The journey from conception to the labour floor and then for some to the NICU is not a straight one. There are times of joy, interspersed with sadness, denial, anger and eventually acceptance, as initial news of being pregnant leads to complications in pregnancy and then eventual admission of an infant to the NICU.
Much has been said in recent years about the building of partnerships with parents and in fact there is a new catchphrase attached to the concept “shared decision making” (SDM). There is no question that in the perfect world this is exactly the relationship that we should be striving for with all of our patients. The world however is not perfect and although this may not be the most popular opinion I have given, I question how applicable this really is in many situations.
A Reality Check
Take for instance the parents who present to the labour floor of their local hospital in advanced labour at 24 weeks. Proponents of this SDM model would suggest that a meeting take place and pertinent information be given to a family and together with the assistance of literature applicable to their situation (possibly a pamphlet) the health care providers and families come to a mutually agreeable decision as to what the best course of action is for them and their unborn infant. This all sounds wonderful but examining the real life situation a little more closely is it actually reasonable to assume we can obtain this? I have not been, nor will I ever be pregnant and certainly have never experienced contractions and felt the veil clouding my vision as the first dose of analgesia enters my veins to deal with the discomfort a woman experiences during labour. Not to mention there are people admitting this couple, taking histories, establishing IV access, scanning bellies and a whole host of other pokes and prods along the way.
My Role Better Defined
Then I come in. Among all this chaos I deliver the information, pass along a pamphlet and do the best job I can to inform said couple of the upcoming decision. The trouble of course is how do we come to this mutual decision in the 15 – 30 minutes I spend with them during this crisis? The answer sadly is we do our best but don’t for a minute think that SDM has occurred. I don’t believe this is possible unless the family has prior experience with a preterm birth or perhaps is a HCP working with newborns or children with disabilities themselves. In fact Boss RD et al in their own research on the subject identified that in hindsight religion, spirituality and hope are what motivated parents rather than what was said at the time. In essence their minds are already made up. It doesn’t mean we shouldn’t strive for the SDM but at least in my opinion, unless their contractions settle, a calmness ensues, they have time to digest the information being given and then meet again under less stressful circumstances, the SDM is a nice idea but for many not a reality.
Shifting To The NICU
I recall a significant moment in my training when I saw how the SDM model can actually cause more grief than help. Dr. Keith Barrington a fellow blogger (if you haven’t discovered him, his work is fascinating over at Neonatal Research) published one of the most impactful pieces of research of the decade during my fellowship. The adverse neuro-developmental effects of postnatal steroids in the preterm infant: a systematic review of RCTs. Following this analysis there was a near moratorium on the use of post natal steroids. The issue this created was that to now receive them you had to be close to the end of the limits of care. At this point you either died (thereby concluding they are of no help) or you survived with disability that was due in part no doubt to how sick you had become (thereby concluding they are dangerous).
The moment I am referring to was a conversation with a family in which the attending managing the unit presented the risks and benefits of postnatal steroids to the family when the FiO2 was at 40% one day. The language used was non directive and the parents asked for another day to decide. The next day and each of the following two days they were unable to choose between giving the steroids and the perceived risk of brain damage versus not and watching the FiO2 climb by about 10% per day. By the time the FiO2 several days later was at 80-90% they were distraught, teary and feeling helpless. What they needed was direction; someone to give them some advice or more simply an educated opinion.
We can strive to share in the decision making but I continue to believe there is a time and place to help our families by taking a stance or side. We can equip them with as much information as we want but is there really any replacement for actually taking care of these infants, experiencing the ups and downs and hearing how they have done in follow-up? We simply can’t expect the average parent to understand the true long term consequences of their decisions. I am not saying we go back to a paternalistic time in medicine but I am saying that one size does not fit all.
We owe it to our families to pursue SDM when we can but we have an equal obligation to recognize when this ideal state is simply not possible. At this point we have to use the experiences and knowledge we have to provide them with the best advice we can. We have gone through medical training, and gone down these paths so many times. We can avoid biased opinion and rely on the facts as they are in our institutions but to not take a stand when it is needed at least for me is doing a disservice to those we are so eager to help.
In the book Outliers by Malcolm Gladwell he talked about 10000 hours being the threshold at which if you practiced or gained that much experience with something you could become an expert. In Sweden the approach to 22 week pregnancies and above is to resuscitate all as a strategy. I wrote about this before in At 22 weeks of gestation does your faith matter most to outcome? The information gleaned from that paper was that if you have a policy of resuscitation at these gestational ages and you compare outcomes to a centre that is selective in who they resuscitate the outcomes are better when you believe all should be resuscitated.
