Answers are grounded in this paper and cite their sources.
Evaluates outcomes for switchers and continuers, suggesting improved bias reduction methods.
When hazard varies over time, evaluating continuers from treatment initiation introduces timing bias that can reverse the apparent direction of treatment effect. SMARTS, combined with confounder adjustment via propensity score matching, IPTW or standardized morbidity ratio weighting, substantially reduced this bias across diverse clinical scenarios and hazard trends.
No takes yet. Share an insight, caveat, or question.
Gao et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: