Font Size: a A A

The Predictive Effect Of HDL On Positive Etiological Examination Of Sepsis And Its Model Construction

Posted on:2023-02-02Degree:MasterType:Thesis
Country:ChinaCandidate:Y ShengFull Text:PDF
GTID:2544307070494254Subject:Clinical medicine
Abstract/Summary:PDF Full Text Request
Objective: To explore the predictive effect of high-density lipoprotein(HDL)on the positive etiological examination in patients with sepsis,and we plan to establish an optimal prediction model,and to explore the difference of prediction efficiency of the prediction model for different etiological types of the positive etiological examination results.Methods: The clinical data of patients diagnosed with sepsis/septic shock with positive etiological examination and sepsis/septic shock with negative etiological examination admitted to the electronic medical record system of Xiangya Hospital of Central South University from February 2018 to September 2021 were retrospectively analyzed,and the differences of clinical indexes between the two groups were compared.Regression analysis was used to determine the indicators that had significant influence on the prediction of sepsis patients complicated with positive etiological examination.The receiver operating characteristic curve(ROC curve)of single index and combined index in predicting sepsis complicated with positive etiological examination were drawn and obtain their area under ROC curve(AUC value).select the model with the best prediction efficiency,evaluate the stability and calibration ability of the model,and obtain the best prediction model of positive etiological examination in patients with sepsis,and carry out internal verification and visualization of the model.And further explore the difference of the predictive efficiency of the optimal prediction model for different etiological types of etiological examination results.Results: A total of 476 patients with sepsis were included in this study,including 236 patients in the positive etiological examination group and 240 patients in the negative etiological examination group.It was found that HDL,C-reactive protein(CRP),body temperature(T)and white blood cell count(WBC)were significant factors affecting the prediction of positive etiological examination in sepsis.The AUC value and 95% confidence interval(CI)of HDL,WBC,CRP and T alone in predicting sepsis with positive etiological examination were 0.730(0.731,0.784),0.552(0.550,0.602),0.623(0.621,0.669)and 0.637(0.637,0.692),respectively.And the AUC value and 95% CI of "HDL+WBC+CRP+T " in predicting sepsis with positive etiological examination were 0.774(0.730,0.818).The cutoff value,specificity and sensitivity of the prediction model and HDL,WBC,CRP and T alone in predicting sepsis with positive etiological examination were 0.461(64.8%,87.2%),0.67(84.7%,62.9%),13.15(57.9%,53.0%),75.07(43.8%,76.7%)and 38.45(62.5%,59.3%).HDL(X1)+ WBC(X2)+ CRP(X3)+T(X4)is the best prediction model of sepsis patients with positive etiological examination,model formula =-17.164-3.169X1+0.033*X2+0.002*X3+0.479*X4.The predicted positive rate of the prediction model was lower than the actual positive rate,which underestimating the risk of sepsis with positive etiological examination.In the subgroup analysis,the AUC values and 95% CI of this prediction model for the prediction of result of etiological examination as Gram-negative bacteria,Gram-positive bacteria,Gram-negative bacteria and Gram-positive bacteria mixed and other pathogen were 0.780(0.731,0.832)、0.749(0.684,0.809)、0.820(0.742,0.900)and 0.820(0.749,0.889),respectively.Conclusion: The level of HDL in patients with sepsis is closely related to the positive etiological examination,and the level of HDL may be used as a positive marker of etiological examination.On this basis,the predictive model "HDL+WBC+CRP+T" constructed in this study has a good predictive effect on the positive etiological examination in patients with sepsis.The application of this model can provide early warning of positive etiological examination in patients with sepsis and may help to improve the prognosis of patients with sepsis.
Keywords/Search Tags:sepsis, HDL, blood culture, NGS, prediction model
PDF Full Text Request
Related items