Part 1: The parameters of ethical research are under pressure from a coterie of stakeholders who blur the distinction between medical treatment and research 

“In real ethics, there are some things that must never be done. Bioethics is…enticed onward by the question, Why not? If it can be done it should be done, or in any event it will be done, and, if it will be done, why not by us rather than by the competition?” We might begin by asking the experts who advocate the crossing of the next moral line, what’s in it for you?”  Richard John Neuhaus[1]

The greatest threat to legitimate medical research and the integrity of medicine is the inordinate pressure exerted by the confluence of powerful interdependent biomedical stakeholders–including academic, commercial and government stakeholders who threaten to weaken, if not eliminate, medical ethics standards embodied in the Hippocratic Oath: “First, do no harm;” the Nuremberg Code: “The voluntary, informed consent of the human subject is essential”; “No experiment should be conducted where there is an a priori reason to believe that death or disabling injury will occur”;  The Declaration of Helsinki: “the well-being of the individual research subject must take precedence over all other interests;” Mandatory Federal requirements of informed consent:  “…an explanation of the purposes of the research…A description of any reasonably foreseeable risks…disclosure of alternative courses of treatment;” [45 CFR 46.116] and restrictions: “Risks to subjects are minimized.” [45 CFR 46.111]

In recent years, a seemingly endless stream of medical research scandals (mostly uncovered through litigation) have shed light on how conflicts of interest have led to widespread research misconduct involving academic physicians at major medical institutions.[2]  Although government-sponsored research is subject to greater regulatory oversight and enforcement by the Office for Human Research Protections (OHRP), this agency’s enforcement has almost exclusively been limited to a “letter of determination” following an investigation, and no penalties. The abuses often exceed those in industry trials; but they are rarely if ever subjected to independent judicial review or legal sanctions.

By shedding light on hidden agendas, conflicts of interest, [3] violations of informed consent,[4] unethical trials that exposed patients to serious risks,[5] and promotional reports that have misled practicing physicians, the scandals have cast the medical research establishment in a bad light. This has spurred a campaign aimed at limiting regulatory oversight; loosening safeguards such as disclosure requirements; loosening regulatory proscriptions by redefining “risk” and what constitutes “research”.  Powerful academic medical research stakeholders and a corps of bioethicists argue that the regulatory process impedes research, adds costs, and causes patients to die by delaying development of life-saving therapies. [6] And they argue that the distinction between medical practice and medical research are unclear; that comparative safety trials of existing treatments—which they deem to be “riskless studies”[7]—should not require full review or informed consent.

A heated debate has been ignited about the parameters of ethical research and the enforcement of informed consent requirements in federally-funded research,[8] [9] [10] following public disclosure of the findings from a two year Federal investigation^[11] by OHRP which was precipitated by a complaint about the conduct of an oxygen supplementation experiment. The experiment, called SUPPORT, was conducted between 2005 and 2009, at 23 major medical centers in the U.S; 1,316 tiny premature babies, who were born between 24 and 27 weeks of gestation, had been enrolled. SUPPORT was sponsored by the Neonatal Research Network (NRN) within the National Institute of Child and Human Development, and approved by the National Institutes of Health (NIH).  The OHRP investigation determined (March 2013) that the SUPPORT Informed Consent documents violated essential ethical and legal Federal research disclosure requirements by failing to disclose to the parents the purpose of the experiment and any of the primary serious risks involved. In particular, parents were not informed about the risk of death, depending on which group their baby was randomly assigned.

Soon after publication of the SUPPORT trial report in The New England Journal of Medicine,[12] (2010) SUPPORT researchers have been publishing articles in the journal Pediatrics[13] [14] in which they stated: “obtaining informed consent requires time and effort; the number of patients enrolled seemed to be much smaller than the number screened…” They acknowledged SUPPORT had encountered resistance from mothers whose consent to enroll their vulnerable babies in the study was difficult to obtain:

“Of the 2228 mothers who were approached, 1219 (54.7%) gave consent… The frequency of approaches [to obtain informed consent] ranged from 1 time to 11 times… It took between 1735 and 2790 hours to obtain consent decisions from the 2228 mothers who were approached, to enroll 611 infants. Our results revealed that 5 families needed to be identified and screened for every 1 infant enrolled successfully.“ ¹³

