Skip to main content
SpringerLink
Log in

Estimating Effects of Dynamic Treatment Strategies in Pharmacoepidemiologic Studies with Time-Varying Confounding: a Primer

  • Pharmacoepidemiology (S Toh, Section Editor)
  • Published:
Current Epidemiology Reports Aims and scope Submit manuscript

Abstract

Purpose of Review

Pharmacoepidemiologists are often interested in estimating the effects of dynamic treatment strategies, where treatments are modified based on patients’ evolving characteristics. For such problems, appropriate control of both baseline and time-varying confounders is critical. Conventional methods that control confounding by including time-varying treatments and confounders in an outcome regression model may not have a causal interpretation, even when all baseline and time-varying confounders are measured. This problem occurs when time-varying confounders are, themselves, affected by past treatment. We review alternative analytic approaches that can produce valid inferences in the presence of such confounding. We focus on the parametric g-formula and inverse probability weighting of marginal structural models, two examples of Robins’ g-methods.

Recent Findings

Unlike standard outcome regression methods, the parametric g-formula and inverse probability weighting of marginal structural models can estimate the effects of dynamic treatment strategies and appropriately control for measured time-varying confounders affected by prior treatment. Few applications of g-methods exist in the pharmacoepidemiology literature, primarily due to the common use of administrative claims data, which typically lack detailed measurements of time-varying information, and the limited availability of or familiarity with tools to help perform the relatively complex analysis. These barriers may be overcome with the increasing availability of data sources containing more detailed time-varying information and more accessible learning tools and software.

Summary

With appropriate data and study design, g-methods can improve our ability to make causal inferences on dynamic treatment strategies from observational data in pharmacoepidemiology.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2

Similar content being viewed by others

References

Papers of particular interest, published recently have been highlighted as: • Of importance •• Of major importance

  1. Ray WA. Evaluating medication effects outside of clinical trials: new-user designs. Am J Epidemiol. 2003;158(9):915–20.

    Article  PubMed  Google Scholar 

  2. Johnson ES, Bartman BA, Briesacher BA, Fleming NS, Gerhard T, Kornegay CJ, et al. The incident user design in comparative effectiveness research. Pharmacoepidemiol Drug Saf. 2013;22(1):1–6.

    Article  PubMed  Google Scholar 

  3. Robins J. A new approach to causal inference in mortality studies with a sustained exposure period—application to the healthy worker survivor effect. Math Model. 1986;7(9–12):1393–512.

    Article  Google Scholar 

  4. Murphy SA, van der Laan MJ, Robins JM. Marginal mean models for dynamic regimes. J Am Stat Assoc. 2001;96(456):1410–23.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Macdougall IC, Bircher AJ, Eckardt K-U, Obrador GT, Pollock CA, Stenvinkel P, et al. Iron management in chronic kidney disease: conclusions from a “Kidney Disease: Improving Global Outcomes” (KDIGO) Controversies Conference. Kidney Int. 2016;89(1):28–39.

    Article  CAS  PubMed  Google Scholar 

  6. Kidney Disease: Improving Global Outcomes (KDIGO) Anemia Work Group. KDIGO clinical practice guideline for anemia in chronic kidney disease. Kidney Int Suppl. 2012;2:279–335.

  7. Locatelli F, Bárány P, Covic A, De Francisco A, Del Vecchio L, Goldsmith D, et al. Kidney disease: improving global outcomes guidelines on anaemia management in chronic kidney disease: a European Renal Best Practice position statement. Nephrol Dial Transplant. 2013;28(6):1346–59.

    Article  CAS  PubMed  Google Scholar 

  8. Kliger AS, Foley RN, Goldfarb DS, Goldstein SL, Johansen K, Singh A, et al. KDOQI US commentary on the 2012 KDIGO clinical practice guideline for anemia in CKD. Am J Kidney Dis. 2013;62(5):849–59.

    Article  PubMed  Google Scholar 

  9. Moist LM, Troyanov S, White CT, Wazny LD, Wilson J-A, McFarlane P, et al. Canadian society of nephrology commentary on the 2012 KDIGO clinical practice guideline for anemia in CKD. Am J Kidney Dis. 2013;62(5):860–73.

    Article  PubMed  Google Scholar 

  10. Krishnan M, Weldon J, Wilson S, Goyhkman I, Van Wyck D. Effect of maintenance iron protocols on ESA dosing and anemia outcomes [Abstract 153]. Am J Kidney Dis. 2011;57(4):B55.

