An evaluation of follow-up methods in the Framingham Heart Study.

Publisher Website
Google Scholar
Abstract
The prospective Framingham Heart Study employs bienniel Study Clinic examinations, hospital and private physicians'' records and death information to diagnose coronary heart disease (CHD). In 10 years of follow-up, the bienniel examinations provided diagnostic information for 3/4 of all 327 cases and 1/2 would have been missed without these examinations. Since, at most, 1/6 of all cases were lost when all follow-up examinations except the last were ignored, it appeared that less frequent examinations would have resulted in very little under-estimation of disease incidence. Hospital and death information were needed for diagnosis of 1/4 of the cases to complete the clinical picture of CHD in the population. Each diagnostic source gave its own distorted view of the distribution of clinical manifestations of CHD. The association of CHD with age, blood pressure, serum cholesterol, relative weight and cigarette smoking, using the cases from each source, closely resembled that found when all cases were studied. Male predominance was exaggerated in the hospital cases and deaths. The generally good agreement noted suggests that 1 source of diagnostic information can be used for a prospective investigation of the associations between attributes and CHD even thought a distorted view of the clinical spectrum will appear.