Analysis Of Clinical Outcomes And Development Of Mortality Prediction Model For Infants With Congenital Diaphragmatic Hernia Under A Multidisciplinary Diagnostic And Treatment Approach

Objective:Congenital Diaphragmatic Hernia（CDH）is a severe congenital structural anomaly.Critically ill infants with CDH often present with progressively worsening refractory hypoxemia and respiratory circulatory failure,which can be managed by extracorporeal membrane oxygenation（ECMO）to support their respiratory and circulatory functions.However,the survival advantage of ECMO remains highly debated due to the occurrence of multiple systemic complications associated with ECMO and variations in practices among different medical centers,along with the substantial cost of hospitalization.The mortality rate of these infants can still reach as high as 50%-70%.Currently,with the advancement of techniques such as pulmonary surfactant administration,inhaled nitric oxide therapy,and high-frequency oscillatory ventilation,coupled with the development of fetal medicine in our center,we have actively explored an integrated multidisciplinary approach in the prenatal,perinatal,and postnatal periods.Through close collaboration with prenatal diagnostic centers,we provide prenatal counseling and perform extracorporeal surgery during delivery.We have also established a"green transport channel"for high-risk neonatal rescue,continuously improving and optimizing clinical management processes to enhance the prognosis of CDH infants.Therefore,the aim of this study is to analyze the clinical outcomes of CDH infants in our center in recent years,examine the manifestations of pulmonary hypertension（PH）,and discuss the risk factors for early postnatal mortality in infants meeting ECMO criteria.Methods:A retrospective analysis was conducted on isolated CDH cases admitted to our hospital from January 2016 to August 2022.The cases were divided into ECMO-indicated and non-ECMO-indicated groups based on whether they met the criteria for ECMO use.The cases were divided into an ECMO-indicated group and a non-ECMO group based on whether they met the criteria for ECMO use within 24 hours of birth.We compared the clinical characteristics of neonates in the ECMO-indicated and non-indicated groups and analyzed and compared the severity of PH between the two groups.In addition,we screened the risk factors for mortality in the ECMO-indicated group through univariate and multivariate logistic regression analyses.Results:（1）A total of 103 cases of CDH infants were included,with 39 cases in the ECMO-indicated group.The overall mortality rate at discharge was 29.13%（95%CI:20.59%-38.9%）,with a mortality rate of 53.85%（95%CI:37.2%-69.9%）in the ECMO-indicated group,of which 38.1%occurred before surgery.The mortality rate in the non-ECMO-indicated group was 14.1%（95%CI:6.64%-25%）,with 11.1%occurring before surgery.Among the surviving infants in the ECMO-indicated group,the median duration of invasive ventilator use was 336.5 hours（IQR,160.5-578.5）,and the median length of hospital stay was 37.5 days（IQR,22.8-51.3）.In the non-ECMO-indicated group,the respective values were 177.0 hours（IQR,121.0-311.0）and 25 days（IQR,19-34）.The ECMO group had a higher proportion of right-sided and bilateral diaphragmatic hernias（48.8%vs.7.8%,p<0.001）,a higher incidence of liver herniation（84.6%vs.7.8%,p<0.001）,and a greater degree of diaphragmatic defect（p<0.05）,with statistically significant differences observed.（2）Within 24 hours of admission,echocardiographic findings revealed that infants in the ECMO group had more severe pulmonary hypertension（p<0.001）,a higher prevalence of right ventricular enlargement（69.7%vs.27.3%,p<0.001）,and left ventricular dysfunction（18.2%vs.0,p=0.002）,with statistically significant differences observed.