a bayesian approach to model the underlying predictors of earlyrecurrence and postoperative death in patients with colorectal cancer

This study, aimed at using a bayesian approach semi-competing risks techniqueto model the underlying predictors of early recurrence and postoperative deathin patients with colorectal cancer (crc). methods: the primary outcomes were theprobability of recurrence, the probability of death without recurrence, and the probabilityof death after recurrence. the patients “recurrence status was determined frompatients” records. the bayesian survival modeling was carried out by semi-competingrisks illness-death models in r 4.1 software. results: the results showed that gender,age at diagnosis, t3 stage, n2 stage, tumor size, grade (differentiation level), and thenumber of chemotherapy, metastasis to other sites was significantly related with recurrence,the death and death after recurrence. conclusion: according to the specificresults obtained for terminal and nonterminal events, appropriate screening strategiesand the earlier detection of crc lead to substantial improvements in the survival ofpatients.