Establishing a forecasting model for outpatient volume--Based on a community hospital

• Main Authors: Mei Reng Chen, 陳美礽
• Other Authors: Tsochiang Ma
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/88230727154900944859

碩士 === 中國醫藥大學 === 醫務管理研究所 === 93 === In the field of medical utilization, most investigators focus on medical expense or individual medical utilization, only few studies had paid attention on OPD utilization. This study try to use a new methodology, time's sequence analyze technology to establish a hospital OPD volume forecasting model. We also try to verify the relationship between hospital characteristics, different health insurance coverages and OPD volume. These results will be valuable information to hospital managers and healthcare policy makers in decision making. This study was a restrospective observational study between January 2000 and December 2004 in a mid-Taiwan regional hospital. Using SCA package software, univariate ARIMA model , transfer function model and intervention function model verified the relationships between hospital characteristics,（such as doctor number, doctor’s average age, department number, OPD number）, policy factors （out-patient service reasonable quantity, hospital's global budget, hospital self-management） and OPD volume. The main results were 1.The result of single variable ARIMA revealed that outpatient volume is strongly correlated to previous OPD volume. Therefore it should be an appropriate model for OPD forecasting. 2.The result of transfer function model shows that the change of doctor number and department number have an impact on the OPD volume two months later. OPD number could influence the policy implemented by NHI, the OPD volume immediately, and doctor’s average age was not statistically significant related to OPD volume. 3.The result of intervention function model revealed that out-patient service reasonable quantity, and hospital's global budget, could not suppress the OPD volume. In the contrary, the OPD volume increased. 4.The transfer function based on doctor's number was the most fitted model, but single variable forecasting was more effective based on previous data. The results of our study have some implication to hospital managers and healthcare policy makers. For the hospital managers 1.Using the doctor’s number and department’s number two months ago could forecasting the OPD volume and adjust the manpower in advance. 2.If the forecasting of OPD volume was in a trend of decreasing, strategy of decrease resources investment, such as layout, replanning OPD number should be taken. 3.The effect of hospital self-management had no effect on OPD volume in this study. However, there was only 6 months duration. A longer observation period is warrant. 4.The forecasting result of univariate ARIMA is more effective than policy intervention. Some unknown influence factors may exist and need further investigation. For the healthcare policy makers 1.Primary physician system and referral system should be performed tightly. Establish leveled medical system and abandon of reasonable OPD quantity. 2.Adjust the payment system, using OPD per person payment system instead of fee for service.