It has been over two years since I have written on this subject and it continues to be something that I get excited about whenever a publication comes my way on the topic. The last time I looked at this topic it was after the publication of a randomized trial comparing in which one arm was provided automated FiO2 adjustments while on ventilatory support and the other by manual change. Automated adjustments of FiO2. Ready for prime time? In this post I concluded that the technology was promising but like many new strategies needed to be proven in the real world. The study that the post was based on examined a 24 hour period and while the results were indeed impressive it left one wondering whether longer periods of use would demonstrate the same results. Moreover, one also has to be wary of the Hawthorne Effect whereby the results during a study may be improved simply by being part of a study.
The Real World Demonstration
So the same group decided to look at this again but in this case did a before and after comparison. The study looked at a group of preterm infants under 30 weeks gestational age born from May – August 2015 and compared them to August to January 2016. The change in practice with the implementation of the CLiO2 system with the Avea ventilator occurred in August which allowed two groups to be looked at over a relatively short period of time with staff that would have seen little change before and after. The study in question is by Van Zanten HA The effect of implementing an automated oxygen control on oxygen saturation in preterm infants. For the study the target range of FiO2 for both time periods was 90 – 95% and the primary outcome was the percentage of time spent in this range. Secondary outcomes included time with FiO2 at > 95% (Hyperoxemia) and < 90, <85 and < 80% (hypoxemia). Data were collected when infants received respiratory support by the AVEA and onlyincluded for analysis when supplemental oxygen was given, until the infants reached a GA of 32 weeks
As you might expect since a computer was controlling the FiO2 using a feedback loop from the saturation monitor it would be a little more accurate and immediate in manipulating FiO2 than a bedside nurse who has many other tasks to manage during the care of an infant. As such the median saturation was right in the middle of the range at 93% when automated and 94% when manual control was used. Not much difference there but as was seen in the shorter 24 hour study, the distribution around the median was tighter with automation. Specifically with respect to ranges, hyperoxemia and hypoxemia the following was noted (first number is manual and second comparison automated in each case).
Time spent in target range: 48.4 (41.5–56.4)% vs 61.9 (48.5–72.3)%; p<0.01
Hyperoxemia >95%: 41.9 (30.6–49.4)% vs 19.3 (11.5–24.5)%; p<0.001
< 90%: 8.6 (7.2–11.7)% vs 15.1 (14.0–21.1)%;p<0.0001
< 85%: 2.7 (1.4–4.0)% vs 3.2 (1.8–5.1)%; ns
Hypoxemia < 80%: 1.1 (0.4–1.7)% vs 0.9 (0.5–2.1)%; ns
What does it all mean?
I find it quite interesting that while hyperoxemia is reduced, the incidence of saturations under 90% is increased with automation. I suspect the answer to this lies in the algorithmic control of the FiO2. With manual control the person at the bedside may turn up a patient (and leave them there a little while) who in particular has quite labile saturations which might explain the tendency towards higher oxygen saturations. This would have the effect of shifting the curve upwards and likely explains in part why the oxygen saturation median is slightly higher with manual control. With the algorithm in the CLiO2 there is likely a tendency to respond more gradually to changes in oxygen saturation so as not to overshoot and hyperoxygenate the patient. For a patient with labile oxygen saturations this would have a similar effect on the bottom end of the range such that patients might be expected to drift a little lower then the target of 90% as the automation corrects for the downward trend. This is supported by the fact that when you look at what is causing the increase in percentage of time below 90% it really is the category of 85-89%.
Is this safe? There will no doubt be people reading this that see the last line and immediately have flashbacks to the SUPPORT trial which created a great deal of stress in the scientific community when the patients in the 85-89% arm of the trial experienced higher than expected mortality. It remains unclear what the cause of this increased mortality was and in truth in our own unit we accept 88 – 92% as an acceptable range. I have no doubt there are units that in an attempt to lessen the rate of ROP may allow saturations to drop as low as 85% so I continue to think this strategy of using automation is a viable one.
For now the issue is one of a ventilator that is capable of doing this. If not for the ventilated patient at least for patients on CPAP. In our centre we don’t use the Avea model so that system is out. With the system we use for ventilation there is also no option. We are anxiously awaiting the availability of an automated system for our CPAP device. I hope to be able to share our own experience positively when that comes to the market. From my standpoint there is enough to do at the bedside. Having a reliable system to control the FiO2 and minimize oxidative stress is something that may make a real difference for the babies we care for and is something I am eager to see.
