The trusted source for
healthcare information and
Electronic grammar check doesn’t ensure consent process effectiveness
Experts offer suggestions for improving informed consent process
The informed consent process has become much more complicated and burdened with legalistic language in recent years, and there is little research available to show what the impact is on subjects’ understanding of their role in studies, experts say.
Some of the studies about informed consent are three and four decades old, and they were conducted with hypothetical subjects rather than people who were enrolled in clinical trials, notes Nancy Kass, ScD, a professor of bioethics and public health at the Bloomberg School of Public Health and Johns Hopkins Bioethics Institute at Johns Hopkins University in Baltimore. Kass spoke about writing effective consent documents at the 2005 annual HRPP conference on ethics and trust across boundaries, sponsored by Public Responsibility in Medicine and Research (PRIM&R) and Applied Research Ethics National Association (ARENA), held Dec. 3-6, 2005, in Boston.
While it’s good news that investigators no longer question the need for informed consent, as might have been the case decades ago, the down side is that their informed consent process typically is guided by regulations, Kass says. "What the regulations tell them to do is disclose to participants a lot of information about the trial," Kass says. "The regulations don’t tell them to find out if participants understand what was said," she adds. "We’ve interviewed research investigators who said, I wish we spent more time figuring out what people understand.’"
When asked how many investigators do some sort of research into what subjects understand, only 16% said they did, Kass says. "This is despite the fact that the vast majority of researchers think we should to more than we do," she notes.
One clever strategy for assessing a subject’s understanding is to eliminate the most used and obvious question of "Do you have any questions?" and replace it with specific questions about the informed consent information discussed, Kass suggests.
"It’s common for researchers to give people a lot of information and then ask, Do you have any questions?’ and participants say, No, I don’t,’ and that’s it," Kass says.
A better approach would be to ask questions, such as these:
"This way you will find out what people understand and what they don’t understand, and you can correct their misunderstandings," Kass says.
Clinical trial sites and IRBs also could put more effort into making informed consent forms readable, and this means going beyond the word processing program’s readability scores, another expert suggests.
"It’s easy to look at an automated Microsoft Word readability score, but it’s easy to fool that system," says Michael Paasche-Orlow, MD, MPH, assistant professor of medicine in the section of general internal medicine, department of medicine, Boston University School of Medicine. Paasche-Orlow also spoke about the consent process at the recent PRIM&R conference.
The MS Word readability tool is based on the Flesch Reading Ease score and Flesch-Kincaid Grade Level index, which score documents according to sentence structure and complexity of language.
The Flesch-Kincaid system was developed to adapt Navy training manuals many years ago, Paasche-Orlow notes. "Their equation has nothing to do with the frequency of words," he says.
Also, the way MS Word uses the system it won’t pronounce any document to be above a 12th-grade reading level, Paasche-Orlow says. This becomes important when an informed consent is checked on a paragraph-by-paragraph basis, because while the document as a whole might have an eighth grade reading level, individual paragraphs could be listed at a 12th grade reading level, he says.
"It’s hard to say what is the appropriate standard and what is the reading level target," he says. "The most common target in the U.S. is an eighth grade reading level, but that’s not based on any empirical evaluation of what the need is."
Also, while an eighth grade reading level might be the appropriate target for some protocols, it might be too high of a reading level for others, Paasche-Orlow says.
"I’ve seen a few IRBs that set targets as low as fifth grade reading level, but the federal standards don’t specify readability levels, they only say the document should be readable by the average subject," he says.
The best approach to assessing an informed consent document’s readability level is to have lay people read it and note the problem areas, he advises. "I recommend that everybody have their form be read by a non-scientist and maybe even somebody who is legitimately in the target audience for their study," Paasche-Orlow says. "They should even have their children read it if they have grade school children who might be able to read it for them."
While automatic readability tools will show investigators where they have gone wrong in writing an informed consent document, they won’t show what is working well and what is not working as well as it could be, Paasche-Orlow says.
Even if investigators do not have non-scientists read through the consent form, they could improve it by concentrating on the areas that are most likely to confuse research subjects.
"Specific concepts in many trials that are harder than others to understand include issues of randomization, therapy misconception, compensation for injury, and things that are easily confused," Paasche-Orlow says.
Research into the informed consent process provides clues about what research participants are most likely to understand and what they are most likely to misunderstand, Kass says.
"For example, they will generally know how often they have to come to clinic and which medications they need to take and how many times a day they have to take them," Kass explains. "What they seem to have the hardest time understanding, first of all, is that the research may not help them clinically."
Research subjects often assume that if their doctor referred them to a trial then it’s because there is helpful medicine given in the trial, Kass says. "And while sometimes that is absolutely true, at other times it may be the trial is testing medicine that may not be helpful, or it may be testing against a placebo," Kass adds.
Clinical trial subjects also have trouble understanding randomization, Kass notes. "Even if they are told very explicitly in a consent process that half the people will get one thing and half get another, and this is decided by flipping a coin, when one interviews patients in a research study, they often have elaborate explanations for why they believe they got the drug," Kass says.
Subjects sometimes say, "My doctor liked me," and they’ll explain how they received the drug instead of the placebo, even if they were told the trial was randomized, Kass adds. "Subjects have a common belief that the study will help them personally, and sometimes that’s a misconception, and sometimes that’s accurate," Kass says.
While patients typically understand that they don’t have to join a study, they have more difficulty understanding that they can drop out of it once they’ve enrolled, Kass says. "It’s more likely they understand they have a choice about enrollment than they understand they have a choice to drop out," she says.
These misunderstandings could be eliminated if investigators would encourage more discussion during the informed consent process and rely less on the consent form, Kass says. "I think they should sit down with someone and say, I would like to talk with you about a research study you might be interested in joining,’" Kass says. "Then they would talk with them about the study and not necessarily look at the consent form."
The investigator or clinical trial coordinator’s language would be informal and possibly include these comments:
The person providing informed consent would emphasize the more difficult parts, instead of treating each section as if it was of equal importance, Kass says. "And at the end of the discussion if it looks like the participant wants to join, they would say, If there are no more questions then I need you to sign this piece of paper that documents that we had this discussion,’" Kass says. "And the person might bring the document home to think about it, depending on what kind of research is involved."