The repeatability at interview of symptoms of angina and possible infarction

Abstract

Reinterviews confined to the ‘Cardiac pain questionnaire’ were conducted on 237 middle-aged men in Oslo, interviewed about 10 months previously in connection with a Finnish-Norwegian lung cancer study. The reinterview sample was stratisfied on the basis of equal numbers of chest pain positives, questionables and negatives. The reinterview findings were applied to the symptom distribution of the interviewed population in Oslo.

Both for angina and possible infarction high rates of conversion to negatives at the reinterview were observed. For angina this rate was lower for those answering Yes to the question on pain when walking on the level. The prevalence of angina was near 4.5 per 100 at both interviews, while the prevalence of possible infarction decreased from 6.1 to 4.4 at reinterview. Repeatability, expressed as per cent of positives both times to those positive one or both times, was 43 for angina and 52 for possible infarction. For men with angina without infarction the repeatability was lower (20).

A mortality follow-up is reported, but the figures were too small for a validation of the reinterview results. The components of variability in a reinterview process are discussed.