As a Neonatologist, there is no question that I am supportive of breast milk for preterm infants. When I first meet a family I ask the question “are you planning on breastfeeding” and know that other members of our team do the same. Before I get into the rest of this post, I realize that while breast milk may be optimal for these infants there are mother’s who can’t or won’t for a variety of reasons produce enough breast milk for their infants. Fortunately in Manitoba and many other places in the world breast milk banks have been developed to provide donor milk for supporting these families. Avoidance of formula in the early days to weeks of a ELBWs life carries benefits such as a reduction in NEC which is something we all want to see.
Mother’s own milk though is known to have additional benefits compared to donor milk which requires processing and in so doing removes some important qualities. Mother’s own milk contains more immunologic properties than donor including increased amounts of lactoferrin and contains bioactive cells. Growth on donor human milk is also reduced compared to mothers’ own milk and lastly since donor milk is obtained from mothers producing term milk there will be properties that differ from that of mothers producing fresh breast milk in the preterm period. I have no doubt there are many more detailed differences but for basic differences are these and form the basis for what is to come.
The Dose Response Effect of Mother’s Own Milk
Breast milk is a powerful thing. Previous studies on the impact of mother’s own milk (MOM) have shown that with every increment of 10 mL/kg/d of average intake, the risk of such outcomes as BPD and adverse developmental outcomes are decreased. In the case of BPD the effect is considerable with a 9.5% reduction in the odds of BPD for every 10% increase in MOM dose. With respect to developmental outcome ach 10 mL/kg/day increase in MOM was associated with a 0.35 increase in cognitive index score.
The same group just published another paper on this cohort looking at a different angle. NICU human milk dose and health care use after NICU discharge in very low birth weight infants. This study is as described and again looked at the impact of every 10 mL/kg increase in MOM at two time points; the first 14 and the first 28 days of life. Although the data for the LOVE MOM trial was collected prospectively it is important to recognize how the data for this study was procured. At the first visit after NICU discharge the caregiver was asked about hospitalizations, ED visits and specialized therapies and specialist appointments. These were all tracked at 4 and 8 months of corrected age were added to yield health care utilization in the first year, and the number of visits or provider types at 4, 8, and 20 months of corrected age provided health care utilization through 2 years.
What were the results?
“Each 10 mL/kg/day increase in HM in the first 14 days of life was associated with 0.26 fewer hospitalizations (p =
0.04) at 1 year and 0.21 fewer pediatric subspecialist types (p = 0.04) and 0.20 fewer specialized therapy types (p = 0.04) at 2 years.” The results at 28 days were not statistically significant. The authors reported both unadjusted and adjusted results controlling for many factors such as gestational age, completion of appointments and maternal education to name a few which may have influenced the results. The message therefore is that the more of MOM a VLBW is provided in the first 14 days of life, the better off they are in the first two years of life with respect to health care utilization.
That even makes some sense to me. The highest acuity typically for such infants is the first couple of weeks when they are dealing with RDS, PDA, higher oxygen requirements etc. Could the protective effects of MOM have the greatest bang for your buck during this time. By the time you reach 28 days is the effect less pronounced as you have selected out a different group of infants at that time point?
What is the weakness here though? The biggest risk I see in a study like this is recall bias. Many VLBW infants who leave the NICU have multiple issues requiring many different care providers and services. Some families might keep rigorous records of all appointments in a book while others might document some and not others. The big risk here in this study is that it is possible that some parents overstated the utilization rates and others under-reported. Not intentionally but if you have had 20 appointments in the first eight months could the number really by 18 or 22?
Another possibility is that infants receiving higher doses of MOM were healthier at the outset. Maternal stress may decrease milk production so might mothers who had healthier infants have been able to produce more milk? Are healthier infants in the first 14 days of life less likely to require more health care needs in the long term?
How do we use this information?
In spite of the caveats that I mentioned above there are multiple papers now showing the same thing. With each increment of 10 mL/kg of MOM benefits will be seen. It is not a binary effect meaning breastfed vs not. Rather much like the medications we use to treat a myriad of conditions there appears to be a dose response. It is not enough to ask the question “Are you intending to breastfeed?”. Rather it is incumbent on all of us to ask the follow-up question when a mother says yes; “How can we help you increase your production?” if that is what the family wants>
Much has been written on the topic of cord clamping. There is delayed cord clamping of course but institutions differ on the recommended duration. Thirty seconds, one minute or two or even sometimes three have been advocated for but in the end do we really know what is right? Then there is also the possibility of cord milking which has gained variable traction over the years. A recent review was published here.
