IRB: Ethics & Human Research
Research on Medical Practices (ROMP): Attitudes of IRB Personnel about Randomization and Informed Consent
Activities such as continuous quality improvement, comparative effectiveness research, and electronic medical record review—collectively referred to as research on medical practices (ROMP)—are posing more and more challenges for meeting regulatory requirements governing research with humans. In 2014, the Office for Human Research Protections issued draft guidance to address inconsistent interpretations of how regulatory rules apply to ROMP that uses a randomized approach. That guidance has not been finalized, however, and differing regulatory interpretations persist. It is unknown whether institutional review board (IRB) personnel view ROMP differently from how they view research evaluating new treatments, particularly when ROMP involves randomization of patients to existing modes of usual care. In this study, we surveyed individuals with experience as IRB personnel about ROMP in order to better understand where there are differing opinions about IRB review and informed consent for ROMP. We used a self-administered online survey to collect data from members of Public Responsibility in Medicine and Research (PRIM&R), a professional organization that includes IRB personnel, researchers, and others interested in the protection of human research subjects. We found that PRIM&R members with experience as IRB personnel expressed varying views when responding to three key issues: what triggers IRB review, who should obtain consent, and when consent can be waived. Additional guidance for IRB personnel regarding activities at the intersection of clinical care and research, including ROMP and quality improvement, is essential to promote consistent interpretation of the regulations across IRBs.
Keywords: research on medical practices (ROMP), research ethics, federal regulations for human subjects research, institutional review boards (IRBs)
There has been increased scrutiny of research taking place in the context of learning health systems as well as increased debate about its ethics. Activities such as continuous quality improvement, comparative effectiveness research, and electronic medical record review—collectively referred to as research on medical practices (ROMP)—are posing more and more challenges for meeting regulatory requirements governing research with humans.1 The fact that ROMP shares some characteristics with clinical research and some with usual clinical practice leads to uncertainty as to how the regulations should be applied to ROMP.2 In particular, ROMP that uses a randomized approach, wherein patients are randomized to different treatments already utilized in usual clinical care so that outcomes can be compared, shares some characteristics with research to evaluate new treatments, such as an explicit goal of generating generalizable knowledge. But unlike in studies of new treatments, all participants in ROMP receive a commonly accepted treatment. Accordingly, there has been some effort to clarify the existing research regulations as they pertain to ROMP. In 2014, the Office for Human Research Protections issued draft guidance to address inconsistent interpretations of how regulatory rules apply to ROMP that uses a randomized approach.3 That guidance has not been finalized, however, and differing regulatory interpretations persist.4
It is unknown whether institutional review board (IRB) personnel (defined in this article to include IRB members, chairs, vice chairs, and staff members) view ROMP differently from how they view research evaluating new treatments, particularly when ROMP involves randomization of patients to existing modes of usual care. Of particular importance are what activities IRBs consider to be research, therefore requiring their review, and whether such research is minimal risk. A second set of questions, for ROMP that utilizes randomization, is whether IRB personnel prefer clinicians or researchers to obtain consent and whether consent or documentation of consent may be waived. Previous focus-group data suggest a lack of consensus among IRB personnel regarding these questions.5 In this study, we surveyed individuals with experience as IRB personnel about ROMP in order to better understand where there are differing opinions about IRB review and informed consent for ROMP.
We used a self-administered online survey to collect data from members of Public Responsibility in Medicine and Research (PRIM&R), a professional organization that includes IRB personnel, researchers, and others interested in the protection of human research subjects. Focus groups and cognitive interviews with IRB personnel were used to guide development of the survey. Additionally, review by content experts in bioethics and clinical research further informed and refined survey development. Specific sampling and recruitment methods are described below. Study data were collected and managed using REDCap electronic data-capture tools hosted at the University of Washington. The study was approved by the University of Washington IRB, with a waiver of documentation of informed consent for the survey.
The sampling frame for the survey was the list of all (approximately 3,200) PRIM&R members who reside in the United States and have self-identified as being engaged in human subjects research issues. We sent email invitations for study participation, along with a link to our survey, to 1,500 randomly selected PRIM&R members, in two waves of 750 each. Nonrespondents were emailed a second invitation 7 days after the first invitation. Via the U.S. Postal Service, remaining nonrespondents were mailed a reminder 16 days after the second invitation and a final email invitation 3 days later. Individuals who completed the survey received a $20 gift card. Data collection occurred from July 2014 to September 2014.
