Optimal dynamic treatment regimes

被引:610
作者
Murphy, SA
机构
[1] Department of Statistics, 439 West Hall, University of Michigan, Ann Arbor
[2] University of Helsinki, Helsinki
[3] University of Bath, Bath
[4] University College London, London
[5] Nuffield College, Oxford
[6] City University, London
[7] University of Utrecht, Utrecht
[8] Carnegie Mellon University, Pittsburgh, PA
[9] Harvard School of Public Health, Boston, MA
关键词
adaptive strategies; causal inference; dynamic programming; multistage decisions;
D O I
10.1111/1467-9868.00389
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
A dynamic treatment regime is a list of decision rules, one per time interval, for how the level of treatment will be tailored through time to an individual's changing status. The goal of this paper is to use experimental or observational data to estimate decision regimes that result in a maximal mean response. To explicate our objective and to state the assumptions, we use the potential outcomes model. The method proposed makes smooth parametric assumptions only on quantities that are directly relevant to the goal of estimating the optimal rules. We illustrate the methodology proposed via a small simulation.
引用
收藏
页码:331 / 355
页数:25
相关论文
共 41 条
  • [1] [Anonymous], 1999, J Consult Clin Psychol, V67, P631
  • [2] Bather John., 2000, DECISION THEORY INTR
  • [3] Bellman R., 1957, DYNAMIC PROGRAMMING
  • [4] Bertsekas D. P., 1996, Neuro Dynamic Programming, V1st
  • [5] BIELZA C, 2001, UNPUB DECISION ANAL
  • [6] BIERMAN KL, 2001, UNPUB EVALUATING USE
  • [7] BROCKWELL AE, 2001, UNPUB GRIDDING METHO
  • [8] Approaches for optimal sequential decision analysis in clinical trials
    Carlin, BP
    Kadane, JB
    Gelfand, AE
    [J]. BIOMETRICS, 1998, 54 (03) : 964 - 975
  • [9] COLLINS LM, 2001, UNPUB DESIGN EVALUAT
  • [10] *COND PROBL PREV R, 1996, J CONSULT CLIN PSYCH, V67, P648