Xiang Liu1, Ju-ying Zhang. 1. Department of Health Statistics, West China School of Public Health, Sichuan University, Chengdu 610041, China.
Abstract
OBJECTIVE: To explore the application of the generalized estimating equation in the ordinal repeated measures data and hence provide methodology reference for the analysis of repeated measures data in the clinical trials. METHODS: An example was illustrated by modeling generalized estimating equation using the GENMOD command in comparison with the independent logistic regression. RESULTS: All parameters and their standard error were estimated, so every factor could be dealt with intuitive estimation of parameter. The standard errors of coefficients in generalized estimating equation are generally greater than that in independent logistic regression. CONCLUSION: Generalized estimating equation can solve the correlation between the dependent data by using working correlation matrix, and it can control strata correlation, repeated measures factor and other confounding factors effectively, so generalized estimating equation provides an effective method for the ordinal repeated measures data.
OBJECTIVE: To explore the application of the generalized estimating equation in the ordinal repeated measures data and hence provide methodology reference for the analysis of repeated measures data in the clinical trials. METHODS: An example was illustrated by modeling generalized estimating equation using the GENMOD command in comparison with the independent logistic regression. RESULTS: All parameters and their standard error were estimated, so every factor could be dealt with intuitive estimation of parameter. The standard errors of coefficients in generalized estimating equation are generally greater than that in independent logistic regression. CONCLUSION: Generalized estimating equation can solve the correlation between the dependent data by using working correlation matrix, and it can control strata correlation, repeated measures factor and other confounding factors effectively, so generalized estimating equation provides an effective method for the ordinal repeated measures data.
Authors: Teresa Schuhmann; Selma K Kemmerer; Felix Duecker; Tom A de Graaf; Sanne Ten Oever; Peter De Weerd; Alexander T Sack Journal: PLoS One Date: 2019-11-27 Impact factor: 3.240