Analysis Of Influencing Factors Of Unplanned Readmission Of Very Premature Infants And Construction Of Risk Prediction Model Based On Graph

ObjectiveTo investigate the current situation of unplanned readmission within one year after discharge from the neonatal intensive care unit in very preterm infants,to explore the related factors of unplanned readmission,and to establish a simple,effective and clinically convenient nomogram risk prediction model,so as to provide a risk screening tool for clinical medical staff to identify high-risk infants at risk of readmission early.MethodsA retrospective case-control study was conducted.According to the inclusion and exclusion criteria,550 very preterm infants who were hospitalized and discharged from the neonatal intensive care unit of the two wards of a tertiary hospital in Shandong Province from January 2020 to October 2021 were selected as the research objects.According to the length of hospital stay,the children were divided into a modeling group(from January 2020 to January 2021)and a validation group(from February 2021 to October 2021)at a ratio of 7:3.The general information,observation data and follow-up data of very preterm infants were investigated by searching databases,searching electronic medical records and telephone follow-up using the self-designed "Influencing factors of unplanned Readmission Questionnaire".The infants were divided into readmission group and non-readmission group according to whether unplanned readmission occurred.SPSS 26.0 and R 4.0.0 statistical software were used to perform independent sample t test,chi-square test,one-way analysis of variance and multivariate Logistic regression analysis on the data of the modeling group to determine the independent influencing factors of unplanned readmission and the regression coefficients of each factor,and a visual nomogram risk prediction model was drawn accordingly.The model was internally validated in the modeling group and externally validated in the validation group.The area under the receiver operating characteristic curve,Hosmer-Lemeshow test and calibration curve were used to evaluate the discrimination and calibration of the model.Results1.A total of 550 very preterm infants were enrolled in this study,and 234 of them had unplanned readmission within 1 year after discharge,with a readmission rate of 42.5%.Among them,161 children in the modeling group had readmission,and the readmission rate was 41.8%.A total of 73 children in the validation group had readmission,and the readmission rate was 44.2%.2.The main causes of unplanned readmission in very preterm infants were respiratory diseases,accounting for 60.6%,followed by digestive system diseases,infectious diseases,hernia,neonatal retinopathy,and hyperbilirubinemia,accounting for 17.0%,9.4%,6.2%,5.1%,and 1.6%,respectively.3.Univariate analysis showed that there were 15 influencing factors related to the occurrence of unplanned readmission in very preterm infants,including:Gestational age,birth weight,gender,polysomnosis,Apgar score 5 minutes,father’s age,neonatal respiratory distress syndrome≥grade Ⅱ,moderate to severe bronchopulmonary dysplasia,intraventricular hemorrhage≥grade Ⅲ,necrotizing enterocolitis≥stage Ⅱ,retinopathy of prematurity,length of first hospitalization,duration of noninvasive mechanical ventilation,duration of invasive mechanical ventilation,and feeding mode outside the hospital.The differences were statistically significant(P<0.1).There were no significant differences in other factors between the readmitted group and the non-readmitted group(P<0.1).4.Multivariate Logistic analysis showed that gestational age of 28～29+6 weeks(OR=3.445,95%CI:1.754-6.767,P<0.001),gestational age<28 weeks of age(OR=5.447,95%CI:2.045-14.511,P=0.001),male(OR=2.331,95%CI:1.339-4.058,P=0.003),first hospital stay ≥60 days(OR=5.833,95%CI:1.782-19.091,P=0.004),invasive mechanical ventilation time≥7 days(OR=4.985,95%CI:2.039-12.183,P<0.001)were the independent risk factors for unplanned readmission in very preterm infants.Exclusive breastfeeding(OR=0.192,95%CI:0.081-0.450,P<0.001)were the protective factors for unplanned readmission in very preterm infants.5.To develop a risk prediction nomogram for unplanned readmission in very preterm infants.In the nomogram risk prediction model,the area under the receiver operating characteristic curve of the modeling group and the validation group were 0.847 and 0.899,respectively,indicating that the model had good discrimination ability.The Hosmer-Lemeshow test showed that the goodness of fit of the two groups was modeling group χ2=6.177,P=0.627;Validation group χ2=4.993,P=0.758,indicating that the predicted value of the model was in good agreement with the actual value.The calibration curves of the two groups showed that the actual fitting curve of the model was well fitted to the standard curve,indicating that the prediction model constructed in thisstudy had good calibration ability.Conclusion1.The unplanned readmission rate of very preterm infants within 1 year after discharge is high,and respiratory diseases are the main reasons for readmission.Gestational age,gender,length of first hospitalization,invasive mechanical ventilation time,and out-of-hospital feeding methods are independent influencing factors for unplanned readmission of very preterm infants.It is suggested that clinical medical staff should strengthen the targeted detection of discharged very preterm infants,control controllable risk factors,and reduce the incidence of unplanned readmission.2.The risk prediction nomogram model for unplanned readmission of very preterm infants constructed in this study has good discrimination and calibration,which is helpful for clinical medical and nursing staff to identify high-risk infants with the risk of readmission early,take targeted preventive measures for intervention in time,and reduce the incidence of unplanned readmission of very preterm infants.