 1 573 985 11 074 8 753 12 847 9 738 82.9 66.4 83.6 73.1 83.5 72.3 44.0 15.1 49.4 11.5 48.7 15.7 52.9 11.5 48.2 15.7 52.6 11.6 Administration Report n= 0 0 17 17 17 17 Teaching Report n= 0 0 5 5 5 5 Research Report n= 0 0 5 1 5 1 aCurrent Training % Report 26.1 b23.8 13.8 b13.0 15.3 14.4 n= aPatient Care % Report M Hours SD M Hours SD 54.8 18.6 c57.8 15.4 57.5 16.4 c59.8 13.8 a57.1 16.8 a59.5 14.1 aAll Active M Hours 4b. 1 50.4 49.3 53.3 49.0 53.0 ap=.00. bp=.ll for reporting full time graduate medical education for gender. cTwo-tail p=.18 for full time women and men in graduate medical education. 155 Full T i m e 53 3 P Women 50.4 Part Time 26. d 23.5 Figure 12. Part and Full Time Practice Hours, by Gender 156 Women worked their longest hours in office practice (Tables A-2 0). Women and men had comparable hours when in office only practice (p=.46). There were significant gender differences in hours spent in hospital-only and hospital combined with office practice (p=00.). Within the latter, again there was no difference in office hours. Men worked more hours in the hospital portion of a mixed practice, thus creating the hour gap. The contribution of specialty to hospital hours was analyzed later. Years of Birth and Graduation Literature has reported most time based comparisons by physicians' age (D'Elia & I. Johnson, 1980; S. R. Kaplan, 1981, 1982). A more sensitive indicator of the influences of training on a physician was the number of years since medical school graduation. Retroactive studies have used graduation dates to divide their subjects (L. M. Arnold et al, 1981; Beil, et al, 1980; Bergquist et al, 1985; Bonar et al, 1982; Cartwright, 1977a, 1977b; Harwood et al, 1981; P. B. Williams, 1978). Graduation year has been used to separate observations subsequent to data collection (E. D. Cohen & Korper, 1976; Hojat et al, 1989) as with this study. Both the year of birth and the year of medical school graduation were obtained for the physicians in this study, enabling the two variables to be tested for sensitivity to change. See Figure 13 which plots decades of birth and graduation for all physicians in the study group, and see Table A-21. Women were younger and more recent graduates. 157 The average year of birth (rounded) of all women physicians was 1947, and for all men 1940, significantly more recent (p=.00). The average year of graduation of all women physicians was (rounded) 1974, and for all men was 1967, significant to p=.00. Women in Table A-21 had a significantly (p=.00) smaller standard deviation in years than men, whether USA or international MDs, or DOs. This reflected the history of women in medicine in the USA, that included relatively few in earlier years and increased since the 1970 actions that opened enrollment to women. The relationship between year of birth and year of graduation was close (Table A-21), higher for men (r=.97) than for women (r=.95). Age may have been a less sensitive representation of professional experience for women, than for men. There was a broader range of birth years in recent years for both women and men, as medical schools admitted more older students more often in recent years. The distribution of the study group by year of graduation (Figure 13) closely reflected the age distribution in Table A-4, showing increments in both women and all doctors in recent decades. 158 ~ M e n MDs • ° ‘ Men DOs ~ B ~ Women MDs Women DOs 100 % ■o. 80% 60% 40% 20% 0%s— 1920s Figure 13. 1930s 1940s 1960s 1950s Decade ol G raduation 1970s 1980s Year Of Graduation For The Study Group, By Gender 159 Testing of the Hypotheses Hypothesis 1 . Women physicians report specialties of family practice, internal medicine, pediatrics and psychiatry more than do men, as their principal specialty. First these four specialties were identified from the AMA list of 85 categories of specialty (Appendix G ) . The specialty list was truncated into these four (pediatrics included adolescent medicine, and psychiatry included child psychiatry). The remainder were grouped as "other" to create a five point scale. See Tables A-12 and A-22 for the principal specialty of women and men physicians. Women were significantly more (Table A-22, p=.00) represented in those historical specialties, 49.1%, compared with 42.3% of men. Women were in internal medicine, pediatrics and psychiatry more than men, but less in family and general practice. Refer to Tables A-5, A-6, A-7 and A-8 where women MDs nationwide were in all specialties, although disproportionately more in primary care. The hypothesis was partially supported, for the aggregate of the four, and for internal medicine, pediatrics and psychiatry. Hypothesis 2 . Women physicians report primary specialties of family practice, general practice, internal medicine, obstetrics and gynecology, and pediatrics, more than do men, as their principal specialty. 160 These five primary care specialties were identified from the AMA list of 85 categories of specialty (Appendix G), re-classified as primary or nonprimary. See Figure 14 and Table 8 for a profile of the women physicians in the study group of whom 42.8% were in primary specialties. Table 8 Primary Care as Principal Specialty bv Gender Principal Specialty Women Men % % n % aFamily/ General 17.1 20.7 3 199 20.3 bInternal Medicine 10.6 10.6 1 673 10.6 aObstetrics/ Gynecology 6.5 5.1 826 5.2 aPediatrics 8.6 3.4 644 4.1 aTotal Primary Care 42.8 39.9 6 342 40.2 aTotal Nonprimary 57.2 60.1 9 428 59.8 1 966 13 804 Total n= Note: ap=.00. Total 15 770 bNo significant differences. The data supported the hypothesis, that women in Michigan in 1986 were more often in primary care specialties than were men, significant to p=.00, for gender and specialty. Women physicians in Michigan in 1986 were 161 nevertheless in nonprimary specialties more than women MDs nationwide in the same year (Table A-7 and A-12). Mean Age - 39 Years M e a n Year ol G ra d u a tio n = 1974 Mean Patient Care Hours • 44 per week MDs - 81% DOs * 19% Training International ■ 19% Specialty Prim ary ■ 43% Nonprim ary ■ 57% 9999999999999999999999999999999999999999999999999999 24144 Location Metropolitan ■ 92% Figure 14. Rural ■ 8% Profile of the Woman Physician 162 Hypothesis 3 . Men physicians report more even distribution across specialties than do women. Gender was one variable and specialty distribution (a nominal scale of 39 categories in Appendix H) of women and men was the other variable. See Table A-12 for the principal specialty of women and men. Men were most commonly in family and general practice, internal medicine, general surgery, obstetrics and gynecology, and psychiatry. Women were most commonly in family and general practice, internal medicine, pediatrics, psychiatry, obstetrics and gynecology. There were no women in colorectal surgery, pediatric allergy, and few in plastic surgery, neurosurgery, and nuclear medicine. was significant to p=.00. than 5 physicians. The chi square There were 11 cells with fewer This weakened the chi square, but the low representation of women, defined as less than 1% of all women, in 22 of the 27 specialties, supported the argument of the hypothesis. Hypothesis 4 . Younger (born after 1945) men and women physicians have similar specialty distributions. Year of birth was one variable, divided as younger (born after 1945) or older (born before 1946) physicians. The AMA 4 broad (abbreviated) categories of specialty (Appendix I) was the other variable. See Table A-23 for the specialty distribution of women and men born after 1945. Gender differences still existed 163 (p=.00) for all specialty groups. The largest difference was that men were 1.8 times more in surgery than women. Differences for gender were narrow in family and general practice, medical specialties, and the "other" specialties. The data failed to support the hypothesis. Hypothesis 5 . More recent medical school graduates (after 1970) men and women physicians have similar specialty distributions. Physicians' year of graduation, divided into recent (after 1970) or early (before 1971) was one variable. The AMA 4 broad (abbreviated) categories of specialty (Appendix I) was the other variable. See Table A-23 for the specialty distribution of women and men graduated after 1970. As for hypothesis 4, the largest gap was for surgeons, and other specialties were separated by no more than 2.3 percentage points. With a chi square significance of p=.00, the data failed to support the hypothesis. A comment is necessary on the meaning of the results using year of birth or year of graduation to analyze specialty. Younger physicians were by definition also recent graduates, but there were also older persons who were recent graduates, more so for women. This was shown by the higher correlation of years of birth and graduation for men than for women (Table A-21). Using graduation year to group 164 physicians will match them with peers from similar medical education and medical practice environments, rather than age peers. Hypothesis 6 . Women physicians report a lower mean total patient care hours worked per week than men. Hypothesis 7 . Women physicians have a larger variance in the total hours worked per week than do men. Gender was treated as the independent variable and total hours per week was a dependent variable. For all analyses of physicians' hours, two factors were monitored as discussed in "Hours Worked", above. Physicians with zero hours were selected out, and those reporting total hours in excess of 80 were excluded. See Tables 7 and A-20, and Figure 11. hypotheses were confirmed. Both of these In Table 7, all women were working an average of 4.7 hours less than men (p=.00), and with a variance 0.6 hours greater than men (p=.00). The data supported the hypotheses. The inclusion of women who specified their practice as part time has possibly lowered the mean and broadened the variance. When women identified their practice as part time on the OHMA survey, they may have been making a judgement that their shorter hours of work did not reflect a coincidentally shorter week but an intentional restriction of their hours. Subsequent analysis was conducted of these 165 hypotheses for women and men who identify themselves as working full time. The results (Table 7) were that the difference between full time women and men was 1.2 hours smaller, only 3.5 hours less for women, but still significant (p=.00). Men's "average" did not rise as much as did the women's when excluding part time physicians. Men's mean hours rose 4.2 hours, and women's rose 5.4 hours. Figure 12 illustrates the crossover of hours of part and full time physicians, where part time women work more hours than part time men. The rise in the "average" women's hours when comparing full time women and men was evidence that women's hours were lengthening, closer to men's or to a physician norm that may have been gender neutral. Hypothesis 8 . Women physicians report a lower mean total patient care hours worked per week than men, within the same specialties. Gender was treated as the independent variable, and for each of the 4 AMA broad specialty categories (Appendix I), total hours per week was a dependent variable. Those in training were excluded as their hours were typically very long, and proportionately more women were in training than were men. Further, women and men in training were distributed among specialties differently from those fully trained. 166 See Table 9. For all physicians the hypothesis was confirmed for family and general practice, medical and other specialties, with a difference significant to p=.00. For surgery, the mean hours and the standard deviation from the mean was not significantly different between women and men (p=.70). This data partially supported the hypothesis. Again, the inclusion of physicians who specified their practice as part time has possibly lowered the mean and broadened the variance of the hours worked. Subsequent analysis of reports by full time, fully trained men and women was to determine if the hours worked varied with gender only. The results were that the hours differences for gender within family and general practice, medical and other specialties were smaller, but still significant to p=.00. Full time women and men surgeons worked the same mean hours and had the same standard deviation (chi square p=.85). This was evidence for the underlying similarities between women and men physicians. This data also partially supported the hypothesis. Hypothesis 9 . Women physicians have a larger variance in total patient care hours worked per week than do men, within the same specialties. Gender was treated as the independent variable, and for each of the 4 AMA broad specialty categories (Appendix I), 167 total hours per week was a dependent variable. Those in training were excluded as their hours were typically very long, and proportionately more women were in training than were men. Further, women and men in training were distributed among specialties differently from those fully trained. See Table 9. When analyzing all women, there was a smaller standard deviation for mean hours for family and general practice, medical and other specialties, and one decimal point more for surgical specialties. The data failed to support the hypothesis. A subsequent analysis of only full time physicians produced a closer standard deviation of the means of women and men. The data still failed to support the hypothesis. 168 Table 9 Total Patient Hours for Specialty by Gender If Fullv-Trained Specialty Category Women Men All Full Time All Full Time aFamily and General M SD 43.1 14.1 48.3 11.0 47.2 16.1 52.7 11.1 aMedical M SD 44.3 16.0 50.3 11.7 49.2 16.1 53.5 11.7 bSurgical M SD 50.9 16.0 54.4 12.6 51.4 15.4 54.7 12.0 a0ther M SD 41.4 13.2 46.4 9.7 46.9 14.8 50.8 10.7 1 083 850 9 244 8 039 n= ap=.00, for all and for full time physicians. bp = .70 for all surgeons, p=.85 for full time surgeons. Hypothesis 10. Younger (born after 1945) women and men physicians have the same mean total hours of patient care. Younger (born after 1945) physicians were selected to test this hypothesis. The independent variable was gender and the dependent variable was the intervally scaled total hours worked per week per physician. See Table A-23 for the total hours of patient care for specialties of women and men. Younger women were in the 169 major child bearing years and their participation in the physician workforce may have been less than full time. To test this, an analysis was first conducted of the co­ variation between gender and full time or part time, as reported in Table 6, now with the two point variable separating year of birth, using Pearson's chi square. The results (Table 10) were that part time work was favored by both women and men who were older, but primarily women in the younger age classification. However women under 40 years of age were 55% more likely to be giving patient care full time than were older women. The results (Table A-24) were that the gender difference for total patient care hours of younger physicians was smaller when part time were excluded, 4.0 hours less for women, still significant to p=.00. There were underlying similarities between women and men physicians professional contributions, when the difference in patient care hours was only 8% of the higher number. The results failed to support the hypothesis. 170 Table 10 Activity Status of Women and Men. For Year of Birth Professional Activity Men Women Part time Patient Care (%) Born < 1946 11.0 2.4 Born > 1945 8.9 10.2 Full time Patient Care (%) Born < 1946 48.7 34.1 Born > 1945 31.5 53.4 Total (%) Born < 1946 n= 59.2 506 34.6 6 483 Born > 1945 n= 40.5 1 200 65.4 5 462 Note: all differences between women and men, and year of birth significant to p=.00. Hypothesis 11. More recent medical school graduates (after 1970) women and men physicians have the same mean total hours of patient care. Recent (after 1970) medical school graduates were selected to test this hypothesis. The independent variable was gender and the dependent variable was intervally scaled total hours worked per week per physician, of greater than zero and less than or equal to 80 hours per week. 171 See Table A-24 for the total hours of patient care for specialty of women and men. The hypothesis was not supported, as the 7.1 hours per week difference was statistically significantly (p=.00) for all women. In addition to the concerns about the inclusion of part time physicians in any test of gender differences in hours worked (Table 9), this analysis has included age, a predictor of differences between the genders. This analysis was redone separating part time from full time physicians, to eliminate any confounding issues. Figure 15 summarizes characteristics of the recent graduates who were full time physicians. The results (Table A-24) were that the gender difference for total patient hours within specialty was smaller when part time were excluded, from 7.1 down to 4.0 hours, but still significantly different to p=.00. This was evidence of a trend by women and men physicians to work similar hours. Considering the sensitivity of graduation year over birth year for hours worked, if analysis was done only for younger women, it will capture all those in child bearing years who were the most likely to work reduced schedules (Table 10). Recent graduates were more diverse in their age range, and may reflect a trend rather than an age range alone. 172 A v e r a g e P atient Ca re Hours = 53.2 p e r we ek Gender = B B Men 86% T ra in in g 11 ■ — J — Worn en 14% .. :_~ m = USA 84% International 16% 35232292 S p ecia lty Prim ary 47% Nonprim ary 53% 999999999 L o c a t io n Metropolitan 88% Figure 15. Rural 12% Profile of Full Time Recent Graduates 173 Hypothesis 12. Younger (born after 1945) women and men physicians have the same mean total hours of patient care, within the same specialties. Younger (born after 1945) physicians were selected to test this hypothesis. The independent variable was gender. The dependent variable was the total hours worked per week per physician in each of AMA's 4 broad specialty categories (Appendix I ) . Results in Table A-24 showed that total hours were 7.2 less for women than men, significant to p=.00. Surgery hours were significantly different between women and men only at the rate of p=.09. All other differences between women and men within the specialties were significant at the p = .00 rate. The data failed to support the hypothesis. This analysis was repeated after separating full and part time physicians, to eliminate any confounding issues as previously noted, and detected in Table 10. The results (Table A-24) were that the gender difference for total patient care hours was smaller when part time were excluded, averaging 4.0 hours, but still significantly different to p=.00. Regarding specialties, for surgery there was not a significant difference in hours between full time women and men born after 1945. Analysis of variance in hours between the hours worked for the four specialty categories and gender was significantly different, by the F-test to p=.00. 174 The widest gap for full time physicians was for family and general practice (6.4 hours), and the narrowest was in surgery (l.l). This data also failed to support the hypothesis. Hypothesis 13. More recent medical school graduates (after 1970) women and men physicians have the same mean total hours of patient care, within the same specialties. Physicians graduated from medical school after 1970 were selected to test this hypothesis. variable was gender. The independent The dependent variable was the total hours worked per week per physician in each of AMA's 4 broad specialty categories (Appendix I). Results in Table A-24 showed that total hours for each specialty group were significantly different for gender, averaging 7.1 hours difference. Surgery hours were significantly different between women and men at the rate of p=.09. All other differences were significant at the p=.00 rate. The data failed to support the hypothesis. This analysis was repeated after separating full and part time physicians, to eliminate any confounding issues as noted previously and detected in Table 10. The results (Table A-24) were that the gender difference for total patient care hours was smaller when part time were excluded, now averaged 4.0 hours, still significantly different to p=.00. 175 For surgery, there was not a significant difference in hours between full time women and men. Analysis of variance in hours between the four specialty categories and gender was not highly significant by the F-test (p=.ll). The widest gap for full time physicians was for family and general practice (5.7 hours), and the narrowest was in surgery (1.2). This data also failed to support the hypothesis. Comparing the results on hours per week for either graduation year (hypothesis 13) and birth year (hypothesis 12), the maxima were lower and the gender differences narrower when comparing by year of graduation. The results supported the use of years since medical school graduation as a more sensitive arbiter of time and change over time, than birth year. However, as the bulk of the literature dividing physicians over time use age as the variable, the analyses continued for both years of birth and of graduation, for comparison with other studies. Hypothesis 14. Younger (born after 1945) women and men physicians have the same percentage reported as professionally inactive. All physicians born after 1945 were selected to test this hypothesis. One variable was gender, and the other variable was the percentage reported professionally inactive or active. See Table A-25 for the rate of professional 176 inactivity for women and men by year of birth. There were no significant differences, and the data supported the hypothesis. Inactivity was not necessarily a permanent professional state, unless the physician reported being retired. In the light of the results of the hypotheses on inactivity, it was useful to review that if women have brief interruptions, primarily for child bearing and family responsibilities, they return to professional activities (Cartwright, 1977b; Ducker, 1986; Ferrier & Woodward, 1982; Geyman, 1980; Heins, et al, 1976; Heins, Smock, L. Martindale, Jacobs & Stein, 1977; Heins, Smock, L. Martindale, Stein and Jacobs, 1977; Lerner, 1981). Hypothesis 15. More recent medical school graduates (after 1970) women and men physicians have the same percentage reported as professionally inactive. All those physicians graduated from medical school after 1970 were selected to test this hypothesis. One variable was gender and the other variable was professional activity. See Figure 15 and Table A-26 for a profile of activity of women and men recent graduates. Men reported inactivity the same as did women (p=.78) and women were as active as men. This confirmed the hypothesis that recent graduates were similar in their professional activity and inactivity. in Hypothesis 16. Older (born before 1946) women physicians have a higher percentage reported as professionally inactive than older (born before 1946) men. All those physicians born before 1946 were selected to test this hypothesis. One variable was gender and the other variable was professional inactivity versus active. See Table A-25 for the rate of professional inactivity for women and men by year of birth. the same as did women (p=.18). Men reported inactivity The data supported the hypothesis. The literature has many different and differing measurements of inactivity. Hojat et al (1987) found 15% of women and 2% of men in their sample were inactive, compared with the AMA finding that in 1986, 6.6% of women MDs nationwide were inactive and that men were less active than women (AMA, 1987; Table 3). Hypothesis 17. Early women medical school graduates (before 1971) have a higher percentage reported as professionally inactive than early men medical school graduates. All those physicians who graduated from medical school before 1971 were selected to test this hypothesis. One variable was gender and the other variable was reported professional inactivity or activity. In Table A-26 the data failed to support the hypothesis, as women reported inactivity no differently than did men (p=.78). 178 There were no detectable differences in gender inactivity totals by either year of birth or graduation (Hypotheses 14, 15, 16 and 17). Hypothesis 18. Women physicians report patient care less often than do men physicians. One variable was gender and the other variable was the nominal scale of reported professional status, recoded into patient care or other. See Tables 6 and 7. In this study women reported patient care at the same rate as did men (p=.38). The data failed to support the hypothesis. It was relevant to analyze the question for full versus part time physicians as part timers may have been in patient care and therefore boosted the patient care proportion for women. Reanalysis for part and full time patient care revealed that women were doing significantly more part time care (Tables 6 and 10) and less full time patient care (chi square p=.00). However, this data still failed to support the hypothesis. Hypothesis 19. Women physicians report inpatient hospital care as the sole professional activity more often than do men. One variable was gender and the other variable was the single category of inpatient hospital care as the practice type without any nonhospital professional activity on the 179 nominal scale of professional activity. Full time physicians were selected to make comparisons relevant, due to more women doing part time patient care. Part time care may have been more likely to be single site, either hospital or office based, but not enough time for both. See Table A-14 in which women did report hospital care alone more than did men for both full and part time physicians. The data supported the hypothesis. This analysis included women in graduate medical education, who were a greater proportion of all women than men in graduate medical education were of all men. These women in training were by definition in hospital based care. There was a cohort of women graduating from medical school and entering graduate medical education (Tables 2, 6, 7, A1, A-2, A-8 and A-15) that was larger than preceding decades. This analysis was redone with physicians in training excluded from the analysis (Table A-20). Women still did report hospital care alone (29%) more than men (20%) for both full and part time physicians (p=.00). These data continued to support the hypothesis. It was interesting to review the data to see how the assumption may have arisen that women work more in hospitals than did men. To begin with (Table A-14) 44% of all women were working only in hospital care. This dropped to 33% when part time women were excluded, and further down to 29% 180 when those in training were excluded. Decades of unexamined observations have failed to acknowledge or differentiate the sub-populations of women physicians when analyzing practice patterns by gender. Hypothesis 20. Women physicians report a lower mean of inpatient hospital care hours worked per week than men, when reporting both hospital and nonhospital practice. One variable was gender and the other variable was the total number of inpatient hours reported worked per week. Fully trained physicians reporting full time patient care were selected to test this hypothesis to permit more meaningful comparisons as women were both more in graduate medical education and in part time practice, more than men. See Table A-20, in which women reported 2.7 fewer inpatient hospital hours when reporting both hospital and nonhospital practice than did men (p=.00). supported the hypothesis. The data There was no difference (p=.63) in office hours when working in both hospitals and offices. These results raised questions regarding why men had more hospital hours when office hours were identical for full time women and men. Subsequent analysis examined this gap for the influence of specialty. Hypothesis 21 . Women physicians report administration as their principal professional activity less than do men. 181 One variable was gender and the other variable was the category of administration reported on the nominal scale of principal specialty (reported or ascribed). See Table 7. No women reported administration as their "principal specialty". This supported the argument of the hypothesis, but it was not possible to test for significance. Seventeen men reported administration. The OHMA instrument was not sensitive enough to detect time spent in administration, as the survey card term "specialty" did not specify the function or activity of administration. The absence of women reporting their principal activity as administration was an eloquent enough indicator to support the argument that women were less in administration than men. Women had the same opportunity to indicate administration as their principal specialty but none did. Women have been underutilized in administration (Rinke, 1981a; S. C. Martin, et al, 1988). Women were also underused in medical administration (S. L. B r o w n & Klein 1982; Costanza, 1989; Dickstein & Nadelson, 1986; Dickstein & Stephenson, 1987; W. Gross & Crovitz, 1975; Loutsch, et al, 1982; Maheux, Dufort & Beland, 1988; and Sullivan, 1974). Administration was reported as a secondary specialty by 40 men and only 1 woman, further supporting the argument. Again, the activity of administration was not specifically identified in the OHMA survey. Even though they were not identical, women had the same option as men to identify 182 "administration" as a primary or secondary specialty, but did not. The data neither supported nor failed to support the hypothesis. Hypothesis 22. Women physicians report a lower mean total hours worked per week in administration than do men. Physicians reporting administration as their principal specialty were to be selected to test this hypothesis. Due to the absence of women reporting administration, this hypothesis could not be tested. In addition, there was uncertainty whether the response of administration as a principal specialty reflected the percentage of physicians in the activity of administration. Hypothesis 23. Women physicians report teaching as a professional activity more often than do men. One variable was gender and the other variable was the number of physicians reporting teaching as either their principal specialty or secondary specialty. See Table 7 in which only 5 men reported teaching as a principal specialty, and 3 men and only 1 woman as a secondary specialty. As noted for hypothesis 21, the question asked by the OHMA survey referred to a specialty, not teaching as a function. However, women had the same opportunity to respond with teaching information as did men, but none did so. 183 More women were expected to report teaching as women have been an increasing proportion of new faculty (AAMC 1988a, 1988c, 1989), and teaching has been recommended for women (Beshiri, 1969; Bowman & D. I. Allen, 1985; C. Eisenberg, 1981). The data neither supported nor failed to support the hypothesis. An additional specific question in the OHMA survey regarding hours in teaching would capture this information more accurately. Hypothesis 24. Women and men physicians report the same mean total hours worked per week teaching, when it was the sole professional activity reported. Physicians reporting teaching as their principal or secondary specialty were to be selected to test this hypothesis. In the light of the low report of teaching (Table 7), the hypothesis was not tested (Table 7). Further there was uncertainty regarding the meaning of these responses of teaching to a question on "principal specialty". The data neither supported nor failed to support the hypothesis. Hypothesis 25. Women physicians report research as a professional activity less than do men. 184 One variable was gender and the other variable was the reporting research activity as either their principal or secondary specialty. The specialties selected were research, pharmaceutical research, or experimental pathology. See Table 7 in which only 4 men reported research as a principal specialty, and 5 men and 1 woman reported it as a secondary specialty. Again, asking for specialty rather than activity may have not gathered representative or meaningful data. However, women had the same opportunity as men to enter research-related specialties or activities but did not, supporting the argument behind the hypothesis. Women MDs were expected to be participating in research at the same rate as men (AMA, 1987; Table A-3). The data neither supported nor failed to support the hypothesis. Hypothesis 26. Women physicians report fewer total hours worked per week in research than do men. Physicians reporting research as principal or secondary specialty were to be selected to test this hypothesis. Without women reporting research (Table 7), and a concern for whether the reporting would be representative or meaningful, this hypothesis was not tested. Hypothesis 27. Women and men physicians are in graduate medical education training at the same rate. 185 One variable was gender and the other variable was the single category of current training reported as professional activity. The total percentage of physicians in the study group who were in training was 15.3%. Detailed information was available on less than half of these, and solely for those no longer on educational limited licenses. The latter license was required for all first year trainees in 1986, and continued for all international graduates up to three years, as well as for any physician not seeking a full license upon completion of the limited license. See Table A-15 in which women were in graduate medical education at a rate 1.9 times that of men (p=.00), 26.1% to 13.8%. The data failed to support the hypothesis as worded. The intention of the hypothesis was to detect whether women had increased to the same proportional level of training as men. As women had exceeded the rate of men, the argument was supported. Hvpothesis 28. Women are less likely to be training at the fellowship level than are men physicians. One variable was gender and the other variable was the single category of fellowship in the single category of current training reported as professional activity. As an educational limited license could be retained in second and subsequent years, there may have been some 186 missing data as these individuals were not surveyed directly in 1986. Their training status would be available as it was required information for licensure. However, full licensure was required for billing medical services to payor sources. Thus there was an economic incentive for sponsors of residencies and fellowships, mostly hospitals, to require conversion to full licensure. The fellowship data was therefore considered reliable and the residency data as representative of second year and later residents. While women were 1.9 times more likely to be in graduate medical education than men, see Table A-15 in which women were in fellowships 35% less than men (p=.00), 6.4% to 9.8% of all in graduate medical education. Nationwide in 1986, women MDs were slightly less represented in the fellowship level (AMA, 1987), and they had not trained at that level in the past (Tables A-5, A-6) but were increasingly doing so (Tables A-7 and A-8). The data supported the hypothesis. Note that fewer DOs and fewer DO women were in fellowships than MDs and MD women. Fellows were most likely to be in the medical specialties (footnote to Table A-15). Hypothesis 29. Women physicians are less broadly distributed among the 83 counties in Michigan than are men. One variable was gender and the other variable was the county designation for their principal location. 