I knew it was a matter of time before a study looking at this strategy came out. Whether you intubate using INSURE or a LISA/MIST technique (passing a semi-rigid catheter through the vocal cords to give surfactant while a baby is on CPAP) there would have to be those that argue the placement of the laryngoscope blade in the mouth and passage of the catheter through the trachea must be uncomfortable. Given such concerns, why wouldn’t you want to provide some sedation to the patient? The main concern would be suppression of respiratory drive and need for intubation or PPV. LISA/MIST usage has been found in systematic reviews to lead to less risk of BPD but what if sedation leads to more PPV especially with uncontrolled tidal volumes on these fragile lungs? Will the benefits remain?
Propofol Before MIST
Dekker et al published Sedation during minimal invasive surfactant therapy: a randomised controlled trial in which they looked at infants receiving surfactant administration by MIST in infants born at 26 – 36 weeks with stratification of results into two groups (26–31+6 and 32–36+6 weeks). The intervention was to give a relatively small dose of propofol 1 mg/kg compared to the typical dose of 2.5 mg/kg prior to using MIST. Physicians were unblinded to the intervention but nurses were asked (they were blinded) to determine the COMFORTneo score as a measure of discomfort or pain. The primary outcome was the percentage of infants with a score <14 during the procedure. A power calculation to determine numbers needed for the study indicated 39 per arm and was based on a previous study (not using propofol though). While it does not appear that a sham was used for a placebo arm, sucrose was utilized for additional comfort in both arms.
The Results Please
Sedation seemed to work even at this lower dose of propofol as the group who received it had a higher percentage with a score <14 (32/42 (76%) vs 8/36 (22%), p<0.001). Moreover, the overall mean scores were also lower (12±3 vs 17±4; p<0.001). When looking at rates of complications though some interesting but perhaps not surprising findings emerge.
A greater risk of desaturation events existed in the group receiving even a low dose of propofol. Not surprisingly with a greater risk of oxygen desaturation there was a need for more nasal intermittent mechanical ventilation (93% vs 47%). Finally, while the numbers are small the incidence of grade 3 or 4 IVH was 0 in the group that received no sedation and 5% in the group that did. While non significant it is worth pointing out that as with all of the listed complications, this study was underpowered for any meaningful secondary outcome conclusions. Given what we know about the interaction of positive pressure ventilation and risks of IVH it is just something to raise an eyebrow at for now.
What is the biggest problem with the study?
As I see it the absence of the placebo group such that the Neonatologists knew who received propofol and who did not makes it difficult to know if there really was a true need for PPV. In the sedation group the percentage of babies that received nasal IMV almost matches perfectly those that experienced desaturation events (93 and 91%) but in the non-sedated group it was 47% and 69%. Were Neonatologists more apt to let the desaturations sit without PPV if they knew the infant did not get propofol and conversely assume they needed PPV if they received a sedative? It seems to me that the study would have been improved with the use of the sham procedure. The scoring by the nurses who were blinded shows that even with a lower dose of propofol than normal though it still provides some sedation.
I would stay tuned in this area as I am sure this will not be the last we hear of this but for now I would suggest that sedation for MIST/LISA should not be routinely done, at least based on this study.
This post has the potential to be polarizing as sustained inflations while common as an approach after delivery in Europe has not been widely adopted in Canada and the United States. Some time ago I wrote about sustained inflations and a reader commented that I should wait for the results of the The Sustained Aeration for Infant Lungs (SAIL) trial before forming a final opinion on whether this is a good strategy or not. The previous blog post on this topic was Is It Time To Use Sustained Lung Inflation in NRP? and was followed by Is expired CO2 the key to making sustained inflation a standard in resuscitation? The first post concluded that there was a concerning trend towards more IVH in those who received sustained inflations (SI) while the second showed both a reduction in BPD and duration of mechanical ventilation with this approach. I suggested that maybe we were really onto something here and then I was asked to wait before coming to a conclusion until the SAIL trial was done. Well that day has come.
