Joint Modeling of Longitudinal Patient Reported Outcomes and Survival Data with Application to an Oncology Clinical Trial
Mark Ernest Boye, Eli Lilly and Company 
Wei Shen, Eli Lilly and Company  
*Ping Wang, Eli Lilly and Company  

Keywords: joint modeling, longitudinal, patient reported outcomes, survival

Joint modeling of longitudinal (repeated measurement) and survival data has been shown to account for the dependencies and associations between these two data types thereby providing a better assessment of treatment effect. More specifically, and under certain circumstances, these models result in more efficient and less biased estimates of treatment effects. Sponsors of oncology clinical trials routinely and increasingly included patient-reported outcome (PRO) instruments to evaluate the effect of treatment on symptoms, functioning, and quality-of-life. Known publications of these trials typically do not include jointly modeled analyses and results. We conducted a post-hoc analysis of a randomized Phase 3 oncology clinical trial to jointly model these data through use of a mixed effect model (longitudinal PRO data) and a Cox proportional hazard model (survival data). The longitudinal and survival components were linked through either random effects or a latent PRO trajectory. We compared the results derived under different model specifications and showed that the use of joint modeling may result in improved estimates of the overall treatment effect.