Cost-effectiveness of echocardiography for evaluation of children with murmurs

Abstract

Echocardiography is widely accepted as an accurate diagnostic test to evaluate heart murmurs in children, however its costs, and the ubiquity of murmurs in children, discourage its universal application. The purpose of this article is to identify some clinical circumstances in which the cost of echocardiography is justified for the evaluation of heart murmurs in infants and children. Eight common clinical problems were selected in which a heart murmur is present and a diagnosis is called for. Effectiveness of echocardiography and less costly clinical diagnostic methods in these settings were compared. In some circumstances, echocardiography is worth the cost, because clinical evaluation is unacceptably insensitive to important disease (the premature infant with a murmur which might represent a patent ductus arteriosus, the infant with a dysmorphic syndrome and a murmur). In others, the expert clinical examination is highly accurate (the asymptomatic child with a heart murmur) and is preferred over the echocardiogram as the initial diagnostic approach on the grounds of cost. When the expert clinical examination suggests minor structural heart disease, a continuum of echocardiographic cost-effectiveness relative to the expert clinical examination exists between these extremes depending on the working diagnosis. A threshold has not yet been defined at any point on this continuum above which the public will demand the greater diagnostic detail available echocardiographically, and below which the public will refuse to accept its greater cost. Quantitative formal cost-effectiveness analysis of echocardiography for evaluation of heart murmur in infants and children is not yet feasible because the benefits of echocardiography are indirect, dependent upon the as yet unmeasured benefits of correct management of congenital heart defects. To go beyond simple comparison of efficacy of echocardiography with less costly methods, further work is required in outcomes research in congenital heart disease.