Take the Guessing Out of the Picture?
Up until the time of birth there is very little pulmonary blood flow. Typically, about 10% of the cardiac output passes through the lungs and the remained either moves up the ascending aorta or bypasses the lungs via the ductus arteriosus. After birth as the lung expands, pulmonary vascular resistance rapidly decreases allowing cardiac output to take on the familiar pattern which we all live with. Blood returning from the systemic venous circulation no longer bypasses the lung but instead flows through pulmonary capillaries picking up oxygen along the way. One can imagine then that if a baby is born and the cord is clamped right away, blood returning from the systemic circulation continues to bypass the lung which could lead to hypoxemia and reflexive bradycardia. This has been described previously by Blank et al in their paper Haemodynamic effects of umbilical cord milking in premature sheep during the neonatal transition.
A group of researchers from the Netherlands published a very interesting paper Physiological-based cord clamping in preterm infants using a new purpose-built resuscitation table: a feasibility study this month. The study centres around a resuscitation table called the Concord that is brought to the mother for resuscitation after birth. The intervention here was applied to infants 26 to 35 weeks gestational age. The cord was clamped after each of the following was achieved for an infant indicating successful transition with opening of the lung and establishment of an FRC.
1. Establishment of adequate breathing (average tidal volume ≥4 mL/kg) on CPAP. They used a mask capable of measuring expired tidal volumes.
2. HR above 100 bpm
3. SpO2 above 25th percentile using FiO2 <0.4
In this way, the cord was only clamped once the baby appeared to have physiologically made the transition from dependence on umbilical cord blood flow to ventilation perfusion matching in the lung. Although 82 mothers consented only 37 preterm infants were included in the end. Exclusion criteria were signs of placental abruption or placenta praevia, signs of severe fetal distress determined by the clinician and the necessity for an emergency caesarean section ordered to be executed within 15 min. This really was a proof of concept study but the results are definitely worth looking at.
How Did These Babies Do?
There are many interesting findings from this study. The mean time of cord clamping was 4 minutes and 23 seconds (IQR 3:00 – 5:11). Heart rate was 113 (81–143) and 144 (129–155) bpm at 1 min and 5 min
after birth. Only one patient developed bradycardia to <60 BPM but this was during a mask readjustement. The main issue noted as far as adverse events was hypothermia with a mean temperature of 36.0 degrees at NICU admission. Almost 50% of infants had a temperature below 36 degrees. Although the authors clearly indicate that they took measures to prevent heat loss it would appear that this could be improved upon!
What stands out most to me is the lengthy duration of cord clamping. This study which used a physiologic basis to determine when to clamp a cord has demonstrated that even at 1 minute of waiting that is likely only 1/4 of the time needed to wait for lung expansion to occur to any significant degree. I can’t help but wonder how many of the patients we see between 26-35 weeks who have a low heart rate after delivery might have a higher heart rate if they were given far more time than we currently provide for cord clamping.
I can also see why cord milking may be less effective. Yes, you will increase circulating blood volume which may help with hemodynamic stability but perhaps the key here is lung expansion. You can transfuse all the blood you want but if it has nowhere to go just how effective is it?
As we do more work in this area I have to believe that as a Neonatal community we need to prepare ourselves for the coming of the longer delay for cord clamping. Do we need to really have the “Concord” in every delivery or perhaps it is time to truly look at durations of 3-4 minutes before the team clamps the cord.
One of the first things a student of any discipline caring for newborns is how to calculate the apgar score at birth. Over 60 years ago Virginia Apgar created this score as a means of giving care providers a consistent snapshot of what an infant was like in the first minute then fifth and if needed 10, 15 and so on if resuscitation was ongoing. For sure it has served a useful purpose as an apgar score of 0 and 0 gives one cause for real worry. What about a baby with an apgar of 3 and 7 or 4 and 8? There are certainly infants who have done very well who initially had low apgar scores and conversely those who had higher apgar scores who have had very significant deleterious outcomes including death. I don’t mean to suggest that the apgar scores don’t provide any useful predictive value as they are used as part of the criteria to determine if a baby merits whole body cooling or not. The question is though after 60+ years, has another score been created to provide similar information but enhance the predictive value derived from a score?