The SAIL trial
This trial led by Dr. H. Kirpalani and involving 18 NICUs in 9 countries was a big endeavour. The paper was just published and is entitled Effect of Sustained Inflations vs Intermittent Positive Pressure Ventilation on Bronchopulmonary Dysplasia or Death Among Extremely Preterm Infants The SAIL Randomized Clinical Trial. The trial compared SI of 15 seconds at a peak pressure of 20 cmH2O, followed if needed by a second SI of 15 seconds
at a peak of 25 cmH2O to traditional PPV for infants who after initial 30 seconds of CPAP required further intervention to establish breathing. These were provided via facemask or nasopharyngeal tune attached to a t-piece resuscitator. In both groups after the initial intervention standard resuscitation steps were carried out. The primary outcome was death or BPD at 36 weeks PMA. A data safety monitoring board (DSMB) was formed as well and it is this group that became very important to the conclusions of the study and led to its early termination. All infants were 23 weeks 0 days’ to 26 weeks 6 days’ GA. Before the study was terminated the final totals were 215 patients in the SI arm and 211 in the traditional PPV group.
The trial was stopped after the DSMB identified an excessive number of early deaths within 48 hours in the SI group. The findings were “11 of 16 early deaths in the sustained inflation group vs 1 of 3 in the standard
resuscitation group were considered possibly related to allocation group”. A number of these deaths occurred in the highest risk group of those born at 23-24 weeks but it was enough to stop recruitment.
With respect to the primary outcome the results showed no difference between the two approaches. In saying this however, if the study did not recruit enough patients as planned to demonstrate a difference one has to question whether the study had enough power to find a benefit.
To answer this question the authors performed a Bayesian Analysis to determine the probability that adding more patients would have led to a different conclusion. That is to determine if they would have found a difference favouring SI. In the end they found that their conclusions would not have changed. Sustained inflations in infants from 23 weeks 0 days’ to 26 weeks 6 days’ GA do not confer a benefit and may be associated with a higher likelihood of death within 48 hours of birth.
What do we do with these results?
I think this is it. I can’t see a research ethics board allowing another study at this point. This by neonatal standards was a big study given the relative scarcity of infants at these gestational ages. The fact that no difference was found in rates of death or survival with BPD for those at highest risk of these outcomes suggests to me that looking at older GA at birth will not produce different results. Sustained inflation to establish FRC and initiate respiration was a good concept backed by animal research. Moreover, clinical work out of Edmonton in recent years suggested potential benefits but with the publication of this study I suspect we will need as a neonatal community to look at other strategies to decrease rates of BPD. Concerns over increased risk of death in my opinion mean this ship has SAILed,
Just about all of our preterm infants born at <29 weeks start life out the same in terms of neurological injury. There are of course some infants who may have suffered ischemic injury in utero or an IVH but most are born with their story yet to be told. I think intuitively we have known for some time that the way we resuscitate matters. Establishing an FRC by inflating the lungs of these infants after delivery is a must but as the saying goes the devil is in the details.
One hundred and sixty five infants comprised this cohort. Overall, 124 (75%) infants were in the high volume group compared to 41 (25%) with a mean VT<6 mL/kg. Median Vt were 5.3 (4.6-5.7) ml/kg for the low group and 8.7
(7.3-10.6) mL/kg which were significantly different. When looking at the rates of IVH and the severity of those affected the results are striking as shown in the table. Hydrocephalus, following IVH developed in 7/49 (14%) and 2/16 (13%) in the >6 mL/kg and <6 mL/kg VT groups. Looking at other factors that could affect the outcome of interest the authors noted the following physiologic findings. Oxygen saturations were lower in the low volume group at 6, 13 and 14 min after birth while tissue oxygenation as measured by NIRS was similarly lower at 7,8 and 25 min after birth (P<0.001). Conversely, heart rate was significantly lower in the VT>6 mL/kg group at 5, 20 and 25 min after birth (P<0.001). Fraction of inspired oxygen was similar in both groups within the first 30 min. Systolic, diastolic and mean blood pressure was similar between the groups. What these results say to me is that despite having lower oxygen saturations and cerebral oxygen saturation at various time points in the first 25 minutes of life the infants seem to be better off given that HR was lower in those given higher volumes despite similar FiO2. Rates of volume support after admission were slightly higher in the high volume group but inotrope usage appears to be not significantly different. Prophylactic indomethacin was used equally in the two cohorts.
Thoughts for the future
Once a preterm infant is admitted to the NICU we start volume targeted ventilation from the start. In the delivery room we may think that we do the same by putting such infants on a volume guarantee mode after intubation but the period prior to that is generally done with a bag and mask. Whether you use a t-piece resuscitator or an anesthesia bag or even a self inflating bag, you are using a pressure and hoping not to overdistend the alveoli. What I think this study demonstrates similar to the previous work by this group is that there is another way. If we are so concerned about volutrauma in the NICU then why should we feel any differently about the first few minutes of life. Impairment of venous return from the head is likely to account for a higher risk of IVH and while a larger study may be wished for, the results here are fairly dramatic. Turning the question around, one could ask if there is harm in using a volume targeted strategy in the delivery room? I think we would be hard pressed to say that keeping the volumes under 6 mL/kg is a bad idea. The challenge as I see it now is whether we rig up devices to accomplish this or do the large medical equipment providers develop an all in one system to accomplish this? I think the time has come to do so and will be first in line to try it out if there is a possibility to do a trial.