They blame the enrollment bias of their study on informed consent requirements—rather than taking responsibility for their failure to design an ethically and scientifically acceptable research protocol:

“use of prenatal consent may not represent adequately the populations that trials are intended to study…the consent process is very inefficient, costly and biased the trial enrollment, such that the mothers and their infants differed in important ways from the available eligible population of very preterm infants.”¹³

Their cost / benefit stipulations led Dr. Simon Whitney of Baylor College of Medicine, Houston, Texas, to write a combative editorial published in the official journal of the American Academy of Pediatrics, in which he argues that regulatory oversight by IRBs (Institutional Review Boards) and OHRP “cost lives as surely as the “Python’s Embrace.”^[15] He asserts that “obtaining consent drained $200,000 from the [SUPPORT] project, and if 75% of the eligible but not enrolled infants had been included, recruitment would have been completed in 18 instead of 48 months.” 

He blames informed consent requirements for impeding the SUPPORT researchers from recruiting “black infants,” which, he says, limit the applicability of the study results to white infants.

The Neonatal Network researchers seek a waiver of informed consent to facilitate the enrollment of premature babies in trials such as SUPPORT, without parental knowledge; indeed, against their will! 

They seek to return to “the good old days,” when “many pediatricians [ ] bnelieved that it was not necessary to obtain permission of parents bnefore using a pediatric patient as a subject in research even if the research was nonotherapeutic…performing nontherapeutic experiments on children without authorization from parents was part of a broader ‘ethos of the time’ in which ‘everyone was a drafter’ in a national war on disease.” [16]

The SUPPORT researchers object to the requirement that informed consent must be obtained BEFORE enrollment in a trial. They insist that “a waiver of consent” should be granted to allow researchers to enroll premature babies at birth without prior parental consent because of the difficulty of obtaining consent; they cite the large number of mothers who refused to provide authorization—i.e., informed consent. They argue (in all seriousness) that parental consent can be obtained AFTER the babies have been enrolled; in other words, AFTER the babies have been subjected to the constraints and risks of the protocol without parental authorization.  (This is a case of medicine adopting the ethics of lawlessness from America’s Old Wild West: “shoot first, ask questions later…” or “shoot first, fill in the blanks later”)

Their naked cynicism is demonstrated by their suggestion that informed consent requirements—protecting our human right to refuse to authorize the enrollment of our babies in research and its risks—be replaced by “postnatal written consent to use the infant’s information.” “This stipulation,” they say, “allows parents to decide whether they want their infant’s information included in the study.”

This is a frontal assault on our inalienable human right to Informed Consent; it is being spearheaded by academics in the Neonatal Research Network who are subsidized by taxpayers. Their extremist demand represents the ultimate downward spiral into the abyss.

SUPPORT: Purpose / Protocol Design /Risks/ Primary End Points / Ethics /Results

Premature babies’ lungs are not fully developed and many function inadequately to meet the needs for oxygen in various organs in their body. Oxygen supplementation is often essential for their survival—and various efforts to provide it for premature babies date back to the 1700s.[17] Without adequate oxygen these babies die. But too much oxygen is believed to cause harm to the eyes and increase the risk of retinopathy of prematurity (ROP), which is sometimes surgically curable, but not always, and may result in blindness. How much supplementary oxygen a premature baby needs during this critical period, when their organs are maturing outside the protection of the mother’s womb, depends upon the prematurity of the baby and other physiological co-morbidity factors. In standard practice in intensive care neonatal units, treating physicians assess each baby’s physical condition and carefully titrate the oxygen levels they breathe. They select oxygen saturation targets (to maintain oxygen levels) within an unrestricted, full range—85%–95%–to maintain each baby at a level of saturation to minimize the risk of too little oxygen which causes brain damage and death, or too much oxygen which is thought to cause retinopathy.

