Mortality and morbidity in oesophageal atresia

Several classification systems exist to predict mortality in oesophageal atresia, the most widely quoted of these being over 20 years old. No classification system exists to predict morbidity. We sought to test whether these classification systems remain relevant and to determine whether they can be useful to predict morbidity. In addition, we aimed to identify independent risk factors for predicting mortality and morbidity.

Neonates presenting with oesophageal atresia over a 20-year period (1990-2010) were retrospectively reviewed. Discriminative statistical analysis compared the performance of current classification systems. Stepwise logistic regression analysis of the influence of perioperative risk factors on mortality and duration of ventilatory support and intensive care unit stay were performed.

All classification systems predicted mortality in this series of 248 neonates. Birth weight, cardiac anomalies and pre-operative pneumonia were independent risk factors for predicting mortality in oesophageal atresia. The Waterston classification is the most useful classification for predicting post-operative morbidity in terms of length of hospital stay and time spent ventilated.

Despite advances in the neonatal care of the very low birth weight infant and those with congenital cardiac disease, these conditions remain relevant in predicting mortality and morbidity in oesophageal atresia.