Flexibility is needed when reviewing pragmatic clinical trials, which typically enroll a broader population of patients and might need more adaptable informed consent than traditional clinical trials.
Researchers have raised questions about how pragmatic trials should be regulated and what IRBs should do to protect participants, but not discourage this type of research.
Here are some suggestions for how IRBs can review pragmatic clinical trials:
• Assess research risk. One of the main ideas behind assessing risk in pragmatic clinical trials is this question: “If the two interventions being compared are within the standard of care or are accepted practice, is that a good way to determine the research risk of these trials?” says Scott Kim, MD, PhD, senior investigator in the department of bioethics at the NIH.
An IRB’s answer to that question will determine what happens with informed consent.
“To waive or alter the traditional requirements, the regulations require the study to be declared minimal risk,” Kim explains. “There is a lot of debate about whether these studies are minimal risk.”
Unless an IRB determines that the clinical trial is minimal risk, researchers cannot significantly change the way they conduct informed consent, Kim says.
“I think it’s quite possible that some pragmatic trials are, indeed, minimal risk,” Kim says. “And for those studies, the regulations allow, if justified, altered informed consent.”
To determine risk, IRBs need to recognize the pragmatic trial’s purpose, says Mitchell Parrish, JD, RAC, CIP, vice president of legal and regulatory affairs at Advarra. “Typically, with a pragmatic trial, you’re comparing two treatments that are approved therapies, so it truly is standard,” Parrish explains. “But it’s a controlled trial, so the physician is not selecting which product is right for you.”
Patients are randomized to either treatment, but both are the standard of care, he adds.
IRBs must decide whether this randomization to two different treatments poses any additional risk — even if each is standard care.
“We have to face the fact that using too general a rule to determine the risk of a pragmatic trial, such as the idea that both arms are used in [usual] practice, making the trial minimal risk, is utterly dangerous,” Kim says. “IRBs should continue to do specific evaluation of the currently existing evidence about risk rather than just relying on a general rule like that.”
This is why it is important for researchers and IRBs to be clear on what usual practice is for purposes of this research, he adds.
• Define and explain usual practice in pragmatic trials. What researchers or IRB members believe is accepted or usual practice might not be usual practice, Kim notes.
“First, you have to prove that the things you are comparing in a trial are truly accepted in practice,” he explains. Then, IRBs should consider how the study has been altered to introduce a research element.
The most challenging part of a pragmatic trial might be to explain to participants their relative risks in a study that compares two treatments that have long been used for their condition.
For example, a pragmatic trial might compare treatment A to treatment B. Both have been used to treat patients with a particular condition, and both are considered accepted practice. Suppose small studies have compared the two treatments, and three of the four studies show that treatment A works better than treatment B. This suggests a more thorough comparison trial would be helpful, Kim explains.
“Here’s the question: Can the researcher assume that treatments A and B are equivalent? No, they can’t, because if they knew the treatments were equivalent, they wouldn’t be doing this study,” he says. “And you have preliminary, but not decisive, evidence that one treatment might be better than the other.”
In that situation, it might be inaccurate to tell potential study participants that if they entered the study, their treatment would be just like going to their doctor with no extra risk, Kim says.
“It is more reasonable to say, ‘Well, you were taking A, and were assigned to B, so there is a chance you might be worse off than if you hadn’t done the study,” Kim explains. “It’s complicated.”
It is up to IRBs and researchers to anticipate research risk and legitimate potential outcomes, looking at these two factors:
- Even though both treatments in a pragmatic trial are used and accepted, it is inaccurate to say there is no risk to an individual entering the study because their course of treatment might be altered by their clinical trial participation, Kim says.
- Each study and its risks should be evaluated by the best current evidence that exists, as well as for the potential different outcomes a trial participant might face if entered in the study, he says.
When it comes to defining usual practice, this type of consideration is ethically more balanced than just saying that if both treatments in a study are accepted by doctors, then there is no risk in participating in the study, Kim adds.
• Consider the study’s vulnerable population. For instance, IRBs should discuss whether the enrollment of vulnerable populations in the study constitutes greater risk in a study that otherwise might be considered minimal risk.
“IRBs should understand the rationale of why there might need to be more vulnerable populations involved in the trial,” Parrish adds. “Once they recognize that rationale, the next thing is to understand, as an IRB, what they can actually offer as an additional protection.”
For example, a pragmatic clinical trial that is assessing treatment for Alzheimer’s disease likely will enroll people who lose some cognitive ability during the length of the trial.
“You might say this population will at some point lose the cognitive ability to do informed consent,” Parrish says. “What can we do to minimize this?”
An IRB might require everyone enrolled in the study to create a plan for if they lose the capacity to decide whether they want to remain in the trial, he adds.
Or, in another example, the study could be about an antianxiety medication for pregnant and postpartum women. IRBs assessing this study should ask the following questions:
- What do these participants need to understand, and how is it explained in the informed consent document?
- Does the informed consent offer suggestions for participants about whom they can see if they experience postpartum or pregnancy issues?
- Could the sponsor provide more information about the site’s plan of care for individuals, including counseling and who to contact if there is a problem?
“IRBs should think about it and be ready to discuss their ideas,” Parrish says.
• View informed consent more flexibly. There is no settled agreement on how informed consent might be handled, altered, or waived when it comes to pragmatic clinical trials.
“My impression from reading the literature is that the practice is unsettled and it varies by IRBs and other situations,” Kim says. “Even if a study is not minimal risk, there are probably studies — from a purely ethical point of view — where it would be permissible to use an informed consent process that is abbreviated and altered from the current practice,” Kim says.
There also have been some pragmatic trials deemed minimal risk, and informed consent procedures were altered, he adds.
Sometimes, IRBs approve the use of deferred consent, giving researchers flexibility to start a pragmatic trial that compares two existing drugs following a health emergency, such as a heart attack, without starting the informed consent process until treatment is underway. At that point, the patients can decide whether they wish to continue.
The main point is that IRBs should make these decisions case by case.
“It’s not good practice to use the reasoning that if you are testing two things already in use in clinics then that automatically makes the study minimal risk,” Kim says. “That reasoning is increasingly being put forward in the literature by people who do clinical trials, and I think that is not a good idea.”
Informed consent in pragmatic clinical trials is tricky, Parrish says.
“If you have someone enrolled in a trial, you have to get their informed consent, but this will screen out some individuals because not everyone will want to participate,” he explains. “And that negates the real-world goal of getting a real-world population, so it’s a big issue that has to be resolved.”
The debate over informed consent for pragmatic trials will continue, although some studies are employing flexible consent strategies, he notes.
For example, some pragmatic trials use cluster randomization, Parrish says. Cluster randomization can be used to study whether to treat an entire group, such as studying immunization, and it can be used to study drug effectiveness. Instead of analyzing individual results, data are analyzed for the entire cluster. (More information is available at: http://bit.ly/2SZrNBV.)
Consent is different in cluster randomization because the trial might be set up to give one certain conventional treatment to every patient with a particular disease who enters hospital A. Then, the same type of patients who visit hospital B would receive a different conventional treatment. The two treatments are compared, but there is no individual randomization, Parrish explains.
“People will have a notification component for individuals who go to this hospital,” he adds. “There still is an issue of informed consent, and IRBs will have to think about the issue of people individually consenting, or doing something more generic.”