Neonatal Research

As I mentioned at the end of my first post about this, the New York Times have jumped into the fray with an editorial entitled ‘An ethical breakdown‘ published today.

They take as ‘gospel’ the OHRP findings concerning the study, and show their lack of understanding of this kind of research, they note the following:

But, in this case, the federal Office for Human Research Protections concluded that the consent forms failed to reveal that there was a greater risk of dying in the low-oxygen group and a greater risk of severe eye damage in the high oxygen group. The form also stated that, because both groups would receive oxygen within the standard of care, there would be no predictable increase in risk no matter which group a baby was in.

But the federal agency found that many infants could have faced greater risks by participating. For example, if a baby whose clinical needs might ordinarily have led doctors to deliver a relatively high level of oxygen was enrolled in the study, the infant might be randomly assigned to receive lower levels of oxygen.

It is obviously much more understandable that a journalist does not comprehend the issues, but they lay out in those two segments the findings of the OHRP, which points out how misguided they are. The consent form could not reveal those different risks, because we didn’t know about them before the trial. The idea that neonatologists somehow know which baby has clinical needs that require a higher level of oxygen, and that the baby might be denied that because of the study would be concerning if it were true, but it is not. Firstly we did not have any reason for choosing one level of oxygen over another, as there had been no prior studies to help us choose. Secondly if there were such a situation the ethical imperative is to take the baby out of the study and treat them according to their clinical needs. There is no reason to believe that would not have been done to these babies; it is reported as a protocol violation, and counted at the end of the study.

Earlier on in the editorial they report the usual saturation limits used in most US NICUs as being 85 to 95%. They don’t seem to realize that the ranges tested in SUPPORT were all within those limits.

I am very concerned about this ruling, (as you can see from 3 posts). I think it is a stain on the reputation of the researchers, it calls into question the integrity of their IRBs, and puts at risk other clinical research in the NICU which might possibly be shown to have unexpected outcomes. Furthermore the other oxygen studies, including COT which will be presented at this year’s PAS meeting, will likely also be tainted by association.

To continue with my diatribe about the OHRP decision concerning the SUPPORT trial:

At one point in the document sent to the University of Alabama, the OHRP quote an editorial by Jay Greenspan:

‘…Although maintaining ranges of hemoglobin oxygen saturation in the vulnerable preterm population in the proximity of 85% to 90% is gaining increasing acceptance, marked variability in opinion exists.”  In short, the research and data analyses that had occurred prior to the SUPPORT study demonstrated that the use of higher versus lower levels of oxygen could significantly affect the likelihood of a premature infant developing ROP and other aspects of morbidity and mortality.’

Although they quote the editorial they clearly show by this statement that they have not understood it. All of the research and the analyses prior to support demonstrated that it was completely uncertain whether the use of higher versus lower levels of oxygen could significantly affect the likelihood of a premature infant developing ROP and other aspects of morbidity and mortality. They also refer to the Cochrane review of oxygen therapy in the preterm infant. Again they have not understood this sentence, the final one of the abstract

‘The results of this systematic review confirm that (the now historical) policy of unrestricted, unmonitored oxygen therapy has potential harms without clear benefits. However, the question of what is the optimal target range for maintaining blood oxygen levels in preterm/LBW infants was not answered by the data available for inclusion in this review.’

As I noted in the previous post, the prior studies were completely uninformative for modern neonatology.

Further on in this letter the OHRP notes that the study

‘specifically enrolled very premature infants and randomized them to one of two levels of oxygen. For many of those infants, the level of oxygen they received was different from what they would have received had they not participated in the study’

This seems to be the crux of the OHRP argument, because we had no idea if either of these ranges were preferable, and therefore some babies were haphazardly being exposed to higher or lower oxygen targets in their NICUs, it was somehow problematic to systematically randomize them. That is, doing this ethically, and collecting the data was somehow worse than allowing the on-going uncertainty. Because some babies who would have had high sat ranges with an increased risk of RoP, would instead have received less oxygen to achieve low saturations with an increased chance of dying. And vice versa.

The OHRP notes that the consent form includes the following: “It is possible that using lower pulse oximeter ranges will result in fewer babies with severe Retinopathy of Prematurity (ROP).” But then they castigate the investigators for not saying ‘that being in the upper range group may result in the greater risk of developing ROP.’ Now I don’t know about you, but if being in the lower range may decrease the risk, then it seems self-evident that the contrary is true!

I know that the early history of clinical research was sullied by several examples of unethical research practice, and that we need to protect human subjects. I know that introducing new medications and innovative practices do need to be carefully regulated to avoid the patients and parents that voluntarily participate in our trials being harmed.  But the standards for consent to participate in research comparing 2 differing standard approaches, either of which is within the limits of usual clinical practice, and which are in current use in NICUs with high quality care, have to be different.

If not how will we determine whether volume ventilation is really better, in terms of important clinical outcomes, than pressure ventilation; or whether permissive hypotension is better than treating everyone with a low blood pressure with dopamine? Sure it may turn out that pressure ventilation is associated with higher mortality, or with lower mortality. We don’t have any reason to believe that right now, but the only way to be sure is to do good research like SUPPORT.