187 Women physicians have been less broadly distributed than men and this was true for women physicians in Michigan in 1986 (overall p=.00). Men had principal locations in every county but women were absent from 19 or 23% of the counties in Michigan. As 66 of the 162 cells contained fewer than 5 physicians, this chi square was rendered weak. The distribution of principal locations among counties was different for women and men. The data supported the hypothesis. Commenting on the county location questions, hypotheses 29 through 35, this data came from the principal location or that for which each physician reported the highest number of hours. This method may have excluded a smaller practice site of a physician in another county, under counting the number of physicians serving the second county. The choice was made to make the individual physician the unit of measurement rather than the parts of physicians serving in several counties. Hence the location data may have been an under count. Hypothesis 30. Women physicians are more likely to practice in metropolitan areas than are men. See Tables A-17 and A-18 and Figures 10 and 16. Hypothesis 31. Men physicians are more likely to practice in rural areas than are women. 188 The counties in Michigan were classified into a nominal scale of metropolitan, rural adjacent, or rural county (Table C-2) location for one variable, and gender was the other variable. Only active physicians were selected for testing these hypotheses. See Tables A-17 and A-18 and Figures 10 and 16 in which 11.2% of active men practiced in a rural county in 1986 compared with 6.0% of women (p=.00). The data supported both of the hypotheses. Hypothesis 32. Women primary care physicians are more likely to practice in rural areas than are women specialists. Selecting all active women physicians, one variable was the classification of principal specialty as primary or nonprimary. The other variable was the three point scale of metropolitan, rural adjacent or rural county (Table C-2), reduced to metropolitan or rural. See Table A-27 for the distribution of active women physicians in metropolitan and rural areas by their specialty type. All active women primary care physicians were 2.1 times more (p=.00) in rural counties than were women specialists. All active primary care physicians (Table A-27) were 1.9 times more likely to be in rural sites than specialists (p=.00). The data supported the hypothesis. 189 Women in specialties were increasing and many (26%, Table A-7, A-20) were still in training in this study group. This was reanalyzed to exclude women in training to determine the proportion of fully-trained women primary care physicians in rural areas. Women primary physicians were still significantly more in rural areas (p=.00, Table A-27), but only 1.6 times more than were specialists. The women in graduate medical education in metropolitan areas had boosted the gap between metropolitan and rural areas. These data also supported the hypothesis. Hypothesis 3 3 . Women primary care physicians are more likely to practice in metropolitan areas than are men primary care physicians. Selecting all active primary care physicians, gender was one variable and the other variable was the metropolitan category of county location. See Table A-27 in which 85.9% of men primary care physicians were practising in metropolitan counties compared with 91.5% of women primary care physicians, even though more women were in primary care than men (Table A-16). This was a significant difference (p=.00), supporting the hypothesis. For metropolitan physicians who were active, and not in training, there was still a significant difference between the proportion of men and women primary care physicians in 190 metropolitan areas (Table A-27, p=.00), with more women in metropolitan areas. The data excluding those in training continued to support the hypothesis. Hypothesis 34. Younger (born after 1945) women physician have a broader distribution between metropolitan counties, rural counties adjacent to a metropolitan county, and rural counties that are not adjacent to a metropolitan county, than older (born before 1946) women physicians. Selecting all women physicians, women in graduate medical education were excluded, as the large number of younger women were likely to be in graduate medical education (26.1%), and the residency sites in teaching hospitals were most likely to be in metropolitan areas. One variable was the scale of younger (born after 1945) and older (born before 1946) physicians. The other variable was the classification of counties into a three point scale of metropolitan counties, rural adjacent and rural counties (Table C-2). In Table A-28 there was no significant difference in distribution for younger women, even with the exclusion of women in graduate medical education (p=.49). The pattern of greater metropolitan distribution of women was sustained for women who were born after 1945. The data supported the hypothesis. 191 Hypothesis 35. More recent medical school graduates (after 1970) women physician have a broader distribution between metropolitan counties, rural counties adjacent to a metropolitan county, and rural counties that are not adjacent to a metropolitan county, than early women medical school graduates (before 1971). Selecting all women, again women in training were excluded. One variable was the scale of those graduated from medical school before 1971 or after 1970. The other variable was the classification of counties into a three point scale of metropolitan counties, rural adjacent and rural counties (Table C-2). See Table A-28 for the location of women physicians by their year of graduation. More recent women medical graduates were practising in a broader geographical distribution (p=.06). The data supported the hypothesis. This mirrors D'Elia and I. Johnson (1980) Illinois findings, and quantifies the location women expressed as preferences in surveys (E. K. Adams & Bazzoli, 1986; R. H. Rosen, et al, 1981) and portrayed in Figure 16. Whether this reflects their medical education in recent decades, or recent trends in medical practice in the state, this study could not explain the increased proportion in a rural location among women who are recent medical school graduates. 192 Hypothesis 36. There are more USA trained women MD physicians among all USA trained MD physicians, than there are women DO physicians among all DJ physicians. One variable was professional degree (MD or DO), and gender was the other variable. See Table A-ll and Figure 14 showing a significant difference (p=.05) in the percentage of women DOs (10.0%) than USA trained MDs (11.5%). This reflected the historical lag for women entering DO schools (Table A-2). It has already been noted that osteopaths were over represented in the study group, but this hypothesis only measured women as a proportion of each profession. The data supported the hypothesis. Women in the study group were not in each profession at the same level. This should be treated with caution due to the possible under representation of women MDs in the study group, who were 16% of the archived 1986 AMA database. The difference in representation of women in the two professions may have been even greater. Hypothesis 3 7 . DO women physicians are more likely to be in primary care than USA trained women MD physicians. All USA trained women physicians were selected, and one variable was professional degree (MD or DO). The other variable was principal specialty as primary or nonprimary. In the study group, 38% of all women were in primary care (Table 11, Figure 8). The evidence for support of the 193 hypothesis was striking, as there was a perfect inversion of percentages of primary physicians between women MDs and DOs. Among MD women, 38% were primary care, and 62% were in nonprimary specialties. However, DO women were 38% in nonprimary specialties, and 62% in primary care. The data supported the hypothesis, with DO women 1.7 times more in primary care specialties than MD women. This was reanalyzed for the effect of recent medical education graduation, as DOs were entering specialties at higher rates than previously. However, 90% of all DO women in the study group had graduated since 1970. The results were that recent DO women graduates were still more in primary care than recent MD women graduates, still significant to p=.00. The data still supported the hypothesis, even for recent graduates. To focus closer on DO differences across time, only those in graduate medical education were examined. Those DOs currently in training would have been most influenced by changes favorable to women within that profession. Sixty five percent of fully trained women DOs were in primary care, but only 32% of trainee DO women were in primary care, significant to p=.00 (Table 11). This almost inverted the 65% / 35% split for fully trained DO women, for primary and nonprimary specialties, respectively, illustrating the change to nonprimary care by the younger DO women. 194 The trends in Table 11 of women's specialty by their nation of training and degree, show that primary care was the principal specialty for 45% of international MDs, down to 21% of MD women (both USA and International) in training. It may take several decades for the change in women's specialty choices to be the norm for all women. However, the trends were vocal that any stereotypes, limited choices, or the restrictions on women in the past have less impact on opportunities for the woman physician in training in 1986 (Tables A-6 and A-7). Table 11 Specialty. Degree and Nation of Training of Women Physicians Degree MD Total % Primary % Specialty n 38 62 1 588 international - All - a,bFully Trained - aTrainees d42 47 21 58 53 79 404 aUSA - All - a,bFully Trained - a,cTrainees d36 48 21 64 52 79 1 184 aMD Trainees - All 21 79 DO Total - a'bFully Trained - a'cTrainees 62 65 32 38 35 68 Total % n= 43 57 378 1 966 195 Note: Significant, p=.00, trainees' specialties different from those in practice. bp=.00, DO fully trained women different from MD fully trained women. cp=.02, DO trainees different from USA MD trainees. dp=.00, more international MD women in primary care than USA MD women. USA MDs n"L184 95%93%95% 1 J In tern atio n al MDs n«*404 DOs n-378 94% 96% 91% 62% 62% Patient Care Figure 16. Primary Nonprim ary Rural Metropolitan Specialty and Location by Nation of Training of Women Physicians 196 Hypothesis 38. There are more internationally trained women among all women MD physicians, than internationally trained men among all men MD physicians. Only MDs were selected to test this hypothesis, with one variable gender and the other variable was nation of training. See Table A-11 and Figure 14 in which there were 35% internationally trained women MDs among all women MDs, and (Table A-ll) 24% internationally trained men among all men MDs (p=.00). The data supported the hypothesis. To test for the more recent access to medical education for women in the USA, this hypothesis was viewed a second time from the perspective of the physicians' year of graduation (Table A-21). The population separated, with more international women as earlier graduates (mean graduation year rounded to 1968) and more USA women as later graduates (mean graduation year of 1976, significant to p=.00). This latter figure reflected Tables 1 and A-l showing the growing proportion of USA women in medicine over recent decades. It also reflected the peak international graduate recruitment and immigration during the 1960s and 1970s physician shortage period in the USA, prior to the increased enrollment of women in USA medical schools. These data continued to support the hypothesis. 197 Table 12 Nation of Training. Year of Graduation of MDs. bv Gender Nation of Training aWomen aMen % Early % Recent % Early % Recent aUSA 40 78 72 82 international 60 22 28 18 567 1 397 7 496 6 305 n= Note. Early graduation was before 1971 and recent after 1970. aAll nation of training and gender differences were significant, p=.00. Hypothesis 39. Internationally trained women MD physicians are more likely to be in primary care than USA trained women MD physicians. After selecting all active women MD physicians, one variable was the nation of physicians' training. The other variable was the classification of principal specialty into primary or nonprimary specialties. See Table 11 in which 42% of internationally trained women MDs were in primary care, compared with 36% of USA trained women MDs (p=.00). The data supported the hypothesis. As a secondary analysis of this question, women in graduate medical education were excluded as there may have 198 been proportionately more USA women in training. International and USA women MDs were all in primary care at the same rate (Table 11). MD women in training were entering a broader range of specialties (Tables 11, A-7, A16) . The data from this subsequent analysis also failed to support the hypothesis. Hypothesis 40. Women physicians are younger on the average than men physicians. Gender and year of birth were tested for differences. Hypothesis 41. Women physicians are more recent medical school graduates on the average than men physicians. Gender and year of medical school graduation were tested for differences. See Table A-21 for gender differences in year of birth and year of graduation from medical school. The data reflected the national figures in Table A-4 of more, younger women physicians in the USA since the 1970 medical school enrollment policy to include more women. Women were significantly younger, with a mean age of 38.8 years to men's 46.2 years (p=.00). Women had fewer years since graduation, a mean of 11.6 years compared with men's mean of 19.3 years, significant to p=.00. The data supported the hypotheses. 199 Gender 13% W om en E 87% M en 73% USA N a t i o n of T r a i n i n g 27% In te r n a tio n a l 37% P rim ary Specialty 63% N o n p r im a r y 91% L o c a tio n 9%R u ral M Atrnnnl Itrrn 53% < 1971 47% ) 1970 G raduation Pr actice m 19% H o s p ita l O n ly I 8% O llic e O n ly 73% M ix e d Figure 17. Profile of Active Full Time MDs 200 Hypothesis 42. Women in nonprimary specialties are younger than women physicians in the primary specialties. Selecting all women physicians, the variable describing principal specialty as primary or nonprimary, and age, computed as the difference between 1986 and the year of birth, were tested for differences. See Table A-29 where the average age for women in primary care was 39.4 years with a standard deviation of 10.7 years. In specialty care the mean age was 38.4 years with a standard deviation of 11.4. The means were significantly different (p=.06). This reflected the limited access of women to specialty training until recent decades. Younger women were training in a broader range of specialties than those women already in practice (Tables A-5, A-6, A-7 and A-16). The data supported the hypothesis. Hypothesis 4 3 . Women physicians are skewed to the younger age groups in specialty care, more than are men in specialty care. All nonprimary specialty care physicians were selected from the classification of principal specialty as primary or nonprimary. Differences in the mean year of birth by gender were tested. See Table A-29 for the gender distribution in primary and specialty care by age. Specialty women's mean age was 38.4 years, standard deviation of 11.4 years, being 6.8 201 years less than 45.2 years for men, whose standard deviation was 13.6 years (p=.00). Men were older and more broadly distributed. This reflected the national figures in Tables 1, A-4 and A-20 of more, younger women physicians since 1970, and the emerging younger women physicians training in a broader range of specialties (Tables A-7 and A-16). The data supported the hypothesis. Hypothesis 44. Women physicians in specialty care are skewed to the more recent medical school graduation years, more than men in specialty care. All nonprimary physicians were selected from the classification of principal specialty as primary or nonprimary. Differences in the number of years since medical school graduation by gender were tested. See Table A-29 for the gender distribution in specialty care by years since graduation. Women's mean years since graduation were 11.0 years, being 7.5 years less (or more recent) than men's, 18.5 years since graduation (p=.00). The data supported the hypothesis. This reflects the national figures in Tables 1, A-l, A2, and A-20 of more women medical graduates since 1970, and the emerging women physicians were training in a broader range of specialties (Tables A-7 and A-16). 202 Hypothesis 45. Certain specialties are more likely to be reported by physicians whose principal activity was research, for both women and men. As an exploratory hypothesis, all physicians were selected who reported research activity on the nominal scales of principal and secondary specialty. The specialties reported (AMA's 39 categories in Appendix H) and percentage of each gender of physicians in each specialty were determined. There were only 10 physicians who reported research as a specialty. This was not probably the same information as research as an activity as reported in the literature. Due to the uncertainty of the data this hypothesis was not analyzed. Subsequent Analyses Geographical Location Year of Graduation Location was re-examined by the year of physicians' graduation for time and gender patterns. Those currently in training were excluded as 99% of their sites were metropolitan and may not have reflected their ultimate practice plans. Metropolitan location had been an increasing choice by male physicians who graduated prior to 1960 (Figure 18). Thereafter male physicians graduating since 1960 have increasingly selected rural locations. Eighteen percent of 203 those men who had graduated in 1980 or afterwards, and who were fully trained, were in rural locations in Michigan (Table A-30). ~ + ~ + Women-Metro Women-Rural Men-Metro ■■*" Men-Rural 100 % 80% 60% 4U% 20 % 0% — 1920s Figure 18. 1930s 1940s 1950s 1960s 1970s 1980-86 Geographical Location by Decade of Graduation and Gender 204 Overall, women physicians had selected metropolitan locations more frequently than men (hypotheses 29, 30 and 31). Women graduating after 1970 increasingly selected rural sites. Sixteen percent of those women who had graduated in 1980 or later, and who were fully trained were in rural locations in Michigan (Table A-30). This could be a visible impact as 75% of all active, fully trained women physicians had graduated since 1970. These trends held for all of USA and Internationally trained MDs and DOs, and for both primary and nonprimary specialties (Hypotheses 32, 33, 34 and 35). Specialty Nonprimary specialists had historically been more concentrated in metropolitan areas, due to the larger population base necessary for specialists and subspecialists. MDs. DOs have been more in primary care than There was an interaction between gender and specialty (hypotheses 1, 2, 3, 4 and 5) and between gender and location. There were interactions between degree (MD or DO) and gender, specialty and location (hypotheses 36, 37 and Tables 11, A-17 and A-18, Figure 18). One subsequent analyses was of the interactions between gender, degree, location (as metropolitan or rural, Table A19) and specialty, using AMA's 4 brief categories (Appendix I). Those in graduate medical education were excluded due to the transient state of their career. 205 Table A-31 described the proportion of each specialty group in which women and men MDs and DOs practiced in rural areas (Appendix B) . were significant. All gender and specialty differences There was no evidence for generalizations as to which physicians have or will practice medicine in the rural counties in Michigan. Family and general practitioners were the most prominent, followed by a range of physicians. In the study group there were 8 men for every woman physician, but there were 12 men for every woman in the rural locations (Table A-32). Relatively speaking, women were highly (1 to 6 men) represented both in general, and in rural sites in the medical and other specialties, than women in family and general practice and in surgical specialties. As a subsequent analysis only full time and fully trained and active physicians were selected. DO women were proportionately more in rural sites (Table A-31) but they represented only 20 of the total of 233 DOs in rural locations. If there were enough women in specialties, they will locate in the rural sites at the same rate as men. By graduation year, younger graduates had moved more into rural areas, as well as being more in specialties than their established predecessors. Tables A-31 and A-32 described the specialty and profession for rural physicians by gender. All differences were significant between specialty, gender and degrees. 206 These results indicated that there are great differences among rural practitioners, and that a wide variety of physicians have selected rural practice. Yet to the extent that women were represented in a specialty, they were as likely to be in rural practice as women in primary care. Patient Care Hours Per Week Women and men physicians worked different patient care hours (hypotheses 6 and 7) even within specialties (hypotheses 8, 9, 12, 13, Tables 9, A-23 and A-24). A subsequent analysis explored total hours reported by the physician's gender and specialty, limited to full time, fully trained physicians. All major interactions were significant (Table A-33) except for surgery. Total hours were different for gender, for specialty, and for total hours for gender within each specialty as well as total hours per specialty within gender. These findings confirmed the separate results of hypotheses 8, 9 and 12, shown in Tables 9, A-23 and A-24. The data illustrated both the diversity within medical practice by gender and specialty, and the virtual impossibility of attributing observed patient care hour differences for gender to a single variable. 207 Practice Type Specialty and gender differences had been observed for the type of practice, whether hospital-only (hypothesis 19 and Tables A-14 and A-20), office-only (Table A-20), or mixed hospital and office practice (hypothesis 20 and Tables A-20), A subsequent analysis for practice type, gender, location and specialty (Appendix I) found an interplay, summarized in Table A-34. Note that this analysis included those in graduate medical education to capture the greater proportion of women in specialties. Hospital Only Practice Hospital only practitioners were more likely to be specialists (80%) and the most prominent was surgical specialties. They were more likely to be women (hypothesis 19, Tables A-14 and A-20) and to be proportionately more MDs. They were more likely (56%) to be recent graduates, capturing those in training. They were the most likely to practice in a metropolitan area (92%). Three forces may have interacted, namely the hospitals hire more specialists, as well as employing trainees, and the hospitals were in metropolitan areas. Office Only Practice Physicians who practice solely in offices (Table A-34) were equally likely to be primary or nonprimary specialists, the most common being family and general practice (38%). They were proportionately more likely to be women (16%) as 208 found previously (hypotheses 1, 2, 3, 4 and 5, and Tables 8, A-12, A-16 and A-22). This was the practice type most dominated by DOs (28% of total). be earlier graduates (75%). They were more likely to They were just as likely, proportionately speaking, to practice in metropolitan or rural offices. Primary care was linked to both women (hypotheses 1, 2 and 3, Table 8 and Figure 14) and to DOs (hypothesis 37 and Table 11). Their interaction with office only practice may have influenced the proportional dominance by women and DOs. Mixed Practice This was the dominant practice type (72%) of active physicians (Table A-34). Physicians who practiced in both hospital and office practice were as proportionately likely to be primary or nonprimary specialists, the most prominent being medical or surgical specialties. They were proportionately as likely to be women or men (12% or 88% respectively). likely to be MDs. They were more proportionately They were more evenly distributed between early and recent graduates than the other practice types. They were as proportionately as likely to be in rural areas as metropolitan areas. Practice Type and Patient Care Hours Those fully trained full time physicians in hospital and office practice were analyzed for patient care hours by their specialty and gender (Table A-35). 209 Differences of total hours were significant for gender for hospital only practice, and for mixed hospital and other practice. There were no significant differences for gender or specialty, or gender and specialty, for the office practice group. For hospital only practice, the two way interaction of specialty and gender was not significant. All patient care hour differences between gender and specialties were significant for the mixed hospital and office practice group. Further analysis of mixed practice (Table A-36) confirmed no significant differences for office hours between women and men, within specialties or for the two way interaction of gender with specialty. The hospital component did have significant differences in hours for gender, and for specialty, except for surgery where there was not a significant difference. However, the two way interaction for specialty and gender was not significant. Again, no clear pattern emerged to explain the observed higher hospital hours for men than women in a mixed practice. As this was the most common practice pattern of all active physicians, there was not the power to explain the differences in hours between the genders. Nation of Training and Specialty Men trained in USA or international medical schools had different patterns of specialty. It was established that men and women were in different specialties (Hypotheses 1, 2, 3, 4, and 5, and Tables A-5, A-12, A-13, A-16, A-22 and 210 A-2 3). A subsequent analysis examined the gender and nation of training differences among specialties (Table A-37). USA men had the principal specialty of medical and surgical specialties, the same as international men. specialties. The proportion of international women MD physicians was greater among women than for men (hypothesis 36, Figure 14 and Table A-ll). However these groups of women were in primary specialties at the same rate, whether fully trained or in training (hypothesis 39, Table 11, figure 15). The same direction existed in Table A-37 for women as for men. More USA women MDs were in medical or family and general practice, and more international women were in medical specialties. As fewer international graduates come to the USA (Fulop, 1986a) and to Michigan (MCGME, 1987), the USA MD pattern (Figure 17) will emerge as the dominant practice of physicians in Michigan and the USA. Length of Career The literature suggested that women planned to work longer than men (Batchelor, 1990; Heins, Smock, L. Martindale, Jacobs and Stein, 1977; Morgan, 1971; J. H. Watson, 1977). available data. This could not be directly measured with the However, an approximation was attempted by analyzing the rate of active versus inactive women and men physicians with respect to the number of years since their graduation. Inactive were grouped into temporary inactive, 211 as designated by the physicians, or presumed inactive by OHMA based on data, and those self-designated as permanently retired. The results (Table A-38) were that both women and men had bimodal distributions of inactivity, between their first and tenth years after medical school graduation, and rising after their thirtieth year after graduation (Figure 19). Activity rates of women were statistically significantly higher than men's for the first two decades after medical school graduation, but not different for the following decades. This finding inverts the historical pattern of women's lower activity rates. A further analysis (Figure 20) of patient care hours for women and men across their years since graduation, illustrated an almost constant, lower difference in hours for women. In the light of the increased hours by younger physicians, this difference may narrow in the decades to come. 212 Active - + ~ Inactive-Temporary 97% Inactive-Permanent 96% 86% 80% WOMEN = =±== 1-10 11-20 21-30 95% 96% 31+ MEN 1-10 Figure 19. 