Respire bébé, respire

Annie Janvier vient de publier un livre. C’est l’histoire de notre fille, et des histoires d’autres familles qu’Annie a rencontré et soigné, et les choses qu’elle a appris de ces familles.

C’est un livre publié pour les familles, les soignants, et tout le monde qui s’intéresse à la question humaine.

Il est publié par Québec Amérique, la plus grande maison d’édition du Québec, qui écrit qu’il est :

un ouvrage percutant, baigné d’humour autant que de larmes

cliquer ici pour leur déscription complète.

et il est disponible:

Directement de Québec Amérique

Dans les magasins de Renaud-Bray et aussi, évidemment, Archambault, Indigo etc.

En version éléctronique: epub, pdf et kindle.

And soon in English!

Posted in Neonatal Research | 4 Comments

What if 5 were the magic number?

Our weeks are made of 7 days. An entirely arbitrary unit of measurement, based on an idea that 7 is magical, so there are 7 continents, 7 seas, 7 days of the week. Or maybe because you can easily divide a lunar month by 4, in fact it isn’t totally clear why there are 7 days in a week. It is hidden in the mists of time, the majority of human cultures use the 7 day week. But what if it had been 5? What if the Aztec system of 20 day periods, divided into 5 day weeks, had taken over?

2 day weekends and 3 days of work sounds like fun. Of course the duration of a pregnancy in Aztec land would be 56 five-day weeks, and then what we now refer to as 25 weeks gestation would become 35 five-day Aztec weeks. Neonatal survival becomes less and less likely as we descend to 34 Aztec weeks, or 33, and so on. But imagine if you were in an Aztec country,and the local paediatric society had recommended that “33 weekers” were not viable, because survival is too low, or impairment is to high (actually they don’t ever explicitly say why 33 weekers should not get active care) . Imagine if you were denied active intervention because you were only at “33 weeks” and 4 days, and told to go away until you hit 34 weeks. You could maybe just fly to Canada to find that you were suddenly viable, as you were 24 weeks and 1 day in 7-day weeks.

You could be forgiven for being a bit angry about your local paediatric society’s guidelines, especially when you realize that your ultrasound performed at 16 Aztec weeks was only accurate to within about one 5-day week, 95% of the time.

I’ve been thinking about this recently. Especially after the publication of the NICHD network data on active treatment, and how it changes by weeks of gestational age. One of the graphs from that publication shows that active intervention at 22 and 23 weeks stays relatively unchanged as a proportion of deliveries until just before the end of the week.

GA and resuscitation

Which is odd, but I think easy to understand. We have become used to thinking of babies in terms of whole numbers of weeks of gestation. So a baby of 23 0/7 is thought of as being the same as 23 5/7. But once the baby is within 48 hours of the next big number, at which point they may have had steroids because of crazy hospital policies, based on 7 being the magical number, our attitudes start to change. I would be fascinated to know when the mothers got their steroids, and if that is really the cause of the jump up in intervention rates.

I would also be fascinated to see what the pattern would be in Aztec land.

One other thing that I think this shows, is that shared decision-making is currently a sham. Surely if mothers were adequately informed, and truly participated in the decisions, rather than having them imposed by physicians or by hospital practice, then there would be a progressive gradual increase in intervention as gestational age advanced. Mothers would have no reason for suddenly deciding at 23 weeks and 5 days that their baby should have active intervention.

I am sure that the physicians who resuscitate 100% of the babies at 22 weeks, and those that resuscitate none of them, all think they practice shared decision-making. But clearly if there are no babies being resuscitated, (once there are significant numbers of mothers in the data) then the values of the physicians and the care team are being imposed on the families. If the decision was truly shared, then there would be at least a few who occasionally received active intervention.

Posted in Neonatal Research | Tagged , , | 2 Comments

PPIs are toxic

PPIs could mean “public-private initiatives” which are highly toxic, as recent experience in the UK and in Quebec has shown, but in this case I mean proton pump inhibitors.

Stark CM, Nylund CM. Side Effects and Complications of Proton Pump Inhibitors: A Pediatric Perspective. The Journal of pediatrics. 2015. This excellent and complete review of PPI toxicities has just been published.

Many people seem to think they are safe, and there has been an explosion in their use in the newborn, with practically no supportive data.

