Deviations from the population-averaged versus cluster-specific relationship for clustered binary data

Abstract

There has been much debate about the relative merits of mixed effects and population-averaged logistic models. We present a different perspective on this issue by noting that the investigation of the relationship between these models for a given dataset offers a type of sensitivity analysis that may reveal problems with assumptions of the mixed effects and/or population-averaged models for clustered binary response data in general and longitudinal binary outcomes in particular. We present several datasets in which the following violations of assumptions are associated with departures from the expected theoretical relationship between these two models: 1) negative intra-cluster correlations; 2) confounding of the response-covariate relationship by cluster effects; and 3) confounding of autoregressive relationships by the link between baseline outcomes and subject effects. Under each of these conditions, the expected theoretical attenuation of the population-averaged odds ratio relative to the cluster-specific odds ratio does not necessarily occur. In all cases, the naive fltting of a random intercept logistic model appears to lead to bias. In response, the random intercept model is modified to accommodate negative intra-cluster correlations, confounding due to clusters, or baseline correlations with random effects. Comparisons are made with GEE estimation of population-averaged models and conditional likelihood estimation of cluster-specific models. Several examples, including a cross-over trial, a multicentre nonrandomized treatment study, and a longitudinal observational study are used to illustrate these modiflcations.

Get full access to this article

View all access options for this article.

References

Williams D. The analysis of binary responses from toxicological experiments involving reproduction and teratogenicity . Biometrics 1975; 31: 949-952 .

Breslow N , Day N. Statistical methods in cancer research, Vol. I. Lyon: International Agency for Research on Cancer (IARC) Scientific Publications No. 32, 1980.

Stiratelli R , Liard N , Ware, J. Random-effects models for serial observations with binary response . Biometrics 1984; 40: 961-971 .

Liang K-Y , Zeger S. Longitudinal data analysis using generalized linear models . Biometrika 1986; 73: 13-22 .

Connolly M , Liang K-Y. Conditional logistic regression models for correlated binary data . Biometrika 1988; 75: 501-506 .

Ten Have T , Reboussin B , Miller M , Kunselman A. Mixed effects logistic regression models for multiple longitudinal binary functional limitation responses with informative drop-out and confounding by baseline outcomes . Biometrics 2002; 58: 137-144 .

Becker M , Yang I. Latent class marginal models for cross-classifications of counts . Sociological Methodology 1998; 28: 293-325 .

Sammel M , Ryan L , Legler J. Latent variable models for mixed discrete and continuous outcomes . Journal of the Royal Statistical Society, Series B 1997; 59: 667-678 .

Reboussin B , Liang K-Y. An estimating equations approach for the LISCOMP model . Psychometrika 1998; 6: 165-182 .

10.

Reboussin B , Liang K-Y , Reboussin D. Estimating for a latent transition model with multiple discrete indicators . Biometrics 2000; 55: 839-845 .

11.

Zeger S , Liang K-Y , Albert P. Models for longitudinal data: a generalized estimating equation approach . Biometrics 1988; 44: 1049-1060 .

12.

Neuhaus J , Kalbfleisch J , Hauck W. A comparison of cluster-specific and population-averaged approaches for analyzing correlated binary data . International Statistical Review 1991; 59: 25-35 .

13.

Graubard B , Korn E. Regression analysis with clustered data . Statistics in Medicine 1994; 13: 509-522 .

14.

Ten Have T , Kunselman A , Zharichenko E. Accommodating negative intracluster correlation in mixed effects models for clustered binary response data . Journal of Biopharmaceutical Statistics 1998; 8: 131-149 .

15.

Coull B , Agresti A. The use of mixed logit models to reflect heterogeneity in capture-recapture studies . Biometrics 1999; 55: 291-301 .

16.

Berlin J , Kimmel S , Ten Have T , Sammel M. An empiricial comparison of several clustered data approaches under confounding due to cluster effects in the analysis of complications of coronary angioplasty . Biometrics 1999; 55: 470-476 .

17.

Neuhaus J , Kalbfleisch J. Between- and within-cluster covariate effects in the analysis of clustered data . Biometrics 1998; 54: 638-645 .

18.

Ten Have T , Landis J , Weaver S. Association models for periodontal disease progression: a comparison of methods for clustered binary data . Statistics in Medicine 1995; 14: 413-430 .

19.

Crouchley R , Davies R. A comparison of population-average and random effects models for the analysis of longitudinal count data with baseline information . Journal of the Royal Statistcal Society, Series A 1999; 162: 331-347 .

20.

Mauritsen R. EGRET reference manual. Seattle: Statistics and Epidemiology Research Corporation (SERC) , 1990.

21.

Ten Have T , Pulkstenis E , Kunselman A , Landis J. Mixed effects logistic regression models for longitudinal binary response data with informative dropout . Biometrics 1998; 54: 367-383 .

22.

Lehmann E. Some concepts of dependence . Annals of Mathematical Statistics 1966; 37: 1137-1153 .