（3）Univariate analysis and multivariate logistic regression analysis revealed that among infants meeting ECMO criteria,gestational age at diagnosis<25 weeks（OR=10.01,95%CI:1.45-69.102,p=0.019）,Pa CO₂（OR=1.121,95%CI:1.021-1.23,p=0.017）,and oxygenation index（OR=1.149,95%CI:1.038-1.273,p=0.007）were independent risk factors predicting mortality in this group.Conclusion:Although ECMO was not used,our center has achieved treatment outcomes comparable to ECMO centers by strengthening the integrated prenatal,perinatal,and postnatal diagnostic and therapeutic approach,optimizing ventilatory management,and implementing precise personalized use of vasoreactive medications.Therefore,it is recommended to exercise caution in the use of ECMO in CDH infants.PH and cardiac dysfunction are adverse prognostic factors in these patients.Additionally,gestational age at diagnosis of<25 weeks,Pa CO₂,and oxygenation index are independent risk factors for early mortality in infants meeting ECMO criteria.Objective: Congenital Diaphragmatic Hernia（CDH）is a profoundly heterogeneous disease,exhibiting varying degrees of severity and diverse clinical manifestations.Despite notable advancements in medical technology and perioperative management in recent years,the mortality rate among CDH patients remains alarmingly high,ranging from 30% to 50%.Early diagnosis and intervention have shown promise in improving the prognosis of CDH infants.This study aims to explore the predictive value of prenatal and early postnatal physiological indicators,as well as laboratory examinations,in determining adverse outcomes in CDH.Moreover,the objective is to develop a predictive model that can aid in the early identification of high-risk infants,offering valuable insights for clinical decision-making.Methods: This study retrospectively collected clinical data including basic information,prenatal diagnostic data,physiological data within 24 hours of birth,and laboratory test results of hospitalized CDH patients from January 2016 to December 2022.The outcome of interest was mortality at the time of discharge.Potential risk factors（p<0.05）were identified through univariate analysis,and logistic regression analysis was conducted on the modeling dataset to determine early predictive factors for mortality.A predictive model was then constructed,and a predictive nomogram was developed.Internal validation was performed using the bootstrap method,and the accuracy of the model was evaluated by calculating sensitivity,specificity,and the area under the curve（AUC）.The model’s calibration ability and clinical effectiveness were evaluated through the Hosmer-Lemeshow goodness-of-fit test,calibration curve,and decision curve analysis（DCA）.Results: A total of 109 CDH patients diagnosed prenatally were included in this study,with a discharge mortality rate of 27.5%（30/109）.Among the cohort,males accountedfor 59.6%（65/109）.The CDH patients were divided into two groups: the survival group（79 cases）and the death group（30 cases）.Univariate analysis was employed to determine candidate variables,and continuous variables were transformed to achieve normality.All candidate variables were then included in a binary logistic regression model,resulting in the establishment of a predictive model for mortality risk.The variables ultimately included in the model were liver herniation（OR=5.11,95%CI: 1.54-16.99,p=0.008）,gestational age/3（OR=0.59,95%CI: 0.38-0.89,p=0.013）,severe pulmonary hypertension（OR=8.14,95%CI: 0.38-0.89,p=0.002）,and the best p H*10 within 24 hours（OR=0.32,95%CI: 0.38-0.89,p=0.003）.The model exhibited a concordance rate of 87.2%,sensitivity of 70.0%,specificity of 82.5%,positive predictive value of 80.8%,and negative predictive value of 89.2%.Furthermore,a nomogram scoring model was constructed,demonstrating good discriminative ability with an AUC of 0.889（95% CI: 0.802-0.975）.The optimal cutoff value was 0.228.Internal validation using bootstrap resampling（1000 times）resulted in a calibrated AUC of 0.827,sensitivity of 87.4%,and specificity of 61.2%.The Hosmer-Lemeshow goodness-of-fit test（p=0.148）?calibration curve and the decision curve analysis indicated satisfactory calibration and a favorable clinical net benefit of the predictive model.Conclusion: This study has preliminarily established a predictive model for mortality risk in CDH patients based on prenatal and early postnatal indicators.It explores inde pendent risk factors for adverse early outcomes and visualizes the model through a no mogram,which can be utilized for early prediction of mortality risk in CDH patients a nd assist clinicians in making timely clinical decisions.