IRB: Ethics & Human Research
Waivers and alterations to consent in pragmatic clinical trials: Respecting the principle of respect for persons
Is it ever ethical to conduct a randomized controlled trial (RCT) without consent?1This surprising question is increasingly being asked2owing to great interest in RCTs that compare widely used “standard” treatments in large pragmatic trials.3Such pragmatic studies may sometimes pose, at most, minimal research risk to participants (or perhaps even net benefits, in some studies)4and promise clinically salient information at low cost.5Thus, pragmatic trials would be an important part of a “learning healthcare system” in which the generation of knowledge is “embedded into the core of the practice of medicine” leading to “continual improvement in care.”6The aim is to embed such pragmatic trials seamlessly into the everyday practice of medicine, with randomization between standard treatments the only research procedure and with the outcomes extracted from the electronic health record.7How can such research be integrated into everyday delivery of patient care? Doubts about the compatibility of current informed consent practices and proposed pragmatic trials have prompted various alternatives to the current informed consent process prescribed by U.S. regulations governing research with humans (hereafter referred to as “regulatory consent”).8Some believe that selected pragmatic RCTs could be conducted ethically without explicit informed consent.9Others have proposed a simple research consent integrated with clinical consent.10Currently, comparative effectiveness RCTs are being planned with various types of departures from regulatory consent.11Any RCT proposing an alternative to regulatory consent must meet specified criteria for a “waiver or alteration of informed consent” (Table 1).
Given that pragmatic comparative effectiveness studies will increasingly be proposed with waivers or alterations to regulatory informed consent, how should the ethical issues relevant to applying these criteria be understood? There is no well-established ethical framework for interpreting the criteria and little specific guidance on how to interpret them. This lack of guidance is evidenced by the fact that, as recently as 2008, the Secretary’s Advisory Committee on Human Research Protections (SACHRP) recommended that the Office for Human Research Protections issue specific guidance on the topic.12Lack of guidance is especially acute in applying the waiver criteria to pragmatic RCTs because historically these criteria have been applied to medical records research or to studies with special research designs (such as experiments requiring deception).13Thus, pragmatic RCTs, which compare individual medical interventions, are a novel setting for waivers or alterations of regulatory consent.
The most important reason that an ethical framework for waiver and alteration of informed consent is needed is that the regulations lump together waivers and alterations, and this obscures important ethical distinctions. Altering regulatory consent by leaving out a single consent element (e.g., whom to contact for information)14is ethically quite different from waiving consent to RCT participation, which could involve actively concealing from someone that he is enrolled in an RCT; yet the regulations require the application of the same four criteria for these twoethicallydifferent departures from regulatory consent. This is problematic, as it opens the door to over-regulation (unnecessarily prohibiting ethically acceptable alterations by holding them to the same standard as for a complete waiver) as well as to under-regulation (erroneously permitting unethical waivers by holding them to the same standard as for a minor alteration). Clearly, a principled ethical framework is necessary to prevent such mistakes.
Assessing Alternative Consent Models: Respect for Persons Framework
What is the appropriate ethical framework governing a waiver or alteration of regulatory consent for pragmatic RCTs? The principle underlying informed consent is respect for persons,15a principle distinct from and independent of considerations of the benefits and risks of participating in research. Reconsideration of informed consent requires an analysis of respect for persons because, although regulatory informed consent is rooted in the principle of respect for persons, the principle itself is broader. It is possible to show respect in ways other than obtaining formal informed consent. Respect also encompasses trust, transparency, and consideration of other legitimate expectations that arise from relationships between clinicians and patients and between clinical investigators and research participants.16Missing from current debates about informed consent for pragmatic RCTs is attention to the following question: do proposed consent models, including waiver of consent for research, conform to the principle of respect for persons?
