Experienced Clinicians Improve Detection of Third and Fourth Heart Sounds by Viewing Acoustic Cardiography

Abstract

Clinical assessment of diastolic heart sounds is challenging. We sought to examine whether visual inspection of acoustic cardiographic tracings augments the accuracy of medical students' and physicians' detection of third and fourth heart sounds (S(3), S(4)) compared to auscultation alone. A total of 90 adults referred for left heart catheterization underwent digital precordial heart sound recordings by computerized acoustic cardiography. Two blinded, experienced readers using a consensus method determined the presence of the S(3)/S(4) on each file. There were 35 subjects from the following 5 groups participating in this study from 1 teaching institution: first-year medical students (n = 5), fourth-year medical students (n = 5), interns (n = 5), medicine residents (n = 5), cardiology fellows (n = 5), and attendings (n = 10). Using a computer module, each subject listened to the heart sounds alone and documented whether an S(3)/S(4) was present. Next, subjects listened to each recording in random order while viewing phonocardiographic tracings, and recorded S(3)/S(4) presence. An S(3) was present in 21 patients (23%) and an S(4) in 31 patients (34%) by consensus overread in 90 recordings. Baseline accuracy for auscultation of S(3)/S(4) did not change with level of experience. While viewing the acoustic cardiogram, first-year medical students had minimal improvement in S(3) (2%) and S(4) (11%) accuracy. More experienced subjects improved S(3) accuracy by 8% to 18% and S(4) by 15% to 32% (P < .05). Accuracy was superior for S(3) compared to S(4) in all ausculatory groups. While listening to heart sound recordings, viewing acoustic cardiography increased subjects' accuracy in detecting diastolic heart sounds, particularly among more experienced subjects. There was greater improvement for S(4) compared to S(3) detection.