Using a web-based application to enhance resident training and improve performance on-call

It is common practice in academic hospitals for radiology residents to provide preliminary interpretations for radiologic examinations performed in the emergency department (ED) during off-hours. In this study, we used a software program called Minerva to identify and track discrepancies between resident and faculty interpretation of ED studies. The objective was to determine if missed case conferences could reduce the number of resident discrepancies related to the types of cases reviewed.

We used Minerva to identify and grade faculty-modified resident preliminary reports as minor or major discrepancies depending on whether the discrepancy had the potential to affect patient management or outcome. Minor and major discrepancy rates were calculated for all residents to evaluate call performance, establish benchmarks, and develop interventions to reduce the number of discrepant cases.

The total discrepancy rate for all residents (n = 22) was 2.6% with a standard deviation (SD) of 0.7%. The average major discrepancy rate for all residents was 1.1% with a SD of 0.4%. Trend analysis of missed cases was used to generate topic-specific resident missed case conferences on acromioclavicular joint separation injuries, elbow joint effusions, and osteochondral fractures, which resulted in an overall 64% decrease in the number of missed cases related to these injuries.

The systematic evaluation of resident discrepancies using a simple software application provides a competency-based metric to assess call performance, establish benchmarks, and develop missed case conferences. This process is expected to result in further reduction in resident discrepancy rates and missed cases.