全民健康保險論質計酬支付制度再評估:以糖尿病論質計酬為例

Abstract

目的： 糖尿病是容易發生其他相關併發症而導致多重慢性病的常見慢性病，對 醫療體系費用帶來極大負擔。糖尿病論質計酬計畫實施後的成效評估，多以醫療費用和利用為主。本研究主要分析病論質計酬方案對於糖尿病病人的照護協調性、併發憂鬱症照護的外溢效果及方案的病人選擇情形。 方法： 本研究使用2007年至2015年健保特別需求申報檔，以準實驗研究設計進行次級資料分析。以加入論質計酬方案之糖尿病病人為研究組，用傾向分數配對出(PSM)從未加入計畫之對照組，再利用照護密度分數(care density score)評估照護協調性；分析方案對糖尿病患併發憂鬱症發生率及相關費用非預期影響的外溢效果。最後分析各年糖尿病併發症嚴重程度指標(DCSI)及共病程度指標(CCI)兩項疾病嚴重度指標在研究組及對照組分布的變化，以評估病人選擇情形。 運用差異中之差異法(DID)、差異中之差異中之差異法(DIDID)及複迴歸以評估糖尿病論質計酬方案的影響。 結果： (1) 加入方案後，照護密度分數顯著高於未加入者。(2) 方案對併發憂鬱症此項未在獎勵指標內的心理疾病，有減少門診費用及整體醫療費用的非預期影響效果。(3) 糖尿病患者納入方案比例有提升，但病人選擇情形仍存在。 結論： 糖尿病論質計酬方案能提升糖尿病病人的照護密度，增加其照護協調性，進而減少病人的醫療費用。對非獎勵指標的心理疾病照護，具有正向的外溢效果，能提高憂鬱症狀的發現和降低相關醫療費用。比較納入與未被納入方案的糖尿病病人，合併症指標分數高的疾病複雜者，更容易被排除在方案之外。 未來如何持續提升方案在整體糖尿病人涵蓋比例，讓更多人受惠，應是糖尿病論質計酬支付制度的重要目標。 關鍵詞：病人選擇、論質計酬、糖尿病、照護密度、憂鬱症

Objective: Diabetes is not only a chronic disease, but also one of the common chronic diseases that is prone to other related complications and lead to multiple chronic diseases, which brings a great burden to the medical care system. The pay for performance payment program for diabetes has been piloted in Taiwan since 2001. It has attempted to encourage medical providers to continuously improve the quality of medical care by increasing benefits from the indicators of payment system structure, process and results. The main purpose of this study is to analyze the continuity of care for diabetic patients, the spillover effect of program on depression care, and evaluation of patient selection. This study mainly used long-term claims data to provide reference for future policy improvement. Methods: Analyzing the claims data of national health insurance from 2007 to 2015 using secondary data analysis with quasi-experimental research design. Diabetic patients who joined the pay for performance program were defined as the study group, and the propensity score matching was used to match the control group who had never joined the program. The care density score was used to evaluate the continuity of care. The incidence of depression complication and related costs in diabetic patients were analyzed to evaluate the spillover effects of program. Two disease severity indexes (DCSI) and comorbidity index (CCI) were analyzed to analyze the change of patient selection. Difference in difference method and difference in difference in difference were used to evaluate the impact of the program. Result: The results found that (1) the care density score of patients in the program was significantly higher than those without enrollment. (2) The program has a positive spillover effect of reducing expenditures for the care of mental illnesses such as depression complication, which is not included in the design of incentive indicators. (3) The proportion of diabetic patients included in the program has increased, but there is still patient selection is still exists. Conclusion: The pay-for-performance program can increase the care density for diabetic patients and reduce the medical expenses of patients. There is a positive spillover effect on improving the detection of depressive symptoms and reducing its related medical costs, benefiting people with diabetes. The results of the study found that there is still existing patient selection, especially in diseases with high comorbidity index scores. Patients with complication are more likely to be excluded from care programs. In the future, the design of program should further establish reward indicators according to the severity of diabetic patients and disease complications, continue to increase the coverage of the plan in the overall diabetic patients, and reduce the medical cost. Keywords: Patient selection, pay for performance, diabetes, care density, depression