    Article  Google Scholar 

  11. Miskulin DC, Tangri N, Bandeen-Roche K, Zhou J, McDermott A, Meyer KB, et al; Developing Evidence to Inform Decisions about Effectiveness (DEcIDE) Network Patient Outcomes in End Stage Renal Disease Study Investigators: intravenous iron exposure and mortality in patients on hemodialysis. Clin J Am Soc Nephrol 2014;9(11):1930–1939.

  12. Schiller B. Implementing an IV iron administration protocol within a dialysis organization. 2014. Available at http://www.nephrologynews.com/implementing-an-iv-iron-administration-protocol-within-a-dialysis-organization/.

  13. Li X, Kshirsagar AV, Brookhart MA. Safety of intravenous iron in hemodialysis patients. Hemodial Int. 2017;21(Suppl 1):S93–103.

    Article  PubMed  Google Scholar 

  14. • Cotton CA, Heagerty PJ. Evaluating epoetin dosing strategies using observational longitudinal data. Ann Appl Stat. 2014;8(4):2356–77. Describes an inverse probability weighting of marginal structural models method to compare multiple dynamic treatment strategies using electronic healthcare databases.

    Article  Google Scholar 

  15. • Zhang Y, Thamer M, Kaufman J, Cotter D, Herńan MA. Comparative effectiveness of two anemia management strategies for complex elderly dialysis patients. Medical Care. 2014;52(Suppl 3):S132–9. Describes an inverse probability weighting approach to compare multiple dynamic treatment strategies using electronic healthcare databases.

    Article  PubMed  PubMed Central  Google Scholar 

  16. • Zhang Y, Young JG, Thamer M, Hernán MA. Comparing the effectiveness of dynamic treatment strategies using electronic health records: an application of the parametric g-formula to anemia management strategies. Health Serv Res. 2017; https://doi.org/10.1111/1475–6773.12718. Describes a parametric g-formula approach to compare multiple dynamic treatment strategies using electronic healthcare databases.

  17. • Li X. Comparative effectiveness of intravenous iron treatment protocols in hemodialysis patients: causal inference with dynamic treatment regimes [PhD Dissertation]. Chapel Hill (NC): University of North Carolina at Chapel Hill; 2017. Describes an approach to identify dynamic treatment strategies that can be realistically supported with the data and an application of inverse probability weighting approach to compare multiple dynamic strategies using electronic healthcare databases.

  18. •• Neugebauer R, Fireman B, Roy JA, O'Connor PJ, Selby JV. Dynamic marginal structural modeling to evaluate the comparative effectiveness of more or less aggressive treatment intensification strategies in adults with type 2 diabetes. Pharmacoepidemiol Drug Saf. 2012;21:99–113. Describes a dynamic marginal structural model approach to compare multiple dynamic treatment strategies using electronic healthcare databases.

    Article  PubMed  Google Scholar 

  19. Neugebauer R, Fireman B, Roy JA, O’Connor PJ. Impact of specific glucose-control strategies on microvascular and macrovascular outcomes in 58,000 adults with type 2 diabetes. Diabetes Care. 2013; https://doi.org/10.2337/dc12-2675.

  20. •• Cain LE, Robins JM, Lanoy E, Logan R, Costagliola D, Hernán MA. When to start treatment? A systematic approach to the comparison of dynamic regimes using observational data. Int J Biostat. 2010;6(2):Article 18. Describes an approach to compare dynamic strategies by emulating randomized trials using observational data and discusses some complications that arise in such practice.

    Article  PubMed  Google Scholar 

  21. Cain LE, Logan R, Robins JM, Sterne JAC, Sabin C, et al. On behalf of the HIV-CAUSAL collaboration. When to initiate combined antiretroviral therapy to reduce rates of mortality and AIDS in HIV-infected individuals in developed countries. Ann Intern Med. 2011;154(8):509–15.

    Article  PubMed  Google Scholar 

  22. •• Hernán MA, Lanoy E, Costagliola D, Robins JM. Comparison of dynamic treatment regimes via inverse probability weighting. Basic Clin Pharmacol Toxicol. 2006;98(3):237–42. Describes a simple inverse probability weighting approach to compare two dynamic treatment strategies.

