是否醫療服務量之大小確實會影響患者之照護結果？ - 以台灣乳癌醫療照護為例，探討醫療服務提供者之技術效率、過程面品質與服務量對於病患復發與存活狀態之相關性

Does service volume really matter to outcome? Examining the association of process quality, technical efficiency, and surgical volume to breast cancer recurrence and survival

郭年真(Nien-Chen Kuo)

指導教授 ： 鍾國彪

共同指導教授 ： 賴美淑(Mei-Shu Lai)

國立臺灣大學/公共衛生學院/醫療機構管理研究所/博士(2010年)

https://doi.org/10.6342/NTU.2010.00210

摘要

研究背景: 醫療照護服務量與照護結果的關係研究在過去數十年間被廣泛的探討，儘管多數的研究支持服務量大的醫院或醫師其照護結果較佳，但仍有研究者獲得不一致的結果並且質疑以服務量多寡用於預測照護的結果是否可靠。相較於服務量，測量醫療服務提供者的效率不僅關注提供醫療照護的多寡，而且也考慮到提供相關照護所耗用的資源。近期美國衛生部醫療服務中心（CMS）執行的「價值基礎的醫療服務購買計畫」（Value-Based PurchasingPlan）亦即是同時注重改善醫療照護提供者的效率與品質。因此，在抑制醫療照護成本與提升照護品質並重的趨勢下，如何將過去探討醫療服務量與照護結果的關係擴展到評估醫療服務供給者的效率、過程面品質對照護結果的影響，對於醫療保險機構、雇主、被保險人，乃至於醫療照護提供者皆是很重要的議題。材料與方法：本研究將採用多層次的Cox＇s 風險比例迴歸模型，探討醫療服務提供者的過程面品質與效率對乳癌病患存活情形的影響。資料來源為合併台灣癌症資料庫、健保資料庫、死亡登記等次級資料，選取2003 及2004 年首次診斷且接受手術治療的乳癌患者進行研究。過程面品質的測量將採用文獻發展的乳癌核心測量指標，而醫療服務提供者的效率則以資料包絡分析法（DEA）進行分析。統計分析過程中亦同時控制病患的共病狀態與疾病嚴重度。結果：由26 家醫院治療的6,396 位女性乳癌患者被納入此研究。統計分析結果發現，在控制病患、醫院及醫師的基本特性後，醫院或醫師的手術量及醫院的效率與乳癌患者的復發或存活無關。過程面品質測量得分較高的患者，其存活較佳且有較低的復發率。多層次分析結果顯示，在醫院-病患的巢狀樣本結構中存在變動效果(random effect)，但是這樣的效果在醫師-病患的樣本結構中並不明顯。本研究並發現醫院或醫師層級的過程面品質，對於病患層次的過程面品質具有調節效果(moderating effect)。
結論：醫院或醫師手術量與乳癌的復發或存活無關，服務量對於過程面品質亦無存在調節效應。探討服務量及照護結果的研究應該將巢狀樣本結果的特性納入考量，並應用多層次的分析方法。未來相關研究應嘗試以縱貫性研究探討，是否可以藉由改變醫療服務提供者的效率來影響醫療照護的結果。

關鍵字

乳癌 ； 醫療品質 ； 手術量 ； 效率 ； 多層次分析 ； 存活分析

並列摘要

Background: The volume-outcome relationship has been widely discussed over past decades. Although previous studies examined the volume-outcome relationship show that the majority of these works support the reverse relationship of volume and outcome on varied procedures, but other studiess reported inconsistent results and raised arguments if volume is robust enough to predict outcome. Comparing to volume, efficiency not only emphasizes on the quantity of service provided by organizations, but also focus on the resources consumed to generate services. The Value-Based Purchasing Plan carried out by CMS recently also emphasizing on improving efficiency as well as clinical quality. Therefore, extending volume-outcome association to measuring efficiency as well as process quality of care, and linking the performance with outcome have been important issues to not only health purchasers, but also providers and consumers when pursuing cost containment and quality improvement in health care. Material and method: This study is going to explore the relationship between process quality, efficiency and survival as well as recurrence in breast cancer care by Cox’s regression and multilevel modeling, with using Taiwan Cancer Database combined with Taiwan National Health Insurance Database (NHID) and death registry. Breast cancer patients diagnosed in 2003 to 2004 and received surgical treatment will be included in this study. Process quality will be measured by a set of core measure indicators, and constant returns to scale (CRS) input oriented data envelope analysis (DEA) method will be used for measuring efficiency of individual hospital. Comorbidity and severity of illness will also be controlled in analysis. Result: 6,396 female breast cancer patients, reported by 26 hospitals, were included in this study. After controlled for patient and provider characteristics, hospital and physician’s surgical volume for breast cancer as well as hospital efficiency are not associated with the patient survival or recurrence. Patients received care with high process quality is associated with better survival and lower recurrence. The multilevel analyses found random effect for the hospital-patients clustered sample, but the effect for physician-patient clusters is not as significant. This study also found the provider’s process quality also has the moderating effect for patient-level process quality. Conclusion: Volume is not associated with breast cancer recurrence or survival, nor it has moderating effect on process quality. The clustered feature of data should be considered for volume-outcome related studies. Future researches should apply a longitudinal design to explore how quality related health outcomes can be affected by a change in provider efficiency.

並列關鍵字

breast cancer ； quality of care ； efficiency ； volume-outcome relationship ； multilevel analysis ； survival analysis

參考文獻

147 .Raudenbush, Stephen W., Bryk, Anthony S., & Congdon, Richard T. (2008). HLM 6: Hierarchical linear and nonlinear modeling Scientific Software International Inc.

69 .Hewitt, Maria Elizabeth, Simone, Joseph V., & National Cancer Policy Board (U.S.) (1999). Ensuring quality cancer care. Washington, D.C.: National Academy Press.

186 .Yang, Min, Rasbash, Jon, Goldstein, Harvey, & Barbosa, Maria (2001). MLwiN Macros for advanced Multilevel modelling Available from http://www.cmm.bristol.ac.uk/MLwiN/download/index.shtml

1 .Aguado, A., Guino, E., Mukherjee, B., Sicras, A., Serrat, J., Acedo, M., et al. (2008). Variability in prescription drug expenditures explained by adjusted clinical groups (ACG) case-mix: A cross-sectional study of patient electronic records in primary care. [Article]. Bmc Health Services Research, 8.

2 .Allgood, P. C., & Bachmann, M. O. (2006). Effects of specialisation on treatment and outcomes in screen-detected breast cancers in Wales: cohort study. [Article]. British Journal of Cancer, 94(1), 36-42.