Multiply robust generalized estimating equations for cluster randomized trials with missing outcomes

被引:1
作者
Rabideau, Dustin J. [1 ,2 ]
Li, Fan [3 ,4 ]
Wang, Rui [5 ,6 ,7 ]
机构
[1] Massachusetts Gen Hosp, Biostat, 50 Staniford St, Boston, MA 02114 USA
[2] Harvard Med Sch, Dept Med, Boston, MA USA
[3] Yale Univ, Sch Publ Hlth, Dept Biostat, New Haven, CT USA
[4] Yale Univ, Ctr Methods Implementat & Prevent Sci, New Haven, CT USA
[5] Harvard Pilgrim Hlth Care Inst, Dept Populat Med, Boston, MA USA
[6] Harvard Med Sch, Boston, MA USA
[7] Harvard TH Chan Sch Publ Hlth, Dept Biostat, Boston, MA USA
关键词
cluster randomized trial; generalized estimating equation; intracluster correlation coefficient; inverse probability weighting; missing data; multiple robustness; EMPIRICAL LIKELIHOOD; EFFICIENCY; IMPLEMENTATION;
D O I
10.1002/sim.10027
中图分类号
Q [生物科学];
学科分类号
07 ; 0710 ; 09 ;
摘要
Generalized estimating equations (GEEs) provide a useful framework for estimating marginal regression parameters based on data from cluster randomized trials (CRTs), but they can result in inaccurate parameter estimates when some outcomes are informatively missing. Existing techniques to handle missing outcomes in CRTs rely on correct specification of a propensity score model, a covariate-conditional mean outcome model, or require at least one of these two models to be correct, which can be challenging in practice. In this article, we develop new weighted GEEs to simultaneously estimate the marginal mean, scale, and correlation parameters in CRTs with missing outcomes, allowing for multiple propensity score models and multiple covariate-conditional mean models to be specified. The resulting estimators are consistent provided that any one of these models is correct. An iterative algorithm is provided for implementing this more robust estimator and practical considerations for specifying multiple models are discussed. We evaluate the performance of the proposed method through Monte Carlo simulations and apply the proposed multiply robust estimator to analyze the Botswana Combination Prevention Project, a large HIV prevention CRT designed to evaluate whether a combination of HIV-prevention measures can reduce HIV incidence.
引用
收藏
页码:1458 / 1474
页数:17
相关论文
共 50 条
[41]   Sample size estimation for stratified individual and cluster randomized trials with binary outcomes [J].
Kennedy-Shaffer, Lee ;
Hughes, Michael D. .
STATISTICS IN MEDICINE, 2020, 39 (10) :1489-1513
[42]   On the estimation of intracluster correlation for time-to-event outcomes in cluster randomized trials [J].
Kalia, Sumeet ;
Klar, Neil ;
Donner, Allan .
STATISTICS IN MEDICINE, 2016, 35 (30) :5551-5560
[43]   Impact of unequal cluster sizes for GEE analyses of stepped wedge cluster randomized trials with binary outcomes [J].
Tian, Zibo ;
Preisser, John S. ;
Esserman, Denise ;
Turner, Elizabeth L. ;
Rathouz, Paul J. ;
Li, Fan .
BIOMETRICAL JOURNAL, 2022, 64 (03) :419-439
[44]   Assessing treatment effect heterogeneity in the presence of missing effect modifier data in cluster-randomized trials [J].
Blette, Bryan S. ;
Halpern, Scott D. ;
Li, Fan ;
Harhay, Michael O. .
STATISTICAL METHODS IN MEDICAL RESEARCH, 2024, 33 (05) :909-927
[45]   A flexible sample size solution for longitudinal and crossover cluster randomized trials with continuous outcomes [J].
Wang, Jijia ;
Cao, Jing ;
Zhang, Song ;
Ahn, Chul .
CONTEMPORARY CLINICAL TRIALS, 2021, 109
[46]   The Role of a Quadratic Term in Estimating the Average Treatment Effect From Longitudinal Randomized Controlled Trials With Missing Data [J].
Yang, Manshu ;
Wang, Lijuan ;
Maxwell, Scott E. .
PSYCHOLOGICAL METHODS, 2024,
[47]   Sample size calculation for cluster randomized trials with zero-inflated count outcomes [J].
Zhou, Zhengyang ;
Li, Dateng ;
Zhang, Song .
STATISTICS IN MEDICINE, 2022, 41 (12) :2191-2204
[48]   Comparison of methods for estimating the intraclass correlation coefficient for binary responses in cancer prevention cluster randomized trials [J].
Wu, Sheng ;
Crespi, Catherine M. ;
Wong, Weng Kee .
CONTEMPORARY CLINICAL TRIALS, 2012, 33 (05) :869-880
[49]   A pattern-mixture model approach for handling missing continuous outcome data in longitudinal cluster randomized trials [J].
Fiero, Mallorie H. ;
Hsu, Chiu-Hsieh ;
Bell, Melanie L. .
STATISTICS IN MEDICINE, 2017, 36 (26) :4094-4105
[50]   What are the appropriate methods for analyzing patient-reported outcomes in randomized trials when data are missing? [J].
Hamel, J. F. ;
Sebille, V. ;
Le Neel, T. ;
Kubis, G. ;
Boyer, F. C. ;
Hardouin, J. B. .
STATISTICAL METHODS IN MEDICAL RESEARCH, 2017, 26 (06) :2897-2908