Explore the Influences and predictionof Comorbidity on the Health Care Utilizations Factor with Chronic Disease: Example of Diabetes Mellitus
| Autor: | Ian-hua Li, 李姸嬅 |
|---|---|
| Rok vydání: | 2013 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 101 The cost of treating chronic diseases has been substantial rising in recent decades. In Taiwan, diabetes is one of the most common chronic diseases. According to the Department of Health, Executive Yuan, R.O.C. (Taiwan), the data show that in 2009, from 1997 to 2009 the diabetic population increased from 538 thousand to 1223 thousand. In 2005, health care costs for treating diabetes went up to 11.5% of the total health care costs. Cost increases were found both in outpatient and inpatient care. In addition, diabetes has more comorbid cases than diabetes only, and these result in massive expenditures、resources or budgets for the health care industry. Comorbidity of diabetes presents a high occurrence adding to the high cost of treating the disease, which can easily lead to extra social and financial burdens being placed on hospitals. This study’s objects are to examine the impact of diabetes comorbidity on the health care industry by using the 2009-2010 National Health Insurance Research Database for analysis. We investigated diabetic patients with comorbid changes to find out their total numbers for each category and to assess the relations to health care utilization. We used the following data mining techniques: 1) Back-propagation neural, 2) Decision tree C5.0 and 3) ENSEMBLE to predict the patients’ length-of-stay in order to reduce waste in the health care industry and to enhance the quality of medical care. The quantitative analysis results show that cardiovascular comorbid diabetes and renal comorbid diabetes has a high prevalence. The average ages for outpatients and inpatients were between 63 years and 69 years. Peripheral vascular comorbid diabetes and neuropathy comorbid diabetes average length-of-stays were longer. Renal comorbid diabetes has a higher average outpatient medical cost. Cardiovascular comorbid diabetes and Peripheral vascular comorbid diabetes have higher average inpatient medical costs. Medical centers and regional hospital have higher medical costs caring for comorbid diabetes patients than do district hospitals. Diabetes payment plans that cover inpatient care have higher medical costs. Prediction model: The back-propagation neural combined with ENSEMBLE has better accuracy. Respectively, ophthalmic comorbid diabetes is 89.63%, renal comorbid diabetes is 82.49%, neuropathy comorbid diabetes is 82.68%, cardiovascular comorbid diabetes is 77.08% and Peripheral vascular comorbid diabetes is 80.82%. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|