• Paediatric Surgery
• Neonatology

It is surgical dogma that bilious (green) vomiting in the neonatal period is indicative of mechanical obstruction until proven otherwise. The proportion of babies with green vomiting who have a surgical cause for their symptoms is, however, well under 100%. Stringer in 2002 reported that bilious vomiting is, in fact, not caused by mechanical obstruction in 62% of patients.1

The work by Mohinuddin et al2 makes a useful contribution to our understanding of this subject. In this study, roughly half of term babies (46%), with bilious vomiting, transferred by the London Neonatal Transfer team for surgical assessment, had a surgical cause for this symptom. Fourteen per cent of babies in this study were found to have a time-critical diagnosis where a perforation was demonstrated or gut viability was potentially compromised.

This paper makes an important contribution by looking at whether clinical signs or X-ray findings can differentiate babies with surgical pathology from those without. It is probably unsurprising that an abnormal abdominal X-ray is strongly associated with surgical pathology. Indeed, many of the conditions reported in this study are associated with characteristic abdominal X-ray findings that would lead to a decision to operate without further investigation being required; the classic ‘double bubble’ of duodenal atresia for example. What is perhaps more striking is the way in which a normal abdominal X-ray does not exclude a surgical cause for bilious vomiting. Mohinuddin et al2 helpfully draw attention to their finding that the specificity of this investigation in the hands of the referring service is only 14%. A further important message from this study is that the clinical signs and X-ray findings did not differ between those babies with time-critical causes for their surgical pathology and those without.

This paper, therefore, is helpful in considering the likelihood of surgical pathology in term infants with bilious vomiting and makes helpful points in relation to the way clinical findings and radiographs can assist with this.

The way these findings are generalised, however, is limited by two points, both clearly acknowledged by the authors. First, the study deals only with those babies with bilious vomiting in whom the referring centre is suspicious of a surgical diagnosis and has referred the baby to a surgical centre for assessment. Second, those babies born at centres with a paediatric surgical service are excluded because they had no need to be transferred.

The correct management of an apparently healthy infant with bilious (green) vomiting is a question that leads to debate when discussed between paediatric surgeons, neonatologists and paediatricians. The paediatric surgeon lives in fear of malrotation volvulus, which can cause irreversible necrosis of the entire small bowel within hours. Many of us have seen infants who appear very healthy early in the illness, and in whom the only presenting symptom is bilious vomiting, and yet are found to have a small bowel volvulus at laparotomy. Anyone who has been in the position of opening the abdomen of a thriving term baby, previously well, to find a completely necrotic intestine that ‘portends death or a life of misery from short-gut syndrome’3 is understandably keen to prevent any other patient from undergoing the same catastrophe. This fear is supported by data reporting that 15% of children undergoing intestinal transplantation have volvulus as the cause of their intestinal failure,4 suggesting that significant small bowel loss from volvulus continues to occur. Paediatric surgeons, therefore, argue that any term baby with bilious vomiting, in whom an alternative surgical diagnosis cannot be made with confidence, should undergo an upper gastrointestinal contrast study to exclude malrotation, and therefore malrotation volvulus, as an emergency.

There is evidence from the literature, however, that this does not represent universal practice. Walker et al5 performed a survey study of neonatologists of whom 80% said they would simply admit a baby with bilious vomiting for observation. More than 50% of neonatologists in this study did not consider a contrast study to be appropriate to investigate a baby with bilious vomiting and one-third felt that surgical referral was not appropriate for a baby following a single bilious vomit.5

A recent audit of practice from Glasgow looked at cases of malrotation in infants <1 month of age. In total, 97% of these babies presented with bilious vomiting, 82% had a volvulus at laparotomy and, perhaps most surprisingly, 72% had symptoms for 24 h before referral.6 These authors make the point that failure to correct malrotation in infancy risks small bowel necrosis during the illness in question and also risks volvulus in later childhood.

The extent to which these studies are generalisable is of course questionable, but it is probably safe to assume that a significant proportion of babies with bilious vomiting are not referred for a surgical opinion. The number of patients who have one or more, conservatively managed, bilious vomits and then have a trouble-free neonatal course is, therefore, unknown.

The paper by Mohinuddin et al is helpful in supporting the view that clinical signs and radiological investigations are not robust enough to exclude surgical pathology on their own, or indeed to distinguish those babies with time-critical conditions from those without. The logical end point of this argument is that prompt surgical referral and investigation is the best way to safely manage these babies. These authors have further added to the body of literature supporting the view that surgical pathology is often the cause of bilious vomiting in term babies. One can only support wholeheartedly the conclusion that babies with bilious vomiting are referred for surgical review and prioritised as time critical.