Xu Qin, Hu Chen-Lu, Xiang Hai-Yan, Yang Jian-Wen, Qian Wei-Ming, Song Jian-Ping
Nursing Department, The Second Affiliated Hospital Zhejiang University School of Medicine, Hangzhou, Zhejiang Province, China.
Medicine (Baltimore). 2025 Jul 4;104(27):e43168. doi: 10.1097/MD.0000000000043168.
This study aims to explore potential biomarkers correlated with the clinical outcome only in patients with intraoperative hypothermia undergoing cardiac surgery. The clinical data of patients with or without intraoperative hypothermia, and 8 blood indicators both in preoperative and intraoperative periods, were obtained from the Medical Information Mart for Intensive Care-IV database. The associations of indicators with the length of hospital stay (LOS) were explored in non-hypothermia and hypothermia groups, respectively. The key indicator showing the association only in hypothermia groups was selected for further analysis. The detailed association between the key indicator and LOS was further determined by a series of analyses, including restricted cubic spline, generalized additive model, generalized linear regression, trend regression, threshold effects analysis, and quartile regression. Their clinical value was explored by the receiver operating characteristic analysis and decision curve analysis. Among 16 indicators, only the intraoperative lactate was identified to correlate with the LOS in patients with hypothermia rather than normal group (all P < .05). Lactate showed a positive and nonlinear association with LOS (all P < .05), and their association was especially observed when log-lactate > 2.420 (all P < .05). Quartile regression showed that the associations of lactate with LOS quartiles from 40% and 80% were stably increased when log-lactate > 2.420 (all P < .05). However, no association was observed in any analyses when log-lactate ≤ 2.420. We also confirmed the favorable prediction performance and clinical net benefit of lactate on the LOS. Intraoperative lactate is significantly associated with the LOS of patients with hypothermia and is a useful biomarker.
本研究旨在仅在接受心脏手术的术中体温过低患者中探索与临床结局相关的潜在生物标志物。从重症监护医学信息数据库-IV中获取了有或无术中体温过低患者的临床数据，以及术前和术中的8项血液指标。分别在非体温过低组和体温过低组中探索指标与住院时间（LOS）的关联。选择仅在体温过低组中显示出关联的关键指标进行进一步分析。通过一系列分析进一步确定关键指标与LOS之间的详细关联，包括受限立方样条、广义相加模型、广义线性回归、趋势回归、阈值效应分析和四分位数回归。通过受试者工作特征分析和决策曲线分析探索它们的临床价值。在16项指标中，仅术中乳酸被确定与体温过低患者而非正常组的LOS相关（所有P<0.05）。乳酸与LOS呈正的非线性关联（所有P<0.05），并且当log-乳酸>2.420时尤其观察到它们的关联（所有P<0.05）。四分位数回归显示，当log-乳酸>2.420时，乳酸与40%至80%的LOS四分位数的关联稳定增加（所有P<0.05）。然而，当log-乳酸≤2.420时，在任何分析中均未观察到关联。我们还证实了乳酸对LOS具有良好的预测性能和临床净效益。术中乳酸与体温过低患者的LOS显著相关，是一种有用的生物标志物。
