運用全民健康保險資料庫進行連續性腎臟替代療法的預後研究

Analysis of Outcomes of Continuous Renal Replacement Therapy Using Taiwan National Health Insurance Research Database

柯畊竹(Keng-Chu Ko)

指導教授 ： 張淑惠

國立臺灣大學/公共衛生學院/公共衛生碩士學位學程/碩士(2016年)

https://doi.org/10.6342/NTU201702923

摘要

背景與研究目的
連續性腎臟替代療法是加護病房中針對急性腎損傷病患重要的治療工具，本研究目的在藉由健保資料庫資料了解加護病房連續性腎臟替代療法的本土性流行病學資料及其存活特性。
研究方法
利用2005-2013年間的健保資料庫資料，探討加護病房中連續性性腎臟替代療法的使用情況與病患特性，以住院死亡及28天死亡為主要結果，入院時間至死亡為存活時間，若出院未死亡或超過28天未死亡則視為設限;並利用Cox比例風險模式 (Cox proportional hazards models) 進行單變項分析及多變項分析，並利用共享脆弱模式 (shared frailty model) 調整未知因素之影響，探討連續性腎臟替代療法對於加護病房病患預後的影響、不同模式的連續性腎臟替代療法對於加護病房病患預後的比較以及升壓劑對於加護病房病患及連續性腎臟替代療法病患預後的影響。
結果
在2005-2013年間，全國執行連續性腎臟替代療法的次數在2005-2010年間有顯著增加，之後維持大致相同的趨勢，各縣市間的執行率則有顯著差異。
連續性腎臟替代療法和加護病房患者的死亡風險增加有關，經調整干擾因子後，住院死亡風險比為1.367 (95%信賴區間: 1.347-1.387)。升壓劑使用上，每日每增加10 defined daily doses (DDD) 的升壓劑使用，住院死亡風險比為1.002 (95%信賴區間: 1.001-1.004)。而升壓劑使用的種類由一種到四種以上，其住院死亡風險比分別為4.754 (95%信賴區間: 4.694-4.814)、14.189 (95%信賴區間: 14.020-14.361)、21.222 (95%信賴區間: 20.959-21.488) 和21.847 (95%信賴區間: 21.450-22.252)。
在連續性腎臟替代療法的族群中，continuous veno-venous hemofiltration (CVVH) 和continuous veno-venous hemodialysis (CCVVHD)為主要模式，此二模式的住院死亡風險差異並不顯著 (CVVH vs. CVVHD 風險比: 0.970，95%信賴區間: 0.937-1.003)。每日每增加10 DDD的升壓劑使用，其住院死亡風險比為1.002 (95%信賴區間: 1.001-1.004)。由一種升壓劑到四種以上升壓劑使用，其死亡風險比分別為3.608 (95%信賴區間: 3.114-4.181)、6.372 (95%信賴區間: 5.538-7.331)、8.280 (95%信賴區間: 7.203-9.518) 和8.022 (95%信賴區間6.966-9.238)。
結論：
連續性腎臟替代療法的執行隨地區別有所不同。在所有加護病房的患者中，有接受連續性腎臟替代療法的病患有較高的死亡風險。CVVH及CVVHD在住院死亡風險上並沒有顯著差異。而升壓劑的種類在所有加護病房病患及有接受連續性腎臟替代療法的病患中，其對死亡風險的影響較升壓劑總量來的大。

關鍵字

全民健保資料庫 ； 連續性腎臟替代療法 ； 加護病房 ； 升壓劑 ； 存活分析 ； 共享脆弱模式

並列摘要

Background and purposes
Continuous renal replacement therapy (CRRT) is an important tool for treatment of acute kidney injury in intensive care units (ICU) .The aim of the study is to explore epidemiologic characteristics of CRRT and its impact of the inpatient survival outcomes in Taiwan.
Methods
We analyzed data extracted from national health insurance research database (NHIRD) from 2005-2013 to investigate demographic data and outcomes of ICU patients. The primary endpoints were in-hospital mortality; and the corresponding censoring events were defined as survival to discharge or survival of the 28th day or discharge. The observed survival time was then from admission to either the primary endpoint or censoring. The impacts of CRRT on the mortality of ICU patients and the effects of vasoactive agents on ICU patients receiving CRRT were studied using the Cox proportional hazard models in the univariable and multivariable analyses.
Results
The incidence of CRRT increased from 2005 to 2010 and kept at a stable level from then on. The incidences of practice were quite different between different counties.
CRRT was associated with a higher in-hospital mortality of ICU patients. The adjusted hazard ratio (aHR) of CRRT and non-CRRT group was 1.367 (95% CI: 1.347-1.387). In all ICU patients, every 10 defined daily doses (DDD) increase of daily vasoactive agent usage was associated with aHR of 1.002 (95%CI: 1.001-1.004). The risk of in-hospital mortality increased as the number of the types of applied vasoactive agents increased. From one type of vasoactive agents to 4 types, the aHRs were 4.754 (95%CI: 4.694-4.814) , 14.189 (95%CI: 14.020-14.361), 21.222 (95%CI: 20.959-21.488), and 21.847 (95%CI: 21.450-22.252), respectively.
For patients receiving CRRT, there is no significant difference between CVVH and CVVHD groups (CVVH vs. CVVHD, aHR: 0.970, 95%CI: 0.937-1.003). Every 10 DDD increase of daily vasoactive agent usage was associated with a higher risk of in-hospital mortality (aHR=1.002,(95%CI: 1.001-1.004). The risk of in-hospital mortality also increased as kinds of vasoactive agents for treatment increased. From 1 kind of vasoactive agents to 4 kinds, aHRs were 3.608 (95%CI: 3.114-4.181), 6.372 (95%CI: 5.538-7.331), 8.280 (95%CI: 7.203-9.518), and 8.022 (95%CI: 6.966-9.238), respectively.
Conclusion
The incidence of CRRT usage was different between counties. In all ICU patients, those who received CRRT had a higher risk of in-hospital mortality. The number of types of vasoactive agents had a strong effect on the risk of in-hospital mortality than the total dosage of vasoactive agents for all ICU patients and CRRT patients.

並列關鍵字

National Health Insurance Research Database ； Continuous renal replacement therapy ； Intensive care units ； Vasoactive agents ； Survival analysis ； Shared frailty model

參考文獻

1. Yang Li. Acute Kidney Injury in Asia. Kidney Dis. 2, 95–102 (2016).

2. Khajehdehi, P. Turmeric: Reemerging of a neglected Asian traditional remedy. J Nephropathol. 1, 17–22 (2012).

3. Amdur, R. L., Chawla, L. S., Amodeo, S., Kimmel, P. L. Palant, C. E. Outcomes following diagnosis of acute renal failure in U.S. veterans: focus on acute tubular necrosis. Kidney Int. 76, 1089–1097 (2009).

4. Cohen, S. D. Kimmel, P. L. Long-term sequelae of acute kidney injury in the ICU. Curr. Opin. Crit. Care 18, 1 (2012).

5. Graham, T. The Bakerian Lecture: On Osmotic Force. Philos. Trans. R. Soc. London 144, 177–228 (1854).