Information anchored reference‐based sensitivity analysis for truncated normal data with application to survival analysis

The primary analysis of time-to-event data typically makes the censoring at random assumption, that is, that—conditional on covariates in model—the distribution event times is same, whether they are observed or unobserved. In such cases, we need to explore robustness inference more pragmatic assumptions about patients post-censoring sensitivity analyses. Reference-based multiple imputation, which avoids analysts explicitly specifying parameters unobserved distribution, has proved attractive researchers. Building results for longitudinal continuous data, show using a Tobit regression imputation model reference-based with right censored log normal information anchored, meaning proportion lost due missing under held constant across We illustrate our theoretical simulation and clinical trial case study.