乳腺癌改良根治术中发生低体温风险列线图预测模型的构建

Construction of a nomogram prediction model for the risk of hypothermia during modified radical mastectomy

目的 分析乳腺癌改良根治术中发生低体温的危险因素,并基于危险因素构建预测乳腺癌改良根治术中发生低体温的风险列线图模型。方法 选取我院收治的383例行乳腺癌改良根治术患者为研究对象,根据是否发生低体温分为低体温组(n=58)和正常组(n=325)。收集患者的临床资料,采用单因素和Logistic回归分析筛选出术中发生低体温的独立危险因素,采用R软件构建预测术中发生低体温的风险列线图模型,并对模型进行验证。结果 2组患者甲状腺功能减退、术前基础体温、术中室温、手术时间、麻醉时间、术中出血量、术中输液量等资料比较差异有统计学意义(P<0.05)。甲状腺功能减退(OR=2.156,95%CI:1.158~4.016,P=0.015)、术前基础体温异常(OR=2.451,95%CI:1.309~4.588,P=0.005)、术中室温<23℃(OR=2.027,95%CI:1.085~3.786,P=0.027)、手术时间>2 h(OR=2.316,95%CI:1.239~4.327,P=0.008)、麻醉时间>3 h(OR=2.264,95%CI:1.206~4.252,P=0.011)、术中输液量>1 500 mL(OR=2.895,95%CI:1.543~5.432,P=0.001)是术中发生低体温的独立危险因素。列线图模型显示,术中输液量>1 500 mL为100分,甲状腺功能减退为93分,麻醉时间>3 h为85分,术中室温<23℃为84分,手术时间>2 h为79分,术前基础体温异常为75分。对列线图模型进行验证,结果显示,C-index为0.834;H-L拟合优度检验χ2=11.854,P=0.078;校正曲线趋近于理想曲线;受试者工作特征曲线下面积为0.812;阈值概率在5%~70%时,净获益值较高。结论 本研究构建的列线图模型能够较准确、可靠地预测乳腺癌改良根治术中低体温发生的风险,满足了临床上对于整合模型的需求,有助于推动个体化医疗的稳步发展。

Objective To analyze the risk factors for hypothermia during modified radical mastectomy,and construct a nomogram model for predicting the occurrence of hypothermia during modified radical mastectomy based on the risk factors.Methods A total of 383 patients received modified radical mastectomy admitted to our hospital were selected and divided into the hypothermia group(n=58)and the normal group(n=325)according to whether hypothermia occurred.The clinical data of the patients were collected,and the univariate analysis and Logistic regression analysis were used to screen out the independent risk factors for intraoperative hypothermia,and a risk nomogram model for predicting intraoperative hypothermia was constructed by R software and verified.Results There were statistically significant differences in the hypothyroidism,preoperative basal body temperature,intraoperative room temperature,operation time,anesthesia time,intraoperative blood loss,and intraoperative infusion between the two groups(P<0.05).The hypothyroidism(OR=2.156,95%CI:1.158~4.016,P=0.015),abnormal preoperative basal body temperature(OR=2.451,95%CI:1.309~4.588,P=0.005),intraoperative room temperature<23℃(OR=2.027,95%CI:1.085~3.786,P=0.027),operation time>2 hours(OR=2.316,95%CI:1.239~4.327,P=0.008),anesthesia time>3 hours(OR=2.264,95%CI:1.206~4.252,P=0.011),intraoperative infusion volume>1500 mL(OR=2.895,95%CI:1.543~5.432,P=0.001)were the independent risk factors for the occurrence of intraoperative hypothermia.The nomogram model showed that the score of intraoperative infusion volume>1500 mL was 100 points,hypothyroidism was 93 points,anesthesia time>3 hours was 85 points,intraoperative room temperature<23℃was 84 points,operation time>2 hours was 79 points,and abnormal preoperative basal body temperature was 75 points.The nomogram model verification results demonstrated that the C-index was 0.834;the H-L goodness-of-fit test showed χ^(2)=11.854 and P=0.078;the calibration curve was close to the ideal curve;the area under the receiver operating characteristic curve was 0.812;and the net benefit value was high at the threshold probability from 5%to 70%.Conclusion The nomogram model constructed in this study can more accurately and reliably predict the risk of hypothermia occurring during modified radical mastectomy,which meets the clinical need for an integrated model and helps to promote the steady development of individualized medicine.