    Article  PubMed  Google Scholar 

  23. •• Young JG, Cain LE, Robins JM, O’Reilly EJ, Hernán MA. Comparative effectiveness of dynamic treatment regimes: an application of the parametric g-formula. Stat Biosci. 2011;3(1):119–43. Describes a parametric g-formula approach to compare dynamic treatment strategies.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Toh S, Hernández-Diaz S, Logan R, Robins JM, Hernán MA. Estimating absolute risks in the presence of nonadherence: an application to a follow-up study with baseline randomization. Epidemiology. 2010;21(4):528–39.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Wang L, Rotnitzky A, Lin X, Millikan RE, Thall PF. Evaluation of viable dynamic treatment regimes in a sequentially randomized trial of advanced prostate cancer. J Am Stat Assoc. 2012;107(498):493–508.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Akdemir G, Markusse IM, Dirven L, Riyazi N, Steup-Beekman GM, Kerstens PJSM, et al. Effectiveness of four dynamic treatment strategies in patients with anticitrullinated protein antibody-negative rheumatoid arthritis: a randomised trial. RMD Open. 2016;2:e000143. https://doi.org/10.1136/rmdopen-2015-000143.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Robins JM, Hernán MA, Siebert U. Effects of multiple interventions. In: Ezzati M, Lopez AD, Rodgers A, Murray CJL, editors. Comparative quantification of health risks: global and regional burden of disease attributable to selected major risk factors. Geneva: World Health Organization; 2004.

    Google Scholar 

  28. Taubman SL, Robins JM, Mittleman MA, Hernán MA. Intervening on risk factors for coronary heart disease: an application of the parametric g-formula. Int J Epidemiol. 2009;38(6):1599–611.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Lajous M, Willett WC, Robins J, Young JG, Rimm E, Mozaffarian D, et al. Changes in fish consumption in midlife and the risk of coronary heart disease in men and women. Am J Epidemiol. 2013;178(3):382–91.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Danaei G, Pan A, Hu FB, Hernán MA. Hypothetical lifestyle interventions in middle-aged women and risk of type 2 diabetes: a 24-year prospective study. Epidemiology. 2013;24:122–8.

    Article  PubMed  PubMed Central  Google Scholar 

  31. García-Aymerich J, Varraso R, Danaei G, Camargo CA, Hernán MA. Incidence of adult-onset asthma after hypothetical interventions on body mass index and physical activity: an application of the parametric g-formula. Am J Epidemiol. 2014;179(1):20–6.

    Article  PubMed  Google Scholar 

  32. Jain P, Danaei G, Robins JM, Manson JE, Herńan MA. Smoking cessation and long-term weight gain in the Framingham Heart Study: an application of the parametric g-formula for a continuous outcome. Eur J Epidemiol. 2016;31(12):1223–9.

    Article  PubMed  Google Scholar 

  33. Wagner EH, Austin BT, Davis C, Hindmarsh M, Schaefer J, Bonomi A. Improving chronic illness care: translating evidence into action. Health Aff. 2001;20(6):64–78.

    Article  CAS  Google Scholar 

  34. Robins J. A graphical approach to the identification and estimation of causal parameters in mortality studies with sustained exposure periods. J Chronic Dis. 1987;40(Suppl 2):S139–61.

    Article  Google Scholar 

  35. •• Robins JM, Hernán MA. Estimation of the causal effects of time-varying exposures. In longitudinal data analysis. In: Fitzmaurice G, Davidian M, Verbeke G, Molenberghs G, editors. . New York: Chapman and Hall/CRC Press; 2009. p. 553–99. Introduces dynamic treatment strategies and methods to estimating effects of these strategies using observational data.

    Google Scholar 

  36. Robins JM, Hernán MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology. 2000;11(5):550–60.

    Article  CAS  PubMed  Google Scholar 

  37. 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(5):561–70.

    Article  PubMed  Google Scholar 

  38. Robins JM. Marginal structural models, 1997 proceedings of the American Statistical Association, Section on Bayesian Statistical Science, 1998 Alexandria, VA American Statistical Association, pg. 1–10.

  39. Pearl J. Causal diagrams for empirical research. Biometrika. 1995;82(4):669–88.

    Article  Google Scholar 

  40. Pearl J. Causality: models, reasoning, and inference. 2nd ed. New York: Cambridge University Press; 2009.

    Book  Google Scholar 

  41. Hernán MA, Hernández-Díaz S, Robins JM. A structural approach to selection bias. Epidemiology. 2004;15(5):615–25.

    Article  PubMed  Google Scholar 

  42. VanderWeele TJ. Concerning the consistency assumption in causal inference. Epidemiology. 2009;20(6):880–3.

    Article  PubMed  Google Scholar 

  43. Orellana L, Rotnitzky A, Robins JM. Dynamic regime marginal structural mean models for estimation of optimal dynamic treatment regimes, Part I: Main content. Int J Biostat. 2010;6(2):Article 8.

