Keep it simple: Make consent documents easy as possible
Readability improves, but many forms still too complex
Although some researchers are doing a better job of presenting informed consent documents in language that subjects can understand, too many still rely on jargon and overly technical language that can stymie people with limited literacy.
That’s the view from those who study the language used in consent documents.
"The average consent form that I’ve looked at very often scores at grade 16 [reading] level — completely inappropriate," says Rima Rudd, ScD, director of educational programs at the Department of Health and Social Behavior, Harvard School of Public Health in Boston.
Rudd says even the target goal of an eighth-grade reading level for documents aimed at a general population may be too high. She says a survey of U.S. adult literacy conducted in the early 1990s showed that the average American adult has what she would term limited literacy skills, particularly in the area of document literacy — the skill required to read a form.
"Document reading is more difficult than other kinds of reading," she says. "You may not know a word in prose, but you can kind of figure out the meaning of the word in a sentence. But if the word is just on a list, just on a chart, it’s more difficult to understand."
Rudd says the average adult probably reads at somewhere between a seventh- and eighth-grade level. "That’s the average," she says. "And then people with more limited literacy skills fall below that."
This has serious implications for the informed consent process, says Holly Taylor, MPH, PhD, assistant research professor in the Department of Health Policy and Management, Bloomberg School of Public Health and Berman Bioethics Institute, Johns Hopkins University in Baltimore.
First and foremost, people who don’t understand the consent form they sign may not fully understand the project they’re volunteering to join and the risks it entails.
In addition, participants who don’t understand the project fully may jeopardize the study itself. "If they don’t understand that they’re required to adhere to a particular prescription or to come into the clinic every week for three weeks for a follow-up checkup, then your trial may not go as well as it should," Taylor says.
Finally, people with limited literacy may simply opt out of clinical trials altogether because they’re intimidated by the complexity of the consent forms — excluding them from any potential benefit. "That’s a bad thing, too," Taylor says.
The Office for Human Research Protections advises that informed consent documents be written in "lay language, (i.e., understandable to the people being asked to participate)."
Better readability lowers anxiety
Studies have shown that improving the readability of informed consent documents has measurable effects on the subject population participating in a clinical trial.
Peter Raich, MD, a senior scientist at AMC Cancer Research Center in Denver, surveyed breast cancer patients using two informed consent documents — a control document used at some 20 institutions and one written at between a seventh- and eighth-grade reading level. His survey was piggybacked on an existing breast cancer treatment trial. A week after subjects completed the forms, they were asked about their knowledge of the trial and their decisions about participating.
Raich said his study found slight improvements in the participation and knowledge of subjects who had used the modified form, but no statistically significant difference.
However, there was significantly less anxiety among subjects who used the lower-literacy form, as well as greater satisfaction with both the form and the entire informed consent process.
Raich said improving the readability in informed consent documents didn’t require sacrificing content. (See chart below.)
How Do You Write Clearly?
|Source: National Institutes of Health. Clear and to the Point: Guidelines for Using Plain Language at NIH. Bethesda, MD; 2001. Web site: www1.od.nih.gov/execsec/guidelines.htm.|
"Instead of a 12th- to 16th-grade reading level, we were down to seventh to eighth grade, mostly by using simpler words, shorter sentences," he says.
Formatting changes also helped — larger print, more white space.
"We divided the content into categories. We did it in a question-and-answer format. When we had to use medical terminology, we tried to explain that. We included some graphics, some pictures, some symbols."
Reich says the regimen subjects must follow was presented in a calendar format that many people found helpful. Interestingly, all the changes in the lower-literacy version made that document twice as long as the standard document. But when asked, fewer subjects reported that the modified document was too long. More, in fact, said the standard version was too lengthy.
"[Subjects receiving the lower-literacy form] perceived it was not as onerous a task wading through these pages," Raich says. "It had a lot to do with the perception of how easy it was to go through it."
He believes that the reason there was little improvement in actual participation was that many cancer patients often come into clinical trials with their minds already made up to join.
"It’s a very critical situation; they feel it’s life-threatening, they know they have to do something," he says. "So I think there’s not as much decision on the part of the patient as far as weighing pros and cons."
