The Value Of Initial Lung Ultrasound Score In Predicting Respiratory Support Mode In Premature Infants With Respiratory Disorders Within 24h After Hospitalization

Objective:With lung ultrasound（Lung ULtrasound,LUS）increasingly used in clinical practice,especially in the diagnosis and treatment of lung disease in newborns,this study aims to find out the value of initial lung ultrasound score（Lung Ultrasound Score,LUSsc）in predicting respiratory support mode in premature infants with respiratory disorders within 24h after hospitalization and help clinicial doctors to choose the respiratory support modes more quickly and accurately.Methods:Premature babies with dyspnea were selected from neonatal department of the First Hospital of Jilin University for this prospective study.There were a total of1,275 cases with dyspnea were admitted to hospital from July 2019 to July 2020.According to the inclusion and exclusion criteria,a total of 450 premature babies with dyspnea were selected to incorporate the study.Then the patients were divided into non-invasive ventilation group and invasive ventilation group according to the highest level of respiratory support mode within 24hour after hospitalization.Respiratory upport mode was determined by the clinical attending physicians without esearchers’intervention.All patients’initial lung ultrasound were performed within 2hours after hospitalization.Then the LUSsc were calculated by two professional ltrasound physician using ultrasound images preserved.A study of clinical data elated to two groups was performed.IBM SPSS 25.0 software were used for single factor correlation analysis and multifactor logistic regression analysis to determine the predictive factors for respiratory support mode.The study then used the receiver perator characteristic（ROC）curve to figure out the best diagnostic cut-off point,and calculated the sensitivity and specificity.Result:（1）In the end,450 premature infants with gestational age of 23-36⁺⁶ week were accepted into the study,including 300 patients in non-invasive ventilation group and150 patients in invasive ventilation group.（2）The results of univariate analysis showed that gestational age,weight,mode of delivery,pregnancy-induced hypertension syndrome,1minute Apgar score,oxygen pressure,BE value,Lac,PEEP,LUSsc and PS after LUS test were the relevant influencers for respiratory support mode（P<0.05）.（3）The results of multifactor logistic regression analysis showed that PS after LUS test（Odds Ratio,OR=0.069,95%Confidence Interval,95%CI:0.034-0.143）,PEEP（OR=1.616,95%CI:1.272-2.052）,LUSsc（OR=1.160,95%CI:1.111-1.210）and weight（OR=0.999,95%CI:0.999-1.000）were the independent influencers of respiratory support mode.（4）The ROC curve showed the area under the curve was 0.885（95%CI:0.895-0.903）,which was proved to be statistically significant（P<0.05）.Therefore 9points was the best demarcation point for non-invasive ventilation and invasive ventilation mode,when LUSsc>9 points to choose mechanical ventilation and when LUSsc≤9 points to choose non-mechanical ventilation,and at this time sensitivity and specificity were 85.0%and 86.6%respectively.Conclusion:（1）LUSsc can predict the initial respiratory support mode of premature babies with dyspnea after hospitalization within 24h.（2）When LUSsc>9 points to choose invasive ventilation and when LUSsc≤9points to choose non-invasive ventilation.It has a high sensitivity and specificity at this time.