Modeling health insurance claim severity in NHIS using parametric probability distribution: (A case study of Amansie East Municipal)

Thumbnail Image
April, 2016
Journal Title
Journal ISSN
Volume Title
General insurance companies face two major problems when they want to use past or present claim amount in forecasting future claim severity. First they have to find the appropriate probability distribution for the large volumes of claim amount. Then test how best these distributions fits the claim data The purpose of this study is to fit a particular distribution suitable for the National Health Insurance Scheme in the Amansie East Municipality. Secondary data was collected from Amansie Municipal Insurance Scheme. The exploratory data analysis technique was used to assist in the identification of the family of distribution which the data might follow. Akike Information Criterion and Kolmogorov- Smirnov used to test the goodness of fit. The diagnostic test probability plot was to used to graphically demonstrate the goodness of fit the distribution. It was found that the Lognormal distribution was appropriate distribution for Fee For Service for all categories of service in the district. However the Burr distribution were considered to be the best distribution for G-DRG in the District level, CHPS compound, Community clinics, Health centers and CHAG hospital. Also lognormal distribution was the best for G-DRG modeling for Private, Chemical and Pharmacy shops and finally the fisk distribution for Public clinics and Maternity. Management at all municipal and district health insurance schemes should be able to apply the appropriate statistical distribution used in this research for management policy prescription to improve their performance.
A thesis submitted to the Department of Mathematics, Kwame Nkrumah University of Science and Technology in partial fulfillment of the requirements for the degree of Msc. Actuarial Science.