The Detection of Heart Disease in Children

Abstract

In a 1 in 6 stratified random sample of Chicago area high school students, urban and suburban, public and non-public, 49,753 of a potential 55,338 students (90 per cent) underwent the heart-sounds tape-recording procedure for the detection of heart disease in children. Based on an extensive educational and organizational effort among appropriate community, professional, voluntary, and public health agencies, the procedure was widely accepted and a high-response rate (90 per cent) obtained. The recorded prevalence rates were 1.4, 0.7, and 2.1 per 1,000 for congenital, rheumatic, and all heart disease, respectively, with detection of 105 students with heart disease, 46 of them (0.9 per 1,000) previously unknown. No significant differences in recorded prevalence of congenital or rheumatic heart disease were observed in strata of the Chicago area high school students, irrespective of age, sex, race, socioeconomic, geographic, urban-suburban, or public-nonpublic school classification. Correspondingly, students with previously undetected heart disease were found among all strata and sections of the community. With appropriate corrections-for validity rate of the procedure, for findings among non-respondents and among students absent at recall, for number of children at home or in hospital or special schools because of organic heart disease-it was possible to make extrapolated estimates of the prevalence rates of organic heart disease among the total Chicago area high school population (about 297,000 students). The estimated prevalence rates were 3.3, 2.8, and 6.1 per 1,000 for rheumatic, congenital, and all heart disease, respectively. Thus, for high school students, in contrast to elementary school students, prevalence rates of rheumatic heart disease were slightly higher than for congenital heart disease. When a history of cardiovascular disease or a gymnasium excuse for cardiovascular disease was present in the school records, students were seen at recall by the pediatric cardiologists, irrespective of the findings reported by the tape readers. As a result, iatrogenic heart disease was diagnosed in 106 students (2.1 per 1,000), and they were "delabeled." This-together with the detection of previously unrecognized heart disease-is a major built-in function and service of the heart-sounds tape-recording procedure. The detection of previously unrecognized heart disease and the correction of misdiagnosed heart disease ("delabeling" of iatrogenic heart disease) were accomplished with a recall rate of 3 per cent, thus reducing to a minimum the number of students seen by pediatric cardiologists. Continuing assessment of validity of the tape-recording procedure, particularly evaluation of the false-negative rate of the physician readers of the tape recordings, reaffirmed the high validity of this procedure (76 per cent or greater). Validity rate was shown to depend upon the precise nature of the material inserted to accomplish the validity assessment. A high level of validity for this mass-detection procedure was again shown to be achievable only with two trained physician readers evaluating the tape recordings. The findings further demonstrate the feasibility, utility, and validity of this method for the large-scale detection of heart disease among elementary and high school students, and for epidemiologic research on congenital and rheumatic heart disease among school populations.