On-call transthoracic echocardiographic interpretation by first year cardiology fellows: comparison with attending cardiologists

BACKGROUND: Transthoracic echocardiograms (TTE) performed and interpreted by cardiology fellows during off-duty hours are critical to patient care, however limited data exist on their interpretive accuracy. Our aims were to determine the discordance rate between TTEs performed and interpreted by cardiology fellows and National Board of Echocardiography certified attending cardiologists and to identify factors associated with discordance. METHODS: Consecutive on-call TTEs acquired and interpreted by 1st year cardiology fellows over 4.6 years at an academic center were prospectively evaluated by attending cardiologists. Fellow interpretations were classified as concordant or discordant with the attending interpretation. We assessed the association of patient, imaging and fellow characteristics with discordance. RESULTS: A total of 777 TTE interpretations (730 patients) were performed/interpreted by 40 first year fellows and overread by 13 attendings. The most common indications were assessment of left ventricular function (40.9%) and pericardial effusion (37.3%). There was a major or minor discordance in 4.1 and 17.4% of studies, respectively with 42.1% of disagreements occurring in assessment of left ventricular size and function. The indication to assess left ventricular function [OR 2.19, 95% CI (1.32, 3.62), P = 0.002 vs. pericardial effusion] and greater duration of echocardiographic image acquisition (OR 1.02, 95% CI 1.01, 1.03, P = 0.004) were independently associated with overall discordance. CONCLUSIONS: In this large prospective study we found that attending cardiologists disagreed with 1 in 5 fellow TTE interpretations. Standardized tools for evaluation of echocardiograms performed by fellows are needed to ensure quality of training and patient safety.

Abstract licence: CC BY-NC-ND

17 more references, whose wording the paper’s licence keeps private.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.