My co-author Dr. Seth Marks from Pediatric Endocrinology spent countless hours reviewing the evidence and fielding seemingly endless questions from reviewers and myself. The support from the CPS was also much appreciated as the back and forth from so many who expressed opinions would have been difficult to manage without their support. What I hope you will find as you read this is the best appraisal of the evidence and directions for care that we could come to in 2019. Where things will stand by the time we make it to the next revision will be interesting to see.
For those who have used the Acute Care of At Risk Newborns (ACORN) program you will be pleased to know that the upcoming new version of this program is completely alligned with the approach outlined in the CPS statement. Given that the first version of ACORN and the CPS were not aligned and caused great discussion and distress at times, we feel this is a big accomplishment.
The algorithm for managing hypoglycemia is clean and easy to follow (I think) and we hope such clarity will greatly help with managing those infants at risk.
Main Questions for the Future
We remain a country divided (much like the recent election) with respect to dextrose gels usage. There are centres which are leary of using adult formulations of dextrose gel in newborns whereas others have adopted such treatments with success. The algorithm and statement address the approach to using dextrose gels or what to do if one wishes to avoid such use. With time, local products or a national brand designed specifically for newborn use may come to be and this will need to be addressed at some point.
Changing the threshold at 72 hours and beyond to require glucose levels of greater than or equal to 3.3 mmol/L may lead some to be worried about a ramp up in admissions but at least locally we have not seen this. In the end those with persistent causes of hypoglycemia will manifest one way or the other and whether it is before or after discharge from the hospital may be a reflection of what threshold you feel comfortable using.
Will local guidelines be affected by adopting the changes outlined in this statement? I suspect so and would welcome feedback before the next version of this document is worked on to determine what if any ramifications (positve or negative) such approaches have had.
No doubt with many changes compared to the previous version of the statement there may be some surprises. Keep an open mind and look at the evidence presented. In other cases the lack of evidence has motivated a change in position. Either way we believe this is the best approach to care for at risk newborns given what we know today.
Any regular reader of this blog will know that human milk and the benefits derived from its consumption is a frequent topic covered. As the evidence continues to mount it is becoming fairly clear that the greater the consumption of mother’s own milk the better the outcomes appear to be with respect to risks of late onset sepsis or BPD as examples. Moving to an exclusive human milk diet has been advocated by some as being the next step in improving outcomes further. While evidence continues to come suggesting that replacement of fortification with a human based instead of a bovine based fortifier may improve outcomes, the largest studies have been retrospective in nature and therefore prone to the usual error that such papers may have.
What is evident though as the science pursues this topic further is that the risk of necrotizing enterocolitis or NEC is not zero even with a human milk diet. Why is that? It might be that some risks for NEC such as intestinal ischemia or extreme prematurity simply are too much to overcome the protective effect of breastmilk. Perhaps though it could be related to something intrinsic in the breastmilk that differs from one mother to another with some producing more protective milk than others.
Secretors vs Non-secretors
When it comes to the constituents of breastmilk, human milk oligosaccharides or HMOs are known to be secreted into breastmilk differently depending on whether a mother has a secretor gene or not. this has been demonstrated recently in HMOs affecting the microbiome in infants Association of Maternal Secretor Status and Human Milk Oligosaccharides With Milk Microbiota: An Observational Pilot Study. HMOs are capable of a few things such as stimulating growth of beneficial microbes and acting as “receptor decoys” for pathogenic bacteria. Previous rat models have also demonstrated their potential to reduce NEC in rat models. Essentially, mothers who have the secretor gene produce more diverse types of HMOs than mothers who are secretor negative.
What came out of the study were a couple very interesting findings. The first is that when analyzing the HMOs present in breastmilk at 2 weeks and comparing those who developed NEC to those who did not there was one significant difference. Lacto-N-difucohexaose I (LNDH I)had a median level of 0 (IQR 0-213) from the milk of those mothers who had infants affected by NEC. There were no differences observed for any other HMOs.
Also of interest was the greater diversity of HMOs present in the breastmilk samples of mothers whose infants did not develop NEC. This was present at all time points.
How Could This Be Useful?
If a broader array of HMOs is associated with less risk of NEC and the presence of LNDH I carries the same association it opens the door to the next phase of this research. Could provision of LNDH I in particular but moreover a wide array of HMOs to mother’s milk reduce the occurrence of NEC? This will need to be tested of course in well designed randomized trials but this type of fortification could be the next step in what we add to human milk to enhance infant outcomes. Given that it may be difficult to determine in short order whether women have these HMOs already a broad based fortification strategy assuming insufficient amounts of HMOs would be best. A quick search on clinicaltrials.gov shows that there are 101 trials in children looking at HMOs at the moment so more information on this topic is certainly on the way. Could HMOs be the magic bullet to help reduce NEC? Just maybe!
