In an observational study with a time-to-event outcome, the standard analytical approach is the Cox proportional hazards regression model. As an alternative to the standard Cox model, in this article we present a method that uses inverse probability (IP) weights to estimate the effect of a baseline exposure on a time-to-event outcome. IP weighting can be used to adjust for multiple measured confounders of a baseline exposure in order to estimate marginal effects, which compare the distribution of outcomes when the entire population is exposed versus when the entire population is unexposed. For example, IP-weighted Cox models allow for estimation of the marginal hazard ratio and marginal survival curves. IP weights can also be employed to adjust for selection bias due to loss to follow-up. This approach is illustrated using an example that estimates the effect of injection drug use on time until AIDS or death among HIV-infected women.
Get full access to this article
View all access options for this article.
References
1.
ColeSR and HudgensMG: Survival analysis in infectious disease research: Describing events in time. AIDS, 2010; 24(16):2423–2431.
2.
GreenlandS and MorgensternH: Confounding in health research. Annu Rev Public Health, 2001; 22(1):189–212.
3.
CoxDR: Regression models and life-tables. J R Stat Soc Ser B Methodol, 1972; 34(2):187–220.
KaplanEL and MeierP: Nonparametric estimation from incomplete observations. J Am Stat Assoc, 1958; 53(282):457–481.
6.
XieJ and LiuC: Adjusted Kaplan Meier estimator and log rank test with inverse probability of treatment weighting for survival data. Stat Med, 2005; 24(20):3089–3110.
7.
ColeSR and HernánMA: Adjusted survival curves with inverse probability weights. Comput Methods Programs Biomed, 2004; 75(1):45–49.
8.
HernánMÁ, BrumbackB, and RobinsJM: Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men. Epidemiology, 2000; 11(5):561–570.
HorvitzDG and ThompsonDJ: A generalization of sampling without replacement from a finite universe. J Am Stat Assoc, 1952; 47(260):663–685.
11.
SatoT and MatsuyamaY: Marginal structural models as a tool for standardization. Epidemiology, 2003; 14(6):680–686.
12.
BaconMC, von WylV, AldenC, SharpG, RobisonE, HessolN, et al.: The Women's Interagency HIV Study: An observational cohort brings clinical sciences to the bench. Clin Diagn Lab Immunol, 2005; 12(9):1013–1019.
13.
LauB, ColeSR, and GangeSJ: Competing risk regression models for epidemiologic data. Am J Epidemiol, 2009; 170(2):244–256.
14.
RobinsJM, HernánMÁ, and BrumbackB: Marginal structural models and causal inference in epidemiology. Epidemiology, 2000; 11(5):550–560.
15.
LinD and WeiL: The robust inference for the Cox proportional hazards model. J Am Stat Assoc, 1989; 84(408):1074–1078.
16.
ColeSR and HernánMA: Constructing inverse probability weights for marginal structural models. Am J Epidemiol, 2008; 168(6):656–664.
17.
CollettD: Modelling Survival Data in Medical Research. CRC Press, Boca Raton, FL, 2003.
18.
EfronB: The efficiency of Cox's likelihood function for censored data. J Am Stat Assoc, 1977; 72(359):557–565.
19.
EfronB and TibshrianiR: An Introduction to the Bootstrap. Chapman Hall, London, 1994.
20.
HoweCJ, ColeSR, WestreichDJ, GreenlandS, NapravnikS, and EronJJJr: Splines for trend analysis and continuous confounder control. Epidemiology, 2011; 22(6):874–875.
21.
HernánMA, BrumbackB, and RobinsJM: Marginal structural models to estimate the joint causal effect of nonrandomized treatments. J Am Stat Assoc, 2001; 96(454):440–448.
22.
D'AgostinoRB, LeeM, BelangerAJ, CupplesLA, AndersonK, and KannelWB: Relation of pooled logistic regression to time dependent Cox regression analysis: The Framingham Heart Study. Stat Med, 1990; 9(12):1501–1515.
23.
