When to inform about subjects' threats is unclear
Weigh laws, confidentiality, and severity of threats
Public health researchers studying sensitive issues — suicide, domestic violence, drug use — sometimes find themselves dealing with more than just the survey questions at hand.
In the course of interviews and surveys, they may learn about threats affecting third parties or participants' suicidal urges.
What is a researcher's duty to inform someone about these threats, which are not the direct result of research but are uncovered by it?
David R. Buchanan, DrPH, a professor of community health education at the University of Massachusetts, Amherst, MA, says the answer isn't as clear-cut as one might think. Laws governing mandatory reporting are vague and vary widely from state to state. It's hard for a researcher conducting a simple survey to know how serious the threat might be. And reporting such information raises as many ethical problems as it potentially solves, since it breaches promised confidentiality to research participants.
Buchanan worries that IRBs are increasingly inclined to simply turn down proposals for this type of research, rather than wade into the murky legal and ethical issues involved.
He tells the story of a researcher he knows who was doing a longitudinal study of Vietnam veterans, one of whom committed suicide during the course of the study.
"Was that her responsibility?" he asks. "Nonetheless, the IRB shut down the research."
Buchanan says incidents such as this have made some researchers reluctant to discuss the ethical quandaries involved in handling non-research related risks for fear of drawing even greater IRB attention to it. But he believes a thorough airing of the problem is necessary, in part to show IRBs that the risks are small and shouldn't hamper important studies.
In a paper published in the Journal of Public Health Policy, he notes that public health researchers have a duty to study leading causes of morbidity such as suicide, homicide and AIDS, which can lead them to work with high-risk populations.
He argues that researchers doing simple surveys and observational studies have limited contact with participants, and aren't in a position to gauge the severity of threats. So, he says it makes no sense for an IRB to deny a proposal based on concerns the researcher and institution might later be sued.
"Is it reasonable to put a burden of a duty to warn on researchers, or post hoc hold them liable, when any reasonable person would say there's no way they could have anticipated (a violent act)?" he says. "If it's not reasonable to expect that a researcher could have anticipated this event, then they should not be held to some standard of a duty to warn."
Buchanan says he first became concerned about the issue of non-research-related risks when he was working on his doctoral dissertation.
He distributed a confidential, but non-anonymous survey questionnaire to a group of adolescents asking about their future intentions to engage in risky behavior such as smoking marijuana. He says that one respondent crossed out one item in the list and wrote in the words "commit suicide."
"I was really torn about what to do," he says. "And I did go talk to a school counselor who, ironically, just shrugged his shoulders and said, what you do you think?"
Buchanan says public health researchers can discover other similar threats to health and safety in the course of conducting studies. The researcher may see an HIV-positive subject sharing a needle with someone he knows to be HIV-negative. A participant in an interview about contraceptive practices could express his intention of beating his partner because he suspects she is unfaithful.
Buchanan had written other articles examining these issues, and a few years ago was invited to participate in a workshop sponsored by the National Institute of Drug Abuse. There he found a number of researchers dealing with the same sorts of questions.
"They were all wrestling with it individually, wrestling with their individual conscience, and deathly afraid of calling further attention to it," Buchanan says. "Because, particularly these days, of the potentially chilling effect it could have on this research."
"But that seems to me to be the wrong direction to take," he says. "Right now, with less known about these issues, the fears are probably overstated and exaggerated. It is an exceedingly rare occurrence. We could do a good job of just documenting the things that could happen and documenting the circumstances in which they're going to happen, if people would be willing to talk about it."
Laws vary, are unclear
Legally, researchers and IRBs confronted with a non-research-related risk must consider several issues:
• Mandatory reporting laws, which vary from state to state, both in terms of who is covered and what kinds of threats they must report. Buchanan says that because these laws tend to cover those health personnel delivering care, it can be confusing as to whether researchers must report threats.
"Say a nurse is doing an interview," he says. "As a nurse, they're a mandatory reporter, but as a researcher are they?"
• The researcher's (and the institution's) liability in the event the threat is carried out. The 1974 case most often cited as laying out a duty to inform is Tarasoff v. the Regents of University of California, in which the California Supreme Court ruled that a psychiatrist should have breached doctor-patient confidentiality to warn a woman that his patient intended to harm her. The woman was later murdered.
Buchanan points out that in the wake of Tarasoff, psychiatrists have been found to have a poor track record in correctly predicting a patient's violent episode. He says a researcher with more limited contact to a subject could be expected to have even less success in correctly gauging the threat the subject poses.
• Informed consent assurances of confidentiality. If there are circumstances under which confidentiality may be breached, the subject must be informed ahead of time in the consent document.
From an ethical standpoint, Buchanan says the researcher's duty to inform hinges on how much he or she knows about the seriousness of the threat. The researcher must weigh that risk against the potential harm involved in breaching confidentiality.
He says that if a researcher's contact with a subject is sustained and in-depth, it might be reasonable to assume that the researcher would be able to make an informed determination about how serious the threat is. But if the subject is merely filling out a questionnaire and handing it in, the researcher wouldn't be in a position to make that judgment.
Ways to assess true risk
Buchanan argues that IRBs reviewing studies that raise questions about non-research-related risk should consider the same question: Will the researcher be able to gauge the seriousness of the threat? If not, it's unlikely he or she could be held legally liable, and so the research shouldn't be hindered.
He suggests other steps that could help in reviewing these types of studies:
• Assessment tools for threats such as suicidality. Buchanan says the tools currently in use have improved to the point where they can be useful in a survey setting.
"You can build into protocols these triggering questions," he says. "If somebody responds 'yes' to the question, then ask the following two questions. If they say 'no,' then skip to the next section."
• Informed consent documents that spell out under what circumstances confidentiality would be breached.
"Basically, it's the same rule of thumb," he says. "If, in assessing the instruments that are going to be used, it looks like the researcher is likely to gain sufficient information that they can make a better than 50/50 assessment that some adverse event is likely to happen, [the IRB] should definitely stipulate that the researcher has to put that into the informed consent: 'Under the following circumstances, we will take actions to try to prevent harm.'"
• Necessary training to ensure staff safety. In addition to participants, IRBs also should consider the possibility that research staff may be exposed to threats because of their involvement in a study, Buchanan says. He says savvy investigators already make sure that staffers who deal with high-risk populations get additional training.
"It's certainly a concern among researchers in the drug world," he says. "I do think that's something that should be pretty routinely required of any research project that's doing field research with potentially dangerous populations."
Buchanan hopes to see the public health research community come up with its own set of criteria for evaluating possible harms encountered in this type of research.
In the meantime, he hopes to persuade IRBs not to dismiss studies too quickly over what he believes are exaggerated risks.
"We need to keep in mind the value of the research — that ultimately, this will have tremendous social value," Buchanan says. "We would be remiss to be overly conservative in failing to approve research with high-risk populations simply for very remote chances that incidents will happen."
"It would certainly be fair for IRBs to ask researchers to address questions of how well do they think they would be able to predict and anticipate some adverse event. But if it seems unlikely that the researcher will have sufficient information to predict some sort of event, then I would strongly urge IRBs to approve the research. Anticipate that their position will be upheld by the courts, which seems to be the direction the courts are going in the post-Tarasoff era."
Source
Buchanan DR. Policy needs regarding the duty to protect in epidemiological research with high-risk populations. Journal of Public Health Policy. 2006;27:293-308.
Public health researchers studying sensitive issues suicide, domestic violence, drug use sometimes find themselves dealing with more than just the survey questions at hand.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