Keller, Daniel M. PHD
Patients often enter a clinical trial hoping that they will benefit from a new drug, device, or procedure. But what they are also consenting to is taking part in a scientific endeavor designed to answer a specific question. Once the results are in, do the study subjects have a right to know the answer, and if so, what are the logistics of how that should be communicated?
One survey of 409 parents of children with cancer and of 86 adolescents with cancer (Fernandez CV et al: JCO 2009;27:878–883) showed that the vast majority of parents (94%) felt that they were entitled to know the research results. As for how, for positive results the parents said they preferred receiving the news by letter, an e-mail summary, or a phone call followed by a letter. If the results had negative implications, though, the parents said they wanted to be contacted in a more personal manner—for example, by phone or in a visit.
Parents were about equally divided between wanting to get the news in person from the researcher (22%) or their oncologist (21%). Others stated a preference for meeting with their family physician or a research nurse. Almost 30% felt they did not need personal contact.
In a review of studies on trial participants’ desire to know the results of the trial, David Shalowitz and Franklin Miller (PLoS Medicine 2008;5:0714-0720) found that a median of 90% of people (range of 20% to 100%) wanted to know the aggregate results or their individual trial results or those of their participating children. More than half of the trials reviewed in that study involved cancer or genetics.
The Fernandez et al survey found that respondents wanted to learn the trial results for two main reasons: First, the results could have direct health care implications for their children, including for surveillance and future decision making. Besides the major outcome of a trial, the majority of participants (83%) said they wanted to hear about long-term sequelae and suggestions for trial participants. A main concern of one third of the parents was about being able to understand the researcher, and about half of the adolescents had this same concern.
The second reason was that they thought that more transparency could lead to a better understanding of trials by the public and the importance of research participation. These participants also viewed themselves as partners in the research and were interested in future research plans (69.5%).
This survey illustrates several points relevant to the issue of communicating clinical trial results to participants: whether they want to know, what they want to know, who should relate the results, and in what manner. Fernandez and colleagues reported that more than one third of United States and Canadian institutional review boards (IRBs) require researchers to address the issue of communicating trial results to participants.
Ann Partridge, MD, MPH of Dana-Farber Cancer Institute and Harvard Medical School, has written on the subject and wrote an accompanying editorial to the Fernandez et al study. “I think that routinely we still don't uniformly give results to most patients,” she said in an interview for this article. We're trying to do it more, and we're instituting through CALGB [Cancer and Leukemia Group B]…a policy where we're posting the results in plain language on the Web site so that it can be acceptable for patients, as opposed to them having to search scientific Web sites and meeting abstracts to find results.” But, she said, “It's a lot of work.”
Pros & Cons
Writing last year on the pros and cons of disclosing the results of cancer genomic studies (Clin Cancer Res 2009;15:4270–4276), Lynn Dressler, DrPH, suggested several criteria for determining which research results to disclose to study subjects:
* The amount of uncertainty that a participant is willing to accept.
* The positive predictive value of a finding.
* The magnitude of harm that may result.
* What would be considered useful information.
Dr. Dressler, Associate Director for Policy and Ethics at the University of North Carolina Institute for Pharmacogenomics and Individualized Therapy, compared the policies on disclosure of five professional, regulatory, and advisory organizations and showed a range of positions on various points, indicating that the field is still in flux. The organizations did agree, though, that if research results are to be disclosed, that the results should be validated both analytically and clinically.
Dr. Dressler listed several arguments supporting disclosure: First, one should respect the autonomy and self-determination of individuals, including the possible value of the research result to the study subject. Second is the principle of “beneficence and nonmaleficence,” or doing good and not doing harm. Receiving trial results can be empowering for participants, such as deciding to make changes in lifestyle or plans, as well as knowing that their participation led to useful results.
“It provides them with some sort of confidence in the system that they've just participated in, that they mean something in this process, they're not just another number in a pile,” Jan Jaeger, RN, MBE, PhD, a fellow in the Center for Bioethics and Assistant Professor in the School of Nursing at the University of Pennsylvania, said in an interview.
But the principle of nonmaleficence may also dictate nondisclosure if disclosure would jeopardize the safety of the participant or a third party or would compromise the validity of the research results, she said.
Several considerations argue against disclosure. First, the participant may have given informed consent that results would not be communicated to him or her. After all, informed consent generally spells out that the point of a trial is to produce generalizable knowledge and not to benefit an individual research subject.
Beneficence/nonmaleficence would come into play if the results were not confirmed, validated, or clinically useful or could do more harm than good, Dr. Jaeger continued. “I think there has to be a great deal of caution in the kind of information that's disclosed to patients because [even though they may want to know,] that doesn't necessarily mean that the trial results or what they learn is going to make them feel good.”
Maurie Markman, MD, Vice President for Clinical Research at the University of Texas M. D. Anderson Cancer Center, has asked if study findings could not in any way benefit an individual participant but could result in considerable emotional distress, should nonmaleficence then “trump considerations of a research participant's ‘right to know.’”