    PubMed  Google Scholar 

  44. Bang H, Robins JM. Doubly robust estimation in missing data and causal inference models. Biometrics. 2005;61(4):962–73.

    Article  PubMed  Google Scholar 

  45. van der Laan MJ. Targeted maximum likelihood based causal inference: Part I. Int J Biostat. 2010;6(2):Article 2.

    PubMed  Google Scholar 

  46. van der Laan MJ. Targeted maximum likelihood based causal inference: Part II. Int J Biostat. 2010;6(2):Article 3.

    PubMed  Google Scholar 

  47. van der Laan MJ, Gruber S. Targeted minimum loss based estimation of causal effects of multiple time point interventions. Int J Biostat. 2012;8(1):Article 9.

    Google Scholar 

  48. • Petersen M, Schwab J, Gruber S, Blaser N, Schomaker M, van der Laan M. Targeted maximum likelihood estimation for dynamic and static longitudinal marginal structural working models. J Causal Inference. 2014;2(2):147–85. Describes how to estimate marginal structural models for dynamic treatment strategies using doubly-robust methods.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Robins JM, Rotnitzky A. Comment on the Bickel and Kwon article, “Inference for semiparametric models: some questions and an answer”. Stat Sin. 2001;11(4):920–36. [“On double robustness”]

    Google Scholar 

  50. Robins JM, Wasserman L. Estimation of effects of sequential treatments by reparameterizing directed acyclic graphs. In: Geiger D, Shenoy P, editors. Proceedings of the thirteenth conference on uncertainty in artificial intelligence, providence Rhode Island. San Francisco: Morgan Kaufmann; 1997. p. 409–20.

    Google Scholar 

  51. Young JG, Tchetgen Tchetgen EJ. Simulation from a known Cox MSM using standard parametric models for the g-formula. Stat Med. 2014;33(6):1001–14.

    Article  PubMed  Google Scholar 

  52. Orellana L, Rotnitzky A, Robins JM. Dynamic regime marginal structural mean models for estimation of optimal dynamic treatment regimes, Part II: Proofs and Additional Results. Int J Biostat. 2010;6:Article 8.

    PubMed  Google Scholar 

  53. van der Laan MJ, Petersen ML. Causal effect models for realistic individualized treatment and intention to treat rules. Int J Biostat. 2007;3(1):Article 3.

    PubMed  Google Scholar 

  54. Robins J, Orellana L, Rotnitzky A. Estimation and extrapolation of optimal treatment and testing strategies. Stat Med. 2008;27(23):4678–721.

    Article  PubMed  Google Scholar 

  55. Cotton CA, Heagerty PJ. A data augmentation method for estimating the causal effect of adherence to treatment regimens targeting control of an intermediate measure. Stat Biosci. 2011;3(1):28–44.

    Article  Google Scholar 

  56. Shortreed SM, Moodie EEM. Estimating the optimal dynamic antipsychotic treatment regime: evidence from the sequential multiple-assignment randomized clinical antipsychotic trials of intervention and effectiveness schizophrenia study. J R Stat Soc Ser C Appl Stat. 2012;61(4):577–99.

    Article  PubMed  PubMed Central  Google Scholar 

  57. Orellana L, Rotnitzky AG, Robins JM. Generalized marginal structural models for estimating optimal treatment regimes. Technical Report, Department of Biostatistics, Harvard School of Public Health. 2006.

  58. Munoz ID, van der Laan M. Population intervention causal effects based on stochastic interventions. Biometrics. 2012;68(2):541–9.

    Article  PubMed  Google Scholar 

  59. Haneuse S, Rotnitzky A. Estimation of the effect of interventions that modify the received treatment. Stat Med. 2013;32(30):5260–77.

    Article  CAS  PubMed  Google Scholar 

  60. Young JG, Herńan MA, Robins JM. Identification, estimation and approximation of risk under interventions that depend on the natural value of treatment using observational data. Epidemiol Meth. 2014;3(1):1–19.

    Article  Google Scholar 

  61. Cole SR, Herńan MA. Constructing inverse probability weights for marginal structural models. Am J Epidemiol. 2008;168(6):656–64.

    Article  PubMed  PubMed Central  Google Scholar 

  62. Petersen ML, Porter KE, Gruber S, Wang Y, van der Laan MJ. Diagnosing and responding to violations in the positivity assumption. Stat Methods Med Res. 2012;21(1):31–54.

    Article  PubMed  Google Scholar 

  63. Schneeweiss S, Avorn J. A review of uses of health care utilization databases for epidemiologic research on therapeutics. J Clin Epidemiol. 2005;58(4):323–37.

