Predictive value of inflammatory markers on the risk of stroke-associated pneumonia in acute ischemic stroke patients / 临床神经病学杂志

ABSTRACT

Objective To explore the predictive value of inflammatory markers for stroke-associated pneumonia(SAP)in patients with acute ischemic stroke(AIS)based on the nomogram model.Methods According to whether pneumonia occurred,259 AIS patients were divided into SAP group(81 cases)and non-SAP group(178 cases).The clinical data of the two groups were compared.The systemic inflammatory response index(SIRI),systemic immunoinflammatory index(SII)and neutrophil to lymphocyte ratio(NLR)were calculated according to the formula.The variables with statistically significant differences were included in the multivariate binary Logistic regression model to screen out the independent risk factors for SAP in AIS patients.The independent risk factors were used to construct a predictive model,and the predictive ability of the two models,which only included traditional factors and included inflammatory indicators at the same time,was further compared from the aspects of discrimination,calibration,clinical practicability and so on.Reclassification analysis was used to evaluate the extent to which the nomogram model improved the predictive value of SAP risk in AIS patients.Results Compared with those in the non-SAP group,the rates of smoking,diabetes,dysphagia,leukocytes,neutrophils,lymphocytes,triglyceride level,NIHSS score on admission,SIRI,SII and NLR were significantly increased in the SAP group,and the rate of hypertension was decreased(all P＜0.05).Diabetes mellitus(OR =2.505,95%CI1.070-5.850,P =0.034),dysphagia(OR =3.492,95%CI1.501-8.119,P =0.004),NIHSS score on admission(OR = 1.310,95%CI1.188-1.446,P＜0.001),SIRI(OR =2.417,95%CI1.327-4.401,P =0.008),NLR(OR =1.434,95%CI1.101-1.860,P =0.007)were independent risk factors for SAP in AIS patients.The area under the curve was 0.788(95%CI0.725-0.852,P＜0.001)for the prediction model without inflammatory factors and 0.884(95%CI0.838-0.930,P＜0.001)for the prediction model with independent risk factors.The calibration curve showed a good consistency between the predicted risk and the observed results.The decision curve showed that the model had a significant net benefit for predicting SAP.In addition,by calculating the net reclassification index(NRI)and the comprehensive discriminant improvement index(IDI),it was found that the nomogram model had a significant improvement in predicting the risk of SAP in AIS patients.Internal verification also proves the reliability of the nomogram model.Conclusions SIRI and NLR are independent predictors of SAP in AIS patients on admission.Adding SIRI and NLR to the traditional model can significantly improve the ability to identify the risk of SAP occurrence in AIS patients.