Objective Admissions Data and In-State Practice: What Can We Really Predict?

Despite the widespread adoption of the holistic review initiative in medical school admissions, admissions committees continue to struggle with identifying applicant attributes that predict future practice location. This challenge has significant practical implications, particularly for institutions committed to addressing regional health care shortages.

We analyzed eight cohorts of applicants who matriculated between 2006 and 2013 and then completed residency. We used univariate and multiple regression models to identify applicant objective attributes linked to later in-state practice. Objective data available in applications included high school location, legal state of residence, Medical College Admissions Test (MCAT) scores, grade point average, college details, graduate school, race/ethnicity, and gender; these were compared by posttraining practice location.

Of 509 graduates in practice, 212 (41.7%) were in-state. In-state graduates were older (P<.001), predominately underrepresented in medicine (42% vs 33%, P<.01), graduated from urban high schools (28% vs 18%, P<.05), attended in-state colleges (66% vs 55%, P<.05), and had lower MCAT scores (56th vs 67th percentile, P<.001). Regression analysis revealed that only age, urban high school, and MCAT score significantly contributed to the model (R 2=0.064, P=.004).

This study shows the limited value of applicant objective data to admissions committees in predicting future in-state practice. The findings emphasize the importance of a holistic admissions process and the value of considering subjective applicant data in efforts to predict future practice location. Additional research may help identify which subjective aspects of an applicant can assist admissions committees in this effort.