Self-report of high cholesterol: determinants of validity in U.S. adults.

摘要: Abstract Background: Hypercholesterolemia is a major cardiovascular risk factor, and cholesterol awareness important in both clinical practice public health. We evaluated the validity of self-reported hypercholesterolemia identified determinants validity. Methods: The study design was cross-sectional survey, from 1988 to 1994, adult participants (N=8236) Third National Health Nutrition Examination Survey for whom self-report serum measurement were available. Sensitivity, specificity, positive predictive value (PPV), negative (NPV) calculated using total ≥5.17 mmol/L (200 mg/dL) and/or taking cholesterol-lowering medication as criterion standard. Results: Overall test characteristics sensitivity, 51%; 89%; PPV, 87%; NPV, 55%. Sensitivity higher among older subjects non-Hispanic whites, specificity with >12 years education, PPV subjects, NPV younger those education. Using thresholds define led lower NPV. Sociodemographic anthropometric predictors by logistic regression. Conclusions: Due low should be used caution, during patient encounter surveillance trends absence measured levels. Specificity consistently much than sensitivity. high may use certain situations. Such validation studies form foundation future research based on self-report.