We have all been there. After an uneventful pregnancy a mother presents to the labour floor in active labour. The families world is turned upside down and she goes on to deliver an infant at 27 weeks. If the infant is well and receives minimal resuscitation and is on CPAP we provide reassurance and have an optimistic tone. If however their infant is born apneic and bradycardic and goes on to receive chest compressions +/- epinephrine what do we tell them? This infant obviously is much sicker after delivery and when the family asks you “will my baby be ok?” what do you tell them? It is a human tendency to want to reassure and support but if they ask you what the chances are of a good outcome it has always been hard to estimate. What many of us would default to is making an assumption that the need for CPR at a time when the brain is so fragile may lead to bleeding or ischemia would lead to worse outcomes. You would mostly be right. One study by Finer et al entitled Intact survival in extremely low birth weight infants after delivery room resuscitation.demonstrated that survival for infants under 750g was better if they had a history of CPR after delivery. The thought here is that more aggressive resusctiation might be responsible for the better outcome by I would presume establishing adequate circulation sooner even if the neonates did not appear to need it immediately.
The Canadian Neonatal Network
In Canada we are fortunate to have a wonderful network called the Canadian Neonatal Network. So many questions have been answered by examining this rich database of NICUs across the county. Using this database the following paper was just published by Dr. A. Lodha and others; Extensive cardiopulmonary resuscitation of preterm neonates at birth and mortality and developmental outcomes. The paper asked a very specific and answerable question from the database. For infants born at <29 weeks gestational age who require extensive resuscitation (chest compressions, epinephrine or both) what is the likelihood of survival and/or neurodevelopmental impairment (NDI) at 18-24 months of age vs those that did not undergo such resuscitation? For NDI, the authors used a fairly standard definition as “any cerebral palsy (GMFCS1), Bayley-III score <85 on one or more of the cognitive, motor or language composite scores, sensorineural or mixed hearing impairment or unilateral or bilateral visual impairment.” Their secondary outcomes were significant neurodevelopmental impairment (sNDI), mortality, a Bayley-III score of <85 on any one of the components (cognitive, language, motor), sensorineural or mixed hearing loss,or visual impairment.sNDI was defined as the presence of one or more of the following: cerebral palsy with GMFCS 3, Bayley-III cognitive, language or motor composite score <70, hearing impairment requiring hearing aids or cochlear implant, or bilateral visual impairment”
What did they discover?
It is a fortunate thing that the database is so large as when you are looking at something like this the number of infants requiring extensive resuscitation is expected to be small. The authors collected data from January 1, 2010 and September 30, 2011 and had a total number of infants born at less than 29 weeks of 2760. After excluding those with congenital anomalies and those who were born moribund they were left with 2587. From these 80% had follow-up data and when applying the final filter of extensive resuscitation they were left with 190 (9.2%) who received delivery room CPR (DR-CPR) vs 1545 who did not receive this.
Before delving into the actual outcomes it is important to note that neonates who did not receive DR-CPR were more likely to be born to mothers with hypertension and to have received antenatal steroids (89 vs 75%). With these caveats it is pretty clear that as opposed to the earlier study showing better outcomes after DR-CPR this was not the case here.
The results are interesting in that it is pretty clear that receiving DR-CPR is not without consequence (higher rate of seizures, severe neurological injury, BPD). Looking at the longer term outcomes though is where things get a little more interesting. Mortality and mortality or neurodevelopmental impairment are statistically significant with respect to increased risk. When you take out NDI alone however the CI crosses one and is no longer significant. Neither is CP for that matter with the only statistically significant difference being the Bayley-III Motor composite score <85. The fact that only this one finding came out as significant at least to me raises the possibility that this could have been brought about by chance. It would seem that while these infants are at risk of some serious issues their brains in the long run may be benefiting for the neurological plasticity that we know these infants have.
The study is remarkable to me in that an infant can have such a difficult start to life yet hope may remain even after dealing with some of the trials and tribulations of the NICU. Parents may need to wade through the troubling times of seizures, long term ventilation and CPAP and then onto a diagosis of BPD but their brains may be ok after all. This is one of the reasons I love what I do!
