Metoder för prediktion av kardiovaskulär sjukdom med njurfunktionen
CalibrationDiscriminationCardiovascular riskCox-regressionStatisticsOperations researchProgrammingActuarial mathematicsStatistikOperationsanalysProgrammeringAktuariematematikMathematics and Statistics
This study examines if the prediction of cardiovascular disease in hypertensive patients can be improved upon when renal function and microalbuminuria are added to the classical risk factors The predictive capability of a model is measured by discrimination, calibration, reclassification and Harrell's C.The results are ambigious. In most cases, microalbuminuria should be included in the model, but the results regarding the other measures of renal function are varied. Therefore, the selection of risk factors to include in the model depends on which measure of prediction one prioritizes.