Abstract
The spatial statistical models or Bayesian methods have been used widely to analyze and study the patterns of geographical distributions for a rare disease or small geographical area. Cancer is the ever-increasing health problem and most common cause of medical deaths around the world. Additionally, cancer has the highest prevalence of diseases in Libya. Libya, as the study area, was selected to perform this research and to estimate its relative risk for liver cancer. Cancer data and population censuses of the country for the time period 2020 were used in this study. The objectives of this study are analyzing and identifying the unusual high and low risk of Libya as a pattern and as the process (the liver cancer mapping) based on the Standardized Morbidity Ratio (SMR), a traditional approach to measuring the relative risk of the disease, and Poisson-gamma model. Also to understand and assess the relationship of liver cancer disease and its high risk of a study area. Regions of the study were extracted by using GIS system. Then, observed liver cancer cases data and the population data were coupled to perform the analysis using WinBUGS software. This study starts with a brief review of these models, starting with the SMR method and followed by the Poisson-gamma model, which we then apply to liver cancer incidence in Libya. All results are compared using maps and tables. The study concludes that the Poisson-gamma model gives better relative risk estimates compared to the classical method. The Poisson-gamma model has can overcome the SMR problem when there is no observed liver cancer in an area