|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Randomized, Controlled Trial of an Easy-to-Read Informed Consent Statement for Clinical Trial Participation: A Study of the Eastern Cooperative Oncology Group
From the Department of Community Medicine, School of Medicine, West Virginia University, Morgantown, WV; Dana Farber Cancer Institute and Harvard School of Public Health, Boston, MA; AMC Cancer Research Center and Colorado Health Outcomes Program, University of Colorado Health Sciences Center, Denver, CO; Prevention Research Center, University of Kentucky, Lexington, KY; Epidemiology Division, University of California, Irvine, Irvine, CA; Mayo Clinic, Rochester; Metro-Minnesota CCOP, Abbott-NW, St. Louis Park, MN; Evanston Northwestern Healthcare, Evanston, IL. Address reprint requests to Cathy A. Coyne, PhD, Department of Community Medicine, School of Medicine, West Virginia University, PO Box 9190, Morgantown, WV 26506; email: ccoyne{at}hsc.wvu.edu.
Purpose: Studies have documented that the majority of consent documents for medical diagnosis and treatment are written at a reading level above that of the majority of the U.S. population. This study hypothesized that use of an easy-to-read consent statement, when compared with a standard consent statement, will result in higher patient comprehension of the clinical treatment protocol, lower patient anxiety, higher patient satisfaction, and higher patient accrual. Methods: A randomized controlled trial was conducted in 44 institutions that were members or affiliates of three cooperative oncology groups. Institutions were randomly assigned to administer either an easy-to-read consent statement or the standard consent statement to patients being recruited to participate in selected cancer treatment trials. Telephone interviews were conducted with a total of 207 patients to assess study outcomes. Results: Patients in the intervention arm demonstrated significantly lower consent anxiety and higher satisfaction compared with patients in the control arm. Patient comprehension and state anxiety were not affected by the intervention. Accrual rates into the parent studies also did not differ significantly between the two study groups. Conclusion: Clinical trial informed consent statements can be modified to be easier to read without omitting critical information. Patient anxiety and satisfaction can be affected by the consent document. The generalizability of these study results is limited by the characteristics of the patient sample. Ninety percent of the sample were white women, and the mean Rapid Estimate of Adult Literacy in Medicine score was approximately 64, indicating a literacy level at or above the ninth grade.
THE INFORMED consent process has recently come under renewed scrutiny.1–3 Although the informed consent document is only one component of the entire consent process, it is an absolute requirement for research posing even a minimal degree of risk. Reading and signing the consent form not only provides a very real reminder of participation in clinical research but is a source for information describing important aspects of the study as defined by Federal statutes.4 Numerous studies have documented that the majority of consent documents for medical diagnosis and treatment are written at or above the twelfth-grade reading level.5–10 This level is considerably higher than that of the majority of the U.S. population, whose average reading abilities are at the eighth-grade level.11 Truly "informed" consent implies that the participant understands the essence of the study and his or her rights and responsibilities as a research participant. Several studies have investigated how well participants comprehend research studies.12–15 A large discrepancy was found between the participant’s understanding about the study and the actual content of the informed consent document the patient was asked to read before his or her participation.12–15 Consent documents were found to include the required information, but participants did not understand what was written, presumably because of limited readability of the documents. Would a simplified and easier-to-read consent statement lead to better comprehension about research projects and, therefore, to a more informed research participant? In the only known prior study on subject comprehension comparing a standard consent (sixteenth-grade level) with a lower literacy consent document (seventh-grade level) for a cancer cooperative group treatment trial, Davis et al15 found that subjects expressed a preference for the modified consent form and found it easier to read. However, no statistically significant difference in comprehension was found. Importantly, the participants for this study were not representative of patients eligible for clinical trials: only some were cancer patients, and none were clinical trial participants. In this study, we tested the effect of an easy-to-read informed consent statement with participants in a cancer treatment trial. Outcomes reported here include differences in comprehension, patient satisfaction with the consent document, patient anxiety, and accrual to the treatment studies.
