European Journal of Epidemiology

, Volume 24, Issue 7, pp 343–349 | Cite as

The impact of unmeasured baseline effect modification on estimates from an inverse probability of treatment weighted logistic model

  • Joseph A. C. Delaney
  • Robert W. Platt
  • Samy Suissa
Methods
First Online:
Received:
Accepted:

Abstract

We present the results of a Monte Carlo simulation study in which we demonstrate how strong baseline interactions between a confounding variable and a treatment can create an important difference between the marginal effect of exposure on outcome (as estimated by an inverse probability of treatment weighted logistic model) and the conditional effect (as estimated by an adjusted logistic regression model). The scenarios that we explored included one with a rare outcome and a strong and prevalent effect measure modifier where, across 1,000 simulated data sets, the estimates from an adjusted logistic regression model (mean β = 0.475) and an inverse probability of treatment weighted logistic model (mean β = 2.144) do not coincide with the known true effect (β = 0.68925) when the effect measure modifier is not accounted for. When the marginal and conditional estimates do not coincide despite a rare outcome this may suggest that there is heterogeneity in the effect of treatment between individuals. Failure to specify effect measure modification in the statistical model appears to results in systematic differences between the conditional and marginal estimates. When these differences in estimates are observed, testing for or including interactions or non-linear modeling terms may be advised.

Keywords

Effect measure modification Statistical models Epidemiology Inverse probability weighting 
This is a preview of subscription content, log in to check access.

Notes

Acknowledgments

RP is the recipient of a Chercheur-boursier award from the Fonds de Recherche en Santé du Québec. SS is the recipient of a Distinguished Investigator Award from CIHR.

