Primary care and non-primary care physicians: a longitudinal study of their similarities, differences, and correlates before, during, and after medical school.

PURPOSE--To investigate similarities and differences between physicians in primary care and non-primary care specialties on performance measures prior to, during, and after medical school, and on demographic characteristics, professional plans and preferences in medical school, professional activities, career satisfaction, perceived problems and research activities, and to predict primary--non-primary care career choices from information obtained in medical school. A questionnaire was mailed to 1,076 physicians who graduated from Jefferson Medical College between 1982 and 1986. Of those who responded (62%), 232 were primary care and 406 were non-primary care physicians (29 physicians in mixed specialties were excluded). Data from the questionnaire concerning professional activities, satisfaction, problems, and research productivities were merged with the college's longitudinal study database. RESULTS--Comparisons of primary care--non-primary care physicians indicated no significant difference between them on performance measures before, during, and after medical school, with the exception that non-primary care physicians had higher scores on quantitative tests before medical school, and primary care physicians scored higher on a licensing examination of general clinical skills and patient management taken during residency training. Also, compared with non-primary care physicians, those in primary care were less likely to be employed full-time, were less likely to locate in metropolitan areas, had a lower rate of academic appointment, and had a higher rate of board certification. Other results showed differences between the groups in terms of age at entrance to medical school, proportion of women, estimates during medical school of anticipated income, career plans during medical school, satisfaction with career and income, and research and scientific activities. A logistic regression model could predict primary care--non-primary care status from specialty interest, professional plans and interests expressed in medical school.