慢性疾病醫療給付改善方案對不同共病症病人併發症之影響—以高血壓及糖尿病為例

The impact of hypertension and diabetes mellitus pay-for-performance programs on complications among patients with different comorbidities

李宣緯(Hsuan-Wei Li)

指導教授 ： 楊銘欽

國立臺灣大學/公共衛生學院/健康政策與管理研究所/博士(2011年)

https://doi.org/10.6342/NTU.2011.00996

摘要

背景：多重慢性病患者照護品質的提升是目前各國健康照護體系所面臨的重要挑戰，隨著論成效計酬政策逐漸被各國醫療體系所採用，其對於多重共病症病人衝擊的瞭解至今仍非常有限。我國慢性疾病醫療給付改善方案已實施多年，然相關研究多針對給付設計評估績效衡量指標之成效，而對於相關併發症之發生及醫療費用之影響則較少見。
目的：高血壓及糖尿病皆為我國高盛行率及高醫療費用支出的疾病，且兩種疾病對於病人的疾病管理目標具有高度相關性。因此本研究同時以兩項方案為例，評估我國慢性疾病醫療給付改善方案在降低或延緩相關併發症發生之成效，及比較不同方案對併發症預防成效之差異，並進一步分析兩項方案對不同共病症病人併發症預防成效及併發症相關醫療費用之衝擊。
方法：以國家衛生研究院100萬人抽樣歸人檔之2001-2009年健保申報資料，運用傾向分數分析法的配對方式，校正論成效計酬方案逆選擇現象可能產生的選樣偏誤後，以羅吉斯迴歸分析評估醫療給付改善方案對急性冠心症、心臟衰竭、腦中風及慢性腎衰竭等四種高血壓及糖尿病的共同且主要併發症之預防成效，以複迴歸分析探討方案對四種併發症1年內相關醫療費用的影響，並以布瓦松迴歸分析方案介入對其醫療利用的改變情形。
結果：糖尿病醫療給付改善方案對於併發症發生機率的降低主要呈現在心臟衰竭及腦中風兩種疾病；急性冠心症雖未達統計顯著水準（p=0.060），但其改變符合預期之方向。而對於併發症相關醫療費用的減少則主要呈現在慢性腎衰竭，且主要是受到門診就醫次數減少的影響。高血壓方案對於併發症的預防成效，雖同時呈現在急性冠心症及腦中風，但進一步區分不同共病症組別，可以發現對於「高血壓+糖尿病+高血脂」組別病人而言，其腦中風發生機率的降低主要仍僅受到糖尿病方案的影響；而高血壓方案對於併發症相關醫療費用的減少，主要呈現在急性冠心症，但較無法明確歸因是受到醫療利用改變的影響。兩項方案對不同共病症病人預防併發症發生的成效，在同時罹患高血壓及糖尿病兩種疾病的共病症組別特別顯著，但在併發症相關醫療費用方面，受到併發症個案數少之限制，僅能針對整體病人進行分析，無法進一步探討對不同共病症病人之影響。此外，兩項方案對於部分併發症預防成效的影響存在效應修飾作用；大致而言，高血壓方案對預防併發症發生及減少相關醫療費用的成效，常需在病人同時有參加糖尿病方案的情況下才能達統計顯著差異，但糖尿病方案的成效，卻能在病人從未參加高血壓方案的情況下呈現統計顯著差異。
結論：兩種慢性疾病醫療給付改善方案對預防併發症發生及減少相關醫療費用皆有影響，但糖尿病方案有較佳的推行成效。本研究根據分析之結果，建議相關主管機關後續應持續發展並推動相關論質計酬方案。

關鍵字

高血壓 ； 糖尿病 ； 論質計酬 ； 共病症 ； 併發症 ； 傾向分數

並列摘要

Introduction: Improving the quality of care delivered to individuals with multiple chronic conditions is among the important challenges our health care system faces. While the use of pay for performance (P4P) in healthcare is increasing, little is known about the impact of these programs on persons with multiple comorbidities. In Taiwan, P4P programs for chronic diseases have been implemented for several years. Most studies evaluated the effects of these programs by measuring the performance indicators of payments, but little was undertaken to explore the impacts on complications and related medical costs.
Objectives: Hypertension and diabetes mellitus (DM) are both diseases with high prevalences, high medical expenditures and high relationship with similar management goals of diseases. Therefore, the objectives of this study were to evaluate the effects of hypertension and DM P4P programs, and explored the impact of these two programs on the incidence rates and the medical costs of related complications among patients with different comorbidities.
Methods: A retrospective analysis of the claims database of the National Health Insurance from 2001 to 2009 was performed. Propensity score matching method was applied to adjust for the characteristics to minimize the effects of selection bias. Logistic regression models, multiple linear regression models and poisson regression models were constructed separately to explore the effects of the two P4P programs on the incidence rates, the medial costs and the medical utilizations of acute coronary syndromes, heart failure, stroke and chronic renal failure.
Results: The DM P4P program decreased the incidence rates of heart failure and stroke. Although the impact on the incidence rate of acute coronary syndromes was not statistically significant, the tendency of prevention effect existed. The DM P4P program could also save the medical costs of chronic renal failure by decreasing related outpatient visits. For hypertension P4P program, it reduced the incidence rates of acute coronary syndromes and stroke, and decreased the medical cost of acute coronary syndromes. The complication prevention effects of the two P4P programs were most significant on patients comorbid with both hypertension and DM. Besides, the effect modification partially existed. The impact of hypertension P4P program on the reduction of the incidence rates and the related medical costs of complications were significant only when patients also participated in DM P4P program. However, the effect of DM P4P program could be statistically significant even when patients did not participate in hypertension P4P program.
Conclusions: Although both the hypertension and DM P4P programs present impacts on complications, the overall impact of DM P4P program was greater than that of hypertension P4P program. Our findings suggest that policy makers could continuously develop and implement related P4P program in the future.

並列關鍵字

hypertension ； diabetes mellitus ； pay for performance ； comorbidity ； complication ； propensity score

參考文獻

3.陳宗泰, 鍾國彪, and 賴美淑, 另一種流行趨勢－論成效計酬的趨勢與展望. 臺灣公共衛生雜誌, 2007. 26(5): p. 353-370.

22.陳都文 and 王俊毅, 糖尿病共同照護網對偏遠地區中老年患者健康狀況之成效. 臺灣公共衛生雜誌, 2009. 28(4): p. 334-343.

40.周碧瑟, et al., 台灣地區糖尿病流行病學. 臺灣公共衛生雜誌, 2002. 21(2): p. 83-96.

43.塗宜育 and 許惠恒, 血糖異常與心血管疾病. 內科學誌, 2008. 19(4): p. 289-296.

45.康志森, et al., 冠心病事件危險預測之新知. 內科學誌, 2006. 17(4): p. 143-154.