11-20 21-30 Length of Career By Gender 31+ 213 W o m e n - P a r t Tim e ■••d t W o m e n - F u ll Tim e M e n -P a rt Time ~B ~ M e n - F u tl Tim e M e a n H o u r s Per W e e k 60 55B 50 45 40 35 -+■* 30 25 20 — 1-10 11-20 21-30 Years S in c e G r a d u a t io n Figure 20. Hours Throughout Career By Gender 3T PROFESSIONAL ACTIVITIES OF WOMEN AND MEN PHYSICIANS IN MICHIGAN Volume II By Margaret Deirdre McNiven A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of College and University Administration 1991 V : SUMMARY AND CONCLUSIONS Summary of The Study The problem to be investigated was whether women physicians in Michigan participated differently in professional activities from men physicians. More women had entered medicine in the USA in the past two decades (Tables A-l and A-2) with the lifting of restrictions by medical schools and universities. There was evidence from the literature that women entered a broader range of specialties in the 1980s (Tables A-7 and A-8, and Figure 5). The professional work hours of women physicians were lengthening at the same time that men physicians were reducing their professional hours. Women medical students and residents indicated the same degree of preference for rural locations as did men, although practising women physicians were still more likely to be in urban areas than men. Nationwide, women physicians were more professionally active than were men. This study was based on a secondary analysis of an amalgamated data base drawn from three archived surveys, the sources of data matched to each other by name, address, professional degree and specialty of the respondents (Figure 7, Table 5, and see chapters III and IV). The amalgamated database (Appendix J) described all professional activities of 15,770 Michigan licensed physicians, MDs and DOs who were surveyed in Michigan in 1986 (Appendices C and D) and for 214 215 whom there was an identical entry in either the AMA (Appendix E) or AOA (Appendix F) Physician Masterfiles. Purpose of the Study Medical schools and their universities have changed the student enrollment of women in the last two decades. The purpose of this study was to describe the professional activities of all women and men physicians (MDs and DOs) in Michigan. Gender differences were analyzed regarding specialty, hours worked in hospitals and offices, geographical distribution, and professional inactivity rates. Nation of training and professional degree differences within gender differences were examined. Secondly, the study examined whether any such differences were related to age or years since graduation. Physicians trained and active during the two decades of increasing opportunities for women in medical education may have different professional activities from the earlier and older physicians (see chapters I and II). This study identified some of the changes in physician activities, intended and unintended, that occurred in Michigan since USA medical schools and universities increased enrollment of women in the last two decades. The generalizations from this research were applicable to no less than 56% of MDs and 89% of DOs practicing in Michigan, based on their representation in the final database. There was a 98% response rate to the source survey, and no less than 99% of the physicians had an AMA or 216 AOA record. However there was a loss of records (Table 5) in the matching process so that 66% of physicians practising in Michigan in 1986 were represented in the final database. A primary contribution of this study was to determine for the first time the professional activities of men and women physicians (MDs and DOs) in Michigan. The inclusion of both MDs and DOs appears to be the first such research on women physicians' professional activities, as no other published statewide study addressed the contribution of DO physicians, neither women nor men. The opportunity for comparison between these two medical professions' contributions by women to a specific state has been a further first study in the research on women physicians in the USA. No study of women DOs has been reported. The research on the difference in the distribution of specialties between men and women physicians created a basis for describing the current supply of each specialty, and the differences over time of supply for each of women and men, and for the two professions (MD and DO) trained in the USA and Michigan. The research on the geographical distribution of women and men physicians provided a planning tool for local and state bodies for their physician needs (Figures 10 and 18) . The research on the professional activities of women and men physicians created a profile of the physician who was in patient care in Michigan in 1986, specifying those 217 solely in a hospital or an office, or in a mixed practice in both hospital and office. Comments on Secondary Analysis There were several drawbacks to the use of existing data. Firstly one could not control the way the data was collected, from conceptualization to the phrasing of questions, what was included and excluded, and how records were entered and maintained. For example hypotheses 21 through 26 on teaching, research and administration could not be reliably or meaningfully addressed by the data as gathered by OHMA. Secondly one could not control the acquisition of the data from the existing sources, nor determine the rules for access. For example, AMA elected not to do the match of the L&R records to their full Physician Masterfile, and hence there was more loss of MDs in the final database (Table 5). Again, the OHMA decision to delete the license number as the identifier for the record withdrew a "real" identifier for follow up studies. Due to the data being on computer tapes, the access was rapid and immediate (after replacement of imperfect tapes). However as formats varied from the three sources, vast statistical support was necessary to resolve the differences and create the final data base. Verification and trust was an important factor when using existing data, unlike the control exercised when one collected the raw data. The 218 underlying standard created was that if the data could not be trusted, such as when there were two physicians records similar but not identical, the data were not used. Standardization was maximized, such as converting AOA data to AMA criteria. The positive experiences in using existing data deserve full mention. Due to the state and national role of the data source organizations, they had rigorous oversight to ensure that their data instruments and collection process were correct and consistent. Their ability to maintain data on 99% of all physicians in Michigan enabled the scope and representativeness of this study. The size of the data base was obtained at a fraction of the cost (time, money and staff) of initiating an original survey. Two operations matters were notable. The role of the statistician / systems analyst in converting and combining existing data was vital to ensure that what was needed was indeed obtained or created. Time was the other matter. Waiting for willing but busy agents of major organizations to create the source tapes to the requested protocols consumed time, which was itself an important resource. 219 Discussion of The Findings Tentative Interpretations Summary Women and men physicians in Michigan in 1986 demonstrated different ranges of professional activities. The body of physicians had graduated from one of three types of medical schools (Table A-ll), USA MD, international MD or a USA DO school. Some differences were detected for these three groups of medical education. However each of these three training groups had differences within them for the proportion and distribution of women (Figure 16) and men, for age and years since graduation (Table A-21), and geographical location (Tables A-17, A-18 and A-19, and Figures 10 and 16). Specialty had the largest influence on professional activities. Primary specialists were more likely to be rural (Table A-21) and work fewer hours as well as proportionately fewer in hospitals (Table A-34). As both women (Tables A-12 and A-16) and DOs (Table A-31) were more likely to be in primary specialties, some of the observed differences in hours worked and office and hospital hours (Table A-20) may have been attributable to the specialty practiced. Those physicians in graduate medical education in 1986 were more in nonprimary specialties, for all men, women and DOs (Table A-16). Full time, fully trained physicians were selected as the ultimate unit of measurement, to eliminate the bias 220 introduced from the uneven distribution of part time physicians, for women of all ages and for older men (Tables 6, A-14, A-15 and A-34). Further, for some variables such as geographical location and hours worked, those still in graduate medical education were excluded (Table A-15), as their hours were longer, and 99% of their locations were in metropolitan counties. In addition, while almost twice the proportion of women MDs were in training than were men MDs, over twice the proportion of women DOs were in training than were men DOs (Table A-16). Profiles of physicians are in Appendix K, namely women, men, rural physicians, USA and International MDs, and DOs. Patient Care Hospital Care Hospital care (Table A-14 and A-35) was provided by 91% of the active physicians of whom 81% were MDs and 19% were DOs. Women were 12%, and 18% of women in hospitals were still in graduate medical education. Seventy eight percent were USA trained, under representing the international graduates. The mean year of birth was 1942 and the year of graduation was 1969, and 51% had graduated since 1970. percent were in rural counties. Ten Sixty two percent were in specialties, with (Appendix I) medical and surgical leading. If providing hospital care, the total patient care hours per week were 52.4, with hospital hours dominating (32.2) and 20.2 hours in nonhospital settings. 221 Nineteen percent of all active physicians were practising solely in a hospital setting (Table A-34). Fourteen percent of the latter were women. specialties dominated. Nonprimary This was also the most metropolitan based practice pattern. Nonhospital or Office Care Office care (Tables A-14 and A-35) was provided by 81% of the active physicians, of which 80% were MDs and 20% were DOs, with a total of 81% USA trained. still in graduate medical education. Sixteen percent were Women were 12%. The mean year of birth was 1940 and year of graduation was 1968, although 53% had graduated since 1970. were the choice of 90%. Metropolitan sites Fifty six percent were in specialties, with (Appendix I) medical and surgical leading. The total patient care hours worked per week were significantly less than for those who predominately provided hospital care, 51.0 total per week (26.3 hours in office and 24.6 in hospital care). Nine percent of all active physicians were in office only practice, 16% of them women, making this the highest density of women in a practice type. There were also proportionately more primary care physicians, more DOs and three times as many early as recent graduates. Forty or More Patient Care Hours Per Week Seventy four percent of the active physicians provided 40 or more hours per week of patient care, including those in graduate medical education, and excluding physicians 222 reporting over 80 hours over week. Women were 11%. MDs were 80% and 20% were DOs, with a total of 19% internationally trained MDs. The mean year of birth was 1943 and year of graduation was 1970, 53% had graduated since 1970, and 21% were in graduate medical education. Sixty two percent were in specialties with (Appendix I) medical and surgical specialties predominating, and 38% in primary care. Ten percent were in rural counties. Their mean hours of patient care were 57.6, with hospital hours almost 50% higher (34.5) than office hours (23.0). Less Than Forty Patient Care Hours Per Week Physicians who provided fewer than 40 hours of patient care per week were 26% of the active physicians. 17%. Women were MDs were 81% with a total of 27% internationally trained and 73% USA trained, and DOs were 19%. Their mean year of birth was 1935 and year of graduation was 1962, with 66% graduated before 1971. They were older and earlier graduates, and more were international MDs than those working more hours. education. A total of 5% were in graduate medical Ninety one percent were in metropolitan and 9% in rural counties. Fifty five percent were in specialty care with (Appendix I) medical and other specialties dominating, and 45% were in primary care. The total patient care hours worked were 30.8, of which 16.6 were in hospital and 14.2 hours were in nonhospital settings. 223 Primary Care Physicians Fourteen percent of primary care physicians (Table 8) were women. Their mean year of birth was 1939 and year of graduation was 1966 (Table A-29), so they were older and earlier graduates, as 55% had graduated before 1971 (Table A-30). MDs were 74% and DOs were 16%, with a total of 80% USA trained, and 8% were in graduate medical education. Only 86% practiced in the metropolitan counties and 16% in rural counties (Table A-27). Within the primary care specialties (Appendix I), 45% were in family and general practice, 41% in the medical specialties, and 14% in surgical specialties. They worked a total of 49.1 patient care hours per week, with office hours 50% higher (29.7) than hospital hours (19.4). Nonprimarv Physicians Nonprimary specialists were 12% women (Table 8), and 84% MD and 16% DO so there were fewer DOs in this category. Internationally trained MDs were 22% of the total. Twenty two percent were still in graduate medical education. The mean year of birth was 1942 and the mean year of graduation was 1969 (Table A-29), with 51% graduated since 1970. Only 7% were in rural areas and 93% were in metropolitan areas (Table A-27). One possible influence was the large percentage in graduate medical education in a metropolitan location. Another influence was the pattern of specialists to practice where there was a larger population of patients, to ensure that they maintain their skills. 224 The dominant specialties (Appendix I) were 48.4% in other specialties, 33.2% were in surgical specialties, and 18.4% were in medical specialties. The total patient care hours worked by nonprimary specialists were 51.6 per week, more than twice as many in hospital settings (35.7) as in nonhospital settings (15.9). Specialties With reference to the AMA's broad categories of specialty (Appendix I), each of the four specialty groups were examined (Tables A-23, A-24, A-32 through A-37) and summarized in Appendix L. Trends Over Time The trends noted over time within the 1986 Michigan physician supply were that since 1970 there were more women who were trained in the USA, 81% of all MD and 95% of all DO women (Footnote, Table A-ll). These recent women MD physicians were more likely to have entered a nonprimary care specialty (Table A-23), and equally likely to have a mixed office and hospital practice as their male counterparts. The trend to train in nonprimary specialties was followed by both women and men physicians (see Figure 15, a profile of full time recent graduates). The movement of physicians into rural areas increased among the recent men and women physicians (Table A-19, A-28, and Figure 18). The differences lessened between women and men in specialty, type of practice (Table A-34), hours 225 worked (Table A-24), inactivity (Table A-26) and geographic location (Figure 18). Specialty and Geographical Location The specialty of a physician was a possible predictor of activity in Michigan in 1986, the following relationships (summarized in Table A-32) emerged. There were 8 active men for every active woman physician in Michigan in 1986. In rural counties there were 12 men to every woman physician. Women surgical specialists were rare (1 to 15 men), and rarer in rural areas, 1 to 68 men surgeons. Overall, women were more represented in medical and other specialties, but there was only 1 woman to every 9 men in the in these specialties in rural areas in Michigan in 1986. Specialty and Practice Type Specialty was a possible determinant of the professional activity of a physician in Michigan in 1986. The literature confirms this direction of influence of specialty over the professional activity, hours worked (and income and status, not variables of this study). Specialty and Hospital Practice There were two different patterns of specialty for gender that led to hospital practice activities. At least 5% of active women physicians practicing some time in hospital settings were in the following specialties (rank ordered): internal medicine, anesthesiology, pediatrics, emergency medicine, psychiatry, pathology, radiology, and obstetrics and gynecology. At least 5% of 226 active men physicians practising some time in hospital settings were in the following specialties (rank ordered): cardiovascular diseases, emergency medicine and internal medicine tied for second, radiology, pathology, and general surgery. Specialty and Office Practice There were overlapping but differing patterns of specialty for gender that led to office practice activities. At least 5% of active women physicians practicing some time in nonhospital (office, clinic, nursing home) settings were in the following specialties (rank ordered): family and general practice, psychiatry, internal medicine, and pediatrics. At least 5% of active men physicians practising some time in nonhospital settings were in the following specialties (rank ordered): family and general practice, psychiatry, occupational medicine, internal medicine, and dermatology. Uncontrolled Variables There were two potentially confounding variables, and two sources of error. Year of birth and year of medical school graduation were possibly confounding variables. It appeared that younger physicians and more recent medical school graduates practiced differently from older physicians and earlier graduates, for both men and women, and MDs and DOs. Systematic bias may have operated on the self-report of specialty or hours worked as a measurement error for gender 227 (Bauder-Nishita, 1980) in opposite directions for the hours men and women reported that they worked. Women may have under reported and men over reported their hours. OHMA did not employ any methods to detect bias or faking. The other source of error was the loss of any subjects (Spector, 1981) without a verifiable, accurate and complete AMA or AOA record for the requested variables, and an identical OHMA record. Loss was less of an error than inclusion of incomplete or inaccurate information on physicians. Relationship Between Findings and Other Researchers Increase in Enrollment of Women The data in this study mirrored the national trends of annually increasing proportions of women entering USA medical schools after 1970 (AAMC, 1988a; AOA, 1989; Dube, 1973a; Levin, 1980; Spieler, 1977; Walsh, 1977; and Tables A-l, A-2 and A-4). Eighty one percent of the USA MD women in Michigan in 1986 had graduated since 1970 (footnote Table A-ll), and 95% of the DO women. Correspondingly, 52% of fully trained internationally trained women MDs practicing in Michigan in 1986 had graduated since 1970, and 48% before 1971. There were fewer international graduates arriving or remaining in Michigan (MCGME, 1987). Lack of Role Models. Mentors and Sponsors Leserman (1981) wrote that "women are to some extent pioneers in a male profession" (p. 38). There were no women in the Michigan study group in 1986 in the specialties of 228 (Table A-12) aerospace medicine, colon and rectal surgery, thoracic surgery or pediatric allergy. There were fewer than 5 women with their principal specialty in 12 of the 37 specialties listed in Table A-12. This confirmed the uneven availability of women physicians to be role models, as well as mentors and sponsors for medical students and trainees (Gruppen & D. R. Brown, 1981; Hsu, 1986; S. R. Kaplan, 1982; McNamara, 1985; Nadelson, 1983; Nadelson & Notman, 1972; Roeske & K. Lake, 1977; I. K. Smith, et al, 1984; Weisman, 1984). Women in Specialties The women in the study group were not in all specialties in Michigan in 1986 (Table A-12), unlike theUSA MDs at large (Table A-5; AMA, 1987). The seven specialties practiced most by women in Michigan in 1986 were (Table A12) family and general practice, internal medicine, pediatrics, child and adult psychiatry, and obstetrics and gynecology. This was comparable with the 1986 national distribution of women MDs in primary care (Table A-5). This also reflected the specialty distribution of women in the 1950s (Dykman & Stalnaker, 1957), the 1960s (Lopate, 1968; Powers et al, 1969), the 1970s (Bobula, 1980a, 1980b; Callan & Klipstein, 1981; Cartwright, 1977a, 1977b; Cohen & Korper, 1976a, 1976b; A. Goldblatt & P. B. Goldblatt, 1976; Harward et al, 1981), and in the 1980s (Mitchell, 1984; Schermerhorn et al, 1986; Silberger et al, 1987). Golden al (1989) compared trends over the 1970s and 1980s which et 229 illustrated broader distribution by women in recent years, also demonstrated in the study group. The seven specialties practiced most by women were the same in 1986 as in 1967, as though some constants operated in specialization by women physicians (Lopate, 1968; Powers, et al, 1969; Spieler, 1977). Women were in subspecialty fellowships 35% less than men (Table A-15). This confirmed a national trend (Kohrman, Lyttle, Andersen, & G. S. Levey, 1989; Table A-5; A. P. Williams, Domnick-Pierre, Vayda, Stevenson, & Burke, 1990). Four fifths of women MDs in graduate medical education in Michigan were in nonprimary specialties with no significant differences from men in selection of primary or nonprimary specialties (Table A-16). This was a reversal of the national trend for over half of women to be in primary care training (Table A-7; AMA, 1987). However, women second year and higher residents were still underrepresented in the surgical specialties (American College of Surgeons, 1988; Dalessandri, 1988; Ramos & Feiner, 1989; Wright, 1967). The trends supported Wunderman's (1980) interpretation that the data reflected the cumulative effect of the past choices by and prospects for women in medicine. The trends by graduation year (Table 11 and Figure 13) indicated that the percentage of women in primary care decreased in recent years. Forty four percent of all international MDs, who were 50% more likely to be earlier graduates than USA MDs (Table 230 12) were in primary care. Thirty eight percent of all USA graduates were in primary care, lowered again to 35% of recent (after 1970) graduates. Only 21% of those in graduate medical education were in primary care, or 1.8 times less than women MDs in practice. DO women showed the same rate of change where trainees were also 1.8 times more likely to be in nonprimary care specialties than DO women in practice. Restrictions (Perun & Bielby, 1981) did appear to have lifted for women in Michigan by 1986. Present Trends Women physicians were entering primary care at the same rate as men (Table 12). This was a reversal of prior trends (AAMC, 1988; Babbott, et al, 1989; Lieu, Schroeder, & Altman, 1989). It supported the argument that convergence in specialty selection between women and men was occurring (E. K. Adams & Bazzoli, 1986; Ferrier & Woodward, 1982; Kutner & Brogan, 1980; Maheux, Dufort & Beland, 1988). Practice Type Hospital based practice was still more common for women than men in Michigan in 1986 (Tables 4 and A-35; Davidson, 1979; Heins et al, 1978; Spieler, 1977). The study did not permit comparisons between private or salaried practice. Patient Hours Worked Per Week This study confirmed that hours per week were lower for women, except for surgery, when comparing full time, fully trained physicians (Table A-33). More authors in recent years have distinguished between full and part time 231 physicians (Bobula, 1980a, 1980b; Callan & Klipstein, 1981; Dennis, et al, 1990; Hojat, et al, 1990; M. J. Lanska, D. J. Lanska, & Rimm, 1984; Mitchell, 1984; W. B. Schwartz, Sloan, & Mendelson, 1988). The over representation of women as part time patient care providers has been recognized as a methodological issue in the research on productivity of physicians. Early studies found larger differences between the hours worked by women and men. Bobula (1980a, 1980b) found a 7.2 hour higher work week reported in 1977 and 1978 by 4,465 men than for 972 women. Using national 1978 and 1979 data Mitchell found only a 3 hour gap in patient care hours for gender. Curry (1983) studied 12,071 Maritime Canadian physicians, 16% of them women, and found one hour difference for total hours, and 4.1 hours difference for patient care, close to this study group's results. The Michigan figures were lower than the national hours per week, compared with Silberger et al's (1987) analysis of 1986 AMA data with a 5.6 hour gender gap. The study group demonstrated convergence of hours between genders, with a 3.3 hour difference. There were no significant differences in hours worked by those solely working in nonhospital settings. It may be that the practice environment shaped the number of hours worked, as hours did not differ between genders, or specialties. 232 Other productivity measures, such as weeks per year and patients per year were not OHMA survey questions. No data was available on the number of dependents to analyze productivity correlates as noted in the literature. Part Time Practice In the OHMA survey, physicians designated themselves as part time (35 hours or less) or full time. Among the study group, 16.5% of women were part time, and 10.5% of men. Part time was likely throughout the women's age groups, but was primarily a status for older men physicians (Figure 9). This was the same direction but a smaller difference noted by Elston (1977), Elliot (1981), Offenbach (1982) and Hojat et al (1987). Part time hours were statistically higher for women than for men (Table 7), a finding not reported in the literature. Inactivity In this study group, women and men were equally inactive, 10.1% for women and 10.4% for men (Tables A-25 and A-26). AMA 1985 data showed that 7.4% women and 8.4% men were inactive (Table 3; AMA, 1987). Both of these findings, Michigan and national illustrated the belief and observation of writers than women were active more than previously believed (Dykman & Stalnaker, 1957; M. A. Elston, 1977; Ferrier & Woodward, 1982; Geyman, 1980; Heins, et al, 1976; Heins, Smock, L. Martindale, Jacobs & Stein, 1977; Heins, Smock, L. Martindale, Stein and Jacobs (1977) ; Lerner, 1981). 233 The pattern of activity was notable (Table A-38, Figure 19) as both women and men in their first decade since graduating from medical school had a reduction in activity, but women were statistically significantly more active (p=.03). This would be interpretable for women as the time in which child bearing and rearing could reduce their participation in medicine. It was not apparent what would reduce the men's activity at this particular time in their career. For both men and women there was a rise in activity levels in their second decade group, still significantly higher for women (p=.00). The activity rates for the 21st and subsequent years after graduation were no different for women and men, confirming the continuity of medical careers for women and men. Length of Career The measurement of length of career was approximated using the years since graduation from medical school. This was not considered an accurate measure of length of career, due to the small number and proportion of women from earlier graduating years (157 women or 4.9% of those who graduated 30 years previously), and the absence of intervening years' data. Inactivity was at the same level for women and men earlier graduates (Table A-26) so women's activity rate had not waned as their career lengthened (Batchelor, 1990; 234 Heins, Smock, L. Martindale, Jacobs and Stein, 1977; Morgan, 1971; J. H. Watson, 1977). The unique data found in this study was the productivity measure of patient care hours per week. While women had lower hours per week throughout their careers (Table A-39 and Figure 20), they actually increased their working hours in their second decade of practice. Men had a slow decline in hours. Convergence in Hours In this study, women and men worked similar but still statistically significantly different hours, throughout their career decades (Table 39). Men's hours were lower and women's hours higher than found in previous studies (Silberger, et al, 1987). This indicated a change in the relative hours worked by women and men physicians as found by Curry (1983), Freiman and Marder (1984), Mitchell (1984), Bowman and M. L. Gross (1986), Hooper (1989) and Woodward, et al, (1990). Geographic Distribution Women physicians were more likely than men to practice in a metropolitan setting (Tables A-17, A-18, A-19, and A27) confirming the research to date (Maheux, Dufort & Beland, 1988; S. C. Martin, et al, 1988; Woodward, 1986). However, women physicians have been working more frequently in rural areas in recent decades (D'Elia & I. Johnson, 1980), measurably so for women who were recent (after 1970) graduates (Tables A-28, A-31, A-33: Figure 18). 235 DOs were more likely to practice in a rural area (Table A31) as noted by Denslow et al (1984). Aae. Year of Medical School Graduation Men and women physicians who were younger, and who were more recent medical school graduates were choosing similar specialties (Tables A-23 and A-30). This mirrored the findings of other researchers (E. K. Adams and Bazzoli, 1986; Ferrier & Woodward, 1982; Kutner & Brogan, 1980; Maheux, Dufort & Beland, 1988; R. H. Rosen, et al, 1981). Tables A-19 and A-31 showed that recent graduates had chosen to practice more in rural areas, as reported by Kutner and Brogan (1980), R. H. Rosen, et al (1981), E. K. Adams and Bazzoli (1986), and Reiman (1989). USA and International MDs In the study group, women MD graduates of international schools were more common than USA trained women MDs (Tables A-ll). This was a proportional difference noted by other researchers (AMA, 1989; Bowman & M. L. Gross, 1986; Mejia, et al, 1980). The specialization of USA and international MDs differed (p=.00, Table A-38), as recorded previously (A. Goldblatt & P. B. Goldblatt, 1976; Mick & Worobey, 1984; Worobey & Mick, 1987). International women graduates were more distributed in metropolitan counties than USA trained MD women (p=.06, Table A-17). They were older and earlier graduates (Table A-21) reflecting their history as immigrants during the shortage periods in the 1960s and 1970s, and a reducing 236 number arriving and staying in the USA in the last decade (Fulop, 1986a). These differences will have less influence in future decades, as the proportion of USA trained physicians increases. Theoretical Implications Specialties Analysis of the specialty of physicians revealed that the renowned primary care practice by women has been a strong historic tendency (Hypotheses 1, 2, 3, 4). it has weakened in recent decades. However The primary care group of women (Hypothesis 2, Table 8) of specialties were only 2.6% more overrepresented compared with men. However the absence of women in many specialties (Hypothesis 3, Table A12) weakened the conclusiveness of the statistical test which found the difference to be significant. Recent (after 1970) graduates had an alternating pattern between the specialties, women more in medicine and other specialties, and men more in surgical and family/general practice specialties (Hypothesis 5, Table A-23). The implication of these findings were that assumptions about women physicians need to be revised to fit the changing decisions by an increasing number of women physicians. They were more likely to be specialists and to practice more in rural counties. These changes should be reflected in recruitment of medical students, and creation 237 of practice opportunities for women. These factors should be reflected in the methodology of future research. Hours Women had worked fewer hours than men physicians (Hypotheses 6 and 7). In 1986 in Michigan, women and men surgeons worked the same hours (Table 9, A-24). The methodological issues for determining comparable hours for women and men (chapter III) were resolved by comparing full time and fully trained physicians with each other. Those still in graduate medical education had typically worked longer hours (Table 7) so they were excluded from study of hours, as 26% of women were in training in this database, and could skew the hours upward. Women and men worked within 3.5 hours of each other each week, a statistical difference (Table A-20). Women's "mean" hours rose significantly when part time women physicians were excluded from the analysis. The difference was maintained across the years of professional experience (Table A-39, Figure 20). The interplay of specialty and hours was an essential analysis. The mean hours per week of each specialty varies over 4.2 hours (Table 9). As women had different specialty distribution from men, some of the hour differential may had been attributable to their specialty rather than a gender factor. 