A second study now under way looks at modified informed consent documents for preventative regimens for prostate cancer. Raich says this may give a better idea of whether a lower literacy document helps form a decision to participate.
IRBs enforce standards
Taylor says she’s seen improvement in the consent forms she reviews as a member of an IRB at Johns Hopkins School of Medicine. But she says it’s still a challenge for many investigators to tailor their writing to a different audience.
"The people writing these things, in general, are MD- or PhD-trained folks who are not asked on a regular basis to express themselves in eighth-grade language. They’re just not trained in writing at that level."
She and Rudd say there are steps IRBs can take to help improve the consent documents that come before them:
• Provide help. The IRB on which Taylor serves employs an informed consent specialist, whose primary responsibility is to review consent forms and consult with investigators about changes to improve readability.
Eventually, the specialist will be available early in the process, so that she can help shape informed consent documents before they ever get to the IRB.
Taylor says that ideally, that person would be skilled in health behavior and health communication in order to help translate complex medical information into format and language that allows participants to better understand it.
• Enforce standards. IRBs should routinely require investigators to run informed consent documents through one of the many available reading level assessments, such as one created by Edward Fry, formerly of the Rutgers University Reading Center, and the SMOG readability formula, which uses counts of polysyllabic words to estimate the grade level of a document.
"They must submit a reading level assessment and tell what assessment form they use," Rudd says. "You can’t compare the SMOG with the Fry — they’re slightly different, they measure in slightly different ways. So the IRB has to say: You must use the SMOG or the Fry. They have to be very clear so there’s consistency."
The IRB also should enforce format requirements such as ragged-right justification, which Rudd says is easier for the eye to follow, as well as sufficient white space and font size.
"I’m not going to read a consent form in 8-point font," Taylor says. "It’s just ridiculous to ask me to do that. And it’s even more ridiculous to ask somebody to enroll in a macular degeneration trial in 8-point font."
In order to enforce easier readability standards, IRBs may have to take on some of their own practices, Taylor says. When studying the boilerplate language required by IRBs of its investigators, she and her colleagues found that language to be complex.
"We should make sure that the standard template the institution requires about injury or privacy should be at the same reading level," she says. "Basically, that’s not the case — that boilerplate in general is at about a 10th-grade level."
• Know the subject population. Taylor says it’s not always necessary, or even appropriate, for an informed consent document to be readable by every conceivable person. But an investigator should know if he or she is recruiting from a population that’s likely to have literacy problems, and respond accordingly.
"It’s the investigator’s responsibility to know who they are going to be recruiting and how many people are likely to be of a lower reading level," she says. "It’s their responsibility to draft a form and a process that meet the lowest common denominator of the folks that they’re recruiting."
She argues against the idea of quizzing people on their reading skills, because it puts illiterate subjects in the embarrassing position of having to admit it.
"You need to find a way to engage the person to get a sense of their reading ability, if you can," she says. "What’s their education — some set of questions that might help you decide whether this person might be illiterate."
• Provide alternatives for lower-literacy subjects. Because low literacy can be a source of shame, many subjects who can’t read often construct elaborate deceptions to mask that fact.
Taylor says that occasionally, as she goes through a consent form with a subject, she is asked to read it aloud.
"A person might say, Oh, I left my reading glasses in the car. Would you mind reading it out loud?’ What I don’t know is, Is it that they can’t read, and this is their way of coping with a medical system that expects every person to be able to read?’" she says.
Reading over a form with a subject, giving him or her ample opportunity to ask questions or review sections, can be a way of helping a lower-literacy subject, she says.
But it can be a time-consuming process, and Rudd warns that the language still has to be understandable.
"Having it read out loud, if it’s the same jargonized speech and scientific terms, is not a help. It’s not a hindrance, but it doesn’t mean that people will necessarily understand it," Rudd says.
Similarly, she says, investigators can employ technology such as video to help show subjects what they’re being asked to do.
"New technology can definitely help out, but we need to have that same respectful attention to language," Rudd says.
Internet tool for improving readability
The Health Literacy Studies program in the Harvard School of Public Health is a research program of the National Center for the Study of Adult Learning and Literacy.
The program’s web site provides a number of resources for IRBs and investigators dealing with readability:
The organization’s web site is www.hsph.harvard.edu/healthliteracy.