Testing equality of risk ratios in multiple 2×2 tables with structural zero

This paper investigates the equality test of risk ratios in multiple 2x2 tables with structural zero, which is formed by a confounding factor such as age, gender, severity of disease, region or other variables of interest. Score statistic, Wald statistic and likelihood ratio statistic for testing equality of risk ratios in multiple 2x2 tables with structural zero are proposed. Three procedures for calculating p-values of the above three statistics based on the MaxT, MinP and Bonferroni methods are developed. An algorithm for determining sample size required for attaining the pre-specified power is presented. Empirical results show that (1) likelihood ratio statistic and score statistic via the MaxT and MinP methods behave satisfactorily in the sense that their type I error rates are close to the pre-given significance level; (2) the Wald-type statistic with logarithmic transformation behaves better than the Wald-type statistic. Simulation studies and a real example are used to illustrate the proposed methodologies.

[1]  Man-Lai Tang,et al.  Statistical inference for correlated data in ophthalmologic studies , 2006, Statistics in medicine.

[2]  O. Kobori,et al.  High frequency of individuals with strong reaction to tuberculin among clinical trainees. , 1999, Japanese journal of infectious diseases.

[3]  K. Lui.  Testing homogeneity of the risk ratio in stratified noncompliance randomized trials. , 2007, Contemporary clinical trials.

[4]  J. Nam.  Power and sample size for stratified prospective studies using the score method for testing relative risk. , 1998, Biometrics.

[5]  K. Lui.  Interval estimation of the risk ratio between a secondary infection, given a primary infection, and the primary infection. , 1998, Biometrics.

[6]  J. Booth,et al.  Resampling-Based Multiple Testing. , 1994 .

[7]  J. Nam.  Power and Sample Size for Testing Homogeneity of Relative Risks in Prospective Studies , 1999, Biometrics.

[8]  Combining 2 x 2 tables that contain structural zeros. , 1995, Statistics in medicine.

[9]  S P Azen,et al.  OBTAINING CONFIDENCE INTERVALS FOR THE RISK RATIO IN COHORT STUDIES , 1978 .

[10]  J. Gart.  Approximate tests and interval estimation of the common relative risk in the combination of 2×2 tables , 1985 .

[11]  Man-Lai Tang,et al.  Exact unconditional inference for risk ratio in a correlated 2 x 2 table with structural zero. , 2002, Biometrics.

[12]  Giovanni Pistone,et al.  Algebraic Bayesian analysis of contingency tables with possibly zero-probability cells , 2006 .

[13]  K. Lui,et al.  Tests for homogeneity of the risk ratio in a series of 2x2 tables. , 2000, Statistics in medicine.

[14]  K. Lui.  Confidence Intervals of the Simple Difference between the Proportions of a Primary Infection and a Secondary Infection, Given the Primary Infection , 2000 .

[15]  A. Agresti,et al.  Categorical Data Analysis , 1991, International Encyclopedia of Statistical Science.

[16]  A. Tamhane,et al.  Multiple Comparison Procedures , 2009 .