Literature DB >> 23280682

Investigating differences in treatment effect estimates between propensity score matching and weighting: a demonstration using STAR*D trial data.

Alan R Ellis1, Stacie B Dusetzina, Richard A Hansen, Bradley N Gaynes, Joel F Farley, Til Stürmer.   

Abstract

PURPOSE: The choice of propensity score (PS) implementation influences treatment effect estimates not only because different methods estimate different quantities, but also because different estimators respond in different ways to phenomena such as treatment effect heterogeneity and limited availability of potential matches. Using effectiveness data, we describe lessons learned from sensitivity analyses with matched and weighted estimates.
METHODS: With subsample data (N = 1292) from Sequenced Treatment Alternatives to Relieve Depression, a 2001-2004 effectiveness trial of depression treatments, we implemented PS matching and weighting to estimate the treatment effect in the treated and conducted multiple sensitivity analyses.
RESULTS: Matching and weighting both balanced covariates but yielded different samples and treatment effect estimates (matched RR 1.00, 95% CI: 0.75-1.34; weighted RR 1.28, 95% CI: 0.97-1.69). In sensitivity analyses, as increasing numbers of observations at both ends of the PS distribution were excluded from the weighted analysis, weighted estimates approached the matched estimate (weighted RR 1.04, 95% CI 0.77-1.39 after excluding all observations below the 5th percentile of the treated and above the 95th percentile of the untreated). Treatment appeared to have benefits only in the highest and lowest PS strata.
CONCLUSIONS: Matched and weighted estimates differed due to incomplete matching, sensitivity of weighted estimates to extreme observations, and possibly treatment effect heterogeneity. PS analysis requires identifying the population and treatment effect of interest, selecting an appropriate implementation method, and conducting and reporting sensitivity analyses. Weighted estimation especially should include sensitivity analyses relating to influential observations, such as those treated contrary to prediction.
Copyright © 2012 John Wiley & Sons, Ltd.

Entities:  

Mesh:

Substances:

Year:  2012        PMID: 23280682      PMCID: PMC3639482          DOI: 10.1002/pds.3396

Source DB:  PubMed          Journal:  Pharmacoepidemiol Drug Saf        ISSN: 1053-8569            Impact factor:   2.890


  28 in total

1.  Marginal structural models as a tool for standardization.

Authors:  Tosiya Sato; Yutaka Matsuyama
Journal:  Epidemiology       Date:  2003-11       Impact factor: 4.822

2.  Results of multivariable logistic regression, propensity matching, propensity adjustment, and propensity-based weighting under conditions of nonuniform effect.

Authors:  Tobias Kurth; Alexander M Walker; Robert J Glynn; K Arnold Chan; J Michael Gaziano; Klaus Berger; James M Robins
Journal:  Am J Epidemiol       Date:  2005-12-21       Impact factor: 4.897

Review 3.  Indications for propensity scores and review of their use in pharmacoepidemiology.

Authors:  Robert J Glynn; Sebastian Schneeweiss; Til Stürmer
Journal:  Basic Clin Pharmacol Toxicol       Date:  2006-03       Impact factor: 4.080

4.  STAR*D: what have we learned?

Authors:  A John Rush
Journal:  Am J Psychiatry       Date:  2007-02       Impact factor: 18.112

5.  Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group.

Authors:  R B D'Agostino
Journal:  Stat Med       Date:  1998-10-15       Impact factor: 2.373

6.  Identifiability, exchangeability, and epidemiological confounding.

Authors:  S Greenland; J M Robins
Journal:  Int J Epidemiol       Date:  1986-09       Impact factor: 7.196

7.  On the use of propensity scores in principal causal effect estimation.

Authors:  Booil Jo; Elizabeth A Stuart
Journal:  Stat Med       Date:  2009-10-15       Impact factor: 2.373

8.  Constructing inverse probability weights for marginal structural models.

Authors:  Stephen R Cole; Miguel A Hernán
Journal:  Am J Epidemiol       Date:  2008-08-05       Impact factor: 4.897

9.  Sequenced treatment alternatives to relieve depression (STAR*D): rationale and design.

Authors:  A John Rush; Maurizio Fava; Stephen R Wisniewski; Philip W Lavori; Madhukar H Trivedi; Harold A Sackeim; Michael E Thase; Andrew A Nierenberg; Frederic M Quitkin; T Michael Kashner; David J Kupfer; Jerrold F Rosenbaum; Jonathan Alpert; Jonathan W Stewart; Patrick J McGrath; Melanie M Biggs; Kathy Shores-Wilson; Barry D Lebowitz; Louise Ritz; George Niederehe
Journal:  Control Clin Trials       Date:  2004-02

10.  Bupropion-SR, sertraline, or venlafaxine-XR after failure of SSRIs for depression.

Authors:  A John Rush; Madhukar H Trivedi; Stephen R Wisniewski; Jonathan W Stewart; Andrew A Nierenberg; Michael E Thase; Louise Ritz; Melanie M Biggs; Diane Warden; James F Luther; Kathy Shores-Wilson; George Niederehe; Maurizio Fava
Journal:  N Engl J Med       Date:  2006-03-23       Impact factor: 91.245

View more
  6 in total

1.  Propensity scores for confounder adjustment when assessing the effects of medical interventions using nonexperimental study designs.

Authors:  T Stürmer; R Wyss; R J Glynn; M A Brookhart
Journal:  J Intern Med       Date:  2014-02-13       Impact factor: 8.989

Review 2.  Propensity score methods to control for confounding in observational cohort studies: a statistical primer and application to endoscopy research.

Authors:  Jeff Y Yang; Michael Webster-Clark; Jennifer L Lund; Robert S Sandler; Evan S Dellon; Til Stürmer
Journal:  Gastrointest Endosc       Date:  2019-04-30       Impact factor: 9.427

3.  Vector-based kernel weighting: A simple estimator for improving precision and bias of average treatment effects in multiple treatment settings.

Authors:  Melissa M Garrido; Jessica Lum; Steven D Pizer
Journal:  Stat Med       Date:  2020-12-16       Impact factor: 2.373

4.  Marginal Structural Models with Counterfactual Effect Modifiers.

Authors:  Wenjing Zheng; Zhehui Luo; Mark J van der Laan
Journal:  Int J Biostat       Date:  2018-06-08       Impact factor: 1.829

5.  Single-arm Trials With External Comparators and Confounder Misclassification: How Adjustment Can Fail.

Authors:  Michael Webster-Clark; Michele Jonsson Funk; Til Stürmer
Journal:  Med Care       Date:  2020-12       Impact factor: 3.178

6.  Propensity score estimation to address calendar time-specific channeling in comparative effectiveness research of second generation antipsychotics.

Authors:  Stacie B Dusetzina; Christina D Mack; Til Stürmer
Journal:  PLoS One       Date:  2013-05-07       Impact factor: 3.240

  6 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.