NEW YORK (GenomeWeb) — As efforts march forward to study the utility, as well as the medical and psychosocial impact of clinical sequencing, the field eagerly awaits best practices for how to obtain patient consent to receive the results of broad genomic analysis.
But defining such recommendations is a complicated task, reflective of the diversity of clinical applications for these new technologies and resulting strategies to engage and inform patients in different clinical settings.
This is one implication of a recently published analysis, by several social scientists and ethicists, of approaches for informed consent and return of results adopted by projects under the National Human Genome Research Institute's Clinical Sequencing Exploratory Research and Return of Results programs.
In the study, published last month in the Journal of Law, Medicine, and Ethics, the authors set out to empirically analyze consent forms used in six of the CSER studies and three projects under the RoR consortium, addressing the similarities and differences in these approaches by using a coding system to directly compare each project's consent strategy.
Gail Henderson, the study's first author and a professor at the University of North Carolina at Chapel Hill, told Clinical Sequencing News that she and her coauthors expected there to be significant diversity in the consent procedures they looked at, but were somewhat surprised by just how much of a lack of convergence they found.
In their analysis, the researchers focused on four specific questions: first, what results the program planned to return to participants; second, how the preferences of patients are taken into account; third, how risks and benefits of sequencing are addressed; and fourth, how programs planned to deal with privacy questions like placement of results in electronic medical records, risk of re-identification, or data sharing.
"With this study we were trying to see if people addressed these four specific areas in ways we could analyze as being similar, and if there were differences, what the nature of the differences was," Henderson said.
"We didn’t have a hypothesis," she added. "We felt that it would probably be pretty diverse, since [some] of us study authors are part of working groups [associated with these projects] and we know the projects are so different." For instance, she said, some projects involve cancer patients, some other diseases, or healthy subjects, while some projects include only children, only adults, or a mix.
"But even though we expected differences, one of the things we wanted to emphasize in this paper was the lack of convergence. They were really so divergent, and there didn't seem to be any significant norms guiding the types of results offered or the ways they were categorized. So it was important for us to put this out there to the broader community to say that it's clear there is not going to be one size fits all."
However, in analyzing the nine programs' consent forms, Henderson and her colleagues did glean some points of recommendation. Overall, for example, the authors concluded based on their analysis that several details appear to be especially important in consent for sequencing, including defining what whole-exome or whole-genome sequencing is and addressing its limitations, as well as describing how sequencing results are reviewed to determine returnable findings. In addition, the authors recommended special attention to educating participants about the meanings of positive, negative, and uncertain results, risks and benefits of participating, and privacy concerns.
In reviewing the written consent forms used by the nine projects, the researchers found a significant diversity of approaches across all four of their target areas.
For example, the researchers found that a common approach among studies was to list different categories of results that might be returned, but what defined or fell into these categories differed from study to study. Only three projects clearly distinguished primary diagnostic from incidental findings, while others mentioned these categories in other ways, and two efforts did not make any distinction.
On the question of participant preference, the approaches were spread from one extreme to the other, including one project where not a single category of finding was governed by participant preference, and another where all categories were so subject.
According to Henderson, this lack of convergence is something inherent to the diversity of the different CSER and RoR study designs, and reflects the fact that future guidelines and best practices for returning results are likely to be diverse, and dependent on the clinical context.
The question of how to obtain patients' consent and how their preferences should be taken into account in returning results depends, for example, on who the patients are. Within the CSER projects, some studies are dealing with critically ill cancer patients, while others are recruiting healthy subjects in the context of normal primary care. Among late-stage cancer patients, there is wide enthusiasm about sequencing and a strong desire for results of all kinds. These programs are seeing near 100-percent response rate in terms of patient recruitment. Others, which deal with healthy subjects, may see only half that, Henderson said.
At the same time, this raises a question about how the results of the CSER and RoR projects can help the field develop a set of best practices or guidelines for future clinical sequencing.
"This is a really important point," Henderson said, "and it's actually a little bit of a tension in the CSER consortium. These are exploratory studies, so when people were writing applications, it was any study design, any patient population is OK — the more diverse the better. So this is really a group of very different studies, not a multi-center trial."
Despite the challenges posed by this diversity though, Henderson said that the program is working hard to identify any areas where there may be commonalities across the studies.
"Early in the consortium, we really tried to come to grips with this by putting together working groups. I am part of the outcomes and measures working group, for example." This team is looking at ways the different projects measure harms and benefits. "We've seen that all the studies think that measuring genomic knowledge and literacy and numeracy are very important, and looking at how well people understand the different outcomes of sequencing, positive, negative, and uncertain results," she explained.
According to Henderson and her colleagues, overall, their analysis has shown clearly that the projects in these programs are exploring multiple and diverse models for consent. Combing through the results of the CSER and RoR consortia to identify useful points for clinical sequencing guidelines best practices may be complicated, but multiple, diverse models will likely be necessary for future clinical sequencing efforts, considering the diversity of clinical applications for sequencing itself.