The survey included questions that probed attitudes toward which activities constitute research, who ethically may obtain informed consent for a research study, who is the preferred person to obtain consent, and when informed consent may be waived. Respondents were randomly assigned to one of two scenarios that described patients with hypertension or atrial fibrillation being randomly assigned one of two drugs with similar side effects, with no blinding of the medications, and with usual clinical follow-up. There were 298 respondents who reviewed the hypertension scenario and 303 respondents who reviewed the atrial fibrillation scenario. Patients with these conditions are at risk for stroke, but the risk differs in likelihood depending on whether they have hypertension or atrial fibrillation. Respondents were also given the opportunity to respond to two open-ended questions, the first focusing on issues regarding informed consent to participate in ROMP and the second asking for any final thoughts they would like to share. Exemplary quotations are presented in the discussion.
Descriptive statistics are reported for specific survey items. The chi-square statistic was used for bivariate analyses. All data management and statistics were performed using SAS® 9.4.
The survey response rate was 40.1% (n = 601/1,500). The majority of respondents (537, or 89.4%) indicated previous or current experience with an IRB. These 537 respondents included 310 (57.7%) reporting experience as IRB members, including chairs and vice chairs, and 224 (41.7%) as IRB staff members. In addition, 255 (47.5%) respondents reported experience as researchers and 104 (19.4%) as clinicians (a nurse, psychologist, or physician). The data in this paper are limited to the 537 individuals who reported having experience as IRB personnel. There were no significant differences between responses for the two scenarios.
What should trigger IRB review? We asked respondents to identify which of seven specific research activities should always trigger full IRB review. As illustrated in Figure 1, a majority of respondents indicated that randomly assigning patients (81.8%) or hospitals or clinics (74.8%) to use specific treatments should always trigger full IRB review. When considering if collecting and analyzing patient data should trigger full IRB review, research intent mattered: 44.2% indicated that IRB review was required when the intention was testing hypotheses for generalizable knowledge, 26.4% when the intention was publishing research results, and 10.1% when the intention was improving future practices within a health system. According to 17.5% of respondents, IRB review was required for sharing deidentified patient data with the intention of testing hypotheses for generalizable knowledge. Few respondents indicated that IRB review should always take place when standard clinical pathways are used to determine patients’ treatments.
Who may or should obtain consent? We asked respondents whether individuals in the following roles may ethically obtain informed consent for a study that randomizes patients to different forms of usual care: the patient’s clinician, an investigator who is not involved with the patient’s care, and a research nurse or study coordinator who is not involved with the patient’s care. As seen in Table 1, approximately half of respondents indicated that all of these types of individuals may ethically obtain informed consent (49.2%). Interestingly, nearly one-third of respondents indicated that all but a patient’s clinician may ethically obtain consent (31.4%), while 9.5% said that only a patient’s clinician may ethically obtain consent.
We also asked respondents to indicate their preference for who should obtain informed consent (Table 2). The patient’s clinician was most preferred by 35.3% of respondents but least preferred by 54.5%. An investigator not involved with the patient’s care was most preferred by 29.9% of respondents and least preferred by 19.8%, and a research nurse or study coordinator not involved with the patient’s care was most preferred by 34.8% of respondents and least preferred by 25.7%.
When can informed consent be waived? Figure 2 shows respondents’ views about the acceptability of waiving informed consent for a study that randomizes patients to different forms of usual care. While 36.6% of respondents indicated that waiving consent was never acceptable, the remainder indicated that a waiver could be acceptable, even if rarely (34.8%).
What are the minimum acceptable and preferred approaches to consent? Respondents were asked to select the minimally acceptable approach to consent in response to our scenario about randomization of individual patients. Most respondents (69.6%) indicated that the minimum acceptable approach includes discussion and written permission (Figure 3). This is compared with 16.0% indicating that it includes a discussion and verbal permission, 11.9% indicating that it includes a general information document, and 2.4% indicating no notification as the minimum acceptable approach. Additionally, respondents were asked to select their preferred approach to notification or consent, and 80.9% preferred a discussion and written permission (p ≤ .0001).
