Preferred practice location at medical school commencement strongly determines graduates' rural preferences and work locations.

Objective To identify factors influencing whether Australian medical graduates prefer to, or actually, work rurally. Design Secondary analysis of longitudinal data from Medical Schools Outcomes Database ( MSOD) using univariate and multivariate logistic regression. Setting Twenty Australian medical schools. Participants Australian or New Zealand citizens and Australian permanent residents who completed MSOD questionnaires between 2006 and 2013. Main outcome measures Preferred and actual work locations 1 ( PGY1) and 3 ( PGY3) years postgraduation. Results Of 20 784 participants, 4028 completed a PGY1 or PGY3 questionnaire. Self-reported preference for rural practice location at medical school commencement was the most consistent independent predictor of whether a graduate would have a rural location preference at PGY1 (odds ratio ( OR) 6.07, 95% confidence interval ( CI) 4.91-7.51) and PGY3 ( OR 7.95, 95% CI 4.93-12.84), and work rurally during PGY1 ( OR 1.38, 95% CI 1.01-1.88) and PGY3 ( OR 1.86, 95% CI 1.30-2.64). The effect of preferred practice location at medical school commencement is independent of, and enhances the effect of, rural background. Graduates of graduate-entry programs or with dependent children were less likely to have worked rurally during PGY1 and PGY3 respectively. Conclusion The most consistent factor associated with rural preferences and work location was students' preferred location of practice at medical school commencement; this association is independent of, and enhances the effect of, rural background. Better understanding of what determines rural preference at medical school commencement and its influence on rural workplace outcomes beyond PGY3 is required to inform Australian medical school selection policies and rural health curricula. [ABSTRACT FROM AUTHOR]