Electrocardiographic interpretation skills of cardiology residents: are they competent?

Achieving competency at electrocardiogram (ECG) interpretation among cardiology subspecialty residents has traditionally focused on interpreting a target number of ECGs during training. However, there is little evidence to support this approach. Further, there are no data documenting the competency of ECG interpretation skills among cardiology residents, who become de facto the gold standard in their practice communities. We tested 29 Cardiology residents from all 3 years in a large training program using a set of 20 ECGs collected from a community cardiology practice over a 1-month period. Residents interpreted half of the ECGs using a standard analytic framework, and half using their own approach. Residents were scored on the number of correct and incorrect diagnoses listed. Overall diagnostic accuracy was 58%. Of 6 potentially life-threatening diagnoses, residents missed 36% (123 of 348) including hyperkalemia (81%), long QT (52%), complete heart block (35%), and ventricular tachycardia (19%). Residents provided additional inappropriate diagnoses on 238 ECGs (41%). Diagnostic accuracy was similar between ECGs interpreted using an analytic framework vs ECGs interpreted without an analytic framework (59% vs 58%; F(1,1333) = 0.26; P = 0.61). Cardiology resident proficiency at ECG interpretation is suboptimal. Despite the use of an analytic framework, there remain significant deficiencies in ECG interpretation among Cardiology residents. A more systematic method of addressing these important learning gaps is urgently needed.