IRB: Ethics & Human Research
Oncology Consent Forms: Failure to Disclose Off-Site Treatment Availability
Many research studies investigate new uses of medications that are commercially available.1 For example, cancer researchers often conduct studies to determine whether a chemotherapy agent approved by the Food and Drug Administration (FDA) to treat one type of cancer is also effective at treating a different type of cancer. The importance of such studies has been growing, as a drop has taken place recently in the number of new drugs being studied in clinical trials and receiving FDA approval.2 These types of studies also provide valuable information about the safety and efficacy of FDA-approved drugs that are prescribed off-label, i.e., prescribed for a condition that is not covered by the FDA approval. According to some estimates, the rate of off-label prescribing ranges from about 40% for adults to 90% for children.3
Since clinical studies of new uses of commercially available treatments play an important role in health care research, it is important that these studies be conducted with appropriate adherence to ethical and regulatory standards. Some of the most important of these rules relate to informed consent. Federal regulations, in addition to applicable ethical standards, require that potential research subjects receive relevant information in order to make informed choices about whether to enroll in clinical trials.4 More specifically, potential subjects are supposed to be given “a disclosure of appropriate alternative procedures or courses of treatment, if any, that might be advantageous to the subject.”5 The alternative courses of treatment may include standard therapies, as well as those under investigation in clinical trials. People often decide to participate in clinical trials precisely because they want access to the new treatments being tested in research studies.6 Yet what are prospective subjects told about alternative courses of treatment? Are they being told what they need to know to make an intelligent choice about whether to enroll in the study? To help answer these questions, we collected and evaluated a sample of consent documents for oncology clinical trials.
Study Methods
We obtained our initial sample of consent documents from ClinicalTrials.gov, a clinical trials registry supported by the National Institutes of Health (NIH).7 The registry contains over 60,000 trials sponsored by government agencies, private industry, and private foundations. Studies on the registry are conducted in all 50 states and in 157 countries. The FDA Revitalization Act of 2007 mandates that clinical trials of controlled, clinical investigations of drugs subject to FDA regulation (other than phase I trials) be registered at ClinicalTrials.gov.8 Other trials are also registered at this site in compliance with medical journal policies that require trial registration in a public trials registry as a condition for consideration of publication of trial results.9
We performed our initial search using the registry’s focused search function. The term “cancer” was entered as the disease or condition; “randomization” was entered in the “additional term” search box; the location was “United States”; and the study phase was either “phase II,” “phase III,” or “phase IV.” We focused on cancer research because oncology clinical trials frequently test FDA-approved drugs for different, unapproved uses, such as new drug combinations or applications of drugs to different types of cancer. Phase I studies were excluded because medications used in this trial phase are usually not commercially available. We also excluded medical device studies. Only trials with adult subjects (aged 18 or older) were included to avoid the assent and parental permission issues that arise in the pediatric context. We included both studies that were recruiting and those that were no longer recruiting subjects. Our study was limited to oncology clinical trials conducted in the United States to ensure that all the clinical trials in the sample were required to comply with U.S. regulations and policies concerning informed consent.
We conducted the initial search on ClinicalTrials.gov on April 14, 2007, which yielded 1,794 studies. We then examined each to determine whether the study medications (drugs, biologics, herbal remedies, or nutrients) were commercially available because they were FDA-approved for at least one use (e.g., drugs or biologics) or did not require FDA approval (e.g., nutrients). Information about the approval status of the drugs or biologics was obtained from the FDA’s Web site. We only included studies if all medications were commercially available at the time the study was initiated. These steps narrowed our study population to 749 studies, or 41.8% (S.E. 1%) of the 1,794 obtained in the initial search.