New Study Expands Data
The same group has published this time around the results of the same cohort from 2006-2015 but this time with developmental outcome data. The paper is entitled Outcomes of a uniformly active approach to infants born at 22-24 weeks of gestation In this paper they discuss outcomes at 234 and 24 weeks respectively in addition to the findings for 22 weeks which they covered in the paper mentioned above. The reason for sharing this study is that if I asked you to imagine going into a room right now and talking to a family at 22 – 24 weeks and predicting the outcome of their infant, my hunch is that you would not provide as rosy a picture as the group in Uppsala, Sweden.
The authors looked at a group of infants with the following breakdown by gestational age.
Not surprisingly as gestational age declines the incidence of complications rise but looking at Table 3 even at 22 weeks there was no difference in rate of NEC or need for PDA surgery. Nor was there a difference in rates of severe IVH/PVL. Rates of BPD were higher and likely attributable to the longer durations of positive pressure ventilation. Are these rates for these complications terrible though when the other option is non-intervention which for certain means death?
What about outcome?
Looking at the outcome at 2.5 years, the rate of cerebral palsy is about 1:10 to 1:20 for all GA. Hearing impairment is almost non-existent and while developmental delay is detected in 50 percent of survivors at 22 weeks only about a quarter of the infants have severe impairment.
Deciding what to do
There is no question that many of you seeing what this post was about would simply say “NO WAY” but in the end isn’t it really about a shared decision with the family? It wasn’t that long ago that we had to have a real paradigm shift in thinking about resuscitating 23 week infants. The amount of mental energy spent for teams worldwide coming to this decision was tremendous and now if you were to suggest compassionate care at 24 weeks you get a look back like you are crazy! It wasn’t that long ago that 24 weeks was considered viability in many places and now that 23 is the new 24 this is the struggle some people have now. Should we go to 22 weeks everywhere as the Swedes have? Clearly this is a decision that institutions need to look at critically and determine if they have the space to accomodate. Each infant should they survive will occupy a bed for at least 6 months. This needs to be looked at before one can just say sure this is a good idea.
In the end what the study demonstrates I believe is that the outcome at 22 and 23 weeks for that matter need not be universally dismal. There certainly is a good chance that a surviving infant will have one or more deficits but in the end I would advocate this is a conversation all units need to be having and critically look at whether each institution has the capactity to consider.
My bet is that five years from now this discussion will be moot as we will be mostly in that direction but at this time I think we are still in evolution.
It has been a few months now that I have been serving as Chair of the Fetus and Newborn Committee for the Canadian Pediatric Society. Certain statements that we release resonate strongly with me and the one just released this week is certainly one of them. Guidelines for vitamin K prophylaxis in newborns is an important statement about a condition that thankfully so few people ever experience. To read the statement on the CPS website click here.
Similar story to vaccinations
Prior to the American Academy of Pediatrics in 1961 proclaiming that all newborns should receive IM Vitamin K at birth the incidence of Vitamin K deficient bleeding was 0.25 – 1.7%. Think about that for a moment. A new parent could expect that 1/100 babies roughly might have intestinal bleeding or worse an intracranial hemorrhage due to an insufficient amount of vitamin K levels in the newborn. The types of bleeding could be categorized into three different time epochs. Early onset (occurring in the first 24 hours post-birth), classic (occurring at days 2 to 7) and late onset (at 2 to 12 weeks and up to 6 months of age).
With a rate that high detractors of providing Vitamin K at birth would say “why should we give it; I haven’t heard of any baby getting such bleeding?” Looking at it another way though, why don’t you see congenital rubella or kids with measles much these days? It’s due to vaccination. Thankfully as a Neonatologist, I don’t see Vitamin K deficient bleeding since most parents provide Vitamin K to their babies at birth. If you went back to the era prior to 1961 when widespread supplementation of Vitamin K began in the US, I imagine it would not have been too uncommon to hear about a baby who had bleeding issues after birth. Just because we don’t hear about German Measles much anymore doesn’t mean the virus causing it doesn’t still exist!
How Effective is Vitamin K?
How effective is Vitamin K administration at birth in preventing hemorrhagic disease of the newborn (HDNB)? Studies estimate an incidence of 0.25 per 100000 live births or 1 in 400000 babies vs the 1/100 risk without any vitamin K. That is one effective intervention! At this point I would ask those families that are still concerned about giving Vitamin K to their infants if this is a risk they can accept? If they refuse Vitamin K and there is a significant bleed how will they react?
