摘要:SummaryBackground Echocardiographic screening for rheumatic heart disease (RHD) can identify individuals with subclinical disease who could benefit from antibiotic prophylaxis. However, most settings have inadequate resources to implement conventional echocardiography and require a feasible, accurate screening method. We aimed to investigate the accuracy of screening by non-expert operators using focused cardiac ultrasound (FoCUS). Methods In this prospective study of diagnostic accuracy, we recruited schoolchildren aged 5 to 15 years in Fiji to undergo two blinded tests. The index test was a FoCUS assessment of mitral and aortic regurgitation, performed by nurses after an 8-week training programme. The reference standard was the diagnosis of {RHD} by a paediatric cardiologist, based on a standard echocardiogram performed by a skilled echocardiographer. The primary outcome was the accuracy of the index test with use of the most sensitive criteria (any regurgitation). Findings We included 2004 children in the study. The index tests were done between September, 2012, and September, 2013, by seven nurses in eight schools in Fiji. The diagnostic accuracy of the screening test (area under receiver operator characteristic curve) was 0·89 (95% {CI} 0·83–0·94). When the primary cut-off point (any regurgitation) was used for analysis, sensitivity was 84·2% (72·1–92·5) and specificity was 85·6% (83·9–87·1). The sensitivity of individual nurses ranged from 66·7% to 100% and specificity 74·0% to 93·7%. Interpretation Screening by briefly trained nurses using FoCUS was accurate for the diagnosis of RHD. Refinements to training and screening test methods should be studied in a range of settings, and in parallel with investigations of the long-term clinical and cost-effectiveness of screening for RHD. Funding Cure Kids, New Zealand; the Fiji Water Foundation provided funding for portable ultrasound equipment; see acknowledgments for further details of funders.