Causal Estimands And Interventions
summary
Proposes separable effects as an estimand that splits a randomized treatment into two manipulable components (one driving the disease pathway, one driving competing events or side effects), formalizes identification via an extended causal graph and the mediation g‑formula, and contrasts the observed two‑arm trial with a four‑arm hypothetical trial where components differ. Applies this to a prostate cancer trial: the estimated direct effect nearly equals the total effect while the indirect (via competing mortality) is small, under strong no‑unmeasured confounding and separability assumptions.
AI-written summary · about
chapters (7)
- 0:00Introduction to estimands
- 1:57Target trial emulation
- 5:02Challenges in randomized trials
- 9:11Concept of separable effects
- 15:01Formalizing causal models
- 20:00Application in survival settings
- 28:10Concluding remarks
related talks
33mSimple Yet Efficient Estimators For Network Causal Inference...
24mA Friendly Introduction To Causality
31mA Gentle Introduction To Causal Inference
41mOn the Path to Causal Inference
44mCausal Reductions
39mCausal Effect Estimation in Practice: Lessons Learned from E-commerce & Banking
11mCausal Inference For Socio-Economic & Engineering Systems
49mCausal Inference in Complex Systems: Network Interference, Strategic Agents, and Beyond