M Hewson

Enteral feeding in hypoglycaemia — reflections and poll results Thanks to everyone who voted, and especially to those who shared how they actually do this. Twenty-two votes is modest numbers but the feedback was interesting: Most push enteral rates well beyond my local guideline 30 mL/kg/day to avoid an IV — 47% to 60 mL/kg/day on day 1, a third to 90+. 50% dextrose is the more common additive (59%) over glucose polymer powders (40%) — though 61% use no polymer at all, and those who do start conservatively at 2–3 g/100 mL. Enteral 10% dextrose is seldom used So there's a wide range in volumes and additives (dextrose or glucose polymers or neither) The comments: roblums emphasised the LGA/IDM infant who has a difficult venous access, sometimes multiple attempts — a strong driver to utilise the gut first. Several favour continuous OG feeds, on the logic that steady delivery smooths the insulin response — essentially imitating an infusion. Of course milk includes protein and fat — substrate for endogenous glucose production. If EBM isn't available, preterm formula gives more carbohydrate and slightly faster bioavailability, since more is glucose polymer than lactose (no hepatic galactose step). Fortification variability is interesting. Dr Sanghvi's unit prepares formula in 10% dextrose rather than water, to 100 mL/kg/day on day one — a simple way to avoid almost all intravenous lines. My back-of-envelope: ~16 g sugar/100 mL, osmolality above 700 mOsm/kg, ~11 mg/kg/min at 100 mL/kg/day but it's well over the ~450 mOsm/kg AAP ceiling, so I worry about mucosal-health/NEC. Glucose polymer adds carbohydrate at much lower osmolality than dextrose, so it's probably the safer option— but the evidence for the real osmolality ceiling (and for the quoted 6–8 mg/kg/min active absorption limit) is thin. lilim's point about changing BGL thresholds - the definition of the cut-off early blood sugar for immediate IV bolus / infusion in asymptomatic infant is quite variable across guidelines: South Australian Guideline mandates IV at <1.5, BAPM uses <1.0, Starship (Auckland NZ) admits at <1.2, Sen et al. trialled 1.1. Is our mandatory-bolus threshold simply too high? My tentative conclusion is to move from using added dextrose 50% (we typically used to add 2g / 100mL milk) to instead use glucose polymers at 5 g / 100 mL when more sugar is required but volume is constrained. Happy generally to go up to 75 to 90 mL/kg/day enteral on day one but aware that we may max out glucose absorption. Consider continuous where tolerance is a problem, and keep feeds running alongside the IV when IV is unavoidable, reserving the bolus for the lowest sugars. FEED1 and Alshaikh 2025 are reassuring on NEC at these volumes. Dr Michael Hewson www.nicutools.org

Hi Dr Sanghvi - that's very interesting. Back of the envelope maths: Your resultant mixture has 16g sugar per 100 mL, and an osmolality of over 700 mOsm/kg. Giving these feeds at 100 mL/kg/day would be a glucose delivery of 11 mg/kg/min - so yes I can see you would avoid most iv insertion. Are you giving these feeds by perfusor, or bolus? Do you see any problems with feed tolerance - any osmotic diarrhoea? The high osmolality would make me worry about mucosal health - and ultimately NEC.

Hi Carlos - I would greatly appreciate any observations / feedback / suggestions including any related to ease of use. Michael.

The challenging IV insertion especially in LGA is a problem -anecdotally some infants require multiple attempts on multiple occasions and that is one factor pushing us to re-consider enteral feeds at higher rates with (if required) added dextrose - to reduce infant pain. Continuous feeds may be better tolerated and theoretically reduce insulin surges (imitating iv fluid administration in that regard). And of course the added value with enteral is the protein and fat - substrate for endogenous glucose production. Which milk to use if EBM unavailable - we usually use a preterm formula - not only more carbs but slightly faster bioavailability as a high percentage is glucose chains rather than lactose (compared to term formula) so the step of hepatic galactose conversion to glucose is nor required.

Hypoglycemia is a common cause of neonatal admission and is often managed with dextrose infusion albeit the early use of 40% glucose gel may have reduced admission rates. Our South Australian guideline recommends IV Dex 10% bolus and infusion at 90/kg/day if BGL is below 1.5 mmol/L, and also recommends iv Dex 10% at 60 ml/kg/day if sugars are between 1.5 and 2.5 despite enteral feeds of 30 mL/kg/day. But what is the role for higher rates of enteral feeding on Day 1 in the otherwise well infant without significant respiratory distress? Breastfeeding and EBM where possible, but also either a term formula (~7.5 g carb/100mL) or preterm (8.3 g/100mL) with the additional option to add dextrose (eg 4 mL of 50% added to 100 mL of formula) or a dextrose polymer (poly-joule) which would allow even greater amounts to be added without excessive osmolality. How high can we go with these options, both in terms of concentration and volume? Do any units use Dextrose 10% solutions enterally to manage the sugars without needing formula? Theoretically we can max-out the gut glucose transporters even when there is no actual gut pathology - quoted max active absorption is 6 to 8 mg/kg/min only but I'm not sure how well-evidenced that is. I'd like to get a sense of what other teams are doing to optimise enteral feed options in hypoglycemia so will try to post a very short survey.

Hi colleagues - many of you will be familiar with nicutools.org neonatal calculators. My new project is a synchronous, multi-participant, instructor-led team simulation in a free browser tool available at neosim.nicutools.org You can set up a virtual neonatal resus case and invite your trainees into the "sim lab" where they can prepare equipment, discuss team roles, then in real time resuscitate the virtual manikin which responds in a life-like fashion to their interventions. T-piece ventilation, intubation, UVC, adrenaline, red cells, needle thoracotomy and more are all modelled. You as teacher / instructor can adapt the sim to suit your teaching requirements and flex the manikin's response mid-sim as you need to support learning. There is a comprehensive debrief timeline provided post sim. Initial user feedback is very positive and is helping with on-going improvements to the project. If you have any trainees interested in the idea of study into the educational value of this specific tool I would be very keen to hear from you as well.

While at Wellington NZ NICU, I created a set of infusion calculators for neonatal use These are mostly fixed rate, and they are built around the specific drug ampoule strengths we stock. Michael

The Kaiser Permanente early onset sepsis calculator gives a great insight into this problem ... http://www.dor.kaiser.org/external/DORExternal/research/InfectionProbabilityCalculator.aspx Anyway from the data provided the EOS risk is very low and observation is all that is needed unless clinical features of sepsis appear. By the way, why wait to stop antibiotics at 48 hours? EOS cultures are almost always positive by 24 hours or not at all.