Study On The Influencing Factors Of Very Low And Extremely Low Birth Weight Infants With Bronchopulmonary Dysplasia

Objective:Through the study of maternal and infant clinical data of the very/extremely low birth weight infants,the influencing factors of BPD were analyzed and a predictive model was established to provide a theoretical basis for early identification of high-risk infants with BPD.Method:From January 1,2016 to December 31,2019,342 very/extremely low birth weight infants with gestational age ≤ 32 weeks who met the inclusion criteria in the NICU of Northwest Women’s and Children’s Hospital were selected as the subjects to collect the clinical data of mothers and infants.Research groups and contents: 1.According to the diagnostic criteria,the infants were divided into BPD group and non-BPD group.The independent risk factors affecting the occurrence of BPD were analyzed,the multi-factor Logistic regression model and nomogram were established,which were verified using the receiver operating characteristic(ROC)curve,Harrell C-Index,and calibration curve.2.According to the severity,the BPD group was divided into mild BPD group and moderate to severe BPD group.The independent risk factors affecting the occurrence of moderate to severe BPD were analyzed,the multi-factor Logistic regression model and nomogram were established,which were verified using the receiver operating characteristic(ROC)curve,Harrell C-Index,and calibration curve.Results:1.The prevalence rate of BPD was 32% to 53.2% in very low birth weight infants hospitalized in our hospital ≥ 28 days with gestational age ≤ 32 weeks from 2016 to2019.2.The risk of developing BPD correlates inversely with the gestational age and birth weight.The younger the gestational age,the lower the birth weight,the higher the prevalence rate of BPD.3.There were significant differences between the BPD group and the non-BPD group in maternal complications,male,gestational age,birth weight,respiratory complications,artificial assisted ventilation,length of hospitalization,clinical outcome,and so on.Multivariate Logistic regression showed that gestational age,birth weight,HCA,mechanical ventilation,and treated with PS were independent risk factors for BPD.Modeling and preliminary verification show that the predictive model has a good degree of discrimination and calibration.4.There were significant differences between moderate to severe BPD group and mild BPD group in maternal anemia,premature rupture of membrane ≥ 18 hours,prenatal use of antibiotics,birth weight,delivery room resuscitation,respiratory complications,artificial assisted ventilation,length of hospitalization,and clinical prognosis.Multivariate Logistic regression showed that 5-minute Apgar≤7,mechanical ventilation ≥ 7 days,and treated with postnatal glucocorticoid were independent risk factors for moderate to severe BPD.Modeling and preliminary verification show that the risk model has a certain degree of discrimination and calibration.Conclusions:1.The risk of developing BPD correlates inversely with the gestational age and birth weight.The younger the gestational age,the lower the birth weight,the higher the prevalence rate of BPD.2.Multivariate Logistic regression showed that gestational age and birth weight were associated with reduced risk of BPD,while Chorioamnionitis,mechanical ventilation,and treated with PS were associated with increased risk of BPD.Therefore,the prevention and treatment of preterm delivery,the control of prenatal infection,and the effort to avoid prolonged mechanical ventilation have a certain protective effect on the occurrence of BPD.3.We have built a predictive model based on the multivariate Logistic regression and conducted preliminary verification.The BPD prediction model has a good degree of discrimination and calibration,which can predict the risk of BPD in the early stage,and provide objective information for clinical decision-making and doctor-patient communication.4.Multivariate Logistic regression showed that 5-minute Apgar ≤ 7,mechanical ventilation ≥ 7 days,and treated with postnatal glucocorticoid were associated with increased risk of moderate to severe BPD.Modeling and preliminary verification show that the prediction model also has a good degree of discrimination and calibration,which means it can provide reference information for the prognosis of premature infants and has certain clinical value.