Ethical Conduct of Palliative Care Research: Enhancing Communication Between Investigators and Institutional Review Boards

Abstract

Palliative care has faced moral and ethical challenges when conducting research involving human subjects. There are currently no resources to guide Institutional Review Boards (IRBs) in applying standard ethical principles and terms – in a specific way – to palliative care research. Using as a case study a recently completed multisite palliative care clinical trial, this paper provides guidance and recommendations for both IRBs and palliative care investigators to facilitate communication and attain the goal of conducting ethical palliative care research and protecting study participants while advancing the science. Beyond identifying current challenges faced by palliative care researchers and IRBs reviewing palliative care research, this article suggests steps that the palliative care research community can take to establish a scientifically sound, stable, productive, and well-functioning relationship between palliative care investigators and the ethical bodies that oversee their work.

Introduction

From the outset of its scientific endeavors (1), palliative care has faced moral and ethical challenges when conducting research involving human subjects (henceforth termed participants) (2–4). The rationale for these ethical concerns has rested on several common perceptions, among them: that the palliative care population is especially fragile and vulnerable, thus warranting extra protection from potential exploitation; that the palliative care population is more likely than other potential study participants to be incapable of understanding research and/or accurately interpreting its conditions; and, that the end of life is a sacrosanct time, into which research activity may present an unwelcome and/or inappropriate intrusion (5–7).

The U.S. Code of Federal Regulations (45 CFR 46) defines an Institutional Review Board (IRB) as “a committee that performs ethical review of proposed research” (8). The role of the IRB is to ensure that two broad standards are upheld: first, that research studies do not place participants at undue risk (beneficence); and second, that participants give – free of coercion – their informed consent to participation (respect/autonomy). In the U.S., an IRB must approve any proposed non-exempt federally-funded research study before it can begin involving human participants; many institutions mandate that the regulations be applied to all research, regardless of funding source.

There are currently no resources to guide IRBs in applying standard ethical principles and terms – in a specific way – to palliative care research. Using as a case study a recently completed multisite palliative care clinical trial, this paper provides recommendations for IRBs and palliative care investigators to facilitate communication and attain a common goal of conducting ethical palliative care research and protecting study participants while advancing the science. Beyond identifying current challenges faced by palliative care researchers and IRBs reviewing palliative care research, this article’s purpose is to suggest steps that the palliative care research community can take to establish a scientifically sound, stable, productive, and well-functioning relationship between palliative care investigators and the ethical bodies that oversee their work.

Ethical Review of Palliative Care Research: Potential Challenges

Implicit in IRBs’ authority to approve, require modifications in, or disapprove any research activity involving human participants, is an important assumption – that IRB members understand the area of proposed investigation sufficiently to evaluate the ethics of the study. Federal regulations stipulate that IRBs must comprise at least five members, one of whom must have “primary concerns in scientific areas.” But the policy stops short of requiring a high level of substantive knowledge or background relevant to the specific study being reviewed, instead stating a general requirement that “The IRB shall be sufficiently qualified through the experience and expertise of its members” (9).

Given the relative nascence of palliative care research, IRBs may not be prepared to review palliative care clinical studies with discipline-specific scrutiny. IRB members may not be familiar with the unique features of palliative care research, such as emphasis on physical, psychosocial, and spiritual experiences of the participant rather than discrete physiological outcomes, including survival. In addition, the study population in palliative care studies is often heterogeneous, medically fragile, and may include family caregivers, characteristics that differ from more homogeneous and less frail study populations to which IRBs may be more accustomed. Unfamiliarity of IRB members with conducting research with the palliative care population coupled with palliative care researchers’ potential lack of appreciation of the IRB’s perspective may result in a convergence of misunderstanding. Interpretation of what is necessary to conduct ethical research and protect participants in palliative care research may differ. For example, IRBs may perceive a risk in a proposed study that is not immediately evident to the researchers. Palliative care researchers, given their prior clinical and research experience and familiarity with existing literature, may not fully explain what the IRB considers requisite protections for study participants. The IRB may request what appear to the investigators to be unnecessary and/or over-protective measures or undue scrutiny that cause delays in the protocol approval process. At the same time, IRBs may be frustrated with perceived lack of attention to potential risk and harm on the part of palliative care investigators (6). IRBs reviewing palliative care protocols have been described as “powerful gatekeepers” (10), at times imposing an “unjustly paternalistic” (6) attitude that denies palliative care patients the opportunity to participate in research.