The Neonatal Resuscitation and Adaptation Score (NRAS)
Back in 2015 Jurdi et al published Evaluation of a Comprehensive Delivery Room Neonatal Resuscitation and Adaptation Score (NRAS) Compared to the Apgar Score. This new score added into a ten point score resuscitative actions taken at the 1 and 5 minute time points to create a more functional score that included interventions. The other thing this new score addressed was more recent data that indicated a blue baby at birth is normal (which is why we have eliminated asking the question “is the baby pink?” in NRP. Knowing that, the colour of the baby in the apgar score may not really be that relevant. Take for example a baby with an apgar score of 3 at one minute who could have a HR over 100 and be limp, blue and with shallow breathing. Such a baby might get a few positive pressure breaths and then within 10 seconds be breathing quite well and crying. Conversely, they might be getting ongoing PPV for several minutes and need oxygen. Were they also getting chest compressions? If I only told you the apgar score you wouldn’t have much to go on. Now look at the NRAS and compare the information gathered using two cardiovascular (C1&2), one neurological test (N1) and two respiratory assessments (R1&2).
The authors in this study performed a pilot study on only on 17 patients really as a proof of concept that the score could be taught and implemented. Providers reported both scores and found “superior interrater reliability (P < .001) and respiratory component reliability (P < .001) for all gestational ages compared to the Apgar score.”
A Bigger Study Was Needed
The same group in 2018 this time led by Witcher published Neonatal Resuscitation and Adaptation Score vs Apgar: newborn assessment and predictive ability. The primary outcome was the ability of a low score to predict mortality with a study design that was a non-inferiority trial. All attended deliveries were meant to have both scores done but due to limited numbers of trained personnel who could appropriately administer both scores just under 90% of the total deliveries were assigned scores for comparison. The authors sought to recruit 450 infants to show that a low NRAS score (0–3) would not be inferior to a similar Apgar at predicting death. Interestingly an interim analysis found the NRAS to be superior to Apgar when 75.5% of the 450 were enrolled, so the study was stopped. What led the apgar score to perform poorly in predicting mortality (there were only 12 deaths though in the cohort) was the fact that 49 patients with a 1 minute apgar score of 0-3 survived compared to only 7 infants with a low NRAS score.
The other interesting finding was the ability of the NRAS to predict the need for respiratory support at 48 hours with a one minute apgar score of 0-3 being found in 39% of those on support compared to 100% of those with a low NRAS. Also at 5 minutes a score of 4-6 for the apgar was found in 48% of those with respiratory support at 48 hours vs 87% of those with a similar range NRAS. These findings were statistically significant while a host of other conditions such as sepsis, hypoglycemia, hypothermia and others were no different in terms of predictive ability of the scores.
An Even Bigger Study is Needed
To be sure, this study is still small and missed just over 90% of all deliveries so it is possible there is some bias that is not being detected here. I do think there is something here though which a bigger study that has an army of people equipped to provide the scoring will add to this ongoing story. Every practitioner who resuscitates an infant is asked at some point in those first minutes to hour “will my baby be ok?”. The truth is that the apgar score has never lived up to the hope that it would help us provide an accurate clairvoyant picture of what lies ahead for an infant. Where this score gives me hope is that a score which would at the very least help me predict whether an infant would likely still be needing respiratory support in 48 hours provides the basic answer to the most common question we get in the unit once admitted; “when can I take my baby home”. Using this score I could respond with some greater confidence in saying “I think your infant will be on support for at least 48 hours”. The bigger question though which thankfully we don’t have to address too often for the sickest babies at birth is “will my baby survive?”. If a larger study demonstrates this score to provide a greater degree of accuracy then the “Tipping Point” might just be that to switching over to the NRAS and leaving the apgar score behind. That will never happen overnight but medicine is always evolving and with time you the reader may find yourself becoming very familiar with this score!
