An associate professor at Columbia Law School in New York City wrote about the handling of research grievances from research studies for the Yale Journal of Health Policy, Law, and Ethics. For her study, Kristen Underhill, JD, DPhil, MSc, spoke with personnel from 30 hospitals and universities to see how they resolve the complaints. The processes for resolving research participants’ concerns are a “curiously unregulated space,” she wrote.1 While the United States court system might recognize claims about physical injuries during a trial or a negligent study design, it usually is left to the IRB to resolve disputes between trial participants and research staff.

In her interviews, Underhill found that, like the IRB at New York State Psychiatric Institute, many IRBs said their complaint resolution processes include “procedural flexibility.” This flexibility allows the complaint process to be tailored to each dispute. For example, a complaint that can be handled through facilitated negotiation can free up time and resources for more complicated disputes. Relevant stakeholders also can be added to the process as needed on a case-by-case basis. Underhill’s respondents also said the processes allows individuals to feel heard and respected. This would be more difficult using a standardized process, they said.

Overall, the respondents said complaint frequency was low, indicating to Underhill a possible problem with access to dispute resolution. Participants might not know they can complain, or how to begin the grievance process. They could fear retaliation or lack of access to the study drug if they complain. This could cause a higher dropout rate for the study, with unhappy participants choosing not to participate in the future. They also might share their perceived negative experiences with others in the community.

Underhill also addressed neutrality. Some respondents were uncomfortable playing the role of a neutral third party, she said. Ties to investigators and institutions complicated the loyalties. The respondents also said they often lacked resources or training to handle disputes, especially for complex ones that involve “mental illness, threats of violence, and volatile interpersonal dynamics.” The respondents said they wished they had access to other dispute resolutions from other institutions for comparison. Accreditation organizations could ask for examples of a written process for dispute resolution but do not set requirements for development or implementation.

Since the frequency of disputes often is low, the respondents wished their processes were more consistent, since they changed based on the type of complaint. The low number of complaints also made it more difficult to gauge the success of the resolution. “Institutions with larger research portfolios with a larger absolute number of complaints are less likely to have this problem, but informants from such institutions still noted difficulties with documenting complaints in a way that allows them to monitor for consistency and systemic problems,” Underhill wrote.

Underhill made several recommendations for how the dispute resolution process can be improved. One is to include the input of participant representatives in the development of the dispute resolution process, a step most respondents did not use. “Having a say in process development is important, in part, as a matter of procedural justice, but also as a matter of improving system accessibility, the durability of resolutions, and perceived legitimacy of the process (and the research institution more generally),” she wrote.

Secondly, the low uptake of the dispute solution process reflects a lack of information, particularly lack of awareness of the forum and the process for dispute resolution, Underhill noted. She suggested better education of subjects during study enrollment and follow-up. “Subjects’ awareness and understanding of protocols and ‘subjects’ rights’ — and thus, their expectations of how they should be treated — will inform whether they recognize wrongs as actionable,” she wrote.

To address the neutrality issues, Underhill said IRBs have several advantages to managing research-related disputes. They can suspend protocols, require revisions or remedies internal to research protocols, or cancel the protocols entirely. They can bring outside assistance into the dispute process, such as department chairs, legal counsel, human resources, and compliance departments. Multisite studies can refer the complaint to either the local IRB or the IRB of record.

Underhill also recommended these changes to the dispute process:

  • increasing publicity and accessibility through informed consent procedures and integration of participant community leaders;
  • compensating participants for physical injuries;
  • building IRB expertise and resources for conflict resolution;
  • using records to identify recurring complaints and improve consistency;
  • providing for advisory third-party review and reconsideration of decisions, even if that review is not binding.

Adjustments to processes for research-related harms are ethically warranted, she concluded. “The Belmont Report and other ethical guidelines have spoken widely on the need to minimize subject harm but have said little about how institutions can (and should) offer redress when they fail to do so,” Underhill wrote. “Participants in human subjects research take on many burdens in the interests of scientific progress; when they experience unintended harms, they should not bear the additional burden of unfair processes.”


  1. Underhill K. Righting research wrongs: An empirical study of how U.S. institutions resolve grievances involving human subjects. Yale J Health Policy Law Ethics 2019;18:2.