Beware of medical research institutions whose physicians violate medical ethics with impunity, exposing vulnerable patients to serious risks of harm without disclosing the nature of the experiment or the risks.
Newly obtained documents from the oxygen experiment (SUPPORT) reveal far more extensive medical ethics violations than those specified in the March 7, 2013 letter by the Office of Human Research Protection to the University of Alabama.
On May 8, 2013, Public Citizen submitted a follow-up letter to Secretary of Health and Human Services and detailed analysis (31 pages) of the SUPPORT research protocol and consent forms approved by each of the 23 IRBs at the participating medical centers at which the experiment—Surfactant, Positive Pressure, and Oxygenation Randomized Trial (SUPPORT) had been conducted on more than 1,300 extremely premature babies.
The review of 22 IRB-approved consent forms for the SUPPORT trial reveals that NONE complied with mandatory Federal disclosure requirements:
- · NONE of the 22 consent forms explained that death was a risk—especially for babies randomized to low oxygen levels;
- · Only two of 22 consent forms disclosed that babies randomized to high oxygen levels were at increased risk of retinopathy or blindness.
- · NONE of the 22 consent forms disclosed that the pulse oximeter readings—that neonatologists rely on to determine an infants’ need for supplementary oxygen—were intentionally altered to provide the treating medical teams with either false high or false low oxygen saturation values (SOP2).
The IRB-approved consent forms falsely assured parents that the study posed no risk: “Because all of the treatments proposed in this study are standard of care, there is no expected increase in risk for your infant.”
In fact, the primary purpose of the experiment—as stated in the protocol —was to assess the effects of two experimental strategies for managing oxygen and two strategies of ventilation therapy in extremely premature babies; the primary endpoints being: death vs. retinopathy (or blindness).
But lacking a comparator group of babies who received standard-of-care which is, individually titrated oxygen, the experiment could not inform physicians whether either of the experimental groups compared favorably or worse than standard-of-care in terms of deaths and rate of retinopathy.
The parents from who consent was obtained were mostly young single mothers [indeed, the authors noted that “guardians” gave consent in some cases.] They were educationally and economically disadvantaged, had little understanding about the complexities of standard care provided to extremely premature babies whose survival depends on carefully calibrated life-support interventions such as oxygen supplementation and intubation (an invasive procedure involving insertion of a tube in to the trachea, the main airway leading to the lungs).
Nor did they have any understanding that experimental treatments posed increased risks for the babies. The deceptive consent documents provided no clue about the true nature, purpose or risks involved. The young mothers trusted the doctors and—as the editor of the NEJM suggested in an editorial defending the dubious consent process for SUPPORT—the mothers were “reassured that giving birth at an academic medical center with a sophisticated nursery for premature newborns and with physicians who have extensive experience with very preterm infants” will provide the best care for their tiny, fragile premature infants.
But, as the protocol and consent documents demonstrate, the mothers (or parents) who consented to entrust the care of their baby to the doctors in this Federally-sponsored trial were badly deceived—they were not infomed about the increased risk of death for babies assigned to the low-oxygen group, and not told that the risk for eye disease (potential blindness) might be increased in babies assigned to the high oxygen grooup–their consents are, therefore, invalid.
In standard care, each baby’s individual needs is assessed by a skilled neonatologist whose professional judgment determines if and when an intervention is necessary. In randomized medical research, individual patients’ needs are subordinated to the requirements of the research protocol, which is one important reason why ethical standards and legal statutes require that “the voluntary informed consent of the human subject is absolutely essential.” [Nuremberg Code, Code of Federal Regulations]
Any research involving incapacitated human beings must meet very stringent standards to justify conducting research on involuntary, non-consensual persons.
Experimental designs requiring randomization to different levels of life-support interventions—independent of need—pose significant risks for patients –especially for premature babies:
In addition to the risk of randomization, the babies in the SUPPORT experiment incurred additional risk from the intentional false oximeter oxygen saturation (SOP2) readings that intentionally misled physicians. As a result, infants assigned to the intentionally inaccurate high (91%-95%) oxygen group were given lower than needed oxygen to breathe—thereby increasing risk of death. And babies assigned to the inaccurate low (85%-89%) oxygen group received more oxygen than they needed—thereby increasing the risk of retinopathy (eye disease). The experiment was designed and conducted as if the vulnerable non-consensual premature babies were lab rats.
Such experiments cannot guide care or adequately monitor safety; they should, therefore, not be done in humans, let alone premature infants. Indeed, the Code of Federal Regulations mandates that “Risks to subjects are minimized.” [45 CFR 46.111(a)(1)] It is the first listed criterion for IRB approval.
In clinical practice treating neonatologists rely on pulse oximeter oxygen saturation readings to continuously monitor extremely premature infants’ oxygen saturation levels (SOP2)—in order to determine each individual baby’s need for supplementary oxygen and other life-saving clinical interventions that may be needed. 
