From the Institute of Health Policy, Management, and Evaluation (M.P.G., B.W.T., B.H., A.B.N.), Department of Surgery (M.P.G., B.H., A.B.N.), Interdepartmental Division of Critical Care Medicine, Department of Medicine (B.W.T., B.H.), University of Toronto, Toronto, Ontario, Canada; and Sunnybrook Research Institute (B.H., A.B.N.), Toronto, Ontario, Canada.
J Trauma Acute Care Surg. 2020 Oct;89(4):743-751. doi: 10.1097/TA.0000000000002892.
Mortality is an important trauma center outcome. With many patients initially surviving catastrophic injuries and a growing proportion of geriatric patients, many deaths might occur following withdrawal of life-sustaining therapy (WLST). We utilized the American College of Surgeons Trauma Quality Improvement Program database to explore whether deaths following WLST might be preventable and to evaluate the impact of excluding patients who died following WLST on hospital performance.
A retrospective cohort study was conducted using data derived from American College of Surgeons Trauma Quality Improvement Program. Adult trauma patients treated at Levels I and II centers in 2016 were included. Three cohorts of deceased patients were created to assess differences in hospital performance. The first included all deaths, the second included only those who died without WLST, and the third included deaths without WLST and deaths with WLST where death was preceded by a major complication. Hospitals were ranked based on their observed-to-expected mortality ratio calculated using each of the three decedent cohorts. Outcomes included absolute change in hospital ranking and change in performance outlier status between cohorts.
We identified 275,939 patients treated at 447 centers who met inclusion criteria. Overall mortality was 6.9% (n = 19,145). Withdrawal of life-sustaining therapy preceded 43.6% (n = 8,343) of deaths and 23% (n = 1,920) of these patients experienced a major complication before death. The median absolute change in hospital performance rank between the first and second cohort was 58 (p < 0.001), between the first and third cohort was 44 (p < 0.001), and between the second and third cohort was 23 (p < 0.001). Hospital performance outlier status changed significantly between cohorts.
The exclusion of patients who die following WLST from benchmarking efforts leads to a major change in hospital ranks. Potentially preventable deaths, such as those following a major complication, should not be excluded.
Epidemiological study, level III.
死亡率是创伤中心的一个重要结果。随着许多患者最初从灾难性损伤中幸存下来，以及老年患者比例的增加，许多患者可能会在停止生命支持治疗（WLST）后死亡。我们利用美国外科医师学院创伤质量改进计划数据库探讨 WLST 后死亡是否可以预防，并评估排除 WLST 后死亡患者对医院绩效的影响。
使用美国外科医师学院创伤质量改进计划数据库进行回顾性队列研究。纳入 2016 年在 I 级和 II 级中心接受治疗的成年创伤患者。创建了三个死亡患者队列，以评估医院绩效的差异。第一个队列包括所有死亡患者，第二个队列仅包括那些没有 WLST 治疗而死亡的患者，第三个队列包括没有 WLST 治疗而死亡的患者和 WLST 治疗而死亡的患者，其中死亡前发生了主要并发症。根据每个队列中观察到的死亡率与预期死亡率之比对医院进行排名。结果包括医院排名的绝对变化和队列间绩效异常状态的变化。
我们确定了 447 家中心收治的 275939 名符合纳入标准的患者。总死亡率为 6.9%（n=19145）。WLST 前有 43.6%（n=8343）的死亡患者，其中 23%（n=1920）的患者在死亡前发生了主要并发症。第一和第二队列之间医院绩效排名的中位数绝对变化为 58（p<0.001），第一和第三队列之间为 44（p<0.001），第二和第三队列之间为 23（p<0.001）。队列间医院绩效异常状态发生了显著变化。
将 WLST 后死亡的患者排除在基准测试之外会导致医院排名发生重大变化。不应该排除潜在可预防的死亡，如发生主要并发症后的死亡。
流行病学研究，III 级。