238 Part Time Physicians The self-described part time physician was found working in one site only, either hospital or office, but rarely both, perhaps due to the limited number of hours available to work. It may had been a bigger factor than gender in influencing the hours contributed to medical practice, along with specialty. The implication of these findings were that women will provide patient care to within 5% to 8% of men. Employers, physician service payors, and the Federal Government should take this almost equal work week into account for medical workforce planning and payment structures. Practice Type Medical work was different for women and men. Women practiced more in hospitals only and less in mixed office and hospital care, even if full time and fully trained. Practice type has been attributed to gender. also inextricably linked to specialty. influence the hours worked. It was Both factors Hospital-only practice was major for "other" specialties, family and general medicine was the leading specialty in office-only practice, and the medical and surgical specialties were the major specialties in the mixed office and hospital practice style. However, no gender differences were found for physicians in office-only practice. One conclusion was that this practice was determined by its setting. It is possible that gender differences in hospital-only and the mixed 239 practice types were related to the historical specialty separation for gender, not to gender per se. Hours were the same for surgery, also supporting the supposition that gender did not determine hour and practice differences, but situational factors in the practice of specialties. The implications were that practice differences subsequent to the specialty practiced may create differences previously attributed to gender, such as location and hours of patient care. MD and DO The most rapid changes among women MDs were summarized in Table 11. The primary care proportions reduced from 45% of all international women MDs who were fully trained down to 21% of women MDs currently in training (both USA and internationally trained). In Table 12, USA women graduates dominate the pool of future women physicians. One implication from these data were that USA women MDs should be the basis for prediction of women physicians' practice patterns in the 21st century. There were also changes noted among DO women as those in graduate medical education halved the two-thirds primary care pattern to one-third (Table 11, Hypotheses 36 and 37). This will have little impact on the larger medical supply predictions, as DOs were 4% of the practicing doctors in the nation (COGME). 240 Geographical Location More recent (after 1970) graduates were more broadly distributed between metropolitan and rural sites, than the greater body of physicians. All physicians in graduate medical education were excluded as the 99% of training sites in Michigan were in metropolitan areas. Hypotheses 30, 31, 32, 33, 34, 35 and Tables A-17, A-18, A-19, A-27 and A-28 confirm the preponderance of metropolitan location for Michigan's physicians. There was a variation of location with specialty, and hence with gender and professional degree as women and DOs were historically more in primary care and less in nonprimary specialties. Analyzing location for recent graduates (Tables A-28, A-30 and Figure 15), more women and more total graduates of the 1970s and 1980s had located significantly more in rural sites, for all specialty categories (Figure 18). The implication of this improved distribution is to inform the medical workforce planners of these trends so they can adjust their programs to recruit rural physicians. Further research is warranted to determine the factors that have broadened the distribution of women and men physicians in Michigan. The second implication of these findings was that the distribution of physicians in rural counties should improve for all specialties, for better health care access for rural residents of Michigan. 241 Recognized Limitations of This Study This statewide study had the largest number of women physicians directly responding since the 1978 survey in California (Bauder-Nishita, 1980) which reported from 3,736 nonfederal women physicians. This was the first study of women DOs. National studies were larger, by conducting secondary analyses of federal data such as HCFA (Mitchell, 1984) and AMA Physician Masterfile (AMA 1987, 1988), without direct responses from all the women physicians. There were two types of limitations. The way the data was collected by each of OHMA, AMA and AOA, did not permit in-depth analyses of questions previously raised in the literature. The second limitation was the way the data base was created which rejected up to one third of instate records due to the criterion of an imperfect match between records. The major difference was the methodological approach of this study compared with that of other statewide studies. The combination of MDs and DOs was a first in the research on women physicians. The combination of data from a statewide survey with national data files was a further innovation, expanding the information from each of the sources. The opportunity to analyze change across time since graduation from medical school was a strength in the present study that better approximates years of practice than physician age. 242 The results of this study have expanded the knowledge of women and men physicians both by the inclusion of DOs and the confirmation of the trends found by others' research. Conclusions Of The Study Women physicians in Michigan in 1986 had different professional activities, specialty, hours per week, and geographic location whether MD or DO. The differences were smaller between recent medical school graduates. There were fewer distinct differences for this population in hours of patient care, specialty practiced and geographical location. Changes for both women and men appear to be toward a genderindependent norm. It was clear from this study that observed differences that have been reported from dozens of studies of women and men physicians were not immutable facts of medicine. They may have been an outcome of underlying forces in medicine that interacted with male/female forces in medical education and society at large, to mold women physicians' practice differently from men's. The observed differences between women and men practising in Michigan in 1986 were possibly based on the job or role, not on gender. The profile was very close for women and men physicians solely in office based care, in their specialty and hours. Again, the observation was made that women were more likely to provide patient care solely in a hospital setting. The rural women and men physicians 243 had a similar profile in specialty and hours, but again, women were less likely to be so-located, an observed difference. Mvths The perception that women were different physicians was persisting even as most measurable differences were narrow (Kasteler & Hulme, 1980; S. C. Martin, et al, 1988). There were still differences in the current physician workforce (Bowman & M. L. Gross, 1986). Primary care was no longer the only option or choice of MD women as 79% of women in training were in nonprimary care, compared with 81% of men. Women had been observed moving into previously under represented specialties (E. K. Adams & Bazzoli, 1986; Burnley & Burkett, 1986; Clavan & Robak, 1978; Nadelson, et al, 1981; Weisman, et al, 1980), just as Szkop-Frankiel (1967) had encouraged 19 years earlier. The length of career, matching graduates from the same decades was similar for women and men. Women were as professionally active as men, in less stereotypical practice patterns (Geyman, 1980). Sex differences in medical practice were diminishing such as in office only practice and surgery (Bergquist, et al, 1985; Table 4; Weisman, et al, 1980). For example, the dominance of women in hospital only practices reduced when part time and earlier graduates were excluded from the analysis. Women MDs who had graduated since 1970 were practising more in rural areas. 244 Women physicians' hours were increasing and men physicians' hours were decreasing (Curry, 1983; Freiman & Marder, 1984; Geyman, 1980; Mitchell, 1984; Weisman, et al, 1980; P. B. Williams, 1978). The difference of less than 10% hours worked by women may not be a meaningful difference, if practice characteristics of women such as efficiency (Roselund & Oski, 1967), patient satisfaction (Freidman, 1988) and lower malpractice rates (Holder, 1979) were considered as gains. Perhaps actual changes in norms were occurring as more women become physicians (Beck, 1988; Bluestone, 1990). The new norm would resemble that of USA trained MDs who had graduated since 1971: they were specializing in similar patterns, working full time and similar hours, and distributing in rural areas more than previously noted. Implications For Practice Women had increased their numbers, and their hours worked and hence the quantity of professional services from women increased. The increasing hours were from women who were USA trained MDs who had graduated since 1970. Specialty care was increasingly the choice of recent graduates, and it may have influenced the hours worked. More recent graduates, both women and men, had located in rural counties, which will increase the geographical distribution of physicians in Michigan. 245 The projection based on those physicians in graduate medical education in 1986 was that the physician workforce will increasingly be in nonprimary specialties. This represents a change for both women and men, and will have a major impact on the medical care available to the people of Michigan and the USA, as the proportion of primary care physicians reduces. The differences in professional activities of women and men physicians in Michigan in 1986 reflect the increasing access for women in medicine since 1970. Younger women physician were practising differently from older women physicians, in sheer numbers, as well as specialty, professional activities, hours worked, and geographical location. Implications For Medical Education Specialty was a key to the ultimate practice of a physician. The recommendation to educators was to encourage and enable women physicians to enter as broad a range of specialties to more equitably distribute women throughout the range of activities and locations of physicians in Michigan. Likewise for men, to better distribute men in specialties and functions. Beyond access and equity, medical educators need to redirect all physicians to promote a better supply of primary specialists in the 21st century, and to sustain the trend toward rural practice. 246 Further, educators need to expand their concerns for women beyond those of students and residents. A comprehensive approach to facilitate and integrate women into the faculty at higher levels and across all disciplines has yet to be realized (AAMC, 1990). Examination of the stereotypes and conseguent limited opportunities for and activities of women physicians was overdue. Now the questions of M. A. Elston (1977) and C. Eisenberg (1981) return: was this the result of long-term prejudice and attenuated opportunities? Greater access and equity for women in medicine have been detected for the woman MD, and similar trends were observed for the woman DO. Recommendations For Further Research Improvement Of This Study The method of this study, a secondary analysis of three archived surveys, allowed an enormous number of physicians' professional activities to be studied for one state, Michigan in one year, 1986. One recommendation was for increasing standardization of data collection by all survey processes, to that of the World Health Organization and the American Medical Association. OHMA modelled its criteria and coding on AMA but still varied from it. Loss of records could have been lower by matching the original L&R records with the AMA masterfile records. Use of existing databases was one way that the body of research on women physicians can be extended. In 247 particular, further analysis of the DO and the internationally trained MD physicians in Michigan would permit an understanding of the differing forces influencing such physicians and especially women. Their professional activities and their professional services provided to the people of Michigan have distinct characteristics that could be better understood for medical workforce planning. Methodological Issues These new findings should be incorporated into future research designs. Future research on physicians should study full time, fully trained physicians for meaningful comparisons between women and men, or any other comparison group. Inclusion of part time physicians who practice differently as well as less, or of residents and fellows who practice differently as well as more than fully trained physicians, have masked differences and trends in prior studies. For example, almost all of the graduate medical education in Michigan occurred in a metropolitan area, so inclusion of training physicians in a study of geographical location of physicians would oversample metropolitan physicians. When women were more likely to be in training than men, this over represents women's location as metropolitan. The second major recommendation is go beyond analysis of the current physician workforce, with its many variations. A new focus is needed on analyzing the patterns of graduate medical education to detect trends in changes in 248 specialty for women and men, and MDs and DOs. This information should then be weighted by the rate of change in each specialty for gender, whether increasing (women in surgery) or decreasing (women and men in internal medicine). Closer monitoring of the emerging medical workforce will allow better planning and design of graduate medical education to balance the specialty supply. The third major recommendation is to use year of medical school graduation to analyze trends over time. Age, the most common physician characteristic used to approximate experience or length of practice is not the same as the latter two. In addition an increasing proportion of older students were entering medical school in the 1980s, rendering age a less sensitive indicator of physician's professional history. Analyses by year of birth and by year of graduation in this study found the year of graduation more sensitive to changes. Fourth, contemporary research questions on patient care hours would be more fruitfully focussed on analysis of the narrowing of the gender hour gap and where and why the physician hours were becoming similar, rather than documenting differences and similarities. Fifth, consensus on a meaningful difference and an acceptable difference in patient care hours worked by women and men physicians should be developed. The difference range for full time fully trained women and men physicians 249 was between minus 0.6% to plus 8.2% of hours, only marginal differences. Sixth, previous research has emphasized that women have seen fewer patients per hour, and per week, and work fewer weeks per year. These productivity measures need further analysis for change. The increase in patient care hours by women need to be interpreted in the larger context of productivity measures (Shank, 1984). Seventh, outcome measures of medical care should be developed, such as patient health status and malpractice rates. These will describe physicians output in a more relevant manner than the number of patients served per hour, and may covary with the latter and with gender. Eighth, the interplay of practice type, subsequent to specialty practiced, needs further analysis by observational or time studies. Gender was not a single determinant of differences in practice, as office-only practice had no hour differences for gender, and surgery had no gender hour differences. These data did not provide the necessary information to separate contributing factors, nor did the source data (AMA, AOA or OHMA). Beyond Research The time has arrived for acceptance of the overriding similarities between women and men physicians by the medical profession, medical educators, and the public. Sustained affirmative action and equal opportunity for women in 250 medical education will allow medicine as well as women to continue to advance. APPENDICES Appendix A Appendix A Tables 251 252 Table A-l Women In MD Medical Schools. 1894-1986 Year Women Students % of all Students Women Graduates % of all Graduates a1894 1,419 * 144 * 1904 1,129 * 254 * 1909 921 * 162 * b1914-15 592 4.0 92 2.6 1919-20 818 5.9 122 4.0 1924-25 910 5.0 204 5.1 1929-20 955 4.4 204 4.5 1934-35 1,077 4.7 207 4.1 1939-50 1,145 5.4 253 5.0 1944-45 1,352 5.6 262 5.1 c1949-50 1,806 7.2 595 10.7 1954-55 1,537 5.4 345 4.9 1959-60 1,710 5.7 405 5.7 1964-65 2,503 7.7 503 6.8 1969-70 3, 390 9.0 700 8.4 1974-75 9,786 18.1 1,706 13.4 1979-80 16,141 25.3 3,497 23.1 1984-85 21,316 31.8 4,904 30.1 1986 22,100 33.4 5,107 32.3 253 Table A-l (continued) Note. aData for 1894-1909 are adapted with permission from Doctors Wanted: No Women Need Apply, p. 240, by Mary Roth Walsh, 1977, New Haven: Yale University Press, Copyright by Yale University Press. with permission. Adapted bData for 1914-1945 are adapted with permission from "Women Students in US Medical Schools: Past and Present Trends" by W. F. Dube, (1973a), Journal of Medical Education 48, (2) 186-189, Copyright 1973 by AAMC. Adapted with permission from AAMC. cData for 1949-1986 are from Women in Medicine Statistics by AAMC, 1988a, Copyright AAMC, Adapted with permission from AAMC. 254 Table A-2 Women In DO Medical Schools. 1956-1986 Year Women Students % of all Students Women Graduates % of all Graduates 1956 43 2.3 9 2.0 1959 38 2.0 9 2.1 1964 38 2.3 10 2.5 1969 59 3.0 12 2.8 1974 267 8.5 44 6.3 1979 837 18.2 192 18.1 1984 1,696 26.2 338 22.9 1986 1,857 28.0 406 25.0 Note. Data are adapted from Yearbook and Directory of Osteopathic Physicians 1987-1988. p. 521, American Osteopathic Association, 1989. Adapted with permission. Copyright Author. 255 Table A-3 MD Physicians' Activities In 1986 Women Total Activity MDs % All Men % Activity % All % Activity 83.6 84.8 92.3 86. 3 < 1.0 3.0 < 1.0 3.9 88.5 74.6 60.8 85.4 70.6 19.2 5.2 Federal Office Hospital aOther Subtotal % Total 1,221 16,200 4,517 21,938 3.5% < 1.0 2.9 < 1.0 3.5 16.4 15.2 7.7 13.7 Nonfederal Office Hospital aOther Subtotal % Total 325,757 118,948 39,107 544,308 96.5% 43.5 35.0 17.8 TOTAL 566,246 15.2 Note. 11.5 25.4 39.2 14.6 84.8 Data adapted from Physician Characteristics and Distribution in the US. 1987 Edition, pp. 74, 89, 90, 112. Copyright by American Medical Association 1987. Adapted with permission. a0ther includes Teaching, Administration, Research, and other (See Table 6). 256 Table A-4 Age Distribution Of MDs In The USA. 1986 Age Groups % of Women % of Men Men as % of Total < 35 41.4 25.2 22.1 76.8 35-44 31.6 16.9 30.0 89.1 45-54 12.6 10.7 19.0 89.3 55-64 7.3 7.8 15.3 92.2 65+ 7.1 7.5 15.6 92.5 Total Women as % of Total 84.8 15.2 n= 86 670 Note. 482 490 Data adapted from Phvsician Characteristics and Distribution in the US. 1987 Edition, do . 7-74. Copyright by American Medical Association 1987. Adapted with permission. 257 Table A-5 Specialty Of Patient Care MDs Bv Gender 1986 Women Specialty n Men % n % 8,724 10.6 56,698 11.8 Pediatrics 13,057 15.0 23,461 4.9 Internal Medicine 15,683 18.1 75,650 15.7 6,136 7.1 25,228 5.2 Family & General Obstetrics/Gynecology Total Primary (43 600) (50.3) (181 037) (37.5) Other Medical 2,991 3.5 31,817 6.6 Other Surgical 4,411 5.1 93,369 19.4 Psychiatry (all) 8,172 9.4 28,430 5.9 16,150 18.6 88,203 18.3 Unspecified 4,946 5.7 19,199 4.0 Inactive 6,400 7.3 40,435 8.4 Other Total 86,670 Note. 15.2 482,490 84.8 Data adapted from Physician Characteristics and Distribution in the US. 1987 Edition, pp. 19, 20, 25. Copyright by American Medical Association 1987. Adapted with permission. 258 Table A-6 Specialty Of Women MDs In Patient Care. 1986 Specialty Women MDs Total MDs Aerospace Allergy Anesthesiology Cardiovascular Child Psychiatry 34 144 3,978 668 1,236 655 1,481 23,146 14,157 3,878 5.2 9.7 17.2 4.7 31.9 * * 0.6 0.1 0.2 Colorectal Surgery Dermatology Diagnostic Radiology Emergency Medicine Forensic Pathology 22 1,235 1,989 1,594 65 835 6,822 13,870 12,297 346 2.6 18.1 14.3 12.9 18.8 * 0.2 0.3 0.3 * Gastroenterology 320 Gen. or Family Practice 8,724 Gen. Preventative Med. 197 General Surgery 2,009 Internal Medicine 15,683 6,469 67,687 937 37,214 91,333 4.9 34.6 21.0 5.4 17.2 * 1.5 * 0.4 2.8 Neurology Neurological Surgery Nuclear Medicine Obstetrics & Gynecology Occupational Medicine 1,205 104 167 6,136 235 8,264 4,126 1,315 31,364 2,678 14.6 2.5 12.2 19.6 8.8 0.2 * * 1.1 * 1,248 326 326 3, 320 13,057 15,180 17,659 7,577 15,515 36,518 8.2 1.8 4.3 21.2 35.8 0.6 2.3 69 152 1,004 243 6,936 373 868 3,531 4,185 32,724 18.5 17.5 28.4 5.8 21.2 * * 0.2 * 1.2 448 403 680 402 23 110 2,010 5,465 8,345 2,464 2,114 8,980 22.3 7.4 10.2 16.3 1.1 1.2 * Ophthalmology Orthopedic Surgery Otolaryngology Pathology Pediatrics Pediatric Allergy Pediatric Cardiology Physical Med. & Rehab. Plastic Surgery Psychiatry Public Health Pulmonary Radiology Radiological Oncology Thoracic Surgery Urological Surgery % Women in % Women Specialty of Total 0.2 * * * 0.1 * * * 259 Table A-6 (continued) Specialty Other Unspecified Unclassified Inactive Not Located Women MDs Total MDs % Women in % Women Specialty of Total 842 1,435 2,837 6,400 664 6,826 6,553 13,661 46,835 2,914 12.4 21.9 20.1 13.7 22.8 0.1 0.3 0.6 1.1 0.1 86,670 482,490 15.2 15.2 TOTAL Note. Data adapted from Physician Characteristics and Distribution in the US. 1987 Edition, pp. 19, 20, 25. Copyright by American Medical Association 1987. Adapted with permission. * = < 0.1%. 260 Table A-7 Specialty Of Women MDs In Practice And Training. 1986 % Women in Practice Specialty Category % Women in Training Primary Care Specialties General or Family Practice 11.3 9.7 Pediatrics 17.1 14.0 Internal Medicine 18.1 23.7 6.9 9.8 53.3 57.0 Obstetrics & Gynecology Total Primary Nonprimary Specialties All Psychiatry 10.0 10.8 Other Medicine 3.3 4.9 Surgery 4.6 8.0 20.5 19.0 Other Specialty Unspecified 8.3 n= 65 924 Note. - 20 746 Data adapted from Physician Characteristics and Distribution in the US. 1987 Edition, pp. 19, 20, 25. Copyright by American Medical Association 1987. Adapted with permission. 261 Table A-8 Specialty Of Women MDs In Training. 1986 n Women Residents Specialty % Women Residents % All Residents * = less than one tenth of one percent Allergy & Immunology Anesthesiology Cardiovascular Child Psychiatry Colorectal Surgery 72 758 51 283 5 0.3 3.7 0.3 1.4 * * 1.0 * 0.4 * Dermatology Dermatopathology Emergency Medicine Endocrinology Forensic Pathology 339 6 279 55 9 1.6 * 1.3 0.3 * 0.4 * 0.4 * * Gastroenterology General or Family Practice Gen. Preventative Medicine General Surgery Hematology 35 2,010 72 927 34 0.2 9.7 0.3 4.5 0.2 * 2.6 * 1.2 * Infectious Diseases Internal Medicine Medical Oncology Neonatal & Perinatal Nephrology 67 4,912 48 142 51 0.3 23.7 0.2 0.7 0.2 * 6.4 * 0.2 * Neurology Neurological Surgery Nuclear Medicine Obstetrics & Gynecology Ophthalmology 352 47 40 2,039 311 1.7 0.2 0.2 9.8 1.5 0.5 * 2.7 0.4 Orthopedic Surgery Otolaryngology Pathology Pediatrics Pediatric Cardiology 140 128 862 2,907 34 0.7 0.6 4.2 14.0 0.2 0.2 0.2 1.1 3.8 * Pediatric Surgery Physical Med. & Rehab. Plastic Surgery Psychiatry 7 218 51 1,956 * 1.1 0.2 9.4 0.3 * 2.5 * * 262 Table A-8 (continued) n Women Residents % Women Residents % All Residents Pulmonary Radiology, Diagnostic Radiology, Therapeutic Rheumatology Thoracic Surgery 45 725 125 39 8 0.2 3.5 0.6 0.2 * * 1.0 0.2 * Urological Surgery Vascular Surgery Transitional Year Others 45 1 342 169 0.2 * 1.6 0.8 * Specialty Total 20,746 Note. * * 0.4 0.2 27.0 Data from "Graduate Medical Education" by A. E. Crowley and S. I. Etzel, (1988), Journal of the American Medical Association. 260 (8), pp. 1093-1101. Copyright 1988 by American Medical Association. Adapted with permission. 263 Table A-9 Academic Rank Of Women (1978 And 1989) And Men (1989) Women Rank 1978 Professor 8.3% Associate Prof Men 1989 1989 9.3% 31.2% 17.9% 19.5% 25.5% Assistant Prof 41.6% 49.3% 34.0% Instructor 24.7% 17.8% 7.2% 4.1 2.1 12 613 50 031 Other/ Missing n= 7.5% 7 153 Note. Data adapted from AAMC Faculty Roster (1990c). Copyright Association of American Medical Colleges, 1990. Used with permission. 264 Table A-10 Gender And Age Of USA Medical School Faculty. 1988 Age % Male % Female 0.2 0.9 30-39 26.1 39.0 40-49 36.2 35.4 50-59 22.6 15.3 60-69 12.4 7.4 > 70 2.3 1.6 Missing 0.3 0.4 11 673 48 535 < 30 n= Note. Adapted from AAMC Faculty Roster System Numbers Book 1988. permission. Copyright AAMC 1988. Adapted with 265 Table A-ll Professional Degree. Nation Of Training Bv Gender Women n a,bMDs USA trained dTotal MDs e,fTotal DOs Total % n % 1 184 60.2 7 914 57.3 29 1.5 191 1.4 375 19.1 2 507 18.2 1 588 80.8 10 612 76.9 378 19.2 3 192 23.1 1 966 12.5 13 804 87.5 cMDs Canadian cMDs International Men Note: aAxnong USA MD women, 81% had graduated since 1970. bWithin USA trained physicians, women MDs were 11.5% of all MDs. cAmong international MD women, 48% had trained since 1970. dAmong MDs, 35% of all women were internationally trained and 24% of all men were internationally trained (p=.00). had trained since 1970. 10.0% of all DOs (p=.00). eAmong DO women, 91% fUSA trained women DOs were 266 Table A-12 Principal Specialty By Gender 0 0 .1 3.2 .6 1.0 .5 3.1 1.9 .3 .5 3.1 1.7 .4 Colon Rectal Surgery Dermatology Diagnostic Radiology Emergency Medicine Forensic Pathology n=0 0 1.1 1.0 2.6 .1 .3 1.2 1.0 2.5 .1 .2 1.2 1.0 2.5 .1 Gastroenterology General Practice General Preventative General Surgery Internal Medicine * .1 17.1 .1 1.7 14.1 .8 20.7 .1 6.4 12.9 .7 20.3 .1 5.8 13.1 .1 W 0 O to n=0 * * % Total U1 Aerospace Medicine Allergy Anesthesiology Cardiovascular Disease Child Psychiatry * a% Men H * n<5 a% Women H*00|-»W'O Specialty .9 .1 6.7 .5 .6 1.3 .1 5.2 .8 .5 1.2 .1 5.4 .8 Ophthalmology Orthopedic Surgery Otolaryngology Anatomic-Clinical Pathology Pediatrics 1.1 .5 .3 2.4 9.8 2.3 3.0 1.3 2.0 3.7 2.2 2.7 1.2 2.1 4.5 Pediatric Allergy n=0 * Pediatric Cardiology Physical Medicine - Rehab. * Plastic Surgery Psychiatry 0 0 0 n=0 * * H 1 to ^ CO to s ] M H O) .1 .7 .6 4.9 .3 .3 1.8 .1 0 .1 .1 .2 .7 2.9 .3 .5 1.6 .2 .2 .6 2.8 .3 .5 1.4 .2 ( 0 CO Ol to VOCTl'OM .6 .7 4.6 m * * .1 1.0 .1 7.2 to Public Health Pulmonology Radiology Therapeutic Radiology Thoracic Surgery Urological Surgery Other Specialty .1 Ul * Neurological Surgery Neurology * Nuclear Medicine Obstetrics and Gynecology Occupational Medicine 267 Table A-12 (continued) Specialty * n<5 a% Women a% Men % Total 24.1 15.4 16.4 1 966 13 804 15 770 Unspecified n= Note; ap = .00, significant differences in principal specialty for women and men. 268 Table A-13 Secondary Specialty By Gender Secondary Specialty % Women Aerospace Medicine Allergy Anesthesiology Cardiovascular Disease Child Psychiatry 0 0 0 0 Colon and Rectal Surgery Dermatology Diagnostic Radiology Emergency Medicine Forensic Pathology 0 0 0 Gastroenterology General Practice General Preventative Med. General Surgery Internal Medicine 0 Neurological Surgery Neurology Nuclear Medicine Obstetrics and Gynecology Occupational Medicine 0 0 0 Ophthalmology orthopedic Surgery Otolaryngology Anatomic / Clinical Path. Pediatrics 0 0 0 0 Pediatric Allergy Pediatric Cardiology Physical Medicine/ Rehab. Plastic Surgery Psychiatry 0 0 0 0 Public Health Pulmonology Radiology Therapeutic Radiology Thoracic Surgery Urological Surgery Other Specialty 0 0 0 0 0 0 0 % Men .1 % Total .1 .1 .1 .7 .1 .1 .2 .2 .7 .2 .1 .1 .1 0 .1 0 .2 0 0 .5 .1 0 .8 .9 0 0 .5 1.9 .1 1.3 4.2 .5 2.0 .1 1.3 4.6 0 0 .1 0 .2 .4 .5 .4 .2 .4 .7 .4 .1 .1 .1 .1 .1 .1 .3 .8 .2 .5 .3 0 0 0 0 .1 .1 .2 .8 .2 .7 .1 .1 .4 .2 0 .1 .4 .2 0 .1 .1 .3 .1 .1 .3 269 Table A-13 (continued) Secondary Specialty aUnspecified/ None Reported Total n= % Women % Men % Total 87.1 83.2 83.8 13 804 15 770 1 966 Note: ap=.00, women report secondary specialties significantly less than do men. 270 Table A-14 Active Physician Practice Bv Gender Men Women Total Practice Type %A11 %Full Time %A11 %Full Time %A11 %Full Time Hospital only a% All 44 33 28 24 31 25 27 29 20 20 21 21 12 10 10 7 10 7 17 12 11 7 11 8 Hospital + nonhospital % All 44 57 62 69 59 68 59 69 73 68 71 1 573 985 11 074 8 753 12 10 88 90 Fully Trained n 1 083 850 9 244 8 039 10 89 90 a% Fully Trained Nonhospital only a% All a% Fully Trained % Fully Trained Patient Care All n b% All b% Fully Trained 56 11 12 847 9 738 10 327 8 889 Practice Locations Of All Patient Care Physicians Median Range Note: ap=.00. 1 1-8 2 1-8 bnot significantly different. 2 1 - 8 271 Table A-15 Gender Differences In Training Status Of Active Physicians Status Women Total Men % % n .2 .1 2 .1 29.4 38.7 888 36.7 6.4 9.8 219 9.1 56.1 44.0 1 127 46.6 7.8 7.5 183 7.6 513 26.1 1 906 13.8 2 419 15.3 15.3 Active, Fully Trained n c% 1 254 64 10 464 76 11 718 74 74.3 Total Physicians n % 1 966 12.5 13 804 87.5 15 770 100 aIntern Resident bFellow aLimited License Other G.M.E. Total in Training n c% Note: % aPhysicians on limited licenses included in 1986 all first year residents and interns from USA schools, and second and third year internationally trained residents, and any other limited license holders. b0f the fellows, 41 were DOs, 4 (or 2% of the total) of them women. MDs were the remaining 147, and 27 (or 12% of the total) were women. cp=.00, significantly more women in graduate medical education than men. 272 Table A-16 Specialty Of Active Physicians In Practice And Training Bv Gender Women Practice aTraining Specialty Category Men Practice aTraining MDs b% Primary 47 21 39 19 b% Nonprimary 53 79 61 81 Total n 1 087 394 8 377 1 438 % 73 27 85 15 DOS b% Primary 65 33 58 b% Nonprimary 35 67 42 81 Total n 167 1191 2 087 466 % 58 42 82 18 Note: 19 a99% of all physicians in training had their principal location in metropolitan areas. bAll differences for practice and training in primary or nonprimary specialties were significant within gender, p=.00. 273 Table A-17 Population Based Location Bv Degree. Nation of Training And Gender Of Active Physicians County Location Women Men Total aActive MDs - USA Trained Metropolitan Rural, large city Rural, medium city Rural, no city Total n 94.1 1.6 3.6 0.1 927 89.3 1.9 7.8 0.8 7 148 89.8 1.9 7.4 0.9 8 075 bActive MDs - Internationally Trained Metropolitan Rural, large city Rural, medium city Rural, no city Total n 95.9 1.2 2.5 0.4 518 92.8 1.4 5.0 0.7 2 417 93.4 1.4 4.6 0.7 2 935 aTotal Active MDS Metropolitan Rural, large city Rural, medium city Rural, no city Total n 94.7 1.5 7.1 0.6 1 445 90.2 1.8 3.2 0.9 9 565 90.8 1.7 6.6 0.9 11 010 88.4 1.2 8.3 2.1 2 389 88.6 1.4 8.0 2.0 2 653 cActive DOs Metropolitan Rural, large city Rural, medium city Rural, no city Total n 90.2 2.7 5.3 1.9 264 aTotal Active Physicians Metropolitan Rural, large city Rural, medium city Rural, no city Total n % 94.0 1.6 3.5 0.8 1 709 12.5 89.8 1.7 7.4 1.1 11 955 87.5 90.4 1.7 6.9 1.1 13 664 100.0 274 Table A-17 (continued) Note; aGender differences significant to p=.00 for MDs who were USA trained, the total of all MDs, and the grand total. bGender differences for internationally trained MDs' location were significant to p=.06. cGender differences for DOs* location were significant to p=.09. 275 Table A-18 Metropolitan Contiguity By Degree And Gender Of Active. Fullv Trained Physicians County Location Women Total Men aActive MDs Metropolitan Rura1, adj acent Rural Total MDs n 93.3 2.9 3.8 88.7 3.7 7.6 89.2 3.6 7.2 1 053 8 126 9 179 85.3 8.7 6.0 86.2 6.6 7.2 86.1 6.7 7.1 150 1 938 2 088 bActive DOs Metropolitan Rural, adjacent Rural Total DOs aTotal Active Physicians Metropolitan Rural, adjacent Rural Total n % Note; 92.3 3.7 4.1 88.2 4.2 7.6 88.6 4.2 7.2 1 203 10.7 10 064 89.3 11 267 100.0 aGender differences for MDs and the total physicians were significant, p=.00. bGender differences for location of DOs were not significant, p = .55. 276 Table A-19 Metropolitan or Rural Location Bv Degree. Gender And Graduation Period Of Active Physicians Location Women Early Men Recent Early Recent aMDs Metropolitan 94.8 94.8 95.5 90.5 Rural 5.2 5.2 5.5 9.5 Total MD n= 479 965 5 335 4 223 bDOs Metropolitan 88.9 87.9 88.9 90.2 Rural 11.1 9.8 11.1 12.1 27 235 1 148 1 239 Total DO n= aTotal Metropolitan 94. 