PPIs increase :

GI infections (including clostridia)

Upper respiratory infections

Lower respiratory infections

Spontaneous Peritonitis

Coeliac Disease

Gastric Fundal Polyps

Malabsorption of calcium

Malabsorption of magnesium

Probably malabsorption of iron and some vitamins

Acute interstitial nephritis

and, of course, Necrotising Enterocolitis.


You should really have good evidence that gastric acid secretion is causing your baby’s problems before prescribing these toxins.


Posted in Neonatal Research | Tagged , | Leave a comment

Holland is changing

A new article from the Groningen group, plus Annie Janvier. Koper JF, et al. Dutch neonatologists have adopted a more interventionist approach to neonatal care. Acta Paediatr. 2015;104(9):888-93. It documents changes in their delivery room and NICU in end-of life practices between two periods 2001-2003 and 2008-2010.

Even with the “more interventionist approach” they only had 1 baby under 25 weeks admitted to the NICU in the later cohort (who died). So they were still, in that period, not offering intervention under 25 weeks.

I wanted to highlight however their table 2: which looks like this, the first column of numbers is the first cohort, and the proportion of the 126 deaths which occurred in each situation. The second column is the 113 babies in the second cohort. The asterisks denote statistical “significance”.

Stillbirth, dead on arrival to the hospital 44 (35%) 37 (33%)
Stillbirth, withholding surgical intervention 29 (23%) 17 (15%)
TOP, congenital malformation  22 (17%) 33 (29%)*
Induction for risk of extreme preterm birth 1(1%) 7 (6%)
Withholding resuscitation (comfort care) 25 (20%) 13 (12%) **

It is basically the table I included in my presentation at the last PAS meeting, and shows a lot more information than just the NICU admission data. The increase in termination of pregnancy (TOP) is probably related to changes in ultrasounds during pregnancy, ultrasound with screening for malformations was introduced as standard practice in Holland between the cohorts, and there was therefore an increase in antenatal diagnosis. This may also partly be the reason for the reduction in comfort care, as there would likely be fewer surprise diagnoses of serious malformations in the delivery room. The decrease in deaths due to “withholding surgical intervention” (which basically means not doing a Cesarean) is probably truly a willingness to intervene more actively for babies.

I think it is important that neonatal survival data are presented with this amount of detail. Otherwise a change in mortality among NICU admissions is difficult to interpret, and comparisons between groups of patients becomes meaningless.

Posted in Neonatal Research | Tagged | Leave a comment

Prebiotics, Probiotics and the microbiome

A series of interesting recent articles:

Yang J, et al. Application of Laser Capture Microdissection and 16S rRNA Gene Polymerase Chain Reaction in the Analysis of Bacteria Colonizing the Intestinal Tissue of Neonates With Necrotizing Enterocolitis. Pediatr Infect Dis J. 2015;34(10):e279-89. These investigators examined intestinal samples resected during NEC surgery. Some controls with intestinal atresia were also studied. They were able to examine bacteria specifically related to the NEC tissue samples. They found several groups of bugs that were not present in the controls. Enterococcus sp. and Escherichia sp. were frequently found in both infants with NEC and controls, whereas Pseudomonas sp., Klebsiella sp., Acinetobacter sp., Clostridium sp., Ochrobactrum sp. and Arcobacter sp. were detected only in NEC tissue samples. In contrast to other studies they actually showed more bacterial diversity in the NEC babies than the controls, but that may be because of the nature of the controls, the controls had intestinal atresia, so they would all have to be very young, and the control tissue all came from the small intestine, whereas the NEC tissue was from large intestine, with some samples from the small.

Carl MA, et al. Sepsis from the gut: the enteric habitat of bacteria that cause late-onset neonatal bloodstream infections. Clinical infectious diseases. 2014;58(9):1211-8. This study report is difficult to interpret, which is a shame as it seems fascinating. The authors don’t say when they started collecting stools, nor how frequently, but they did find, in many infants who subsequently developed sepsis, that stools that had been collected beforehand were often colonized with organisms which seem identical to the organisms that they isolated from the blood. Those organisms included E coli (2) Serratia (2) and 3 cases of late onset sepsis caused by GBS (which is an unusually high proportion). A sepsis caused by Klebsiella and 2 MRSA were not preceded by intestinal colonization, and another baby had GBS colonization and was infected by a different GBS. Babies who did not develop sepsis did not have these particular organisms. The authors don’t say much about the microbiome of the non-septic babies.