Design and Sample This study, conducted from March 1998 through January 2000, was a randomized controlled trial comparing the effects of easy-to-read informed consent statements and standard consent statements on patients enrolled in cancer treatment trials. Member institutions and affiliates of three cooperative oncology groups, Eastern Cooperative Oncology Group (ECOG), North Central Cancer Treatment Group (NCCTG), and Cancer and Leukemia Group B (CALGB), were randomly assigned to either intervention (use of the easy-to-read consent statement) or control (use of the standard consent statement). Because many affiliate institutions share institutional review boards (IRBs), the unit of randomization was the IRB, with the exception of CALGB institutions. CALGB main institutions and their affiliated institutions were assigned together, regardless of shared IRBs. The IRB was chosen as the unit of randomization in non-CALGB institutions to eliminate potential contamination and confusion from using two different consent statements at a single institution. Three intergroup treatment trials (parent treatment studies: ECOG Protocol E1594: a phase III trial in metastatic non–small-cell lung cancer; CALGB Protocol C9741: a randomized phase III trial of sequential chemotherapy using doxorubicin, paclitaxel, and cyclophosphamide or concurrent doxorubicin and cyclophosphamide followed by paclitaxel at 14- or 21-day intervals in women with node-positive stage II/III breast cancer; and ECOG Protocol E2197: a phase III study of adriamycin/taxotere v adriamycin/cytoxan for the adjuvant treatment of node-positive or high-risk node-negative breast cancer) were selected as parent treatment studies for the purpose of this consent study. Treatment trials were selected based on anticipated high numbers of patients being accrued and the timing of protocol activation being consistent with the activation of this study. The two primary hypotheses of the study were that use of an easy-to-read consent statement, when compared with a standard consent statement, would result in (1) higher patient comprehension of the clinical treatment protocol, and (2) lower patient anxiety. Two secondary hypotheses tested were that use of an easy-to-read consent statement, when compared with a standard consent, will result in (1) a higher level of patient satisfaction, and (2) higher patient accrual to the parent treatment studies. Patients being recruited by participating institutions to one of the parent treatment studies were eligible to participate in the consent study. After the physician, clinical research associate (CRA), or research nurse explained the parent treatment study to the patient, the CRA or research nurse explained the consent study and asked whether the patient was willing to participate in it. If the patient considered participating in the treatment study and agreed to participate in the consent study, the patient was then asked to sign an informed consent statement for the consent study. The CRA or nurse would then give the patient either the easy-to-read or the standard consent statement for the parent treatment study to read, depending on prior institutional randomization.
Assessment
Approximately 1 to 2 weeks later, but before treatment began, patients were contacted by telephone by trained interviewers. Twenty-three true/false and multiple-choice questions were used to assess patient comprehension of the consent process and treatment regimens. Content validity for the measure was high, as evidenced by the judgments of a panel of experts who reviewed the questionnaire. Patient state anxiety was assessed using the state subscale of the State-Trait Anxiety Inventory (STAI). The STAI’s state subscale17 consists of 20 items that assess the patients’ current feelings of apprehension, worry, nervousness, and tension. Scale scores range from 20 (low state anxiety) to 80 (high state anxiety). For the purpose of scoring in this study, STAI scores were divided by the number of scale items. Thus, a score of 1 indicates low state anxiety, and a score of 4 indicates high state anxiety.
Patient satisfaction with the ease of reading and understanding the consent statement was assessed using four items. This four-item scale had high internal consistency as measured by Cronbach’s coefficient alpha (
Description of the Consent Statements Revisions to the easy-to-read versions of the consent statements included alterations in text style, page layout, font size, and vocabulary. Content of the consent statements was not altered. Readability was reduced to the seventh- to eighth-grade level. Design elements of the easy-to-read consent statements were modified and evaluated using the Suitability Assessment of Materials criteria.19 Words containing only one or two syllables and shorter, less complex sentences were incorporated into the revised consent statements. Other efforts to improve comprehension included use of less technical jargon, parallel sentence structure, incorporation of one idea per paragraph, and use of personal rather than impersonal pronouns. A question-and-answer format was used to organize the information in the consent statement. A check box and additional white space was provided to the right of each question and response to enable patients to indicate that they did not understand a response and to write notes to discuss with the CRA or physician. A treatment calendar was included that used symbols to indicate the method of administration for each drug used in the treatment. Because a larger font size and a greater amount of white space were used, the easy-to-read versions of the consent statements contained 16 pages, over twice as many as the standard consent statements. Easy-to-read consent statements were reviewed by the study’s expert advisory panel and, in addition, by oncologists, research nurses, CRAs, and patient advocates before use.
