Two recent reports remind their readers how difficult it can be to ensure that a person making a decision or expressing a preference about his or her medical care is doing so with proper information.
First, in the journal IRB: Ethics and Human Research, Kim et.al. ask the question, “Research Participants’ ‘Irrational’ Expectations: Common or Commonly Mismeasured?” (Article free to the public, registration required.) They cite the oft-raised concern of “therapeutic misconception” in clinical research: people who volunteer for certain clinical trials often misunderstand that the primary goal of the research may not be to demonstrate a treatment is effective, or they think that their enrollment has a better chance of benefitting them, or even a known chance, when that is not the case. Or, they do not grasp fundamental features of the research, including (or, better, especially) random assignment. In their study, Kim et.al. found what they suggest is evidence that people may understand randomization perfectly well, but not apply it to their situation or appreciate its meaning for them. They studied people with Parkinson’s disease enrolling in a randomized trial of gene therapy vs. a sham procedure, and they found that while study subjects could readily demonstrate that they knew what randomization was, and what it meant for the likelihood of being assigned to one study arm or the other, when they were asked what group they thought they personally would be in, many professed ignorance or some level of certainty that they would be assigned to the arm they preferred—viz., the treatment arm. The study suggests (and the authors say so) that this is perfectly understandable and reasonable human behavior, and that the discrepancy does not mean the subject has been misled or is not intelligent. Put another way, researchers should not look down on people who appear to overestimate their likelihood of personal benefit, or conclude that such an overestimate necessarily implies a deficiency in the informed consent process. It also can be read as a bit of fresh air for conscientious clinical researchers who worry that—or are besieged with accusations that—they are taking undue advantage of sick people who want to get better.
The second article, in the Journal of Clinical Oncology, comes from the “Video Images of Disease for Ethical Outcomes” consortium—“VIDEO” for short. (Everything has to have a slick acronym.) In this study (subscription required), investigators from four major cancer centers studied 150 people with advanced cancer and an expected survival of less than a year. They asked the cancer patients whether they would want cardiopulmonary rescuscitation (CPR) performed if their heart stopped beating. Everyone was read a scripted description of CPR with an estimate of its likelihood of success, then half the people were randomly assigned to also view a 3-minute video that showed simulated chest compressions on a mannequin, and images of an actual ventilated patient receiving medications, with the script also included via a narrator. The underlying premise was that words alone are insufficient to give people an appreciation of what CPR entails. People who saw the video were more likely to say “NO” to CPR. (The study did not include whether anyone actually had CPR performed.) Women, white people, and people who had higher health literacy (an uncommon trait in our society) were also more likely to say “NO.” Three-fourths of the people who saw the video said afterward that they were “very comfortable” watching it. The study authors describe steps they took to keep the video or narrative from being alarmist or unduly influencing, and state that pushing a patient one way or another is a real concern for decisions like whether to plan to do CPR at the end of life. (I couldn’t help but wonder what would happen if people with operable cancers were shown videos of their proposed surgery before deciding to have it—would they be less likely to consent to a procedure with known benefits?) But when it comes to informed consent issues in general, my impression is that audio-visual tools to aid the decision process are generally viewed as helpful, and we should anticipate greater use of them. It seems to me that context and equipoise are critical—context, in that any decision-aid tool should be used in the setting of a relationship of open communication and trust between patient and physician, and equipoise in the sense that the physician must have the patient’s welfare and free choice clearly in view, without competition from subordinate goals (like cost-control, public or private).