比较NSQIP手术风险计算器、TRISS与ASA分级对老年创伤患者手术死亡率的预测
该研究为四家一级创伤中心前瞻性收集的老年(65岁及以上)创伤手术患者数据,比较ASA-PS、NSQIP-SRC、TRISS及两种联合模型对死亡、住院时长和并发症的预测能力。结果122例中9例(7.4%)死亡,平均住院时长12.9天(范围1–110天)。NSQIP-SRC预测死亡的AUC为0.978,优于ASA-PS的0.768和TRISS的0.903(p=0.007);预测住院时长R²为25.9%(TRISS 20.5%、ASA-PS 13.3%),预测并发症R²为34.0%(TRISS 29.4%、ASA-PS 22.6%)。将TRISS加入NSQIP-SRC可小幅改善并发症预测(R² 35.5% vs 34.0%,p=0.046),但加入ASA-PS或TRISS未改善死亡或住院时长预测。
要点
- 在122例老年创伤手术患者中,NSQIP-SRC预测死亡的AUC为0.978,优于ASA-PS的0.768和TRISS的0.903(p=0.007)。
- NSQIP-SRC预测住院时长的R²为25.9%,高于TRISS的20.5%和ASA-PS的13.3%。
- NSQIP-SRC预测并发症的R²为34.0%,高于TRISS的29.4%和ASA-PS的22.6%。
- 将TRISS加入NSQIP-SRC可改善并发症预测(R² 35.5% vs 34.0%,p=0.046),但未改善死亡或住院时长预测。
- 该研究证据等级为Level III(预后与流行病学)。
适用人群与场景
- 65岁及以上老年创伤患者,在入院24小时内接受手术。
- 数据来自四家一级创伤中心,前瞻性收集。
研究设计与关键结果
- 研究类型为四家一级创伤中心前瞻性收集数据的预后比较研究,共纳入122例老年创伤手术患者,9例(7.4%)死亡,平均住院时长12.9天(范围1–110天)。
- 比较五种评分模型:ASA-PS、NSQIP-SRC、TRISS、TRISS-ASA-PS、TRISS-NSQIP-SRC,采用logistic回归,以Brier分数和ROC曲线下面积比较死亡预测,以调整R²和均方根误差比较住院时长和并发症数量预测。
- NSQIP-SRC预测死亡的AUC为0.978,优于ASA-PS的0.768和TRISS的0.903(p=0.007)。
- NSQIP-SRC预测住院时长的R²为25.9%,高于TRISS的20.5%和ASA-PS的13.3%。
- NSQIP-SRC预测并发症的R²为34.0%,高于TRISS的29.4%和ASA-PS的22.6%。
- 将TRISS加入NSQIP-SRC后并发症预测R²为35.5%,优于单用NSQIP-SRC的34.0%(p=0.046)。
- 将ASA-PS或TRISS加入NSQIP-SRC未改善死亡或住院时长的预测能力。
证据类型、等级与局限
- 证据类型:前瞻性多中心预后比较研究;证据等级:Level III(预后与流行病学)。
- 局限:原文未报告局限性。
对床旁实践的改变
- 作者认为,包含合并症和功能状态的NSQIP-SRC在预测老年创伤手术患者的死亡、住院时长和并发症方面优于单用TRISS。
提炼说明
- 本条依据论文摘要整理;摘要未涉及「诊断标准与关键阈值」「处置流程与药物剂量」「红旗征与不能漏的情况」「指南或共识出处」,故未列出这些栏目;全文请点「阅读原文」。
英文原文(摘要)· 点击展开
BACKGROUND: The Trauma and Injury Severity Score (TRISS) uses anatomical and physiologic variables to predict mortality. Elderly (65 years or older) trauma patients have increased mortality and morbidity for a given TRISS, in part because of functional status and comorbidities. These factors are incorporated into the American Society of Anesthesiologists Physical Status (ASA-PS) and National Surgical Quality Improvement Program Surgical Risk Calculator (NSQIP-SRC). We hypothesized scoring tools using comorbidities and functional status to be superior at predicting mortality, hospital length of stay (LOS), and complications in elderly trauma patients undergoing operation.
METHODS: Four level I trauma centers prospectively collected data on elderly trauma patients undergoing surgery within 24 hours of admission. Using logistic regression, five scoring models were compared: ASA-PS, NSQIP-SRC, TRISS, TRISS-ASA-PS, and TRISS-NSQIP-SRC.Brier scores and area under the receiver operator characteristics curve were calculated to compare mortality prediction. Adjusted R2 and root mean squared error were used to compare LOS and predictive ability for number of complications.
RESULTS: From 122 subjects, 9 (7.4%) died, and the average LOS was 12.9 days (range, 1-110 days). National Surgical Quality Improvement Program Surgical Risk Calculator was superior to ASA-PS and TRISS at predicting mortality (area under the receiver operator characteristics curve, 0.978 vs. 0.768 vs. 0.903; p = 0.007). Furthermore, NSQIP-SRC was more accurate predicting LOS (R2, 25.9% vs. 13.3% vs. 20.5%) and complications (R2, 34.0% vs. 22.6% vs. 29.4%) compared with TRISS and ASA-PS. Adding TRISS to NSQIP-SRC improved predictive ability compared with NSQIP-SRC alone for complications (R2, 35.5% vs. 34.0%; p = 0.046). However, adding ASA-PS or TRISS to NSQIP-SRC did not improve the predictive ability for mortality or LOS.
CONCLUSION: The NSQIP-SRC, which includes comorbidities and functional status, had superior ability to predict mortality, LOS, and complications compared with TRISS alone in elderly trauma patients undergoing surgery.
LEVEL OF EVIDENCE: Prognostic and Epidemiologic; Level III.
参考来源
- Stopenski S, Kuza CM, Luo X, et al. Comparison of National Surgical Quality Improvement Program Surgical Risk Calculator, Trauma and Injury Severity Score, and American Society of Anesthesiologists Physical Status to predict operative trauma mortality in elderly patients. J Trauma Acute Care Surg. 2022. PMID: 34882598. doi:10.1097/TA.0000000000003481
供执业医师抢救与诊疗参考,不替代临床判断;剂量请结合体重、肝肾功能和本院药品规格复核,以指南原文为准。