Censoring Balancing Functions for Undetected Probably Significant Effects in Cox Regression

Abstract

Weighted Cox regression models were proposed as an alternative to the standard Cox proportional hazards models where consistent estimators can be obtained with more relative strength compared to unweighted cases. We proposed censoring balancing functions which can be built in a way that allows us to obtain the maximum possible significant treatment effects that may have gone undetected due to censoring. The harm caused by this is compensated and new weighted parameter estimates are found. These functions are constructed to be monotonic because even the hazard ratios are not exactly constant as in the ideal case, but are violated by monotonic deviations in time. For more than one covariate, even the interaction between covariates in addition to censoring can lead to the loss of significance for some covariates’ effects. Undetected significant effects of one covariate can be obtained, still keeping the significance and approximate size of the remaining one(s). This is performed by keeping the consistency of the parameter estimates. The results from both the simulated datasets and their application to real datasets supported the importance of the suggested censoring balancing functions in both one covariate and more than one covariate cases.