5 93.9 89.9 89.9 Rural 5.5 6.1 10.1 10.1 Total n= 506 1 200 6 483 5 462 Note: aGender differences for location of MDs were all significant, as for the total of physicians (p=.00). bGender differences for location of DOs were not significant for either "early" (p=.99) or "recent" (p=.31) graduates. 277 Table A-20 Patient Care Hours For Full Time. Fullv Trained Physicians Bv Gender Patient Care Hours aHospital Only n % Gender Report Women Men Total 235 29 1 578 21 1 813 22 44.7 12.4 48.3 13.9 47.8 13.8 97 12 566 8 663 8 42.7 8.5 43.5 9.8 43.4 9.6 477 59 5 385 71 5 862 70 M SD 52.1 11.9 54.5 11.5 54.3 11.5 aHospital Portion M SD 20.6 13.8 23.3 14.7 23.0 14.7 bOffice Portion M SD 31.5 12.6 31.2 13.3 31.3 13.2 809 7 529 8 338 49.4 11.5 52.9 11.5 52.6 11.6 M SD bOffice Only n % Gender Report M SD aHospital + Office n % Gender Report Total Patient Care n aM SD 278 Table A-20 (continued) Note; aAll hospital and total mean hours reported were significantly different for gender (p=.00). bThere were no significant gender differences for hours spent in "office" patient care, whether "office only" (p=.46) or the office component of "hospital + office" (p=.63). 279 Table A-21 Years Of Birth And Graduation. Bv Gender aWomen Year n= Birth M SD Graduation M SD n= Birth M SD Graduation M SD MDs - USA Trained 1 184 aMen 7 914 1948.6 11.3 1939.2 14.7 1976.1 11.1 1966.0 14.9 MDs - Internationally Trained 404 2 698 1943.3 9.0 1939.7 10.2 1968.3 8.8 1965.6 10.0 378 3 192 1950.0 12.1 1941.3 12.9 1978.9 10.2 1969.5 12.8 DOS n= Birth M SD Graduation M SD Total n= bBirth M SD bGraduation M SD 1 966 13 804 1947.2 11.1 1939.8 13.6 1974.4 11.1 1966.7 13.7 Note: aAll gender differences were significant, p=.00. Correlation of Years of Birth and Graduation =.95 for women and =.97 for men. 280 Table A-22 Historical Specialties Bv Gender Specialty Women % Total Men % n % aFamily Practice 17.1 20.7 3 199 20.3 aInternal Medicine 14.1 12.9 2 062 13.1 aPediatrics 9.8 3.7 709 4.5 aPsychiatry 8.1 4.9 840 5.3 50.9 57.7 8 960 56.8 1 966 13 804 15 770 aOther n= Note: ap=.00. 281 Table A-23 Specialty If Born After 1945. Or Graduated After 1970 Specialty Women Men Total Born after 1945 Family or General 17.4 18.7 18.5 Medical Specialties 24.1 22.1 22.5 9.8 17.2 15.8 18.4 17.3 17.5 Surgical Specialties Other Specialties Uncodable 25.7 n= 1 332 5 753 7 085 Graduated After 1970 Family or General 18.0 19.8 1 498 Medical Specialties 24.1 21.8 1 709 9.5 17.3 1 222 18.3 17.8 1 381 Surgical Specialties Other Specialties Uncodable n= 1 892 1 394 6 308 7 702 Note: All oender differences for specialty are significant, p=. 00, whether born after 1945 , and graduated after 1970. 282 Table A-24 And Graduation Specialty Category Women Men Full Time All All Full Time Total n Born After 1945 Family & General M Medical M aSurgical M Other M Total M n= 43.6 48.2 52.6 54.6 1 308 44.8 51.0 51.4 54.4 1 591 53.0 a55.2 55.8 a56.3 1 120 41.7 47.0 48.9 50.7 1 242 44.7 49.9 51.9 53.9 7 085 1 163 613 5 058 3 468 bGraduated After 1970 54.4 1 498 50.4 51.8 54.6 1 709 CO • 1 221 52.4 in n= 48.7 55.6 c56.0 1 222 47.2 48.8 50.6 1 381 49.8 51.9 53.8 7 702 653 5 562 3 892 u Family and General M 43.9 Medical M 45.0 cSurgical M 52.3 Other M 42.1 Total M 44.8 Note; aNot significant, p=.50. bF-test of the interactions between specialty and gender was significant only to p=.ll. cNot significant, p=.47. 283 Table A-25 Physician Inactivity Bv Year Of Birth And Gender Age Group Inactivity Status < 1946 > 1945 % Men 2.6 5.6 Active 29.6 52.7 Inactive 7.5 4.8 Active 60.3 36.9 1 966 13 803 10.1 199 10.4 1 435 89.9 1 767 89.6 12 368 Inactive n= Total Inactive n= Total Active n= % Women Note: There were no significant gender differences. 284 Table A-26 Physician Inactivity By Year Of Graduation And Gender Period of Graduation < 1971 > 1970 Inactivity Status a,b a,c % Women % Men Inactive 2.4 Active 26.5 49.0 Inactive 7.7 5.1 63.4 40.6 Active 1 964 13 801 Inactive 10.1 199 10.4 1 436 Active 89.9 1 765 89.6 12 365 n= Total n= 5.3 n= Note; aThere were no statistically significant gender differences. bActivity and year of graduation of women was significant to p=.08. cActivity and year of graduation of women was significant to p=.01. 285 Table A-27 And Gender Location Primary % Women Specialty % Men % Women % Men Metropolitan All n= Fully Trained n= 91.5 85.9 96.0 92.4 657 4 006 950 6 734 90.5 84.9 94.0 90.7 542 3 651 568 5 232 8.5 14.1 4.0 7.6 61 588 40 551 9.5 15.1 6.0 9.3 57 650 36 537 bRural All n= Fully Trained n= Note: aAll gender differences for metropolitan location, and location differences for specialty and gender, were significant, p=.00. bThe differences within the rural location, for all or only fully trained physicians were not significant, p=.22. 286 Table A-28 Location Of Fully Trained Women By Age. Year Of Graduation a,bAge County Location c'Graduation % Old % Young Metropolitan 93.0 91.7 94.4 90.7 Rural, adjacent 3.7 3.6 2.6 4.4 Rural 3.3 4.7 3.0 4.9 n= 533 662 494 699 % Early % Recent Note: aOld group born before 1946, and young after 1945. bAge differences in location were not significant, p=.49. cEarly group graduated before 1971 and recent group after 1970. dLocation was significantly different for graduation periods, p=.06 287 Table A-29 Specialty. Age. Years Since Graduation By Gender aWomen aMen bPrimary Care Age in Years M SD 39.4 10.7 47.7 13.5 46.7 13.4 Years Since Graduation M 12.4 SD 10.3 20.5 12.8 18.3 12.7 n= 733 4 727 5 460 bNonprimary Care Age in Years M SD 38.4 11.4 45.2 13.6 44.4 13.5 Years Since Graduation M 11.0 SD 11.2 18.5 13.8 18.3 13.7 7 377 8 376 38.8 11.1 46.2 13.6 45.3 13.5 Years Since Graduation M 11.6 SD 11.1 19.3 13.7 18.4 13.6 12 104 13 836 n= 999 Total Age in Years M SD Total n= 1 732 Note: aAll gender differences were significant, p=.00. bAll gender differences between primary and nonprimary specialty were significant, p=.00. 288 Table A-30 Percentage Of Graduates Of Each Decade In A Rural Location Bv Specialty And Gender Of Fully Trained Patient Care Physicians Decade of Graduation Women Primary Nonprimary Total n Total % Rural n Men Primary Nonprimary Total % Rural n % Total % Rural nRural 1920s 0 0 0 0 19 21 11 0 1930S 0 0 7 0 127 16 88 16 1940s 3 20 23 0 393 11 441 7 1950s 20 12 52 4 1 010 12 1 126 7 1960S 57 7 141 4 1 063 13 1 652 9 1970S 135 6 228 8 1 276 18 1 655 11 1980- 1986 247 17 61 10 401 22 255 13 Rural n= % Total n= 40 57 a9 .5 Note: aSignificance p=.00. 650 b6.1 490 a1 5 .2 a9 .4 bNo significant geographical distribution of women specialists, p=.23. 289 Table A-31 Percentage Of All Active MDs And DOs. Bv Specialty And Gender. Showing Those In A Rural Location Specialty Category DO MD % All Women % All Men % All Women % All Men 16.3 26.3 25.0 19.8 % Medical 6.3 8.9 10.0 5.6 % Surgical 3.4 11.1 0 9.1 % Other 7.6 8.9 13.0 10.2 % Total Rural 7.5 11.9 17.7 14.1 238 2 113 1 369 9 996 % Family/General Total n= Note: All gender differences within specialties and within degrees significant p=.00. 290 Table A-32 Active Physicians By Specialty And Location. Bv Gender Specialty Women to Men Women Men Total n Urban Rural Urban Rural 82% 18% 79% 21% 2 293 1:9 1:11 95% 5% 93% 7% 3 009 1:6 1:9 98% 2% 90% 10% 2 708 1:13 1:68 94% 6% 92% 8% 2 735 1:6 1:9 aTotal 93% 7% 89% 11% 11 267 1:8 1:12 1:8 1 187 93 8 877 1 110 Family /General Medical Surgical Other n= 1:9 1:6 1:15 1:6 Note: All differences for gender and location for each specialty were significant, p=.00. aTotal included uncodable specialties, not reported in the above categories. 291 Table A-33 Hours Bv Specialty And Gender Of Full Time. Fully Trained Physicians. By Gender Specialty Category Women Men Men - Women Hours % aFamily/General 48.4 52.7 + 4.3 + 8.2 aMedical 50.4 53.5 + 3.1 + 5.8 bSurgical 54.8 54.5 - 0.3 - 0.6 aOther 46.6 50.8 + 4.2 + 8.3 aTotal 49.6 52.9 + 3.3 + 6.2 809 7 529 n= Note: aGender differences significant for family/general, medical and other specialties, p=.00. bNo significant gender differences for surgery. 292 Table A-34 Practice Type and Characteristics of Active Physicians % Hospital % Office Only Only Gender % Women % Men % Hospital + Office Total % a14 86 a16 84 12 88 13 87 4 21 19 56 38 18 10 24 23 33 30 14 21 29 24 26 a20 80 b49 51 a44 56 39 61 44 56 75 25 50 50 52 48 Geographical Location % Metropolitan a92 % Rural 8 91 9 90 10 90 10 Degree3 % MD % DO 82 18 72 28 81 19 81 19 2 488 1 204 9 473 19 9 72 Specialty Category % Family/Gen % Medical % Surgical % Other % Primary % Nonprimary Graduation Period3 % < 1971 % > 1970 n= % Active Physicians Note; Significant differences, p=.00. differences, p=.l4. bNo significant 293 Table A-35 Hours For Practice Types bv Specialty And Gender For Full Time. Fullv Trained Physicians Women Men aHospitalbOf f ice Hospital aHospitalbOf fice Hospital Only Only Only + Office Only + Office Specialty Category Family/Gen n= Medical n= Surgical n= Other n= cTotal n= 43.5 44.3 50.6 48.2 43.6 54.3 12 38 112 54 224 1 329 46.6 42.6 52.0 50.6 44.3 54.3 55 21 225 196 83 1 795 52.1 45.0 55.4 56.1 40.0 54.5 12 2 109 204 38 1 844 46.6 45.0 48.6 50.0 42.1 53.3 139 33 73 1 001 199 733 c46.7 b4 3 .1 C51.9 c50.9 b44.3 c54.2 218 94 519 1 455 544 5 701 Note: aHospital only, two way differences for specialty and gender were not significant. bOffice only, no significant differences between specialties, or genders, or for the two-way interactions of hours for specialty with gender. cSignificant differences (p=.00) between specialties, and between genders within specialties. 294 Table A-36 Hours By Specialty And Gender For Full Time. Fully Trained Mixed Practice Physicians aSpecialty a,b,cTotal Office Houri 'c»dTotal Hospital Hours Category Women Men Women Men Family/ General 37.5 39.9 13.0 14.5 Medical 31.2 32.0 20.7 22.3 Surgical 30.5 26.8 27.7 27.7 Other 21.2 22.8 27.4 30.5 Total a31.0 a30.9 d21.0 d23.3 519 5 701 519 5 701 n= Note; aThe interactions between specialty and gender were not significant, F=.87). significant for gender, F=.66. bTotal office hours not cThere were significant F=.00 differences between specialties, and gender and specialty interactions. dTotal hospital hours were significantly different for each of gender and specialties, F=.00. 295 Table A-37 Distribution Of Specialty Bv Gender And Nation of Training Of Active MDs Specialty Category Women Men % USA % International % USA% International Family/ General 11 10 13 9 Medical 29 35 26 31 Surgical 12 11 26 25 Other 21 32 22 28 Uncodable 27 12 13 9 Total 74 26 80 20 931 520 7 170 2 428 n= Note: All differences for specialty by nation of training and gender were significant, p=.00. 296 Table A-38 Activity After Graduation bv Gender Years After Graduation n % Active % faTemporarily% Permanently Inactive Inactive bWomen 1 032 86.1 11.8 .6 ell-20 417 96.6 1.0 1.0 f21-30 222 95.5 .9 2.3 f31+ 157 80.3 1.3 11.5 dl-10 cMen dl-10 4 032 84.9 12.6 .2 ell-20 3 288 95.3 2.7 .2 f29-30 2 914 95.5 1.4 1.0 f31+ 3 074 82.4 1.8 11.6 Note: Percentaae do not add up to 100% as uncodable activity rates not listed. “Includes presumed temporarily inactive physicians. women = 35. ep=.00. bTotal uncodable cTotal uncodable men = 342. dp=.03 fNo significant difference in activity. 297 Table A-39 Hours Worked After Graduation Bv Gender Years After Graduation aMean Hours Patient Care Women Men Part Time Full Time Part Time Full Time 1-10 26.7 52.1 25.5 55.5 11-20 25.8 48.6 26.9 53.7 21-30 30.2 49.8 25.2 52.8 31+ 23.0 47.5 22.0 50.4 241 914 1 217 8 067 n= Note: aAll differences in mean hours were significant, p=.00, for the years since graduation, for full and part time, and for gender. Appendix B 298 Appendix B Michigan Department of Commerce Local Development Services Rural Development Office Classification of all Counties in the State of Michigan, 1990. Explanation of Terms Table B-l Classification of 83 Counties in Michigan as Metropolitan or Rural Metropolitan A metropolitan county has a city with a population over 50.000. There are 22 metropolitan counties in Michigan. Rural, with a large place A non-metropolitan county with a city or village with a population of 15,000 or more is in this category. There are four such counties in Michigan. Rural, with a midsized place This is a non-metropolitan county with at least one city or village with a population greater than 2,500 and less than 15.000. There are 35 such counties in Michigan. Completely rural This is a rural county with no place, city or village as large as 2,500 people. counties in Michigan. There are 22 completely rural 299 FTPS ZIPS This is the Federal Postal Service nationwide classification of counties numbered in alphabetical order, using odd numbers only. So the 83 counties in Michigan are numbered from 1 through 165. OHMA CODE This is the numbering of counties in alphabetical order, from 1 to 83, by the Office of Health and Medical Affairs. Classification of 83 Counties in Michigan with Reference to Metropolitan Proximity (Table B-2). Metropolitan Counties, of which there are 22. Rural Counties, adjacent to a metropolitan countv This refers to any of the three subclassifications of rural counties by population size. The county must be next to one of the 22 metropolitan counties. There are 19 rural counties adjacent to metropolitan counties in Michigan. Rural Counties, not adjacent to a metro county This refers to any of the three subclassifications of rural counties by population size. with any metropolitan county. The county must not border There are 42 rural counties not adjacent to a metropolitan county in Michigan. 300 Table B-l Classification of Michigan Counties as Metropolitan or Rural County Classification FIPS Code OHMA Code Metropolitan Bay Berrien Calhoun Clinton Eaton Genesee Ingham Jackson Kalamazoo Kent Lapeer Livingston Macomb Midland Monroe Muskegon Oakland Ottawa Saginaw St. Clair Washtenaw Wayne 17 21 25 37 45 49 65 75 77 81 87 93 99 111 115 121 125 139 145 147 161 163 9 11 13 19 23 25 33 38 39 41 44 47 50 56 58 61 63 70 73 74 81 82 73 91 103 155 37 46 52 78 Rural Counties with a Larger Place Isabella Lenawee Marquette Shiawassee 301 Table B-l (continued) County Classification FIPS Code OHMA Code Rural Counties with a Mid-sized Place Alger Allegan Alpena Barry Branch Cass Charlevoix Cheboygan Chippewa Clare Delta Dickenson Emmet Gogebic Grand Traverse Gratiot Hillsdale Houghton Huron Ionia Iosco Mackinac Manistee Mason Mecosta Menominee Montcalm Newaygo Otsego Presque Isle Schoolcraft St. Joseph Tuscola Van Buren Wexford 3 5 7 15 23 27 29 31 33 35 41 43 47 53 55 57 59 61 63 67 69 97 101 105 107 109 117 123 137 141 149 153 157 159 165 2 3 4 8 12 14 15 16 17 18 21 22 24 27 28 29 30 31 32 34 35 49 51 53 54 55 59 62 69 71 75 77 79 80 83 302 Table B-l (continued) County Classification FIPS Code OHMA Code Completely Rural Counties Alcona Antrim Arenac Baraga Benzie Crawford Gladwin Iron Kalkaska Keweenaw Lake Leelanau Luce Missauke Montmorency Oceana Ogemaw Ontonagon Osceola Oscoda Roscommon Sanilac 1 9 11 13 19 39 51 71 79 83 85 89 95 113 119 127 129 131 133 135 143 151 1 5 6 7 10 20 26 36 40 42 43 45 48 57 60 64 65 66 67 68 72 76 303 Table B-2 Adjacency of Michigan Rural Counties to Metropolitan Counties County Classification FIPS Code OHMA Code Adjacent Allegan Arenac Barry Branch Cass Gladwin Gratiot Hillsdale Ionia Isabella Lenawee Montcalm Newaygo Oceana St. Joseph Sanilac Shiawassee Tuscola Van Buren 5 11 15 23 27 51 57 59 67 73 91 117 123 127 149 151 155 157 159 3 6 8 12 14 26 29 30 34 37 46 59 62 64 75 76 78 79 80 304 Table B-2 (continued) County Classification FIPS Code OHMA Code Not Adjacent Alcona Alger Alpena Antrim Baraga Benzie Charlevoix Cheboygan Chippewa Clare Crawford Delta Dickenson Emmet Gogebic Grand Traverse Houghton Huron Iosco Iron Keweenaw Kalkaska Lake Leelanau Mackinac Manistee Marquette Mason Mecosta Menominee Missauke Montmorency Ogemaw Ontonagon Osceola Oscoda Otsego Presque Isle Roscommon Schoolcraft Wexford 1 3 7 9 13 19 29 31 33 35 39 41 43 47 53 55 61 63 69 71 79 83 85 89 97 101 103 105 107 109 113 119 129 131 133 135 137 141 143 153 165 1 2 4 5 7 10 15 16 17 18 20 21 22 24 27 28 31 32 35 36 40 42 43 45 49 51 52 53 54 55 57 60 65 66 67 68 69 71 72 77 83 Appendix C Appendix C Office of Health and Medical Affairs 1986 Physician Census Survey Questionnaire. GRADUATE MEDICAL EDUCATION • 1havecompleted Internship, Residency, and/or Fellowship training in: S P E C IA LT Y H O S P IT A L CITY state •SjSKSfJ1 Int. Rat. Rat. Fall. ACTIVITY STATUS In tha remainder ol 1986,1 expad to ba: 1 1Professionally inactive ot retired Q Active, no 'ima spent :n patient care. O Active, lata than 35 hours par weak in patient care. □A ctive, 35 or more hours par weak in patient cars. • ■d in u u n e m i y □ internship u n d e r g o in g (cnaotone n applicants) □ Residency S P E C IA LT Y a □ Fellowship training in: H O S P IT A L CITY in the future 1Dlan to undertake: (cheek ona * applicable) □ internship □ Residency u Fellowship training in: S P E C IA L T IE S PLERSE COMPLETE BOTH SIDES OF THIS CRRD In fo rm a tio n obtained from this card w ill be aggregated and used to describe th e distribution o f physicians and to an alyze h ealth resources. Id e n titie s o f indiuidual physicians w ill not be m ade public. SPECIALTY I EXPECT MY PRACTICE IN 1986 TO BE IN THEFOllOWING SPECIALTIES: PRIMARY SPE C IA LTY : PRACTICE LOCATIONS (To ba completed by all physicians who provida cars lo palianti.) I EXPECT TO PROVDE HOSPITAL CARE TO PATIENTS IN THE FOUOWWG LOCATIONS DURING 1986. (H y o udo not know rip coda, plaaaa writs nama and aty ol hospital.) H OSPITAL AVERAGE ZIP CODE H OUR8/W EEK SECONOARY SPEC IA LTY: (Optional) 305 I EXPECT TO PROVIDE N ON -HO S P. CARE TO PATIENTS IN THE FOLLOWING LOCATIONS DURING 1986. (oHIca, dinto. ate.) AVERAGE ZIP CODE HOURS/WEEK m CD m m m m • m CD Appendix D Appendix D Office of Health and Medical Affairs Technical Notes on Physician Census Results, 3/20/90. 306 307 Number of Physicians. When added across all counties or specialties, these figures represent unduplicated counts of physicians in Michigan, without regard to full time or part time status. In computing the "number of physicians," each patient care physician who practices in more than one location is fractionally allocated among those locations based on hours of hospital and non-hospital practice. (For example, a doctor who provided patient care services for 15 hours per week in a Lapeer county office, 15 hours per week in a Shiawassee County office, and 20 hours per week in a Genesee County hospital would contribute 0.3 to the number of physicians in Lapeer County and Shiawassee County, and 0.4 to the number of physicians in Genesee County.) Full-Time Equivalents. FTE’s are calculated according to the definition used by the U.S. Bureau of Health Professions in designating Health Manpower Shortage Areas. Fully-trained physicians providing patient care for 40 or more hours per week are counted as full FTE’s, and those providing patient care for less than 40 hours per week are counted as fractional FTE's. Physicians who practice in more than one location are fractionally allocated among those locations based on their non-hospital hours. However, physicians practicing exclusively in the hospital setting are allocated among their practice locations according to their hospital hours, and physicians undergoing graduate medical education are counted at one tenth the level of other physicians. (For example, a fully-trained physician who proviued patient care services for 15 hours per week in a Lapeer county office, 15 hours per week in a Shiawassee County office, and 20 hours per week in a Genesee County hospital would be counted as 0.5 FTE in Lapeer County and 0.5 FTE in Shiawassee County.) Patient Care Physicians. In order to reflect all patient care resources, the tabulations of patient care physicians include all doctors who report spending any time providing patient care services in Michigan. This differs from the practice of the American Medical Association and the American Osteopathic Association, which classify doctors only according to their major professional aciiviiy and their primary state of practice. Consequently, tables published by the AMA and the AOA generally report a slightly lower number of physicians in patient care and a slightly higher number in non-patient care activities. (For example, a doctor who spends 10 hours per week providing patient care services in Monroe County and 30 hours per week doing administrative work in Ohio would be classified by the AMA or AOA as a full-time non-patient care physician in Ohio. However, that doctor would be counted as one Michigan physician in tabulations of the "number of physicians," and as 0.25 in tabulations of patient care FTE's.) County. Physicians report their practice locations by zip code. These zip codes have been aggregated according to their associated county, as designated by the U.S. Postal Service. These aggregations of zip codes do not correspond exactly to the actual county boundaries or county populations. Therefore, when computing population-based ratios, it is necessary to use county population estimates which are also based on aggregations of zip codes. 308 Specialty Designations. Each self-designated principal specialty was recorded as indicated by the respondent, resulting in a list of 295 specialties, procedures, and professional activities. These were then re-coded to correspond to the specialty categories used by the American Medical Association. (See attached list.) The "No Response" category in the tables represents physicians who did not respond to the survey or who did not report their specialty. The "Unspecified" category represents responses which could not be associated with one of the standard specialties, e.g. "residency," "administration," "research," etc. Most of the physicians in the "Unspecified" category are interns or residents with limited educational licenses. Primary Care Physicians. For purposes of these tables, doctors are classified as primary care physicians if they reported principal specialties in family practice, general practice, pediatrics, internal medicine, obstetrics, gynecology, obstetrics/gynecology, adolescent medicine, geriatrics, primary care, public health patient care, urgent care, ambulatory care, student health, or hospice medicine. Doctors whose specialties are not known are assigned a fractional value based on the proportion of respondents who practice in primary care specialties. It should be noted that the classification of respondents as primary care physicians is independent of their assignment to the specialty categories recognized by the AMA, which serve as the basis for the tabulations by specialty. Some of the respondents assigned to the AMA categories of family practice, pediatrics, internal medicine, or obstetrics/gynecology are practitioners of subspecialties of those fields that are not considered part of primary care, such as pediatric gastroenterology, pediatric rheumatology, sports medicine, and menopausal management. Such doctors are not included in tabulations of primary care physicians. 309 Graduate Medical Education. Physicians in GME include those undergoing internship, residency, or fellowship training. GME Ratio. The GME Ratio is the number of physicians undergoing graduate medical education per 100,000 population. Total Ratio. The Total Ratio is the number of physicians per 100,000 population; it reflects both physicians undergoing GME training and physicians outside of training programs. Response Rates. The total number of active physicians in Michigan can be estimated from data produced by the American Medical Association, the American Osteopathic Association, and the Michigan Department of Licensing and Regulation. Using these figures as a base, the response rate of the Physician Census with regard to activity status is 95%. One or more practice locations are known for 93% of Michigan's active physicians. Patient care hours are known for 90% of the active physicians who are not reported to be in GME training, and for 79% of active physicians overall. Specialties are known for 93% of the active physicians who are not reported to be in GME training and for 84% of active physicians overall. » 310 Adjustment for Non-Responses. Physicians who did not report their current primary specialty of practice were classified according to the specialty of their most recent GME training, if available. Physicians who did not indicate their practice locations were counted as being located in the zip code of their mailing address on file with the Department of Licensing and Regulation. Hours in patient care were estimated where necessary based on the average hours of similar respondents. (Similar respondents were identified based on any available information on the physician's degree, training status, specialty, and full time or part time patient care status.) Non-respondents whose licenses lapsed in 1987 were assumed not to have been active in Michigan in 1986. Other physicians who did not report their professional activity were assigned a fractional value based on the proportion of similar respondents who provided patient care services; this fractional value was used in computing the Adjusted Count and the count of FTE's. Non-respondents were also assigned a fractional value representing the likelihood that they were undergoing GME training in 1986. (The total number of non-respondents in GME training was estimated by subtracting the total number of GME respondents from the number of filled GME positions in Michigan reported by the Michigan Council on Graduate Medical Education. These GME positions were then allocated among the non-respondents based on the proportion of reported GME trainees in each zip code.) Rounding Error. Fractional values in these tables are rounded to the nearest tenth. Since the statewide totals are based on the figures prior to rounding, they may not be equal to the sum of the rounded figures. Appendix E Appendix E American Medical Association Record of Physicians1 Professional Activities (PPA). 311 312 OFFICIAL RECORD IMMEDIATE RESPONSE REQUESTED RECORD OF PHYSICIANS' PROFESSIONAL ACTIVITIES Copyright 1986AMA DIRECTORYOFPHYSICIANS INSTRUCTIONS CONTENTS 1, Please answer every question on the basis of your current activities. 2. INDICATE ANY CHANCES. Information which you previously reported is printed as "Prior Census" data. Check to see if these data still accurately reflect your activities. Make changes or corrections where necessary. 3. PLEASE CH ECK YOUR MAILING ADDRESS printed In the top right corner of this form. Changes to your preferred professional mailing address (i.e., the address at which you wish to receive medieally*related materials including drugs, other samples, and printed material) should be made in the space provided. 4. Upon completion, return the questionnaire at your earliest convenience, using the preaddressed envelope. The questionnaire is divided into five sections: I. Specialization - Your primary, secondary, and tertiary specialties and how many hours you spend in each, during a typical week. II. Hospital Information • How many hours you spend in a hospital during a typical week, and the name and address of your primary hospital. III. Professional Activities • How many hours you spend in various professional activities during a typical week. IV. Present Employment • How many hours you spend working for various employers or under various types of employment arrangements during a typical week. V. Address and General Information - Mailing address changes, office address, office telephone number, and race/elhnicily. COPES FOR StLf»PESIGNATCO PRACTICE SPECIALTIES” ADI AM A Al AN CD CCM D DMP DlA DU IM END f p s ft ! c u t Geriatrics ! I s CE CP CRM Adolescent Medicine Aerospace Medicine Allergy Allergy and Immunology Anesthesiology Cardiovascular Diseases Critical Cart Medicine Dermatology Oermatopathology Diabetes Diagnostic Laboratory immunology Emergency M e d ian * Endocrinology facial Nettie Surgery, Otolaryngology family Practice Gastroenterology Ceneral Practice CO Gynecological Oncology GYN Gynecology HEM Hematology 1 C Immunology IP ID IM IM MFM MM NPM NEP N CHN NA NM NR NTR OIS OBG OM ON OPH OTO PTH ATP Immunopathology Infectious Diseases Internal Medicine Legal Medicine Maternal and fetal Medicine Medical Microbiology Neonatal—Pennatal Medicine Nephrology Neurology Neurology. Child Neuropathology Nuclear Medicine Nuclear Radiology Nutrition Obstetrics Obstetrics and Gynecology Occupational M edian* Oncology Ophthalmology Otolaryngology Pathology, Anatomic/Clinical Pathology. Anatomic • I I Pathology, flo o d Banking CMP CLP FOP RIP PO PDA PDC POE PHO PNP PA PM P CHP PYA PH PUD R DR PDR TR REN RHU Pathology. Chemical Pathology. Clinical Pathology, forensic Pathology, Radioisotopic Pediatries Pediatric Allergy Pediatric Cardiology Pedieirtc Endocrinology Pediatric Hematology—Oncology Pediatnc Nephrology Pharmacology, Clinical Physical Medicine and Rehabilitation Psychiatry Psychiatry. Child Psychoanalysis Public Health Pulmonary Dneases Radiology Radiology. Diagnostic Radiology. Pediatric Radiology, Therapeutic Reproductive Endocrinology Rheumatology ABS CDS CRS CS HS HNS NS ORS POS PS TS TRS U v s Surgery, Surgery. Surgery, Surgery, Surgery, Surgery, Surgery. Surgery, Surgery, Surgery, Surgery, Surgery, Surgery, Surgery, Abdominal Cardiovascular Colon and Rectal Ceneral Hand Head and Neck Neurological Orthopedic Pediatric Plastic Thoracic Traumatic Urological Vascular IN ADOITION TO THE ABOVE, THE FOLLOWING SPECIALITY DESIGNATIONS ARE ALSO USED: OS US Other, i.e.. a speciality other than those appearing above Unspecified speoahy 313 RECORD OF PHYSICIANS’ PROFESSIONAL ACTIVITIES III. PROFESSIONAL A C IIV IIIL S A V b How many houn per week do yuu ty|flwS!il#**itKaad> ci the following proMonai m M i M M*uu tp tn d in AhoaphaJ and/or offtco In a hospital............ 4) MEDICAL TEACHING?. Include hour* spent in medical reachirtf or iratmnf at wed at m preparation lor tubrects (auehr m medical tchooh. nurm g Khooii. other hospital schools. Octet, umvemtiei. or any other educartonal inueunons. •hU 2) What k tha nam t and addr— of tho hotpHal w hart you admh m oatof your p a titn o f (If you do not admh patitnts ghrt lha mom and addraat of your primary hoepfcaL) 5) MEDICAL RESEARCH). Indude acdehes (whether funded ot unfunded) performed lo medical knowledfe, potentially leadmf so publicarion •mmeimmod . fasti adbtm r a **"** «4-M 0 OTHER MEDICAL ACTIVITIES (not Ihud above) NOT Involving direct can of patients!......................................... 7)AU ntOTESSIONAL ACTIVITIES(lha total ot Qswsslons 14).................................... . If you indicated 20 hours or less on Question 7, or if none of the above categories applies to you, please answer Question S. I) Are you: re n t JtNet □Mittn ffM S c . □» 314 IV. I’ RLSENI I M l’ L O Y M I NT V. ADDRESS & GENERAL IN F O R M A T IO N Ulan bidkoi i thecucmntnumboro lh o u n pf w—fc io f M d ld f A lfo lA llflf M flo y ifw M V lO )1 M fllV llll|IIM n lL (N ott: Thh ihouid NOT ba corduaea wMl where your houn era v e n t) ttum ameer IVW ('quaadon 1 th n i §. Hyou 1) Profniional mailing addreu: - I aL. aLIa* ■- --A-_ ^mnv>Mviviicn|minHOT«NiPfi(Wiin| nfo(0^houn> '•V•Hr'! •HOUMWKWCK 3 3 1) S*H*mploy«d 10I0pnak» I) Two phyildan ptacdce ■□t fu ll o r part owner Q 2 Employee. Please enter any addreti corrections below: 3) Croup practice. kM M A m *APMw ea «eeW en an tm lrmmm*.m t A* > AMA.A^AaiaimfeAO>rtAArAMMOA ■mNwMtar. km.Dwnvbm Hmt,*k. A imm A w iiA ym a W w i* W earea*. cht um lip CROUP NAME AND ADDRESS; 2) b your prolawlonal trialBog addraaa (Ham t l above) also your office addreasl r a Q i YES □ a NO — *• (H NO, pleaae provide your office addreai). PLEASE ENTER CORRECTIONS BELOW: tW.WW 4) MetScal School .(or patent unhenlty). •?- - V C : ---- 3) Office telephone number: I 5) Non-Go**mment Hospital. )_______ CAaoi/ilvnt ffi Cfty/County/StMe Government e)Homltal........................ . A tCAAAKCasrwP t J A A C V C A COAMV Indicate your reca/ethnkhy: b) Other than hospital. in □ i Whke, non-Mbpanlc □ r Black, non-Hhpanlc OiHhpanlc 7) UA l •) Hospital. □ a Aden/Oriental p i Other bt Othor than hoaphal. Indicate Federal Apency: 'flew •■•;••• .. S) y.v*v •%.; OtherpnBoot a re employment ^ . • «—. * - B) Other non padaeai taaaemployment HOT ■ Ibaad above. • B'ValVetjia^p^Vfd'iIpti i»* Appendix F Appendix F American Osteopathic Association 1989 Census of the Osteopathic Medical Profession. 