A MULTIVARIATE MODEL FOR SPECIALTY PREFERENCE BY MEDICAL-STUDENTS

Background. Numerous studies have demonstrated differences among stude nts regarding their preference of a medical specialty. The goal of the present research was to develop a model for the selection of a primar y care specialty (ie, family practice, general internal medicine, medi cine/pediatrics, and general pediatrics). Methods. A self-administered questionnaire was mailed to 822 first-year through fourth-year medica l students at the University of Michigan Medical School in Ann Arbor. Students listed their first preference for medical specialty, anticipa ted income and work hours, and the influence of attitudinal and social factors on their preference. A total of 645 (78.5%) students responde d. Average age was 25; 58% were male, and 77% resided in Michigan. Res ults. Overall, 34.3% of the medical students who responded to the ques tionnaire expressed a preference for a surgical specialty; 27.3%, prim ary care; 19.9%, a hospital-based practice; and 18.5%, nonprimary care and non-hospital based practice. A multiple logistic regression model developed on preference for a primary care specialty achieved a class ification accuracy of 82%. The most important factors influencing spec ialty preference were sex, expected income, attitudes about general me dicine issues, attitudes about surgery, and the influence of other peo ple. Conclusions. No single factor dominates a student's preference fo r primary care. Students preferring primary care were most strongly in fluenced by their perceptions of practice variations. Students preferr ing nonprimary care specialties were more interested in income, presti ge, and hospital-based practice. Medical school faculty had no signifi cant impact on the preferences of either group of students.