The trusted source for
healthcare information and
Comparative effectiveness research (CER) has the potential to improve outcomes for individual patients as well as groups of patients, as well as lower costs.1
"I am encouraged to see that it is gaining traction more and more," says Ryan Spellecy, PhD, associate professor of bioethics at Medical College of Wisconsin in Milwaukee. "The key ethical concern is how to best convey to patients that this is research."
Patients becoming research subjects after enrolling in a research study when coming to a medical center or doctor’s office for clinical care is nothing new, but it is common in CER to compare two treatments that are commonly used.
"Since both treatments could be considered standard of care, or perhaps within the range of standard of care, it is that much harder to distinguish research from clinical care," says Spellecy.
In clinical care, a physician prescribes a course of treatment based on the best evidence available. In CER, a subject might be randomized to one of two treatment arms. "There is now good evidence showing that when it is not known which treatment is better or worse, randomization does not increase risk," says Spellecy.
However, he cautions, the research subject still needs to know that he or she is participating in research, and that the goals of research are different from the goals of clinical care.
"I worry that many voices in the current debate over CER focus primarily on risk and ignore respect for persons and the right of research subjects to informed consent," says Spellecy. "Informed consent matters, regardless of risk."
Spellecy says the challenge is how to best present the research subject with enough information, in the proper manner, to decide whether or not to participate in the research.
There is a risk that patients may become overwhelmed with the length of the consent form. "Unfortunately, consent forms are getting longer and longer, and there comes a point at which, in the name of completeness, we sacrifice comprehension," says Spellecy.
There have been some novel proposals to include in CER consent forms only the information that is unique to the research study.
"That is, if the standard of care information is already covered in clinical conversations, the consent form and conversation should focus on the risks, benefits, and other aspects that are unique to the research," says Spellecy.
Chris Feudtner, MD, PhD, MPH, the Steven D. Handler endowed chair of medical ethics at The Children’s Hospital of Philadelphia, says that thinking through the appropriate informed consent process for CER, and obtaining it, is more complicated than for a more routine clinical trial, since CER compares two potentially useful therapies to each other and not to a placebo.
"But I wouldn’t think of the standards for CER informed consent any differently than the standards we hold ourselves to in other clinical studies," says Feudtner.
While in some cases CER poses minimal risk and informed consent is potentially waivable, says Feudtner, in other cases, fully obtained informed consent is "absolutely essential," just like with any other clinical trial.
"In other words, one size fits all’ will not be suitable across the range of what a CER study might investigate," says Feudtner. "We have already have thought through, in human subjects protections, conditions in which informed consent can be waived, and I think that guidance still applies."
Feudtner adds, "most of the hullabaloo has occurred because people are not fully aware of all of the clinical decisions that are made in routine, and quite standard, clinical practice, based on a provider’s personal habits or a hospital’s clinical care pathways. These inevitably entail risks, but rarely get fully described to families as such."
In a sense, he argues, CER is simply revealing the deficiencies of informing people about the prognosis and risks of treatment and potential benefits in regular clinical practice.
One circulating argument is that CER will grind to a halt if providers go through the process of obtaining consent in all cases, because many participants would refuse to participate.
"I would never accept the argument that CER is so important that consent always has to be waived," says Feudtner. "That is a bad argument at many levels."
In some cases, however, CER clearly imposes no significant marginal increased risk of a bad outcome for a patient, which is one of the hallmarks of a situation in which consent could be waived.
If the risks posed by the two treatments under study are roughly the same, and researchers are trying to sort out whether one treatment might be more effective than the other, or the benefits might be greater, there is an argument for waiving consent, Feudtner says.
"I don’t think CER should get a free pass — it’s not so important that we have to facilitate it by uniformly waiving consent," Feudtner says. "But on the other hand, I don’t think that CER truly warrants more scrutiny than other clinical trials."
A broader solution for how to appropriately manage the process of informed consent for minimal, marginal risk CER is to have a global consent at the time of entering a hospital. "This would alert people that treatments are continually compared to determine if one is more effective than another, and that consent will be sought if it is believed that there is a difference in the risk or benefit," says Feudtner. The consent could also state that most of the time, treatments are well-established, and in those cases, patients will not be approached for consent.
"There would be a transparency in doing this," says Feudtner. "We are letting people know that we do CER and QI work, and that they can ask us questions about that."
Feudtner says that this is sufficient to protect ethical standards, which include both the protection of human subjects and the prudent improvement of clinical practice to improve patient outcomes.
A number of issues have been identified with CER that require careful ethics and regulatory consideration, says Robert M Califf, MD, vice chancellor of clinical and translational research at Duke University in Durham, NC. He says these are several of the most important:
• How should it be determined that a CER protocol adds minimal risk to what is encountered in usual clinical care?
• If consent is modified to make it simpler and more feasible for large studies, how should that be done?
The NIH Health Care Systems Research Collaboratory and the Patient Centered Outcomes Research Network provide forums to sort through these issues. "They are far from settled," notes Califf.
Califf says there is no short, easy answer to the question of how trials that randomize to two or more "standard of care" interventions should be designed and reviewed, other than "carefully and with major expertise."
Involving patients and their families will likely make the studies both more relevant and more feasible, according to Califf.
"We are embarking on several major surveys of the American population very soon to find out what they think as they become better informed about the uncertainties of clinical practice and the promise of developing better evidence to improve clinical care," he reports.