Don't Skip a Beat: Deliberate Practice and Mastery Learning of Pediatric Electrocardiograms.

Abstract

To develop an online pediatric electrocardiogram (ECG) educational intervention, to examine pediatricians' diagnostic skill development as they progressed to achieve a performance-based standard, and to determine the frequency of pediatric ECG findings at highest risk for diagnostic error. This multicenter, prospective cohort study included a convenience sample of pediatricians. There were 400 cases in the intervention, and for each case, participants first determined whether an actionable abnormality was present or absent. If present, participants categorized abnormalities as rate/rhythm, anatomical/technical, or Q-wave/repolarization, and selected the most appropriate specific diagnosis from a drop-down list of options. Immediate feedback was provided after each case, and practice continued until a minimal passing standard was achieved. A total of 345 pediatricians performed 46 649 pediatric ECG case interpretations. Initial accuracy was 82.1% in identifying ECG with actionable findings, 70.9% for correct categorization of actionable findings, and 45.2% for selecting most actionable specific diagnosis. There were learning gains for each of these diagnostic tasks: +14.6% (95% CI 13.1, 15.8), +15.0% (95% CI 13.2, 16.8), and +19.8% (95% CI 17.5, 22.1), respectively. Furthermore, 65% achieved the minimal passing standard in a median of 325 cases (IQR 198, 496) or 6.5 hours (IQR 3.4, 10.0) of practice. Among the 46 649 case interpretations, 7675 (16.5%) were incorrect interpretations. ECG findings consistent with ventricular hypertrophy, Brugada sign, prolonged QTc, and ischemia/pericarditis were among the most challenging diagnoses. Structured practice with feedback can feasibly improve pediatrician ECG interpretation skills and derive data to identify ECG findings prone to diagnostic error.