After drawing a random sample of 250 studies from the 749, we attempted to contact people (such as study chairs or coordinators) who might be able to provide us with consent forms. Since reliable contact information was often not available from the information provided at ClinicalTrials.gov, we searched for contact information on the Internet using GoogleTM, PubMed, and other search locations. We eventually obtained consent forms for 113 studies; 21 of the documents were excluded from our sample because they were duplications or they did not meet our eligibility criteria (due, for example, to misinformation that was provided at ClinicalTrials.gov). This left us with 92 consent forms to analyze.
We examined the forms and categorized them as follows: (a) whether the consent form mentioned that subjects may be able to obtain the treatments being investigated in the study without participating in the study; and (b) whether the consent form stated that the treatments being investigated in the study are commercially available. If a study mentioned that a treatment was FDA-approved for at least one use, we categorized this as a “yes” for purposes of (b). We were interested in knowing what percentage of the consent forms fell within each of these two categories. We were additionally interested in determining whether three factors—the source of funding, type of forms used (model vs. local), and number of words in the document—made any difference with respect to whether consent forms belonged to category (a) or (b). Two of us (JA and NW) read all of the consent forms and recorded the data and consulted with DR on how to classify studies and record data. We used logistic regression in SAS version 9.1 for performing the analyses reported in this article.
Consent Form Language
Of the 92 consent forms in our sample, 38 (41.3%) were privately funded, 40 (43.5%) were publicly funded, and 14 (15.2%) had both public and private funding. In our statistical analysis, we grouped all studies that had some public funding into a single category. Forty-six consent forms (50%) were model forms—i.e., forms generated by investigators to be used at different research sites; the other 46 (50%) were local forms—i.e., forms approved by particular institutional review boards (IRBs) for use at one site. The approval dates of the forms ranged from less than a year before the inception of our study (e.g., 2007 date) to 11 years before the study (e.g., 1996 date), with an average age of 2.29 years. The number of words in the consent forms ranged from 970 to 10,458, with an average of 3,750. The number of pages ranged from three to 24, with an average of 9.87 pages.
The most significant finding was that only a small proportion of the consent forms stated that prospective subjects may be able to obtain the treatments being investigated in the study without participating in the study (17.4%, S.E. = 3.95%). There were no statistically significant differences between source of funding (p = 0.965), model vs. local (p = 0.832), and number of words in the document (p = 0.360) with respect to this category. An example of a passage from a consent document stating that the treatments being investigated in the study may be available without participating in the study is as follows:
Instead of being in this study, you have other options. You may receive gemcitabine or the combination of docetaxel and gemcitabine without being included in this study, since both drugs are available nationwide.
Twenty-two of the consent forms (23.9%, S.E. = 4.44%) contained language telling prospective subjects that the treatments under investigation are commercially available. None of the three factors mentioned above were statistically significant at p = 0.05 with respect to this category. There were no statistically significant differences between source of funding (p = 0.410), model vs. local (p = 0.816), or the number of words in the document (p = 0.958). Almost all of these forms included information about commercial availability in a part of the form other than in a specific section marked “alternatives” or “other options,” usually at the beginning of the document. Significantly, of the 22 forms that stated that treatments were commercially available, only five (22.7%) informed subjects that they could obtain the treatments without participating in the study. Ten of those 22 forms (45%) included misleading statements, which we defined as an unqualified statement (i.e., without accompanying language somewhere in the consent form specifically noting off-study availability) that a layperson might likely interpret as implying that he or she could notobtain the treatments except by participating in a research study. Here is an example of such a misleading statement: “Pemetrexed is a commercially available drug. However, when used in combination as directed by this study, pemetrexed is considered experimental.”
Discussion
The fact that less than one in five of the consent forms in our sample disclosed to potential subjects that the study treatments might be available to them off-study has important implications for the ethics of clinical research. The main reason for disclosing this information is to allow prospective research participants to make an informed choice about whether they want to take the chance of not getting the commercially available drug as a study participant. Because many people enroll in clinical trials to get access to a new treatment, not informing them that they might be able to get the drug off-study denies them a key piece of information to make an informed decision about research participation.