Since principles do not operate in a vacuum, how to meet the need to respect persons depends on the context. For this discussion, imagine a pragmatic RCT in a learning health care system comparing two or more widely used “standard” treatments for hypertension that appear to have similar side effects. As part of the learning health care system’s goal of continuous improvement in care, assume the RCT is to be integrated into a clinic’s daily routine to maximize generalizability at very little cost. The researchers argue that the RCT’s aim of deriving clinically generalizable and relevant data would not be achievable with regulatory consent, because such consent could not be integrated into the daily routine of the clinic. In order to assess the ethical impact of a proposal to waive informed consent or to alter some element of regulatory consent, it is necessary to assess the implications that the waiver or the alteration would have for the principle of respect for persons in this context. This can be accomplished by addressing three issues.
First, to what degree does the alternative to regulatory consent meet legitimate patient expectations, especially of transparency and trustworthiness? By its very purpose, the pragmatic RCT in a learning health care system seeks to integrate research into the ordinary treatment context.17In such a context, the doctor-patient relationship embodies expectations that are essential to patient decision-making. Sick or injured patients (and even healthy persons, who will, eventually, all become sick or injured) must be able to place implicit trust in their doctors for many medical decisions. Thus, for example, patients give tacit consent for routine perioperative antibiotics,18and parents of infants in the neonatal intensive care unit entrust the doctors with decisions about oxygenation parameters.19But this implicit trust is rooted in the doctor-patient relationship. When the underlying basis for choice of antibiotics or oxygenation settings deviates from ordinary clinical practice and reasoning, as in pragmatic RCTs, patients have a legitimate expectation of transparency. In addition, for preference-sensitive decisions20(e.g., between two treatments of equivalent efficacy but different side effects, a decision in which the patient’s personal preference should carry a lot of weight) or for recommended treatments with significant risks and burdens, patients rightfully expect to have the final, informed say in the treatment choice. By contrast, a patient entering a hospital does not expect to have a say in how a hospital decides to staff its intensive care unit.21
Second, the right of patients to make their own health care or research decisions implies that they have an opportunity to authorize a decision. A consent alternative that lacks transparency about research participation does not give individuals any opportunity to directly authorize their participation in research. Specifically, even when a patient explicitly consents to the treatment assigned in an RCT, that consent does not authorize participation in research if she is unaware of the research purpose that governed the treatment selection. Failure to give due weight to this ethical consideration is a weakness of some commentaries arguing that consent is not necessary for some RCTs of individual treatments.22However, it does not follow that regulatory consent is required. For some pragmatic trials, even a simple verbal consent may suffice to authorize participation. While a learning health care system that advertises, educates, and sends letters to inform its patients regarding RCTs could be providing an opportunity of a sort, the nature of that opportunity will depend on the degree of effort patients must exert and the significance of consequences they must accept if they choose not to participate. Institutional review boards (IRBs) would need to evaluate the meaningfulness and adequacy of the various types of opportunities for authorization.
Third, what is the fit between the type, amount, and mode of information provided and the nature of the decision? A more lengthy informed consent process does not imply a more valid or meaningful consent. Consider, for example, an RCT for acute treatment of myocardial infarction (MI). Patients in the emergency room who are potential research participants will be uncomfortable and frightened but not necessarily incapable of giving valid informed consent to enroll in a study.23Is a detailed and lengthy process that is typical of clinical trials more respectful than an abbreviated consent process tailored to communicate a few high-priority items? An abbreviated consent process that focuses only on certain key elements (and forgoing, for example, an explicit discussion of confidentiality and privacy that is already implicit in a hospital setting) might be more meaningful. Excessive information insensitive to the context can confuse individuals rather than enhance their understanding about the research in which they are being recruited to participate.24Moreover, the need for rapid treatment precludes regulatory consent. However, some may argue that prospective consent should be waived entirely when enrolling patients in studies involving acute treatment for MI.25Consideration needs to be given to which option—a simplified consent, an opportunity for “informed refusal,”26a waiver of consent, or another alternative, if any—best satisfies respect for persons while promoting socially valuable research. In any case, some alterations of consent have the potential to be better than regulatory consent from an ethical perspective.
