BAYESIAN SURVIVAL ANALYSIS OF DIABETES MELLITUS PATIENTS: A CASE STUDY IN TIKUR ANBESSA SPECIALIZED HOSPITAL, ADDIS ABABA, ETHIOPIA
Diabetes is a complex, chronic illness that occurs either when the pancreas does not produce enough insulin or when the body cannot effectively use the insulin it produces. Globally, 415 million (340-536 million) people have diabetes in 2015 with regional prevalence of 8.8% (7.2-11.4%) by 2040 this figure will expect rise to 642 million (521-829 million) with predicted prevalence rate of 10.4% (8.5-13.5%) and more than 22 million people in the African Region; by 2040 this figure will almost double (IDF, 2015).The statistical result of World health organization estimated that the number of cases of diabetics in Ethiopia to be about 796,000 in 2000, and projected that it would increase to about 1,820,000 by the year 2030 (WHO, Diabetes estimates and Projections, 2003).But,according to the report of international diabetes federation atlas in 2017 there were around 2,567,900 [1,094,000-3,795,400] million diabetes cases in Ethiopia in 2017 (IDF atlas, 2017). The general objective of the study is to identify the determinant risk factors for the survival of Diabetic mellitus patients. From 2474 patients a sample of 451 diabetes patients administered the treatment in Tikur Anbessa specialized Hospital between September 11/ 2008-9/5/2014 were included in the study. The data were analyzed using classical and Bayesian Accelerated failure time model because of the failure in proportional hazard assumption. Bayesian Accelerated failure time model was better model than Classical Accelerated failure time model because it contains smaller AIC. Descriptive statistics and the Kaplan-Meier survival curves were used to estimate and compare the survival time of diabetes patients among different categorical characteristics of the patients. From the result, the survival time until death is significantly related to the age category, BMI, types of diabetic disease, alcohol use, diabetic complication, blood pressure, cholesterol level, family history of diabetic, fasting blood sugar, comorbidity, density lipoprotein, triglyceride level and smoking habit. The patients should keep their normal body weight and change their life style such as smoking habit, alcohol consumption, and take care of on their lipid cholesterol level.