Given that many preterm infants as they near term equivalent age are ready to go home it is common practice to discontinue caffeine sometime between 33-34 weeks PMA. We do this as we try to time the readiness for discharge in terms of feeding, to the desire to see how infants fare off caffeine. In general, most units I believe try to send babies home without caffeine so we do our best to judge the right timing in stopping this medication. After a period of 5-7 days we generally declare the infant safe to be off caffeine and then move on to other issues preventing them from going home to their families. This strategy generally works well for those infants who are born at later gestations but as Rhein LM et al demonstrated in their paper Effects of caffeine on intermittent hypoxia in infants born prematurely: a randomized clinical trial., after caffeine is stopped, the number of intermittent hypoxic (IH) events are not trivial between 35-39 weeks. Caffeine it would seem may still offer some benefit to those infants who seem otherwise ready to discontinue the medication. What the authors noted in this randomized controlled trial was that the difference caffeine made when continued past 34 weeks was limited to reducing these IH events only from 35-36 weeks but the effect didn’t last past that. Why might that have been? Well it could be that the babies after 36 weeks don’t have enough events to really show a difference or it could be that the dose of caffeine isn’t enough by that point. The latter may well be the case as the metabolism of caffeine ramps up during later gestations and changes from a half life greater than a day in the smallest infants to many hours closer to term. Maybe the caffeine just clears faster?
Follow-up Study attempts to answer that very question.
Recognizing the possibility that levels of caffeine were falling too low after 36 weeks the authors of the previous study begun anew to ask the same question but this time looking at caffeine levels in saliva to ensure that sufficient levels were obtained to demonstrate a difference in the outcome of frequency of IH. In this study, they compared the original cohort of patients who did not receive caffeine after planned discontinuation (N=53) to 27 infants who were randomized to one of two caffeine treatments once the decision to stop caffeine was made. Until 36 weeks PMA each patient was given a standard 10 mg/kg of caffeine case and then randomized to two different strategies. The two dosing strategies were 14 mg/kg of caffeine citrate (equals 7 mg/kg of caffeine base) vs 20 mg/kg (10 mg/kg caffeine base) which both started once the patient reached 36 weeks in anticipation of increased clearance. Salivary caffeine levels were measured just prior to stopping the usual dose of caffeine and then one week after starting 10 mg/kg dosing and then at 37 and 38 weeks respectively on the higher dosing. Adequate serum levels are understood to be > 20 mcg/ml and salivary and plasma concentrations have been shown to have a high level of agreement previously so salivary measurement seems like a good approach. Given that it was a small study it is work noting that the average age of the group that did not receive caffeine was 29.1 weeks compared to the caffeine groups at 27.9 weeks. This becomes important in the context of the results in that earlier gestational age patients would be expected to have more apnea which is not what was observed suggesting a beneficial effect of caffeine even at this later gestational age. Each patient was to be monitored with an oximeter until 40 weeks as per unit guidelines.
So does caffeine make a difference once term gestation is reached?
A total of 32 infants were enrolled with 12 infants receiving the 14 mg/kg and 14 the 20 mg/kg dosing. All infants irrespective of assigned group had caffeine concentrations above 20 mcg/mL ensuring that a therapeutic dose had been received. The intent had been to look at babies out to 40 weeks with pulse oximetry even when discharged but owing to drop off in compliance with monitoring for a minimum of 10 hours per PMA week the analysis was restricted to infants at 37 and 38 weeks which still meant extension past 36 weeks as had been looked at already in the previous study. The design of this study then compared infants receiving known therapeutic dosing at this GA range with a previous cohort from the last study that did not receive caffeine after clinicians had determined it was no longer needed.
The outcomes here were measured in seconds per 24 hours of intermittent hypoxia (An IH event was defined as a decrease in SaO2 by ⩾ 10% from baseline and lasting for ⩾5 s). For graphical purposes the authors chose to display the number of seconds oxygen saturation fell below 90% per day and grouped the two caffeine patients together given that the salivary levels in both were therapeutic. As shown a significant difference in events was seen at all gestational ages.
