Continuous diagnostic tests are often used to discriminate between diseased and healthy populations. Therefore, it is important to look for measures to assess the effectiveness of diagnostic markers in discriminating between these two states. The overlap coefficient (OVL), which is defined as the area of intersection of the graphs of two probability density functions, is considered as a summary measure of diagnostic accuracy. Often additional information is available on some of the covariates which are known to influence the accuracy of such measures. Consequently, in this work we incorporate covariate information using linear regression models with normally distributed errors, enabling the construction of the OVL estimator based on the ordinary least squares estimates of the regression coefficients. We further develop a method for constructing generalized confidence intervals for the covariate-specific overlap coefficient. A simulation study is conducted to assess the finite-sample performance of the proposed generalized confidence interval method and to compare it with other procedures. Finally, an illustration to assess the age-specific accuracy of glucose as a biomarker of diabetes is provided.

翻译：暂无翻译