The Change in this CPS Statement From the Past
In the last statement on Vitamin K, the authors suggested that the oral route was a reasonable option. Instead of giving 1 mg of Vitamin K IM one would dose it as 2 mg orally and then repeat at 2-4 weeks and then 6-8 weeks. In looking at the effectiveness though it is worth noting that while we can assure that families will get the first dose, as with any medication that needs repeat dosing there is the risk of forgetfulness leading to missed dosing down the road. In fact when the authors looked at the risk of late HDNB they found the following “The relative risk for VKDB, when comparing PO versus IM vitamin K administration in these two studies, was 28.75 (95% CI 1.64 to 503.45) and 5.97 (95% CI 0.54 to 65.82), respectively .”
The outcome of course remains rare but the risk based on two studies was almost 30 times higher than if IM dosing was given.
On this basis IM is recommended.
Having said all this I recognize that despite all this information, some families will choose for a number of reasons to still opt for the oral dose. As the statement suggests we need to encourage such use when a family refuses IM vitamin K. The 30 fold risk compared to IM administration is magnitudes lower than the approximate 1/100 risk of giving nothing at all!
In the end I believe that one case of intracranial hemorrhage from inadequate vitamin K is too much. This one vitamin indeed could save a life.
A few weeks back I wrote about the topic of intubations and whether premedication is really needed (Still performing awake intubations in newborns? Maybe this will change your mind.) I was clear in my belief that it is and offered reasons why. There is another group of practitioners though that generally agree that premedication is beneficial but have a different question. Many believe that analgesia or sedation is needed but question the need for paralysis. The usual argument is that if the intubation doesn’t go well and the patient can’t spontaneously ventilate could we be worse off if the patient loses their muscle tone.
Neonatal Intubation Registry
At the CPS meeting last month in Quebec City. I had the pleasure of listening to a talk by Dr. Elizabeth Foglia on the findings from a Neonatal intubation registry that many centres have been contributing to. The National Emergency Airway Registry for Neonates (NEAR4NEOs), records all intubations from a number of centres using an online database and allows for analysis of many different aspects of intubations in neonates.
This year, J. Krick et al published Premedication with paralysis improves intubation success and decreases adverse events in very low birth weight infants: a prospective cohort study. This study compared results from the registry of two centres, the University of Washington Medical Center (UWMC) and Seattle Children’s Hospital where the former rarely uses paralysis and the latter in almost all instances of non-emergent intubation. In all, 237 encounters were analyzed in the NICU for babies < 1500g with the majority of encounters (181) being from UWMC. The median PMA at intubation was 28 completed weeks (IQR: 27, 30), chronological age was 9 days (IQR: 2, 26) and weight was 953 g (IQR: 742,1200). The babies were compared based on the following groups. Premedication with a paralytic 21%, without a paralytic 46% and no premedication 31%.
This was an observational study that examined the rates of adverse events and subdivided into severe (cardiac arrest, esophageal intubation with delayed recognition, emesis with witnessed aspiration, hypotension requiring intervention (fluid and/or vasopressors), laryngospasm, malignant hyperthermia, pneumothorax/pneumomediastinum, or direct airway injury) vs non-severe (mainstem bronchial intuba- tion, esophageal intubation with immediate recognition, emesis without aspiration, hypertension requiring therapy, epistaxis, lip trauma, gum or oral trauma, dysrhythmia, and pain and/or agitation requiring additional medication and causing a delay in intubation.).
How did the groups compare?
It turns out paralysis seems to be a big deal (at least in this group of infants). Use of paralysis resulted in less attempts to intubate (median 1 attempt; IQR: 1, 2.25 vs. 2; IQR: 1, 3, p < 0.05)). In fact success was no different between the groups with no paralysis or no premedication at all! When it comes to tracheal intubation adverse events the impact of using paralysis becomes more evident. Paralysis does make a difference in reducing the incidence of such events and moreover when only looking at the rate of severe adverse events as defined above the finding was that none occurred when paralysis was used vs 9 when no paralysis was employed and 5 when no premedication was used at all. The rate of bradycardic events was less in the paralytic group but rates of oxygen desaturation between the three arms were no different.
How do we interpret the results?
