| 摘要: | 「全民健康保險醫療給付改善方案」是全民健保醫療費用支出改採論質計酬方式的開始，健保局於民國九十年首推子宮頸癌、乳癌、肺結核、糖尿病及氣喘等五項疾病醫療給付改善方案至今，許多研究結果雖顯示該方案有不錯成效，但大多是學界的分析結果，且囿於資料取得不易，其資料涵蓋率往往限於某一醫院或某一地區。健保局的申報資料涵蓋所有參與給付改善方案的全國性照護資料，不僅有個案管理照護系統資料，亦有門診、住院及急診等醫療利用資料，為使個案管理照護系統資料能有效達到疾病管理計畫的功能，應加以運用。因此本研究將以方法學回顧並整理國內外有關以健保申報資料相關文獻之病患配合度的研究進展，以健保申報資料為基礎，建構適當疾病管理指標且分析結果，以作為評估健保局給付方案或發現問題之參考。本研究將取得全民健康保險之資料庫與衛生署死亡檔，同時蒐集文獻並整理各種疾病之病患配合度指標，根據整理之指標，處理資料庫和初步資料分析以計算出病患配合度數據。接下來根據各個疾病進行專家座談，討論經文獻整理適合本土且健保局資料庫可衍生之配合度指標的完整性與適用性，經補充與修改後，再次進行專家座談討論配合度指標之應用，希望藉由不同專業專家，使得本研究所發展之指標能適用於疾病管理之監控。接下來進行病患配合度指標分布之描述及其與照護結果變項相關之統計分析，最後，完成結論與建議。
Program evaluation provides the information regarding the extent to which the proposed program meets its defined objectives or whether the aspects of the proposed program is superior in costs or effectiveness to those in an existing program. Therefore, evaluation of a new program is necessary. There are two specific objectives of this study. One is to develop adherence measures using administrative claims data of National Health Insurance in patients participated Diabetes Managed Care Program, Asthma Managed Care Program, or Tuberculosis Managed Care Program implemented by Bureau of National Health Insurance in 2001. The other is to evaluate the managed care programs by describing the distributions of adherence, utilization of inpatient and emergency care, length of stay, mortality, and outcome measures, and examining the relationship between them. This is a longitudinal cohort study of patients with diabetes, asthma or tuberculosis enrolled in managed care programs before 2002. They have to complete a health status assessment and are enrolled in the managed care programs continuously for one and half years. The baseline assessment includes items related to demographics, insurance variables, life style, disease status (severity, comorbidity, or complications), health care service utilization in the year before enrollment, and outcome measures. Demographic, clinical, and utilization-related economic variables will be retrieved from the administrative claims data of the Bureau of National Health Insurance. Mortality status will come from Death Dataset of Department of Health. Associations will be examined by Chi-square, t-test, analysis of variance, multiple linear regression, and Cox？H？Hs proportional hazard model when appropriate. |