315 NUMBER OF PATIENT* PER WEEK 10C. Patients per week TREATED in EMERGENCY i— ii— 1 i— i ROOM OR OUTPATIENT CLINICS?................................................................- I __ 11___11_! 100 Patients per week TREATED in THE PATIENT'S HOME (House Calls)? i— 1 1— 11— j - I__ 11___11_I 10E. Patients per week TREATED In NURSING/ I— »j— | 1— i CONVALESCENT/EXTENDED CARE?.......................................................... - I __ I I ___II _I 10F. Patients per week TREATED in OTHER TREATMENT SETTINGS NOT LISTED ABOVE? - Aincrictm(isteitpnlhicAsaaciutum 143 ONTARIO STREET • CHICAGO. ILLINOIS 60S11 l- W M S t im « (3131SBO-MOO 198 9 CENSUS OF THE OSTEOPATHIC MEDICAL D D O B B It lB in U Section I. This section is your opportunity to review the information the AOA maintains in its physician master file. II this information is not correct please make the necessary changes and return H to us so your record may be corrected; if information is missing please write in your response. Please complete all four pages of the questionnaire. Is this your 1. Home Address □ 2. Office Address □ 3 Hospital Address a 4. Other Address □ [----if -----1 (---I 1__ 11__ 11_I ItMMSearW--------------------- I 11. Do you have HOSPITAL ADMITTING PRIVILEGES? ............................ - YesD No □ 12. If question eleven is answered "YES*, please indicate the names of the major lospitats, the cities in which they are located and the number of patients you admit to each in a TYPICAL WEEK Date of Birth. Gender 12A. Hospital Nam e__________________________ C ity_______________________________ Major Practice Focus*: Minor Practice Focus*: 12B Hospital N am e__________________________ e s s k s *•□□□□□□□ ** 12C Hospital Nam e__________________________ S " City_____________________________ ..□□□ 10 * For Definition of practice focus and a list of the practice foci see the glossary on the reverse of Ihe cover letter. Medical Education: City_____________________________ s-UULJUiJUU Board Certification: 13 Please list the states in which you are currently licensed to practice medicine, and provide the license number and year ol issue lor each. i3a. 138. ..cu cu .cu cu □ □ cu cu cu cu cu .cu □ □ □ □ □ □ □ .D D i 3 D . .D D ,« « « .cu cu .cu cu Graduate Medical Education (osteopathic and allopathic)**: J II I 13C. 3 » ...CUCUCUCU CUCUCUCU ... , 14. Please indicate your race/ethnicity by checking the appropriate box. 14A. White. Non-Hispanlc 14D. Asian/Pacific Islander - □ 14B Black, Non-Hispanic . □ TIC. Hispanic » □ 14E. American Indian/ Alaskan Native 15. Please enter your social security number. . □ □ □ / □ □ / □ □ □ a THANK YOU FOR YOUR ASSISTANCE. Please relum in Ihe enclosed POSTAGE PAID ENVELOPE. ’’ Institution names are CURRENT names —CONTINUED- Section II. 9. This section asks foe descriptive information about your practice. Most questions ask for an estimate of how many patients you treat or hours you practice in a typical week. Some quest ions may be answered •YES’ or-NO". 1. PROFESSIONAL ACTIVITIES: How many hours per week do you TYPICALLY SPEND In EACH of the following professional activities? If you do not spend any time in a particular activity, please enter a zero in this appropriate space. PRESENT EMPLOYMENT: How many hours per week are you TYPICALLY EMPLOYED in EACH of the following employment arrangements? If you ARE NOT involved In a particular employment arrangement, please enter a zero in the appropriate space. E iam pht e l A n e w ** Hour* - 3?houn tyrant-»% Number • 12 patients EStS[®J(H [Jk (DEE 9A. Hours per week in SELF-EMPLOYED SOLO PRACTICE?......... ................... □ □ □ « 98. Hours per week in TWO-PHYSICIAN PRACTICE (Partnership)? ... .............. □ □ □ „ 9C Hours per week in GROUP PRACTICE?..................................................... ... C D C D C D m (i.e. three or more physicians organized to provide medical care through the joint use of equipment and with income from the practice distributed according to predetermined methods.) 1A. Hours per week in POSTDOCTORAL training program as either an INTERN. RESIDENT or FELLOW?..................................................... .............................._ 9D. Hours per week in OTHER GROUP PRACTICE? (e.g. HMOs. freestanding ambulatory care centers, etc.) IB. Hours per week in DIRECT PATIENT CARE?......................................... Exclude time spent in training, teaching or research. Include time spent on routine office work and travel time pertaining to your patients. 9E. Hours per week in HOSPITAL. NON-GOVERNMENTAL?........ .................. . □ 9F. Hours per week In CITY/COUNTY/STATE GOVERNMENT? .. . . ................. □ This care (IB) b predominantly given in what type of setting? 181. Short Term »G 1B2. Intermediate —□ 1B3. Long Term • □ BQ. Hours per week in FEDERAL GOVERNMENT? .□□CL 1D. Hours per week in MEDICAL EDUCATION?.......................................... 105. Other * a 1D2. Director of Medical Education 1D4. Adjunct Facufty ............................... . □ II the response (9G) is greater than zero, check setting: 9G1. Hospital 9G2. Non-Hospital If the response (9G) is greater than zero, check agency: 9G3. Army *%□ 9G4. Navy 9G5 Air Force ~ □ 9G6 Veterans Administration 9G7. USPHS ~ □ 9G8 Other .□ .□ i w > ___________________________________ 91. Hours per week in OTHER PATIENT CARE EMPLOYMENT?...................... nwo*r Simi fritn - Yes □ N oD 3. Are you affiliated with a PPO (preferred provider organization)?........... - Yes D No □ 4. Are you affiliated with an HMO (health maintenance organization)?...... - Yes □ NoQ 8 It question seven was answered "NO" then ESTIMATE the percent of MEDICARE assignments you TYPICALLY accept each week................. —CONTINUED— . □ . □ □ « i— i r— 11 — i -□□□% -----—-- EZZiC U CDm i 9J. Hours per week in OTHER NON-PATIENT CARE EMPLOYMENT?.............(e g. pharmaceutical companies, insurance carriers, osteopathic org.) 10. Have you signed a participation agreement tor fees with MEDICARE? ... A participation agreement states ASSIGNMENTS will be accepted from AU. MEDICARE PATIENTS. □ □ ___________________ J_II II_L 2 Are you affiliated with an IPA (individual practice association)?............. 7. □ «. O 1F. Hours per week In OTHER MEDICAL ACTIVITIES NOT INVOLVING PATIENT CARE?................................................................................. In a TYPICAL WEEK, estimate the PERCENTAGE ot your patients who are enrolled In MEDICARE...................................................................... . - Q 9H. Hours per week In MEDICAL SCHOOL or other ACADEMIC INSTITUTION? 6. □ « ,□ 1E Hours per week In MEDICAL RESEARCH?. 5. In a TYPICAL WEEK, estimate the PERCENTAGE of your patients who are enrolled in MEDICAID........................................................................ □ II the response (9F) Is greater than zero, check setting: 9F1. Hospital -□ 9F2. Non-Hospital -□ 1C Hours per week In ADMINISTRATIVE ACTIVITIES?.............................. Are these hours (ID ) as w-o 101. Full Time Faculty - □ 1D3. Part Time Faculty □ PATIENT CARE: How many patients per week do you TYPICALLY TREAT (or provide services if a Pathologist or Radiologist) in EACH of the following treatment settings? Count one visit each time you see a patient (or their specimen or film), even if you see the same patient more than once per week. If you ARE NOT treating in a particular setting, please enter zero In the appropriate space. .-□□□% . -Y«aO NUMBER OF PATIENTS PER WEEK NoD 10A. Patients per week TREATED In OFFICE?.............. ..................................... . □ (include outpatient surgery 11done in office.) .-□□□% 106. Patients per week TREATED in SHORT TERM ACUTE CARE HOSPITALS? i— □ □ 11— 11— i ..................J II II I -CONTINUED— Appendix G Appendix G American Medical Association List of Self-Designated Practice Specialty Codes (85 Categories). SELF-DESIGNATED PRACTICE SPECIALTY CODES ON THE AMA PHYSICIAN MASTERFILE Self-designated practice specialties (SDPS) listed on the AMA Physician Masterfile have historically related to the record-keeping needs of the Ameri­ can Medical Association and do not imply 'recogni­ tion* or 'endorsement* of any field of medical practice by the Association. The AMA Division of Survey and Data Resources manages the Masterfile and records a physician’s SDPS based on physician response to the Physicians' Professional Activities Questionnaire (PPA Census). The full list of Master­ file SDPS codes is pre-printed on the questionnaire and physicians indicate the average number of hours spent during a typical week in those special­ ties to which they limit their practice. The SDPS with the greatest number of hours becomes the pri­ mary SDPS, the SDPS with the next greatest num­ ber of hours becomes the secondary SDPS, and so on. The fact that a physician chooses to designate a given s p e c ia lty o n o u r re c o rd s d o e s n o t n e c e s s a rily mean that the physician has been trained or has special competence to practice the SDPS. Appendix B-2 lists codes for the SDPS included on the AMA Physician Masterfile. On the other hand, American Specialty Board certification indicates that a physician has received preparation in accordance with established educa­ tional standards of one of the 23 member boards of the American Board of Medical Specialties (ABMS). It is not uncommon for a physician to be board-certified in one field of medicine while lim ­ iting practice to another. As of December 31,1986, 47,602 (8.4%) of all board-certified physicians in the United States were certified by a board other than that corresponding to their Primary SDPS, and 254,637 (44.7%) were not board-certified. In addi­ 318 tion to maintaining physicians’ SDPS on the AMA Physician Masterfile, the AMA collects primary specialty board certification data through the American Board of Medical Specialties. Because both board certification and SDPS are important, these data are included in the individual physi­ cian!) record as well as in the American Medical Directory. Neither SDPS nor certification by a member board of ABMS should be confused with the fact that a physician has successfully completed a pro­ gram (or programs) of accredited graduate medical education. Accreditation is the process whereby the Accreditation Council for Graduate Medical Educa­ tion (ACGME) grants public recognition to a special­ ized program which meets certain established educational standards as determined through ini­ tial and subsequent periodic evaluations by one of ihe 24 Residency Review Committees. AMA collects physicians’ residency training data through an an­ nual Census of all ACGME-accredited residency training programs. These data, along with SDPS and certifications by a member board of ABMS ap­ pear under separate headings on the AMA Physi­ cian Profile. AMA is taking steps to insure that the list of Masterfile SDPS codes is not misused to represent certification by a member board of ABMS or ACGME-accredited residency training in a given specialty. We ask that you help in this effort. Source: American Medical Directory, 30th edition. Division of Survey and Data Resources, American Medical Association, 1986. 319 LIST OF 85 SELF-DESIGNATED PRACTICE SPECIALTY CODES* AM A AN CD CHP CRS D DR EM FOP GE A erospace M edicine Allergy A nesthesiology C ardiovascular Diseases C hild P sychiatry Colon and Rectal Surgery Dermatology Diagnostic Radiology Em ergency M edicine Forensic Pathology G astroenterology GP G eneral P ractice— Includes: FP Fam ily P ractice GP G eneral P ractice GPM GS OM OPH ORS O ccupational M edicine O phthalm ology O rthopedic S urgery OTO Otolaryngology— Includes: FPS Facial P lastic Surgery, Otolaryngology OTO Otolaryngology PTH A n ato m ic/C lin ical Pathology—Includes: ATP A natom ic Pathology PTH A n atom ic/C linical Pathology BLB B loodbanking CLP C linical Pathology CMP C hem ical Pathology DMP Dermatopathology IP Im m unopathology MM M edical Microbiology NA N europathology RIP R adioisotophic Pathology PD Pediatrics— Includes: ADL A dolescent M edicine NPM N eonatal-Perinatal M edicine PD Pediatrics PDE Pediatrics, Endocrinology PHO Pediatric Hematology-Oncology PNP Pediatrics, Nephrology PDA PDC PM PS Pediatric Allergy Pediatric Cardiology Physical M edicine & Rehabilitation Plastic Surgery P P sychiatry— Includes: P Psychiatry PYA Psychoanalysis PH PUD P ublic H ealth P ulm onary Diseases R Radiology— Includes: NR N uclear Radiology PDR Pediatric Radiology R Radiology G eneral P reventive M edicine General AS CDS GS HS HNS PDS TRS VS Surgery— Includes: A bdom inal Surgery C ardiovascular S urgery G eneral Surgery H and Surgery Head and N eck S urgery Pediatric Surgery T raum atic Surgery V ascular Surgery Internal AI DIA DLI M edicine— Includes: A llergy an d Im m unology D iabetes Diagnostic Laboratory Imm unology Endocrinology G eriatrics Hematology Im m unology infectious DisCuCCs Internal M edicine Nephrology N utrition M edical Oncology R heum atology END GER HEM IG ID IM NEP NTR ON RHU NS N eurological Surgery N Neurology— Includes: CHN C hild N eurology N N eurology RO TS U Radiation Oncology Thoracic Surgery Urological Surgery NM N uclear M edicine OS OBG O bstetrics & Gynecology— Includes: GYN Gynecology GO Gynecological Oncology MFM M aternal an d Fetal M edicine OBS O bstetrics OBG O bstetrics & Gynecology REN R eproductive Endocrinology O th er Specialty— Includes: PA Clinical Pharmacology CCM Critical Care M edicine LM Legal M edicine OS O th er Specialty US U nspecified NOTE: Therapeutic Radiology has been changed to Radiation Oncology. 'Above listinggives the conversion of the 85 Self-Designated Practice SpecialtyCodes (as listedon the PPA questionnaire and on the A M A Physician Masterfile) into the 39 Specialty Codes used for statistical purposes by the Association. Appendix H Appendix H American Medical Association List of Practice Specialty Codes (39 Categories). AM A AN CD CHP Aerospace Medicine Allergy Anesthesiology Cardiovascular Disease Child Psychiatry CRS D DR EM FOP Colon and Rectal Surgery Dermatology Diagnostic Radiology Emergency Medicine Forensic Pathology GE GP GPM GS IM Gastroenterology General Practice General Preventative Medicine General Surgery Internal Medicine NS N NM OBG OM Neurological Surgery Neurology Nuclear Medicine Obstetrics and Gynecology Occupational Medicine OPH ORS OTO PTH PD Ophthalmology Orthopedic Surgery Otolaryngology PDA PDC PM PS P Pediatric Allergy Pediatric Cardiology Physical Medicine and Rehabilitation Plastic Surgery Psychiatry PH PUD R TR TS Public Health Pulmonology Radiology Therapeutic Radiology Thoracic Surgery U OS US Urological Surgery Other Specialty Unspecified 7\ M AW T"> -> 4 -V ^ a 1 ac* t Pediatrics 320 Appendix I Appendix I American Medical Association List of Specialty Board Abbreviations (4 Categories). GENERAL PRACTICE FP Family Practice GP General Practice MEDICAL SPECIALTIES A Allergy Cardiovascular Diseases CD Dermatology D Gastroenterology GE Internal Medicine IM PD Pediatrics PDA Pediatric Allergy PDC Pediatric Cardiology PUD Pulmonary Diseases SURGICAL SPECIALTIES CRS Colon and Rectal Surgery General Surgery GS NS ’ Neurological Surgery OBG Obstetrics and Gynecology OPH Ophthalmology ORS Orthopedic Surgery OTO Otolaryngology PS Plastic Surgery TS Thoracic Surgery U Urological Surgery ' OTHER SPECIALTIES AM Aerospace Medicine Anesthesiology AN CHP Child Psychiatry DR Diagnostic Radiology EM Emergency Medicine FOP Forensic Pathology GPM General Preventive Medicine N Neurology NM Nuclear Medicine Occupational Medicine OM P Psychiatry Public Health PH PM Physical Medicine and Rehabilitation PTH Pathology, Anatomic/Clinical R Radiology Radiation Oncology RO Other Specialty OS Unspecified US 321 Appendix J Appendix J Final Data Base. 322 323 Table J-l Codebook for Individual Physician File LABEL COLUMNS DESCRIPTION FORMAT ID 1-5 RANDOM UNIQUE IDENTIFICATION F5.0 LISTATUS 7 LICENSE STATUS 4 categories F1.0 LITYPE 9 LICENSE TYPE 3 categories A1 GENDER 11 0=MEN, l=WOMEN F1.0 BIRTHYR 13-16 YYYY-YEAR OF BIRTH F4.0 STATE 18-20 COUNTRY/STATE OF MEDICAL SCHOOL GRADUATION AMA CODES F3.0 GRADYR 22-25 YY-YEAR OF GRADUATION FROM MEDICAL SCHOOL F4.0 SPEC 27-29 SPECIALTY from AMA=85, AOA=99 categories A3 B87 31-32 BOARD 43=MD, 51=DO F2.0 AA 34-35 ASCRIBED ACTIVITY STATUS 8 categories F2.0 ST 37-38 REPORTED ACTIVITY STATUS 17 categories F2.0 CT 40 CURRENT TRAINING STATUS 6 categories F1.0 TCT 42 Temporary file Training Status 3 categories Fi.0 APS 44-46 ASCRIBED PRINCIPAL SPECIALTY F3.0 PS 48-50 REPORTED PRINCIPAL SPECIALTY F3 .0 SS 52-54 REPORTED SECONDARY SPECIALTY F3.0 TH 56-61 # TOTAL HOURS F6.0 THH 63-68 # TOTAL HOSPITAL HOURS F6.0 TOH 70-75 # TOTAL OFFICE HOURS F6.0 324 HH1 77-82 HOSPITAL HOURS 1st location F6.0 HH2 84-89 HOSPITAL HOURS 2nd location F6.0 HH3 91-96 HOSPITAL HOURS 3rd location F6.0 HH4 98-103 HOSPITAL HOURS 4th location F6.0 OH2 105-110 OFFICE HOURS 2ND location F6.0 OH3 112-117 OFFICE HOURS 3RD location F6.0 OH4 119-124 OFFICE HOURS 4TH location F6.0 APS 38 126-128 AMA 38 specialties 41 categories F3.0 APS 82 130-132 AMA 82 specialties 86 categories F3.0 APS 5 134-136 AMA 5 specialties 7 categories F3.0 SS38 138-140 AMA 38 specialties 41 categories F3.0 SS82 142-144 AMA 82 specialties 86 categories F3.0 SS5 146-148 AMA 5 specialties 7 categories F3.0 LENGTH OF RECORD = 148 / BLOCK = 14800 325 Table J-2 Codebook for Physician Records By Location LABEL COLUMNS DESCRIPTION FORMAT ID 1-5 RANDOM UNIQUE IDENTIFICATION F5.0 LISTATUS 7 LICENSE STATUS 4 categories F1.0 LITYPE 9 LICENSE TYPE 3 categories A1 GENDER 11 0=MEN, l=WOMEN F1.0 BIRTHYR 13-16 YYYY-YEAR OF BIRTH F4.0 STATE 18-20 COUNTRY/STATE OF MEDICAL SCHOOL GRADUATION AMA CODES F3.0 GRADYR 22-25 YY-YEAR OF GRADUATION FROM MEDICAL SCHOOL F4.0 SPEC 27-29 SPECIALTY AMA= 85 categories, AOA=99 A3 B87 31-32 BOARD (MD or DO) A2 AA 34-35 ASCRIBED ACTIVITY STATUS 8 categories F2.0 ST 37-38 REPORTED ACTIVITY STATUS 17 categories F2.0 CT 40 CURRENT TRAINING STATUS 6 categories F1.0 42-44 ASCRIBED PRINCIPAL SPECIALTY F3.0 ss 46-48 REPORTED SECONDARY SPECIALTY F3.0 TH 50-55 # TOTAL HOURS F6.0 THH 57-62 # TOTAL HOSPITAL HOURS F6.0 TOH 64-69 # TOTAL OFFICE HOURS F6.0 H 71-76 HOURS IN EACH LOCATION: 4 each site, 6 categories F6.0 ZT 78 ZIP TYPE FOR EACH LOCATION 7 categories F1.0 HT 80 HOUR TYPE FOR EACH LOCATION 7 categories FI. 0 PT 82 PRACTICE TYPE 3 categories FI. 0 326 PFTE 84-89 # PHYSICIAN FTE for all locations F6.0 FTE 91-96 # FULL TIME EQUIVALENT at individual location F6.0 COUNTY 98-100 83 counties, FIPS code 1-165 F3.0 APS38 102-104 AMA 38 specialties 41 categories F3.0 APS 82 106-108 AMA 82 specialties 86 categories F3.0 APS5 110-112 AMA 5 specialties 7 categories F3.0 SS38 114-116 AMA 38 specialties 41 categories F3.0 SS82 118-120 AMA 82 specialties 86 categories F3.0 SS5 122-124 AMA 5 specialties 7 categories F3.0 LENGTH OF RECORD = 125 / BLOCK = 12500 Appendix K Appendix K Physician Profiles The Woman Physician A profile of women physicians in Michigan in 1986 (Figure 13) was that 84% were MDs (35% of them trained overseas) and 16% were DOs (Table A-ll). Seventy percent had graduated since 1970, fully 40% since 1980 which was only 6 years prior to when this data was collected. Their mean year of birth was (rounded to) 1947 and mean year of graduation was 1974, each significantly later than for men (Table A-21). Women physicians were primarily (94%) in metropolitan areas (Table A-17, A-18 and A-19, Figure 17). Fifty eight percent were in nonprimary specialties (Table 8) with (Appendix I) medical and other specialties predominating. Ninety four percent of active women were in patient care, and 29% were still in graduate medical education (Table 6). The mean patient care hours worked (Table 7), with an upper limit of 80 hours, totalled 49.6 hours per week, 18.2 in nonhospital settings and 31.4 in hospital settings, all significantly lower than for men. The inactive rate (10.1%) was not statistically significantly different from men's rate (Table A-25). 327 328 Earlv Women Graduates The active woman physicians who graduated before 1970 were 95% MD of which 38% were USA trained and 62% international graduates, and 5% were DO. graduation was 1961. Their mean year of Ninety five percent practiced in metropolitan and 5% in rural counties (Table A-28). Fifty four percent were primary specialists with (Appendix I) medical and other specialties dominating. Their total patient hours were 44.8 per week, balanced between office (22.3) and hospital (22.5) hours. Recent Women Graduates The active women physicians who had graduated since 1970 were markedly different. MDs were 79%, 77% of which were USA trained and 23% were internationally trained, and 21% were now DOs. The mean year of birth was 1952 and year of graduation was 1980, with 41% still in graduate medical education. Ninety four percent practiced in metropolitan and 6% in rural counties, perhaps skewed to the metropolitan by the large percentage still in training. Fifty nine percent were in nonprimary specialties (Tables A-2 3 and A24) with (Appendix I) medical specialties predominating. Their total patient care hours averaged 54.4 per week (including those in training) with over twice as many in hospital (37.5) as in office (16.9) settings. 329 Men Physicians MDs were 79% of all men, 25% internationally trained and 75% US trained, and 21% were DOs (Table A-ll). percent of men were active (Table A-28). Ninety Men's mean year of birth was 1940 and year of graduation was 1967, older and earlier than women (Table A-21). Fifty five percent had graduated before 1970. Ninety percent were active in metropolitan areas (Table A-17, A-18 and A-19, and Figure 17). Sixty one percent were in nonprimary specialties (Table 8) with (Appendix I) medical specialties leading, and 39% were in primary care. Ninety five percent were in patient care, but 15% were still in graduate medical education (Table 6). The mean patient care hours worked (Table 7, with an upper limit of 80) totalled 50.8 hours, 21.8 in nonhospital settings, and 29.0 in hospital settings. Early Men Graduates Men who graduated prior to 1971 were 82% MD of which 70% were USA trained and 30% were internationally trained, and 18% were DOs. Their mean year of birth was 1931 and their year of graduation was 1958, earlier than for early women. Ninety percent practiced in metropolitan and 10% in rural counties. Fifty eight percent were nonprimary specialists with (Appendix I) surgical specialties dominating. Their mean total patient care hours were 49.0 per week, balanced between office (24.9) and hospital (24.1) settings. 330 Recent Men Graduates Men who graduated after 1970 illustrated the changes in enrollment, as 76% were MDs and 81% of those USA trained and 19% international graduates, but 24% were DOs. There were more women in MD schools, and while the total enrollment in DO schools grew rapidly after 1971, there were fewer women admitted to DO than MD schools. Their mean year of birth was 1951 and year of graduation was 1979, one year earlier on each account than for recent women graduates. Thirty four percent were still in graduate medical education, a proportion lower than for women. Ninety percent were in metropolitan and 10% in rural counties, reflecting the one-third in training in metropolitan areas. Sixty four percent were in nonprimary specialties (Tables A-23 and A-24) with (Appendix I) medical specialties leading. Their mean total patient care hours were 58.3 per week (including those still in training), twice as many in hospitals (39.3) as in offices (19.0), and 3.9 hours more than all active women. Rural Physicians Rural Michigan was a principal practice site for 9.6% of all active physicians (Tables A-17, A-18, A-19 and A-27). Women were 8%. Seventy seven percent were MDs, 82% of them USA trained and 18% internationally trained, and 23% were DOs. Only 1% of those in graduate medical education were in rural areas. (Table A-28). Fifty three percent had graduated before 1970 The mean year of birth of an active rural 331 physician was 1940 and year of graduation was 1967, older and earlier than the total active population. Fifty five percent were in primary care (Tables A-27, A-31, A-32 and A-33). The mean patient care hours worked (with an upper limit of 80 hours) totalled 50.7 hours per week, 26.2 in nonhospital settings, and 24.5 in hospital settings. An original finding was the fall, then the continuing increase in the proportion of both women and men (Table A30, Figure 17) practising in rural areas, from each decade of graduation. Up to 18% of men 16% of women fully trained and graduating since 1979 were in rural locations. USA Trained MDs USA MDs were 59% of all active physicians, and 12% of them were women. Their mean year of birth was 1941, and year of graduation was 1968 (Table A-21), 51% of them graduating since 1970. Ninety five percent of active physicians were in patient care, and 64% were in specialties, half in either (Appendix I) medical or surgical specialties (Table A-37). The mean total patient care hours were 50.3 per week by USA MDs, more in hospitals, 29.0 hours, than in nonhospital settings, 21.3 hours. The practice differences noted between women and men USA trained MDs were that women were more likely to be in primary care in metropolitan areas and to be younger and more recent graduates than men (Table A-17). 332 Internationally Trained MDs International MDs were 31% of total active and 18% were women (Table A-ll). physicians, The mean year of birth was 1940 and year of graduation was 1966 (Table A-21). Two thirds had graduated before 1971, reflecting the history of international medical practitioners as immigrants during the USA physician shortage in the 1960s and 1970s. They have had a lessening impact on physician services in the USA, as fewer international MDs were arriving in the USA, or remaining after completion of training (Fulop, 1986a). Ninety four percent were in patient care, 63% in specialties (Table A-37) of whom (Appendix I) 60% were in either medical or other specialties. The mean total patient care hours worked were 50.9 hours per week, balanced between hospital (18.7) and nonhospital (19.8) settings. DOs DOs were 20% of the active physicians, and 25% of all USA trained physicians (Table A-ll). total of DOs. Women were 10% of the Ninety five percent were in patient care, with 51% in primary care, 40% of which (Appendix I) were in family and general practice. The mean year of birth was 1942 and the year of graduation was 1970 (Table A-21), so they were older, but more recent graduates than MDs. The mean hours worked per week were 51.5 (Figure 18), more in the hospital setting (28.1) than in the office (23.4). They were proportionately more in office only practices than MDs. 333 Between the professions (MD and DO) practising in Michigan in 1986, a DO was most likely to be in primary care and significantly more in rural settings (Tables A-17, A-18, A-19 and A-31). Primary care was more commonly practiced in rural areas for all physicians (Table A-31), so the practice patterns DOs have had a steady impact in the USA, as well as in a DO rich state like Michigan. Nationwide in 1984, fifty percent of osteopathic physicians were practising in a rural county (Denslow, et al, 1984). Specialty training patterns for DOs were changing over the years of graduation. Women and men DOs were significantly (p=.00) diverging from the historical primary specialty choices of DOs in patient care (Table A-16). 334 10% W om en M en Gender 90% 51% P r im a r y Specialty 4 9 % N o n p r im a r y M e tr o p o lita n 89% L o c a t io n R u ra l 43% G rad ua tion < 1971 57% ) 1970 17% H o s p ita l O n ly P r a c t ic e 13% O llic e O n ly 7 0 % M ix e d Figure 21 . Profile of Active DOs Appendix L Appendix L Characteristics of Specialties With reference to the AMA's broad categories of specialty (Appendix I), each of the four specialty groups were examined (Tables A-23, A-24, A-32 through A-38). Family and General Practice This was the most distinct specialty group. DOs were 45% of this specialty and 55% were MD, with 90% USA trained, the highest for each of DO and USA trained MDs, two underlying characteristics of this specialty. were women. Ten percent The mean year of birth was 1937 and the mean year of graduation was 1965 and 57% had graduated before 1971, making them older and earlier graduates than other specialty groups. Twenty one percent were in rural counties Tables A-31 and A-32), another record for a specialty group, one woman to every 11 men. The total patient care hours worked were 47.3 per week, with office hours almost three times (34.4) that in hospitals (12.9), another identifying characteristic of family and general practitioners (Table A33). They were the least represented specialty in hospital only practice, and the highest in office only practice. Medical Specialists These were a varied group of practitioners, 15% of them women (Table A-32), 90% of them MD of which 23% were internationally trained and 77% USA trained. 335 The mean year 336 of birth was 1941 and year of graduation was 1968, and 49% had graduated since 1970, and 11% were still in graduate medical education. Ninety three percent practiced in metropolitan counties and 7% in rural, one women to every 9 men. Within the medical specialties, 65% were primary including pediatrics and general internal medicine, while 35% were nonprimary. They worked a total of 49.7 hours per week, with a balance between office (24.2) and hospital (25.6) hours (Table A-33). They were the highest represented specialty in hospital and office practice, and the second lowest in office only practice. Surgical Specialists Surgery was dominated by male MDs (Table A-37). Men were 93% of the total and women were 7%, making surgery the lowest representation of women in a specialty group. MDs were 86% (18% internationally and 82% USA trained) and DOs were 14% of the total. The mean year of birth was 1939 and the mean year of graduation was 1965, with 64% graduated before 1971, making them the second oldest specialty group, while 9% were still in graduate medical education. Ninety percent practiced in metropolitan and 10% in rural counties (Table A-31 and A-32). Using the primary and nonprimary division, 27% were primary practitioners, as this category included the obstetric and gynecology specialists. The total hours worked were 52.9 per week, with 22.3 hours in office sites and 30.6 in hospital settings. 337 Surgical specialists were the lowest represented specialty in office only practice, and the second lowest in hospital only practice. Surgery was a practice in both office and hospital. Other Specialists This specialty group were broadly distributed in part due to being a broad category. Women were 13% and men 87%, 84% were MDs (27% internationally trained and 73% USA trained) and 16% were DOs. The mean year of birth was 1940 and year of graduation was 1967, with 10% still in graduate medical education. Ninety two percent practiced in metropolitan and 8% in rural counties, with one woman to every 9 men. Their mean total hours were 46.8 per week, with almost three times as many (34.6) in hospital than office (12.2) settings. They were the highest represented specialty in hospital only practice. BIBLIOGRAPHY BIBLIOGRAPHY Abir-Am, P. G . , & Outram, D. (Eds.)* (1987). Uneasy careers and intimate lives. Women and science. 1789-1979. New Brunswick: Rutgers University Press. Abram, R. J. (Ed.). (1985). Send us a ladv physician: Women doctors in America. 1835-1920. New York: W. W. Norton. Accreditation Council for Graduate Medical Education. (1990). Manual of Structure and Functions for Graduate Medical Education Review Committees. Chicago: Author. Accreditation Council for Graduate Medical Education. (1991). ACGME and Residency Review Committees change standards for resident work hours. Chicago, Illinois: Author. Ackerman-Ross, F. S., & Sochat, N. (1980). Close encounters of the medical kind: Attitudes toward male and female physicians. Social Science and Medicine. 14A. 61-64. Adams, E. K . , & Bazzoli, G. J. (1986). Career plans of women and minority physicians: Implications for health manpower policy. Journal of the American Medical Women's Association. 41(1), 17-20. Adams, J, W, (1977), Patient discrimination against women physicians. Journal of the American Medical Women's Association. 32.(7) , 255-261. Aguirre, B. E., Wolinsky, F. D . , Niederauer, J . , Keith, V . , & Fann, L.