English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

In semicompeting risks problems, nonterminal time-to-event outcomes such astime to hospital readmission are subject to truncation by death. These settingsare often modeled with illness-death models for the hazards of the terminal andnonterminal events, but evaluating causal treatment effects with hazard modelsis problematic due to conditioning on survival (a post-treatment outcome) thatis embedded in the definition of a hazard. Extending an existing survivoraverage causal effect (SACE) estimand, we frame the evaluation of treatmenteffects in the context of semicompeting risks with principal stratification andintroduce two new causal estimands: the time-varying survivor average causaleffect (TV-SACE) and the restricted mean survivor average causal effect(RM-SACE). These principal causal effects are defined among units that wouldsurvive regardless of assigned treatment. We adopt a Bayesian estimationprocedure that parameterizes illness-death models for both treatment arms. Weoutline a frailty specification that can accommodate within-person correlationbetween nonterminal and terminal event times, and we discuss potential avenuesfor adding model flexibility. The method is demonstrated in the context ofhospital readmission among late-stage pancreatic cancer patients.

Survivor average causal effects for continuous time: a principal stratification approach to causal inference with semicompeting risks