Background: Colorectal cancer (CRC) is an important cause of mortality and morbidity in many communitiesworldwide. This population based study was conducted to assess determinants of colorectal mortality in Iranianpatients. Materials and
Methods: A cohort of 1,127 cases of confirmed colorectal cancer registered in a populationbased registry covering 10 referral hospital in Tehran, Iran, were followed for five years. Information about tumorcharacteristics, smoking status and family history were collected at base line and survival status were followedevery six months by contacting patient or next of kin (if patients died during the follow-up). The cause of deathfor each case was validated by verbal autopsy and referring to patient medical records at the time of death. Thedata were analyzed by Stata software using univariate and multivariate analysis (Cox regression). In building themodel a p value of less than 5% was considered as significant.
Results: The age at diagnosis was 53.5±14 years.Sixty one percent were male. Colorectal mortality among the patients was 96.9 person-years among men and 83person-years among women. Seventy five percent of patients lived for 2.72 years, 50% for 5.83, and 25% for 13years after the diagnosis of colorectal cancer. The age at diagnosis was significantly different between men andwomen (p<0.03). Higher tumor grade predicted higher death rate; the adjusted hazard ratios were 1.79 (95%CI,0.88-3.61), 2.16 (95%CI, 1.07-4.37), and 3.1 (95%CI, 1.51-6.34) for grades II, III, and IV respectively when theywere compared with grade I as reference. Ethnicity, marital status, family history of cancer, and smoking wererelated to survival with different degrees of magnitude.
Conclusions: Among many factors related to survivalamong the colorectal patients, tumor grade and smoking showed the highest magnitudes of association.