Oversight of research is changing significantly with the Human Biomedical Research Act; IRBs, researchers and research institutions need to know its import and adapt accordingly as it comes into force over the next year or so. To this end, CENTRES and the Funding Initiative for Capacity Development in Research Ethics hosted a workshop on the Act, delving deeply into 9 cases to understand the Act’s provisions and implications. We held two back-to-back workshops on the topics, with a strong turnout totalling about 130. Three topics were discussed: Scope of the Act, consent requirements, and IRB/researcher/institutional responsibilities. Below, I’ll summarize the cases and discussion from each of them, interspersed with some photos of the event.
Part I: Scope
The main criteria for inclusion within the scope of the Act are intention (roughly, to study a disease or disorder; plastic surgery; or study performance or endurance) and intervention (roughly, having an intentional mental, physical or physiological effect on the body, or using identifiable tissue or health information).
Case 1: Nutrition and Pregnancy Study
The case involved a test of a nutritional packet on various health outcomes for expecting mothers and their offspring. Attendees agreed that it would be within scope, as health outcomes are ultimately studied because of their correlations with disease and disability outcome, and a nutritional packet has a clear effect on the body. Some wondered whether it would count as a clinical trial, however, and therefore covered by a different legal regime; it was advised that researchers with such doubts contact the Health Services Authority for a determination.
Case 2: Adolescent sleep study
Adolescents were to be sequestered for 15 days, with sleep deprivation induced and its mental and physiological effects measured. Again, there was consensus that it was within scope. Depression and anxiety measures were used as outcomes, clearly relating to health; sleep deprivation is a form of performance or endurance; and sleep deprivation has a clear physiological effect on the body.
Case 3: Mood Enhancer
The third case caused the most disagreement in the scope section. This involves a non-invasive tDCS stimulant device, designed not to treat a condition but rather improve healthy people’s moods. It clearly fit the second part of the definition, involving a direct effect on the body of the subjects. But did it meet the first? No disease or disorder was being studied. However, some participants thought that the device’s indirect effects on anxiety and the like may bring it within scope. More felt that improved mood may be a form of mental performance. But others countered that being in a good mood is hardly a “performance” – more like a state of being. We emphasized that, in cases of doubt, IRBs and researchers would do well to play it safe and assume the study is within the scope of the Act.
Case 4: Fitness band
Drawing on a well-known policy initiative, this case described a test of a wristband that measured paces per day and paid $10 to your account if you hit more than 10,000, with the goal of improving healthy behaviours. Opinions were still somewhat split, but leaned towards being out of scope by virtue of measuring behaviour rather than health outcomes. However, if the study involved explicit health outcomes (e.g., taking blood pressure measurements, checking for cardiac incidents, etc.), the numbers flipped and most agreed it would count as within scope. Some still wondered, though, whether a monetary incentive could really count as having an effect on the body.
Part II: Consent
If a study is within scope, the Act delineates various requirements for consent, including a “dual consent” model for consent of minors. It also allows waivers of consent in some informational studies, but only if strict criteria are met: minimal risk; study is impracticable without the waiver; the waiver wouldn’t otherwise violate the subjects’ rights; and the research is in the public good.
Case 5: Health Information Study
To test the requirements for a waiver of consent, we presented a health information study seeking such a waiver. The researcher argued that the security of the data made it minimal risk, it was socially valuable and therefore in the public good, and obtaining consent would mean less than 100% coverage, biasing the results and making the study scientifically invalid. Participants were generally agreed that the study was indeed minimal risk, but not sure about the impracticability requirement. It did not seem acceptable to say consent is impracticable because people would say ‘no’ if asked. However, if people were simply unreachable (contact information outdated, or they had died), that may be a valid reason to waive the requirement. Some suggested that sheer numbers of subjects may make getting consent from everyone too difficult, but it was clarified that “impracticable” is a very high bar well beyond just ‘impractical’. Furthermore, it is unlikely the study would meet the public good requirement, as that is intended to cover studies of national interest such as those involving national registries (clarity on the public good criterion will be provided by MOH at a later date).
Case 6: Adolescent sleep study, revisited
We returned to Case 2 above to discuss its consent implications, given that participants were minors aged 13-18. Dual consent of both minors and a parent is required when the minor has sufficient intelligence and understanding of the research; a common opinion among participants was that 13-year-olds typically meet this criterion. And even if they did not, given the onerous nature of the study (15 days away from their friends and family), consent should be obtained from the minor anyway on ethical grounds (and from both parents). This is a good reminder that ethics and the law can come apart; IRBs and researchers may want to be more strict than the law, when ethics demands it.
Part III: Responsibilities
The Act imposes various duties on researchers, IRBs and research institutions. Researchers have direct responsibilities to ensure consent is obtained, approval given, and participants’ rights respected. IRBs are tasked with approving or exempting research from review, while research institutions hold ultimate responsibility for oversight and control of the research (including appointment of said IRBs). Researchers and research institutions are potentially liable to fines or imprisononment for certain violations of the law, though IRBs are shielded from such sanctions.
Case 7: Collaborative Trial
We imagined that institution A (which lacks an IRB) sent a medical worker over to institution B (which has an IRB) to test out a new therapeutic technique, with a researcher at a foreign institution C analysing the data. How should research be governed? It was clear from the law that institution B had to oversee the research; A would not qualify as a research institution because it functionally does not exercise oversight and control (evidenced by its lack of IRB), and C would not qualify because it is a foreign entity. This means that institution B holds all the legal requirements (and liability), though the medical worker and foreign researcher are individually responsible to ensure their research is conducted responsibly (and liable if they fail to do so).
Case 8: Recruitment error
What if a researcher accidentally recruits a minor into a study, when the protocol specified only adults would be enrolled? Researchers are required by the Act not to deviate from the approved protocol, and the research institution is required to report any violation of the Act to the Director of Medical Services; furthermore, it was agreed that it was good institutional practice for researchers to report any deviations to both their IRB and research institution, though a ‘triage’ system may be needed to avoid flooding the DMS with minor deviation reports.
Case 9: New outcome measure
In a variation on the above case, what if a researcher decides post-approval measuring a new outcome would improve the study? As outcome measures are part of protocols, any new outcome measures would indeed count as deviation; the researcher should therefore go to their IRB before measuring the new outcome, and get an amended protocol approved (or exempted from review). If the researcher failed to get it approved ahead of time, and only informed the IRB or research institution after they had begun measuring the new outcome, they would be in violation of the Act.
This workshop was, by design, focused on particular legal ramifications of the new Act. But we should remember that researchers, IRBs and research institutions have ethical obligations that operate whether or not they are compelled by the law. So we encourage those involved in research not to think that simply adhering to the letter of the Act’s requirements means their oversight is sufficient. They should rather be cognizant of ethical concerns that may arise in the course of human subjects research even when the law is silent.