Are third parties due protection?
Social-behavioral research often involves others
Much of social-behavioral research is, at its heart, the study of social relationships — between ourselves and our family, our friends, our coworkers, and others in our lives.
When IRBs review protocols for this research, the focus of their efforts is to protect the subjects who are speaking to investigators or filling out surveys. But IRBs also must be alert to the potential risks that such research may pose to others who are named in subjects’ discussions. These third parties may have sensitive personal information revealed about themselves without ever signing a consent form or even knowing they have been discussed.
The key to dealing with third parties, as it is with subjects themselves, is to carefully think through what information is elicited and from whom at the outset of a study and to vigorously protect the resulting data, say those who have looked at the issue.
Felice Levine, PhD, executive director of the American Educational Research Association in Washington, DC, was co-chair of a working group that examined this issue and developed recommendations for investigators and IRBs.
"The investigator really must wrestle with what is the focus of the study," Levine says. "Who and what is this project about? From whom do we need to gain consent?"
In the review process, the IRB’s role is "to ensure that all data have strong confidentiality protections and for an IRB to further address the question of whether indeed the study is about multiple parties and not just about what the investigator thought was the identified party," she says.
The consequences of failing to secure third-party protections can be severe. In 2000, human subjects research at Virginia Commonwealth University was shut down for nearly a year after complaints by the father of a research subject in VCU’s Mid-Atlantic Twin Registry.
The father had opened an envelope addressed to his then college-age daughter and found a survey that included sensitive questions about the daughter, her twin and himself. He filed a complaint with OHRP. Part of the father’s complaint was that he had not given permission for sensitive information to be gathered about him.
While the third-party complaint was not the only issue leading to the temporary shutdown at VCU, "it was a precipitating event, and clearly it raised for all of us some awareness," says Ann Nichols-Casebolt, PhD, associate dean at VCU’s School of Social Work and chair of the social-behavioral IRB at the Richmond-based university.
"We’re prompted by health care providers and encouraged in a public health sense to be aware of our family health history and to have all that information," she says. "But that’s very different than when you’re in a research context and someone is asking you information about someone else. "We’ve become much more sensitized and sensitive to that."
Do we have a human subject?
When examining issues of third parties referenced in research, the principal question IRBs must ask is whether the third party is considered to be a human subject afforded the rights and protections of the Common Rule.
In 2001, the NIH examined this issue and issued a list of recommendations on how to treat third parties.1 The panel found that while third parties mentioned in research aren’t human subjects per se, they could be seen as human subjects for the purposes of the Common Rule if individually identifiable private information is collected about them.
If both parts of that requirement are met — the information sought is individually identifiable and private — then the researcher should gain informed consent from the person or seek a waiver.
Individually identifiable information would include such things as full names, addresses and other contact information, Social Security numbers and identifiable photos. However, the NIH definition would not define relationships that are identified only by an association, such as a spouse, father, mother, sister, friend, as "readily" identifiable information for the purposes of the Common Rule.
Those associations, the NIH recommendations state, would be considered "possibly" or "potentially" identifiable, requiring effort to ascertain a person’s identity.
While private information could include many kinds of health information about individuals, more general information, such as age or ethnic or cultural descriptors, would not be.
In addition, a researcher may ask a subject about his or her family’s health history, in order to understand the subject’s own behavior or decisions. The NIH guidelines suggest that type of information would not be considered "private," and Levine agrees.
"If my father had an early heart attack, asking me how that might have shaped my experience is very different than asking me to hand over his medical record," she says.
The NIH panel recommends that investigators and IRBs carefully evaluate the relevance of the information being asked of one party about another. If verification of the third party’s information is important, then that person should be recruited as a research subject.
Levine co-chaired a social and behavioral sciences workgroup of the National Human Research Protections Advisory Committee (NHRPAC), which in 2002 issued its own clarification on the status of third parties identified in research.2
The group’s recommendations state that referring to a third party in a study doesn’t necessarily mean that party should be treated as a research subject. But investigators and IRBs should be mindful about whether the information collected is sensitive and could cause harm to a third party. If a third party is identified as a subject and must be contacted for consent, the NHRPAC recommendations note that the IRB must consider whether such a disclosure could harm the original subject of the study.
"One of the issues about migrating a third party to a human subject is that when one does that, if one didn’t do that with a high degree of caution, that could potentially change the nature of the promises and assurances provided to the primary subjects of the research," Levine says.
"For example, in a family where there is spouse abuse, one certainly would not entertain the notion that one should not only speak to the spouse but the abusing spouse and inform the abusing spouse that this research is going on," she says.
Data protection is key
Levine says the other important point for investigators and IRBs is ensuring that all information collected in the course of the research, including information about third parties, is kept confidential.
"There’s a primary issue that all data, all information provided by subjects, whether only about oneself or that which might pertain to others, needs to be protected with the highest degree of security and data protection plans," she says.
OHRP has not yet published its own guidance on the handling of third parties in social-behavioral research; spokeswoman Patricia El-Hinnawy says the agency is in the process of creating the guidance, but no timeline is yet available for its completion.
At VCU, the social-behavioral IRB takes great care with research proposals that might involve third-party information.
"In general what we have said is if it is private identifiable information about another person, that the general rule is no, we cannot have that individual provide that information about a third party," Nichols-Casebolt says.
When considering whether information is identifiable, she says her IRB goes so far as to include relationships such as parents or spouses.
"If you ask, Does your mother or father have this?’ then you’re not asking that third party’s name, but if you have the name of the individual you’re interviewing, they only have one mother," Nichols-Casebolt says.
If the researcher is asking for third-party information as a means of recruiting other subjects — e.g., seeking out family members for research with a genetic component or other members of a group sharing characteristics such as HIV infection— the IRB requires that the researcher give the subject the original subjects information about the study and ask them to pass it along to others, she says.
She does note that there are times when the situation isn’t clear-cut, particularly when a researcher is asking a subject about their perceptions of another person’s behavior, in order to learn the effects of the behavior on the subject.
"I think that in those situations, we’ve tried to look at every instance as a separate one and look at the risks and the benefits. What is the purpose for doing it?" she says.
Nichols-Casebolt says she believes there’s enough flexibility in the guidelines to allow for research of groups such as survivors of abuse or adult children of alcoholics, where information about third parties provides important context.
"Some of these issues are so important — we want those subjects’ perspective," she says. "We just try to think it through and put ourselves in the perspective of the third party would I want my child or my parent to participate? But we’re also researchers and trying to strategize about ways to respect that third party but help facilitate the research."
References:
- The National Institutes of Health’s Recommendations on Protection of Third Party Information in Research: www.nih.gov/sigs/bioethics/nih_third_party_rec.html.
- The National Human Research Protections Advisory Committee’s Clarification of the Status of Third Parties When Referenced by Human Subjects in Research: www.aera.net/aera.old/humansubjects/NHRPAC_Final_Third_Parties.pdf.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content