Putting it into context
The scale used I find interesting and I can’t help but wonder if it was done intentionally to provide impact. The outcome here is measured in seconds and when you are speaking about a mean of 1200 vs 600 seconds it sounds very dramatic but changing that into minutes you are talking about 20 vs 10 minutes a day. Even allowing for the interquartile ranges it really is not more than 50 minutes of saturation less than 90% at 36 weeks. The difference of course as you increase in gestation becomes less as well. When looking at the amount of time spent under 80% for the groups at the three different gestational ages there is still a difference but the amount of time at 36, 27 and 38 weeks was 229, 118 and 84 seconds respectively without caffeine (about 4, 2 and 1 minute per day respectively) vs 83, 41, and 22 seconds in the caffeine groups. I can’t help but think this is a case of statistical significance with questionable clinical significance. The authors don’t indicate that any patients were readmitted with “blue spells” who were being monitored at home which then leaves the sole question in my mind being “Do these brief periods of hypoxemia matter?” In the absence of a long-term follow-up study I would have to say I don’t know but while I have always been a fan of caffeine I am just not sure.
Should we be in a rush to stop caffeine? Well, given that the long term results of the CAP study suggest the drug is safe in the preterm population I would suggest there is no reason to be concerned about continuing caffeine a little longer. If the goal is getting patients home and discharging on caffeine is something you are comfortable with then continuing past 35 weeks is something that may have clinical impact. At the very least I remain comfortable in my own practice of not being in a rush to stop this medication and on occasion sending a patient home with it as well.
As someone with an interest in neonatal abstinence (NAS) I am surprised that I missed this study back in May. Anyone who says they aren’t interested in NAS research must be turning a blind eye to the North American epidemic of patients filling neonatal units or postpartum wards in need of treatment for the same. News feeds such as CNN have covered this story many times with concerning articles such as this published “Opioid Crisis Fast Facts” even the Trump White House has officially declared it as an emergency at this point. With NICU resources stretched and care providers fatigue levels wearing thin (these patients are typically very challenging to take care of due to the crying and agitation with neurological excitability that is at the core of the symptoms, something needs to be done. The vast majority of neonatal care providers treat such patients with an approach that promotes first non pharmacologic strategies such as keeping mom and baby together when possible, breast feeding and disturbing these infants as little as possible to name a few points. For those patients though who require pharmacologic support though, the mainstay has been oral morphine. At least in our units though once a patient is admitted and undergoes treatment we are still looking at anywhere from 3-4 weeks on average that they will occupy a hospital bed. If only there was a better way.
Could Buprenorphine do the trick?
While morphine is widely used to treat NAS symptoms unresponsive to other non pharmacologic methods of control, buprenorphine has a similar profile as an opioid but has less risk of respiratory depression as a partial agonist. A small but important trial has been published directly comparing the use of morphine to buprenorphine for treatment of NAS symptoms with the primary outcome being days of treatment and the second important point being length of stay. The trial, Buprenorphine for the Treatment of the Neonatal Abstinence Syndrome.by Kraft WK et al was entitled the BBORN trial for short. This was a single centre trial in which a double blind/double dummy approach was used. By double dummy this meant that after randomization those babies randomized to morphine received morphine plus a buprenorphine placebo and the other arm received a buprenorphine dose and a morphine placebo. In total 33 infants were randomized to buprenorphine and 30 to morphine (hence my comment about this being a small study). Their power calculation had called for 40 infants per arm to detect a 28% difference in the primary end point of duration of treatment but in the end that didn’t matter so much as they found a significant difference exceeding their estimate anyway. A lack of power would have become important mind you had they not found a difference as they wouldn’t have actually had the numbers to do so.
A strength of the study up front was that all care providers scored NAS symptoms the same way (need to take into account there is some subjectivity in scoring altogether though) and escalations and decreases of medication were done following a strict protocol both ways. In both arms, once a maximal dose of 60 mcg/kg of body weight for buprenorphine and 1.2 mg/kg for morphine was reached phenobarbital was added. When comparing the two groups at the outset there were no significant differences in characteristics so two generally similar populations of infants were being treated.
The Results Were Indeed Impressive
Before launching into the table, there were 21 babies in both groups that were bottle fed and 12 in the burprenorphine group and 9 in the morphine group that breastfed.
|Median days of treatment
|Hospital stay in days
No difference was seen in those who needed phenobarbital. Looking at the table, a couple things really stand out to me. They were looking for a 28% reduction in days of treatment. The results came in far excess of that at a 46% reduction. Curiously, breastfeeding which has classically been associated with a reduction in scores and therefore faster weaning due to less symptoms seemed to have the opposite effect here. Does this imply that breastfeeding slows down both duration of treatment and length of stay as a result? With a study this small it is difficult to say with so few breastfed babies but if I had to guess I would suggest those mothers that worked at breastfeeding may have had longer stays.