Based on the results from the registry it looks like paralysis is a good thing here when electively intubating infants. If we try to determine the reason for it I suspect it may have much to do with the higher likelihood of success on the first attempt at placing an ETT. The longer it takes to place the ETT or the more number of attempts requiring intermittent PPV in a patient who truly needs a tube the greater the likelihood that you will see adverse events including bradycardia. It may simply be that a calm and still patient is an easier intubation and getting the tube in faster yields a more stable patient.
I am biased though and I think it is worth pointing out another possible reason for the differing results. One hospital in this study routinely used premedication and the other did not. Almost 3/4 of the patients came from one hospital which raises the possibility that skill set could be playing a role. If the skill of providers at the two hospitals differed, the results could reflect the variable skill in the practitioners versus the difference in the medications used themselves. What I don’t know though is whether the two share the same training program or not. Are the trainees the same at both sites (google maps says the two sites are 11 minutes away by car)? The difference still might be in local respiratory therapists or Neonatologists intubating as well. Regardless, the study provides evidence that paralysis makes a difference. To convince those out there though who remain skeptical I think we are going to need the registry to take part in a prospective trial using many centres. A format in which several centres that don’t use paralysis are compared to several who do routinely would help to sort out the concern in skill when looking only at two centres. This wouldn’t be randomized of course but I think it would be very difficult at this point to get a centre that strongly believes in using paralysis to randomize so a prospective study using groups chosen by the individual centre might be the next best thing. If anyone using the registry is reading this let me know what you think?
If I look back on my career there have been many things I have been passionate about but the one that sticks out as the most longstanding is premedicating newborns prior to non-emergent intubation. The bolded words in the last sentence are meant to reinforce that in the setting of a newborn who is deteriorating rapidly it would be inappropriate to wait for medications to be drawn up if the infant is already experiencing severe oxygen desaturation and/or bradycardia. The CPS Fetus and Newborn committee of which I am a member has a statement on the use of premedication which seems as relevant today as when it was first developed. In this statement the suggested cocktail of atropine, fentanyl and succinylcholine is recommended and having used it in our centre I can confirm that it is effective. In spite of this recommendation by our national organization there remain those who are skeptical of the need for this altogether and then there are others who continue to search for a better cocktail. Since I am at the annual conference for the CPS in Quebec city I thought it would be appropriate to provide a few comments on this topic.
Three concerns with rapid sequence induction (RSI) for premedication before intubation
1. “I don’t need it. I don’t have any trouble intubating a newborn” – This is perhaps the most common reason I hear naysayers raise. There is no question that an 60-90 kg practitioner can overpower a < 5kg infant and in particular an ELBW infant weighing < 1 kg. This misses the point though. Premedicating has been shown to increase success on the first attempt and shorten times to intubation. Dempsey 2006, Roberts 2006, Carbajal 2007, Lemyre 2009
2. “I usually get in on the first attempt and am very slick so risk of injury is less.” Not really true overall. No doubt there are those individuals who are highly successful but overall the risk of adverse events is reduced with premedication. (Marshall 1984, Lemyre 2009). I would also proudly add another Canadian study from Edmonton by Dr. Byrne and Dr. Barrington who performed 249 consecutive intubations with predication and noted minimal side effects but high success rates at first pass.
3. “Intubation is not a painful procedure”. This one is somewhat tough to obtain a true answer for as the neonate of course cannot speak to this. There is evidence available again from Canadian colleagues in 1984 and 1989 that would suggest that infants at the very least experience discomfort or show physiologic signs of stress when intubated using an “awake” approach. In 1984 Kelly and Finer in Edmonton published Nasotracheal intubation in the neonate: physiologic responses and effects of atropine and pancuronium. This randomized study of atropine with or without pancuronium vs control demonstrated intracranial hypertension only in those infants in the control arm with premedication ameliorating this finding. Similarly, in 1989 Barrington, Finer and the late Phil Etches also in Edmonton published Succinylcholine and atropine for premedication of the newborn infant before nasotracheal intubation: a randomized, controlled trial. This small study of 20 infants demonstrated the same finding of elimination of intracranial hypertension with premedication. At the very least I would suggest that having a laryngoscope blade put in your oral cavity while awake must be uncomfortable. If you still doubt that statement ask yourself whether you would want sedation if you needed to be intubated? Still feel the same way about babies not needing any?
4. What if I sedate and paralyze and there is a critical airway? Well this one may be something to consider. If one knows there is a large mass such as a cystic hygroma it may be best to leave the sedation or at least the paralysis out. The concern though that there might be an internal mass or obstruction that we just don’t know about seems a little unfounded as a justification for avoiding medications though.