PRIM&R members with experience as IRB personnel expressed varying views when responding to three key issues: what triggers IRB review, who should obtain consent, and when consent can be waived. Respondents expressed diverse opinions about whether certain types of research should always trigger IRB review, including a small minority who indicated that full IRB review was always required for “using standard clinical pathways to determine patients’ treatments,” in spite of the fact that this type of activity does not normally fall under an IRB’s purview. Our study reveals that respondents had different interpretations of what qualifies as research and further illustrates the difficulty in reconciling the blending of research and clinical practices in ROMP. The challenges of determining when ROMP activities constitute research have been discussed for more than a decade,6 and yet IRBs continue to struggle without clear guidance. Protecting research participants while simultaneously allowing important quality-improvement initiatives to continue efficiently with proportional oversight will require ongoing effort on the part of IRB personnel, regulators, quality experts, and members of the bioethics community.
Respondents also expressed diverse opinions about who may and should obtain informed consent. Some expressed concerns about clinicians obtaining informed consent, whereas others believed only clinicians may ethically obtain informed consent. Furthermore, while clinicians were the least preferred option for approximately half of respondents, one-third chose clinicians as their most preferred. This difference is illustrated by two responses to the open-ended questions. One respondent expressed concerns surrounding clinician-obtained consent: “It seems nearly impossible for a patient to not experience undue influence when recruited for research participation by a provider involved in his/her care.” On the contrary, another respondent expressed reasons for preferring a clinician: “I have a slight preference for the treating MD because this person can more knowledgeably address the subject’s concerns regarding the appropriateness of being randomized to a [standard of care] therapy vs. receiving a therapy that is best for the subject.”
Interestingly, other data suggest that the majority of potential research participants prefer to be asked to participate in research by their own physician.7 For low-risk research in clinical settings, consent by physicians might be more efficient and might also encourage patients to be more engaged in the consent process and the decision to join the study.8 Given the potential conflict of interest that this situation poses, however, it is not surprising that this issue remains unsettled among members of the bioethics community.9
Finally, respondents expressed diverse views when addressing the acceptability of waiving informed consent for a study that randomizes patients to different forms of usual care. Although a third of respondents considered waiving consent in this situation never acceptable, most respondents considered this rarely or sometimes acceptable. The open-ended responses illustrated the range of views regarding the ethical underpinnings of waiving informed consent. One respondent said it would be “hard to justify a waiver of informed consent if you are going to have the opportunity to see the potential participant face to face on multiple occasions. There would need to be a very strong justification for waiving.” Another respondent suggested, in contrast, that waiving consent can often be
ethically justified. According to this respondent, “[W]aiving traditional informed consent does not necessarily prevent transparency, notification, and even some form of opt out. The ethical principle of respect for persons may be best manifest[ed] by different kinds of processes and procedures in different research contexts in which all subjects/patients are treated with standard care practices.”
According to the federal research regulations, in order to waive or alter consent, an IRB must determine that the research involves no more than minimal risk, the waiver will not adversely affect the rights and welfare of subjects, the research could not practically be carried out without the waiver or alteration, and when appropriate, subjects will be provided with additional pertinent information after participating.10 In recent years, there have been proposals to place less weight on the standard regulatory approach to informed consent for ROMP.11 These new proposals, however, seem to be at odds with the views of many of our study respondents regarding consent requirements for ROMP that uses randomization.
One study limitation is the 40% response rate and potential for nonrespondent bias. However, our sample is sufficient to establish the validity of our observations of variability. A second limitation is that PRIM&R members who have experience as IRB personnel may not be representative of all IRB members and staff members. PRIM&R members, however, are self-selected by their dedication to the protection of human subjects in research, and the group as a whole may be made up of a high proportion of IRB leaders. Finally, to identify any differences in participants’ responses to the survey questions that may be attributable to their perception of condition severity, we randomly assigned participants to review the hypertension scenario or an atrial fibrillation scenario. However, in retrospect, because the risks associated with the two medical conditions were similar in kind, any differences in participants’ responses between the two scenarios may have been too difficult to detect.
