Competency-Based Medical Education: Do the Cardiac Imaging Training Guidelines Have it Right?

Abstract

Objectives: This study sought to: 1) determine if the current training volume guidelines are reasonable for attaining competence for interpreting myocardial perfusion imaging (MPI); and if not, 2) identify potential thresholds for training volumes and competence.

Background: There is a growing desire to adopt competency-based medical education (CBME). As such, the implementation of CBME will require new and novel methods of defining, measuring, and assessing clinical competence. The potential use of CBME in cardiac imaging has not been well studied.

Methods: Consecutive MPI studies were interpreted independently by trainees, and expert readers reviewed the same studies. Studies were quantified using summed scores and % left ventricular (LV) ischemia and the kappa agreement between trainee and expert were measured every 50 cases. Agreement for all MPI and abnormal MPI cases was calculated.

Results: A total of 24 trainees interpreted 9,668 MPI studies over 37 months. Agreement between trainees and expert readers increased with MPI case volumes but at different rates. The threshold for competence was set at 2 SDs below expert interobserver agreement. The average trainee surpassed this threshold for both summed stress score and %LV ischemia after 800 studies and after 400 abnormal MPI studies. Trainees learned at different rates and surpassed the competence threshold after different case volumes.

Conclusions: The use of CBME within nuclear cardiology appears to be feasible. Our results suggest that current guidelines may be insufficient to ensure competence and would support the need to increase the MPI case volumes. The use of CBME principles would suggest that trainees may achieve competence at different rates and our results suggest a shift in focus from volume-based learning toward target agreement thresholds.

Keywords: cardiac imaging training guidelines; competency-based medical education; myocardial perfusion imaging.