What explains the finding that so few consent forms disclose off-study availability of drugs and biologics? One possibility is the concern that significantly fewer people will enroll in the oncology trials if this information is provided than would do so if it were omitted.10 Although that concern is understandable, in our view it does not justify failing to include such information in consent documents. It’s also possible that both researchers and the IRBs that approve consent documents may resist informing prospective subjects about off-study availability of interventions because they believe that off-label use of interventions outside of a research study would not be a wise or even acceptable practice. Because there may be limited or no information about the risks and benefits of treatments prescribed off-label for nonindicated conditions, some might argue that informing potential subjects about the availability of these medications outside of a clinical trial is inappropriate because it might not be in subjects’ best interests to take medications for which there is limited risk-benefit information.
But there is a major logical flaw in this argument.11 If it is not in the best interests of individuals to receive treatments that are not FDA-approved for their condition outside of a clinical trial, then having a 50% chance of receiving the off-label treatment in a randomized clinical trial might also not be in their best interests. Yet it is unlikely that consent forms for oncology trials include language like the following:
Participation in this study is NOT in your best interests. You will be helping us learn important information, but if you want to do what is best to treat your cancer, then you should not be in this study. The treatment being evaluated is much too risky, and we would never let a doctor actually give it to you while claiming that it is a reasonable way to treat you.
There is a narrower version of this argument, which claims that for certain studies, the actual risk of harm from an intervention is lower (or the benefits greater) when participating in a clinical trial than when receiving the same treatment outside of the study.12 This might be due, for example, to study activities—for instance, additional monitoring, tests, or procedures. But this argument also will rarely hold up. Almost all of the studies we reviewed involved merely administration of standard types of chemotherapy combined with common clinical tests. All of these procedures could readily be duplicated by a doctor in private practice. As to the claim that somehow participation in clinical trials produces better outcomes through some type of inclusion benefit, a recent analysis of this claim demonstrated no such benefit for participants in oncology clinical trials.13 No doubt there is the rare clinical trial in which some procedure can be better accomplished in the study than outside it, but that uncommon circumstance does not support the broader proposition that it is generally inappropriate to provide off-label treatments outside of clinical trials.
Our study has a number of limitations, including the fact that our sample may not be representative of all clinical research. We focused on consent forms from cancer clinical trials conducted in the United States. Consent forms from other types of research might be very different. Thus, it is possible that in other types of research on commercially available treatments, a higher percentage of consent forms describe specific alternatives, mention that treatments being investigated in the study may be available without participating in the study, or state that the treatments are commercially available. There does not appear to be any good reason to suspect, however, that the consent forms we studied are very different in this regard from those found in other types of clinical research on commercially available treatments, since the research design of the studies we examined is similar to the design of other studies involving tests of new uses of approved treatments. In research on HIV/AIDS, heart disease, diabetes, and many other diseases, investigators examine new uses of approved drugs, new drug combinations, new dosing regimens, and so on.14 Nonetheless, extending our study would be useful to more definitively determine if similar conclusions would hold for the consent forms used in other medical fields.
Another limitation of our study is that we examined consent forms, not the entire process of consent. The document is, of course, but one part of the process of consent, which also includes candid discussions between investigators (or others who conduct the consent process) and prospective subjects. Therefore, those who conduct the consent process may sometimes—and perhaps even often—inform prospective subjects about the option of obtaining treatments off-study, even when the consent form does not include this information.
That possibility—which of course is only a possibility, and in our view, an unlikely one at that—should not overshadow the fact that this information is not provided in consent forms, which we believe is a clear wrong. There are powerful reasons behind the core requirement in the federal regulations governing research with humans that potential subjects be provided with certain information in writing. Until such time as the process of obtaining consent is routinely documented—such as by audiotaping or videotaping informed consent discussions—then providing information in a consent form about the availability of the study drug off-trial is the only way to have minimal proof that this information is given to prospective research subjects. This information may be of crucial importance to many, if not most, potential subjects as they decide whether to enroll in a clinical trial. Moreover, we believe that not providing prospective subjects with information about off-study availability of treatments could violate the regulatory requirement that information about alternatives to participation must be provided in the consent form.
