Estimating the Distribution of Quality Adjusted Life With Censored Data

Anastasios A. Tsiatis

Quality of life is an important aspect in evaluations of clinical trials of chronic diseases, such as cancer and AIDS. Quality adjusted survival analysis is a method that combines both the quantity and quality of life into one single measurement. In most clinical trials quality adjusted life (QAL) may be censored due to incomplete follow-up. Estimation of the distribution of QAL using the Kaplan-Meier estimator is biased because of induced informative censoring. We propose a method for consistent estimation of the distribution of QAL based on the use of weighted estimating equations. Asymptotic properties for these estimators are derived. This includes asymptotic normality as well as issues of efficiency. These methods can also be used for estimating and comparing the mean QAL among competing treatments. Simulation experiments are conducted to evaluate the small sample performance of these methods.