Palliative care researchers may not effectively communicate with IRBs about distinct methodological approaches such as adaptation of study instruments to minimize respondent burden, use of study recruitment procedures that are sensitive to potential vulnerabilities during serious illness and after the death of a loved one, and engagement with proxy respondents (11–16). Exacerbating the problem is the lack of participation of palliative care researchers on IRB review committees, perpetuating the lack of content-specific expertise.

Rather than being viewed as a burden or intrusive, clinical research has proven to be consistent with many palliative care patients’ goals of care and personal wishes. In several studies, a majority of palliative care patients welcome the opportunity to participate in research (17–19), reporting benefits that include a sense of contribution to greater community good, a sense of meaning to life, a feeling of pride, and the opportunity to reflect on life and the illness experience (20–21). According to the Institute of Medicine 1997 report, Approaching Death: Improving Care at the End of Life, “Altruism is an important value for some people faced with irreversible illness” (22). Many people see participation in research as an opportunity to engage in an altruistic endeavor in the limited time they have remaining (23–24).

Also exacerbating potential misunderstandings between IRBs and palliative care researchers is the lack of clear terminology utilized in palliative care research; the field has yet to articulate standardized definitions for basic but commonly used terms such as “end of life,” “terminally ill” and “serious illness”, or even who is included in a “palliative care population.” The lack of a discipline-specific lexicon thwarts coordination and communication even between palliative care researchers; hence it is not surprising that there may be different interpretations by IRB members and investigators of the language used in palliative care research protocols.

Does Palliative Care Research Require Its Own Set of Ethical Principles?

The palliative care literature abounds with letters, commentaries, and reviews addressing ethical concerns related to palliative care research (1, 25–29). The field is deeply committed to 1) building its evidence base to reduce suffering and improve quality of life through thoughtful investigation, 2) upholding its focus on humane and compassionate care for patients and caregivers, and 3) identifying and resolving issues that could impede conduct of ethical research. For these purposes, palliative care investigators must develop strategies for proactively addressing – with efficiency, integrity, and rigor –ethical concerns that pertain to conduct of research in this population.

In 2002, the National Institute of Nursing Research (NINR) and the Office of Rare Diseases (ORD) co-sponsored a workshop, the purpose of which was to gather expert perspectives on whether or not palliative care research poses ethical issues that are unique, or different from, those arising in research involving other populations, and which would warrant special guidelines, procedures, or restrictions. Workshop participants organized ethical concerns regarding the design and conduct of palliative care studies into two broad categories: those related to informed consent, and those related to the potential impact of the study on participants in terms of risk, benefit (or lack thereof), and burden (30). These categories, and the general ethical issues they encompass, they found, were not unique to palliative care. Voluntariness of consent, with adequate attention given to decision-making capacity, and the balance of risks, burdens and potential benefits, represent the central ethical concerns in palliative care studies, just as they do for investigations conducted by other disciplines. The workshop’s conclusion that palliative care research should uphold the same high standards of ethical conduct as are applied to all other clinical research supported long-standing expert opinion in the field (31–33).

Concerns About Ethical Challenges May Negatively Influence the Conduct of Relevant, Generalizable Palliative Care Research

Research activity in palliative care has increased dramatically. Over the last 30 years, the proportionate number of palliative care publications increased fourfold, from 0.1% of all Ovid Medline citations to 0.4%. The preponderance of this activity, however, does not reflect interventional research. In 2005, only 7% of published palliative care and hospice articles (33) reported on clinical trials; a follow-on review of studies published in three leading discipline-specific journals in 2007 found that 81% of articles reported on studies that were not interventional (37). Although there are multiple potential contributors to the lack of intervention studies in palliative care, including the relative nascence of the field, until recently little dedicated research funding in palliative care and methodological challenges, it is also important to consider whether perceptions of ethical roadblocks are deterring applied clinical research in palliative care.

Of concern is the extent to which anticipation of difficulty obtaining IRB approval may dissuade investigators from aggressively pursuing important and clinically relevant palliative care clinical studies. Despite their clinical importance, topics perceived as ethically “difficult” may simply be avoided. Or, investigators may practice “IRB avoidance,” in which they deliberately design their protocols to minimize IRB scrutiny (36), thus avoiding what should be practice-driven and clinically meaningful research pursuits in an effort to avert ethical complaints. In addition, anticipated ethical concerns may influence study design, potentially decreasing generalizability of study findings. For example, concern about certain patients being “too sick” to participate in research may lead to exclusion of those patients – who may be the ones most representative of the population that stands to benefit from the evidence generated (37).