Statistical Methods The study had 80% power to detect a difference of 0.6 SD in the comprehension and anxiety scores between the two arms (significance level 0.05 two-sided), while taking into consideration the correlation of subjects within a randomization unit.23
Study Population and Randomization Forty-four institutions agreed to participate in the study. Randomization resulted in 24 control institutions and 20 intervention institutions. The study accrued a total of 226 patients. One hundred thirty-seven were from the control institutions, and 89 were from the intervention institutions. Accrual ranged from one to 38 patients per institution, accounting for the unequal numbers of patients in the two study arms. The analysis methods included consideration of the correlation among patients from the same randomization unit.
Most patients (75%) were interviewed within 1 week of signing the consent statement for participation in this study. Sixteen percent of the patients were interviewed within 1 to 2 weeks of signing the study’s consent statement, and fewer than 9% were not reached within this time period. Interviews required 30 to 45 minutes to complete. Of the 19 patients who did not complete the interview, four refused, one was too ill, 10 were unable to be contacted after at least six attempts, one was ineligible for the parent treatment study, and three were ineligible for the consent study because interviewers did not receive information from the site within 1 week of signing the study’s consent statement. Demographic characteristics of the 207 patients completing the interview are shown in Table 1
Intervention Effect One of the primary hypotheses was that patients in the intervention arm would have greater comprehension than patients in the control arm. As shown in Table 2  , this hypothesis was not supported (P = .21).
The study’s outcomes did support the second primary hypothesis that use of an easy-to-read consent statement will result in lower patient anxiety compared with a standard consent statement when consent anxiety was specifically assessed. Estimated score differences after adjusting for correlation within the same randomization units on the consent anxiety scale were significant between the intervention and control arms (P = .016). Lower mean scores in the intervention group indicate lower consent anxiety among the patients using the easy-to-read consent statement (Table 2  ). In contrast, differences in state anxiety between the two arms as measured by the STAI were not significant (P = .89). Consent anxiety and STAI scores were positively associated (P < .0001).
Use of the easy-to-read consent statement was associated with higher patient satisfaction compared with use of the standard consent statement (P = .004), supporting a secondary hypothesis of the study (Table 2 Among the patients who had made a treatment decision by the time of the telephone interview, over 85% in both groups reported that they had decided to participate in the treatment trial. Actual accrual rates onto the parent treatment studies, however, did not differ significantly between the two study groups (P = .32). Patient characteristics found on multiple regression to be associated with study outcomes include REALM score, education level, and age. Patient characteristics positively associated with comprehension were REALM score (P = .007) and education level (P = .008); age was negatively associated with comprehension (P = .0006). REALM score (P = .028) was significantly associated with consent anxiety, whereas it was found to be only weakly associated with state anxiety score (P = .11).
Effects on Comprehension Items
Among breast cancer patients, the proportion of those with correct responses to questions about the treatment protocol itself was low (Table 4 ). Fewer than two thirds of the breast cancer patients selected the correct multiple-choice response to items describing the number of cancer-fighting drugs to be used in the study. In responding to true/false items, a quarter or fewer patients in either arm correctly responded to items regarding the time period patients will be receiving chemotherapy and whether all patients will be receiving the same chemotherapy drugs. A slightly higher proportion of patients in the intervention arm correctly responded to the item inquiring about how it is decided what treatment group they are in compared with those in the control arm. This difference was significant among the breast cancer patients. A large proportion of patients in both arms correctly responded to the questions regarding side effects, choice, and financial aspects of participating in the treatment trial. Differences in percentage of correct responses to items regarding benefit to others between lung and breast cancer patients were noted, with lung cancer patients having higher scores than breast cancer patients.