Writing in Cancer (2006;106:1421–1424), he gave a hypothetical example of a patient who was now doing well but who three years earlier had received what was recently shown to be inferior treatment. In this case, there would be no way to turn back the clock to give the better therapy—for example, a post-surgical adjunctive regimen.
The point of a clinical trial is to see what works best. But in the case of a patient being included in an inferior treatment group, “What does that do to the patient's trust in the doctor?” Dr. Partridge asked. “That's not been studied very well.”
Dr. Markman proposed a possible compromise solution to balance the right of the patient to know with the principle of nonmaleficence. He suggested including in the IRB-approved consent forms a section that discusses why a participant may or may not want to choose to learn the study results. By having this discussion before trial entry, the participant can weigh the conflicting ethical and practical arguments and risks in general terms before anyone has any knowledge of the study outcome. However, the person should be made aware that any findings that could affect his or her current or future management will always be provided.
Another potential problem is the “cognitive dissonance” that may result when an individual patient is doing well but the group results were inferior or negative. Dr. Partridge, who noted that she has had mixed experience in terms of how well patients can understand clinical trial results, gave an example of one of her patients who just could not understand how the trial results could have come out negative since she herself was doing well.
In the review by Shalowitz and Miller, 10 studies assessed participants’ preferred methods for receiving results. As Fernandez et al found, study participants thought that it would be OK to receive positive or neutral results by mail, but 53% preferred to receive results with negative implications in person.
The authors also reviewed investigators’ attitudes toward communicating trial outcomes. In four of the five trials reviewed, a substantial majority of researchers supported communicating results to participants. However, cancer investigators identified the problems of cost and the time involved to prepare a lay summary and the difficulty of contacting study subjects as major barriers to communicating aggregate results. There was also a concern that communicating the results could bias future follow-ups.
Some investigators saw possible negative psychological consequences as a barrier to telling participants the results. But from their review of patients’ attitudes, Shalowitz and Miller said that despite the potential for negative psychological consequences, participants do still want the opportunity to receive research results and that investigators’ fear of psychological harm should therefore not be a reason to withhold research results, unless there are situations where there is clear evidence of a threat to participants’ safety.
Dr. Partridge and Eric Winer, MD, in an editorial accompanying the Fernandez et al paper, noted that in a study by her group, breast cancer patients who had been randomly selected to receive what turned out to be an inferior treatment were more dissatisfied with how the results were shared and reported more anxiety about their disease after learning the results. For such patients, Drs. Partridge and Winer recommended personal contact with a medical professional who can explain the results, answer questions, and address emotional distress. Mailing, e-mailing, or posting results to a web site are less desirable than direct contact.
Investigators need to assume that study participants will find out the results one way or another, through published articles, the news media in the case of large or important trials, general searching of the Internet, or via advocacy groups.
Drs. Partridge and Winer recommend, therefore, that researchers should include in the design of all Phase III trials a plan to share results, and the plan should form a two-step process as part of the informed consent. First, participants should be offered an opportunity to learn the results. Second, they should be offered various methods and media by which to learn the results, including in person, to allow for different learning styles and preferences.
These authors also recommend that investigators communicate trial results to study participants just before or at the time that the results are released to the media. They noted that some participants have expressed anger at learning the results through the media and not from their health care providers.
Dr. Jaeger, the bioethicist, agreed: “From where I sit, if the results are in, the study is closed, the data's been analyzed, the first people that probably should be informed are the participants,” she said.
Drs. Partridge and Winer recommended that sufficient evidence exists to incorporate disclosure of results into trial designs now, while continuing to do more research on the matter.
Shalowitz and Miller highlighted some areas for ongoing research. They determined that evidence is lacking on the cost of outreach to trial participants. They also suggested more research on whether communicating results will affect participants’ perception of biomedical research or influence them to enroll in future studies.
Dr. Partridge said the main idea of communicating results to participants is to treat them as partners in the research process since they fill the critical role of helping to improve care for future patients. Communication may also help to make the research endeavor more transparent and credible to the general public. “It may serve to improve rates of clinical trial accrual to have people more aware of the critical importance of clinical trials in terms of improving medical care in this country and in this world,” she said.
The research process has gotten a black eye at times when a trial has gone wrong and the result has sometimes been a loss of public confidence. For example, the University of Pennsylvania was accused of being less than transparent and of making some misrepresentations in the well-known1999 trial of gene therapy in which a research participant died four days after receiving an injection of an adenovirus vector.
“It doesn't take much,” Dr. Jaeger said. “It takes one research-related incident, one research-related death that makes it to the media…and there's public outcry.
“So, restoring confidence or trying to maintain the confidence of the public in science and in research is really very important. We need to do this research, we need to make drugs available and devices…and we need the public's support to do it.”
© 2010 Lippincott Williams & Wilkins, Inc.