In Support of SUPPORT ??? A View from the NIH
June 5, 2013
Each year in the United States, nearly 500,000 infants — 1 in every 8 — are born prematurely, before 37 weeks of gestation. Despite substantial advances in their care, premature infants face a daunting array of challenges; they are at high risk for death in infancy and face severe and lifelong health problems if they survive.1 The National Institutes of Health (NIH) has a legal and moral responsibility to do research in partnership with scientists and families to optimize the care of these highly vulnerable infants. In recent weeks, a major public debate has arisen regarding a study designed to do just that. And the ramifications go well beyond this one study: the outcome of this debate could affect how we conduct and communicate about critical research on interventions that are within the standard of care for all diseases and conditions.
The Surfactant, Positive Pressure, and Oxygenation Randomized Trial (SUPPORT), carried out at more than 20 sites between 2004 and 2009, sought to identify, in infants born very prematurely at 24 to 27 weeks' gestation, the oxygen-saturation level within the range considered the standard of care that would minimize the risk of retinopathy of prematurity (ROP), a complication of oxygen therapy that can result in vision loss.2 When the study began, targeting an oxygen-saturation range of 85 to 95% was becoming standard clinical practice, and the American Academy of Pediatrics (AAP) later recommended this range in its 2007 guidelines. The SUPPORT researchers and institutional review boards (IRBs), practicing clinicians, and the AAP had no scientific evidence to expect a difference in mortality between the two treatment groups in SUPPORT — one with the oxygen saturation target of 85 to 89%, the other with the target of 91 to 95%.
An important finding of the study was a reduced incidence of ROP in the lower oxygen-saturation range. However, contrary to what was known at the time, the study also showed a slightly but significantly increased incidence of death — 19.9% versus 16.2% (P=0.04) — among infants assigned to the lower as compared with the upper range. As a result, last year the AAP amended its guidelines, citing SUPPORT, and physicians treating very premature infants are starting to use higher saturation rates to reduce the risk of death, even with the potentially higher risk of ROP at these levels. Studies such as SUPPORT that compare two alternatives, both within current standard clinical practice, often lead to critical improvements in medical care.
The federal Office for Human Research Protections (OHRP), which is charged with providing leadership in the protection of the rights, welfare, and well-being of persons involved in research conducted or supported by the U.S. Department of Health and Human Services (DHHS), asserted in March 2013, on the basis of its own examination of the evidence, that the SUPPORT researchers failed to provide prospective parents sufficient information about the risks posed by the study. After a detailed review of the protocol, the relevant consent documents, and the research literature, we respectfully disagree with the conclusions of the OHRP, which we believe resulted from a fundamental difference in interpretations of how the regulations should apply to the state of scientific understanding when the SUPPORT study commenced. Moreover, there is a larger issue here: how risks should be conveyed in the informed-consent process when research is comparing interventions that are all considered to be the standard of care.
In a letter dated March 7, 2013, the OHRP asserted that the study's consent form failed to convey 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.”3 That finding was influenced by research conducted in the 1950s, but in our view, it failed to assign proper weight to studies conducted in premature infants in the 2000s, which used more sophisticated oxygen-monitoring and oxygen-measurement devices, similar to those used in SUPPORT.4 The more recent studies showed no increased risk of death or neurodevelopmental impairment at saturation levels as low as 70%.5
Given these data, the investigators had no reason to foresee that infants in one study group would have a higher risk of death than would those in the other group. The babies included in SUPPORT were, of course, facing substantial risks because of prematurity — the same risks as premature babies who were not enrolled in the study — but their care was never compromised for the sake of the study. The sample consent form for SUPPORT stated that each of the “possible combinations of treatments is considered by some units to represent their desired approach” (www.nih.gov/icd/od/foia/library/Records.htm). This statement describes the clinical equipoise at the time of the study, which was, in fact, the justification for conducting a clinical trial. Although the OHRP took issue with the consent form, it stated that the study design was ethical — a conclusion worth emphasizing. The increased risk of death was a significant and unexpected finding of the study; if it had been known before the study began, standard clinical care would not have encompassed the lower oxygen range, and it would have been unethical to conduct the study.
The NIH is committed to ensuring that prospective research participants — and the people who speak for and love them — are given clear, complete, and accurate information about the risks and benefits of participating in research. We are strongly committed to supporting critical research studies like SUPPORT, which inform clinical care by providing rigorous evidence for use in daily practice. This controversy has alarmed some of the parents of infants who were in the study, confused the biomedical research community, and befuddled IRBs. Several other studies seeking new insights to improve care for these vulnerable infants have been put on hold as the field tries to understand the OHRP findings.
But controversies such as this are also an opportunity to advance shared understanding, provide clarification, and encourage progress. The public debate surrounding the SUPPORT study has set the stage for a substantive national dialogue with the research, advocacy, and ethics communities on how best to respect and protect participants in research studies conducted within the standard of care and how to define “reasonably foreseeable risks” in this setting. The timing is critical — the clinical research community, bioethicists, regulators, IRBs, and prospective research participants are paying close attention now. The NIH is happy to work with all stakeholders to advance this important dialogue and its translation into clear guidance, in accordance with the plan just announced by the DHHS (www.hhs.gov/ohrp/). In addition, a new letter to the University of Alabama at Birmingham from the OHRP, stating its intention to put all compliance actions on hold until the process of producing appropriate guidance is completed, is available now on the OHRP website (www.hhs.gov/ohrp/detrm_letrs/YR13/jun13a.pdf).
Going forward, the NIH strongly and unequivocally supports the importance of the role of the OHRP in the oversight of human subjects research. But the community will benefit from an explicit description of the process the OHRP follows for investigating complaints. For example, when questions are raised about reasonably foreseeable risks and the state of the science relevant to a particular clinical trial, appropriate independent experts might need to be consulted. Finally, we are pleased to see the DHHS plans to ensure that investigators and IRBs will have a fair and transparent process for appealing OHRP findings and compliance actions, in those situations in which reasonable people disagree about the actions taken.
The circumstances surrounding the SUPPORT study have unquestionably created controversy in the research community, but the situation has created an opportunity for a better understanding of the scientific and ethical issues that must be addressed when designing such studies in the future. We look forward to working with the OHRP, the research community, and patient advocates to improve the effectiveness and ethical standards of research involving human participants.