Upgrade to Pro — share decks privately, control downloads, hide ads and more …

G-methods for time-varying treatments (Causal inference: What if, Chapter 21-1)

38e2af7f8bdad4f2087ab3d42b627e33?s=47 Shuntaro Sato
November 25, 2020

G-methods for time-varying treatments (Causal inference: What if, Chapter 21-1)

Keywords: 因果推論, Time-varying, G-formula, IP weighting, Doubly robust estimation

38e2af7f8bdad4f2087ab3d42b627e33?s=128

Shuntaro Sato

November 25, 2020
Tweet

Transcript

  1. None
  2. ・G-methods for time-fixed treatments 本日の内容 ・The g-formula for time-varying treatments

    ・IP weighting for time-varying treatments ・A doubly robust estimator for time-varying treatments
  3. ・G-methods for time-fixed treatments 本日の内容 ・The g-formula for time-varying treatments

    ・IP weighting for time-varying treatments ・A doubly robust estimator for time-varying treatments
  4. Stratification effect measure modification (-) effect measure modification (+) Mantel-Haenszel

    method 別々にオッズ比を報告
  5. Why model? effect measure modification (-) effect measure modification (+)

    別々にオッズ比を報告(1つの効果を報告できない) g-methods
  6. g-formula A=1を代入 A=0を代入

  7. IP weighting marginal structural model

  8. Conditional or Marginal? outcome regression saturated parametric stratification g-formula IP

    weighting g-estimation or algebraically equivalent
  9. Time-varying treatment g-methods

  10. ・G-methods for time-fixed treatments 本日の内容 ・The g-formula for time-varying treatments

    ・IP weighting for time-varying treatments ・A doubly robust estimator for time-varying treatments
  11. 前提 ・本章ではidentifiability conditions(sequential exchangeability, positivity, and consistency)のviolationが ないものとする。 ・static treatment

    strategies (always treat vs. never treat) の効果を推定する。
  12. g-formula (weighted average) ・time-fixed treatment (A1 の反実アウトカム) ・time-varying treatment

  13. g-formula (weighted average)

  14. g-formula (weighted average)

  15. g-formula (simulation) のシミュレーション と

  16. g-formulaの注意点 ・DAGに基づいたcovariates L1 をモデルに含める ・static sequential exchangeabilityが成立すればstatic treatment strategyの効果はidentify可能

  17. g-formulaの一般化 ・static treatment strategy ・dynamic treatment strategy linear regression logistic

    regression
  18. ・G-methods for time-fixed treatments 本日の内容 ・The g-formula for time-varying treatments

    ・IP weighting for time-varying treatments ・A doubly robust estimator for time-varying treatments
  19. IP weighting (weights) ・nonstabilized IP weights ・ stabilized IP weights

  20. IP weighting (non-stabilized)

  21. Stabilized weights non-stabilized weights: stabilized weights: Lと独立であればよい Lと独立であればよい

  22. IP weighting (stabilized)

  23. IP weightingの一般化 ・nonstabilized IP weights ・ stabilized IP weights logistic

    regression logistic regression (misspecifiedでも可)
  24. Marginal Structural Model ・2K > Nのときは推定できない ・marginal structural mean model

    stabilized IP weightsを使って推定 misspecified??
  25. Effect Measure Modification ・baseline variable VによるEMMがある場合、marginal structural modelは以下の通り(parametric) stabilized IP

    weightsを使って推定 Vに入れて良いのはbaseline variableだけ
  26. ・G-methods for time-fixed treatments 本日の内容 ・The g-formula for time-varying treatments

    ・IP weighting for time-varying treatments ・A doubly robust estimator for time-varying treatments
  27. Doubly Robust Estimator ・g-formula ・ IP weighting

  28. 1. Doubly Robust (time-fixed) 2. 3. A=1とA=0でそれぞれ を推定 を推定 ,

    をLについて標準化
  29. 1. Doubly Robust (time-varying) 2. 3. を推定 からパラメータ を求める。 を求めておく

    を推定し、Aの値に応じた を求める。 これを繰り返して を求める。 always treat
  30. Discussion