When I first went to one of the hospitals I have worked at I discovered that the nurses were routinely performing gastric lavage of infants in the NICU who had meconium aspiration, or a history of meconium in the amniotic fluid. I put a stop to the practice as I couldn’t see any rationale, and I had worked in other hospitals where this was not done, and babies in general seemed fine without it. I now find that this is a not uncommon practice, frequent enough for randomized controlled trials and a systematic review, performed by my friend and colleague Sanjay Patole and his co-workers. The SR suggests that maybe there is a benefit in terms of reduced feeding intolerance, but most babies did not have significant feeding problems, and the only individual trial which showed a significant benefit was of poor quality, caregivers were unmasked, and allocation was unmasked, and the outcome is rather subjective. Basically this is an unpleasant procedure for the babies, with no reliable evidence that it does anything.
No.
OK, I was just going to stop after the “No” but I can’t be that brief, or dismissive. 12 babies with BPD off respiratory support and 34 babies without BPD had esophageal impedance monitoring for an average of about 24 hours. There were no differences in the numbers or characteristics of reflux events. Only when a multivariate regression was performed did the authors find that one of the 5 indices that they calculated was different between groups. Which is pushing it, as they say. The association could easily be due to chance, the authors have really inflated the risk of a type 1 error, and should have been made to tone down their discussion.
They also showed that most “symptoms of reflux” were not temporally related to reflux (even though one of the symptoms was regurgitation, which probably was related!) Which is again not how they report the data.
The authors talk about anti-reflux medication, and the importance of proving reflux before using the medications. I would agree…. if there were any such thing as anti-reflux medication. The only therapies used for reflux in the newborn are toxic placebos. Please stop.
Anti-acid medications, on the other hand, may well be effective for reducing acid production, but the only clearly demonstrated effect of them is to increase sepsis, chest infections and necrotizing enterocolitis.
If you click on gastro-oesophageal reflux on the word cloud to the right, you can see some of my other posts on these agents which are used for reflux.
Many of you will know that I think the clinical value of routine MRI imaging of very preterm infants is unproven. Maybe I should say that differently; there is good evidence that routine MRI imaging has very poor positive predictive value for adverse outcomes in very preterm infants.
On the other hand this is the kind of research that can be done with fancy forms of MRI, that can give some indication of the pathophysiology of brain injury in the preterm infant. The authors looked at fractional anisotropy (I guess I am going to have to learn how to say that word) of the cortico-spinal tracts on MRI from babies with cystic PVL and without. FA was higher in the controls than among the PVL babies, and among those PVL babies was not associated with the presence or severity of cerebral palsy (which as you will know is the major long term consequence of cystic PVL). The later a scan was done, in terms of post-menstrual age (that is, if it was done at or after term equivalent age) the smaller was the thalamus. Which suggests that the damage to the corticospinal tracts occurs when the PVL is developing, and the thalamic abnormality is a consequence of that. As usual the babies with cystic PVL were not generally extremely preterm, they were moderately preterm infants who often had perinatal sepsis, or other inflammatory conditions.
Schuurmans J, et al. Neonatal morphine in extremely and very preterm neonates: its effect on the developing brain – a review. The Journal of Maternal-Fetal & Neonatal Medicine. 2015;28(2):222-8. This is an excellent review, the unfortunate thing is that they are unable to reliably address the issue in the title, really, as there is little good quality evidence. They note:
Considering all above-mentioned results together, it can be concluded that morphine does not seem to have a negative effect on neurodevelopmental outcome. None of these studies were prospectively designed to investigate the long-term neurodevelopmental outcome, thus, the power of these studies to show a more specific effect of morphine on neurodevelopment was too low. Also, selection bias and loss to follow up could play a key role in the evaluation of the results, since most studies used a smaller subset from the original population.
Neonatal Research 