More research is certainly needed on this topic. Interviews with investigators and IRB members may help us understand why they would choose to withhold information on the availability of medications off-study—for instance, whether some may feel that telling prospective subjects about the availability of medications off-study is inappropriate. Interviews with prospective as well as current and former research subjects may help us to better understand the importance of disclosing this information to them. Observations of informed consent discussions would allow us to know whether information about the availability of treatments off-study is sometimes communicated to prospective subjects, even when the information is not contained in the consent document. Until such information is obtained, it appears from our sample of consent forms that potential subjects in clinical oncology trials may not be provided with information that is crucial to making an informed decision about research participation.
Acknowledgments
This research was supported by the intramural program of the National Institute of Environmental Health Sciences and the Office of the Director, Office of Intramural Research, National Institutes of Health. It does not represent the views of the NIH. The authors have no conflicts of interest to disclose. David Resnik had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The authors are grateful to Adil Shamoo and Grace Kissling for reviewing an earlier draft of this manuscript.
This research was classified as not subject to human subjects research regulations by the Office of Human Subjects Research, National Institutes of Health.
David B. Resnik, JD, PhD, is Bioethicist, National Institute of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, NC; Shyamal Peddada, PhD, is Principle Investigator, Biostatistics Branch, National Institute of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, NC; Jason Altilio is a graduate student, City College of New York, New York, NY; Nancy Wang is an undergraduate student, Duke University, Durham, NC; and Jerry Menikoff, MD, JD, is Director, Office of Human Subjects Research, National Institutes of Health, Bethesda, MD.
References
1. Menikoff J.What the Doctor Didn’t Say: The Hidden Truth about Medical Research. New York: Oxford University Press, 2006.
2. Chong C, Sullivan Jr. D. New uses for older drugs.Nature2007;448:645-646.
3. Beck J, Azari E. FDA, off-label use, and informed consent: Debunking myths and misconceptions.Food Drug Law Journal1998; 53:71-104; Gazarian M et al. Off-label use of medicines: Consensus recommendations for evaluating appropriateness.Medical Journal of Australia2006;185:544-548.
4. Emanuel E, Wendler D, Grady C. What makes clinical research ethical?JAMA2000;283:2701-2711.
5. Department of Health and Human Services. Protection of Human Subjects. 45 CFR 46.116(a)(4)(2005).
6. See ref. 1, Menikoff 2006; Daugherty C, Banik D, Janish L, Ratain M. Quantitative analysis of ethical issues in phase I trials: A survey interview study of 144 advanced cancer patients.IRB: Ethics & Human Research2000;22(3):6-14; Grady C. Patient or guinea pig? Dilemma of clinical trials.New York Times, January 5, 1999, D1.
7. https://clinicaltrials.gov.
8. Drazen J, Morrissey S, Curfman G. Open clinical trials.NEJM2007;357:1756-1757.
9. De Angelis C et al. Clinical trial registration: A statement from the International Committee of Medical Journal Editors.NEJM2004;351:1250-1251.
10. See ref. 1, Menikoff 2006.
11. See ref. 1, Menikoff 2006.
12. Lippman M. High dose chemotherapy plus autologous bone marrow transplantation for metastatic breast cancer.NEJM2000;343:1119-1120.
13. Peppercorn J, Weeks J, Cook E, Joffe S. Comparison of outcomes in cancer patients treated within and outside clinical trials: Conceptual framework and structured review.Lancet2004;363:263-270.
14. See ref. 1, Menikoff 2006.
David B. Resnik, Shyamal Peddada, Jason Altilio, Nancy Wang, and Jerry Menikoff, “Oncology Consent Forms: Failure to Disclose Off-Site Treatment Availability,”IRB: Ethics and Human Research 30, no. 6 (2008).