B D Skinner1, W P Newton. 1. Department of Family Medicine, University of North Carolina, Chapel Hill, USA. theroo@med.unc.edu
Abstract
BACKGROUND AND OBJECTIVES: Knowing which factors predict National Resident Matching Program (Match) success is critical for medical students, student advisors, and residency faculty, but published reports have focused on a limited number of program characteristics or a small number of years. METHODS: We obtained information about region, affiliation, hospital size, population served, intern salary, and intern call schedule from the Directory of Family Practice Residency Programs for all US family practice programs from 1984-1998 and correlated them with initial Match results, using both univariate and multivariate techniques. RESULTS: Over 15 years, there have been substantial increases in programs and positions offered, salaries, proportion of larger programs, and reduced intern call frequency. Western region, larger program size, and community hospital affiliation were associated with a higher initial Match rate, but only 7%-19% of the variance was explained. In the late 1980s, salary had an effect, as did call in the worst Match years; in recent years, suburban programs have filled better and rural programs less well. Hospital size had no relationship to fill rates. CONCLUSIONS: Program characteristics have a real but modest impact on initial Match rates; the significance may increase in times of decreased interest in family medicine, such as the late 1980s and early 1990s.
BACKGROUND AND OBJECTIVES: Knowing which factors predict National Resident Matching Program (Match) success is critical for medical students, student advisors, and residency faculty, but published reports have focused on a limited number of program characteristics or a small number of years. METHODS: We obtained information about region, affiliation, hospital size, population served, intern salary, and intern call schedule from the Directory of Family Practice Residency Programs for all US family practice programs from 1984-1998 and correlated them with initial Match results, using both univariate and multivariate techniques. RESULTS: Over 15 years, there have been substantial increases in programs and positions offered, salaries, proportion of larger programs, and reduced intern call frequency. Western region, larger program size, and community hospital affiliation were associated with a higher initial Match rate, but only 7%-19% of the variance was explained. In the late 1980s, salary had an effect, as did call in the worst Match years; in recent years, suburban programs have filled better and rural programs less well. Hospital size had no relationship to fill rates. CONCLUSIONS: Program characteristics have a real but modest impact on initial Match rates; the significance may increase in times of decreased interest in family medicine, such as the late 1980s and early 1990s.