A joint longitudinal and survival model for dynamic treatment regimes in Presence of Competing Risk Analysis

Abstract

Background: In oncology setup, the personalized medicine with the Dynamic Treatment Regime(DTR) is the attractive tool for treatment management. It is based on sequence rule by changing the required treatment by looking into patients dynamic condition repeatedly. This repeatedly measured condition generates the longitudinal data and the time-to-event data jointly. The time-to-event data is generated with competing risks. Now handling the joint longitudinal and survival data with competing risks is itself challenging. It becomes more challenging to decide the best effective treatment strategy while we work with DTR approach in presences of competing risks through joint longitudinal and survival model. Methods: This article is dedicated towards development of statistical methodology to handle competing risk analysis for DTR in repeatedly measured survival data. In this study a simulated dataset is used to resemble the observed data distribution seen in a motivating cancer trial data. Results: This algorithm is prepared through Bayesian analysis to obtain the best effective therapeutic regimen. The OpenBUGS function is prepared, which provides the therapeutic effect comparison and there after estimations between different treatment sequences. Conclusion: This developed method is easy to handle for personalized medicine context in oncology for supportive decision rule.