Vascular and cardiac functional and structural screening to identify risk of future morbid events: preliminary...

Risk factors have served to identify patients in need of antihypertensive and lipid-lowering therapy. Because of their limited sensitivity and specificity, we developed a screening program using noninvasive testing and a scoring system aimed at detecting functional and structural cardiovascular abnormalities in asymptomatic individuals. Ten cardiovascular tests were performed in 1 hour by a single technologist. Tests were scored as normal (0), borderline abnormal (1), or abnormal (2). Total disease score (DS) could range from 0 (all tests normal) to 20 (all tests abnormal). Scores of 0-2 were classified as normal, 3-5 as early disease, and 6+ as advanced disease. Morbid events during follow-up of 6 months to 8 years were determined from mailed questionnaires. Framingham risk scores (FRS) were calculated using published algorithms. Thirty-five morbid events (1 of 169 in the "normal" group, 8 of 214 in the "early disease" group, and 26 of 230 in the "advanced disease" group) occurred during the follow-up period among the 613 individuals who completed the questionnaire. Risk for morbid events was highly significantly different between the Kaplan-Meier curves based on disease detection (log rank 21.75, P ≤ .0001). FRS were significantly different but less discriminating, with five morbid events in the 227 subjects with FRS <10, eight in 162 with FRS 10-13, and 22 of 227 with FRS >13 (log rank 9.80, P = .0074). The area under receiver operating characteristic curve for DS (0.74) surpassed that of FRS (0.66) and was not improved when both were included in the model. Neither blood pressure levels nor low-density lipoprotein cholesterol levels provided adequate discrimination. Identifying early disease in asymptomatic individuals provides a better guide to the need for preventive therapy than traditional risk factor assessment.