    Article  PubMed  Google Scholar 

  64. Murray MD. Use of data from electronic health records for pharmacoepidemiology. Curr Epidemiol Rep. 2014;1(4):186–93.

    Article  Google Scholar 

  65. Foley RN, Collins AJ. The USRDS: what you need to know about what it can and can’t tell us about ESRD. Clin J Am Soc Nephrol. 2013;8(5):845–51.

    Article  PubMed  Google Scholar 

  66. Schmidt M, Schmidt SA, Sandegaard JL, Ehrenstein V, Pedersen L, Sorensen HT. The Danish National Patient Registry: a review of content, data quality, and research potential. Clin Epidemiol. 2015;7:449–90.

    Article  PubMed  PubMed Central  Google Scholar 

  67. Andrade SE, Bérard A, Nordeng HME, Wood ME, van Gelder MMHJ, Toh S. Administrative claims data versus augmented pregnancy data for the study of pharmaceutical treatments in pregnancy. Curr Epidemiol Rep. 2017;4(2):106–16.

    Article  Google Scholar 

  68. Logan RW, Young JG, Taubman S, Lodi S, Picciotto S, Danaei G, et al. The GFORMULA SAS macro 2017. https://www.hsph.harvard.edu/causal/software./

  69. Lendle S, Schwab J, Petersen ML, van der Laan MJ. Ltmle: an R package implementing targeted minimum loss-based estimation for longitudinal data. J Stat Softw. 2016.

  70. Daniel RM, De Stavola BL, Cousens SN. gformula: estimating causal effects in the presence of time-varying confounding or mediation using the g-computation formula. Stata J. 2011;11(4):479–517.

    Google Scholar 

  71. •• Daniel RM, Cousens SN, De Stavola BL, Kenward MG, Sterne JAC. Methods for dealing with time-dependent confounding. Stat Med. 2013;32(9):1584–618. Tutorial for g-methods.

    Article  CAS  PubMed  Google Scholar 

  72. •• Causal Inference Methods for PCOR using Observational Data Conference. CIMPODS 2017. Retrieved from http://www.cimpod2017.org. Includes tutorials on g-methods for dynamic treatment strategies.

  73. •• Comparative effectiveness research based on observational data to emulate a target trial (CERBOT). Retrieved from http://cerbot.org/. Includes tutorials on g-methods for dynamic treatment strategies.

  74. •• Hernán MA, Robins JM. Using big data to emulate a target trial when a randomized trial is not available. Am J Epidemiol. 2016;183(8):758–64. Outlines a framework to obtain causal effect in comparative effectiveness research using observational data.

    Article  PubMed  PubMed Central  Google Scholar 

  75. Hernán MA, McAdams M, McGrath N, Lanoy E, Costagliola D. Observation plans in longitudinal studies with time-varying treatments. Stat Methods Med Res. 2009;18(1):27–52.

    Article  PubMed  Google Scholar 

  76. Li X, Cole SR, Kshirsagar AV, Fine J, Stürmer T, Brookhart MA. Identification of dynamic treatment regimes in complex longitudinal data.[unpublished, revise and resubmit].

  77. Brookhart MA. Counterpoint: the treatment decision design. Am J Epidemiol. 2015;182(10):840–5.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, 401 Park Drive, Suite 401 East, Boston, MA, 02215, USA

    Xiaojuan Li, Jessica G. Young & Sengwee Toh

Authors
  1. Xiaojuan Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jessica G. Young
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Sengwee Toh
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Xiaojuan Li.

Ethics declarations

Conflict of Interest

Xiaojuan Li is supported by a Thomas O. Pyle Fellowship from the Harvard Medical School & Harvard Pilgrim Health Care Institute.

Jessica G. Young is supported by a faculty grant from the Harvard Pilgrim Health Care Institute.

Sengwee Toh is partially supported by the National Institute of Biomedical Imaging and Bioengineering (U01EB023683).

Human and Animal Rights and Informed Consent

This article does not contain any studies with human or animal subjects performed by the authors.

Additional information

This article is part of the Topical Collection on Pharmacoepidemiology

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Li, X., Young, J.G. & Toh, S. Estimating Effects of Dynamic Treatment Strategies in Pharmacoepidemiologic Studies with Time-Varying Confounding: a Primer. Curr Epidemiol Rep 4, 288–297 (2017). https://doi.org/10.1007/s40471-017-0124-x

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s40471-017-0124-x

Keywords

Search

Navigation