How can we try and get our babies colonized with the good bugs?

Skin to skin contact with the baby’s mother, that’s how. Hendricks-Munoz KD, et al. Skin-to-Skin Care and the Development of the Preterm Infant Oral Microbiome. American journal of perinatology. 2015(EFirst). They showed associations between skin to skin care and development of the microbiome of the babies’ saliva, with “an accelerated pace of oral microbial repertoire maturity”.

Also don’t give them (or their mothers) antibiotics. Arboleya S, et al. Intestinal Microbiota Development in Preterm Neonates and Effect of Perinatal Antibiotics. The Journal of pediatrics. 2014(0). Because, the infants not exposed to antibiotics (either directly or via their mothers) had higher relative amounts of Comamonadaceae, Staphylococcaceae, and unclassified Bacilli . Infants not exposed to antibiotics also had significantly higher percentages of Bifidobacteriaceae, Streptococcaceae, unclassified Actinobacteria, and unclassified Lactobacillales and lower of Enterobacteriaceae than both groups of infants whose mothers received antibiotics.

Also give them human milk, its packed with goodies.

Underwood MA, et al. Human Milk Oligosaccharides in Premature Infants: Absorption, Excretion and Influence on the Intestinal Microbiota. Pediatr Res. 2015. Or at least it is if your mother is not a non-secretor. It seems that some mothers are homozygous for a gene that prevents them secreting fucosylated glycans into their breast milk. It also looks like this is quite common, 6 out of 1 mothers in this study were non-secretors, and this status had a big influence on the intestinal microbiome development. Also the sialylated oligosaccharides were very variable and some mothers produced milk that had oligosaccharides with little fucose or sialic acid.; those mothers’ babies had more dysbiosis than the others.

So choose your mother carefully if you are going to be born prematurely.

Barrett E, et al. The neonatal gut harbours distinct bifidobacterial strains. Archives of disease in childhood Fetal and neonatal edition. 2015. In this study from healthy full term infants they identified over 170 different bifidobacteria strains. Some of the babies were being breast fed, some receiving a formula without prebiotics, and some a formula with added prebiotics. In this case the prebiotics were galacto-oligosaccharides and poly-fructose. They showed that dietary prebiotic supplementation was associated with an increased prevalence of B. longum in infants in addition to increased strain diversity
Endo A, et al. Long-term monitoring of the human intestinal microbiota from the 2nd week to 13 years of age. Anaerobe. 2014;28(0):149-56. Ten subjects donated stool on multiple occasions during the first 13 years of their lives. They changed. Most of the change was in the first 12 months of life. The major changes were int eh relative abundance of the families of bugs, major species did not change dramatically. Probiotic supplementation in early life didn’t seem to have a lasting effect on the microbiome.

Hickey L, et al. Cross-colonization of infants with probiotic organisms in a neonatal unit. Journal of Hospital Infection. 2014;88(4):226-9. In this data from the ProPrems trial, only 3 of the 38 controls tested were cross-colonized with the probiotics they were testing. Which is lower than some other studies, maybe in Melbourne they are better at washing their hands.

And finally, two very nice review articles covering the issues.

Walker WA, Iyengar RS. Breast milk, microbiota, and intestinal immune homeostasis. Pediatr Res. 2015;77(1-2):220-8.

Pacheco AR, et al. The Impact of the Milk Glycobiome on the Neonate Gut Microbiota. Annual Review of Animal Biosciences. 2015;3(1):419-45.


Posted in Neonatal Research | Tagged , , | Leave a comment

Neonatal updates

Ong KK, et al. Postnatal growth in preterm infants and later health outcomes: a systematic review. Acta Paediatrica. 2015;104(10):974-86.

This interesting systematic review confirms that the observational studies are fairly consistent, improved growth, and improved head circumference growth are associated with improved outcomes. As for the interventional studies the evidence is much weaker. Of the small number of trials some were only performed after discharge from the NICU, a couple are quite old trials with control groups who received what would now be considered to be inadequate intakes.

I guess the problem now is that it would be difficult to justify a study where one group received recommended intakes, and the other received less nutrition. We know that if you get close to recommended intakes you can get close to intra-uterine growth rates, so randomizing babies to have sub-optimal growth would be problematic.

