A generalized partially linear mean-covariance regression model for longitudinal proportional data, with applications to the analysis of quality of life data from cancer clinical trials

Abstract

Motivated by the analysis of quality of life data from a clinical trial on early breast cancer, we propose in this paper a generalized partially linear mean-covariance regression model for longitudinal proportional data, which are bounded in a closed interval. Cholesky decomposition of the covariance matrix for within-subject responses and generalized estimation equations are used to estimate unknown parameters and the nonlinear function in the model. Simulation studies are performed to evaluate the performance of the proposed estimation procedures. Our new model is also applied to analyze the data from the cancer clinical trial that motivated this research. In comparison with available models in the literature, the proposed model does not require specific parametric assumptions on the density function of the longitudinal responses and the probability function of the boundary values and can capture dynamic changes of time or other interested variables on both mean and covariance of the correlated proportional responses. Copyright © 2017 John Wiley & Sons, Ltd.

A bootstrap semiparametric homogeneity test for the distributions of multigroup proportional data, with applications to analysis of quality of life outcomes in clinical trials

“…As noticed in the description of quality of life assessment in CO.17 trial, multigroup proportional data are collected longitudinally at many timepoints from the same subjects . Development of statistical procedures for the comparisons of longitudinal proportional outcomes over time between treatment groups is an interesting and important problem which requires further investigations.…”

“…Once the distribution is specified, likelihood based methods may be used for statistical inference, such as comparison of distributions of the proportional data among different treatment groups. There are, however, two problems when applying this parametric approach to statistical practice: (i) These distributions require the range of data are confined in an open interval such as (0, 1), instead a closed interval [0,1]; but, in practice, boundary values 0 and 1 may be observed, for example, in the QoL assessments from some patients who may consider themselves as having the best QoL and some others who may feel they had the worst QoL; (ii) a specific parametric distribution may be too restrictive to fit the proportional data which may be skewed without a specific pattern for its distribution as noticed by several authors …”

Spatiotemporal heterogeneity of SARS-CoV-2 diffusion at the city level using geographically weighted Poisson regression model: The case of Bologna, Italy

“…Most studies in medical research are based on classical regression models like ordinary least square (OLS) regression and generalized linear models (GLM) (Choi et al, 2017;Takele et al, 2019;Zheng et al, 2017) but these classical models produce bias by the resulting, average parameters over the whole study area without considering geographical variation. This kind of global regression models cannot detect non-stationary phenomena and may thus obscure differences in relationships between predictors and the outcome variable.…”

The effect of sociodemographic factors on COVID-19 incidence of 342 cities in China: a geographically weighted regression model analysis

“…In medical research, most studies utilize conventional regression models, such as ordinary least square regression and generalized linear models (GLM) [10][11][12]. However, these conventional regression models generate bias by producing average parameters over the whole studied regions without considering the potential geographical variation.…”