- J. (1989). Occupational prestige in the health care delivery system. Journal of Health and Social Behavior. 30(8), 315-329. Allen, D. I. (1989). Women in medical specialty societies. Journal of the American Medical Association. 262(24), 34393443 . Allen, D. I. (1990). Women and specialties. [Letter to the Editor]. Journal of the American Medical Association. 263(17), 2298. 338 339 Allison, P. D. (1984). Event for longitudinal event data. Quantitative Applications in Beverly Hills & London: Sage history analysis. Regression Sage University Paper Series on the Social Sciences, 07-046. Publications. American College of Surgeons. (1988). Socio-economic factbook for surgery 1988. Chicago, IL: Author. American Council on Education. (1987). The new agenda of women for higher education: A report of the ACE Commission on Women in Higher Education. Washington DC: Author. American Medical Association. (1967). The directory of approved internships and residencies 1967-68. Chicago, IL: Author. American Medical Association. (1987). Physician characteristics and distribution in the U.S. Chicago, IL: Author. American Medical Association. (1988). Physician characteristics and distribution in the U.S. Chicago, IL: Author. American Medical Association. (1989). Physician characteristics and distribution in the U.S. Chicago, IL: Author. American Medical Association. (1988). The new woman physician: Her growing numbers are not yet represented in organized medicine. Hawaii Medical Journal 47(9), 444-445. American Medical Association Center for Health Policy Research. (1988). Physician supply and utilization bv specialty: Trends and projections. Chicago, IL: Author. American Medical Women's Association. (1968). The fuller utilization of the woman physician. Report of a conference on meeting medical manpower needs. Washington DC: Women's Bureau, U.S. Department of Labor. American Osteopathic Association. (1987). Year book and directory of osteopathic surgeons. 1987-1988. Chicago, IL: Author. American Osteopathic Association. (1988). Year book and directory of osteopathic surgeons. 1988-1989. Chicago, IL: Author. American Osteopathic Association. Papers. Chicago, IL: Author. (1989). AOA Position 340 Andrews, F. M . , Klein, L., Davidson, T. N., O'Malley, P. M . , & Rodgers, W. L. (1981). A guide for selecting statistical techniques for analyzing social science data. (2nd ed.). Ann Arbor, MI: Survey Research Center, Institute for Social Research, The University of Michigan. Angell, M. (1981). Women in medicine: Beyond prejudice. New England Journal of Medicine. 304(19), 1161-1163. Angell, M. (1982). Juggling the personal and professional life. Journal of the American Medical Women's Association. 37(3), 64-68. Anonymous. (1986). (No Title). Journal of the American Medical Women's Association. 41(2), 58-60. Antonelli, M. A. (1986). III. Patient care issues. Hawaii Medical Journal. £5(12), 428-429. Antoine, L. (1966). Demand. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 21(10), 842. Are there too many doctors? (1986). World Health Organization Chronicle. 40(6), 240-241. Arnold, L. M . , Willoughby, T. L., Calkins, V., & Jensen, T. (1981). The achievement of men and women in medical school. Journal of the American Medical Women's Association. 36(7) , 213-221. Association of American Medical Colleges. (1988a). Women in medicine statistics. Washington DC: Author. Association of American Medical Colleges. Report. 2.(35) . Washington, DC: Author. (1988b). Weekly Association of American Medical Colleges. (1988c). Faculty roster system. Women and minorities on U.S. medical school faculties. Washington DC: Author. Association of American Medical Colleges. (1990a). Building a stronger women's program. Washington, DC: Author. Association of American Medical Colleges. Report. 4(13). Washington, DC: Author. (1990b). Weekly Association of American Medical Colleges. (1990c). AAMC Data Book. Statistical Information Related to Medical Education. Washington, DC: Author. 341 Association of American Medical colleges. Report. 5(14). Washington, DC: Author. (1991). Weekly Avery, M. E. (1981). Women in medicine, 1979: What are the issues? Journal of the American Medical Women's Association. 36(2), 79-81. Babbott, D . , Baldwin, D. C . , Jr., Killian, C. D., & Weaver, S. 0. (1989). Racial-ethnic background and specialty choice: A study of U.S. medical school graduates in 1987. Academic Medicine. 64(10), 595-599. Barclay, W. R. (1973). The future for medical education and women in medicine. Journal of the American Medical Women's Association. 28(2), 69-70. Barondess, J. A. (1981). Are women different? Some trends in the assimilation of women in medicine. Journal of the American Medical Women's Association. 36(3), 95-104. Barthauer, L. (1989). The medical mystique. The New Physician. 38(9), 15-19. Batchelor, A. J. (1990). Senior women physicians: the question of retirement. New York State Journal of Medicine. 90(6), 292-294. Battle, C. U. (1966). Wanted: more cheap, docile women doctors? Message from the president. Journal of the American Medical Women's Association. 41(5), 132,141. Baucom-Copeland, S., Copeland, E. T . , & Perry, L. L. (1983). The pregnant resident: Career conflict? Journal of the American Medical Women's Association. .38(4), 103-105. Bauder-Nishita, J. (1980). Gender specific differentials of medical practice in California. Women and Health. 5(3), 516. Bauknecht, R. (1967). Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22(6), 407-410. Baum, A. Z. (1980, February). The wall crumbles - ever so slowly. Medical Economics. 104-110. Baumgartner, L. (1966). Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 21,(10), 832-837. 342 Baxter Foundation. (1989). Women and health care. Health Management Quarterly. Third Quarter. Deerfield, IL: Author. Beck, L. (1988). The coming sex change in medical practice. Connecticut Medicine. 52(5), 307. Becker-Manheimer, 0. (1967). Utilization of medical womanpower in the specialty of radiology. In (No editor) Optimal utilization of medical womanpower. Proceedings of the xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22(7), 480-482. Beil, C., Sisk, D. R . , & Miller, W. E. (1980). A Comparison of specialty choices among senior medical students using BEM Sex-Role Inventory Scale. Journal of the American Medical Women's Association. 35(7), 178-181. Bennett, N. M., & Nickerson, K. G. (1990a). Letter to the Editor. New England Journal of Medicine. 322(18), 1316-1317. Bennett, N. M . , & Nickerson, K. G. (1990b). Women, families and academic medicine. New York State Journal of Medicine. 90(6), 284-285. Bergquist, S. R . , Duchac, B. W . , Schalin, V. A., Zastrow, J. F., Barr, V. L. , & Borowiecki, V. (1985). Perceptions of freshman medical students of gender differences in medical specialty choice. Journal of Medical Education. 6 0 (5), 379383. Berliner, H. S. (1987). Strategic Factors in U.S. Health Care Human Resources. Capital and Technology. Conservation of Human Resources Studies in Health Policy. Boulder CO: Westview Press. Bernstein, A. E. (1989). Gender equity. Journal of the American Medical Women's Association. 44(3), 84-86. Beshiri, P. H. (1969). The woman doctor. Her career in modern medicine. New York: Cowles Book Company. Bickel, J. (1988). Women in medical education. New England Journal of Medicine. 319(24), 1579-1584. Black, R. R . , II, Schmittling, G., & Stern, T. L. (1980). Characteristics and practice patterns of family practice residency graduates in the United States. The Journal of Family Practice. 11(5), 767-778. Blain, D. (1968). Women in psychiatry. In Roche Laboratories (Eds.). Careers in psychiatry. 111-127. New York: Macmillan Company. 343 Blanchette, P. L. (1986). I. Women in medicine: Patient care issues. Hawaii Medical Journal. 45(12), 426-427. Bluestone, N. R. (1990). Mainstreaming the minority and assumptions of rightful place. New York State Journal of Medicine. 90(6), 283-284. Blumberg, P., Flaherty, J. A., & Morrison, A. E. (1984). Social support networks and psychological adaptation of female and male medical students. Journal of the American Medical Women's Association. 39(5), 165-167. Bobula, J. D. (1980a). Income differentials between male and female physicians. In G. L. Glandon, & R. J. Shapiro (Eds.), Profile of medical practice 1980. Decade of the 1970s - the changing profile of medical practice, (pp. 123-133). Monroe, Wisconsin: American Medical Association. Bobula, J. D. (1980b). Work patterns, practice characteristics, and incomes of male and female physicians. Journal of Medical Education. 55(10), 826-833. Bodel, P. T . , & Short, E. M. (1972). Women in medicine: Views from a medical school. Clinical Research. 20(2), 125129. Bonar, J. W . , Watson, J. A., & Koester, L. S. (1982). Sex differences in career and family plans of medical students. Journal of the American Medical Womenfs Association. 37.(11) , 300-304. Boris, E . , & Honey, M. (1988). Gender, race and the policies of the Labor Department. Monthly Labor Review. 111(2), 2636. Bourne, P. G . , & Wikler, N. J. (1978). Commitment and the cultural mandate: Women in medicine. Social Problems. 25. 430-440. Bowers, J. Z. (1966). Women in medicine. An international study. The New England Journal of Medicine. 275 ( 7) . 362-365. Bowers, J. Z. (1967). Wife, mother and physician. Journal of the American Medical Women's Association. 22.(10), 760-764. Bowers, J. Z. (1968). Special problems of women medical students. Journal of Medical Education. 43.(5), 532-537. Bowman, M. A. (1986). Specialty choice of black physicians. Journal of the National Medical Association. 78(1), 13-15. Bowman, M. A . , & Allen, D. I. (1985). Stress and women physicians. New York: Springer-Verlag. 344 Bowman, M. A., & Gross, M. L. (1986). Overview of research on women in medicine - issues for public policymakers. Public Health Reports. 101(5), 513-521. Brodkin, A. M . , Shrier, D. K., & Buxton, M. (1982). Parenting and professionalism - a medical school elective. Journal of the American Medical Women's Association. 17(9), 227-230. Brown, E. P. (1962). (Moderator) Medical womanpower - can it be used more efficiently? Journal of the American Medical Women's Association. 17(12), 973-985. Brown, S.- L . , & Klein, R. H. (1982). Woman-power in the medical hierarchy. Journal of the American Medical Women's Association. 37(6), 155-164. Bulkley, B. H. (1981). Guest editorial: Woman M.D.: How, not may I. Journal of American Medical Women's Association. 36(3), 94. Bunker, J. P., & Pool, J. G. (1971). The case for more women in medicine: The Stanford program. New England Journal of Medicine. 285(1), 53. Burnley, C. S., & Burkett, G. L. (1986). Specialization: Are women in surgery different? Journal of the American Women's Association. 4^(5), 144-152. Butter, K . , Carpenter, E., Kay, B., & Simmons, R. (1985). Sex and status: Hierarchies in the health workforce. Public Health Policy Series. Washington, D.C.: American Public Health Association. Callan, C. M . , & Klipstein, E. (1981). Women physicians in Connecticut: A survey. Connecticut Medicine. 45(8) , 494-496. Calkins, E. V., Arnold, L. M . , & Willoughby, T. L. (1987). Gender differences in predictors of performance in medical training. Journal of Medical Education. 62(8), 682-685. Calmes, S. H. (1985). The first workshop for women in anesthesiology. Journal of the American Medical Women's Association. 40(5), 153-154. Campbell, M. A. (a pseudonym). (1973). Why would a girl go into medicine? Medical education in the United States: A guide for women. The Clearinghouse on Women's Studies. Old Westbury, New York: The Feminist Press. Campos-Outcalt, D., & Senf, J. H. (1989). Characteristics of medical schools related to the choice of family medicine as a specialty. Academic Medicine. 64(10), 610-615. 345 Cartwright, L. K. (1972a). Conscious factors entering into decisions of women to study medicine. Journal of Social Issues. 28(2), 201-215. Cartwright, L. K. (1972b). The personality and family background of a sample of women medical students at the University of California. Journal of the American Medical Women’s Association. 22(5), 260-266. Cartwright, L. K. (1977a). Personality changes in a sample of women physicians. Journal of Medical Education. 52(6), 467-474. Cartwright, L. K. (1977b). Continuity and noncontinuity in the careers of a sample of young women physicians. Journal of the American Medical Women's Association. 32.(9) , 316-321. Cartwright, L. K. (1987). Role montage: Life patterns of professional women. Journal of the American Medical Women's Association. 42(5), 142-148. Chapman, J. E. (1969). A common denominator in the equation toward a medical education. Journal of the American Medical Women's Association. 24.(7) , 561-565. Clark, E. J . , & Rieker, P. P. (1986). Gender differences in relationships and stress of medical and law students. Journal of Medical Education. 61(1), 32-40. Clavan, S., & Robak, N. (1978). Changing sex roles and the profession of medicine. Ethics in Science and Medicine. 5, 15-24. Clever, L. H. (1986). VI. The "superwoman" syndrome.Hawaii Medical Journal. 4jj(1 2 ), 436, 439. Coffin, S. E., & Babbott, D. (1989). Early and final preferences for pediatrics as a specialty: A study of US medical school graduates in 1983. Academic Medicine. 64(10), 600-605. Cohen, E. D., & Korper, S. P. (1976a). Women in medicine: A survey of professional activities, career interruptions, and conflict resolutions. Trends in medical education and specialization. (First of two parts). Connecticut Medicine. 40(2), 103-110. Cohen, E. D., St Korper, S. P. (1976b). Women in medicine: A survey of professional activities, career interruptions, and conflict resolutions. Trends in career patterns. (Second of two parts). Connecticut Medicine. 40(3), 195-200. 346 Cohen, M . , Woodward, C. A . , & Ferrier, B. M. (1988). Factors influencing career development: Do men and women differ? Journal of the American Medical Women's Association. 43.(5) , 142,147-154. Coombs, R. H . , & Hovanessian, H. C. (1988). Stress in the role constellation of female resident physicians. Journal of the American Medical Women's Association. 43.(1), 21-27. Coryllos, E. (1984). Making a difference in practice. In (No editor) Women in medicine. Making a difference. Proceedings of the Third regional conference. Journal of the American Medical Women*s Association. 39(3), 80, 106. Costanza, M. E. (1989). Women in medical leadership. Journal of the American Medical Women’s Association. 44.(6) , 185-186. Coste, C. (1975). Women in medicine: Progress and prejudice. The New Physician. 24(H)/ 25-33. Coulter, M. P. (1972). Residency programs for women. Journal of Medical Education. 47(10), 836-837. Council on Graduate Medical Education. (1989). First Report of the Council. Volume II. (DHHS Publication No. HRS-P-OD90-1). Washington DC: U.S. Government Printing Office. Cox, B. G . , & Lewis, L. A. (1981). Responses of medical women to a survey on sexist language in medical publications. Journal of the American Medical Women's Association. 36(1), 15-18. Crosse, V. M. (1967). Interrelationships. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22.(6), 410-412. Crovitz, E. (1980). Women entering medical school: The challenge continues. Journal of the American Medical Women's Association. 35(12), 291-298. Crowley, A. E., & Etzel, S. I. (1988). Graduate Medical Education in the United States. Journal of the American Medical Association. 260(88), 1093-1101. Cuca, J. M. (1979). The specialization and career preferences of women and men recently graduated from U.S. medical schools. Journal of the American Medical Women's Association. 34. 425-435. 347 Curry, L. (1983). The effect of sex on physician work patterns. In Research in Medical Education. 1983. Proceedings of the Twenty-Second Annual Conference. 144-150. Washington DC: Association of American Medical Colleges. Custer, W. S., & Dimon, D. E. (1987). Income differentials between women and men physicians. Paper prepared for presentation at the Western Economic Association meeting, Vancouver, July 7-11, 1987. Unpublished Manuscript. Dalessandri, K. (1988). The surgical work force and women surgeons. Journal of the American Medical Women's Association. 43(6), 169-170. Darga, Kenneth. (1988). Update on OHMA's physician census. OHMA News. 1(2), p. 3. Davidson, L. R. (1979). Choice by constraint. The selection and function of specialties among women physicians-in­ training. Journal of Health Politics. Policy and Law. 4(2), 200-2 2 0 . Day, P. (1982). Women doctors. Choice and constraint in policies for medical manpower. London: King's Fund Centre. D'Elia, G., & Johnson, I. (1980). Women physicians in a nonmetropolitan area. Journal of Medical Education. 55(7), 580-588. Dennis, T . , Harris, I., Petzel, R . , Lofgren, R . , Rich, E., McCollister, R . , Foley, J . , & Lifto, J. (1990). Influences of marital status and parental status on the professional choices of physicians about to enter practice. Academic Medicine. 65(12), 775-776. Denslow, J. S., Hosokawa, M. C., Campbell, J. D., Roberts, c. R . , & Samuels, M. E. (1984). Osteopathic physician location and specialty choice. Journal of Medical Education. 59(8), 655-661. Dial, T. H . , Bickel, J., & Lewicki, A. M. (1989). Sex differences in rank attainment among radiology and internal medicine faculty. Academic Medicine. 64.(4) , 198-202. Dickstein, L. J. (1990). Female physicians in the 1980s: Personal and family attitudes and values. Journal of the American Medical Women's Association. 45(4), 122-126. Dickstein, L. , & Nadelson, C. (Eds.). (1986). Women physicians in leadership roles. Washington DC: American Psychiatric Press. 348 Dickstein, L. J . , & Stephenson, J. J. (1987). A national survey of women physicians in administrative roles. Journal of the American Medical Women's Association. 42.(4) , 108-111. Donoghue, G. D. (1988). Eliminating salary inequities for women and minorities in medical academia. Journal of the American Medical Women's Association. 42(1), 28-29. Dralle, P. W . , Daum, D. V., Heim, D. D., & Elston, R. C. (1987). Sex role stereotypes in freshman medical students' perceptions of self and physicians. Journal of the American Medical Association. 42(3), 75-81. Dube, W. F. (1969). Women in graduate education. Journal of the American Medical Association. .210(8) , (Excerpt only) 1551-1553. Dube, W. F. (1973a). Women students in U.S. medical schools: Past and present trends. Journal of Medical Education. 42(2), 186-189. Dube, W. F. (1973b). U.S. medical student enrollments, 19681969 through 1972-1973. Journal of Medical Education. 48(3), 293-297. Dube, W. F., Stritter, F. T . , & Nelson, B. C. (1971). Study of U.S. medical school applicants, 1970-71. Journal of Medical Education. 46(10), 837-857. Ducker, D. G. (1978). Believed suitability of medical specialties for women physicians. Journal of the American Medical Women's Association. 33(1), 25-32. Ducker, D. G. (1986). Role conflict in women physicians: A longitudinal study. Journal of the American Medical Women's Association. 41(1), 14-16. Ducker, D. G. (1988). The professional context for women physicians: Collegial relations, role models, and mentors. Journal of the American Medical Women's Association. 43 (4), 106-108. Dykman, R. A., & Stalnaker, J. M. (1957). Survey of women physicians graduating from medical school 1925-1940. Part Two. Journal of Medical Education. 32.(3), 3-38. Eccles, J. S. (1987). Gender roles and women's achievementrelated decisions. Psychology of Women Quarterly. 11. 135172. Ehrenreich, B. (1975, November/December). The health care industry: A theory of industrial medicine. Social Policy. 411. 349 Eisenberg, C. (1981). Similarities and differences between men and women as students. Journal of the American Medical Women's Association. 36(2), 45-50. Eisenberg, C. (1983). Women as physicians. Journal of Medical Education. 58 (7), 534-541. Eisenberg, C. (1989). Medicine is no longer a man's profession, or, when the men's club goes coed it's time to change the regs. New England Journal of Medicine. 321(22). 1542-1544. Eisenberg, L. (1971a). Medical womanpower: A statistic goes "astray". American Journal of Orthopsychiatry. 41(3), 348349. Eisenberg, L. (1971b). Canard about women physicians. New England Journal of Medicine. 284(13), 734-735. Eisenberg, L. (1981). The distaff of Aesculapius - the married woman as physician. Journal of the American Medical Women's Association. J36(2) , 84-88. Elkin, E. M. (1971). Male thoughts of the female doctor's dilemma. The New England Journal of Medicine. 284(19). 1106. Elliot, C. M. (1981). Women physicians as workers. Journal of the American Medical Women's Association. 36(3), 105-108. Ellsbury, K . , Schneeweiss, R., Montano, D. E., Gordon, K. C . , & Kuykendall, D. (1987). Gender differences in practice characteristics of graduates of family medicine residencies. Journal of Medical Education. 62(11), 895-903. Elston, M. A. (1977). Women in the medical profession: Whose problem? in M. Stacey, M. Reid, C. Heath, & R. Dingwall (Eds.), Health and the division of labor. London: Croom Helm. England, M. J. (1987). Leadership for women: Medical academia. Journal of the American Medical Association. 42(1), 4. Engleman, E. G. (1974). Attitudes toward women physicians a study of 500 clinic patients. The Western Journal of Medicine, 120(2). 95-100. F., W. C. (1986). Women in medicine: Stamping out stereotypes. The Internist. 27(3), 5-6. Fagin, C. (1984). (No Title). Journal of the American Medical Women's Association. 39(3), 88-90. 350 Farrell, K . , Witte, M. H., Holguin, M . , & Lopez, S. (1979). Women physicians in medical academia. Journal of the American Medical Association. 241(26). 2808-2812. Fay, M. (1966). Women physicians in the twenty-first century. Goals. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 21(10), 843-846. Fennema, K., Meyer, D. L., & Owen, N. (1990). Sex of physician: Patients' preferences and stereotypes. Journal of Family Practice. 30(4), 441-446. Ferrier, B. M., & Woodward, C. A. (1982). Career choices, work patterns and perceptions of undergraduate education of McMaster medical graduates: Comparison between men and women. Canadian Medical Association Journal. 126. 1411-1414. Fett, I. (1976). The future of women in Australian medicine. The Medical Journal of Australia. 2(Suppl. 2), 33-39. Finseth, K. A. (1978). Overcoming barriers to women in medicine: A model program. Journal of the American Medical Women's Association. 21(1)/ 33-37. Fiorentine, R. (1987). Men, women, and the premed persistence gap: A normative alternatives approach. American Journal of Sociology. 92(5), 1118-1139. Fleischer, R. A . , & Abney, E. (1990). New York State Journal of Medicine. 90(6), 335-336. Freedman, A. M . , Kaplan, H. I., & Sadock, B. J. (1980). Comprehensive textbook of psychiatry III. Baltimore: williams it Wilkins. Freeman, R. M . , & Waickman, L. A. (1988). Influence of gender on completing an internal medicine residency. Journal of Medical Education. 62(12), 914-916. Freiman, M. P., & Marder, W. D. (1984). Changes in the hours worked by physicians, 1970-1980. American Journal of Public Health. 74(12), 1348-1352. Friedman, E. (1988). Women and medicine: From tension to truce. The Western Journal of Medicine. 149(6), 681-682. Fryer, G. C. (1986). V. Career choices and tensions: The woman internist. Hawaii Medical Journal. 45(12), 432, 435. 351 Fulop, T. (1986a). Health personnel for "health for all": progress or stagnation? Part I. WHO Chronicle. 40(5), 194199. Fulop, T. (1986b). Health personnel for "health for all": progress or stagnation? Part II. WHO Chronicle. 40(6), 222225. Geyman, J. P. (1980). Increasing number of women in family practice: An overdue trend. The Journal of Family Practice. 10(2), 207-208. Ginzberg, E. (Ed.). (1986). From physician shortage to patient shortage. The uncertain future of medical practice. Boulder, CO: Westview Press. Glandon, G. L., & Shapiro, R. J. (Eds.). (1980). Profile of medical practice 1980. Decade of the 1970s - the changing profile of medical practice. Monroe, Wisconsin: American Medical Association. Goldblatt, A., & Goldblatt, P. B. (1976). The status of women physicians: A comparison of USMG women, USMG men, and FMGs. Journal of the American Medical Women's Association. 21(8), 325-328. Golden, W. E. (1989). Initial career choices of medical school honors graduates in the early 1970s and 1980s. Academic Medicine. 64(10), 616-621. Goldstein, M. Z., Bromet, E. J., Hanusa, B. H . , & Lasell, R. L. (1981). Psychiatrists* life and work patterns: a statewide comparison of women and men. American Journal of Psychiatry. 138(7), 919-924. Gomez, T. (1967). Demand for Philippine medical womanpower. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22(7), 478-480. Goodwin, N. J. (1985). The black woman physician. New York State Journal of Medicine. 85(4), 145-147. Gough, H. G. (1975). Specialty preferences of physicians and medical students. Journal of Medical Education. 50(6), 581588. Graham, R. (1984). (No Title). Journal of the American Medical Women's Association. 2£(3), 85-86. 352 Grant, L. (1988). The gender climate of medical school: Perspectives of women and men students. Journal of the American Medical Women's Association. 42(4), 109-110, 115119. Graves, P. L . , & Thomas, C. B. (1985). Correlates of midlife career achievement among women physicians. Journal of the American Medical Association. 254(6), 781-787. Gray, J. D. (1983). The married professional woman: An examination of her role conflicts and coping strategies. Psychology of Women Quarterly. 7(3), 235-243. Gray, M. J . , & Ackerman, J. (1978). Attitudes of women medical students toward obstetrics and gynecology. Journal of the American Medical Women's Association. 32(4), 162-164. Griffen, W. 0., & Schwartz, R. W. (1990). Controllable lifestyle as a factor in choosing a medical career. The American Journal of Surgery. 159 (2 ), 189-190. Gross, L. (1990). An interview with Nancy E. Gary, MD, a medical school dean. New York State Journal of Medicine. 90(6), 321-324. Gross, W . , & Crovitz, E. (1975). A comparison of medical students' attitudes toward women and women medical students. Journal of Medical Education. 50(4), 392-394. Gruppen, L. D., & Brown, D. R. (1981). Longitudinal changes in specialty preferences: Sex differences. Research in Medical Education: 1981. Proceedings of the Twentieth Annual Conference, 100-103. Washington DC: Association of American Medical Colleges. Haar, E . , Halitsky, V . , & Strieker, G. (1975). Factors related to the preference for a female gynecologist. Medical Care. 12(9), 782-790. Hafferty, F. W. (1986). Physician oversupply as a socially constructed reality. Journal of Health and Social Behavior. 27(12), 358-369. Hammond, J. M. (1981). Social support groups, women's programs, and research on gender differences: The bad press for women in medical education literature. Journal of the American Medical Women's Association. 26(5), 162-166. Harris, J. (1984). Making a difference in health policy. In (No editor) Women in medicine. Making a difference. Proceedings of the Third regional conference. Journal of the American Medical Women's Association. 29(3), 77-79. 353 Harris, M. B . , & Conley-Muth, M. A. (1981). Sex role stereotypes and medical specialty choice. Journal of the American Medical Women’s Association. 36(8), 245-252. Harris, M. B., & Conley-Muth, M. A. (1983). Sex role stereotypes and choice of physician. Journal of the American Medical Women's Association. 3JJ(6) , 149-154. Harward, D. H . , Lyons, C. M . , Porter, C. Q . , & Hunter, R. S. (1981). A comparison of the performance of male and female medical students and residents. Journal of Medical Education. 56(10), 853-855. Hayes, M. D. (1981). The impact of women physicians on social change in medicine: The evolution of humane health care delivery systems. Journal of the American Medical Women's Association. 36(2), 82-84. Heilman, M. E., & Kram, K. E. (1983). Male and female assumptions about colleagues' views of their competence. Psychology of Women Quarterly. 7(4) , 329-337. Heins, M. (1985). Update: Women in medicine. Journal of the American Medical Women's Association. 40(2), 43-50. Heins, M . , Smock, S., Jacobs, J . , & Stein, M. (1976). Productivity of women physicians Journal of the American Medical Association. 236(17), 1961-1964. Heins, M . , Smock, S., & Martindale, L. (1978). Current status of women physicians. International Journal of Women's Studies. 1(3), 297-305. Heins, M . , Smock, S., Martindale, L., Jacobs, J., & Stein, M. (1977). Comparison of the productivity of women and men physicians. Journal of the American Medical Association. 237(23), 2514-2517. Heins, M . , Smock, S., Martindale, L., Stein, M . , & Jacobs, J. (1977). A profile of the woman physician. Journal of the American Medical Women's Association. 32.(11), 421-427. Hellstedt, L. McG. (1967). Interrelationships. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22.(6), 413-414. Hendrickson, R. M. (1975, July/August). Women physicians: Where they rank on the money tree. Prism, (pp. 20-21). 354 Henkel, R. E. (1976). Tests of significance. Sage University Paper Series on Quantitative Applications in the Social Sciences, 07-004. Beverly Hills & London: Sage Publications. Henry, G. (1967). Inter-relations. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22.(6) , 415-419. Hilberman, E., Gispert, M . , & Harper, J. (1976). Impact of a district branch task force on women. American Journal of Psychiatry. 133(10), 1159-1164. Hildebrand, D. K . , Laing, J. D., & Rosenthal, H. (1977). Analysis of ordinal data. Sage University Paper Series on Quantitative Applications in the Social Sciences, 07-008. Beverly Hills & London: Sage Publications. Hoferek, M. J., & Sarnowski, A. A. (1981). Feelings of loneliness in women medical students. Journal of Medical Education. 56(5), 397-403. Hojat, M . , Gonnella, J. S., Moses, S., & Veloski, J. J. (1987). Comparisons of male and female physicians on their practice patterns, professional activities, and perceptions of professional problems. Research in Medical Education 1987. Proceedings of the Twenty-Sixth Annual Conference. 2327. Washington DC: Association of American Medical Colleges. Hojat, M . , Gonnella, J. S., Veloski, J. J., & Moses, S. (1990). Differences in professional activities, perceptions of professional problems, and practice patterns between men and women graduates of Jefferson Medical College. Academic Medicine, 65(12), 755-761. Holder, A. R. (1979). Women physicians and malpractice suits. Journal of American Medical Women's Association. 34(6), 239-240. Holmstrom, M. (1967). Optimal utilization of medical womanpower. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22.(5), 336-337. Holton, S. C. (1969) . The woman physician: A study of role conflict. Journal of the American Medical Woman's Association. 24.(8), 638-645. Hooper, J. (1989). Full-time women general practitioner - an invaluable asset. Journal of the Roval College of General Practitioners. 3^(324), 289-291. 355 Hsu, J. (1986). IV. Playing by his rules. Hawaii Medical Journal. 45(12), 431-432. Iverson, G. R . , & Norpoth, H. (1976). Analysis of variance. Sage University Paper Series on Quantitative Applications in the Social Sciences, 07-001. Beverly Hills & London: Sage Publications. Jacob, H. (1984). Using published data. Errors and remedies. Sage University Paper Series on Quantitative Applications in the Social Sciences, 07-042. Beverly Hills & London: Sage Publications. Janus, C. L . , Janus, S. S., Price, S., & Adler, D. (1983). Residents: The pressure's on the women. Journal of the American Medical Women's Association. 3j£(l) , 18-21. Johnson, D. G. (1983). Physicians in the making. Personal, academic, and socioeconomic characteristics of medical students from 1950 to 2000 (1st Ed.). San Francisco: JosseyBass Inc. Johnson, D. G . , & Hutchins, E. B. (1966). Doctor or dropout? A study of medical student attrition. Journal of Medical Education. (Special Issue). Jussim, J . , & Muller, C. (1975). Medical education for women: How good an investment? Journal of Medical Education. 50(6), 571-580. Kaplan, H. I. (1972). Part-time residency training: An approach to the graduate training of some women physicians. Journal of the American Medical Women's Association. 27(12), 648-650. Kaplan, H. I., Kaplan, H. S., & Freedman, A. M. (1964). Residency training in psychiatry for physician mothers. Journal of the American Medical Association. 