References

  1. 1.
    Rothman KJ, Greenland S. Modern epidemiology. Philadelphia: Lippincott Williams & Wilkins; 1998.Google Scholar
  2. 2.
    Clayton D, Hills M. Statistical models in epidemiology. Oxford: Oxford University Press; 1993.Google Scholar
  3. 3.
    Suissa S. Novel approaches to pharmacoepidemiology study design and statistical analysis. In: Strom BL, editor. Pharmacoepidemiology. Sussex: Wiley; 2000. p. 785–805.CrossRefGoogle Scholar
  4. 4.
    Robins JM, Hernán MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology. 2000;11:550–60. doi: 10.1097/00001648-200009000-00011.PubMedCrossRefGoogle Scholar
  5. 5.
    Hernán MA, Brumback B, Robins JM. Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men. Epidemiology. 2000;11:561–70. doi: 10.1097/00001648-200009000-00012.PubMedCrossRefGoogle Scholar
  6. 6.
    Cole SR, Hernán MA, Robins JM, Anastos K, Chmiel J, Detels R, et al. Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models. Am J Epidemiol. 2003;158:687–94. doi: 10.1093/aje/kwg206.PubMedCrossRefGoogle Scholar
  7. 7.
    Mortimer KM, Neugebauer R, van der Laan MJ, Tager IB. An application of model-fitting procedures for marginal structural models. Am J Epidemiol. 2005;162:382–8. doi: 10.1093/aje/kwi208.PubMedCrossRefGoogle Scholar
  8. 8.
    Brumback BA, Hernán MA, Haneuse SJ, Robins JM. Sensitivity analyses for unmeasured confounding assuming a marginal structural model for repeated measures. Stat Med. 2004;23:749–67. doi: 10.1002/sim.1657.PubMedCrossRefGoogle Scholar
  9. 9.
    Hernán MA, Taubman SL. Does obesity shorten life? The importance of well-defined interventions to answer causal questions. Int J Obes Lond. 2008;32(Suppl 3):S8–14. doi: 10.1038/ijo.2008.82.PubMedCrossRefGoogle Scholar
  10. 10.
    Cole SR, Hernán MA. Constructing inverse probability weights for marginal structural models. Am J Epidemiol. 2008;168:656–64. doi: 10.1093/aje/kwn164.PubMedCrossRefGoogle Scholar
  11. 11.
    Hernán MA, Robins JM. Estimating causal effects from epidemiological data. J Epidemiol Community Health. 2006;60:578–86. doi: 10.1136/jech.2004.029496.PubMedCrossRefGoogle Scholar
  12. 12.
    Bryan J, Yu Z, van der Laan MJ. Analysis of longitudinal marginal structural models. Biostatistics. 2004;5:361–80. doi: 10.1093/biostatistics/kxg041.PubMedCrossRefGoogle Scholar
  13. 13.
    Stürmer T, Rothman KJ, Glynn RJ. Insights into different results from different causal contrasts in the presence of effect-measure modification. Pharmacoepidemiol Drug Saf. 2006;15:698–709. doi: 10.1002/pds.1231.PubMedCrossRefGoogle Scholar
  14. 14.
    Gail MH, Wieand S, Piantadosi S. Biased estimates of treatment effect in randomized experiments with nonlinear regressions and omitted covariates. Biometrika. 1984;7:431–44. doi: 10.1093/biomet/71.3.431.CrossRefGoogle Scholar
  15. 15.
    Austin PC. The performance of different propensity score methods for estimating marginal odds ratios. Stat Med. 2006;26:3078–94. doi: 10.1002/sim.2781.CrossRefGoogle Scholar
  16. 16.
    Austin PC, Grootendorst P, Normand SL, Anderson GM. Conditioning on the propensity score can result in biased estimation of common measures of treatment effect: a Monte Carlo study. Stat Med. 2007;26:754–68. doi: 10.1002/sim.2618.PubMedCrossRefGoogle Scholar
  17. 17.
    Greenland S, Robins JM, Pearl J. Confounding and collapsibility in causal inference. Stat Sci. 1999;14:29–46. doi: 10.1214/ss/1009211805.CrossRefGoogle Scholar
  18. 18.
    Greenland S. Interpretation and choice of effect measures in epidemiologic analyses. Am J Epidemiol. 1987;125:761–8.PubMedGoogle Scholar
  19. 19.
    Newman SC. Commonalities in the classical, collapsibility and counterfactual concepts of confounding. J Clin Epidemiol. 2004;57:325–9. doi: 10.1016/j.jclinepi.2003.07.014.PubMedCrossRefGoogle Scholar
  20. 20.
    Kurth T, Walker AM, Glynn RJ, Chan KA, Gaziano JM, Berger K, et al. Results of multivariable logistic regression, propensity matching, propensity adjustment, and propensity-based weighting under conditions of nonuniform effect. Am J Epidemiol. 2006;163:262–70. doi: 10.1093/aje/kwj047.PubMedCrossRefGoogle Scholar
  21. 21.
    Burton A, Altman DG, Royston P, Holder RL. The design of simulation studies in medical statistics. Stat Med. 2006;25:4279–92. doi: 10.1002/sim.2673.PubMedCrossRefGoogle Scholar
  22. 22.
    Maldonado G, Greenland S. The importance of critically interpreting simulation studies. Epidemiology. 1997;8:453–6.PubMedGoogle Scholar
  23. 23.
    Martens EP, Pestman WR, Klungel OH. Conditioning on the propensity score can result in biased estimation of common measures of treatment effect: a Monte Carlo study (p n/a). Stat Med. 2007;26:3208–10. doi: 10.1002/sim.2878.PubMedCrossRefGoogle Scholar
  24. 24.
    Robins JM, Hernán MA, Rotnitzky A. Invited Commentary: effect modification by time-varying covariates. Am J Epidemiol. 2007;166:994–1002. doi: 10.1093/aje/kwm231.PubMedCrossRefGoogle Scholar
  25. 25.
    Petersen ML, Deeks SG, Martin JN, van der Laan MJ. History-adjusted marginal structural models for estimating time-varying effect modification. Am J Epidemiol. 2007;166:985–93. doi: 10.1093/aje/kwm232.PubMedCrossRefGoogle Scholar
  26. 26.
    Delaney JAC, Daskalopoulou SS, Suissa S. Traditional versus marginal structural models to estimate the effectiveness of β-blocker use on mortality after myocardial infarction. Pharmacoepidemiol Drug Saf. 2008;. doi: 10.1002/pds.1676.PubMedGoogle Scholar
  27. 27.
    Schneeweiss S. Sensitivity analysis and external adjustment for unmeasured confounders in epidemiologic database studies of therapeutics. Pharmacoepidemiol Drug Saf. 2006;15:291–303. doi: 10.1002/pds.1200.PubMedCrossRefGoogle Scholar
  28. 28.
    Greenland S. Basic methods for sensitivity analysis of biases. Int J Epidemiol. 1996;25:1107–16. doi: 10.1093/ije/25.6.1107-a.PubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media B.V. 2009

Authors and Affiliations

  • Joseph A. C. Delaney
    • 1
  • Robert W. Platt
    • 2
  • Samy Suissa
    • 2
  1. 1.Collaborative Health Studies Coordinating Center, Department of BiostatisticsUniversity of WashingtonBldg. 29, Suite 310, 6200 NE 74th StreetSeattleUSA
  2. 2.Department of Epidemiology, Biostatistics and Occupational HealthMcGill UniversityMontrealCanada

Personalised recommendations