In spite of these limitations, the variability observed among respondents about randomization in ROMP suggests discomfort with research conducted within the clinical setting, which poses a challenge for researchers, policy-makers, and IRBs themselves. IRBs across the country are responsible for reviewing, approving, and monitoring research. Yet, for ROMP, they are charged with this task without clear rules or guidelines, resulting in inconsistent oversight for similar research scenarios. Variability in the areas of what should trigger IRB review, who should obtain consent, and when consent can be waived is problematic for two main reasons. Some IRBs may be more restrictive in interpreting current regulations, potentially slowing or preventing research that could eventually benefit patients. By contrast, having a less restrictive IRB could result in the failure to appropriately protect research participants. Finding the right balance is an ongoing challenge for IRBs, as illustrated by one respondent who chairs an IRB at an academic medical institution. This respondent admitted, “[W]e are struggling with this issue. We must solve [this issue,] as I think the future of healthcare and improved quality with decreased cost rides on the marriage of the clinical and research enterprises!”
Additional guidance for IRB personnel regarding activities at the intersection of clinical care and research, including ROMP and quality improvement, is essential to promote consistent interpretation of the regulations across IRBs. There have been attempts to refine and clarify how to apply regulations to these activities,12 but it is not clear at this time if these attempts at guidance will be implemented.13 Furthermore, if there is guidance, it needs to be carefully drafted to balance the interests of patients, IRBs, and the research community. Additional policy deliberation among IRB personnel, the research community, and the public to strike a balance that both respects participants and improves health care will be a necessary next step.
Kathryn M. Porter, JD, MPH, is a research associate at the Treuman Katz Center for Pediatric Bioethics at Seattle Children’s Research Institute; Mildred K. Cho, PhD, is a professor in the Department of Pediatrics and the associate director at the Stanford Center for Biomedical Ethics at Stanford University; Stephanie A. Kraft, JD, is a senior fellow at the Treuman Katz Center for Pediatric Bioethics at Seattle Children’s Research Institute and in the Department of Pediatrics at the University of Washington; Diane M. Korngiebel, DPhil, is an assistant professor in the Department of Biomedical Informatics and Medical Education at the University of Washington; Melissa Constantine, PhD, is a research associate in the Division of Health Policy and Management at the University of Minnesota; Sandra Soo-Jin Lee, PhD, is a senior research scholar at the Stanford Center for Biomedical Ethics at Stanford University; Maureen Kelley, PhD, is an associate professor at the Ethox Centre in the Nuffield Department of Population Health at the University of Oxford; Cyan James, PhD, is an AAAS Science and Technology policy fellow in Washington, DC; Ellen Kuwana, MS, is a senior communications specialist at the Treuman Katz Center for Pediatric Bioethics at Seattle Children’s Research Institute; Adrienne Meyer, MPA, is the assistant director for researcher support for the Human Subjects Division at the University of Washington; Douglas Diekema, MD, MPH, is a professor in the Department of Pediatrics at the University of Washington and the director of education at the Treuman Katz Center for Pediatric Bioethics at Seattle Children’s Research Institute; Alexander M. Capron, LLB, is a university professor and the Scott H. Bice chair in healthcare law, policy and ethics at the Gould School of Law, a professor of law and medicine at the Keck School of Medicine, and the codirector of the Pacific Center for Health Policy and Ethics at the University of Southern California; David Magnus, PhD, is a professor of medicine and biomedical ethics and the director of the Stanford Center for Biomedical Ethics at Stanford University; and Benjamin S. Wilfond, MD, is a professor in the Department of Pediatrics at the University of Washington and the director of the Treuman Katz Center for Pediatric Bioethics at Seattle Children’s Research Institute.
The authors thank Elisa Hurley and Kimberly Hensle Lowrance at PRIM&R for their help in coordinating the PRIM&R survey and all of the PRIM&R members who completed the survey, Steven Joffe at the University of Pennsylvania for his contributions to survey design and manuscript review, Philip Lavori at Stanford University and Thomas Gallagher at the University of Washington for their contributions to survey design, Isabelle Wijangco at Stanford University for her project management and research support, and Emily Rosenthal at Stanford University for her invaluable research assistance. This work was supported by the National Center for Advancing Translational Sciences at the National Institutes of Health, grants UL1 TR000423-07S1 (University of Washington and Seattle Children’s Research Institute) and UL1 TR001085 (Stanford University).
