Family medicine group suggests guidelines
Recommendations cover three areas
Research in family medicine is a diverse field, encompassing everything from practice-based research to educational studies to research involving communities such as American Indian reservations. Because of this diversity, and because much family medicine research differs from the typical clinical trial, many researchers in the field find existing guidelines for protecting research subjects hard to apply, says William Hueston, MD, professor and chair of the Department of Family Medicine at the Medical University of South Carolina, Charleston.
In an effort to address that problem, he says the Society of Teachers of Family Medicine (STFM) received funding from the Association of American Medical Colleges (AAMC) and the Office of Research Integrity (ORI) to create its own human subject protection recommendations for primary care researchers.
"In primary care disciplines, particularly in family medicine, we tend not to do as many randomized clinical trials, but instead focus on things that involve participants in the community," Hueston says.
"What we proposed to the ORI was to pull out those areas that family physicians in particular and primary care doctors more generally do in research," he says. "Let’s see what the issues are that have to do with both data integrity and participant protection."
The resulting guidelines were presented last year at STFM’s annual conference, andHueston says they have been accepted for publication in an upcoming issue of the society’s journal, Family Medicine.
The panel that convened to write the guidelines focused on three areas requiring attention: educational research, practice-based research networks (PBRNs) and community-based participatory research.
Of the three, Hueston says the issues involved in educational research — the study of different educational interventions and their effects on students — were the most straightforward and most easily resolved.
"I think most people would agree that people participating in those endeavors need to be informed of what’s going on, to give permission for the use of the data, etc.," Hueston says. "But it’s a bit of a stretch for some of the educational folks who wonder; Wait a minute, if I change a lecture do I have to get permission from students?’"
While the requirements are not that strict, Hueston says students who are the subjects of data collection in research need to be advised of that fact and considered research participants.
"You need to give informed consent where applicable and this needs to go to IRBs," he says.
Hueston notes that an audit of educational research studies submitted for publication in Family Medicine in 2004 showed that 90% of them had not undergone review by or received exemption from an IRB. He says he was not surprised at those numbers.
The new guidelines call for all educational research studies to be submitted to an IRB for review, and for researchers in this field to receive training about regulations regarding human subjects protection.
Family Medicine itself announced last year that authors submitting a manuscript to the journal must include a statement indicating that it has been reviewed or exempted by an IRB.
Practice-based research issues
Practice-based research networks (PBRN) of physicians’ offices give rise to other issues of research participant protection. Hueston says that many PBRNs operate at very different levels of sophistication. At the high end of the scale are sophisticated national research networks that work with IRBs to ensure projects are well-designed, physicians and staff are well-trained, and the resulting studies are scientifically valid.
"But there are many smaller research networks with only a few practices or a handful of groups, where the level of sophistication may not be very high, and they may be engaging in research that really isn’t going to produce anything that’s terribly valid," he says. "In which case, there’s really no justification of any risk."
Another issue in research conducted within PBRNs is the fuzzy line between what’s considered research and what’s considered quality improvement in a practice.
"I think at an individual practice level, people are beginning to get a handle on what’s QI and what’s research," Hueston says.
"The problem occurs when you bind people together in a network and collect data from multiple sites. Now is it QI or is it research? It’s a difficult question."
In general, he says, when a network applies existing knowledge to its patients to improve care, that’s quality improvement. An example would be implementing a new technique that has proved to be successful in published studies.
"But generally, people use the criterion that if it is intended to be published, to demonstrate new knowledge, or to disseminate the application of existing knowledge, then it’s research," Hueston says.
The PBRN guidelines offered by the STFM panel include:
— assuring that PBRNs’ research projects will provide valid and meaningful scientific information;
— streamlining informed consent, where appropriate, in projects with reduced risk, and allowing for central IRB approval or acceptance of consent forms approved by other certified IRBs;
— divulging any conflicts of interest by all investigators in a PBRN project;
— collaboration by investigators, clinicians, and regulators to better define the boundary between research requiring IRB approval and quality improvement projects.
"I think if we set these as guideposts, then even the less-experienced networks can say, OK, we really need to work towards them,’" Hueston says.
"I’m hoping this improves the overall caliber of PBRN research, particularly among those smaller networks that may not have as much research sophistication or the resources that they’ve been able to put into this in the past."
Community research raises tough issues
Of the three areas the STFM panel focused on, Hueston says the issues raised by community-based participatory research were the most difficult.
An example of this type of research might include researchers going into a community such as a Native American reservation or a small, rural town and conducting a needs assessment with residents to identify health care problems and study possible solutions. Researchers might study the success of a stop-smoking campaign or a project in schools to teach healthy behaviors.
The introduction of that third party — the community, in addition to researchers and individual subjects — creates problems in designing studies, Hueston says.
"The town’s perspective may not be shared by the researchers, and how you actually implement that intervention can be ethically complex," he says. "Does everybody in the town have to consent before you try to improve the health of the community? Who has to say yes, and who speaks for the community? There’s no elected health spokespeople, you have to go with key informants, people who seem like they’re leaders and opinion experts and hope that you’ve got the right folks."
When it comes time to interpret the data, the community may have a different take on the data collected and may have a stake in wishing to protect its reputation, Hueston says.
"Is it the researchers’ data or the community’s data? And how do you collaborate to report what happened, keeping confidentiality and protecting the reputation of the community as well?" he says.
"Those are all issues you don’t find in the manuals for controlled randomized trials or benchtop research." These tend to be issues that IRBs have little or no experience with, he says, because their emphasis is on risk to individuals.
"IRBs will have no idea what’s going on here," he predicts. "The town doesn’t have an IRB. The subjects or participants may have different perspectives than the IRB in an institution many miles away in a different cultural milieu."
All of these issues caused considerable debate on the panel, Hueston says.
"How far do you allow the community to decide what gets published and what doesn’t? Do they have veto power if it doesn’t look good to them? Where really do you draw the line?"
The panel’s recommended guidelines include:
— including protection of the community and its possible stigmatization as part of IRB oversight of such studies.
— requiring review and approval by community representatives in addition to the IRB.
— review of the plan for disseminating findings from the study, particularly results that may be sensitive.
— protecting data from the research, and making it available for additional analysis only with consent of the participating community and researchers.
Hueston says that this area will require further consideration and debate.
"I think that’s the area that will require the most work in the future to be sure that these are played out and that people adhere to them," he says. "It has the potential for a lot of harm for participants if not ethically performed."