Establishment And Validation Of Nomogram For Patients With Sepsis Based On NEWS Score

Objective A prediction model of various indicators,combined with NEWS,is to be created to forecast the prognosis of those suffering from sepsis,thereby furnishing medical personnel with precise and expeditious prediction technology and enhancing the prognosis of sepsis patients.Methods Patients with sepsis admitted to the Department of Critical Care Medicine of the Second Hospital of Lanzhou University from January 2019 to August2022 were collected as the research objects.The clinical data of all patients with sepsis within 24 hours after admission to the Intensive Care Unit were collected,including vital signs within 24 hours after admission,NEWS,laboratory biochemical indicators,comorbidities,underlying diseases,infection sites,length of hospital stay and treatment methods.Randomly dividing patients who met the diagnostic criteria of sepsis into training and validation groups at a ratio of 7:3,logistic regression analysis was employed to calculate the independent factors influencing the prognosis of the training group,based on the 30-day mortality index as the outcome index.Three groups of prediction models were constructed.By comparing the prediction efficiency and feasibility analysis of the three groups of models,model 2 was selected as the prediction model and a nomogram graph was drawn.By drawing the calibration curve,calculating the area under ROC curve of the nomogram and NEWS,and drawing the decision curve analysis,the consistency,prediction accuracy and clinical application value of the nomogram model were effectively evaluated.Results（1）In this study,457 patients who fulfilled the diagnostic criteria of sepsis were randomly split into two groups:the training group（320 cases）and the validation group（137 cases）.There were no significant difference in gender,age,infection sites,complications,treatment methods,vital signs,and laboratory indicators between the two groups（P>0.05）.There were little differences in length of hospital stay（P=0.049）,NEWS score（P=0.042）,hypertension（P=0.021）and D-dimer（P=0.014）between the two groups.And the difference between the death numbers and the survival numbers of patients in varying NEWS scores was striking（P<0.001）.（2）Univariate and multivariate analysis according to the outcome indicators in the training group showed that age（OR=1.032,95%CI:1.007-1.059）,NEWS（OR=1.563,95%CI:1.340-1.821）,Pa O₂（OR=0.958,95%CI:0.934-0.983）and lactate（OR=1.211,95%CI:1.073-1.368）were independent risk factors for 30-day prognosis of patients with sepsis.The AUCs of model 1,model 2 and model 3 in the training and validation groups were 0.927,0.921,0.881,0.943,0.941 and 0.870 respectively.Finally,model 2（the above four independent influencing factors）was selected to construct a nomogram.（3）The nomogram model’s calibration curve was found to be in agreement with the ideal curve between the two groups after evaluation and verification.The ROC curve of the nomogram model and NEWS in both the training and validation groups yielded values of 0.921 and 0.871 for the training group,0.941 and 0.859 for the validation group respectively,indicating that NEWS had a positive predictive effect on the prognosis of patients with sepsis and the prediction model combined with NEWS would be more effective.Additionally,decision curve analysis revealed that the nomogram model had a high clinical application value in both groups.Conclusions NEWS has important clinical value in predicting 30-day mortality of ICU patients with sepsis.A prognosis nomogram for sepsis patients,based on NEWS,was created through multivariate analysis and the amalgamation of independent influencing factors.And internal verification proved the feasibility and clinical effect of the model.It provides clinicians with a simple and effective way of assessment method,thus greatly improving the accuracy of prediction.