Bayesian Competing Risks Model: An Application to Breast Cancer Clinical Trial with Incomplete Observations

Presences of Lost to follow up observations of patient in any clinical trial is natural and difficult to handle. Conclusion about the effective therapeutic region becomes conflicted in presence of lost-to follow up observations particularly in death-event data. It becomes more tedious to find out the best therapy under consideration. To make certain about compatible treatment, it is important to handle the data into proper and suitable manner. Using data from randomized clinical trial for patients with node negative breast cancer, we illustrated the approach to handle lost to follow up observation. The primary events are considered as death/alive with equal priority to the censored observations. Nelson-Aalen estimator and Bayesian approach are considered to handle the lost to follow up in the data set. Therapeutic differences on survival event were observed among the treatment under consideration. Competing risks model is found suitable choice to deal with death-event data in presences of lost-to follow up observations. Bayesian with non-informative priors and Nelson-Aalen estimator as distribution free approach are compatible to deal with death-event data.