Two‐sample density‐based empirical likelihood tests for incomplete data in application to a pneumonia study

摘要: In clinical trials examining the incidence of pneumonia it is a common practice to measure infection via both invasive and non-invasive procedures. context recently completed randomized trial comparing two treatments procedure was only utilized in certain scenarios due added risk involved, given that level surpassed threshold. Hence, what observed bivariate data with pattern missingness variable dependent upon value observation within pair. order compare exhibiting this we developed semi-parametric methodology utilizing density-based empirical likelihood approach provide non-parametric approximation Neyman-Pearson-type test statistics. This novel has parametric components. The component utilizes observations for non-missing cases, while tackle case where are missing respect variable. method illustrated through its application actual obtained study shown be an efficient practical method.