The primary purpose of the experiment—as stated in the SUPPORT protocol[18] and in the published article in The New England Journal of Medicine (2010)12— was to compare the rates of death vs. ROP in premature babies whose oxygen intake was determined by which of two protocol-restricted, oxygen saturation target ranges they had been randomly assigned. Babies assigned to low oxygen saturation target ranges (85-89%) received less oxygen, and babies assigned or high target ranges (91-95%) received more oxygen. The risk of death from restricted oxygen intake was well known to the SUPPORT researchers who reported: “the combined risk difference in the trials from the 1950s was an absolute increase of 4.9% in the oxygen-restricted group…the practice of restricting the fraction of inspired oxygen (FIO2) to no more than 0.50 [in the 1960s] was estimated to result in an excess of 16 deaths per case of blindness prevented.”[19] “Despite the risk in mortality when restrictive oxygen supplementationwas used in the 1950s and 1960s and the limited data from observational studies, [20] it is becoming common practice to lower oxygen supplementation in an effort to reduce the risk of retinopathy.”  

Ethical standards dictate that an experiment that may increase the risk of death for babies should not be done.
But if the uncertain clinical information is deemed to be essential for improving babies’ survival rates, the experiment must be so designed as to ensure that babies’ lives are not sacrificed by the dictates of a protocol. Experiments that are designed to compare two fixed-dose regimens in randomized patient groups are expedient, and may provide a statistically significant difference between two regimens. However, such trials may seriously undermine patients’ safety—in particular patients whose medical condition is critical.  Newborn premature babies are in critical condition; they are born with unstable, varying clinical conditions requiring constant monitoring and individually adjusted titrated treatments for their survival. If they are randomly assigned to a treatment, there is an added risk of being consigned to a treatment contrary to their individual need; this is called “practice misalignment”—which is a risk of randomization.[21]  The experimental design used in the SUPPORT trial restricted the usual target range of oxygen saturation targets to either low (85-89%) or high (91-95%)—thereby exposing the defenseless premature babies to increased risks of serious harm. The primary endpoints of the SUPPORT experiment were Death and retinopathy (ROP); but death was concealed from the parents. .

Yet, the experimental design had been approved by NIH (the government funding agency) and by the institutional ethics review boards (IRBs) at the 23 participating medical centers, and it is being defended by researchers, the editor of the NEJM, the NIH leadership, and “bioethicists.”

Risks of harm for babies increased with an untested experimental deviation from clinical practice:
In intensive care units where premature babies are cared for, five vital signs are carefully monitored by physicians: pulse rate, blood pressure, breathing rate, temperature, and oxygen saturation levels. Pulse oximeters monitor each baby’s arterial oxygen saturation levels, providing vital information that guide treating physicians about how much supplemental oxygen each baby needs.  But the oximeters used in the SUPPORT experiment were deliberately set to provide inaccurate oxygen saturation levels—3% too high for babies randomized to the low oxygen group, or 3% too low for babies randomized to the high oxygen group.[22]  What responsible physician in clinical practice would put patients’ lives at risk by using an incorrect pulse oximeter to make critical clinical decisions? Imagine an intensive care unit whose thermometers had been set to mislead physicians, showing a temperature reading of 100 when the actual temperature was 103.

The designers of the SUPPORT protocol lost sight of their primary obligation as doctors—they failed to consider the foreseeable risks for the babies. 

Partial consequences of misleading the treating physicians about the actual oxygen saturation level:
babies in both groups were at increased risk of being severely harmed by this deviation from usual care;
2) many of the very fragile babies in the SUPPORT experiment received too little or too much oxygen, depending on which of the two experimental groups the baby was randomized;
3) physicians were kept in a state of ignorance about what the actual oxygen saturation level was in the critically ill babies in their care;
4) physicians did not know whether they were giving too much or too little oxygen to each baby;
5) physicians who were fooled into giving too little oxygen to help babies breathe, may have caused those babies preventable death or brain damage;
6) physicians who were fooled into giving too much oxygen to a baby who didn’t need it, may have caused that baby ROP requiring surgery;
7) significantly more babies were killed–23–in the low oxygen group, a difference of 3.7%;
8) lacking a control group of usual practice, the experiment lacks external validity;
9) the findings have no value for clinical practice—except to demonstrate what NOT to do.

The OHRP investigation findings: “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.¹¹

The stated rationale for incorporating the reckless experimental strategy—that misinformed physician about the babies’ oxygen saturation levels, was to maintain the study blind.¹⁷ Babies—whose lives were put at increased risk by this deception—were deemed less important than preventing potential bias, which should have been addressed differently.  The approval of this experiment reveals how far medical research has deviated from medicine’s foremost ethical principles; “Above all, do no harm,” and “the well-being of the individual research subject must take precedence over all other interests.” The IRB gatekeepers at major academic medical centers, whose responsibility is to review research design and informed consent documents—to ensure that they comply with ethical / legal requirements—have given a green light of approval for this grossly unethical government-funded experiment in which safeguarding the lives of the defenseless human subjects gave way to the supremacy of “Research u`ber Alles.”

