Public Citizen: Oxygen Experiment On Premature Babies Is Highly Unethical
Health Letter article, June 2013
On April 10, 2013, Public Citizen’s Health Research Group (HRG) wrote a scathing letter to Secretary of Health and Human Services Kathleen Sebelius condemning a highly unethical government-funded clinical trial involving 1,316 very premature infants. These babies were exposed to an increased risk of blindness, brain damage and death without their parents being informed through the consent forms they signed of these risks or the true purpose and nature of the research. We urged Sebelius to personally apologize to the parents of the critically ill babies enrolled in this high-risk experiment held throughout the U.S. because it was funded by the National Institutes of Health (NIH), the premier research agency within the Department of Health and Human Services (HHS).
The study, known as the Surfactant, Positive Pressure, and Oxygenation Randomized Trial (SUPPORT), was conducted between 2005 and 2009 by approximately two dozen prominent medical research centers throughout the country, including those affiliated with Stanford University; Yale University; Brown University; Duke University; the University of California, San Diego; and the University of Alabama at Birmingham (UAB), which was the lead institution. The participating institutions are part of a multicenter group known as the Neonatal Research Network, which was established in 1986 by the Eunice Kennedy Shriver National Institute of Child Health and Human Development, part of NIH, to conduct research studies on preterm and term newborns.
Overview of oxygen treatment in premature babies
To understand why the conduct of the SUPPORT study was so unethical, one needs to first understand how oxygen therapy is normally administered in premature babies.
Because premature babies have immature lungs, they usually require treatment with supplemental oxygen to survive and to prevent brain damage and other problems caused by oxygen deficiency. In many cases, premature babies also need to undergo intubation (insertion of a breathing tube into the trachea, the main airway leading to the lungs) and treatment with a ventilator (an automated breathing machine). More than 50 years of medical research have demonstrated that for premature babies, treatment with too little oxygen can cause brain injury or death, whereas treatment with too much oxygen can lead to damage to the retina of the eye and blindness.
As part of routine care, the amount of oxygen given to each premature baby is individually adjusted based on a continuous assessment of many clinical factors by the baby’s medical team, which includes neonatologists, nurses and a variety of medical specialists. One of the most important factors used to guide oxygen therapy is the oxygen saturation level, a measure of blood oxygen content. Oxygen saturation levels have become so important in the care of critically ill patients that it is sometimes referred to as the “fifth vital sign” (the first four vital signs being pulse, blood pressure, breathing rate and temperature). Since the 1980s, technology has been widely available to monitor oxygen saturation levels continuously using a probe placed on a finger or elsewhere on the skin, which is then connected to a device called a pulse oximeter. All premature babies cared for in neonatal intensive care units (NICUs) in the U.S. have been monitored with these devices for decades.
For the medical centers participating in the SUPPORT study, routine care of premature infants not in the study typically involved giving enough oxygen to maintain oxygen saturation levels somewhere within the range of 85 to 95 percent. (Healthy children and adults normally have an oxygen saturation level in the 97 to 99 percent range.) However, within this broad range, the oxygen level target for each premature baby is individually determined at any particular time based on numerous considerations, including:
- diagnostic test results that indicate whether enough oxygen is being delivered to the baby’s body (for example, a high blood acid level would indicate insufficient oxygen delivery and the need to increase the oxygen concentration in the air the baby is breathing);
- clinical signs of inadequate oxygen delivery to a particular organ (for example, seizures may suggest that the baby’s brain is not receiving enough oxygen); and
the wishes of the baby’s parents regarding the balancing of specific risks (for example, balancing the risk of death with the risk of blindness) and what the parents think would be in the best interests of their baby.
The SUPPORT study oxygenation experiment
The babies in this experiment were extremely premature, having been born at 24 to 28 weeks gestational age and weighing on average less than two pounds. Such babies are critically ill, experience a wide range of complications and have a high mortality rate (although with gradual improvements in neonatology, the survival rate has gradually improved).
The researchers randomly divided these vulnerable subjects into two experimental groups. For one group, the researchers tried to maintain the infants’ blood oxygen levels in a low target range (oxygen saturation level of 85 to 89 percent), and for the other group in the more conventional, high target range (oxygen saturation level of 91 to 95 percent), rather than adjust each baby’s oxygen levels within the broader range of 85 to 95 percent to meet his or her individual needs as would have been the case had the baby not been in the study. The researchers then measured the impact of the two target ranges of oxygen levels for premature babies — specifically, whether infants in one group were more likely to die, suffer brain damage, or develop eye disease and blindness in comparison to the other group.
