The other part of the criticisms of Public Citizen, included in their letter to the secretary of the HSS in the USA, in which they call for the TOP (transfusion of preterms) trial to be stopped immediately, are regarding the consent forms.

The letter reviews the two previous trials as I have presented them in part one, and repeatedly states that there were:

‘less favorable outcomes seen in the IOWA and PINT studies for subjects in the restrictive transfusion group’

Public Citizen are willfully misrepresenting the results of those two trials: they did not show less favorable outcomes, as I have explained in part 1.

That being the case, what are the additional risks of being in the TOP trial, compared to being in an NICU and not participating in the trial?

Most babies under 1kg get transfused during their hospital stay. They may be transfused according to thresholds which are similar to the high TOP thresholds or the low TOP thresholds, or some other number, based on whatever the doctor is feeling like that day. So there are great variations around the world, and the procedures to be studied in TOP are well within the range of current clinical practice. There is therefore a very good justification for saying that the research poses no additional risks. It is dangerous to be an extremely preterm baby, transfusions on the other hand are really rather safe (safer than being ventilated, or having an arterial line). Studies in adults in ICU showed unexpectedly that lower transfusion thresholds were preferable, except maybe if you have unstable coronary artery disease. In the PICU transfusing at lower thresholds was not different from higher thresholds. The prior studies in the preterm baby have shown no difference, but are under-powered for the long term outcomes.

By the way, according to the Public Citizen standards those two previous trials were unethical; neither the TRICC trial, nor the TRIPICU trial had a ‘carry-on-doing-whatever-it-is-you-already-do,-and-transfuse-when-you-feel-like-it’ group (thankfully).

Public Citizen seem to think that the TOP IRBs were all misled or weren’t doing their job, and that they should have required the consent forms to list all possible complications associated with transfusions and/or prematurity as things that might possibly be different between the groups.

Their specific issues with the consent forms are as follows:

(1) As previously discussed, as part of routine care outside the research context, the hemoglobin level at which a particular premature infant would be transfused is routinely based on consideration of many individual patient factors, only some of which are taken into consideration in the experimental algorithms for the liberal and restrictive transfusion groups. Note also that an experimental  algorithm based in part on a poll of hemoglobin thresholds that would be acceptable to neonatologists in the context of a randomized clinical trial is not the same as what the hemoglobin thresholds for blood transfusion would be in usual care outside of a clinical trial. None of the consent forms clearly describe how the research interventions deviate from the usual individualization of transfusion care in extremely premature infants not enrolled in the study.

This is just evidence that they don’t know how we decide to transfuse patients.

It is also an egregious misunderstanding of clinical research. It is actually essential that the thresholds being tested are thresholds which would be used outside of a trial. If not the trial is a waste of time. If they were to compare transfusion at 20 to transfusing at 2, then that would be useless, as we won’t ever do that. If the thresholds being tested would be unacceptable in clinical care that really would be unethical, and you should just save your money and go home.

(2) Only two of the 17 IRB approved consent forms identified the restrictive group as being the usual approach for infants not enrolled in the research at that institution.  None of the other consent forms explained how the thresholds used for the two experimental groups compared to those used at the institution where the infant would be hospitalized if not enrolled in the research.

Not everyone would agree that the restrictive group is the usual approach, I would certainly say that both the restricted and the liberal approach are within usual care. Like many hospitals practice in my NICU is very variable, and you could find babies in my unit today who have received transfusions at the upper or lower limit, so to make such a statement would be misleading.

(3) Seven consent forms included the following misleading statement or one very similar to it:

This study does not alter the routine care of your baby.

The protocol changes nothing about routine care, other than obviously the transfusion threshold. That is what that statement means.

(4) Sixteen consent forms included the following uninformative and misleading statement that blurred the distinction between the two research interventions being tested and the individualized transfusion decisions that would occur for infants not enrolled in the research:

Both of these [hemoglobin threshold] levels [for determining when to transfuse blood] are in the usual range used by doctors in the NICU.

That is neither misleading nor uninformative, it is an accurate statement.

Public Citizen go on to state:

Finally, it seems unlikely that any parent who fully understands the results of the prior clinical trials, as well as the true risks, purpose, and nature of the experiment, would be willing to enroll their premature infant in this study.

Well I would. I am a parent of a premie, who understands these issues much better than Public Citizen, and I would certainly have been willing to enroll my baby.

They are also inconsistent, they note that the restricted transfusion thresholds are described, in the protocol, as being closer to the current practice of more neonatologists than the higher thresholds, then repeatedly claim that the lower thresholds are already known to be more dangerous (which is not true). Don’t they understand that would be really good evidence that we need to do this trial? And surely they can’t think that their own cursory perusal of the published evidence is so much more perceptive than the world experts on transfusion of the preterm who have got together to do this study?

Unfortunately Public Citizen get media attention. The NPR website for example has a news item ‘Another Study Of Preemies Blasted Over Ethical Concerns‘ which is not too unbalanced but does quote some of the dangerous misconceptions of Public Citizen. It seems that trying to destroy the confidence of parents in the ethical oversight of research in the preterm is one of their goals, if they succeed, they will seriously harm many thousands of babies, whose treatment will be improved by this kind of trial.

The last word again to Dr Lantos

What a topsy-turvy world! Today, neonatologists around the world want to carefully and collaboratively study the risks and benefits of common therapies. They publically announce their intentions and seek feedback from research review boards. Their study designs are rigorously scrutinised, and their consent forms are reviewed for accuracy and understandability. Parents are accurately informed of the reasonably foreseeable risks and benefits of the research. The studies are carefully monitored. Babies in the studies are protected from the risks of research and the risks of treatment with non-validated therapies. Their outcomes are better than those of babies not in the studies.

But then these neonatologists are criticised by advocacy groups and by the federal agencies whose mandate is to ensure the responsible conduct of research. These groups suggest that babies would be better off if doctors would misrepresent the risks of the studies, frighten parents away, prevent responsible research and continue to treat babies based on their non-validated beliefs about what is best.

We cannot let that happen.