First off I should let you know that we do not do transpyloric feeding for our infants with BPD. Having said that I am aware of some units that do. I suspect the approach is a bit polarizing. A recent survey I posted to twitter revealed the following findings:
I think the data from this small poll reveal that while there is a bias towards NG feeds, there is no universal approach (as with many things in NICU).
Conceptually, units that are using transpyloric feeds would do so based on a belief that bypassing the stomach would lead to less reflux and risk of aspiration. The question though is whether this really works or not.
New N of 1 Trial
I don’t think I have talked about N of 1 trials before on this site. The trials in essence allow one patient to serve as a study unto themselves by randomizing treatments over time for the single patient. By exposing the patient to alternating treatments such as nasogastric or nasoduodenal feedings one can look at an outcome and get a sense of causality if a negative or positive outcome occurs during one of the periods consistently. That is what was done in the study Individualising care in severe bronchopulmonary dysplasia: a series of N-of-1 trials comparing transpyloric and gastric feeding by Jensen E et al from the Children’s Hospital of Philadelphia. The authors in this study determined that using a primary outcome of frequency of daily intermittent hypoxaemic events (SpO2 ≤80% lasting 10–180 s) they would need 15 patients undergoing N of 1 trials between nasogastric and nasoduodenal feeding. Included infants were born at <32 weeks and were getting positive airway pressure and full enteral nutrition at 36 0/7 to 55 6/7 weeks PMA. Infants who were felt to be demonstrating signs of reflux or frank regurgitation were enrolled.
Thirteen of 15 enrolled patients completed the study. The two who did not complete did so as their oxygen requirements increased shortly after starting the trial and the clinical team removed them and chose their preferred route of feeding. Randomization looked like this:
Of the 13 though that completed and using an intention to treat analysis of the other two the findings were somewhat surprising. Contrary to what one might have thought that transpyloric would be a lung protective strategy, the findings were opposite.
Overall the combined results from these 15 patients demonstrated that nasogastric feedings were protective from having intermittent hypoxic events.
How can this be explained?
To be honest I don’t really know but it is always fun to speculate. I can’t help but wonder if the lack of milk in the stomach led to an inability to neutralize the stomach pH. Perhaps distension has nothing to do with reflux and those with BPD who have respiratory distress with some degree of hyperinflation simply are prone to refluxing acid contents due to a change in the relationship of the diaphragmatic cura? It could simply be that while the volume in the stomach is less, what is being refluxed is of a higher acidity and leads to more bronchospasm and hypoxemic events.
What seems to be clear even with this small study is that there really is no evidence from this prospective trial that transpyloric feeding is better than nasogastric. Given the size of the study it is always worth having some degree of caution before embracing wholeheartedly these findings. No doubt someone will argue that a larger study is needed to confirm these findings. In the meantime for those who are routinely using the transpyloric route I believe what this study does at the very least is give reason to pause and consider what evidence you have to really support the practice of using that route.
The Ortolani and Barlow manouvers are probably the two most requested parts of the physical exam that students ask to be shown. We line up several medical students who take turns applying the steps of abduction and then adduction, testing the stability of the hips. We routinely give oral sucrose, position in kangaroo care or breastfeed while performing other noxious stimuli such as heel lancing but at least in my centre give nothing for manipulating the hips in such a fashion.
So the authors set out to compare findings in these 28 infants and standardized the study as much as they could by having one Neonatologist perform all hip exams and having a video recording of the infant’s face during the procedure assessed by two independent reviewers in order to assign the PIPP-R scores. While not a randomized trial, for the type of intervention being studied this was the right approach to take to determine the answer to their question.
Interesting findings indeed. Statistically significant differences were noted in bilateral changes in oxygen extraction during hip examination as well as for the GSR small peaks. The PIPP-R scores as well were vastly different between the two groups suggesting that the areas of the brain responsible for perception of pain were indeed activated more so with manipulation of the hip than with auscultation of the heart.
What can we take from this?
The hip exam may elicit responses indicating pain but there remains the question of how much is actually elicited. Nonetheless, the authors findings are intriguing as they certainly challenge the notion that this is a quick exam that should be just done and gotten over with. Clearly bundling or Kangaroo Care are not an option here due to the nature of what is being done. The next time you are planning on doing such tests though should you at least consider non-nutritive sucking on a pacifier or sucrose solution if readily available? If not readily available then should it be?