I think that all we can do is to keep trying to get to optimal growth, and keep an eye on our outcomes.

Jering K, et al. Parenteral Nutrition as an Unexpected and Preventable Source of Mercury Exposure in Preterm Infants. The Journal of pediatrics. 2015;166(6):1533-5. In other nutritional news, TPN can be source of evil as well as good. Aluminium, manganese, and now mercury can be found in it. Although the levels were low, there was detectable mercury in the TPN in this NICU, maybe from sharing equipment for preparation of TPN for adults.

Jary S, et al. Less severe cerebral palsy outcomes in infants treated with therapeutic hypothermia. Acta Paediatr. 2015. Some babies treated with therapeutic hypothermia still develop cerebral palsy. This cohort study shows that they are less severely affected than a historical comparison group before hypothermia.

Brock JW, et al. Bladder Function After Fetal Surgery for Myelomeningocele. Pediatrics. 2015.  Children from the MOMS trial of antenatal surgical closure for meningomyelocele were followed. The bladder function outcomes were all better in the antenatal surgery group. The proportion who were using intermittent catheterisation at 30 months of age was 38% vs 51%, which was not statistically significant, but is quite a substantial difference. The other outcomes showed significantly less trabeculation and less open bladder neck with antenatal surgery.

The authors give a summary of the other results of the MOMS trial.

 In summary, prenatal surgery was associated with less need for cerebrospinal fluid shunt at 12 months and a better composite score for mental development and motor function at 30 months. Prenatal surgery also revealed benefit in several key secondary outcomes including hindbrain herniation, ability to walk unaided, and a better score on the Bayley II Psychomotor Development Index. These results were tempered by an increase in preterm birth and the risk of uterine dehiscence in the prenatal surgery group

Posted in Neonatal Research | Leave a comment

How much phosphorus does a preterm baby need?

Brener Dik PH, et al. Early hypophosphatemia in preterm infants receiving aggressive parenteral nutrition. J Perinatol. 2015;35(9):712-5. This group of neonatologists in Buenos Aires routinely give 3 g/kg/d of lipid, 3 to 3.5 g/kg/d of amino acids, 40 mg/kg/d of calcium gluconate and 20 mg/kg/d of glycerophosphate starting on day 1 to their babies under 1250 grams. By day 6 the babies were often hypercalcemic and hypophosphatemic, especially the IUGR babies. Babies who were more unwell in the first days of life were also more likely to have a low phosphate. 40 mg of calcium gluconate is slightly less than 0.1 mmol of calcium, 20 mg/kg/d of sodium glycerophosphate is about 0.1 mmol/kg/d of phosphorus.

Boubred F, et al. Extremely preterm infants who are small for gestational age have a high risk of early hypophosphatemia and hypokalemia. Acta Paediatr. 2015. In Sweden they give less nutrition on day 1; 2 g/kg/d of protein, 1 g/kg/d of lipid and 6 g of glucose, and then ramp it up over 4 days. On day 1 they give very little phosphorus, thereafter the babies were receiving about 0.6 mmol/kg/d of calcium, and between 0.5 and 0.6 mmol/kg/d of phosphorus. The IUGR babies were much more likely to develop low serum phosphates, and again the peak seems to be about day 4.

Moe K, et al. Administering different levels of parenteral phosphate and amino acids did not influence growth in extremely preterm infants. Acta Paediatr. 2015;104(9):894-9. In this observational study from Copenhagen they report 3 cohorts of preterm babies, in the middle cohort there was an error in their TPN software, so they gave much less phosphorus to the babies, especially during the first 3 days of life, when there was about 0.07 mmol/100 ml of TPN. The babies who got this regime had much more hypophosphatemia, especially if they were IUGR; the authors didn’t find a difference in growth during the first month of life.

Three other recent articles address these issues also :

Bonsante F, et al. Initial amino acid intake influences phosphorus and calcium homeostasis in preterm infants–it is time to change the composition of the early parenteral nutrition. PLoS One. 2013;8(8):e72880. This article (free access) reports a cohort study with varying protein intakes in preterm infants, the babies were all treated in the University hospital in Dijon, but, interestingly none of the authors’ current affiliations are with that hospital, they are from Réunion, Italy and Belgium. They show that the babies who received more amino acids had more hypophosphatemia, even though they also received more phosphorus. The high AA group received 21 mg/kg/d of phosphorus which is about 0.67 mmol/kg/d.