Respect Framework for Waiver or Alteration Criteria
A “respect framework” based on the principle of respect for persons provides a systematic justification and a principled guide for implementing the criteria for a waiver or alteration of consent. The four criteria should not be treated as independent items on a checklist but should, rather, be understood and implemented in a way that integrates the principles underlying them. This can be done in three steps.
First, theminimal riskcriterion, based on the principle of beneficence (specifically, regarding research participants’ welfare), is used to set the risk threshold above which no alternative to regulatory consent is compatible with respect for persons—defined in the regulations as the level of risk commensurate with everyday life or that encountered during routine physical or psychological examinations or tests.27In assessing minimal risk, IRBs should focus on the incremental risks of the research, as distinct from the risks involved in standard medical care. In some comparative effectiveness RCTs, it may be that participation in research poses no more than minimal increase in risk to individuals than what they would be exposed to if they were not in the study.
Second, once a protocol’s eligibility for consideration of waiver or alteration of regulatory consent is established, the next step is to determine to what degree the principle of respect for persons is affected by the proposed alternative. The criterion that a proposed alternative to regulatory consent should not adversely affect the rights and welfare of research subjects is firmly rooted in the principle of respect for persons. As a 2008 report by the SACHRP notes, in applying the criterion, one should ask, would research participants either object to the waiver or consider that such a waiver has potential to cause adverse consequences for them?28This focus on the expectations of research participants is crucial in understanding why the mention of welfare in this criterion is not redundant; in other words, given the minimal risk criterion, one might think that a further provision about welfare is not needed. However, the minimal risk criterion refers to the research risks, determined by the IRB according to the regulations. That determination does not include input from research subjects. In essence, an IRB’s determination of acceptable risk simply means that a study’s risks in relation to its benefits are such that individuals can be invited to enroll in the study. If individuals provide informed consent, then in effect they are personally accepting the risks of research participation from their individual perspectives. However, when a deviation from informed consent is proposed, what is the status of the individual perspective in the proposed alteration or waiver of consent? The rights and welfare criterion requires an assessment of this question. We agree with the SACHRP’s interpretation of the criterion such that the notion of welfare here is best interpreted as arising from the prospective research participant’s perspective and expectations.
The requirement of providing to participants “whenever appropriate . . . additional pertinent information after participation” is also clearly rooted in the principle of respect.29In evaluating consent processes for pragmatic RCTs, this criterion should be interpreted as an obligation to maintain respect throughout participation in an RCT. Thus, after a person is enrolled in the study via an altered consent process,30a more detailed written information sheet could be provided. Efforts by a learning health care system to advertise and educate patients about the integration of learning and care in a health system can also be seen as arising from this obligation.31
The third step is to judge whether the “research could not practicably be carried out without the waiver or alteration” (Table 1). This practicability criterion introduces considerations of scientific validity and feasibility, which are closely intertwined in the context of pragmatic RCTs.32If implementing traditional regulatory consent would make a pragmatic RCT’s results invalid, then in effect the consent makes the intended study not possible to carry out. However, practicability should not be determined solely on the basis of “convenience, cost, or speed.”33But for pragmatic RCTs testing individual treatments, validity and cost are intricately intertwined since the very point of the research—that of integrating research into the practice setting itself to maximize its clinical relevance—is to maximize external validity at minimal cost.
The best way to understand the task for the IRB in applying this criterion is to consider the task as resolving a dilemma. Researchers who propose a waiver or alteration are asserting that there is a dilemma: conducting the proposed pragmatic RCT is desirable, and regulatory consent is generally required, but we cannot have both. It is the IRB’s task to first determine if a genuine dilemma exists: if it does not (i.e., the RCT and consent are compatible), then no waiver or alteration is granted. If there is a genuine dilemma, then there are only two choices: the RCT can be revised to preserve sufficient scientific and social value that is compatible with regulatory consent, or, if the first option is not possible, then the IRB must evaluate whether waiver or alteration of regulatory consent is permissible. It could turn out that no waiver or alteration is permissible, and the final determination would be that the proposed pragmatic RCT is not permissible.
