A comparison of methods for estimating the causal effect of a treatment in randomized clinical trials subject to noncompliance

Biometrics. 2009 Jun;65(2):640-9. doi: 10.1111/j.1541-0420.2008.01066.x. Epub 2008 May 28.


We consider the analysis of clinical trials that involve randomization to an active treatment (T = 1) or a control treatment (T = 0), when the active treatment is subject to all-or-nothing compliance. We compare three approaches to estimating treatment efficacy in this situation: as-treated analysis, per-protocol analysis, and instrumental variable (IV) estimation, where the treatment effect is estimated using the randomization indicator as an IV. Both model- and method-of-moment based IV estimators are considered. The assumptions underlying these estimators are assessed, standard errors and mean squared errors of the estimates are compared, and design implications of the three methods are examined. Extensions of the methods to include observed covariates are then discussed, emphasizing the role of compliance propensity methods and the contrasting role of covariates in these extensions. Methods are illustrated on data from the Women Take Pride study, an assessment of behavioral treatments for women with heart disease.

Publication types

  • Comparative Study
  • Evaluation Study
  • Research Support, N.I.H., Extramural

MeSH terms

  • Algorithms
  • Biometry / methods*
  • Causality*
  • Computer Simulation
  • Data Interpretation, Statistical*
  • Epidemiologic Research Design*
  • Models, Statistical
  • Outcome Assessment, Health Care / methods*
  • Patient Compliance / statistics & numerical data*
  • Proportional Hazards Models
  • Randomized Controlled Trials as Topic / methods*