Clinicopathological study of the heart and coronary arteries of autopsied cases from the community of Hisayama during a 10-year period. Part V. Comparison of autopsy findings with electrocardiograms--Q.QS items of the Minnesota Code.

In a longitudinal study of a general population in Hisayama, Japan, 339 persons aged 40 years or over at death were autopsied during the period November 1, 1961-October 31, 1971. In 308 of these people, electrocardiograms taken at the periodic examinations were available, and Q.QS items of the Minnesota Code were recorded in 49 persons. The sensitivity of item 1-1 to autopsy-proven old myocardial infarction was 0.32 and the specificity was 0.95. Using the estimated prevalence of old myocardial infarction in this community--77 per 100,000--the proportion of persons without old myocardial infarction among those with item 1-1 in the general population (PF+) was 0.9954 and the proportion of persons with old myocardial infarction among those without item 1-1 (PF-) was 0.0005. The sensitivity of items 1-1 and 1-2 to old myocardial infarction was 0.43 and the specificity was 0.94. PF+ of items 1-1 and 1-2 was 0.9949 and PF- was 0.0005. These large values of PF+ mean that more than 99 per cent of persons with these items are probably false positives if these items are used as a screening test for this disease in the general population of this community, i.e., that these items cannot be defined as a definite myocardial infarction.