- Faden RR, Kass NE, Goodman SN, et al. An ethics framework for a learning health care system: A departure from traditional research ethics and clinical ethics. Hastings Center Report 2013;43:S16-S27; Whicher D, Kass N, Saghai Y, et al. The views of quality improvement professionals and comparative effectiveness researchers on ethics, IRBs, and oversight. Journal of Empirical Research on Human Research Ethics 2015;10:132-144.
- Candib LM. How turning a QI project into “research” almost sank a great program. Hastings Center Report 2007;37(1):26-30; Lynn J, Baily MA, Bottrell M, et al. The ethics of using quality improvement methods in health care. Annals of Internal Medicine 2007;146:666-673; Kim SYH, Miller FG. Waivers and alterations to consent in pragmatic clinical trials: Respecting the principle of respect for persons. IRB: Ethics & Human Research 2016;38(1):1-5.
- U.S. Department of Health and Human Services. Draft guidance on disclosing reasonably foreseeable risks in research evaluating standards of care. 2014. https://federalregister.gov/a/2014-29915.
- Weiss EM, Joffe S. Promoting informed decision making for comparative effectiveness randomized trials. JAMA Pediatrics 2015;169:803-804; Lantos JD. U.S. research regulations: Do they reflect the views of the people they claim to protect? Annals of Internal Medicine 2015;162:731-732; Shepherd L, Macklin R, King NM, et al. Informed consent for research on medical practices. Annals of Internal Medicine 2015;163(9):725.
- Lee SS, Kelley M, Cho M, et al. Adrift in the gray zone: IRB perspectives on research in the learning health system. AJOB Empirical Bioethics 2016;7(2):125-134.
- Casarett D, Karlawish JHT, Sugarman J. Determining when quality improvement initiatives should be considered research: Proposed criteria and potential implications. JAMA 2000;283:2275-2280; see ref. 2, Lynn et al. 2007; see ref. 1, Whicher et al. 2015.
- Cho MK, Magnus D, Constantine M, et al. Attitudes toward risk and informed consent for research on medical practices: A cross-sectional survey. Annals of Internal Medicine 2015;162:690-696; Kelley M, James C, Kraft SA, et al. Patient perspectives on the learning health system: The importance of trust and shared decision making. American Journal of Bioethics 2015;15(9):4-17; Kraft SA, Cho MK, Constantine M, et al. A comparison of institutional review board and patient views on consent for research on medical practices. Clinical Trials 2016;13(5):555-565.
- See ref. 4, Lantos 2015.
- Taylor HA, Kuwana E, Wilfond BS. Navigating parental permission for neonatal research. American Journal of Bioethics 2015;15(4):76-76; Eisenberg LR. Research involving premature infants: Timing is everything. American Journal of Bioethics 2015;15(4):79-80; Fiore RN, Cushman R. Informed consent and parental permission for research: Rules, roles, and relationships. American Journal of Bioethics 2015;15(4):77-78; Shah A, Porter K, Juul S, et al. Precluding consent by clinicians who are both the attending and the investigator: An outdated shibboleth? American Journal of Bioethics 2015;15(4):80-82.
- 45 CFR 46.116(d) (2005).
- Kim SYH, Miller FG. Informed consent for pragmatic trials—the integrated consent model. NEJM 2014;370:769-772; see ref. 4, Weiss and Joffe 2015; see ref. 2, Kim and Miller 2016.
- U.S. Department of Health and Human Services. Notice of proposed rulemaking for revisions to the Common Rule. Federal Register 2015;80:53931-54061; see ref. 3.
- Joffe S, Magnus DC. A flawed revision of the Common Rule. Annals of Internal Medicine 2016;165(2):143-145; Magnus D, Wilfond BS. Research on medical practices and the ethics of disclosure. Pediatrics 2015;135(2):208-210.