Assessing artificial censoring weights for clone censor weighting studies: A simulation-based approach

Author: David Pritchard

Challenge

Clone-censor-weight studies require estimation of artificial censoring weights, but no simulation-based framework existed for evaluating the performance of different censoring weight estimation approaches.

Solution

Target RWE biostatisticians developed a simulation framework to assess artificial censoring weight estimation methods for CCW studies, demonstrating that Cox proportional hazards-based estimation substantially outperforms logistic regression.

Impact

Establishing Cox-based censoring weight estimation as the preferred approach provides Target RWE with a validated analytic standard, ensuring that CCW studies are built on methodologically sound foundations.