Do we have the right cocktail?
The short answer is “I don’t know”. What I do know is that the use of atropine, an opioid and a muscle relaxant seems to provide good conditions for intubating newborns. We are in the era of refinement though and as a recent paper suggests, there could be alternatives to consider;Effect of Atropine With Propofol vs Atropine With Atracurium and Sufentanil on Oxygen Desaturation in Neonates Requiring Nonemergency IntubationA Randomized Clinical Trial. I personally like the idea of a two drug combination for intubating vs.. three as it leaves one less drug to worry about a medication error with. There are many papers out there looking at different drug combinations. This one though didn’t find a difference between the two combinations in terms of prolonged desaturations between the two groups which was the primary outcome. Interestingly though the process of intubating was longer with atropine and propofol. Given some peoples reluctance to use RSI at all, any drug combination which adds time to the the procedure is unlikely to go over well. Stay tuned though as I am sure there will be many other combinations over the next few years to try out!
One of the benefits of operating this site is that I often learn from the people reading these posts as they share their perspectives. On a recent trip I was reunited with Boubou Halberg a Neonatologist from Sweden whom I hadn’t seen in many years. I missed him on my last trip to Stockholm as I couldn’t make it to Karolinska University but we managed to meet each other in the end. As we caught up and he learned that I operated this site he passed along a paper of his that left an impact on me and I thought I would share with you.
When we think about treating an infant with a medicinal product, we often think about getting the right drug, right dose and right administration (IV, IM or oral) for maximum benefit to the patient. When it comes to nutrition we have certainly come a long way and have come to rely on registered dietitians where I work to handle a lot of the planning when it comes to getting the right prescription for our patients. We seem comfortable though making some assumptions when it comes to nutrition that we would never make with respect to their drug counterparts. More on that later…
A Swedish Journey to Ponder
Westin R and colleagues (one of whom is my above acquaintance) published a seven year retrospective nutritional journey in 2017 from Stockholm entitled Improved nutrition for extremely preterm infants: A population based observational study. After recognizing that over this seven year period they had made some significant changes to the way they approached nutrition, they chose to see what effect this had on growth of their infants from 22 0/7 to 26 6/7 weeks over this time by examining four epochs (2004-5, 2006-7, 2008-9 and 2010-11. What were these changes? They are summarized beautifully in the following figure.
Not included in the figure was a progressive change as well to a more aggressive position of early nutrition in the first few days of life using higher protein, fat and calories as well as changes to the type of lipid provided being initially soy based and then changing to one primarily derived from olive oil. Protein targets in the first days to weeks climbed from the low 2s to the mid 3s in gram/kg/d while provision of lipid as an example doubled from the first epoch to the last ending with a median lipid provision in the first three days of just over 2 g/kg/d.
While figure 3 from the paper demonstrates that regardless of time period there were declines in growth across all three measurements compared to expected growth patterns, when one compares the first epoch in 2004-2005 with the last 2010-11 there were significant protective effects of the nutritional strategy in place. The anticipated growth used as a standard was based on the Fenton growth curves.
What this tells us of course is that we have improved but still have work to do. Some of the nutritional sources as well were donor breast milk and based on comments coming back from this years Pediatric Academic Society meeting we may need to improve how that is prepared as growth failure is being noted in babies who are receiving donated rather than fresh mother’s own milk. I suspect there will be more on that as time goes by.
Knowing where you started is likely critical!
One advantage they have in Sweden is that they know what is actually in the breast milk they provide. Since 1998 the babies represented in this paper have had their nutritional support directed by analyzing what is in the milk provided by an analyzer. Knowing the caloric density and content of protein, carbohydrates and fats goes a long way to providing a nutritional prescription for individual infants. This is very much personalized medicine and it would appear the Swedes are ahead of the curve when it comes to this. in our units we have long assumed a caloric density of about 68 cal/100mL. What if a mother is producing milk akin to “skim milk” while another is producing a “milkshake”. This likely explains why some babies despite us being told they should be getting enough calories just seem to fail to thrive. I can only speculate what the growth curves shown above would look like if we did the same study in units that actually take a best guess as to the nutritional content of the milk they provide.
This paper gives me hope that when it comes to nutrition we are indeed moving in the right direction as most units become more aggressive with time. What we need to do though is think about nutrition no different than writing prescriptions for the drugs we use and use as much information as we can to get the dosing right for the individual patient!