Risk Overload and Risk Misdirection in the Consent Process
Mark Hochhauser, Ph.D.
Golden Valley, Minnesota
Understanding risks is a key part of the informed consent process for prospective subjects. If they don't understand the trial's risks they can't make informed decisions about whether to participate in the trial or not. For example, research organizations typically explain to prospective subjects how important it is for them to understand clinical trial risks (italics added):
1) "Informed consent is the process of ensuring that people understand all of the benefits and risks involved in a clinical trial before deciding whether or not to participate."
2) "Before you agree to join a clinical trial, researchers must be sure that you understand everything about the study - risks, benefits, obligations, etc."
3) "All subjects in clinical trials must give informed consent, which is a process that helps provide full and complete understanding of the risks and benefits of the study before and after research participation begins."
Can anyone understand 160 risks?
As desirable as it is for prospective subjects to understand all of a trial's risks, can they realistically achieve "complete understanding?" Clinical trials involving drug combinations are complicated, with long and detailed risk descriptions. Based on four clinical trials that came to our IRB, I counted the risks for each trial. Because I chose to count as one risk those risks that were separated by a comma, semi-colon or colon, the total risks might have been even higher using a different counting strategy. For example: Is "nausea and/or vomiting" one risk or two? Is "skin ulcer, hives or welts" one risk or three? Is "loss of energy or fatigue" one risk or two?
Trial #1 had four drug combinations, Drug A had 26 risks, Drug B had 34, Drug C had 45 and Drug D had 50 = 155 risks. Trial #2 used six drugs: Drug A had 42, Drug B had 36, Drug C had 18, Drug D had 20, Drug E had 19 and Drug F had 27 = 160 risks. Trial #3 used three drugs: Drug A had 31 risks, Drug B had 17 and C had 34 = 82 risks. Trial #4 used 3 drugs: Drug A had 21 side effects, Drug B had 10 "important" side effects and Drug C had 43 side effects = 74 side effects.
Signing "I understand…"
In the signature section of all three trials, prospective subjects signed under the sentence "I have read this consent form and understand the possible risks and benefits of my participation in this study." While agreement with that statement may make subjects, researchers and sponsors believe that subjects actually do understand the possible risks of the study, it's unlikely that subjects could remember many of the risks or their likelihood, severity or importance if later asked.
Incomprehensible Risk Categories
Risk categories present another problem. That is, it's one thing to count the number of risks, it's another to place trial risks into meaningful risk categories. Table #1 shows the risk categories for these four clinical trials.
Risk categories in Table #1 aren't helpful because there's no standardization of risk language; #1) likelihood, #2) frequency and commonness, #3) severity + commonness, #4) mixed frequency and importance will have different meanings for different researchers and subjects. Plus, it's not clear if "risks" and "side effects" have the same meaning for prospective subjects; are "side effects" less dangerous than "risks?"
Trial #1 categorized risks as likely, less likely, but serious, and rare, but those terms are meaningless without some kind of quantitative explanation—how "likely" is "likely?" Trial #2 tries to be quantitative but without much success. Not all subjects know that < and > means "less than" and "greater than." Trial #3 combines severity and commonness in categories that can only be described as incomprehensible. Trial #4 used a good combination of quantitative and quality risks for two of the drugs, but the third drug included 10 "important side effects" with no quantitative or qualitative descriptors.
How can subjects be expected to "completely understand" 74 to 160 risks in categories that aren't of much help? Ideally, risks should be described as the percentage of subjects who might experience those risks as done in Trial #4 for two of the three drugs. Unfortunately, such data isn't always available for the experimental drugs being tested, although sponsors may have that kind of data from similar drugs already studied or approved by the FDA.
A recent United Kingdom study1 looked at people's risk perceptions for a hypothetical clinical trial by giving them either a verbal risk description (common, uncommon, rare) or a combined risk description (common = 1% - 19%; uncommon = .1% - 1%; rare – 0.1% - .1%) based on European Union (EU) guidelines. As shown in Table #2, subjects given only the verbal risk description rated their likelihood of experiencing a risk 3 times higher than those receiving a combined risk description. Note that even those who were given the actual EU risk description still overestimated the likelihood that they would experience risks! That is, even though they were told that a common risk would affect 1% - 10% of subjects, they believed that their likelihood of experiencing a common risk would be 19%. These undefined Verbal Risk Descriptions create risk misdirection in which subjects perceive risks to be more common than they actually are.
Compared to those getting the Combined Risk Description, those getting Verbal Risk Description were more likely to perceive the study as having a risk to health, but were less likely to be satisfied with the information, saw less possible benefit to their health, and had less intent to participate, listing "lack of adequate information" as a reason for not participating in the trial.
Combining risk overload with ambiguous risk categories means that prospective subjects will overestimate clinical trial risks—which may cause them to decide not to participate in that trial. More accurate portrayal of clinical trial risks isn't just an informed consent communication issue, but a recruiting issue. If you're having a hard time recruiting subjects, perhaps it is because subjects perceive more risks in the clinical trial than actually exist.