Bayesian Estimation of Colorectal Cancer Mortality in the Presence of Misclassification in Iran

Abstract

Background: Colorectal cancer is the third most common cause of cancer-related deaths in the world.Addressing the burden of related mortality is therefore important but according to the Iranian mortality registry,about 20% of death statistics were still recorded in misclassified categories. The aim of this study is to reestimatethe CRC mortality rate for Iranian population, using a Bayesian approach in order to revise thismisclassification.
Methods: National Death Statistics Reported by the Ministry of Health and Medical Education(MOH&ME) from 1995 to 2003 were included in this analysis. The Bayesian approach to correct and accountfor misclassification effects in Poisson count regression was employed to estimate the mortality rates by age andsex group.
Results: According to the Bayesian re-estimate there were between 30 to 40 percent underreportedmortality records in death due to colorectal cancer and the rate for related mortality had moderately increasedthrough recent years.
Conclusion: Our findings suggest a substantial undercount of colorectal cancer mortalityin the Iranian population. Therefore healthcare policy makers who determine research and treatment prioritiesfrom death rates as indicators of public health problems should pay notice to this underreporting.

Keywords