Optimal Permutation Tests for the Analysis of Group Randomized Trials

Two facts complicate the comparison of interventions in group randomized trials (GRT's), a family of clinical trials in which each member of a particular group receives the same treatment assignment. First, individual outcomes within each group are often correlated. Second, the number of groups in a GRT is often not sufficient to make asymptotic approximations possible. Therefore, the tests used with methods, such as generalized estimating equations (GEE) and penalized quasi likelihood (PQL), originally developed for longitudinal studies, may not be valid. As an alternative, a class of permutation tests is derived to maximize the power of testing for an intervention effect in a GRT while maintaining a nominal test size. The test uses a statistic that is a weighted sum of residuals, with the weights based on the group sizes and the variability of each individual outcome. Through simulation, we demonstrate the importance of weights to a permutation test's power and compare the power of permutation tests, GEE, and PQL. Last, we apply our methods to an actual GRT to study smoking cessation, discuss the findings based on permutation tests, and compare those findings with traditional methods.

[1]  R J Carroll,et al.  On design considerations and randomization-based inference for community intervention trials. , 1996, Statistics in medicine.

[2]  C Probart,et al.  Work site-based cancer prevention: primary results from the Working Well Trial. , 1996, American journal of public health.

[3]  N. Jewell,et al.  Hypothesis testing of regression parameters in semiparametric generalized linear models for cluster correlated data , 1990 .

[4]  S. Zeger,et al.  Longitudinal data analysis using generalized linear models , 1986 .

[5]  A. Alesina,et al.  Inequality and Happiness: Are Europeans and Americans Different? , 2001 .

[6]  D. Firth,et al.  Estimating Intraclass Correlation for Binary Data , 1999, Biometrics.

[7]  D. Pierce.  The Asymptotic Effect of Substituting Estimators for Parameters in Certain Types of Statistics , 1982 .

[8]  Joseph P. Romano.  On the behaviour of randomization tests without the group invariance assumption , 1990 .

[9]  Scott L. Zeger,et al.  [Inference Based on Estimating Functions in the Presence of Nuisance Parameters]: Rejoinder , 1995 .

[10]  Bing Li,et al.  A deviance function for the quasi-likelihood method , 1993 .

[11]  S G Thompson,et al.  The design and analysis of paired cluster randomized trials: an application of meta-analysis techniques. , 1997, Statistics in medicine.

[12]  R Brookmeyer,et al.  Person-time analysis of paired community intervention trials when the number of communities is small. , 1998, Statistics in medicine.

[13]  P. Sen,et al.  Nonparametric methods in multivariate analysis , 1974 .

[14]  Charles Stein,et al.  On the Theory of Some Non-Parametric Hypotheses , 1949 .

[15]  J. Kalbfleisch,et al.  The estimating function bootstrap , 2000 .

[16]  Ronald H. Randles,et al.  On the Asymptotic Normality of Statistics with Estimated Parameters , 1982 .

[17]  M. Piedmonte,et al.  On some small sample properties of generalized estimating equationEstimates for multivariate dichotomous outcomes , 1992 .

[18]  E. B. Manoukian.  Modern concepts and theorems of mathematical statistics , 1985 .

[19]  C. Heyde,et al.  Quasi-likelihood and its application , 1997 .

[20]  Asymptotic Behavior of Linear Permutation Tests under General Alternatives, with Application to Test Selection and Study Design , 2000 .

[21]  Z. Feng,et al.  A comparison of statistical methods for clustered data analysis with Gaussian error. , 1996, Statistics in medicine.

[22]  R. W. Wedderburn.  Quasi-likelihood functions, generalized linear models, and the Gauss-Newton method , 1974 .

[23]  N. Breslow,et al.  Bias correction in generalised linear mixed models with a single component of dispersion , 1995 .

[24]  N. Breslow,et al.  Approximate inference in generalized linear mixed models , 1993 .

[25]  D B Abrams,et al.  Cancer control at the workplace: the Working Well Trial. , 1994, Preventive medicine.

[26]  D. Abrams,et al.  The Working Well Trial: baseline dietary and smoking behaviors of employees and related worksite characteristics. The Working Well Research Group. , 1995, Preventive medicine.

[27]  Mitchell H. Gail,et al.  Community Intervention Trial for Smoking Cessation (COMMIT): I. cohort results from a four-year community intervention. , 1995, American journal of public health.

[28]  Z Feng,et al.  Correlated binomial variates: properties of estimator of intraclass correlation and its effect on sample size calculation. , 1992, Statistics in medicine.