推估高風險手術集中化政策對照護結果影響之成效探討-以心血管處置為例

The Implementation of Centralization for High-risk Surgery-Estimating Potentially Avoidable outcome of cardiac procedure

陳靖怡(Ching-Yi Chen)

指導教授 ： 郭年真 ； 鍾國彪

國立臺灣大學/公共衛生學院/健康政策與管理研究所/碩士(2016年)

https://doi.org/10.6342/NTU201600318

摘要

研究背景：醫療照護服務量與照護結果關係的研究(volume-outcome research)發展至今已相當成熟，這些研究結果也是集中化衛生政策重要的決策依據，期望將病患導引至高服務量或高專業性的醫療機構內接受照護服務，並且使病患獲得最佳的照護並改善整體照護結果。目前有許多國家已針對高風險手術實施集中化政策，同時文獻也指出，多數國家實施手術集中化後對於整體照護結果確實有顯著之正面影響。反觀台灣，近年來同樣也有大量的實證證據顯示手術服務量與照護結果之關係，但目前仍未有集中化相關衛生政策介入。因此集中化政策是否適用於我國之醫療體系內並且有助於提升手術處置照護品質，需要進一步評估手術集中化政策可能帶來之效益。

研究目的：本研究以兩項高風險之心臟處置為例，探討若台灣實施集中化政策，只允許高服務量醫院執行處置，(1)對於病患接受處置後30日之死亡、非計畫再住院及併發症之影響；(2)對於民眾就醫可近性的影響；以及(3)對於低服務量醫院之住院醫療收入之影響。

研究方法：本研究採回溯性研究法，主要以2009年至2012年全民健康保險資料庫進行資料分析。研究對象為2010至2012年各年度第一次接受經皮冠狀動脈氣球擴張術(PTCA)以及各年度第一次接受冠狀動脈繞道手術(CABG)之病患，以多階層羅吉斯迴歸探討醫院服務量與照護結果之關係，並且以G-Computation來估計可避免不良照護結果之事件發生人數。病患就醫可近性影響之方面，以區域變異法來探討跨區就醫情形並使用Open GeoDa軟體來預測集中化後各醫療次區域跨區就醫比例及醫院家數分布情形。最後，以醫院為分析單位，瞭解醫院執行心臟處置之醫療收入占當年度住院醫療收入之比例來推測低服務量醫院在住院醫療收入的影響程度。

研究結果：在2010年至2012年間，共有92,370人曾接受PTCA及9,530曾接受CABG，並且分別有29,689及4,150人至低服務量接受處置。集中化後，預估PTCA可避免死亡人數總共有442人(95% CI: 274-693)、可避免再住院人數共有620人(95% CI: 471-776)、可避免併發症發生人數有766人(95% CI: 607-921)；在CABG部分可避免死亡人數共有127人(95% CI: 72-218)、可避免再住院人數共有66人(95% CI: 30-125)、可避免併發症人數共有271人(95% CI: 201-299)，然而，研究結果發現醫院服務量對於CABG術後30日再住院及2010術後併發症無顯著之關係。在病人就醫可近性方面，研究預估PTCA在集中化後病患需跨區就醫的比例約在10-12%間；而在CABG部分則約在18%。此外在低服務量醫院住院醫療收入影響上，CABG部分平均各醫院每年可能將損失2%之住院醫療收入；而PTCA方面則約損失5.5%-6.6%。

結論與建議：整體來說，根據結果顯示集中化政策對於整體照護結果可以帶來正向之影響，並且對於大部分之病患就醫可近性及低服務量醫院財務之影響有限，因此，本研究建議台灣未來在擬定提升及改善心臟相關處置之照護品質之策略時，可以將集中化政策納入可行方案之中，並且深入做更精確之評估。

關鍵字

集中化政策 ； 高風險處置 ； 醫療品質 ； 可避免不良照護結果發生人數 ； 病人就醫可近性 ； 醫療收入

並列摘要

Background: Volume-outcome research techniques are well-developed, and a great deal of evidence from volume-outcome studies has indicated that surgical volume is negatively associated with adverse healthcare outcomes. Many countries have thus initiated health policies which encourage the centralization of surgery. The goal of these policies is to have patients treated at centralized, high-volume surgical centers which employ highly specialized physicians. As these centers typically have access to abundant medical resources, they can provide optimal care and also improve post-care outcomes. We reviewed the operation and effectiveness of surgery centralization policies in many countries and found that they had a positive impact on health care. In recent years, several studies have also revealed a positive correlation between health outcomes and the volume of specific types of surgeries or procedures in Taiwan. Currently, Taiwan does not have a policy of centralization for high-risk surgeries, and whether such a policy can positively impact the healthcare system in Taiwan has not been confirmed. Therefore, the impact of centralization policies must be further evaluated.

