临床衰弱量表作为短期死亡率的预测指标：关于诊断试验准确性研究的系统评价和荟萃分析
Clinical Frailty Scale as a predictor of short-term mortality: A systematic review and meta-analysis of studies on diagnostic test accuracy.
作者信息
Lee Ji Hwan, Park Yoo Seok, Kim Min Joung, Shin Hye Jung, Roh Yun Ho, Kim Ji Hoon, Chung Hyun Soo, Park Incheol, Chung Sung Phil
机构信息
Department of Emergency Medicine, Yonsei University College of Medicine, Seoul, Republic of Korea.
Department of Biomedical Systems Informatics, Biostatistics Collaboration Unit, Yonsei University College of Medicine, Republic of Korea.
出版信息
Acad Emerg Med. 2022 Nov;29(11):1347-1356. doi: 10.1111/acem.14493. Epub 2022 Apr 25.
BACKGROUND
The Clinical Frailty Scale (CFS) is a representative frailty assessment tool in medicine. This systematic review and meta-analysis aimed to examine whether frailty defined based on the CFS could adequately predict short-term mortality in emergency department (ED) patients.
METHODS
The PubMed, EMBASE, and Cochrane libraries were searched for eligible studies until December 23, 2021. We included studies in which frailty was measured by the CFS and short-term mortality was reported for ED patients. All studies were screened by two independent researchers. Sensitivity, specificity, positive likelihood ratio (PLR), and negative likelihood ratio (NLR) values were calculated based on the data extracted from each study. Additionally, the diagnostic odds ratio (DOR) was calculated for effect size analysis, and the area under the curve (AUC) of summary receiver operating characteristics was calculated. Outcomes were in-hospital and 1-month mortality rate for patients with the CFS scores of ≥5, ≥6, and ≥7.
RESULTS
Overall, 17 studies (n = 45,022) were included. Although there was no evidence of publication bias, a high degree of heterogeneity was observed. For the CFS score of ≥5, the PLR, NLR, and DOR values for in-hospital mortality were 1.446 (95% confidence interval [CI] 1.325-1.578), 0.563 (95% CI 0.355-0.893), and 2.728 (95% CI 1.872-3.976), respectively. In addition, the pooled statistics for 1-month mortality were 1.566 (95% CI 1.241-1.976), 0.582 (95% CI 0.430-0.789), and 2.696 (95% CI 1.673-4.345), respectively. Subgroup analysis of trauma patients revealed that the CFS score of ≥5 could adequately predict in-hospital mortality (PLR 1.641, 95% CI 1.242-2.170; NLR 0.580, 95% CI 0.461-0.729; DOR 2.883, 95% CI 1.994-4.168). The AUC values represented sufficient to good diagnostic accuracy.
CONCLUSIONS
Evidence that is published to date suggests that the CFS is an accurate and reliable tool for predicting short-term mortality in emergency patients.
背景
临床衰弱量表（CFS）是医学领域中一种具有代表性的衰弱评估工具。本系统评价和荟萃分析旨在研究基于CFS定义的衰弱能否充分预测急诊科（ED）患者的短期死亡率。
方法
检索了PubMed、EMBASE和Cochrane图书馆，以查找符合条件的研究，检索截至2021年12月23日。我们纳入了使用CFS测量衰弱情况并报告了ED患者短期死亡率的研究。所有研究均由两名独立研究人员进行筛选。根据从每项研究中提取的数据计算敏感性、特异性、阳性似然比（PLR）和阴性似然比（NLR）值。此外，计算诊断比值比（DOR）用于效应量分析，并计算汇总受试者工作特征曲线下面积（AUC）。结局指标为CFS评分≥5、≥6和≥7的患者的住院死亡率和1个月死亡率。
结果
总体而言，共纳入17项研究（n = 45,022）。虽然没有证据表明存在发表偏倚，但观察到高度异质性。对于CFS评分≥5的患者，住院死亡率的PLR、NLR和DOR值分别为1.446（95%置信区间[CI] 1.325 - 1.578）、0.563（95% CI 0.355 - 0.893）和 2.728（95% CI 1.872 - 3.976）。此外，1个月死亡率的合并统计量分别为1.566（95% CI 1.241 - 1.976）、0.582（95% CI 0.430 - 0.789）和2.696（95% CI 1.673 - 4.345）。创伤患者的亚组分析显示，CFS评分≥5能够充分预测住院死亡率（PLR 1.641，95% CI 1.242 - 2.170；NLR 0.580，95% CI 0.461 - 0.729；DOR 2.883，95% CI 1.994 - 4.168）。AUC值表明诊断准确性足够好。
结论
迄今为止发表的证据表明，CFS是预测急诊患者短期死亡率的准确可靠工具。
Zotero 插件