In this study, we improved the readability of standard consent forms for ongoing clinical treatment trials by making extensive revisions in wording and layout without changing the content, but these changes were not associated with a statistically significant difference in comprehension. Despite a negative finding, this study demonstrated that certain topics important to informed consent are not well understood by patients. It is of concern that only a small proportion of patients understood the level of privacy provided to participants in a research study. Few patients understood that individuals other than staff from the treatment facility, such as representatives from the Federal Drug Administration and the National Cancer Institute, may have access to their medical records. Although the consent statements list who may see their medical records, patients did not seem to understand the level of confidentiality afforded them. This lack of understanding may be because the issue of confidentiality of medical records is not thoroughly addressed during oral consultation with the physician or nurse. Other topics such as the treatment protocol and treatment side effects are likely to be considered more important to discuss with patients and more germane to their decision making than patient confidentiality. Other investigators have also shown that the consent process is often performed in an incomplete manner and that comprehension of key concepts is seldom assessed.24 Patients, especially those with lung cancer, demonstrated a lack of understanding when addressing the extent of personal benefit the treatment studies may provide. Approximately half of the lung cancer patients in either arm responded incorrectly to the item that examined their understanding that a cure was unlikely to result from their participation in the treatment study. Responses to the item examining their knowledge of the purpose of the study indicate that about two thirds of the lung cancer patients believed that the study was being performed to find a treatment that would cure their cancer. These results may mean that these patients truly did not understand that the possibility of cure was remote or that patients understood the limited benefit to treatment, yet needed to retain a sense of hope to cope effectively with the disease. Such unsupported expectations are especially common among patients’ perceptions of potential benefits of treatment in phase I and II cancer clinical trials.13 The differences in anxiety measurements between the two study arms are limited specifically to consent anxiety. Patients in the intervention arm demonstrated significantly lower consent anxiety than did those in the control arm. However, the slight difference in state anxiety between the two groups was not significant. Indeed, the level of state anxiety among the cancer patients in this study was lower than normative data with general medical and surgical patients reported by Spielberger.17 This may have been related to the fact that patients were at least 1 week postdiagnosis and were interviewed while at home. Thus, state anxiety may have been relatively low because the patients were beyond the treatment decision-making stage and had a period of time to cope with the diagnosis. The results of this study demonstrate that consent anxiety and satisfaction with the consent statement were clearly related to the consent modifications. One strength of this study is that it involved cancer patients actually being recruited onto treatment studies. This enabled comprehension to be assessed in patients for whom the information provided in the consent statement was relevant. However, because this study did not assess the information shared during the discussion with the physician, research nurse, or CRA, the study results cannot determine the amount of knowledge gained from these interactions compared with the knowledge gained from reading the consent statement. This is an important limitation of this study, which needs to be addressed in future research. Randomization should have controlled for differences in physician consultation, although this cannot be verified. Our data collection instrument did include questions measuring understanding of all of the essential elements to informed consent, which similar studies have not measured.15 This study did not find general anxiety to be associated with comprehension; however, there were distinct differences in a specific form of anxiety related to the consenting process between the two study groups. Although involvement of patients currently being recruited onto clinical trials is a strength of the study, it also is a limitation. Because patients recruited onto clinical research studies tend to have higher levels of education than the general population, this study was not able to include a significant number of individuals with low literacy levels. This limits the generalizability of this study’s results to more highly literate patients, but not to those who might most benefit from easy-to-read consent statements. However, this supports what previous investigators have reported;25 simple materials do not offend well-educated participants. To assess the effect of easy-to-read consent statements with a more representative patient population, one method might be to conduct such research with patients who are undergoing treatment, but are not on a trial. Unfortunately, this would not address some of the key elements required of informed consent to participate in research. As greater numbers of individuals from underserved populations are recruited for clinical research, either for prevention or treatment, studies should be conducted on ways to meet all the needs of the informed consent process, including the written document, for such a diverse population base. Another drawback of this study was that more than 90% of the participants were white women. Further studies should substantiate the study results in men and minority populations. In conclusion, this study indicates that easy-to-read informed consent statements are associated with reduced patient consent anxiety, an increased satisfaction with the informed consent document, but not with improved patient comprehension. Additional research is needed to examine factors that will improve patient comprehension while preserving patients’ need for effective coping and decision making.26,27
This study was coordinated by the ECOG (Robert L. Comis, MD, Chair). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.
Supported in part by Public Health Service grants CA23318, CA21115, CA13650, and CA72592, and grants from the National Cancer Institute, National Institutes of Health, and the Department of Health and Human Services.
