糖尿病論質計酬之成果與成本分析-系統性回顧與統合分析

摘要研究目的根據國民健康署統計，糖尿病為國人十大死因之ㄧ，每年有近萬人因糖尿病死亡，全國約有200多萬名糖尿病的病友，且每年以25,000名的速度持續增加，糖尿病及其所引發的併發症嚴重影響國人健康，醫療負擔也急遽增加。根據上述研究背景與動機，本研究將探討在論質計酬模式下，糖尿病病人醫療資源使用及醫療結果之差異。研究方法研究架構為糖尿病病人經由論質計酬策略的介入措施，對於糖尿病控制成果、成本效果與無介入措施間的差異，使用多個資料庫去搜尋文獻，並進行系統性整合相關研究，文獻的搜尋必須是18歲以上成人之糖尿病族群，介入措施為論質計酬，時間設定為2001年至2018年文獻，本研究針對臨床試驗的文章做統合分析，排除系統性回顧文章；排除導讀文章；排除臨床指引等。藉由Comprehensive Meta-Analysis (CMA)軟體做統合分析。研究結果本研究依研究目的與假說，搜尋相關文獻進行文章評讀與統合分析，得出糖化血色素控制良好的比例在論質計酬顯著性高於非論質計酬(OR=1.14, P<.001)。膽固醇控制良好的比例在論質計酬比非論質計酬好，但無統計學上顯著差異(OR=1.034, P=0.43)。血壓控制良好的比例在論質計酬顯著性高於非論質計酬(OR=1.733, P<.05)。糖尿病死亡率在論質計酬顯著性低於非論質計酬(OR=0.46, P=0.008)。整體而言，論質計酬模式下，糖尿病病人有較好的照護品質。8篇糖尿病照護成本進行統合分析，得出糖尿病直接醫療成本，論質計酬顯著性高於非論質計酬(d=0.308，P<.001)。雖然，論質計酬對糖尿病照護品質有所提升，但直接醫療成本也有所增加。結論與建議本研究經統合分析後，發現糖尿病在論質計酬下比非論質計酬組有較好的照護品質，糖化血色素與血壓控制良好比例亦有顯著性增加，死亡率顯著性減少。但是，在論質計酬下，直接醫療成本增加，糖尿病併發症減少、住院率及住院天數減少，相對病人生活品質提升，醫療耗用減少，所以糖尿病論質計酬實施有其正面效果。可能有不同因素影響論質計酬的成果，將獎勵的對象變為糖尿病病人自己，或許能改變病人對糖尿病控制的行為，此研究成果將可提供衛生主管機關於相關政策制定之參考。

關鍵字

糖尿病；論質計酬；成本分析；臨床成果

並列摘要

Abstract Purpose In Taiwan, diabetes is the top ten cause of death. Nearly 10,000 people die from diabetes every year. There are more than 2 million diabetic patients, and the increasing rate is 25,000 annually. Diabetes complications have seriously affected the health of people, and the economic burden was also increased rapidly. Based on the research background and motivation, this study purposed to compare the differences between pay-for-performance (P4P) and non-pay-for-performance (non-P4P) in the medical resources utilization and medical outcomes among diabetic patients. Research Methods The research framework is the intervention of diabetes patients through the quality compensation strategy. We evaluated the effect of diabetes outcomes and cost between P4P and non-P4P. Multiple databases are used to search the literatures, systematically integrate relevant research. The literature search must under the condition of older than 18-year-old diabetes patient and the time period is from 2001 to 2018. This study conducts an integrated analysis of articles for clinical trials, excludes systematic review articles, guided articles, and clinical guidelines. Integrated analysis was performed by Comprehensive Meta-Analysis (CMA) software. Results According to the research purpose, it showed that the proportion of good HbA1c control is significantly higher than non-P4P (OR=1.14, P<.001). The proportion of good cholesterol control was better than that of non-P4P, but there was no statistically significant difference (OR=1.034, P=0.43). The proportion of well-controlled blood pressure was significantly higher than that of non-P4P (OR=1.733, P<.05). Diabetes mortality was significantly lower than non-P4P compensation (OR=0.46, P=0.008). According to the above analysis, under the P4P mode, diabetic patients have better quality of care. Eight integrated literature about diabetes medical costs were analyzed. The direct medical cost under P4P was significance higher than non-P4P (d=0.308, P<.001). Although P4P mode has improved the quality of diabetes care, direct medical costs have also increased. Conclusions and Recommendation Under the P4P program, diabetes patient had better quality of care, including the increased ratio of good HbA1c and blood pressure control than non-P4P. Under the P4P program, the direct medical cost had increased, but there were many studies had explored the quality of the reward, such as the reduction of diabetes complications, hospitalization rate and hospitalization days, improvement of quality of life and decreased medical resource, therefore, the quality of diabetes implementation has its positive effects. There are different factors affecting the results of the quality of the rewards. The target of the reward into the diabetes patients may change the behavior of the patient for diabetes treatment. The results of this research will provide a reference for the health policy.