Are residents' work schedules related to their in-training examination scores?

To study the relationship between residents' pre-test available study time and subsequent examination performance.

We compiled rotation schedules and in-training examination performances for all residents enrolled in the two University of Texas Medical School-Houston obstetrics and gynecology residencies between 1993 and 1995. Services on which the residents had rotated were ranked by surveying all enrolled residents as to their perceived available study time while on those services. A mean available study time rating was computed for each service, which was used to describe each resident's schedule. Pearson correlations between available preparation-time preceding the in-training examination and examination performance were calculated. "Tough" and "easy" rotation subgroups were defined; two-tailed t tests were performed to compare the in-training examination performance of these subgroups.

No significant correlation was found between available study time and subsequent examination scores beyond that which could be accounted for by year of training. Scores of residents in the "tough" subgroup were not significantly different from those in the "easy" subgroup on any examination performance measure. For correlations at the one-tailed .05 level, statistical power to detect relationships predicting 16% of variance ranges from approximately 85% for samples of 42 to 68% for samples of 27.

Residents' in-training examination performance was not significantly affected by their service obligations. Altering resident monthly schedules is not likely to improve in-training examination performance.