The National Institutes of Health (NIH) recently announced that one of its studies showed testosterone treatments can improve anemia in elderly men. But in the same JAMA issue where the study results appeared, an editorial questioned the ethics of the study. The editorial criticized the study’s protocol because it included no means to alert test subjects in case the testosterone treatment had worsened instead of improved their anemia levels.
The author argued that in this study the researchers had an ethical duty to return test results to the test subjects. The protocol did exclude any potential subject who already had anemia, but it included no mechanism to track or alert a specific person if his hemoglobin levels fell too low during the study. The hemoglobin testing happened at an offsite facility and only after the study’s procedures had ended.
The question of returning results is always a difficult one. Although the JAMA article indicates that IRBs had opportunity to identify the issue during their review of the study, other experts would suggest that the IRB’s role in this area is a matter of debate—or is at least evolving . So far, research regulations say little to nothing about returning results; Europe and Canada have done a bit more than the United States, but overall official guidance is sparse. Still, many IRBs and other ethics review committees realize the importance of these questions and have established their own guidelines for assessing whether, when and how study participants should receive results from studies they are a part of.
SACHRP (the Secretary’s Advisory Committee on Human Research Protections) has also been wrestling with these questions about releasing study results. So far, it has presented recommendations for three of the four situations that “returning results” might describe.
(The odd category out is return of incidental findings, and SACHRP may release comments on that this year.)
JAMA’s editors relied on the most recent of these recommendation, on returning individual study results, to bolster its position that the testosterone study was unethical.
In fact, SACHRP’s endorsement of returning individual results was not unequivocal. While its main recommendation supported returning study results to individuals in general, the discussion also raised conditions which might in certain circumstances argue against it. This is not to second-guess the JAMA editors, or the researchers and institutions that conducted the research, but it’s interesting to review what SACHRP identifies as potentially mitigating factors.
- Deciding to return results to participants could risk the blind in a specific study, and SACHRP acknowledged this could be a material issue in decision-making. (In the testosterone study, the investigators did not learn about hemoglobin levels because an outside laboratory conducted the tests after the study visits ended.)
- Shifting the purpose of any lab test from tracking experimental results to directing healthcare decisions raises a concern for CLIA certification. Research-oriented laboratory procedures fall outside of CLIA enforcement, but if any results could lead to medical responses, it would need to be measured in compliance with CLIA requirements. (It was not clear from the journal article whether the hemoglobin tests were CLIA-compliant.)
- Asking a researcher to track and report on a participant’s health could risk therapeutic misconception. The boundary may blur, at least in the eyes of the participant, between helping advance scientific knowledge and receiving primary medical care. (As noted above, this study’s design prevented the researchers from learning any specific lab results. In this specific case, the risk of therapeutic misconception was lessened, as was any potential opportunity to help an individual.)
The questions around returning results are important ones and, as responses to this NIH research shows, there’s not necessarily consensus around the answers. Pressure is growing for the public release of all study results (Regulations are further developed along these lines in Canada and Europe than in the United States.) In response to the JAMA editorial, the NIH-backed researchers agreed to inform all of the test subjects directly about the results. SACHRP may add more insight to the issue this year if it does release recommendations about incidental findings.
Even if the regulations say little about returning results, we who design, conduct, and review human subject trials might take a lesson from this study to assess when and how to return results.