A Bayesian Case-Control Study to Study the Effect of Covariates in Gall Bladder Carcinoma

Abstract

The paper provides a Bayesian study to see the effect of different covariates in developing gallbladder carcinoma from a group of patients examined in the form of case-control scenario. The cases are those who have developed carcinoma from a group of gallbladder patients and the control are those who belong to the latter category but not in the former. The study has also been extended to evaluate the odds ratio, a ratio that measures the odds of exposure for cases compared to controls. Two different realistic modelling formulations-one based on beta-binomial and the other based on its extension to multinomial-Dirichlet setup are entertained for the intended study. The results are finally illustrated by means of a real data based example taken from a Polish study where the effect of factors such as age, sex, eating habits, etc. has been examined in developing carcinoma of the gallbladder.