肝脓肿致脓毒症的危险因素分析及基于PNI、SII预测模型的建立与验证OA北大核心CSTPCD

Objective:To explore the risk factors of sepsis caused by liver abscess,and to construct a nomogram prediction model based on the Prognostic Nutrition Index(PNI)and systemic immune inflammation index(SII).Methods:We retrospective-ly selected 215 patients who were admitted to the Second Affiliated Hospital of Hainan Medical University from January 2021 to December 2023 who were diagnosed with bacterial liver abscess.They were divided into group(n=71 cases)and non-sepsis group(n=144 cases)sepsis according to whether they developed sepsis.The clinical data of the patients were collected and com-pared.The patient's PNI value and SII value was calculated,and the best boundary value of sepsis in patients with liver abscess was predicted by drawing a receiver operating characteristic(ROC)curve,and it was converted into a binary variable;univariate analysis and multivariate logistic regression analysis was used to select independent predictors of sepsis in patients with liver ab-scess and a nomogram prediction model was constructed;the prediction model was evaluated by further drawing the area under the ROC curve(AUC),calibration curve and decision curve analysis(DCA)of the prediction model.Results:Among the 215 pa-tients with bacterial liver abscess,70 cases(70/215)did not undergo liver abscess puncture,8 cases(8/215)underwent puncture culture but the etiology was negative,and 137 cases(137/215)underwent puncture culture and the etiology was positive.);in-cluding 114 cases of Klebsiella pneumoniae(114/137),which was the most common strain in patients with liver abscess in this study;through ROC curve analysis,the AUC of PNI in diagnosing sepsis in patients with liver abscess was 0.760(95%CI:0.697～0.816),the sensitivity and specificity are 0.803,0.681(P<0.05)respectively;the AUC of SII is 0.756,(95%CI:0.693～0.812),the sensitivity and specificity are 0.732,0.722(P<0.05);the study showed through single-factor and multi-factor logistic regression analysis that there were 6 independent predictors of sepsis in patients with liver abscess(all P<0.05),including gender,PCT,PNI,SII,treatment method,diabetes,the nomogram model constructed with the above 6 predictors predicts the risk of sepsis in patients with liver abscess.The nomogram ROC curve analysis results show that the AUC of the model are 0.886(95%CI:0.837～0.934),and the corresponding sensitivity and specificity are 0.845 and 0.806,respectively.Conclusion:Klebsiella pneumoniae is the most common type of liver abscess.Diabetes,biliary tract disease,elevated PCT value,and elevated SII are risk factors for sepsis in patients with liver abscess.Anti-infection combined with abscess puncture and drainage and elevated PNI value a protective factor for patients.The nomogram prediction model constructed based on this study has good accuracy,calibra-tion,and clinical practicability for patients with liver abscess complicated by sepsis.