腹腔镜结直肠癌手术术中低体温预警模型的构建与验证

Construction and validation of hypothermia early warning model during laparoscopic colorectal cancer surgery

目的对腹腔镜结直肠癌患者术中低体温危险因素的预警模型进行构建并验证,旨在指导临床预防措施。方法选取该院2020年1月至2021年12月行腹腔镜手术治疗的结直肠癌475例患者,术中出现低体温83例为验证组,其余非低体温患者为对照组(n=392)。收集两组一般资料,采用单因素分析及多因素分析腹腔镜结直肠癌患者术中低体温危险因素,构建预警模型并验证其准确度。结果475例结直肠癌患者术中低体温发生率为17.47%(83/475)。经单因素与多因素logistic回归分析,术中低体温独立危险因素包括:麻醉时间大于3 h、手术室温度小于或等于23℃、BMI≤24 kg/m^(2)、年龄大于60岁、术中补液量大于1000 mL(P<0.05)。预警模型:logistic(P)=-2.213+2.580×(年龄)+1.608×(BMI)+1.460×(手术室温度)+1.254×(麻醉时间)+1.119×(术中输液量)。受试者工作特征曲线下面积为0.709(95%CI:0.631~0.789),校准曲线显示其与实际曲线一致性良好,表明该模型具有较好的区分度,预测拟合度较好(P<0.001)。Hosmer-lemeshow结果:χ^(2)=162.181,P<0.001,表明预测较准确。结论构建预警模型具有良好的精确度和区分度,可以有效预测术中发生低体温的概率。

Objective To construct and verify the early warning model of the hypothermic risk factors during laparoscopic surgery in the patients with colorectal cancer in order to guide the clinical preventive measures.Methods A total of 475 patients with colorectal cancer undergoing laparoscopic surgery in this hospital from January 2020 to December 2021 were selected.Eighty-three patients with hypothermia during surgery served as the validation group(n=83) and the other served as the non-hypothermia control group(n=392).The general data in the two groups were collected.The univariate analysis and multivariate analysis were adopted to analyze the risk factors of hypothermia during laparoscopic surgery in the patients with colorectal cancer.Then the early warning model was constructed and its accuracy was verified.Results The incidence rate of hypothermia during surgery in 475 patients with colorectal cancer was 17.47%(83/475).The univariate and multivariate logistic regression analysis results showed that the anesthesia time >3 h,operating room temperature ≤23 ℃,BMI≤24 kg/m^(2),age>60 years old,intraoperative fluid infusion volume >1 000 mL were the independent risk factors for intraoperative hypothermia(P<0.05).The early warning model:logistic(P)=-2.213+2.580×(age) +1.608×(BMI)+1.460×(operating room temperature)+1.254×(anesthesia time)+1.119×(intraoperative fluid infusion volume).The area under the curve(AUC) of the receiver operating characteristic(ROC) curve was 0.709(95%CI 0.631-0.789).The calibration curve showed that the calibration curve was in good agreement with the actual curve,indicating that the model had good discrimination and good prediction fitting(P<0.001).The result of Hosmer-lemeshow test was χ^(2)=162.181,P<0.001,indicating that the prediction was more accurate.Conclusion The constructed early warning model has good accuracy and discrimination,which could effectively predict the probability of hypothermia during surgery.