SAN FRANCISCOMany patients and some physicians do not understand the purpose and fundamental concepts of clinical trials, according to a study of informed consent.
In one question, participating physicians displayed less understanding about the purpose of clinical trials than did patients, reported Steven Joffe, MD, of Dana-Farber Cancer Institute in Boston. More than half (54%) of practitioners disagreed with the statement, "The main reason clinical trials are done is to improve the treatment of future cancer patients." Among patients, 75% agreed that the statement is indeed correct.
On other questions, however, physicians displayed a clearer understanding than did patients about standard vs nonstandard therapy and incremental risk. Only 26% of patients but 86% of physicians disagreed with the statement, "All the treatments and procedures in my clinical trial are standard for my type of cancer." Similarly, only 30% of patients but 82% of physicians disagreed with the statement, "The treatment being researched in my clinical trial has been proven to be the best."
The gap narrowed slightly with the statement, "Compared with standard treatments, my trial does not carry any additional risks or discomforts"; 37% of patients and 71% of physicians disagreed. Both did better with, "There may not be direct medical benefit to me by my participation in this clinical trial"; 71% of patients agreed, as did 95% of physicians.
Another multigroup investigation found that giving easy-to-read, informed consent documents to patients considering clinical trials reduced their anxiety and increased their satisfaction in comparison to traditional documents. Peter C. Raich, MD, of AMC Cancer Research Center in Denver reported that the simplified paperwork did not, however, increase patients’ comprehension. Many patients did not understand what was written in either the simplified or standard format with respect to the privacy of medical records and degree of personal benefit to expect from a trial.
In presenting the two studies, Dr. Joffe and Dr. Raich created a complex picture of how health care providers might improve the informed consent process. For ethicist Bernard Lo, MD, of the University of California, San Francisco, they also raised the question of whether informed consent is realistic as a universal model for all cancer patients making the emotional decision of whether or not to enter a clinical trial.
"Informed consent is primarily a rational, decision-making model. This model may not hold. There may be serious limitations," Dr. Lo said, raising such factors as hope, altruism, and different cultural expectations of physicians in a discussion of the two investigations. "These are inescapable features of how we as human beings make decisions, think, and respond to grave situations like cancer," he said.
For the study reported by Dr. Joffe, researchers created a generic standardized questionnaire, called the Quality of Informed Consent (QuIC), to assess whether patients really understood the documents they were signing. "One important problem has been called the ‘Therapeutic Misconception,’ the mistaken belief among research subjects that every aspect of the research study was designed to benefit them directly," Dr. Joffe said.
The survey was mailed from June 1999 to January 2000 to 287 consecutive adult cancer patients within 14 days of their signing an informed consent at Dana-Farber Cancer Institute, Massachusetts General Hospital, or Brigham and Women’s Hospital, all in Boston. The 207 subjects who responded represented 73 clinical trials and were seen by 77 consenting physicians who were also asked to complete the questionnaire.
Patient characteristics were as follows:
the mean age of the patients was 55 years;
55% were female;
91% were white;
95% spoke only English at home;
53% were college educated;
half were in phase II trials, 24% in phase I trials, and 26% in phase III trials;
more than half (53%) had relapsed or had progressive disease; and
65% had a 10% or less chance of 5-year, disease-free survival.
Talking It Over
Discussion with the investigating physician was an important part of the consent process, according to Dr. Joffe. He noted that 48% of patients described the discussion as lasting 1 hour or more; 84% had an adult friend or relative present; and 37% reported a nurse being present. Most patients (72%) reported signing the consent form at a second or subsequent discussion rather than at the first discussion with the provider.
About a third (31%) received simplified forms based on a consent template that the National Cancer Institute issued in 1998, and 84% of all patients reported reading the consent form carefully. Dr. Joffe reported what he called a disconnect between information received and information used, however: "73% said the consent form was an important source of information," he said, "but only 37% found the consent form to be important to the decision."
Overall, 90% of patients reported being satisfied with the informed consent process, and most considered themselves to be well informed. "Self-assessment scores correlated weakly but significantly with knowledge scores, suggesting that subjects have a limited ability to assess their own understanding," Dr. Joffe said. College-educated people and those who spoke English at home scored best on the knowledge questions, he said. Factors associated with a smaller increment in scores included the NCI template, patients taking time to sign, presence of a nurse, and careful reading of the documents.
Conflicting Physician Roles
Asked why so many practitioners gave answers suggesting therapeutic misconceptions, Dr. Joffe cited the conflicting roles played by physicians who do research. "On one hand we have clinical obligations to our patients as physicians that are very deeply rooted in us from our training. On the other hand when we become physician investigators, we are asked to represent at the clinical trial objectives which are to create generalizeable knowledge," he said. "One way to resolve this conflict is to make it go away by believing that everything about the research is actually in the interest of the present subjects, who are there in the room with us."
Two implications of the study, he concluded, are that physicians need to do more to educate patients and that "we may actually need to educate ourselves about these issues as well."
The study reported by Dr. Raich took place at 44 institutions from March 1998 through January 2000, with participation by three cooperative groupsthe Eastern Cooperative Oncology Group, the North Central Cancer Treatment Group, and the Cancer and Leukemia Group B. The 207 patients who were assessed came from one metastatic lung cancer study and two adjuvant breast cancer studies.
Randomization was by site. At 24 sites, 129 patients received standard consent forms written at the 12th to 14th grade level. A version written at the 7th to 8th grade level was given to 78 patients at 20 sites. "Modifications included changes in wording, in font size, in layout, but not in the content," Dr. Raich said.
The mean age in both groups was 53 years. More than 90% of participants were white women, the majority of whom had at least some college education. The mean reading level was 9th grade or higher. All participating patients were interviewed by telephone one week after the consent process.
Dr. Raich said that the overall difference in comprehension between the two groups was not significant. Comprehension was 72% for the easy-to-read form vs 69% for the standard form. Fewer subjects than expected in both arms answered questions correctly, according to Dr. Raich. For example, only half of lung cancer patients acknowledged that the treatment could not cure their cancer and less than two thirds of breast cancer patients could identify the duration of their treatment or how many chemotherapy agents they were to receive.
The simplified form did not make a difference in accruals; 68% of the standard group and 75% of those who received the simplified form enrolled in a trial. The easy-to-read form had its biggest effect on patient anxiety and satisfaction. "Satisfaction with length actually favored the easy-to-read version, and this was interesting because the easy-to-read version had essentially twice as many pages as the standard version," Dr. Raich said.
Noting gaps in understanding among patients in both groups, he concluded, "The information provided in the consent statement and by the consent process does not meet all the needs of patients participating in cancer treatment trials. Additional research should be done to address these needs, especially in individuals with low reading skills and from diverse cultures."