We propose the following rule to guide researchers and IRBs in their task of resolving the above dilemma: Any permissible alternative to regulatory consent should use procedures that preserve the greatest degree of respect for persons that is compatible with practicability of the RCT. This is analogous to a reasoning process that researchers and IRBs already follow in the analysis of research risks. Just as research risks to subjects should be scientifically necessary and minimized, any alterations to the informed consent requirement should both be necessary and preserve the greatest degree of respect for subjects that is compatible with the completion of a valid study.
Thus, judgments regarding scientific impracticability should be relatively flexible, using a low threshold, to the extent that the proposed altered consent process respects potential research participants with due sensitivity toward their rightful expectations, gives them a meaningful opportunity for authorization, offers a consent process that is congruent to the nature of the decision, and provides pertinent information in various other ways (e.g., letters and advertisements). This is because, despite the deviation from regulatory consent, the ethical principle of respect for persons is still strongly upheld and not compromised. Indeed, it is likely that simplifying the consent process so that the information is much more congruent to the task may even improve participants’ understanding of the research study.
However, for pragmatic RCT protocols that propose waivers of consent involving lack of transparency and no substantive opportunity for authorization, the threshold for impracticability should be set quite high. This is because such a process may well be incompatible with the principle of respect for persons. Complete waivers of consent should not be granted if there are feasible alterations to regulatory consent consistent with the practicability of pragmatic RCTs.
The compelling promise of comparative effectiveness RCTs in learning health care systems, their relatively low risk in some instances, and their likely incompatibility with traditional or regulatory informed consent naturally raise the question of whether waiver or alteration of consent for such studies is appropriate. Answering this question requires an ethical framework that takes into account the principle of respect for persons. The goal of preserving this principle to the greatest degree feasible should guide decisions about waivers or alterations of informed consent. For pragmatic RCTs of sufficiently low risk, any departures from regulatory consent should be the minimum necessary for the successful completion of the trial.
Scott Y. H. Kim, MD, PhD,is a senior investigator in the Department of Bioethics in the Clinical Center at the National Institutes of Health, andFranklin G. Miller, PhD,was, at the time of writing this article, a member of the senior faculty in the Department of Bioethics in the Clinical Center at the National Institutes of Health.
The ideas and opinions expressed are the authors’ own. They do not represent any position or policy of the National Institutes of Health, the Department of Health and Human Services, or the U.S. government.
The authors thank Neal Dickert, Steven Joffe, David Wendler, and Alan Wertheimer for comments on earlier drafts.
1. Pletcher MJ, Lo B, Grady D. Informed consent in randomized quality improvement trials: A critical barrier for learning health systems.JAMA Internal Medicine. 2014;174(5):668-670.
2. Faden R, Kass N, Whicher D, et al. Ethics and informed consent for comparative effectiveness research with prospective electronic clinical data.Medical Care2013;51:S53-S57; Platt R, Kass NE, McGraw D. Ethics, regulation, and comparative effectiveness research: Time for a change.JAMA.2014;311(15):1497-1498.
3. Eapen ZJ, Lauer MS, Temple RJ. The imperative of overcoming barriers to the conduct of large, simple trials.JAMA2014;311(14):1397-1398.
4. See ref. 1, Pletcher et al. 2014.
5. Staa TP, Goldacre B, Gulliford M, et al. Pragmatic randomised trials using routine electronic health records: Putting them to the test.BMJ2012;344:e55.
6. Olsen L, Aisner D, McGinnis J.The Learning Healthcare System. IOM Roundtable on Evidence-Based Medicine.Washington, DC: Institute of Medicine of the National Academies, 2007.