The experiment clearly failed to meet medical, ethical, and scientific standards.
Legal requirements for valid informed consent, the foremost mandatory ethical and legal mandate for permissible research involving human subjects, require:disclosure of the purpose of the research which indlude:

• *description of current practice;
• in what respect does the research differ from current practice;
• *what the known and foreseeable risks are;
• how those risks will be minimized.

The SUPPORT consent documents[23] were grossly deceptive; [24]  they failed to disclose most of the important Federally mandated elements of informed consent: They failed to disclose that for babies randomly assigned to be maintained at restricted low oxygen saturation targets, the risk of death and brain damage from inadequate oxygen supplementation is increased significantly.The experiment failed to meet fundamental ethical limitations on permissible medical research: “No experiment should be conducted where there is an a priori reason to believe that death or disabling injury will occur.” [Nuremberg Code] The experiment shyould not have been approved.

The parents were deceived; the foreseeable increased risk of death for babies randomized to the restricted low oxygen group was concealed from them. The consent documents deceived them with false assurances that their babies would received “standard of care” with no significant risk: “Because all of the treatments proposed in this study are standard of care, there is no expected increase in risk for your infant.”[24] [25]

In standard care at the medical centers at which SUPPORT was conducted, oxygen intake levels and oxygen saturation targets (to maintain those levels) are selected from a broad range of values, allowing physicians to exercise clinical judgment. Oxygen levels are individually adjusted using an accurate pulse oximeter to help guide physicians’ decisions which are focused on the baby’s best interest. That standard of care is very different from the high or low oxygen saturation targets that babies were randomly assigned in the experiment. The protocol and published articles^{12 18} refer to the high oxygen target level as “more conventional”—an indication that the researchers considered the low oxygen target as less “conventional”—i.e., experimental, which by definition is acknowledged to pose risk until proven safe.  In their published article in the NEJM (2010)¹² the researchers disclosed that death vs. retinopathy were the primary outcome measures comparing “more conventional” high target levels to experimental low oxygen saturation target levels—and they acknowledged that “the safety of low target levels of oxygen saturation remains a concern,” citing numerous reports in the literature.

The mothers (mostly single) who gave their consent were approached at the time of admission to the hospital prior to the birth of their baby.  These mothers placed their trust in the doctors. But as the protocol and consent documents reveal, the mothers’ trust was misplaced. The consent forms were crafted to misinform them about the experimental nature and the risks involved. In essence, the researchers denied the mothers their fundamental human right to decide what’s best for their baby.

The invalid consent documents violated the essential ethical and legal requirements of valid informed consent; they were approved by NIH and the IRBs participating medical centers.

Why did the researchers withhold vital information from the mothers whose consent they sought?

Given that disclosure of such vital information is mandatory under Federal law, the researchers must have had a competing interest against disclosing the true facts to the mothers. A reasonable conclusion is that the researchers realized that if the mothers were informed about the purpose of the experiment; the difference between standard of care compared with being randomized to a restricted, protocol-dictated oxygen level; that some babies might possibly receive insufficient oxygen levels; and that the risk of insufficient oxygen is brain damage and death—no mother would have agreed to enroll her baby.

It is a testament to the protective maternal instinct that even the misleading, deceptive consent documents were unpersuasive for almost half of the mothers who were approached to authorize their baby’s enrollement in the experiment, refused to give their consent, even after numerous requests.¹³ The revelation that vital information was concealed from mothers, earned the SUPPORT research teams a stinging indictment by the editorial board of The New York Times which called “An Ethical Breakdown.”26]

    Vera Sharav

   End of Part I of 4

Read Part 2 Confluence of Self-Interest Groups: Government / Academic / Bioethics

https://ahrp.org/cms/content/view/927/9/

Part 3 Bioethicists Promote the Business of Medicine, Not the Ethics of Medicine: https://ahrp.org/cms/content/view/928/9/