Success and failure with neonatal tracheo-oesophageal anomalies.

In seven and a half years, one surgical team treated 67 consecutive neonates with oesophageal atresia and/or tracheo-oesophageal fistula. According to Waterston's classification, 28 were in group A, 12 in group B and 27 in group C. The mortality rate during the initial admission was 10 per cent, all seven deaths being unavoidable in infants in group C with multiple anomalies. Birthweight alone had no bearing upon the chances of survival. Primary oesophageal repair, including one suture-fistula procedure and one delayed primary repair, was attempted in 54 (84 per cent) of the 64 patients with atresia and was successful in 46 (85 per cent). All three H-type tracheo-oesophageal fistulae were successfully divided in infants in group A. Recurrent tracheo-oesophageal fistula developed in four (7 per cent) infants, one of whom (group A) underwent successful repair. One disrupted anastomosis was successfully resutured (group A), so an intact oesophagus was finally achieved in 51 patients, of whom six (12 per cent) developed anastomotic strictures and 21 (41 per cent) underwent surgery for gastro-oesophageal reflux. Of the 60 early survivors, 10 (17 per cent) underwent aortopexy for tracheomalacia. Whenever possible, primary repair is advocated in all infants. Even for those in group C with multiple, severe associated anomalies, the combined early and late mortality was no greater following primary repair (7 died of 12 operated) than after staged repair (4 died of 7 operated), but major anastomotic complications were more common in infants in group C (5 out of 19) than in those in groups A and B (3 out of 38).