The most egregious deviation from medical practice that undoubtedly increased risk for the tiny babies was the use of pulse oximeters that had been deliberately altered to give either false high or false low oxygen readings. The intent was to mislead the medical treatment teams caring for the tiny babies by providing inaccurate pulse oximeter readings. [See p. 17 and 18, SUPPORT protocol.]
Thus, the tiny babies whose lives were on the line were in the care of physicians and nurses who had been provided inaccurate information about each baby’s oxygen needs. By design, the neonatologists were asked to provide life-support interventions not only blindfolded, but with one hand tied behind their back.
The protocol-required deception was first brought to public attention by AHRP on April 24, 2013. See, https://ahrp.org/cms/content/view/918/9/
This dubious deception (ostensibly to ensure the study blind) sacrificed the babies’ welfare, impacting vital clinical decisions. Some babies may have undergone protocol-driven, clinically unnecessary invasive interventions—thereby increasing risks, trauma and discomfort for the vulnerable babies.
For example, some babies in the high-oxygen group may have been unnecessarily intubated (an invasive procedure involving insertion of a tube in to the trachea, the main airway leading to the lungs) or may have been treated with an artificial breathing machine. Those who received too much oxygen were exposed to increased risk of eye damage and blindness. Babies in the low-oxygen group whose needed treatment was delayed—due to the false oximeter reading—may have been exposed to inadequate oxygen supply. Those babies were put at increased risk of neurological damage and death.
It is unconscionable for any professional medical care giver to agree to be a participant in such an experiment.
The SUPPORT experiment was not designed to support or improve the babies’ survival rate or to decrease the risk of retinopathy; indeed, multiple elements of the protocol sacrificed rather than supported the babies’ survival needs.Furthermore, the experiment was not designed to provide useful information to improve premature infants’ care.
How could the reviewers at NIH and the local institutional (ethics) review boards (IRBs) have failed to consider the increased, protocol-driven risks for the babies?
The willingness of academic physicians at 23 medical centers to participate in an experiment that required them to treat extremely premature babies who were struggling to survive—without the benefit of having accurate information to guide their clinical decisions, decisions that can—and did—lead to preventable deaths and preventable blindness—is shocking.
It would appear that the experiment was designed to obtain a sufficiently robust statistical difference between the two experimental arms in order to improve the likelihood of publication in The New England Journal of Medicine .
The physicians who designed the experiment, the IRBs that approved it, and the medical teams at 23 medical centers who conducted the experiment–that lacked a current practice control arm, that provided treating physicians with deliberately inaccurte pulse oximeter readings; and consent forms that deceived parents–were accomplices in an illegitimate medical experiment.
It is unlikely that animal care committees at the same medical centers would have approved the experimental design as conducted on the infants; the animal care committee would have insisted that the experimental design had to be scientifically valid to provide useful results and not waste animals.
These complicit medical professionals violated Federal statutory prohibitions against increasing risk to human subjects [45 CFR 46.111(a)(1)]; what’s more, they violated and bedrock international ethical standards of medical research: “In medical research involving human subjects, the well-being of the individual research subject must take precedence over all other interests.” [Declaration of Helsinki]
A fair question to ask government officials at the National Institutes of Health and the National Institute of Child and Human Development (NICHD) is: What is the Institute’s approval criteria for neonatal research? What was the justification for providing millions of dollars for the fatally flawed SUPPORT experiment? Dr. Rosemary Higgins, program scientist for the NIH Neonatal Research Network (NRN) has defended the trial in the media: What was Dr. Higgins’ responsibility in the approval of the dubious SUPPORT protocol and consent form?
What assurance do citizen-taxpayers have that the 4,500 neonates who are currently being sought as subjects for seven different NIH-funded experiments in neonates that are scheduled to be conducted at some of the same NRN medical centers will not similarly be abused in unethical experiments? See, https://ahrp.org/cms/content/view/916/9/
The conduct of the professionals involved in this unconscionable experiment pose a very serious threat to legitimate medical research; their conduct should be of concern to conscientious physicians who are committed to the “do no harm” Hippocratic principle of medicine. In truth, this experiment was a premeditated medical assault on fragile non-consensual infants who were struggling to live.
A chorus of like-minded medical academics and “bioethicists” in government and universities, led by Dr. Jeffrey Drazen, the editor of The New England Journal of Medicine, are committed to the utilitarian proposition that the interests of research supersede the interest of individual patient/subjects. 
They can always be counted on to rally to the defense of experimental research atrocities such as the oxygen SUPPORT experiment, an experiment that was designed, approved, and conducted by unconscionable medical teams at 23 medical centers without any regard for the safety of premature babies whose lives were put at increased risk.
