Establishment And Verification Of In-hospital Mortality Risk Nomogram Of Sepsis Patients Based On APACHE Ⅱ Score

Objective:In this study,through the analysis of hospitalized sepsis patients,independent risk factors of in-hospital death of sepsis patients were screened,and a nomogram model of 7-day,14-day and 28-day mortality risk of sepsis patients was constructed,so as to provide an efficient risk level screening method for sepsis patients,reduce the mortality of sepsis patients and reduce the economic pressure of patients.Methods:1.According to the inclusion and exclusion criteria,505 patients with sepsis who received treatment in Shanxi People’s Hospital from January 2018 to October 2022 were included.505 patients with sepsis were randomly divided into a training set(n=405 cases)and a test set(n=100 cases)according to the inclusion and exclusion criteria.2.Variables were screened by LASSO regression for baseline data,illness,treatment and laboratory results of patients in the training set.Multivariate Cox regression was used to analyze the screened variables,and independent risk factors for death in sepsis patients were finally screened out(P<0.05).3.Independent risk factors were determined by Cox regression analysis(P<0.05)The7-day,14-day and 28-day mortality risk nomogram prediction models of sepsis patients were constructed with R language 4.2.2 software.4.C-index was used to evaluate the accuracy and ability of the prediction model.Area under ROC curve(AUC)was used to evaluate the model.The calibration curve was used to evaluate the calibration ability,and then the prediction effect of the model was obtained.Evaluation of clinical practicability of the model by decision curve,Kaplan-Meier curve was used to evaluate the actual effect of the model.5.The test set was used for external verification,and C-index,ROC curve,calibration curve,decision curve and Kaplan-Meier curve were also used to evaluate the predictive ability of the model Results:1.By LASSO regression and Cox regression analysis,10 independent risk factors of death in sepsis patients were selected,including septic shock,neutrophil count,platelet distribution width,albumin,indirect bilirubin,D dimer,urinary nitrite,urinary specific gravity,lactic acid,APACHE Ⅱ score(P<0.05).2.Based on the above selected risk factors,the nomogram prediction models of 7-day,14-day and 28-day mortality risk of sepsis patients were successfully constructed.3.The C-index of the training set at 7 days,14 days and 28 days was 83.0,83.0 and81.8,respectively.The C-index of the test set at 7 days,14 days and 28 days were 78.7,78.7and 80.1,respectively.The ROC curve AUC of the model at 7 days,14 days and 28 days of the training set were 0.854,0.844 and 0.812,respectively.The ROC curve AUC of the model at 7 days,14 days and 28 days in the test set were 0.799,0.794 and 0.853,respectively,indicating that the nomogram has a good predictive performance for hospital death risk of sepsis patients.The calibration curves of the training set and the test set showed good consistency between the in-hospital death risk predicted by the nomogram and the actual inhospital death risk,indicating that the calibration and prediction ability of the nomogram were good.The nomogram shows good clinical practicability in the decision curve of both the training set and the test set.Kaplan-Meier curves showed that sepsis patients in both the training set and the test set differed in whether they died during the course of the disease(P<0.05),it can be concluded that the constructed nomogram has good predictive value.Conclusions:1.Septic shock,neutrophil count,platelet distribution width,albumin,indirect bilirubin,D dimer,urinary nitrite,urinary specific gravity,lactic acid,APACHE Ⅱ score were independent risk factors for death in sepsis patients.Strategies that clinicians use to improve these factors after sepsis patients are admitted can improve hospitalization survival.2.The hospital death risk diagram of sepsis patients constructed by independent risk factors screened out shows good predictive efficacy and clinical practicability.In clinical work,clinicians can use the diagram to calculate and get the high risk groups of sepsis patients to die,make targeted treatment,and provide a simple and convenient prognostic means for clinical.