It is notable that the SUPPORT study’s protocol (a document that includes the rationale for doing the study and a detailed description of the predefined set of procedures for the experiment) stated that the higher oxygen saturation target range was “more conventional” than the lower oxygen saturation target range for the usual standard of care of premature babies. This implicitly means that neonatologists in the U.S. most commonly tried to maintain premature babies in the 91 to 95 percent range at the time the study was conducted.
Another part of the oxygen experiment was the use of specially altered pulse oximeters to monitor the oxygen saturation of the study babies. For babies in the high-oxygen group, the pulse oximeters were intentionally altered to read inaccurately low, whereas for babies in the low-oxygen group, they were intentionally altered to read inaccurately high whenever their actual oxygen saturation levels were greater than 85 percent or less than 95 percent. For example, when the actual oxygen saturation levels of babies in the low-oxygen group were 85 to 89 percent, the study pulse oximeters indicated the level to be 88 to 92 percent. When the actual oxygen saturation levels of babies in the high-oxygen group were 90 to 95 percent, the study pulse oximeters indicated the level to be 88 to 92 percent. Thus, when the altered pulse oximeters read 90 percent, the actual oxygen saturation level was 87 percent for the low-oxygen group and 93 percent for the high-oxygen group, a clinically significant, 6 percent difference.
The medical teams caring for babies enrolled in the study were only allowed to use these inaccurately reading pulse oximeters when caring for babies in the study. The combined experimental procedure of randomly assigning babies to low or high target oxygen levels without respect to their individual clinical needs, as well as providing intentionally inaccurate information to the entire medical team about blood oxygen levels — a vitally important parameter used to care for babies —represented a considerable deviation from usual standard of care. It is important to know actual oxygen saturation levels because that is a key parameter in deciding when premature babies should be intubated and placed on a ventilator and when they can be safely taken off a ventilator and be allowed to breathe on their own.
Providing the medical team with inaccurate information about oxygen levels could have adversely affected these critical clinical decisions. For example, the inaccurate oxygen level readings could have led the medical team to intubate and artificially ventilate some babies who did not need these medical procedures, thus unnecessarily exposing the babies to the risks of intubation and artificial ventilation. On the other hand, the inaccurate oxygen level readings could have led the medical team to not intubate and artificially ventilate other babies who did need these medical procedures, thus exposing them to risks of oxygen deficiency.
For many of the SUPPORT study babies, the overall level of oxygen they received was different from what they would have received had they not participated in the study. Many babies in the low-oxygen group predictably received less oxygen than they would have otherwise received, potentially increasing risk of brain injury and death. On the other hand, many babies in the high-oxygen group predictably received more oxygen than they would have otherwise received, potentially increasing the risk of eye disease and blindness.
Indeed, results of the study published in the New England Journal of Medicine revealed that babies in the high-oxygen group were twice as likely to develop the serious eye disease associated with prematurity as those in the low-oxygen group — 18 percent versus 9 percent. Not surprisingly, babies in the low-oxygen group had a higher death rate, with 20 percent of babies in that group dying before discharge compared to 16 percent in the high-oxygen group.
Inadequate consent forms
Given the nature of the experimental interventions in the SUPPORT study and their serious, potentially life-threatening risks, one may wonder how the parents of more than 1,300 premature infants were willing to consent to enrolling their extremely premature infants in this experiment. A review of the final consent forms that were approved by the institutional review board (IRB) — a committee charged with conducting an ethical review of human subjects research — at each study institution reveals that parents were not informed about the true purpose, nature or risks of the study, thereby providing a plausible explanation for why many parents gave their consent.
For example, regarding the purpose of the oxygen experiment, most consent forms simply stated that the study would try to determine whether use of the lower oxygen range would lower the rate of eye disease and blindness. However, all failed to mention that the other primary purpose was to see if babies randomly assigned to the low-oxygen group would have a higher (or lower) death rate in comparison to the high-oxygen group.
When describing the experimental procedures involved in the research, the consent forms in general included language like the following:
The babies in this study will also be placed randomly (again, like the flip of a coin) into a group monitored with lower oxygen saturation ranges or higher oxygen saturation ranges. Oxygen saturation is measured on a baby with a machine called a pulse oximeter .... The babies in the lower range group will have a target saturation of 85-89%, while the babies in the higher range group will have a target saturation of 91-95%. All of these saturations are considered normal ranges for premature infants. If the saturation falls below 85% or goes higher than 95% then the pulse oximeter will alarm so that the doctors and nurses know when to turn your baby’s oxygen up or down. [emphasis added]
The short statement that all of these saturations are considered normal ranges for premature infants was very misleading in several regards. First, it failed to communicate to the parents that the higher range was the more conventional range (as was stated in the protocol). Second, it failed to explain the true complexities of oxygen management in premature babies and how oxygen targets are normally individually adjusted based on many clinical factors. Third, and perhaps most important, the consent forms failed to inform parents that the pulse oximeters were altered to intentionally provide inaccurate information across almost all of the 85 to 95 percent range, depriving the entire medical team of accurate information vitally important to the care of premature babies.
