Using resident performance on Step 2 to predict surgical residency success

To determine whether performance on Step 2 CK can be used to predict adverse academic outcomes in surgical residency.

Surgical resident data from a single institution was used. Step 2 scores of each resident were normalized against the average Step 2 CK score for each cohort's application cycle (ie, resident who scored 246 on a year with a national average of 246 would have a normed score of 1; resident with a score >246 would have a normed score >1 etc). All adverse events were classified, and logistic regression analysis was used to determine the effects of the Step 2 scores. Adverse academic outcomes were defined.

Ninety-two surgical residents were evaluated with 7,182 faculty evaluations reviewed. Lower Step 2 CK scores demonstrated an increased likelihood of lower American Board of Surgery In-Training Examination performance <25% (P = .011). The remaining adverse events, including passing the Qualifying Examination (P = .998), passing the Certifying Examination (P = .778), early withdrawal from the surgical residency program (P = .565), failure on mock orals during postgraduate years 4 and 5 (P = .731; P = .600), as well as selection of administrative chiefs (P = .565), all demonstrated no correlation.

Emphasis has historically been placed on board examinations like Step 2 CK to predict overall success in residency and is used as an important metric during the interview process. However, these findings demonstrate that although lower scores on Step 2 CK correlate with a lower quartile on the American Board of Surgery In-Training Examination during general surgery residency, the examination cannot predict success in all areas of surgical residency.