bimalc

I have never practiced with a group that did INSURE, but I've often wondered about premedication, especially after the study from Albany with an astonishingly high failure rate when using morphine as the premedication. Helpfully, that same group now provides data on use of remifentanyl instead: https://www.ncbi.nlm.nih.gov/m/pubmed/29789465/ Personally, the biggest change for me once we go to INSURE (assuming we don't just skip to MIST/LISA) is that I've routinely muscle relaxed for intubation for a number of years.

I also have an exceedingly low success rate in SUA cases (I am credentialing at a new facility and reviewed 3 years of my umbilical lines and I had only one successful UAC in SUA listed over a 3 year period)

The approach these authors suggest is really only implementable for moderate to late preterm infants. The overall approach is tempting to extend to more extreme premature infants, but I'm not aware of anyone pulling together the data to do this (Presumably the NRN in the US has the underlying data and VON may as well, though both are limited datasets in terms of post-discharge data. CHNC has a much higher risk and outworn population but does have a validated PHIS linkage as well as (theoretically) neonatal follow-up data). Stefan - when you say individual fortification, are you looking solely at patient characteristics or are you assessing (for milk fed babies) the caloric density and relative macronutrient composition of the milk (mother's own or donor) in order to guide degree of fortification? At the Pediatric Academic Society Meeting in the US this year there was a Swedish vendor (https://www.mirissolutions.com) selling an FDA-approved solution for human milk analysis: Do you have experience with this device?

I would object to the interpretation that the CIs crossing 1 for several secondary outcomes suggests that these results are likely due to chance. On the contrary, the available evidence suggests that it is more likely these are real associations (though potentially with small effect sizes particularly for non-motor outcomes). First, all of the presumably mechanistically related outcomes have the same qualitative effect (ie CPR has the worse outcome) whereas if these were truly random, we might see some outcomes favor the CPR group. Second, for the outcomes where the CI does cross 1, most of the 'mass' of the CI still favors the no-CPR group. Third, the one contrary study you cited aside, the prevalence of pre-existing evidence favored the no-CPR before this study was even done (ie the prior was not non-informative).

I've never seen this, but my first instinct would be to work the child up for diabetes (assuming this isn't sepsis or iatrogenic). If a biochemical diagnosis was not immediately obvious and you are in the UK, I would see if the genomic medicine service in Cambridge would accept your patient for neonatal whole genome sequencing. In terms of actually lowering the blood glucose, you could try increasing protein intake (if calories are adequate already, you could even reduce carbohydrate calories). Branched chain AAs are a key signal for energy balance and can stimulate a more robust insulin.

I have never found the need to use the quantitative ratio of insulin and glucose concentrations (and I assure you, your lab is measuring an insulin concentration in the samples you provide. If you are getting reports of insulin levels that are not concentrations you need to call your lab director). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4141553/ is a review that discusses one demonstration of a modified ratio as well as its limitations, though not in neonatal data.

What do you mean by 'early intervention'?

The use of a ratio without units implies that both quantities should be in the same units before taking their ratio. which units you choose to use is irrelevant. The purpose of the ratio is to detect inappropriately high insulin levels during hypoglycemia.

This is perhaps not the answer she was hoping for, but in her own multi center cohort study on bleeding risk and transfusion from 2017 there was at least one center that administered concentrated platelets ...

https://pdfs.semanticscholar.org/82f2/72696e66a81605427eafadd87d307573d4b1.pdf In general, thrombocytosis in neonates is reactive. Above is a case series emphasizing this point. Almost anything that stresses the marrow and/or induces hypoxia can cause thrombocytosis. Treatment should be directed at the underlying cause, NOT the platelets themselves. If you remove the stressor on the marrow, the platelet count will improve. As the case series above attests, clinically meaningful thrombotic events are exceedingly uncommon even at quite high platelet counts and, in fact, hemorrhagic occurrences are more common in this age group. I would not pursue anti platelet therapy unless I had evidence of meaningful clot burden and even then I would very carefully consider the relative risks and benefits of ant platelet or anticoagulant therapy.

What degree of thrombocytosis are you concerned about and what is your goal? I admit to having no experience in ordering anti platelet agents neonates, but I must also say I cannot think of a circumstance in which thrombocytosis would lead me to want to administer an anti platelet agent .... Can you provide more clinical context?

Thank you, Stefan, for reminding me of my one major criticism of the PINT trial: They did not include NEC (or at least surgical NEC) in their composite outcome. As for why we, as a profession, continue to worry about TANEC, it is mostly recall bias (we all remember the case of NEC shortly after transfusion but we never stop to think of all the transfusions we have given without NEC appearing) I've often wondered if there isn't already enough data out there to answer these questions (or at least get a start). Supervised learning techniques have advanced sufficient that if you were willing to invest in combining and curating multiple datasets of ELBWs you could probably train a binary classifier to predict risk of NEC with or without transfusion.

I would recommend reaching out to Susan.Slattery@northwestern.edu who helped develop weight-based standardized feeding pathways with me when I was in Chicago. The broad outlines of our approach were weight-based stratification, q12h advancement of volumes by nurses without waiting for the rounding team to arrive (in babies with expected course). When I left that unit, I think we were advancing ELBWs ~24mL/kg/d after a few days of trophic feeds, but I heard through the grape vine that they either slowed that down for all ELBWs or subdivided the ELBWs around 600-700g and slowed down just the absolute smallest babies, but either way, she would have the most recent information. Outcomes showing improvements in enteral nutrition without increased NEC were presented at PAS last year.

My differential in cases like this includes meconium peritonitis (obviously not what your baby has), infection/TORCH, chromosomal (quite common once you've excluded the previous two), Metabolic (Lysosomal storage disorders and peroxisomal biogenesis disorders most notably), endocrinopathy (primary or paraneoplastic) with ectopic calcification, vascular calcification (of which some may be due to mendelian disorders). Additional history I would request would be pregnancy exposures/infections and family history including potential consanguinity. In terms of ICU management, in resource limited settings, almost none of these evaluations need to be done in the ICU. Assuming the child was near intact and well appearing, feeding and otherwise ready to discharge soon, I would check an ECHO (if available) for structure, function and effusion and review X-rays for evidence of metabolic disease. The app face2gene has a function where you can upload X-ray images if you do not have local expertise in reviewing x-rays for these signs. You said 'labs WNR' but I would double check that the Ca and P were not borderline and in need of repeat before discharge, and that transaminases and the total and direct bill are reassuring. I would collect the relevant infectious studies during the index hospitalization, especially CMV, in order to better interpret congenital vs acquired exposures. Things you could defer to outpatient (many of which I would do in the US during the index hospitalization) would be karyotype or (preferably) microarray with SNP probes for potential regions of homozygosity, metabolic studies for lysosomal storage disease and peroxisomal defects, Endocrine evaluation of PTH axis, vascular imaging and consideration of related single gene disorders (most notably GACI, generalized arterial calcification of infancy).

I know this is not what you mean, but the single most common cause of this phenomenon in my experience is actually reversal of the probes (ie the pre- and post- are mislabeled/swapped). But, yes, I have seen several cases of the actual physical phenomenon you describe, often self resolved in a short period of time and sometimes causing an annoyed call from a cardiologist asking why I got an ECHO