Should we all jump on the buprenorphine train?
For now I would give this a big maybe. One of the concerns about burprenorphine is that it comes as a solution of 30% alcohol. Giving multiple doses (3 per day in this study) of such a solution could in part contribute to these results of lower NAS symptoms. Is giving alcohol to reduce symptoms a good idea here? Not sure if there are any long term effects and moreover if the cumulative dose of this medication would be of a concern. Definitely something to check with your local pharmacist before rolling this out. On the other hand if the dose of alcohol provided was truly significant I might have expected the burprenorphine group to be poorer feeders due to intoxication which we certainly did not see.
With increasing volumes of newborns afflicted with symptoms of NAS we do need to find a way to stem the tide. Ideally, primary preventative strategies would be best but until that solution is found could burprenorphine be the next step in tackling this epidemic?
We can always learn and we can always do better. At least that is something that I believe in. In our approach to resuscitating newborns one simple rule is clear. Fluid must be replaced by air after birth and the way to oxygenate and remove CO2 is to establish a functional residual capacity. The functional residual capacity is the volume of air left in the lung after a tidal volume of air is expelled in a spontaneously breathing infant and is shown in the figure. Traditionally, to establish this volume in a newborn who is apneic, you begin PPV or in the spontaneously breathing baby with respiratory distress provide CPAP to help inflate the lungs and establish FRC.
Is there another way?
Something that has been discussed now for some time and was commented on in the most recent version of NRP was the concept of using sustained inflation (SI) to achieve FRC. I have written about this topic previously and came to a conclusion that it wasn’t quite ready for prime time yet in the piece Is It Time To Use Sustained Lung Inflation In NRP?
The conclusion as well in the NRP textbook was the following:
“There are insufficient data regarding short and long-term safety and the most appropriate duration and pressure of inflation to support routine application of sustained inflation of greater than 5 seconds’ duration to the transitioning newborn (Class IIb, LOE B-R). Further studies using carefully designed protocols are needed”
So what now could be causing me to revisit this concept? I will be frank and admit that whenever I see research out of my old unit in Edmonton I feel compelled to read it and this time was no different. The Edmonton group continues to do wonderful work in the area of resuscitation and expand the body of literature in such areas as sustained inflation.
Can you predict how much of a sustained inflation is needed?
This is the crux of a recent study using end tidal CO2 measurement to determine whether the lung has indeed established an FRC or not. Dr. Schmolzer’s group in their paper (Using exhaled CO2 to guide initial respiratory support at birth: a randomised controlled trial) used end tidal CO2 levels above 20 mmHg to indicate that FRC had been established. If you have less CO2 being released the concept would be that the lung is actually not open. There are some important numbers in this study that need to be acknowledged. The first is the population that they looked at which were infants under 32 6/7 weeks and the second is the incidence of BPD (need for O2 or respiratory support at 36 weeks) which in their unit was 49%. This is a BIG number as in comparison for infants under 1500g our own local incidence is about 11%. If you were to add larger infants closer to 33 weeks our number would be lower due to dilution. With such a large number though in Edmonton it allowed them to shoot for a 40% reduction in BPD (50% down to 30%). To accomplish this they needed 93 infants in each group to show a difference this big.
So what did they do?
For this study they divided the groups in two when the infant wouldn’t breathe in the delivery room. The SI group received a PIP of 24 using a T-piece resuscitator for an initial 20 seconds. If the pCO2 as measured by the ETCO2 remained less than 20 they received an additional 10 seconds of SI. In the PPV group after 30 seconds of PPV the infants received an increase of PIP if pCO2 remained below 20 or a decrease in PIP if above 20. In both arms after this phase of the study NRP was then followed as per usual guidelines.
The results though just didn’t come through for the primary outcome although ventilation did show a difference.
|Duration of mechanical ventilation (hrs)
The reduction in hours of ventilation was impressive although no difference in BPD was seen. The problem though with all of this is what happened after recruitment into the study. Although they started with many more patients than they needed, by the end they had only 76 in the SI group and 86 in the PPV group. Why is this a problem? If you have less patients than you needed based on the power calculation then you actually didn’t have enough patients enrolled to show a difference. The additional compounding fact here is that of the Hawthorne Effect. Simply put, patients who are in a study tend to do better by being in a study. The observed rate of BPD was 33% during the study. If the observed rate is lower than expected when the power calculation was done it means that the number needed to show a difference was even larger than the amount they originally thought was needed. In the end they just didn’t have the numbers to show a difference so there isn’t much to conclude.
