What is the purpose of using a time-dependent Cox model in cohort studies?

• A. It ignores time-varying covariates
• B. It uses time-dependent methods to handle time-varying exposures and covariates
• C. It assumes no censoring
• D. It assumes proportional hazards with fixed covariates

Answer: (B)

Time-dependent Cox models are designed to handle variables that change during follow-up. In cohort studies with time-to-event data, exposures or covariates (like treatment, smoking status, or a biomarker) may not stay fixed from the start. The time-dependent approach updates covariate values as they change and uses those current values to estimate the hazard at each moment. This means the risk set is evaluated with the covariate status at that time, and person-time is split into intervals where covariates are constant. By doing this, the model can assess how changing exposure affects the hazard and produce unbiased estimates when exposures vary over time. Ignoring time-varying covariates would misclassify exposure and bias results, and the other options either misstate censoring assumptions or mischaracterize the role of time-varying covariates in the model.