The CVDS experience has been reviewed with regard to several methodologic problems in the study of CVD. In light of this experience the following conclusions seem warranted: 1. Community surveillance studies of CVD should limit their inquiries to hard endpoints, i.e. those that are fatal or hospitalized. Community-physician reporting should be used only as an adjunctive surveillance tool because of its low additional yield when used in combination with hospital and death certificate surveillance. Physician reporting can be utilized along with computer checks of the records of insurance carriers and state hospital associations as internal quality control devices to detect lapses in the primary surveillance techniques. 2. The perusal of death certificates is an indispensable surveillance tool for community studies of CVD and is quite reliable for case finding. 3. Spouses or other relatives of patients can be relied upon to give fairly accurate information concerning chest pain prior to a CHD event, but are apt to disagree with patients' reports about other symptoms. 4. A longitudinal cohort study may be able to make diagnoses of MI and CI more frequently than a study without long-term follow-up information at its disposal; however, fatal events are diagnosed equally by the two types of study.
ASJC Scopus subject areas