Search:

Cancer Control Research

Abstract

Assessment of quality of life is becoming more and more common, if not standard, in clinical cancer trials. Although the primary endpoints remain survival and tumor response, quality of life endpoints have become vital in interpreting the overall benefits of cancer treatments. The most popular method to measure quality of life is with instruments that utilize several multiple-item subscales, in which each item is scored on a Likert scale. Most statistical methods for the analysis of quality of life data in clinical trials do not explicitly consider the properties and psychometric features that were of interest in scale development. In this regard, the measurement and statistical summarization of QOL data, along with the clinical interpretation, can be somewhat disjoint from the psychometric concerns of the development process. The aim of this proposed research is to address the complex issues present in analyzing multiple-item ordinal quality of life data in clinical trials while maintaining fidelity to the psychometrical foundations upon which quality of life instruments are built. Accomplishing this will require the development of item response models which recognize the longitudinal aspects of clinical trial designs as well as the potential problems of informatively missing data and multiple comparisons. The proposed research will address these issues through the following interrelated goals: (1) Develop a general item response modeling approach for longitudinal multiple-item quality of life data measured on ordinal scales with model components for missing data mechanisms and latent trait regression on treatment indicators and other covariates; (2) Extend the models and inference techniques to include multiple latent traits for instruments which measure several dimensions of quality of life; (3) Incorporate robust variance estimation for situations when the proposed item response model may offer a useful summary of the underlying process, but the likelihood may be slightly misspecified; (4) Develop flexible Markov chain Monte Carlo algorithms for parameter estimation, which can easily be adapted for changes in one or more components of the model; and (5) Conduct simulation studies to assess the effects of model misspecification on parameter estimation and statistical hypothesis testing.