Based on Logistic Regression Model to Analyze the Influence Factors of Pulmonary Infection Occurred in Preterm Neonates

ZHAO Chenjing; CAO Huili

doi:10.3969/j.issn.1004-437X.2025.01.010

Abstract：Objective To explore the risk factors of lung infection in premature infants and the related prediction models.Methods Clinical data of 2 157 premature infants in the Third Affiliated Hospital of Zhengzhou University from January 2021 to October 2023 were collected for retrospective study.They were divided into infected group and non-infected group according to whether pulmonary infection occurred.The clinical data of the two groups were compared,and the influencing factors of pulmonary infection were analyzed by single and multiple factors to build a prediction model.Receiver operating characteristic(ROC)curve and area under curve(AUC)were used to analyze the predictive model to predict pulmonary infection.Results(1)Incidence of pulmonary infection:the infection rate of preterm infants was 1.62％.(2)Univariate analysis:the proportion of preterm infants with gestational age＜32 weeks,birth mass＜1 500 g,1 minute Agar score ≤7 points,underlying diseases,and poor mental response was higher in the infected group than in the non-infected group,albumin was higher than in the non-infected group,and the duration of mechanical ventilation,intravenous nutrition,and oxygen intake was longer than that in the non-infected group(P＜0.05).(3)Logistic regression analysis:gestational age(OR=0.226,95％CI:0.073-0.698),birth weight(OR=0.259,95％CI:0.096-0.701),duration of mechanical ventilation(OR=6.698,95％CI:2.754-16.291),psychiatric response(OR=3.864,the 95％CI:1.267-11.786),and deep vein cannulation(OR=3.116,95％CI:1.291-7.523)were independent risk factors for developing pulmonary infection in preterm infants(P＜0.05).(4)Logistic prediction model:logit(P)=-9.276-0.226X,-0.259X2+6.698;X3+3.864X4+3.116X5.The X1 was gestational age,X2 was the birth weight,X3 was the mechanical ventilation time,X4 was mental reaction,X5 was a deep vein catheter.(5)ROC analysis:the AUC for predicting lung infection in preterm infants was 0.912(95％CI:0.869-0.947),with a predictive sensitivity of 83.57％and a specificity of 89.11％.Conclusion The higher risk of lung infection in preterm infants is associated with younger gestational age,psychiatric response,deep venous cannulation,lower birth mass,and longer duration of mechanical ventilation,according to which the construction of logistic regression model has a reliable predictive value,which can be used to guide the diagnosis and treatment for the clinic.