| Objectives1.To construct a risk prediction model for intraoperative hypothermia in patients undergoing radical gastrointestinal cancer surgery.2.To validate the efficacy of the intraoperative hypothermia risk prediction model in patients undergoing radical gastrointestinal cancer surgery.Methods1.Through a systematic literature search on the influencing factors of intraoperative hypothermia in patients undergoing radical gastrointestinal cancer surgery,relevant influencing factors were initially identified.A correspondence questionnaire was compiled,and the influencing factors included in this study were finally determined after two rounds of expert correspondence.Based on the requirements of the sample size of the clinical risk prediction model,276 patients who underwent radical gastrointestinal cancer surgery from November 2020 to June 2021 in a tertiary care hospital in Qingdao were selected as the modeling group.And 138 patients who underwent radical gastrointestinal cancer surgery from July to September 2021 were selected as the validation group.Using the data of the validation group for external validation of the model.2.Using univariate analysis,the general data of 130 hypothermic patients in the modeling group were compared with 146 non-hypothermic patients to initially screen out the influencing factors for intraoperative hypothermia.The influencing factors with statistically significant differences in the results of the univariate analysis were included in the binary logistic regression analysis to screen out the independent influencing factors of intraoperative hypothermia,and then construct the risk prediction model.The R language was applied to draw a column line graph to visualize the model results for clinical application.3.138 patients in the validation group were selected to validate the model differentiation and calibration.The ROC curve was plotted,and the area under the ROC curve was used to evaluate the discrimination of the model and determine the prediction effect of the model.The Hosmer-Lemeshow test was used to determine the fitting effect of the model,and P > 0.05 was considered as no statistically significant difference.The calibration degree of the model was tested by combining the Brier score and calibration curve.Results1.Based on a literature search,factors influencing intraoperative hypothermia in patients undergoing radical gastrointestinal cancer were initially identified.Then using the Delphi method,after two rounds of expert consultation,22 variables were included: age,BMI,basal body temperature,American Society of Anesthesiologists classification,nutritional risk screening,preoperative hemoglobin level,preoperative albumin level,duration of dietary abstinence,basal metabolic rate,mode of anesthesia,mode of surgery,duration of surgery,duration of anesthesia,bleeding volume,blood transfusion volume,flushing fluid volume,fluid infusion volume,crystalloid volume,colloidal fluid volume,the operating room temperature,surgical site,and insulation method.2.Univariate analysis of the general data of 276 patients in the modeling group showed that age,BMI,basal body temperature,basal metabolic rate,operation time,anesthesia time,bleeding volume,flushing fluid volume,infusion volume,colloid fluid volume,operating room temperature,operation site,and insulation method were the influencing factors of intraoperative hypothermia in radical gastrointestinal cancer surgery.Binary logistic regression analysis showed that age,BMI,operation time,bleeding volume,colloid fluid volume,operation site,and insulation method were independent influencing factors for the occurrence of intraoperative hypothermia in patients undergoing radical gastrointestinal cancer surgery(P<0.05).Among them,advanced-age,long operation time,high bleeding volume,high colloid fluid volume,and rectal surgery being independent risk factors,and high BMI and insulation equipment being independent protective factors.3.Based on the results of binary logistic regression analysis,a column line graph model was drawn using Software.The probability of intraoperative hypothermia was obtained by adding all the scores together and the probability corresponding to the total score was the probability of intraoperative hypothermia.The higher the total score,the higher the probability of intraoperative hypothermia.4.138 patients were included in the validation group;72 patients did not experience intraoperative hypothermia and 66 patients did.The constructed risk prediction model predicted that 77 patients would not experience hypothermia and 61 patients would experience hypothermia.Comparing the prediction results with the actual results,the sensitivity of the prediction model constructed in this study was 69.70%,the specificity was79.17%,and the accuracy was 74.64%.The area under the ROC curve of the risk prediction model in the validation group was 0.820,with a 95% confidence interval of [0.752,0.889],which had good discrimination.The results of the Hosmer-Lemeshow test showed P =0.139 > 0.05,with no statistically significant difference,indicating a good fit.The calibration curve showed that the Brier score of the model was 0.174 score(<0.25),and the predicted and measured values are in good agreement.Conclusions1.Advanced-age,long operation time,high bleeding,high colloidal fluid volume,and rectal surgery are independent risk factors for the occurrence of intraoperative hypothermia in patients undergoing radical gastrointestinal cancer surgery,and high BMI and holding device are independent protective factors2.The risk prediction model for intraoperative hypothermia in radical gastrointestinal cancer patients constructed in this study has good predictive efficacy with good differentiation and calibration.This model is convenient to guide medical and nursing staff to identify the risk factors and high-risk groups of intraoperative hypothermia in radical gastrointestinal cancer patients at an early stage.And to intervene in high-risk patients as early as possible to prevent the occurrence of intraoperative hypothermia. |
PDF Full Text Request