Christmann V, et al. Early postnatal calcium and phosphorus metabolism in preterm infants. Journal of pediatric gastroenterology and nutrition. 2014;58(4):398-403. In this study from Nijmegen babies received much more calcium and phosphorus, quickly getting up to 3 mmol/kg/d of calcium and 1.92 mmol/kg/d of phosphorus, by day 3. By day 4 to 5 many of their babies were hypophosphatemic, and hypercalcemic, at which time they had almost no phosphorus in the urine. So although they were getting lots of phosphorus, it looks like the ratio was not correct.

Pereira-da-Silva L, et al. Early High Calcium and Phosphorus Intake by Parenteral Nutrition Prevents Short-term Bone Strength Decline in Preterm Infants. Journal of Pediatric Gastroenterology & Nutrition. 2011;52(2):203-9. This is the only reference in today’s post that is actually an RCT, from Lisbon this time. About 40 preterm babies per group were randomized to different intakes of Calcium and Phosphorus in their TPN. They either got 45 mg/kg/d of calcium (1.1 mmol) with 36 mg/kg/d of phosphorus (1.16 mmol)  or 75 mg/kg/d of calcium (1.9 mmol) and 44 mg/kg/d of phosphorus (1.42 mmol), which usually started on day 1. The enteral feeds were not changed, so by the end of the first week there was little difference between the groups, and the macronutrient supply was also similar. From week 3 to week 6 there was a progressive reduction in bone strength in the low mineral group, and no reduction in the high group. The authors do not report the incidence of hypophosphatemia.

One thing this review has taught me is that the reporting of mineral intakes and metabolism in preterm infants is often really unclear, even when the study is concentrating on minerals. Could we all please report intakes and balances in mmol?

The other things that are clear, (and I must claim clairvoyance because our TPN standards here have said this for years) is that the requirements for calcium and phosphorus in the first few days of life are not the same as later on in the life of the preterm. The ratio between calcium and phosphorus should be higher for the first few (3-4?) days. Phosphorus is important for cellular metabolism as well as for bone growth. In those first few days we should aim to give enough phosphorus for those requirements, and to avoid hypophosphatemia. At the same time avoiding hypercalcemia and hypocalcemia are important. We might need to give as much as a 1:1 ratio of calcium to phosphorus just after birth, and then progressively switch to a ratio of about 1.6:1 (in mmol).

In fact the paper by Bonsante (a great name for a physician!) suggests something similar, they propose that the appropriate P intake (in mg/kg/d) might be calculated by dividing the Ca intake (also in mg/kg/d) by 2.15 and adding the amino acid intake (in g/kg/d) -1.3 multiplied by 9.8. If a baby is getting 2.5 g/kg/d of amino acids, and 40 mg (1 mmol)/kg/d of calcium this would work out to about 31 mg of phosphorus, which is almost exactly 1 mmol. If they are getting more protein, they would receive a bit more phosphorus, which is consistent with the data from those studies above.

Once the baby starts to grow consistently, then more phosphorus is laid down in bone, and we should probably aim for a ratio of nearer to 1.66 :1 (in mols), which is the ratio of calcium to phosphorus in new bone.

The recommendations of ESPGHAN published in 2005 (available free on-line) only discuss requirements during the growing phase of the preterm, and not during the first few days. They refer to a calcium phosphorus ratio of between 1.3:1 and 1.7:1 (in mols). The total intake they suggest is to vary according to weight gain, and to give 4 mmol of calcium for every 20 g of weight gain. Which I think is difficult to put into practice, as this suggests that you can only decide on how much calcium and phosphorus to give after the babies have already received the TPN and you see how well they grow!

The increased risk of hypophosphatemia in IUGR babies is similar to a “refeeding syndrome” although Bonsante’s group suggest not using that term, they suggest “placental interrupted feeding syndrome” instead, in order to note that you don’t have to be severely malnourished to develop these findings, they occur in AGA preterm also. Do IUGR babies, in the first days of life, need more phosphate? Do they need a higher ratio of phosphorus to calcium? Or both? I think some more observational studies with a higher P administration and an appropriate ratio might help us to know.

Posted in Neonatal Research | Leave a comment