Factors associated with short- and long-term survival in oesophageal atresia with tracheoesophageal fistula

Background: Risk stratification scores for oesophageal atresia with tracheoesophageal fistula (OA-TOF) are useful to predict survival. Scores from single-centre series may lack external validity. We aimed to model survival in OA-TOF and validate a new risk score using an international, multicentre database of patients. Method: Retrospective data (2000-2022) from 4 tertiary centres were used; a multivariable logistic regression model for survival to discharge was derived from single centre data, and then validated on the whole dataset using bootstrapping. In addition, Cox regression was used to analyse determinants of longer-term survival in those infants who had survived to one year of age. Results: Overall survival to discharge was 94% (668/708). After internal validation, mortality prior to discharge was significantly predicted by lower birthweight (aOR 1.174/100g; p=0.007), major cardiac lesion requiring surgery; (aOR 12.8; p<0.001) or an additional associated major structural anomaly or syndrome (aOR 5.12; p<0.001). A further 19 patients died after hospital discharge at a median 453 days [range 60-2640]. Factors associated with long-term mortality in those who had survived to one year of age were: major renal disease (bilateral or unilateral structural anomaly with impaired function, aOR 11.08 p<0.001) or an additional major structural anomaly or syndrome (aOR 6.41 p<0.001). We propose a new Oesophageal Atresia Risk (OAR) Score. Conclusion: Low birthweight infants and those with cardiac disease are at risk of early mortality; these factors are less significant than the presence of major renal disease for those infants who survive to discharge. Major structural anomalies and syndromic associations remain significant determinants of mortality throughout.