Predicting the risk of hospitalization to six diagnosis with highest costs based on outpatient claims

The amount of medical expenses incurred in health insurance is usually determined by the maximum frequency limit and the cost for each component. If this year’s total outpatients for the diagnosis of typhoid fever is 500 patients and 239 patients are hospitalized with total incurred up to Rp 1.396.388.877, how the total incurred if next year there is an increase risk incidence of hospitalization? This problem is interesting to discuss as to predict the risk of occurrence in hospitalization based on outpatient claims data. Limitation problem in this study is limited to six diagnosis of coronary disease that are atrial disease, typhoid fever, dengue hemorrhagic fever, chronic renal failure, other specified diabetes mellitus, and diarrhea and gastroenteritis of infectious origin presumed as the highest cost of disease occurrence and frequency of hospitalization often occurred based on secondary data obtained from a third party administrator firm PT XX. To predict, we will be used the binary logistic regression analysis and to see the pattern of the spread or the characteristics of the prediction, we will be used multiple correspondence analysis (MCA). Binary logistic regression show an increase in the incidence of hospitalization for typhoid fever and dengue hemorrhagic fever are influenced by the frequency of outpatient and age. MCA showed an increased risk of hospitalization tend to occur in RS Awal Bros Bekasi, RS Eka Hospital BSD, RS Bina Sehat, RS Hermina Arcamani, RS Graha Permata Ibu, RSIA Tumbuh kembang, and RS H Darjad.