Finally, with two exceptions, none of the consent forms disclosed the risks of the experiment comparing high and low oxygen target ranges. Two consent forms noted that babies in the high-oxygen group might have an increased risk of eye damage. Two even went so far as to state, “There is no known risk to your baby from monitoring with the pulse oximeters used for this study.” This extremely misleading statement disregarded the risks of brain injury, death and eye disease depending on the randomized group assignment of each baby, instead leading parents to essentially believe there were no risks. It also failed to mention the risks of giving intentionally inaccurate information about the babies’ oxygen saturation levels to the medical teams caring for the babies. These included risks associated with either inappropriate delays in intubation or unnecessary intubations. Ironically, the only pulse oximeter risk mentioned — possible skin breakdown — was not even a risk of the research, because these babies needed monitoring with pulse oximeters anyway.
The failure to disclose critically important information regarding the purpose, nature and risks of the research to parents of the SUPPORT study babies represented an egregious violation of research ethics. These failures undoubtedly directly affected parents’ decisions to enroll their premature babies in this study. It is highly likely that had they been appropriately informed, many, if not most, parents would have declined to enroll their babies.
A federal watchdog first identifies consent-form violations
In February and again in March, the Office for Human Research Protections (OHRP), a watchdog division in HHS that oversees human subjects research, sent little-publicized letters to the UAB stating that the study had violated “the regulatory requirements for informed consent, stemming from the failure to describe the reasonably foreseeable risks of blindness, neurological damage and death.”
The February OHRP letter to UAB was first brought to Public Citizen’s attention when a reporter from a small monthly trade-press publication contacted Dr. Michael Carome, then the deputy director of Public Citizen’s Health Research Group (HRG), seeking comment on the letter. (For 10 years prior to joining Public Citizen, Dr. Carome was on the OHRP staff, serving most recently as its associate director for regulatory affairs for eight years).
Actions taken by Public Citizen
After reviewing the OHRP letters and published medical journal articles about the SUPPORT study and related research, it became readily apparent to HRG that not only was OHRP right about the SUPPORT study consent forms failing to disclose serious risks regarding the oxygen experiment, but also that OHRP failed to identify other consent-form deficiencies related to the purpose and nature of the study.
On April 10, after consulting with experts in critical care medicine, pediatrics and ethics, Dr. Carome and Dr. Sidney Wolfe, then the director of HRG, sent a letter urging Sebelius, along with NIH Director Francis Collins, to personally apologize to the parents of the 1,316 babies enrolled in the SUPPORT study because they were never informed — and still may not know — that their babies faced serious danger from the SUPPORT study and that some may have died or suffered from serious eye disease unnecessarily. Public Citizen also called for a broad investigation into the study and other corrective actions and has collected more than 16,000 signatures on a petition with these demands. (The petition can be viewed and supported at http://bit.ly/13MZ06L).
Public Citizen’s letter prompted widespread media coverage of the story by outlets ranging from The New York Times, The Washington Post and the Wall Street Journal to National Public Radio, “PBS NewsHour,” CNN, “The Diane Rehm Show” and many more.
In a subsequent, April 15 letter to Sebelius, Public Citizen urged HHS to publicly release the protocols and consent forms for seven current and upcoming HHS-funded randomized trials on premature babies being conducted by the Neonatal Research Network — the same group of institutions that conducted the SUPPORT study. Public Citizen also demanded that Sebelius order a suspension of enrollment of infants into these trials until such documents were released and reviewed by independent experts to ensure that the study designs and consent forms were ethical. So far, HHS has not responded to Public Citizen.
In response to the damaging disclosures about the unethical conduct of the SUPPORT study, the investigators and various commenters speaking out in support of the investigators — including the editors of the New England Journal of Medicine and some renowned bioethicists — mounted an aggressive campaign to defend the ethics of the study and discredit the findings made by OHRP.
Subsequently in late April, Public Citizen obtained the complete SUPPORT study protocol, followed by the complete IRB-approved consent forms used by at least 22 involved institutions in early May. From these documents, we learned crucial new information about the use of the intentionally inaccurate pulse oximeters and how this unusual experimental procedure posed additional dangers to the babies. In a May 8 letter and report sent to Sebelius, Public Citizen outlined this new information and renewed its request for further action by HHS to address this matter. As this article went to press, a response from the Secretary had not been received.