Operational Risk Research on Social Pooling Fund Under Diseases Score Settlement System

Objectives: To find out the inner outer risks and its influence on social pooling fund under diseases score settlement (DSS).Methods: To Use step multiple linear regression analysis, the risk factors of the fund have been screened out. The selected risk factors have been taken into BP artificial neural network (BPANN). Results: In 12,724 insured inpatients, chronic diseases accounted for 24.89%.The average medical expense per inpatient was 11,950.88RMB and per hospitalization expenditure of social pooling fund was 7,665.81RMB. The 10 variables such as age, sex, unit type, hospital level, individual pays,medicine fee, medical fee, operation fee, nurse expense, bed fee and other expense were statistically significant. Conclusion: The growing aging population, changes in disease spectrum, increasing medical costs are all risks of non-controllable running outside the system. Moral hazard and the defective design of the system belong to the system controllable risks. The results from BPANN were compatible with multiple linear regression analysis. The payment system plays an important role in health insurance [1]. Good payment can control the hospitalization expenditures in a reasonable scope, while an imperfect one can throw a monkey-wrench into the system.The diseases score settlement (DSS) is payment system of Huai’an in China. This article develops two simple models (step multiple linear regression analysis and back-propagation artificial neural network(BPANN)) to illustrate the risks both inside and outside DSS and explore the risk control function of DSS. BPANN are the most widely used networks and are considered to be the workhorse of ANNs because of its simplicity and its power to extract useful information from samples [2].Due to its strong learning ability and generalization capability, BP networks have been successfully used in forecasting some financial problems, for example, predicting stock market returns [3], loan risk warning [4] and forecasting bankruptcy firms [5].