The DHHS Office for Human Research Protections has just issued a ridiculous ruling. According to them the consent forms for SUPPORT were not sufficiently clear about risks of blindness in the higher oxygen group, and the risks of death in the lower oxygen group.

To give a little background the SUPPORT trial was a ground-breaking trial, the arm of the study which the OHRP are concerned about was a randomized comparison of 2 different target ranges of oxygen. The higher oxygen target was between 91 and 95% and the lower was between 85 and 89%. The study found unexpectedly that there was a higher mortality in the lower saturation group, and confirmed what was suspected, but not known, that having very low saturations gave less retinopathy than having modestly low saturations. Healthy babies at sea level have saturations over 95% so even the ‘high’ saturation level was already lower than ‘normal’. There was no good evidence how much lower than normal we should keep the babies saturations.

We must remember that the previous trials of oxygen therapy that showed the harm to the eyes of too much oxygen were in no way comparable to SUPPORT. There were 3 performed in the 50s, without any way to continuously monitor the babies oxygen concentration. As one example the study of Patz, 1954 compared the Experimental group (restricted oxygen) in which infants received oxygen only for clinical indications, and to a maximum FiO₂ of 0.4. The range of duration of oxygen in this group was 1 day – 2 weeks.  The comparison was the Control group (liberal oxygen): infants were placed in supplemental oxygen of 60-70% for 4-7 weeks, then weaned over one week. (Quoted from Lisa Askie’s Cochrane review).

So in one group the maximum FiO2 was 40% even if the baby was blue and bradycardic, in the other group the babies got over 60% oxygen whether they needed it or not! Despite these features, there was no difference in mortality.

The only other RCT data we had before SUPPORT were 2 small trials (from 1968 and 1972) that did not measure RoP, were mostly concerned with mortality and with resolution of lung disease. These two trials used intermittent blood oxygen monitoring, (not used in the 3 trials from the 50’s) and found no difference in mortality.

The other factor to remember is that prior to the results of SUPPORT there was a very wide variability of saturation targets in use in different NICUs across the country. Some were allowing saturations as low as 82-85%, others did not let the saturations go below 94%. So a baby in an NICU in the USA could be exposed to a saturation that was similar to the 2 ranges tested, or below or above the ranges tested in this trial.

So what did the NICHD network do? They did the most ethically appropriate thing, and launched a randomized trial. The trial protocol was developed and approved by the network centers, it was funded by the network, and was sent to ethics review committees at 18 institutions who approved the trial. Ethics review committees, IRBs usually have little to say about the scientific basis for a trial that has already been peer-reviewed and received funding. They mostly concentrate on the consent form to ensure that it is clear, understandable and describes any specific risks to participants.

The results of the trial as I mentioned were a little surprising, there was more mortality in the low oxygen group. The high oxygen group had more retinopathy, and more treatment for retinopathy with laser, but no difference in blindness.

So what did the OHRP say? First of all that the babies in the study were exposed to increased risk compared to babies treated according to ‘standard care’. They state the following

‘According to the study design, on average, infants assigned to the upper range received more oxygen than average infants receiving standard care, and infants assigned to the lower range received less.’

This is totally ridiculous, and clearly shows that the assessors were incompetent. You can’t average all the oxygen exposure of all the babies in the 2 groups, and then compare to some non-existent standard care. There was no standard of care, which the OHRP do not seem to be aware of. A child outside of the trial could well have had either of these target ranges, or ranges even more variable.

They also state the following

‘It would have been appropriate for the consent form to explain (i) that the study involves substantial risks, and that there is significant evidence from past research indicating that the level of oxygen provided to an infant can have an important effect on many outcomes, including whether the infant becomes blind, develops serious brain injury, and even possibly whether the infant dies; (ii) that by participating in this study, the level of oxygen an infant receives would in many instances be changed from what they would have otherwise received, though it is not possible to predict what that change will be; (iii) that some infants would receive more oxygen than they otherwise would have, in which case, if the researchers are correct in how they suppose oxygen affects eye development, those infants have a greater risk of going blind; and (iv) that the level of oxygen being provided to some infants, compared to the level they would have received had they not participated, could increase the risk of brain injury or death.’

This again shows the lack of insight of the OHRP into these issues. Yes, being a very preterm baby in an NICU carries significant risks. We now know and did not know before the study, that having a saturation just below 90% increases mortality, compared to just above 90%. We now know that having a saturation in the low 90’s increases RoP compared to the high 80’s but that doesn’t increase blindness.

The OHRP thinks we should have know that before doing the study and included it in the consent forms. If we had known it we would not have done the study. (I say we, as I was involved in the Canadian version of this trial).

The OHRP is using an impossible standard. It is of note that they say nothing about the other arm of this trial which compared 2 other interventions, early CPAP to immediate intubation for surfactant. As that trial did not have a significant difference in the outcomes, they say nothing about the consent for that arm, but the potential differences were at least as great as for the O2 comparison. The only reason they have focused on the O2 is that there was a difference in the outcomes. So presumably someone complained and the OHRP don’t like the idea that the network funded a trial which had more deaths in one group. They have to realize that that is going to be a risk in funding research in critical populations, and if they don’t like it we will just have to stop doing research.

This is already a longer post than my usual, so I will take a break, but there is even more stupidity to come in this ruling. Which is now being repeated, including in an editorial in the New York Times.