189(1), 99-102. Kaplan, H. S. (1962). A new concept of graduate training for women physicians. Journal of the American Medical Women's Association. 17.(10), 820-821. Kaplan, S. R. (1981). Motivations of women over 30 for going to medical school. Journal of Medical Education. 56(10), 856-858. Kaplan, S. R. (1982). Medical school and women over 30. Journal of the American Medical Women's Association. 37(2), 39-50. 356 Kasteler, J. M . , & Hulme, S., III. (1980). Attitudes toward women physicians in a Mormon community. Journal of the American Medical Women's Association. 35(2), 37-41. Kehrer, B. H. (1976). Factors affecting the incomes of men and women physicians: An exploratory analysis. The Journal of Human Resources. 11(4), 526-545. Keyserling, M. D. (1969). "'Waste no talent - neglect no skill'". Journal of the American Medical Women's Association. 24(2). 146-153. Kiecolt, K. J . , & Nathan, L. E. (1985). Secondary analysis of survey data. Sage University Paper Series on Quantitative Applications in the Social Sciences, 07-053. Beverly Hills & London: Sage Publications. Kinder, B. K. (1985). Women and men as surgeons: Are the problems really different? Current Surgery. 42.(2), 100-104. Kirchner, S. G . , Kossoff, J . , & Pickens, D. R. III. (1984). Women in radiology in the United States: 1982 survey of their professional practices. American Journal of Roentgenology. 142.(5), 1055-1059. Kletke, P. R . , Marder, W. D., & Silberger, A. B. (1990). The growing proportion of female physicians: Implications for US physician supply. American Journal of Public Health. 80(3). 300-304. Kletke, P. R., Schleiter, M. K., & Tarlov, A. R. (1987). Changes in the supply of internists: The internal medicine population from 1978 to 1998. Annals of Internal Medicine. 107(1), 93-100. Kohrman, C. H., Lyttle, C. S., Andersen, R. M . , & Levey, G. S. (1989). National study of internal medicine manpower. XIV: patterns of residency and fellowship over time. 1987 update. Archives of Internal Medicine. 149(10). 2179-2185. Kosa, J . , & Coker, R. E., Jr. (1965). The female physician in public health. Conflict and reconciliation of the sex and professional roles. Sociology and Social Research. 49 (4), 294-305. Kris, K. B. (1985). Developmental strains of women medical students. Journal of the American Medical Women's Association. 40(5), 145-148. Kritzer, H., & Zimet, C. N. (1967, January). A retrospective view of medical specialty choice. Journal of Medical Education. 42., (pp. 47-53) . 357 Kroser, L. S. (1987). The growing influence of women in medicine. In D. B. Nash (Ed.), Future practice alternatives in medicine, (pp. 109-137). New York: Igaku-Shoin. Kutner, N. G., & Brogan, D. R. (1980). A comparison of the practice orientations of women and men students at two medical schools. Journal of the American Medical Women's Association. 15(3), 80-86. Lake, A. (1967). Drop those prejudices against women doctors. Journal of the American Medical Women's Association. 22.(6), 402-406. Langwell, K. M. (1982). Differences by sex in economic returns associated with physician specialization. Journal of Health Politics. Policy and Law. 6(4), 752-761. Lanska, M. J., Lanska, D. J., & Rimm, A. A. (1990). Effect of rising percentage of female physicians on projections of physician supply. Journal of Medical Education. 59(11), 849927. LeFevre, M. L . , & Colwill, J. M. (1983). Practice location as a function of medical school and residency location: Implications for resident selection. The Journal of Family Practice. 16(6), 1157-1160. Lerner, M. R. (1981). The women: Who, where, when and why. Journal of the American Medical Women's Association. 36(1), 5-12. Leserman, J. (1980). Changes in the professional orientation of medical students: A follow-up study. Journal of Medical Education. 55(5), 415-422. Leserman, J. (1981). Men and women in medical school. How they change and how they compare. New York: Praeger Publishers. Levey, B. A., Gentile, N. O., Jolly, H. P., Beaty, H. N., & Levey, G. S. (1990). Comparing research activities of women and men faculty in departments of internal medicine. Academic Medicine. 65(2), 102-106. Levin, B. S. (1980). Women and medicine. Metuchen, New Jersey: The Scarecrow Press. Levinson, W . , Tolle, S. W . , & Lewis, C. E. (1989). Women in academic medicine. Combining career and family. New England Journal of Medicine. 321(22), 1511-1517. 358 Levinson, W . , Tolle, S. W . , & Lewis, C. E. (1990). Letter to the Editor. New England Journal of Medicine. 322(18), 13161317. Lewis-Beck, M. S. (1980). Applied regression. An introduction. Sage University Paper Series on Quantitative Applications in the Social Sciences, 07-022. Beverly Hills & London: Sage Publications. Lieu, T. A., Schroeder, S. A., & Altman, D. F. (1989). Specialty choices at one medical school: Recent trends and analysis of predictive factors. Academic Medicine. 64(10), 622-629. Linn, L. S., Cope, D. W., & Leake, B. (1984). The effect of gender and training of residents on satisfaction ratings by patients. Journal of Medical Education. 59.(12) , 964-966. Lisoskie, S. (1986). Why work fewer hours. Journal of the American Medical Women's Association. 41(3), 73,90. Lloyd-Green, L. (1967). Supply. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22.(5) , 337-338. Lopate, C. (1968). Men and women in medical school. Baltimore: The Johns Hopkins Press. Lorber, J. (1975). Women and medical sociology: Invisible professionals and ubiquitous patients. In M. Millman & M. Kanter (Eds.), Another voice. Feminist perspectives on social life and social science. Garden City, New York: Anchor Books. Lorber, J. (1981). The limits of sponsorship for women physicians. Journal of the American Medical Women's Association. 36(11), 329-338. Lorber, J. (1982). How physician spouses influence each other's careers. Journal of the American Medical Women's Association. 37(1), 21-26. Lorber, J. (1984). Women physicians careers, status and power. New York & London: Tavistok Publications. Lorber, J. (1986). Sisterhood is synergistic. Journal of the American Medical Women's Association. 41 (4 )r 116-119. Lorber, J. (1987). A welcome to a crowded field: Where will the new women physicians fit in? Journal of the American Medical Women's Association. 42.(5), 149-152. 359 Lorber, J., & Ecker, M. (1983). Career development of female and male physicians. Journal of Medical Education. 58(6)» 447-456. Loutsch, E. , Yandow, V., Semlitz, L., & Berger, H. (1982). Women leaders in psychiatry. Journal of the American Medical Women's Association. 37.(10), 255-259. Lovelace, J. C. (1985). Career satisfaction and role harmony in Afro-American women physicians. Journal of American Medical Women's Medical Association. 40(4), 108-110. Lowenstein, L. M. (1971). [Letter to the Editor]. New England Journal of Medicine. 284(13), 735. Lyacki, H. , Josef, N . , & Chapin, K. (1983). Sex bias in the training of psychiatric residents. Journal of the American Medical Women's Association. 28(4), 105-107. Maheux, B. , Dufort, F . , & Beland, F. (1988). Professional and sociopolitical attitudes of medical students: Gender differences reconsidered. Journal of the American Medical Women's Association. 43.(3), 73-76. Maheux, B., Dufort, F . , Lambert, J . , & Berthiaume, M. (1988). Do female general practitioners have a distinctive type of medical practice? Canadian Medical Association Journal. 139. 737-740. Maheux, B., Dufort, F . , Lambert, J . , & Levesque, A. (1989). The professional attitudes and practice characteristics of male and female specialists. Journal of the American Medical Women's Association. 44(5 ), 154-158. Markert, R. J. (1987). Gender differences in the choice of a primary care medical specialty. Journal of Medical Education. (52(7), 601-603. Marks, G., & Beatty, W. K. (1972). Women in white. New York: Charles Scribner's Sons. Marquart, J. A., Franco, K. N., & Carroll, B. T. (1990). The influence of applicants' gender on medical school interviews. Academic Medicine. 65(6), 410-411. Martin, S. C., Arnold, R. M . , & Parker, R. M. (1988). Gender and medical socialization. Journal of Health and Social Behavior. 29(12), 333-343. Mattera, M. D. (1980, February). Female doctors: Why they're on an economic treadmill. Medical Economics. 98-101. 360 Matteson, M. T., & Smith, s. V. (1977). Selection of medical specialties: Preferences versus choices. Journal of Medical Education. 52(7), 548-554. McGrath, E., & Zimet, C. N. (1977a). Female and male medical students: Differences in specialty choice selection and personality. Journal of Medical Education. 52(47), 293-300. McGrath, E . , & Zimet, C. N. (1977b). Similarities and predictors of specialty interest among female medical students. Journal of the American Medical Women's Association. 32.(10), 361-373. McNamara, M. F. (1985). Women surgeons: How much of an impact? Current Surgery. 42.(2), 94-99. Mejia, A., Pizurki, H., & Royston, E. (1980). Foreign medical graduates. The case of the United States. Lexington, Mass: D.C. Heath and Company. Michigan Association of Governing Boards of State Universities. (1988). Report on status of women faculty in Michigan public universities. Haslett, Michigan: Author. Michigan Association for Medical Education. (1991). Regulation of Resident Hours: The Status of Michigan Residency Programs in 1990. East Lansing, Michigan: Author. Michigan Council on Graduate Medical Education (MCGME). (1987). Annual Report. Lansing, Michigan: Author. Michigan Department of Commerce. (1990). (Untitled Classification of Michigan counties into rural and urban). Lansing, Michigan: Author. Mick, S. S., & worobey, J. L. (1984). Impact of women and foreign graduates on specialty distribution of U.S. house officers. Journal of Medical Education. 59(12), 921-927. Mitchell, J. B. (1984, Winter). Why do women physicians work fewer hours than men physicians? Inguirv. 21 361-368. Mitchell, J. B., Schurman, R . , & Cromwell, J. (1988). The changing nature of physicians' office visits. Health Services Research. 23.(4), 575-591. Moldow, G. (1987). Women doctors in gilded-age Washington. Race, gender and professionalisation. Urbana, Illinois: University of Illinois Press. 361 Moore-West, M . , Lucero, S., Christy, J., & Kaufman, A. (1982). A descriptive study of women physicians in rural practice. Journal of the American Medical Women's Association. 3J7 (10) , 267-270. Morantz, R. M . , Pomerleau, C. S., & Fenichel, C. H. (Eds.). (1982). In her own words; Oral histories of women physicians. Westport, Connecticut: Greenwood Press. Morantz-Sanchez, R. M. (1985). Sympathy and science. Women physicians in American medicine. Oxford: Oxford University Press. Morgan, B. C. (1971). Admission of women into medical schools in the United States: Current status. The Woman Physician. 26(6), 305-309. Morrow, L. E. (1973). Statistics on women in medicine and addendum to publicity and public relations. 1972 Annual Report. Journal of the American Medical Women’s Association. 28(11), 600-602. Myers, D. (1986). II. Patient care issues. Hawaii Medical Journal. 45(12), 427-428. Myers, M. F. (1985). The medical students of 1985: Toward meeting their personal needs. Journal of the American Medical Women's Association. 40(5), 141-144. Nadelson, c. (1983). The woman physician: Past, present and future. In J. P. Callan (Ed.), The physician. A professional under stress (pp. 261-276). Norwalk, Connecticut: AppletonCentury-Crofts. Nadelson, c . , & Notman, M. T. (1972). The woman physician. Journal of Medical Education. 47(1), 176-183. Nadelson, C. , Notman, M. T., & Lowenstein, P. (1981). A follow-up study of Harvard medical school graduates, 19671977. Journal of the American Medical Women's Association. 36(2), 51-62. Nemir, R. L. (1978). Women in medicine during the last half century. Journal of the American Medical Women's Association. 31(5), 201-206. Nickerson, K. G . , Bennett, N. M . , Estes, D., & Shea, S. (1990). The status of women at one academic medical center. Breaking through the glass ceiling. Journal of the American Medical Association. 264(14), 1813-1817. 362 Norcini, J. J. , Fletcher, S. W., Quimby, B. B . , & Shea, J. A. (1985). Performance of women candidates on the American Board of Internal Medicine certifying examination. Annals of Internal Medicine. 102(1), 115-118. Norris, F. S. (1969). Medical womanpower. [Letter to the Editor]. New England Journal of Medicine. 281(16), p. 910. Notman, M . , & Nadelson, C. C. (1973). Medicine: A career conflict for women. American Journal of Psychiatry. 130(10), 1123-1127. Ochberg, R. L. , Barton, G. M . , & West, A. N. (1989). Women physicians and their mentors. Journal of the American Medical Women's Association. 44(4), 123-126. O'Connell, M. A . , & Beighton, F. (1979). Students' career plans and the medical profession. Journal of Medical Education. 54(6), 509-511. Offenbach, B. (1982). Four-letter words to avoid. Journal of the American Medical Women's Association. 37.(4), 103-104. Ogle, K. S., Henry, R. C., Durda, K . , & Zivick, J. D. (1986). Gender-specific differences in family practice graduates. Journal of Family Practice. 23.(4) , 357-360. Palac, D. M . , Lee, M. A., & Collins, J. F. (1990). Letter to the Editor. New England Journal of Medicine. 322(18). 1317. Parkhouse, J. (1989). Half our medical future. Medical Education, 23., 1-2. Parrish, J. B. (1971). Women in medicine: What can international comparisons tell us? The Woman Physician. 26(7), 352-361. Pearse, W. H. (1990). Women and specialties [Letter to the Editor]. Journal of the American Medical Association. £63.(17) , p. 2298. Pennell, M. Y., & Renshaw, J. E. (1972). Distribution of women physicians, 1970. Journal of the American Medical Women's Association. 27(4), 197-203. Pennell, M. Y., & Renshaw, J. E. (1973). Distribution of women physicians, 1971. Journal of the American Medical Women's Association. 28(4), 181-186. Perun, P. J. , & Bielby, D. D. V. (1981). Towards a model of female occupational behavior: A human development approach. Psychology of Women Quarterly. 6(2), 234-252. 363 Phelps, C. E. (1968). Women in American medicine. Journal of Medical Education. 43.(8), 916-924. Pirami, E. (1967). Supply. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22.(5), 338-339. Potter, R. L. (1983). Issues for women in psychiatric residency training. In (No editor) Resident, woman, wife, mother: Issues for women in training. Journal of the American Medical Women's Association. 38(4), 98-102. Powers, L . , Parmelle, R. D., & Wiesenfelder, H. (1969). Practice patterns of women and men physicians. Journal of Medical Education. 44(6), 481-491. Psathas, G. (1968). Toward a theory of occupational choice for women. Sociology and Social Research. 52(2), 253-268. Quadagno, J. (1976). Occupational sex-typing and internal labor market distributions: An assessment of medical specialties. Social Problems. 23., 442-453. Radius, S. M . , Becker, M. H . , Smith, D. R . , & Katatsky, M. E. (1979). The rejected medical school applicant; Sex differences in attitude and outcomes. Journal of the American Medical Women's Association. 34, 208. Ramey, E. (1984). Making a difference in research. In (No editor) Women in medicine. Making a difference. Proceedings of the Third regional conference. Journal of the American Medical Women's Association. 39(3), 74-76. Ramos, S. M . , & Feiner, C. J. (1989). Women surgeons: A national survey. Journal of American Medical Women's Association. 44(1)/ 21-25. Raymond, J. G. (1982). Medicine as a patriarchal religion. The Journal of Medicine and Philosophy. 7(2), 197-216. Reinhardt, U. E. (1975). Physician productivity and the demand for health manpower. Cambridge, Massachusetts: Ballinger Publishing Company. Reiman, A. S. (1980). Here come the women. The New England Journal of Medicine, 302(22), 1252-1253. Reiman, A. S. (1989). The changing demography of the medical profession. The New England Journal of Medicine. 321(22), 1540-1542. 364 Renshaw, J. E., & Pennell, M. Y. (1971). Distribution of women physicians, 1969. The Woman Physician. 26(4), 187-195. Reynolds, H. T. (1977). Analysis of nominal data. Sage University Paper Series on Quantitative Applications in the Social Sciences, 07-007. Beverly Hills & London: Sage Publications. Rinke, C. M. (1981a). The Economic and academic status of women physicians. Journal of the American Medical Association. 245(22), 2305-2306. Rinke, C. M. (1981b). The professional identities of women physicians. Journal of the American Medical Association. 245(23), 2419-2421. Roberts, S. M. H. (1967). Demand. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22(7), 482-484. Roeske, N. A., & Lake, K. (1977). Role models for women medical students. Journal of Medical Education. 52(6), 459466. Roessler, R . , Collins, F., & Mefferd, R. B. (1975). Sex similarities in successful medical school applicants. Journal of the American Medical Women's Association. 30(6), 254-265. Romm, S. (1982). Women doctor in search of a job. Journal of the American Medical Women's Association. J37(l) , 11-15. Roos, N. P., Gaumont, M . , & Colwill, N. L. (1977). Female and physician: A sex role incongruity. Journal of Medical Education. 52(4), 345-346. Rosen, G. (1972). Changing attitudes of the medical profession to specialization. In E. Freidson & J. Lorber (Eds.), Medical men and their work. Chicago: AldineAtherton. Rosen, R. H . , Heins, M . , & Martindale, L. J. (1981). Practice plans of today's medical students. Journal of Medical Education. 56(1), 57-59. Rosenlund, M. L . , & Oski, F. A. (1967). Women in medicine. Annals of Internal Medicine. 66(5), 1008-1012. Rosenthal, P. A., & Eaton, J. (1982). Women MDs in America: 100 years of progress and backlash. Journal of the American Medical Women's Association. 37(5), 129-133. 365 Ruben, R. J. (1972). Women in medicine: Past, present and future. Journal of the American Medical Women's Association, 27(5), 251-259. Santilli, V. C. (1990). Perspectives on women physicians in New York State: past, present, and future. New York State Journal of Medicine. 90(6), 317-321. Scadron, A. (1980). AMWA's experiment in planned change: A report on the "women in medical academia" project. Journal of the American Medical Women's Association. 35(12), 299301. Schaller, J. G. (1990). The advancement of women in academic medicine. Journal of American Medical Association. 264(14), 1854-1855. Schermerhorn, G. R . , Colliver, J. A., Verhulst, S. J . , & Schmidt, E. L. (1986). Factors that influence career patterns of women physicians. Journal of the American Medical Women's Association. 41(3), 74-78. Schindler-Baumann, J. (1967). Interrelationships. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22.(6) , 419-421. Schroeder, L. D. , Sjoquist, D. L., & Stephan, P. E. (1986). Understanding regression analysis. An introductory guide. Sage University Paper Series on Quantitative Applications in the Social Sciences, 07-057. Beverly Hills & London: Sage Publications. Schwartz, N. B. (1973). Why women form their own professional organizations. Journal of the American Medical Women's Association. 2jJ(l) , 12-15. Schwartz, R. W . , Jarecky, R. K., Strodel, W. E., Haley, J. V., Young, B., & Griffen, W. 0., Jr. (1989). Controllable lifestyle: A new factor in career choice by medical students. Academic Medicine. 64(10), 606-609. Schwartz, W. B. , Sloan, F. A., & Mendelson, D. N. (1988). Why there will be little or no physician surplus between now and the year 2000. The New England Journal of Medicine. 218(14), 892-897. Sclabassi, S. E. (1984). Sex differences in medical students' perceptions of clinical teaching activities. Journal of the American Medical Women's Association. 2£(5), 163-164. 366 Scully, A. L. (1982). From the halls of academia. Journal of the American Medical Women's Association. 37(10), 265. Shank, S. E. (1988). Women and the labor market: The link grows stronger. Monthly Labor Review. (3), 3-8. Shapiro, C. S., Stibler, B.- J., Zelkovic, A. A., & Mausner, J. S. (1968). Careers of women physicians: A survey of women graduates from seven medical schools, 1945-1951. The Journal of Medical Education. 43(10), 1033-1040. Shapiro, E., & Driscoll, S. (1977). Reduced-schedule residencies: Innovations in graduate medical education. Journal of the American Medical Women's Association. 32.(8) , 291-292. Shapiro, E. C., Lane, M. S., Driscoll, S. G . , & Books, E. F. (1980). The supply of reduced-schedule residencies. Journal of the American Medical Women's Association. 35(2), 42,4750. Shapiro, J. (1981). Children during residency: It's easier if you're a man. Journal of the American Medical Women's Association. 36(7), 227-231. Shapiro, J. (1982). Pregnancy during residency: Attitudes and policies. Journal of the American Medical Women's Association. 37(4), 96-103. Shaw, L. B . , & Shapiro, D. (1987). Women's work plans: contrasting expectations and actual work experience. Monthly Labor Review. (11), 7-13. Shore, E. G. (1984). Making a difference in academia. In (No editor) Women in medicine. Making a difference. Proceedings of the third regional conference. Journal of the American Medical Women's Association. 39(3), 81-83,106. Silberger, A. B., Marder, W. D., & Willke, R. J. (1987, Winter). Data watch. Practice characteristics of male and female physicians. Health Affairs. 104-109. Sinai, S., Weavil, P., & Camp, M. G. (1988). Survey of women physicians on issues relating to pregnancy during a medical career. Journal of Medical Education. 63.(7), 531-538. Sirridge, M. S. (1985). The mentor system in medicine - how it works for women. Journal of the American Medical Women's Association. 40(2), 51-53. 367 Smith, I. K . , Lancaster, C. J., & Fleming, G. A. (1984). Medical students' feelings related to isolation and school environment. Journal of the American Medical Women's Association. 19(5), 159-162. Spector, P. E. (1981). Research designs. Sage University Paper Series on Quantitative Applications in the Social Sciences, 07-023. Beverly Hills & London: Sage Publications. Spellacy, W. N. (1982). Survival of departments of obstetrics and gynecology: Ten points for Association of Professors of Gynecology and Obstetrics action in the 1980s. American Journal of Obstetrics and Gynecology. 1 4 4 (6 ), 626629. Spieler, C. (Ed.). (1977). Women in medicine - 1976. Baltimore, Maryland: Josiah Macy, Jr. Foundation, Waverly Press. Spiro, H. M. (1975). Myths and mirths - women in medicine. The New England Journal of Medicine. 292 1 7) , 354-356. SPSS Inc. (1990). SPSS reference guide. Chicago: SPSS Inc. Spurlock, J. (1984). (No Title). Journal of the American Medical Women's Association. 39(3), 85,89. Stern, M. S. (1976). Impacts of changes in the physician manpower sex ratio. Journal of Medical Education. 51(12), 1012-1013. Stiles, R. G. (1990). Letter to theEditor. Journal of Medicine. 322.(18), 1318. Stohl, W. (1990). Letter to the Editor. of Medicine. 322(18), 1318. New England New England Journal Storrie, V. M. (1966). Supply. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 21(10), 837-839. Sullivan, M. P. (1974). A new era: Challenges for the woman physician. Journal of the American Medical Women's Association. 29(1), 9-11. Surawicz, F. G., Norton, J. C., & Abies, B. S. (1976). Comparative study of male and female medical students at the University of Kentucky. Journal of the American Medical Women's Association. .31(5) , 205-210. 368 Swanson, A. G . , Randlett, R. R., Haynes, R. A., & Killian, C. D. (1989). AAMC Data Report. Present activities of 1989 U.S. medical graduates. Academic Medicine. 64(10), 630-633. Szkop-Frankiel, S. (1967). Supply. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22.(5), 338, 340. Tamburrino, M. B . , Franco, K. N., Evans, C. L., & Seidman, P. A. (1985). Finances: A hidden cost to women medical students. Journal of the American Medical Women's Association. 40(5), 149-151. Thomas, C. B. (1976). What becomes of medical students: The dark side. The Johns Hopkins Medical Journal. 138(5), 185195. Thomas, M. A. (1970). "Should a girl be encouraged to become a physician?". The Women Physician. 25(10), 629-630. Thomas, M. A. (1971). Is there a need for postgraduate parttime training programs? The Woman Physician. .26(11) , 551552. Titus-Dillon, P. Y., & Johnson, D. G. (1989). Female graduates of a predominantly black college of medicine: Their characteristics and challenges. Journal of the American Medical Women's Association. 44(6), 175-182. Tucker, J. B., St Margo, K. L. (1987). Profile of women family practice residents. Family Medicine, 19(4), 269-271. Uhlenberg, P., St Cooney, T. M. (1990). Male and female physicians: Family and career comparisons. Social Science and Medicine. 30(3), 373-378. United States Department of Health and Human Services. (1987). Characteristics of physicians: Michigan. December 31. 1985. (ODAM Report No. 123-87). Washington DC: U.S. Government Printing Office. Vaschak, M. R. (1970). Household help - the woman doctor's gordian knot. The Woman Physician. 25(6), 383-385. Vaschak, M. R. (Chair of the Committee). (1972). Special report of the medical education and research committee of the American Medical Women's Association. Journal of the American Medical Women's Association. 27.(6), 324-327. Vogel, M. (1985). Another doctor in the house. The New Physician. 34.(11), 19-23. 369 Wallace, H. M. (1980). Women in medicine. Journal of the American Medical Women's Association. 35(4), 201-211. Wallace, J. (1969). Part-time internships and residencies: Programs to be encouraged. Journal of the American Medical Women's Association. 24(7), 566-570. Wallis, L. A., Gilder, H . , St Thaler, H. (1981). Advancement of men and women in medical academia. Journal of the American Medical Association. 246(20), 2350-2353. Walsh, M. R. (1976). Feminism: A support system for women physicians. Journal of the American Medical Women's Association. 31(6), 247-250. Walsh, M. R. (1977a). Doctors wanted: No women need apply. Sexual barriers in the medical profession. 1835-1975. New Haven St London: Yale University Press. Walsh, M. R. (1977b). Women physicians and World War II. Journal of the American Medical Women's Association 12(5), 189-193. Walsh, M. R. (1990). Women in medicine since Flexner. New York State Journal of Medicine. 59(6), 302-308. Watson, J. H. (1977). A survey of two cohorts of women who graduated from a Canadian medical school. Canadian Medical Association Journal. 117(8). 892-894. Waxman, M. (1988). Women in medicine and the medical sciences: Problems, progress, and prospects. Connecticut Medicine. 52(12), 717-720. Weaver, J. L . , & Garrett, S. D. (1978). Sexism and racism in the American health care industry: A comparative analysis. International Journal of Health Services. 8(4), 677-703. Weinberg, E. (1969). Part-time residency training. Journal of the American Medical Association. 210(8), 1435-1437. Weisman, C. S. (1984). Gender composition of medical schools and specialty choices of graduates. Journal of Medical Education. 59(4), 347-349. Weisman, C. S., Levine, D. M . , Steinwachs, D. M . , & Chase, G. A. (1980). Male and female physician career patterns: Specialty choices and graduate training. Journal of Medical Education. 55(10), 813-825. Westling-Wikstrand, H . , Monk, M. A . , & Thomas, C. B. (1970). Some characteristics related to the career status of women physicians. Hopkins Medical Journal. 127(11), 273-286. 370 Weyrauch, K. F., Boiko, P. E., & Alvin, B. (1990). Patient sex role and preference for a male or female physician. Journal of Family Practice. 30^(5), 559-562. Wheeler, R . , Candib, L . , & Martin, M. (1990). Part-time doctors: Reduced working hours for primary care physicians. Journal of the American Medical Women's Association. 45(2), 47-54. White, K. L. (1964). General practice in the United States. Journal of Medical Education. 39(4), 333-345. Whiting, B. E., & Bickel, J. (1990). AAMC data report. Women on faculties of U.S. medical schools, 1978-1989. Academic Medicine. 65(4), 277-278. Wilkinson, C. J., & Linde, H. W. (1986). Status of women in academic anesthesiology. Anesthesiology. 64(4), 496-500. Williams, A. P., Domnick-Pierre, K., Vayda, E., Stevenson, H. M . , & Burke, M. (1990). Women in medicine: practice patterns and attitudes. Canadian Medical Association Journal. 143(3), 194-201. Williams, M. J. (1969). Underemployment of women physicians. Annals of Internal Medicine. 71(4), 862-863. Williams, P. A. (1971). Women in medicine: Some themes and variations. Journal of Medical Education. 46(2), 584-591. Williams, P. B. (1978) . Recent trends in the productivity of women and men physicians. Journal of Medical Education. 53(5), 420-422. Wilson, M. P. (1981). The status of women in medicine: Background data. Journal of the American Medical Women's Association. .36(2), 62-79. Wilson, M. P. (1987) . Making a difference - women, medicine, and the twenty-first century. The Yale Journal of Biology and Medicine. 60, 273-288. Witte, M. H., Arem, A. J., & Holguin, M. (1976). Women physicians in United States medical schools: A preliminary report. Journal of the American Medical Women's Association, 31(5), 211-213. Woodside, N. B. (1975). Women too often on the fringes. Annals of Internal Medicine. 82(3), 418-420. 371 Woodward, C. A. (1986). Career choices of men and women in medicine: A study of a cohort of recent medical graduates. Research in Medical Education 1986. Proceedings of the Twenty-Fifth Annual Conference. 109-114. Washington DC: Association of American Medical Colleges. Woodward, C. A., Cohen, M. L., & Ferrier, B. M. (1990). Career interruptions and hours practiced: Comparison between young men and women physicians. Canadian Journal of Public Health. 81(1), 16-20. Worobey, J. L . , & Mick, S. S. (1987). How sex and country of medical education affect FMG specialty choice. Journal of the American Medical Women's Association. 42(3), 85-89. Wright, K. W. (1967). Supply - the Americas. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22.(5), 340-342. Wunderman, L. E. (1980). Female physicians in the 1970s: Their changing roles in medicine. In G. L. Glandon & R. J. Shapiro (Eds.), Profile of medical practice 1980. Decade of the 1970s - the changing profile of medical practice (pp. 51-66). Monroe, Wisconsin: American Medical Association. Yamazaki, R. (1967). The demand for medical women. In (No editor) Optimal utilization of medical womanpower. Proceedings of the Xth congress of the Medical Women's International Association. Journal of the American Medical Women's Association. 22.(7), 485-487. Zimet, C. N., & Held, M. L. (1975). The development of views of specialties during four years of medical school. Journal of Medical Education. £50(2) , 157-166. Zimny, G. H., & Shelton, B. R. (1982). Sex differences in medical specialty preferences. Journal of Medical Education. 57(5), 403-405. General References 372 General References American Psychological Association. (1989). Publication manual of the American Psychological Association. Washington, DC: Author. Anderson, R. B. W. (1975). Choosing a medical specialty: A critique of literature in the light of "curious findings". Journal of Health and Social Behavior. .16(2) , 152-162. Chaff, S. L . , Haimbach, R . , Fenichel, C., & Woodside, N. B. (1977). Women in medicine: A bibliography of the literature on women physicians. Metuchen, N.J.: Scarecrow Press. Hadley, J. (1977). An empirical model of medical specialty choice. Inquiry. 14.(12), 384-401. Herman, M. W . , & Veloski, J. J. (1981). Premedical training, personality characteristics and performance in medical school. Journal of Medical Education. 15. 363-367. Jolly, P., Higgins, E. J . , Goodson, M. P. (1980). Trends in medical school faculty characteristics new faculty and continuing faculty 1968-1978. Washington DC: Association of American Medical Colleges. Jolly, H. P., & Larson, T. A. (1976). Participation of women and minorities on U.S. medical school faculties. Washington DC: Association of American Medical Colleges. Reynolds, R. A., & Abram, J. B. (Eds.). (1985). Socioeconomic characteristics of medical practice 1983. Chicago Illinois: American Medical Association. United States Department of Health and Human Services. (1981). The report of the Graduate Medical Education National Advisory Committee to the Secretary. Department of Health and Human Resources (sic). (DHHS Publication No. (HRA) 81-651). Washington DC: U.S. Government Printing Office. Whitbeck, C. (1982). Women and medicine: An introduction. The Journal of Medicine and Philosophy. 7(2), 119-133.