Predicting success: What medical student measures predict resident performance in neurology?

Many medical school metrics are used by residency programs to differentiate residency applicants. The importance of each metric in the field of neurology is unclear.

This is a single-site retrospective evaluation of characteristics that predict resident quality. Several measures from all 57 adult neurology residents over 8 years were obtained including Step I scores, college and medical school rankings, in-service training examination scores, advanced degrees, and number of publications during residency. Two program directors, blinded to these data and each other's ratings, rated the quality of all residents at the end of the residency. The data were then anonymized for all analyses.

There was no significant relationship between Step I scores and resident quality, though Step I scores correlated significantly with in-service training examination scores. Medical students with PhDs did not perform differently in terms of resident quality, number of publications in residency, or in-service training examination scores. Resident quality was correlated with the ranking of each applicant's undergraduate college, but not the ranking of their medical school.

While Step I is used by many residency programs in ranking potential residents, it does not correlate with overall resident quality, although Step I scores may predict success on future standardized medical examinations. Students with PhDs do not differ from other residents across several metrics. Applicants from highly selective colleges, though not highly selective medical schools, had significantly higher quality ratings. Further research is needed to determine characteristics of medical students that predict performance during neurology residency.