Department of Surgery, University of Pittsburgh, Pittsburgh, PA.
Wolff Center at UPMC, University of Pittsburgh Medical Center, Pittsburgh, PA.
Ann Surg. 2021 Dec 1;274(6):e1230-e1237. doi: 10.1097/SLA.0000000000003808.
The goal of this project was to first address barriers to implementation of the Risk Analysis Index (RAI) within a large, multi-hospital, integrated healthcare delivery system, and to subsequently demonstrate its utility for identifying at-risk surgical patients.
Prior studies demonstrate the validity of the RAI for evaluating preoperative frailty, but they have not demonstrated the feasibility of its implementation within routine clinical practice.
Implementation of the RAI as a frailty screening instrument began as a quality improvement initiative at the University of Pittsburgh Medical Center in July 2016. RAI scores were collected within a REDCap survey instrument integrated into the outpatient electronic health record and then linked to information from additional clinical datasets. NSQIP-eligible procedures were queried within 90 days following the RAI, and the association between RAI and postoperative mortality was evaluated using logistic regression and Cox proportional hazards models. Secondary outcomes such as inpatient length of stay and readmissions were also assessed.
RAI assessments were completed on 36,261 unique patients presenting to surgical clinics across five hospitals from July 1 to December 31, 2016, and 8,172 of these underwent NSQIP-eligible surgical procedures. The mean RAI score was 23.6 (SD 11.2), the overall 30-day and 180-day mortality after surgery was 0.7% and 2.6%, respectively, and the median time required to collect the RAI was 33 [IQR 23-53] seconds. Overall clinic compliance with the recommendation for RAI assessment increased from 58% in the first month of the study period to 84% in the sixth and final month. RAI score was significantly associated with risk of death (HR=1.099 [95% C.I.: 1.091 - 1.106], p < 0.001). At an RAI cutoff of ≥37, the positive predictive values for 30- and 90-day readmission were 14.8% and 26.2%, respectively, and negative predictive values were 91.6% and 86.4%, respectively.
The RAI frailty screening tool can be efficiently implemented within multi-specialty, multi-hospital healthcare systems. In the context of our findings and given the value of the RAI in predicting adverse postoperative outcomes, health systems should consider implementing frailty screening within surgical clinics.
本项目的目标首先是解决在大型多医院综合医疗服务系统中实施风险分析指数（RAI）的障碍，然后证明其用于识别高危手术患者的效用。
先前的研究表明 RAI 评估术前虚弱的有效性，但尚未证明其在常规临床实践中的实施可行性。
RAI 作为虚弱筛查工具的实施始于 2016 年 7 月匹兹堡大学医学中心的一项质量改进计划。RAI 评分通过集成到门诊电子健康记录中的 REDCap 调查工具收集，然后与来自其他临床数据集的信息相关联。在 RAI 后 90 天内查询 NSQIP 合格的手术程序，并使用逻辑回归和 Cox 比例风险模型评估 RAI 与术后死亡率之间的关联。还评估了住院时间和再入院等次要结果。
2016 年 7 月 1 日至 12 月 31 日，五个医院的外科诊所共完成了 36261 名独特患者的 RAI 评估，其中 8172 名患者接受了 NSQIP 合格的手术。平均 RAI 评分为 23.6（SD 11.2），术后 30 天和 180 天的总体死亡率分别为 0.7%和 2.6%，收集 RAI 的中位数时间为 33 [IQR 23-53] 秒。总体而言，诊所对 RAI 评估建议的遵守率从研究期间的第一个月的 58%增加到第六个月（最后一个月）的 84%。RAI 评分与死亡风险显著相关（HR=1.099[95%CI：1.091-1.106]，p<0.001）。在 RAI 截断值≥37 的情况下，30 天和 90 天再入院的阳性预测值分别为 14.8%和 26.2%，阴性预测值分别为 91.6%和 86.4%。
RAI 虚弱筛查工具可以在多专科、多医院的医疗保健系统中有效地实施。根据我们的研究结果和 RAI 在预测不良术后结果方面的价值，卫生系统应考虑在外科诊所中实施虚弱筛查。
