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

Journal of Internal Medicine - Tập 275 Số 6 - Trang 570-580 - 2014
Til Stürmer‎1, Richard Wyss1, Robert J. Glynn2, M. Alan Brookhart1
1Department of Epidemiology, UNC Gillings School of Global Public Health University of North Carolina at Chapel Hill Chapel Hill NC USA
2Division of Preventive Medicine & Division of Pharmacoepidemiology and Pharmacoeconomics Brigham and Women's Hospital Harvard Medical School Boston MA USA

Tóm tắt

Abstract

Treatment effects, especially when comparing two or more therapeutic alternatives as in comparative effectiveness research, are likely to be heterogeneous across age, gender, co‐morbidities and co‐medications. Propensity scores (PSs), an alternative to multivariable outcome models to control for measured confounding, have specific advantages in the presence of heterogeneous treatment effects. Implementing PSs using matching or weighting allows us to estimate different overall treatment effects in differently defined populations. Heterogeneous treatment effects can also be due to unmeasured confounding concentrated in those treated contrary to prediction. Sensitivity analyses based on PSs can help to assess such unmeasured confounding. PSs should be considered a primary or secondary analytic strategy in nonexperimental medical research, including pharmacoepidemiology and nonexperimental comparative effectiveness research.

Từ khóa


Tài liệu tham khảo

10.1056/NEJM200006223422507

10.1002/sim.4780020222

10.1002/sim.4780030421

10.1093/aje/kwi223

Walker AM, 1996, Confounding by indication, Epidemiology, 7, 335

10.1093/biomet/70.1.41

10.1016/j.jclinepi.2005.07.004

10.1093/aje/kwi106

10.1002/sim.2739

10.1093/aje/kwg231

10.1111/j.1742-7843.2006.pto_293.x

10.1093/aje/kwp435

10.1093/aje/kwp436

10.1093/aje/kwq198

10.1097/EDE.0b013e3181875e61

10.1097/MLR.0b013e3181dbebe3

10.1016/S0146-0005(97)80013-4

10.1056/NEJM199805213382106

10.1097/00001648-200111000-00017

Rothman KJ, 2008, Modern Epidemiology

10.1056/NEJMoa032709

10.1200/JCO.2011.41.1140

10.1056/NEJM198907203210301

10.1056/NEJMoa050613

Pearl J, 2000, Causality: Models, Reasoning, and Inference

10.1097/00001648-200009000-00011

10.1097/00001648-200009000-00012

10.1002/pds.1231

10.1002/pds.1555

10.1002/sim.3782

Ellis AR, 2011, Propensity scores from logistic regression yielded better covariate balance than those from a boosted model: a demonstration Using STAR*D Data, Pharmacoepidemiol Drug Saf, 20, S129

ImaiK RatkovicM.Covariate balancing propensity score.2012; Working paper available athttp://imai.princeton.edu/research/CBPS.html.

10.1093/aje/kwn391

10.1097/00001648-200105000-00011

10.1002/pds.2074

10.1093/aje/kwj149

10.1093/aje/kwr364

10.1002/pds.2098

10.1097/EDE.0b013e3181a663cc

10.1093/aje/kwr001

10.1002/pds.2152

Mantel N, 1959, Statistical aspects of the analysis of data from retrospective studies of disease, J Natl Cancer Inst, 22, 719

Stürmer T, 2009, Matching in cohort studies: return of a long lost family member, Am J Epidemiol, 169, S128

10.1080/00220671.1941.10881036

10.2307/2985543

Ellis AR, 2013, Investigating differences in treatment effect estimates between propensity score matching and weighting: a demonstration using STAR*D trial data, 138

10.1002/bimj.200810488

ParsonsLS.Reducing bias in a propensity score matched‐pair sample using greedy matching techniques 2001. (http://www2.sas.com/proceedings/sugi26/p214-26.pdf).

10.1097/01.EDE.0000081989.82616.7d

10.1093/aje/kwj047

10.1136/jech.2002.006361

10.1093/aje/kwn164

10.1093/ije/dyi274