A real change is coming and with this post you will get a glimpse into where the next big thing in Neonatology is likely to be. A catchy title for sure and also an exaggeration as I don’t see us abandoning the endotracheal tube just yet. There has been a lot of talk about less invasive means of giving surfactant and the last few years have seen several papers relating to giving surfactant via a catheter placed in the trachea (MIST or LISA techniques as examples). There may be a new kid on the block so to speak and that is aerosolized surfactant. This has been talked about for some time as well but the challenge had been figuring out how to aerosolize the fluid in such a way that a significant amount of the surfactant would actually enter the trachea. This was really a dream of many Neonatologists and based on a recently published paper the time may be now for this technique to take off.
A Randomized Trial of Aerosolized Surfacant
Minocchieri et al as part of the CureNeb study team published Nebulised surfactant to reduce severity of respiratory distress: a blinded, parallel, randomised controlled trial. This trial set out to obtain a sample size of 70 patients between 29 0/7 to 33 6/7 weeks to demonstrate a difference in need for intubation from 30% down to 5% in patients treated with CPAP (30% was based on the historical average). The authors recognizing that the babies in this GA bracket might behave differently, further stratified the randomization into two groups being 29 0/7 – 31 6/7 weeks and 32 0/7 to 33 6/7 weeks. Those babies who were on CPAP and met the following criteria for intubation were either intubated in the control group and given surfactant (curosurf) using the same protocol as those nebulized or had surfactant delivered via nebulisation (200 mg/kg: poractant alfa) using a customised vibrating membrane nebuliser (eFlow neonatal). Surfactant nebulisation(100 mg/kg) was repeated after 12 hours if oxygen was still required. The primary dichotomous outcome was the need for intubation within 72 hours of life, and the primary continuous outcome was the mean duration of mechanical ventilation at 72 hours of age.
Criteria for intubation
1. FiO2 >0.35 over more than 30 min OR FiO2 >0.45 at
2. More than four apnea/hour OR two apnea requiring BVM
3. Two cap gases with pH <7.2 and PaCO2 >65 mm Hg (or) >60 mm Hg if arterial blood gas sample).
4. Intubation deemed necessary by the attending physician.
Did It Work?
Eureka! It seemed to work as 11 of 32 infants were intubated in the surfactant nebulisation group within 72 hours of birth vs.22 out of 32 infants receiving CPAP alone (RR (95% CI)=0.526 (0.292 to 0.950)). The reduction though was accounted for by the bigger babies in the 32 0/7 to 33 6/7 weeks group as only 1 of 11 was intubated when given nebulized surfactant compared to 10 of 13 managed with CPAP. The duration of ventilation in the first 72 hours was not different between the groups: the median (range) 0 (0–62) hour for the nebulization group and 9 (0–64) hours for the control group (p=0.220). It is important in seeing these results that the clinicians deciding whether infants should be intubated for surfactant administration were blind to the arm the infants were in. All administration of curosurf via nebulization or sham procedures were done behind a screen.
The total number of infants randomized were 66 so they did fall shy of the necessary recruitment but since they did find a difference the results seem valid. Importantly, there were no differences in complications although I can’t be totally confident there really is no risk as this study was grossly underpowered to look at rarer outcomes.
Breaking down the results
This study has me excited as what it shows is that “it kind of works“. Why would larger babies be the ones to benefit the most? My guess is that some but not a lot of surfactant administered via nebulization reaches the alveoli. Infants with lesser degrees of surfactant deficiency (32 0/7 to 33 6/7) weeks might get just enough to manage without an endotracheal tube. Those infants (in particular less than 32 0/7 weeks) who have more significant surfactant deficiency don’t get enough and therefore are intubated. Supporting this notion is the overall delay in time to intubation in those who were intubated despite nebulization (11.6 hours in the nebulization group vs 4.9 hours in the control arm). They likely received some deposition in the distal alveoli but not enough to completely stave off an endotracheal tube.
One concerning point from the study though had to do with the group of infants who were intubated despite nebulization of surfactant. When you look at total duration of ventilation (hours) it was 14.6 (9.0–24.8) in the control arm vs 25.4 (14.6–42.2) p= 0.029*. In other words infants who were intubated in the end spent about twice as long intubated as those who were intubated straight away. Not a huge concern if you are born at 32 weeks or more but those additional thousands of positive pressure breaths are more worrisome as a risk for CLD down the road.
