Structural Nested Models for Cluster-Randomized Trials

In clinical trials and epidemiologic studies, adherence to the assigned components is not always perfect. In this book chapter, we are interested in estimating the causal effect of cluster-level adherence on an individual-level outcome. Two different methodologies will be provided, based on ordinary and weighted structural nested models (SNMs). We also applied the jackknife to construct confidence intervals. The computation is straightforward with application of instrumental variables software, and the programming schemes are developed for both ordinary and weighted structural nested models. Simulation studies under ordinary structural nested models with different link functions (loglinear SNM, logistic SNM, and linear SNM) were conducted to validate our methods. We then applied the methods to a school-based water, sanitation, and hygiene study to estimate the causal effect of increased adherence to intervention components on student absenteeism. The results calculated from these two methodologies are quite close.

[1]  B. Brumback,et al.  Assessing the impact of a school‐based water treatment, hygiene and sanitation programme on pupil absence in Nyanza Province, Kenya: a cluster‐randomized trial , 2011, Tropical medicine & international health : TM & IH.

[2]  Marshall Joffe,et al.  Causal logistic models for non‐compliance under randomized treatment with univariate binary response , 2003, Statistics in medicine.

[3]  Stephen Burgess,et al.  Improving bias and coverage in instrumental variable analysis with weak instruments for continuous and binary outcomes , 2012, Statistics in medicine.

[4]  Joshua D. Angrist,et al.  Identification of Causal Effects Using Instrumental Variables , 1993 .

[5]  Tihomir Asparouhov,et al.  Intention‐to‐treat analysis in cluster randomized trials with noncompliance , 2008, Statistics in medicine.

[6]  P Gustafson,et al.  Use of instrumental variables in the analysis of generalized linear models in the presence of unmeasured confounding with applications to epidemiological research , 2008, Statistics in medicine.

[7]  S. Vansteelandt,et al.  Sense and sensitivity when correcting for observed exposures in randomized clinical trials , 2005, Statistics in medicine.

[8]  J M Robins,et al.  Correction for non-compliance in equivalence trials. , 1998, Statistics in medicine.

[9]  M Alan Brookhart,et al.  American Journal of Epidemiology Practice of Epidemiology Instrumental Variable Analysis for Estimation of Treatment Effects with Dichotomous Outcomes , 2022 .

[10]  Dylan S. Small,et al.  Bounds on causal effects in three‐arm trials with non‐compliance , 2006 .

[11]  Dylan S Small,et al.  Random effects logistic models for analysing efficacy of a longitudinal randomized treatment with non‐adherence , 2006, Statistics in medicine.

[12]  S. Burgess.  Identifying the odds ratio estimated by a two‐stage instrumental variable analysis with a logistic regression model , 2013, Statistics in medicine.

[13]  J. Robins.  Correcting for non-compliance in randomized trials using structural nested mean models , 1994 .

[14]  Elizabeth A Stuart,et al.  On the use of propensity scores in principal causal effect estimation , 2009, Statistics in medicine.

[15]  Jay Bhattacharya,et al.  Estimating probit models with self‐selected treatments , 2006, Statistics in medicine.

[16]  J. Robins,et al.  Instruments for Causal Inference: An Epidemiologist's Dream? , 2006, Epidemiology.

[17]  N. Nagelkerke,et al.  Estimating treatment effects in randomized clinical trials in the presence of non-compliance. , 2000, Statistics in medicine.

[18]  D. Rubin,et al.  Principal Stratification in Causal Inference , 2002, Biometrics.

[19]  Xihong Lin,et al.  Estimating causal effects in trials involving multitreatment arms subject to non‐compliance: a Bayesian framework , 2010, Journal of the Royal Statistical Society. Series C, Applied statistics.

[20]  J. Roy,et al.  Causal models for randomized trials with two active treatments and continuous compliance , 2011, Statistics in medicine.

[21]  Stijn Vansteelandt,et al.  Causal inference with generalized structural mean models , 2003 .

[22]  S. Vansteelandt,et al.  On Instrumental Variables Estimation of Causal Odds Ratios , 2011, 1201.2487.

[23]  Babette A Brumback,et al.  Using structural‐nested models to estimate the effect of cluster‐level adherence on individual‐level outcomes with a three‐armed cluster‐randomized trial , 2014, Statistics in medicine.

[24]  Dylan S Small,et al.  Two‐stage instrumental variable methods for estimating the causal odds ratio: Analysis of bias , 2011, Statistics in medicine.

[25]  N M Laird,et al.  Correcting for non-compliance in randomized trials: an application to the ATBC Study. , 1999, Statistics in medicine.

[26]  J. Albert.  Estimating efficacy in clinical trials with clustered binary responses , 2002, Statistics in medicine.

[27]  S. Greenland.  An introduction To instrumental variables for epidemiologists , 2000, International journal of epidemiology.

[28]  Jeffrey M. Wooldridge,et al.  Solutions Manual and Supplementary Materials for Econometric Analysis of Cross Section and Panel Data , 2003 .

[29]  Marshall M Joffe,et al.  Weighting in instrumental variables and G‐estimation , 2003, Statistics in medicine.

[30]  Tihomir Asparouhov,et al.  Cluster randomized trials with treatment noncompliance. , 2008, Psychological methods.