7. See ref. 5, Staa et al. 2012.
8. See ref. 1, Pletcher et al. 2014; Faden RR, Beauchamp TL, Kass NE. Informed consent, comparative effectiveness, and learning health care.NEJM2014;370(8):766-768; Kim SYH, Miller FG. Informed consent for pragmatic trials—the integrated consent model.NEJM2014;370(8):769-772.
9. See ref. 2, Faden et al. 2013.
10. See ref. 8, Kim and Miller 2014.
11. NIH Collaboratory. Demonstration project ethics and regulatory documentation. https://www.nihcollaboratory.org/demonstration-projects/Pages/default.aspx.
12. Secretary’s Advisory Committee on Human Research Protections (SACHRP). Recommendations related to waiver of informed consent. https://www.hhs.gov/ohrp/sachrp/sachrpletter013108.html. 2008.
13. Elliott MM. Research without consent or documentation thereof. In: Amdur R, Bankert E, eds.Institutional Review Board: Management and Function. Sudbury, MA: Jones and Bartlett, 2002:250-255.
14. Department of Health and Human Services. Protection of Human Subjects. 1991. 45 CFR 46. Section 116d.
15. The Belmont Report: Ethical principles and guidelines for the protection of human subjects of research. Washington, DC: U.S. Government Printing Office, 1978.
16. Dickert NW. Re-examining respect for human research participants.Kennedy Institute of Ethics Journal2009;19(4):311-338.
17. Grumbach K, Lucey CR, Johnston S. Transforming from centers of learning to learning health systems: The challenge for academic health centers.JAMA2014;311(11):1109-1110.
18. Truog RD, Robinson W, Randolph A, et al. Is informed consent always necessary for randomized, controlled trials?NEJM1999;340(10):804-807.
19. Magnus D. The SUPPORT controversy and the debate over research within the standard of care.American Journal of Bioethics2013;13(12):1-2.
20. O’Connor AM, Mulley AG, Wennberg JE. Standard consultations are not enough to ensure decision quality regarding preference-sensitive options.Journal of the National Cancer Institute2003;95(8):570-571.
21. Kerlin MP, Small DS, Cooney E, et al. A randomized trial of nighttime physician staffing in an intensive care unit.NEJM2013;368(23):2201-2209.
22. See ref. 2, Faden et al. 2013; see ref. 18, Truog et al. 1999.
23. Dickert NW, Llanos A, Samady H. Re-visiting consent for clinical research on acute myocardial infarction and other emergent conditions.Progress in Cardiovascular Disease2012;55(3):251-257.
24. Singer E, Vonthurn DR, Miller ER. Confidentiality assurances and response: A quantitative review of the experimental literature.Public Opinion Quarterly1995;59(1):66-77; Walkup J, Bock E. What do prospective research participants want to know? What do they assume they know already?Journal of Empirical Research on Human Research Ethics. 2009;4(2):59-63.
25. Shaw D. HEAT-PPCI sheds light on consent in pragmatic trials.Lancet2014;384(9957):1826-1827.
26. See ref. 23, Dickert et al. 2012.
27. Department of Health and Human Services. Protection of Human Subjects. Subpart A. Basic HHS Policy for the Protection of Human Subjects. 45 CFR 46.102.
28. See ref. 12, SACHRP 2008.
29. See ref. 12, SACHRP 2008.
30. See ref. 8, Kim and Miller 2014.
31. See ref. 8, Faden et al. 2014; Largent EA, Joffe S, Miller FG. Can research and care be ethically integrated?Hastings Center Report2011;41(4):37-46.
32. See ref. 12, SACHRP 2008.
33. See ref. 12, SACHRP 2008.
Kim SYH and Miller FG. Waivers and alterations to consent in pragmatic clinical trials: Respecting the principle of respect for persons.IRB: Ethics & Human Research2016;38(1):1-5.