What I do like though
I have a feeling or a hunch that with a larger sample size there could be something here. Using end tidal pCO2 to determine if the lung is open is in and of itself I believe a strategy to consider whether giving PPV or one day SI. We already use colorimetric devices to determine ETT placement but using a quantitative measure to ascertain the extent of open lung seems promising to me. I for one look forward to the continued work of the Neonatal Resuscitation–Stabilization–Triage team (RST team) and congratulate them on the great work that they continue doing.
A grenade was thrown this week with the publication of the Australian experience comparing three epochs of 1991-92, 1997 and 2005 in terms of long term respiratory outcomes. The paper was published in the prestigious New England Journal of Medicine; Ventilation in Extremely Preterm Infants and Respiratory Function at 8 Years. This journal alone gives “street cred” to any publication and it didn’t take long for other news agencies to notice such as Med Page Today. The claim of the paper is that the modern cohort has fared worse in the long run. This has got to be alarming for anyone reading this! As the authors point out, over the years that are being compared rates of antenatal steroid use increased, surfactant was introduced and its use became more widespread and a trend to using non-invasive ventilation began. All of these things have been associated with better short term outcomes. Another trend was declining use of post-natal steroids after 2001 when alarms were raised about the potential harm of administering such treatments.
Where then does this leave us?
I suppose the first thing to do is to look at the study and see if they were on the mark. To evaluate lung function the study looked at markers of obstructive lung disease at 8 years of age in survivors from these time periods. All babies recruited were born between 22-27 completed weeks so were clearly at risk of long term injury. Measurements included FEV1, FVC, FVC:FEV1 and FEF 25-75%. Of the babies measured the only two significant findings were in the FEV1 and ratio of FEV1:FVC. The former showed a drop off comparing 1997 to 2005 while the latter was worse in 2005 than both epochs.
This should indeed cause alarm. Babies born in a later period when we thought that we were doing the right things fared worse. The authors wonder if perhaps a strategy of using more CPAP may be a possible issue. Could the avoidance of intubation and dependence on CPAP for longer periods actually contribute to injury in some way? An alternative explanation might be that the use of continuous oximetry is to blame. Might the use of nasal cannulae with temporary rises in O2 expose the infant to oxygen toxicity?
There may be a problem here though
Despite everyone’s best efforts survival and/or BPD as an outcome has not changed much over the years. That might be due to a shift from more children dying to more children living with BPD. Certainly in our own centre we have seen changes in BPD at 36 weeks over time and I suspect other centres have as well. With concerted efforts many centres report better survival of the smallest infants and with that they may survive with BPD. The other significant factor here is after the extreme fear of the early 2000s, use of postnatal steroids fell off substantially. This study was no different in that comparing the epochs, postnatal glucocorticoid use fell from 40 and 46% to 23%. One can’t ignore the possibility that the sickest of the infants in the 2005 cohort would have spent much more time on the ventilator that their earlier counterparts and this could have an impact on the long term lung function.
Another question that I don’t think was answered in the paper is the distribution of babies at each gestational age. Although all babies were born between 22-27 weeks gestational age, do we know if there was a skewing of babies who survived to more of the earlier gestations as more survived? We know that in the survivors the GA was not different so that is reassuring but did the sickest possible die more frequently leaving healthier kids in the early cohorts?
This bigger issue interestingly is not mentioned in the paper. Looking at the original cohorts there were 438 in the first two year cohort of which 203 died yielding a survival of 54% while in 1997 survival increased to 70% and in 2005 it was 65%. I can’t help but wonder if the drop in survival may have reflected a few more babies at less than 24 weeks being born and in addition the holding of post natal steroids leading to a few more deaths. Either way, there are enough questions about the cohorts not really being the same that I think we have to take the conclusions of this paper with a grain of salt.
It is a sensational suggestion and one that I think may garner some press indeed. I for one believe strongly though as I see our rates of BPD falling with the strategies we are using that when my patients return at 8 years for a visit they will be better off due to the strategies we are using in the current era. Having said that we do have so much more to learn and I look forward to better outcomes with time!