As it stands, if you had an infant who was 33 weeks and grunting with an FiO2 of 35% might you try this if you could get your hands on the nebulizer? It appears to work so the only question is whether you are confident enough that the risk of such things as pneumothorax or IVH isn’t higher if intubation is delayed. It will be interesting to see if this gets adopted at this point.
The future no doubt will see a refinement of the nebulizer and an attempt to see how well this technique works in infants below 29 weeks. It is in this group though that prolonging time intubated would be more worrisome. I don’t want to dismiss this outright as I see this as a pilot study that will lead the way for future work that will refine this technique. If we get this right this would be really transformative to Neonatology and just might be the next big leap.
The modern NICU is one that is full of patients on CPAP these days. As I have mentioned before, the opportunity to intubate is therefore becoming more and more rare is non-invasive pressure support becomes the mainstay of therapy. Even for those with established skills in placing an endotracheal tube, the number of times one gets to do this per year is certainly becoming fewer and fewer. Coming to the rescue is the promise of easier intubations by being able to visualize an airway on a screen using a video laryngoscope. The advantage to the user is that anyone who is watching can give you some great tips and armed with this knowledge you may be better able to determine how to adjust your approach.
For those of you who have followed the blog for some time, you will recall this is not the first time video laryngoscopy has come up. I have spoken about this before in Can Video Laryngoscopy Improve Trainee Success in Intubation. In that piece, the case was made that training residents how to intubate using a video laryngoscope (VL) improves their success rate. An additional question that one might ask though has to do with the quality of the intubation. What if you can place a tube using a video laryngoscope but the patient suffers in some way from having that piece of equipment in the mouth? Lucky for us some researchers from the Children’s Hospital of Philadelphia have completed a study that can help answer this additional question.
Video Laryngoscopy may work but does it cause more harm than good?
Using a video laryngoscope requires purchasing one first and they aren’t necessarily cheap. If they were to provide a better patient experience though the added cost might well be worth it. Pouppirt NR et al published Association Between Video Laryngoscopy and Adverse Tracheal Intubation-Associated Events in the Neonatal Care Unit. This study was a retrospective comparison of two groups; one having an intubation performed with a VL (n=161 or 20% of the group) and the other with a standard laryngoscope (644 or 80% of the group). The study relied on the use of the National Emergency Airway Registry for Neonates (NEAR4NEOs), which records all intubations from a number of centres using an online database and allows for analysis of many different aspects of intubations in neonates. In this case the data utilized though was from their centre only to minimize variation in premedication and practitioner experience.
Tracheal intubation adverse events (TIAEs) were 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 intubation, 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.
Looking at the patient characteristics and outcomes, some interesting findings emerge.
Patients who had the use of the VL were older and weighed more. They were more likely to have the VL used for airway obstruction than respiratory failure and importantly were also more likely to receive sedation/analgesia and paralysis. These researchers have also recently shown that the use of paralysis is associated with less TIAEs so one needs to bear this in mind when looking at the rates of TIAEs. There were a statistically significant difference in TIAEs of any type of 6% in the VL group to 19% in the traditional laryngoscopy arm but severe TIAEs showed not difference.
Given that several of the baseline characteristics might play a role in explaining why VL seemed superior in terms of minimizing risk of TIAEs by two thirds, the authors performed a multivariable analysis in which they took all factors that were different into account and then looked to see if there was still an effect of the VL despite these seemingly important differences. Interestingly, us of VL showed an Odds ratio of 0.43 (0.21,0.87 95% CI) in spite of these differences.
What does it mean?
Video laryngoscopy appears to make a difference to reducing the risk on TIAEs as an independent factor. The most common TIAE was esophageal intubation at 10% and reducing that is a good thing as it leads to fewer intubation attempts. This was also sen as the first attempt success was 63% in the VL group vs 44% in the other.
Now we need to acknowledge that this was not a randomized controlled trial so it could indeed be that there are other factors that the authors have not identified that led to improvements in TIAEs as well. What makes this study so robust though is the rigour with which the centre documents all of their intubations using such a detailed registry. By using one centre much of the variability in practice between units is eliminated so perhaps these results can be trusted. Would your centre achieve these same results? Maybe not but it would certainly be interesting to test drive one of these for a period of time see how it performs.