Approximately 1,300 infants were enrolled in a clinical research study to determine optimal oxygen administration to low birth weight infants.  The Department of Health and Human Services recently informed the researchers that they failed to properly inform the parents of the infants of the risks of the research study. 1 The study was conducted from 2004 to 2009 in approximately 23 clinical research sites and was reviewed by at least 23 institutional review boards (IRBs).2 Infants enrolled in the study were randomized to either a low level of oxygen administration or a high level.  The purpose of the study was stated to be to determine the optimal level of oxygen to either prevent blindness from retinopathy of prematurity (high level of oxygen) or survival and neurologic development (low level of oxygen). The concern was that in spite of several clinical trials previously conducted on oxygen levels in low birth weight infants and the impact on the outcome,  there were lingering concerns that low levels of oxygen might impact the morbidity and mortality of infants.  Importantly however there was consensus in the medical community prior to the study that high levels of oxygen was a cause of blindness in low birth weight infants and low levels were associated with neurologic abnormalities.

In the clinical study titled “The Surfactant, Positive Pressure, and Oxygenation Randomized Trial (SUPPORT) a noninvasive device called a pulse oximeter was used to measure the blood oxygen saturation level to determine the amount of oxygen delivered needed to keep the oxygen saturation at a low range of 85% or a high range of 95%.  In their “Risk and Benefits” section of the research protocol the investigators did not identify any risks related to randomizing infants to either the high or the low dose of oxygen.2

The informed consent used in the study stated “Because all of the treatments proposed in the this proposed study are standard of care, there is no predictable increase in risk for your baby.”  The 6 page informed consent did not indicate to the parents that there were important health consequences of being involved in the study including severe eye disease, poor neurologic development or brain damage leading to death.  The most egregious part of the study was that the parents who enrolled their infants were informed that their infant was receiving “standard of care” while standard of care at the institutions was to closely monitor the infants and maintain oxygen levels between 85% and 95%, not keep them at one extreme or the other.  Thus, rather than receiving standard of care,  infants were assigned to either the low oxygen value or the high oxygen value, a randomization that would increase the likelihood for the study having significant statistical outcomes – either an increase in retinopathy of prematurity or neurologic complications.  The endpoints of the study were skewed towards complications rather than benefits.

The results of the study were reported in the New England Journal of Medicine in 2010.2  The infants treated with high oxygen had a statistically significant higher rate of severe eye disease while the infants treated with low oxygen had a statistically significant higher percentage of deaths before discharge.  The study placed over 1,300 infants at risk for complications of treatment simply confirm what some felt was already know from previous clinical trials. The study did not provide any additional insight into precisely what levels of oxygen might prevent either retinitis of prematurity or increased neurologic complications. 3,4

This study reveals that many IRBs do not have the expertise to thoroughly review the science and ethics of clinical trials placing thousands of research subjects at risk for complications of clinical research.6,7  We believe that an investigation by an independent review committee into the composition of IRBs, the expertise of their members and how they apply scientific and ethical principles to their review should be conducted, especially as the number, size and complexity of clinical trials increases and places an increased number of research participants at risk.


  1. Human Research Protection under Federal Wide Assurance (FWA) 5960.
  2. SUPPORT Study Group of the Eunice Kennedy Shriver NICHD Neonatal Research Network.  Target ranges of oxygen saturation in extremely preterm infants. N Engl J Med. 2010 362:1959-69.
  3. Study of Babies Did Not Disclose Risks, U.S. Finds.
  4. National Eye Institute.  Facts about Retinopathy of Prematurity (ROP).\rop
  5. Greenspan JS, Goldsmith JP. Oxygen therapy in preterm infants: hitting the target. Pediatrics 2006. 118:1740-1[volume]+AND+1740[page]+AND+Greenspan[author]&cmd=detailssearch
  6. Ammann AJ, Gough K, Caplan.  Were the interests of the vulnerable truly served? The predictable failure of HIVIG. J Acquir Immune Defic Syndr. 2012 Sep 1;61(1):e8-10.[author]+AND+hivig&TransSchema=title&cmd=detailssearch
  7. Ammann, AJ  From scarcity to abundance. Who decides the priority for clinical trials in resource poor countries?