Objectives: We sought to estimate how many deaths, unplanned readmissions, and comorbidities could potentially be avoided in Taiwan if a centralization policy was implemented for cardiac procedures. We further examined how a policy which mandated that patients be referred to high-volume hospitals could affect accessibility to care as well as financial losses in low-volume hospitals.

Methods: Data were obtained from Taiwan’s National Health Insurance Research Database. Our retrospective cohort design included patients who underwent Percutaneous Transluminal Coronary Angioplasty (PTCA) or Coronary Artery Bypass Grafting (CABG) for the first time between 2010 and 2012. We used multilevel logistic regression and G-computation to examine volume-outcome relationships and to estimate the number of potentially avoidable adverse healthcare outcomes. To examine the effects on patent accessibility to care, we performed geographic variation and used free Open GeoDA software to predict the cross-boundary flow of cardiac procedures (i.e. patients who would be required to visit a high-volume center as a result of centralization policy). Finally, we predicted the impact that surgery centralization would have on inpatient revenue in low-volume hospitals by calculating the ratio of total cardiac procedure inpatient revenue to total hospital inpatient revenue.

Results: We found that 29,689 of 92,370 patients who underwent PTCA and 4,150 of 9,530 patients who underwent CABG were admitted to low-volume hospitals between 2010 to 2012. We estimated that centralization of PTCA could reduce the number of deaths by 442 (95% confidence interval [CI]: 274-693), the number of readmissions by 620 (95% CI: 471-776), and the number of comorbidities by 766 (95% CI: 607-921). We further estimated that centralization of CABG could reduce the number of deaths by 127 (95% CI: 72-218), the number of readmissions by 66 (95% CI: 30-125), and the number of comorbidities by 271 (95% CI: 201-299). However, we did not find a significant relationship between hospital volume and 30-day readmission or 30-day comorbidity for CABG in 2010. Considering accessibility to care, 10-12% of patients who underwent PTCA and 18% of patients who underwent CABG had to travel farther to access medical care. Finally, we determined that low-volume hospitals incurred financial losses of approximately 5.5-6.6% and 2%, respectively, by referring PTCA and CABG patients to higher-volume surgical centers.

Conclusions: Centralization policies should positively impact healthcare outcomes and only have minor impacts on accessibility to care for patients and minor financial consequences for low-volume hospitals. We therefore suggest that Taiwan consider centralization when developing healthcare-related policies to improve quality of care and the healthcare outcomes for cardiac procedures.

並列關鍵字

Centralization ； high-risk surgery ； quality of care ； potentially avoidable adverse healthcare outcomes ； accessibility to care ； medical revenue

參考文獻

李虹映, 黃信忠, 許怡欣, & 林文德. (2014a). 台灣急重症跨區就醫之變化情形：2001－2010年. [Change in the Cross-Boundary Flow of Emergent Care in Taiwan, 2001-2010]. 臺灣公共衛生雜誌, 33(1), 64-74. doi:10.6288/tjph201433102055

李虹映, 黃信忠, 許怡欣, & 林文德. (2014b). 台灣急重症跨區就醫之變化情形: 2001-2010 年. 臺灣公共衛生雜誌, 33(1), 64-74.

林民浩, 楊安琪, & 溫在弘. (2011). 利用地區差異與人口學特徵評估全民健保資料庫人口居住地變項之推估原則.

林維娟, 張鴻仁, 王本仁, 周穎政, & 李丞華. (2004). 影響住院病患跨區利用之因素. [Cross-Region Hospitalization Behavior and Its Related Factors in Taiwan]. 臺灣公共衛生雜誌, 23(6), 453-461. doi:10.6288/tjph2004-23-06-05

洪維河, 鄭守夏, 張睿詒, & 江東亮. (1998). 臺灣醫療區跨區住院比例之變遷，1985-1995. [Changes in the Proportion of cross-region Admissions in Taiwan, 1985-1995]. 中華公共衛生雜誌, 17(5), 388-394. doi:10.6288/cjph1998-17-05-04