×
Williamson, Elizabeth; Morley, Ruth; Lucas, Alan; Carpenter, James

Propensity scores: from naïve enthusiasm to intuitive understanding. (English) Zbl 1242.62124

Stat. Methods Med. Res. 21, No. 3, 273-293 (2012).
Summary: Estimation of the effect of a binary exposure on an outcome in the presence of confounding is often carried out via outcome regression modelling. An alternative approach is to use propensity score methodology. The propensity score is the conditional probability of receiving the exposure given the observed covariates and can be used, under the assumption of no unmeasured confounders, to estimate the causal effect of the exposure. We provide a non-technical and intuitive discussion of propensity score methodology, motivating the use of the propensity score approach by analogy with randomised studies, and describe the four main ways in which this methodology can be implemented. We carefully describe the population parameters being estimated - an issue that is frequently overlooked in the medical literature. We illustrate these four methods using data from a study investigating the association between maternal choice to provide breast milk and the infant’s subsequent neurodevelopment. We outline useful extensions of propensity score methodology and discuss directions for future research. Propensity score methods remain controversial and there is no consensus as to when, if ever, they should be used in place of traditional outcome regression models. We therefore end with a discussion of the relative advantages and disadvantages of each.

Cited in 4 Documents

MSC:

62P10 Applications of statistics to biology and medical sciences; meta analysis
92C50 Medical applications (general)

Keywords:

confounding; inverse probability weighting; matching; observational study; propensity score; stratification
PDF BibTeX XML Cite
\textit{E. Williamson} et al., Stat. Methods Med. Res. 21, No. 3, 273--293 (2012; Zbl 1242.62124)
Full Text: DOI
  • logo of hbz
  • logo of WorldCat

References:

[1] DOI: 10.1093/biomet/70.1.41 · Zbl 0522.62091
[2] DOI: 10.1002/sim.3133
[3] DOI: 10.1002/(SICI)1097-0258(19981015)17:19<2265::AID-SIM918>3.0.CO;2-B
[4] DOI: 10.1093/oxfordjournals.aje.a010011
[5] DOI: 10.1002/sim.2328
[6] DOI: 10.1093/aje/kwj047
[7] Austin PC, Stat Med 27 pp 2066– (2008)
[8] DOI: 10.1002/sim.3150
[9] Rodgers B, Dev Med Child Neurol 20 pp 421– (1978)
[10] Weschler D, Weschler Intelligence Scale for Children, Anglicized revised edition (1974)
[11] DOI: 10.1080/01621459.1986.10478354
[12] DOI: 10.1162/003465304323023651
[13] DOI: 10.1093/aje/kwn164
[14] DOI: 10.2307/2529981 · Zbl 0463.62015
[15] Little R, Ann Rev Public Health 21 pp 121– (2000)
[16] Hullsiek KH, Biostatistics 3 pp 179– (2002) · Zbl 1133.62361
[17] DOI: 10.1002/sim.1903
[18] DOI: 10.1080/01621459.1984.10478078
[19] Baser O, Value Health 9 pp 377– (2006)
[20] Hill J, Stat Med 27 pp 2055– (2008)
[21] Zhao Z, Rev Econ Stat 86 pp 91– (2004)
[22] Rubin DB, Pharmacoepidemiol Drug Safe 13 pp 855– (2004)
[23] Weitzen S, Pharmacoepidemiol Drug Safe 13 pp 841– (2004)
[24] DOI: 10.2307/2532266
[25] Zhao Z, Econ Lett 98 pp 309– (2008) · Zbl 06132279
[26] DOI: 10.1037/1082-989X.9.4.403
[27] DOI: 10.1002/pds.1555
[28] Judkins DR, Stat Med 26 pp 1022– (2007)
[29] Gu XS, J Comput Graph Stat 2 pp 405– (1993)
[30] DOI: 10.1093/aje/kwj149
[31] Austin PC, Pharmacoepidemiol Drug Safe 17 pp 1218– (2008)
[32] Austin PC, Pharmacoepidemiol Drug Safe 17 pp 1202– (2008)
[33] Stuart EA, Stat Med 27 pp 2062– (2008)
[34] Stuürmer T, Am J Epidemiol 162 pp 279– (2005)
[35] Weitzen S, Pharmacoepidemiol Drug Safe 14 pp 227– (2005)
[36] DOI: 10.1007/978-1-4757-3692-2
[37] Heckman J, Econometrica 66 pp 1017– (1998) · Zbl 1055.62573
[38] DOI: 10.1002/sim.2618
[39] DOI: 10.1016/j.jclinepi.2007.07.011
[40] DOI: 10.1214/ss/1009211805 · Zbl 1059.62506
[41] Martens EP, Stat Med 26 pp 3205– (2007)
[42] Forbes A, Stat Med 27 pp 5556– (2008)
[43] DOI: 10.1002/sim.2781
[44] DOI: 10.1093/oxfordjournals.aje.a114593
[45] DOI: 10.1093/biomet/87.3.706 · Zbl 1120.62334
[46] Williamson E. Inference from estimates of exposure effects using stratification on the propensity score. PhD thesis, London School of Hygiene &Tropical Medicine, London, 2008.
[47] DOI: 10.1002/sim.2277
[48] Tu WZ, Health Serv Outcome Res Methodol 3 pp 135– (2002)
[49] Lucas A, Lancet 339 pp 261– (1992)
[50] Lucas A, Archives of Disease in Childhood 59 pp 722– (1984)
[51] Lucas A, Lancet 335 pp 1477– (1990)
[52] DOI: 10.1002/(SICI)1097-0258(20000515)19:9<1141::AID-SIM479>3.0.CO;2-F
[53] Robins JM, Biometrics 48 pp 479– (1992) · Zbl 0768.62099
[54] Fitzmaurice G, Nutrition 22 pp 1214– (2006)
[55] DOI: 10.1016/j.jclinepi.2004.10.016
[56] DOI: 10.1080/01621459.2000.10474233
[57] Mattei A, Stat Methods Appl 18 pp 257– (2009) · Zbl 1405.62018
[58] Qu Y, Stat Med 28 pp 1402– (2009)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.