If NIH-funded neonatal research at the Neonatal Research Network comes to a halt, blame those responsible for the SUPPORT experiment and their retinue of supporters such as: Dr. Jeffrey Drazen, Editor, NEJM; bioethicists, Arthur Caplan and David Magnus, Stamford University one of the SUPPORT experiment sites. They are apologists for the medical establishment’s wrongdoing and the enemy of honest research who would pervert the values of our society if they are not repudiated.
The SUPPORT experiment is an example of the worst case of institutional betrayal of trust after the adoption of statutory fedeeral protections. An apology is due to every parent whose infant was enrolled in the SUPPORT experiment. The National Institute of Child and Human Development and the medical center at which each infant was subjected to this illegitimate experiment should apologize.
Vera Sharav
Excerpts from Public Citizen’s detailed review of the SUPPORT protocol and revelations in the informed consent documents (obtained through Freedom of Information statute).
“The new information highlighted in Public Citizen’s report affirms the appropriateness of OHRP’s determination in its March 7, 2013, letter to UAB that the UAB IRB -approved consent form failed to mention the serious, reason ably foreseeable risks related to the part of the study comparing two experimental strategies for managing oxygen in extremely premature infants. Those risks, correctly identified by OHRP, included increased risks of brain injury; an eye disease called retinopathy of prematurity, which can lead to blindness in severe cases; and death, depending on the randomized group assignment of each baby…Our review of all IRB-approved consent forms for the study reveals that none explained that death was a risk of the oxygen experiment and only two disclose d that eye disease or blindness was a risk of exposure to high oxygen levels.
Moreover, the new information demonstrates that the deficiencies of the UAB IRB-approved consent form were far more significant than those discussed in OHRP’s March 7 letter. In particular, the IRB-approved consent forms in many, if not all, cases either did not disclose at all or did not accurately describe the following:
1. The IRB-approved consent forms stated: “There is no known risk to your baby from monitoring with the pulse oximeters used for this study.”
The consent forms failed to mention that the oximeters would be miscalibrated to provide the medical teams caring for the premature babies in the study with oxygen saturation readings that were either inaccurately low or inaccurately high.
2. The experimental procedure of using pulse oximeters —devices used to continuously monitor blood oxygen levels —that were intentionally miscalibrated to provide the medical teams caring for the premature babies in the study with oxygen saturation readings that were either inaccurately low or inaccurately high. Only 11 consent forms disclosed in some way the plan to use this procedure, but none explained that it deceived the treating medical team or such deception could have impacted important clinical decisions related to the babies’ care.
3. The substantial, reasonably foreseeable risks of harms from intentionally providing the medical teams caring for the babies in the study with inaccurate information regarding the babies’ oxygen saturation levels. This unprecedented deceptive procedure may have adversely impacted important clinical decisions regarding whether to intubate a baby and start mechanical ventilation (treatment with an artificial breathing machine) or whether to discontinue mechanical ventilation.
Because of this deceptive experimental procedure:
(a) Some babies in the high-oxygen group may have undergone protocol-driven intubations and been placed on mechanical ventilation when such procedures were not clinically indicated. This could have unnecessarily exposed some babies to increased risk of:
(i) trauma to the mouth and gums during intubation;
(ii) trauma to the trachea, resulting in bleeding and puncture of the airway during intubation; (iii) pneumothorax (collapsed lungs, possibly resulting in the need for insertion of chest tubes); (iv) pneumonia during mechanical ventilation; and (v) death (see example on page 11 of the enclosed report).
(b) Some babies in the low-oxygen group may have had actual clinical indications for intubation and mechanical ventilation, but because of inaccurate oxygen saturation levels, these treatments may have been inappropriately delayed. This could have unnecessarily exposed some babies in the low -oxygen group to increased risk of prolonged hypoxemia (oxygen deficiency) with inadequate oxygen delivery to the brain, resulting in neurological damage and possibly death.
4. None of the IRB-approved consent forms disclosed to parents that the investigators characterized the high-oxygen target levels in the protocol as being “more conventional”; by implication, the low -oxygen target levels were considered less conventional. Only two consent forms suggested an oxygen saturation range that was most commonly used in routine practice.
5. Neither the protocol nor the consent form explained how the experimental procedures for managing the babies’ oxygen therapy deviated from the usual “more conventional” standard of care the babies would have received had they not been enrolled in the study.
Public Citizen notes that the absence of these critical elements of information about the purpose, nature, and risks of the SUPPORT study’s complex oxygen experiment, combined with the inclusion of statements indicating that the experimental procedures had no risk, denied the parents of babies enrolled in the trial the opportunity to make an informed decision when they gave consent for the research. “The agency should have cited UAB and the other SUPPORT study institutions for additional serious deficiencies in the IRB-approved consent form regarding the lack of disclosure of critically important information about the purpose, nature, and risks of the oxygen experiment.”