Recently the practice of keeping ELBW infants with a midline head position for the first three days of life has been recommended to reduce IVH as part of a bundle in many units. The evidence that this helps to reduce IVH has been somewhat circumstantial thus far. Studies finding that decreased sagittal sinus blood flow, increased cerebral blood volume with increased intracranial pressure all occur after head turns would theoretically increase the risk of IVH. Raising the head of the bed would help in theory with drainage of the venous blood from the head and in fact systemic oxygenation has been shown to improve with such positioning. This presumably is related to increased cardiac output from better systemic venous return.
Bringing it to the bedside
Interestingly, some of the above studies are from over thirty years ago. We now have some evidence to look at involving this practice. Kochan M et al published Elevated midline head positioning of extremely low birth weight infants: effects on cardiopulmonary function and the incidence of
periventricular-intraventricular. The study involved maintaining ELBW infants in an elevated midline head position (ELEV- supine, head of bed elevated 30 degrees, head kept in midline) versus standard head positioning (FLAT–flat supine, head turned 180 degrees every 4 h) during the first 4 days of life to see if this would decrease in the incidence of IVH. Ninety infants were randomized into both arms of the study. In terms of baseline characteristics, BW of 725g in the FLAT vs 739 in ELEV were comparable as well as GA both at 25 weeks. Two differences on the maternal side existed of 40% ELEV vs 24.4% FLAT of mothers having preeclampsia and 23.3% FLAT vs 10% ELEV having prolonged rupture of membranes both of which were statistically significant.
What did they find?
Ultrasounds were performed at entry into the study and then daily for days 1-4 and then on day 7 with abnormal scans repeated weekly. In terms of IVH the authors noted no overall difference in rate of IVH. What they did find however was a statistically significant reduction in the rate of Grade IV IVH. The p value for the finding of lower rates of Grade IV IVH was 0.036 so not strikingly significant but different nonetheless. Given that the venous drainage of the head is also dependent on the resistance to flow from the pressure in the thorax one can’t infer that the intervention alone is responsible for this without ensuring that that respiratory findings are similar as well. Similarly without knowing inflow of blood into the head as measured by blood pressure it is difficult to say that the reduction in IVH isn’t related to differences in blood pressure.
The authors helpfully looked at both of these things. For those infants on high frequency ventilation the mean airway pressure was higher on day one being 11.5 cm H2O (FLAT) vs 9.9 cm H2O (ELEV) neither of which are high although different. The rest of the three days were no different. For those on conventional ventilation the only difference was on day 4 where the MAP was higher for ELEV at 8 vs 7.4 cm H2O which again is fairly mild. Interestingly, as was found in other studies that oxygenation was improved with elevation of the head, the maximum FiO2 for the two groups was different on day 1 being 46% in the FLAT vs 37.5% in the ELEV.
Looking at the hemodynamic side of things there were differences in the lowest mean BP recorded on day 1 and 3 but otherwise the groups were similar. It would have been nice to see mean results during this time rather than lowest but this is what we have.
In terms of complications of preterm birth there were no differences found in rates of sepsis (important given the increase rate of prolonged rupture in the FLAT group), NEC or ROP.
Although length of stay was no different 92 vs 109 days ELEV (NS), survival to discharge was at 88% vs 76% (p=0.033) which also may explain the longer length of stay.
What Can We Learn From This
Don’t worry. I am not about to throw the results out. There are a couple observations though that need to be addressed. The first is the increased rate of preecampsia in the ELEV group. This finding could have impacted the results. We know that fetuses exposed to this condition are stressed and are often born with better lungs than their non-exposed counterparts. The endogenous increase in steroids due to this stress is attributable and may explain the better oxygenation and lower mean airway pressures needed in the ELEV group rather than improvements in flow alone from positioning. The second issue is adherence to the protocol as there were some infants in the ELEV group who were placed flat for the final 1-2 days of the study. Having said that, this would serve to dilute the effect rather than strengthen it so perhaps it makes the results more believable.
So where does this leave us? This study demonstrates improved survival and a reduction in Grade IV IVH without an overall reduction in IVH. There was nothing found to suggest that the intervention is harmful. Given the background studies demonstrating improved systemic oxygenation, reductions in ICP and cerebral blood volume the finding of reduced severe IVH seems plausible to me. This could be a practice changing study for some units who have perhaps only adopted midline positioning in the first few days of life. It will be interesting to see if this takes off but is certainly worth a good look at.