Echocardiography is the diagnostic test of choice for latent rheumatic heart disease. The utility of echocardiography for large-scale screening is limited by high cost, complex diagnostic protocols, and time to acquire multiple images. We evaluated the performance of a brief hand-held echocardiography protocol and computer-assisted auscultation in detecting latent rheumatic heart disease with or without pathological murmur.

A total of 27 asymptomatic patients with latent rheumatic heart disease based on the World Heart Federation criteria and 66 healthy controls were examined by standard cardiac auscultation to detect pathological murmur. Hand-held echocardiography using a focussed protocol that utilises one view – that is, the parasternal long-axis view – and one measurement – that is, mitral regurgitant jet – and a computer-assisted auscultation utilising an automated decision tool were performed on all patients.

The sensitivity and specificity of computer-assisted auscultation in latent rheumatic heart disease were 4% (95% CI 1.0–20.4%) and 93.7% (95% CI 84.5–98.3%), respectively. The sensitivity and specificity of the focussed hand-held echocardiography protocol for definite rheumatic heart disease were 92.3% (95% CI 63.9–99.8%) and 100%, respectively. The test reliability of hand-held echocardiography was 98.7% for definite and 94.7% for borderline disease, and the adjusted diagnostic odds ratios were 1041 and 263.9 for definite and borderline disease, respectively.

Computer-assisted auscultation has extremely low sensitivity but high specificity for pathological murmur in latent rheumatic heart disease. Focussed hand-held echocardiography has fair sensitivity but high specificity and diagnostic utility for definite or borderline rheumatic heart disease in asymptomatic patients.