Misclassification in registry of cancer statistics is a major problem for developing countries, often leading to underestimate health risks and burden of diseases. Two statistical approaches are recommended to overcome misclassification. The first one uses a small validation sample and the second one is Bayesian analysis. In this paper, a simulation study was conducted to evaluate the efficiency of Bayesian model and estimate the misclassified parameter, which included two main scenarios -  first using informative priors and second with non-informative ones. Misclassified parameter was simulating for range from 0.1 to 0.7 to provide weak to strong misclassification. Mean square error was calculated for efficiency criteria as the mean difference of Bayesian estimation of misclassified parameter and simulated ones. The results indicate a good Bayesian estimation for misclassified parameters. Also, according to efficiency analysis, the Bayesian estimation with informative priors was more efficient than non-informative ones. With small sample size and lower variances, the efficiency increases.

Bayesian, burden, mortality, incidence, registration.