Development of Chronic Kidney Disease Risk Prediction and Management System

Abstract

Chronic kidney disease is one of a major global public health issue, affecting over 10% of the population worldwide. It is the leading cause of death in 2016 ranking 16th and is expected to rise to 5th rank by 2040. Due to the significant morbidity, mortality and growing prevalence of the disease, there is a need to identify high risk subjects to avoid a greater burden. Identification of factors leading an individual to chronic kidney disease is also essential, as some risk factors can be improved, can stop or slow down progression to chronic kidney disease, and enhance the ability of health care providers to prevent kidney failure. Different risk factors have been identified for different countries, and risk prediction models also have been developed depending on risk factors worldwide for different countries to identify risk groups. This identified risk factor has shown variations in different countries. In Ethiopia identification of risk factors in general people is limited. Variety of the studies done in Ethiopia identified risk factors only for on one or two disease affected population, even though the disease affects any person. In addition, since risk factors vary in different countries due to life styles and other factors, prediction models are needed to be developed specifically. In Ethiopia there is lack of developing predictions systems which has considered Ethiopian people. Moreover, studies have indicated there is high prevalence and low awareness of chronic kidney disease in Ethiopia. In this research, a system that that can, estimate probability of having CKD , identify risk level of CKD, and recommends management of CKD risk factors is developed. Additionally, significant risk factors are also identified in the study. The study uses expert knowledge and statistical analysis to identify risk factors, to develop risk prediction and management system. From multivariable logistic regression analysis it‟s observed that male gender, overweight, hypertension, diabetes, experiencing injury on kidney, smoking above four years and family history of kidney disease were found to be significantly associated with chronic kidney disease. The system has showed 63.3 %, 65.3 % and 77.5% accuracy at 14%, 24% and 34% cut off percent respectively in estimating probability. This study will have significance in preventing chronic kidney disease at early stage and creating awareness.