Receiver Operating Characteristic Curves To Evaluate Oxidized Low Density Lipoprotein For Coronary Heart Disease Screening

Receiver Operating Characteristic curves to evaluate Oxidized Low Density Lipoprotein for Coronary Heart Disease screeningObjective Evaluate the clinical value and discriminatory power of oxidized low density lipoprotein (OxLDL) and other 4 biomarkers to diagnose Coronary Heart Disease and determine the Optimal Operating Point(OOP) using the Receiver Operating Characteristic(ROC) curves in different statistical modeling method. To increase the diagnostic accuracy of the screening test and get rid of the influence of covariates, some new methods were used to plot the ROC curve.Method Empirical curve, the binormal model and proportional odds model were applied and compared to plot the ROC curve, the proportional odds model and predictors based on generalized linear models were presented at the same time to control the influence of covariates.Result The different methods show conformably that the OxLDL, HDL and TG are suitable for Coronary Heart Disease screening, the OOP of the diagnostic kit was determined and interpreted with higher sensitivity and specificity. The theory, ideas and modeling methods are discussed and compared with the example at the same time.Conclusion When assessing the performance and determining the OOP of a screening test using ROC curve, the availability and convenience, conditions and covariates control should be concerned to select a more aptitude method.