GreenlandS: Absence of confounding does not correspond to collapsibility of the rate ratio or rate difference. Epidemiology, 1996; 7(5):498–501.
24.
GreenlandS, RobinsJM, and PearlJ: Confounding and collapsibility in causal inference. Stat Sci, 1999; 14(1):29–46.
25.
HernánMA: The hazards of hazard ratios. Epidemiology, 2010; 21(1):13–15.
26.
HoweCJ, ColeSR, ChmielJS, and MuñozA: Limitation of inverse probability-of-censoring weights in estimating survival in the presence of strong selection bias. Am J Epidemiol, 2011; 173(5):569–577.
27.
KalbfleischJD and PrenticeRL: The Statistical Analysis of Failure Time Data. Wiley-Interscience, Hoboken, NJ, 2002.
28.
RobinsJM and FinkelsteinDM: Correcting for noncompliance and dependent censoring in an AIDS clinical trial with inverse probability of censoring weighted log rank tests. Biometrics, 2000; 56(3):779–788.
29.
HernánMA, Hernández-DíazS, and RobinsJM: Randomized trials analyzed as observational studies. Ann Intern Med, 2013; 159(8):560–562.
30.
VanderWeeleTJ: Concerning the consistency assumption in causal inference. Epidemiology, 2009; 20(6):880–883.
31.
ColeSR and FrangakisCE: The consistency statement in causal inference: A definition or an assumption?. Epidemiology, 2009; 20(1):3–5.
32.
PearlJ: On the consistency rule in causal inference: Axiom, definition, assumption, or theorem?. Epidemiology, 2010; 21(6):872–875.
33.
ColeSR and HernánMA: Fallibility in estimating direct effects. Int J Epidemiol, 2002; 31(1):163–165.
34.
PearlJ: Direct and indirect effects. Presented at Proceedings of the Seventeenth Conference on Uncertainty in Artificial Intelligence, Seattle, 2001.
35.
ColeSR, HernánMA, RobinsJM, AnastosK, ChmielJ, DetelsR, et al.: Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models. Am J Epidemiol, 2003; 158(7):687–694.
36.
RobinsJM: Marginal structural models versus structural nested models as tools for causal inference. In: Statistical Models in Epidemiology, the Environment, and Clinical Trials. Springer, New York, 2000, pp. 95–133.
37.
VanderWeeleTJ: Marginal structural models for the estimation of direct and indirect effects. Epidemiology, 2009; 20(1):18–26.
38.
MickeyRM and GreenlandS: The impact of confounder selection criteria on effect estimation. Am J Epidemiol, 1989; 129(1):125–137.
39.
VanderWeeleTJ and ShpitserI: A new criterion for confounder selection. Biometrics, 2011; 67(4):1406–1413.
40.
GreenlandS, PearlJ, and RobinsJM: Causal diagrams for epidemiologic research. Epidemiology, 1999; 10(1):37–48.
41.
BrookhartMA, SchneeweissS, RothmanKJ, GlynnRJ, AvornJ, and StürmerT: Variable selection for propensity score models. Am J Epidemiol, 2006; 163(12):1149–1156.
42.
LeeBK, LesslerJ, and StuartEA: Weight trimming and propensity score weighting. PloS One, 2011; 6(3):e18174.
43.
LeeBK, LesslerJ, and StuartEA: Improving propensity score weighting using machine learning. Stat Med, 2010; 29(3):337–346.
44.
KishL: Weighting for unequal P. J Official Stat, 1992; 8(2):183–200.
45.
HernánMA and ColeSR: Invited commentary: Causal diagrams and measurement bias. Am J Epidemiol, 2009; 170(8):959–962.
PetoR and PetoJ: Asymptotically efficient rank invariant test procedures. J R Stat Soc Ser A, 1972; 135(2):185–207.
48.
BreslowN: Covariance analysis of censored survival data. Biometrics, 1974; 58(3):643–649.
49.
Hertz-PicciottoI and RockhillB: Validity and efficiency of approximation methods for tied survival times in Cox regression. Biometrics, 1997; 53(3):1151–1156.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