1. Kefalides PT: Research on humans faces scrutiny: New policies adopted. Ann Intern Med 132:513–516, 2000
2. Burman W, Reves RR, Cohn DL, et al: Breaking the camel’s back: Multicenter clinical trials and local institutional review boards. Ann Intern Med 134(2):152–157, 2001
3. Woodward B: Challenges to human subjects protections in US medical research. JAMA 282 (20):1947–1952, 1999
4. Code of Federal Regulation: Protection of Human Subjects, 45 CFR 5. Goldstein A, Frasier P, Curtis P, et al: Consent form readability in university-sponsored research. J Fam Pract 42:606–611, 1996[Medline] 6. Philipson S, Doyal M, Grabram G, et al: Informed consent for research: A study to evaluate readability and processability to effect change. J Investig Med 43:459–467, 1995[Medline] 7. Lawson S, Adamson H: Informed consent readability: Subject understanding of 15 common consent form phrases. IRB 17:16–19, 1995[Medline]
8. Hopper K, TenHave T, Hartzel J: Informed consent forms for clinical and research imaging procedures: How much do patients understand? AJR Am J Roentgenol 164:493–496, 1995 9. Ogloff I, Otto R: Readability on informed consent forms used in research. Ethics Behav 1:239–252, 1991[CrossRef][Medline] 10. Murgatroyd R, Cooper R: Readability of informed consent forms. Am J Hosp Pharm 48(12):2651–2652, 1991[Medline] 11. Kirsch IS, Jungeblat A, Jenkins L: Adult Literacy in America. Washington, DC: U.S. Department of Education, 1992
12. Riecken HW, Ravich R: Informed consent to biomedical research in Veterans Administration hospitals. JAMA 248(3):344–348, 1982 13. Daugherty C, Kiolbasa T, Sieger, M: Informed Consent (IC) in clinical research: A study of cancer patient (pt) understanding of consent forms and alternatives of care in phase I clinical trials. Am Soc Clin Oncol, 1997 (abstr)
14. Williams MV, Parker RM, Baker DW, et al: Inadequate functional health literacy among patients at two public hospitals. JAMA 274(21):1677–1682, 1995
15. Davis TC, Holcombe RF, Berkel HJ, et al: Informed consent for clinical trials: A comparative study of standard versus simplified forms. J Natl Cancer Inst 90(9):668–678, 1998 16. Davis TC, Long SW, Jackson RH, et al: Rapid estimate of adult literacy in medicine: A shortened screening instrument. Fam Med 25:391–395, 1993[Medline] 17. Spielberger CD: Manual for the State-Trait Anxiety Inventory STAI (form Y) "Self-Evaluation Questionnaire." Palo Alto, CA: Consulting Psychologists Press, Inc, 1983 18. Doak CC, Doak LG, Root JH: Teaching Patients with Low Literacy Skills. New York: J.B. Lippincott, 1985 19. Doak CC, Doak LG, Root JH: Teaching Patients with Low Literacy Skills (ed 2). Philadelphia, PA: JB Lippincott, 1995, p 212 20. Laird NM, Ware JH: Random effects models for longitudinal data. Biometrics 38:963–974, 1982[CrossRef][Medline]
21. Liang KY, Zeger SL: Longitudinal data analysis using generalized linear models. Biometrika 73:13–22, 1986 22. Zeger SL, Liang KY: Longitudinal data analysis for discrete and continuous outcomes. Biometrics 42:121–130, 1986[CrossRef][Medline] 23. Donner A, Klar N: Design and Analysis of Cluster Randomization Trials in Health Research. London: Arnold, 2000 24. Titus SL, Keane MA: Do you understand?: An ethical assessment of researchers’ descriptions of the consenting process. J Clin Ethics 7(1):60–68, 1996[Medline]
25. Weiss B, Coyne C: Communication with patients who cannot read. N Engl J Med 337(4):272–274, 1997 26. Lavori PW, Sugarman J, Hays MT, et al: Improving informed consent in clinical trials: A duty to experiment. Control Clin Trials 20(2):187–193, 1999[CrossRef][Medline] 27. Raich PC, Plomer KD, Coyne CA: Literacy, comprehension, and informed consent in clinical research. Cancer Invest 19 (4):437–445, 2001[CrossRef][Medline] Submitted July 3, 2002; accepted November 12, 2002.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
= 0.85), with scale scores ranging from 1 (low satisfaction) to 4 (high satisfaction). The consent anxiety scale, developed for this study, included four items tailored to the specific aspects of distress thought to be correlated with giving consent to treatment. Scores range from 1 (low consent anxiety) to 4 (high consent anxiety). Cronbach’s coefficient alpha (
46, 1991
