Since time immemorial, therapists have wondered how much food their patients should have. The question is even more difficult to answer if patients are vomiting, have recently had abdominal surgery and are unable to speak.
Historically, the long journey towards proving what is the best treatment may well have started with astute observations that engendered a rule, later to be enforced by guilds and experts. We are still on that journey, but we are now building a body of evidence-based knowledge brick by brick. Yet much uncertainty remains in every-day clinical practice. The study by Elveos et al. which is published in this edition of the Journal of the Norwegian Medical Association provides an important contribution to paediatric surgery (1) – a contribution that gives pause for thought to all of us who in good faith have been following rules that have been unsupported by evidence.
In 1717, Patrick Blair gave an accurate description of hypertrophic pyloric stenosis after performing an autopsy on an infant who had vomited to death (2). The condition came to be feared and was terminal without surgery. Even with surgery, the outcome was often fatal. In 1912, Conrad Ramstedt described the principle of treatment as we know it today: surgery during which the serosa and the thickened muscle are incised to mucosa and the gap is left open (3). This remains one of the most frequently conducted operations in neonates. Only the level of access has changed since then, and the intervention is now speedy and safe, with good outcomes, as highlighted in the article by Elveos et al. This is rightly considered one of our discipline's victories.
A hundred years after the best surgical procedure was identified, the post-operative regime remained unclear
How should the child be fed after surgery? A hundred years after the best surgical procedure was identified, the post-operative regime remained unclear (4). There was profound disagreement in many hospital departments where different regimens were favoured. In North America, 15 % of surveyed institutions practised 'ad lib' postoperative feeding, often with certain restrictions imposed. So, for around 100 years, we have been starving our patients. A common protocol-based regimen would be six hours of fasting, followed by small amounts of thin glucose solution at intervals of a few of hours, followed by diluted milk, and then eventually milk. There was no consensus about how early after the operation the feeding should start, nor about how fast the feeding regimen should be escalated, and what the babies should be fed.
Elveos et al. refer to the only existing meta-analysis about post-pyloromyotomy feeding from 2016, the only one there is (5). Only three randomised studies and 11 comparative studies were identified for the analysis. This showed significantly shorter hospitalisation periods with ad lib feeding, but some of the studies used complications as exclusion criteria. The meta-analysis also concluded that there was more vomiting in cases of early feeding and rapid escalation. Many questions remained unanswered.
But had the infant been asked? Or the mother? Nurses will also have reflected on the matter. Before Elveos et al. conducted their study, a master's thesis was produced in the same department. This describes 'the frustration felt by infants, their parents and nurses because the child was stressed from hunger.' The thesis concluded as follows: 'Let us agree to let babies feed ad lib' (6). The department at Oslo University Hospital took the brave decision to break with the paradigm. When all players on the team are allowed to contribute, the results tend to be better. The outcomes may well be valuable – such as better treatment plans, and studies like this one.
The thesis concluded as follows: 'Let us agree to let babies feed ad lib'. The department at Oslo University Hospital took the brave decision to break with the paradigm
Elveos et al. have reviewed the consequences of the department changing their regimen to allowing infants to regulate their own intake. It is a great responsibility to be given to a five-week-old baby. Allowing for the documentation problems of a retrospective design, they nevertheless found that the little ones know best. The infants will eat what they can, perhaps slightly adjusted by mother's reduced production if breastfeeding, and they will recover faster and without further complications. So, the study also teaches us that if we had asked the patient sooner, and listened to the patient's wordless utterances, we would have saved many people from unnecessary hunger and discomfort. If we keep asking, more answers are sure to come our way.
Thank you to the team in Oslo who chose to listen to the infants, and who in doing so, have added another fine brick to the body of knowledge.