目標：本文利用2000-2005年全民健保資料庫之冠狀動脈繞道手術(CABG)病患資料，探討醫院手術量與病患療效之關係。二者的關係是因醫院的熟能生巧效果？亦或是各醫院的醫療品質差異？方法：本文利用Cox比例危險模型探討病患死亡危險率之影響因素。另外，為考量各醫院因醫療品質差異對病患療效之影響，在病患的存活模型中加入各醫院的個別效果。結果：在病患的存活模型中發現每增加1%的醫院手術量將使病患死亡的危險率下降17.5%。但是，若在病患的存活模型中加入各醫院的個別效果，以醫院手術量與病患療效的時間變化量衡量二者關係，將發現二者的關係不再顯著。結論：醫院手術量與病患療效的關係主要是反映各醫院醫療品質的差異，並非是熟能生巧效果。當醫院有較佳的醫療品質時，不僅使其病患有較佳的療效，也因常被選擇與推薦而擁有高手術量，使得手術量與病患療效產生正向關係。

Objectives: To investigate the surgical volume-outcome relationships for patients undergoing coronary artery bypass graft (CABG) surgery. Methods: We analyzed the records in Taiwan’s National Health Insurance Research Database for CABG patients during the period 2000-2005. Cox proportional hazard was used to identify factors relevant to patient survival. The hospitalspecific effects are also included in the model to account for quality differences between hospitals. Results: Our initial estimates showed that surgical volume was negatively associated with hazard rate such that a 1% increase in surgical volume was associated with a 17.5% lower probability of dying in the hospital. However, when we re-estimated the model accounting for quality differences between hospitals by including hospital-specific effects, the effect of volume disappeared. Conclusions: The volume-outcome relationship for CABG patients appears to reflect quality differences between high-and low-volume hospitals, rather than the principle of "practice makes perfect."