Case Example: The “Statin Study”

The “statin study,” a multisite, randomized, controlled trial of continuing vs. discontinuing statins in palliative care patients with limited prognosis, was the first clinical trial initiated by the Palliative Care Research Cooperative Group (PCRC). The PCRC is a national cooperative group funded in 2010 by the NINR to 1) advance high-quality, collaborative, comparative effectiveness and other clinically relevant research in palliative care; 2) establish a national infrastructure for palliative care research, to support high-quality, efficient, and timely studies; and 3) develop standardized, replicable, processes and methodologies that support the success, productivity, and sustainability of research cooperative groups across disciplines. The study rationale, PCRC founding principles and its development have been reported elsewhere (38–40).

Eligible participants for the statin trial were adults, diagnosed with advanced life-limiting illness, with an estimated prognosis of one to six months who were taking statins for primary or secondary prevention. The primary objective was to determine whether or not discontinuation of statins affects 60-day survival. Secondary objectives were to determine the impact of discontinuing vs. continuing statins on 1) incidence of clinically significant cardiovascular events, and 2) quality of life, performance status, anxiety and depression, symptoms, polypharmacy, satisfaction with care, and costs. Data were collected in person at baseline, and thereafter by telephone weekly through Week 4, every other week from Week 5 until death or six months (24 weeks), and monthly from Week 25 until death.

The study was designed for implementation in a palliative care population, addressing documented challenges to recruitment and retention. In addition to recognized barriers to enrollment (41–44), the investigators expected the study topic to evoke emotional responses, specifically, an emotional reaction to the concept of discontinuing a medication that the patient, caregiver, and/or provider believe to be protective or life-sustaining. Preparation for the trial included special training of study personnel and development of consent forms that attended to this and other anticipated concerns.

In this section we describe the lessons learned in the course of IRB review of this study at one of the two lead sites. Our experiences underscore the point that, ten years after the NINR/ORD-sponsored conference, and despite considerable increase in palliative care research activity, misunderstandings and communication gaps between IRBs and palliative care investigators persist. Ethical issues raised by the IRB, which were explored and resolved through ongoing dialogue between the IRB and study investigators, included: 1) consent, 2) participant burden, and 3) adverse events. These issues were amicably resolved to the satisfaction of both the IRB and the investigators through direct communication between the investigators and the IRB director and review panel chair. These extremely helpful face-to-face meetings entailed explanation and discussion of both the IRB perspective and the investigator perspective, clarification of language and definitions used in the protocol, and sharing of information from prior trials and the literature about conducting research with this study population. Specific issues and their resolution are detailed in Table 1.

Table 1.

Ethical Issues Raised by the IRB for the Statin Study, and Their Resolution

Issue	Resolution
Consent-related issues, in the setting of anticipated cognitive/physical decline
Assurance of ongoing informed consent (repeated consent at all data collection time points to ensure continued willingness to participate vs. consenting at baseline only)	Discussed the relative burden vs. benefit of repeating full consent process (which is what was requested) at each data collection point.
	Explained that the study design required following patients over time to capture the effects of the intervention on patient outcomes throughout their illness trajectory, including the primary study outcome, survival, necessarily spanning a time when the participant would be unable to make decisions and act as his/her own advocate.
	Clarified that the primary treating physician would continue to act as the study participant’s advocate throughout the course of the study, and would have the ability to override study randomization.
	Modified consent form to include language stating that a family member of the study participant should approach the treating physician and/or study team with any concerns about the study participant’s involvement in the study.
	Emphasized that all participants are followed closely for safety through the adverse event monitoring and reporting process.
Impaired functional status, cognition intact (cannot sign form but able to give consent)	Clarified that the clinical research coordinator (CRC), as part of obtaining informed consent, would assess for cognitive deficits or confusion. If the participant who is unable to physically sign the consent form provides his/her verbal consent, then the CRC and a witness would sign the consent form, and an explanation at the bottom of the consent form would indicate that compromised functional status (not cognitive status) made it impossible for the patient to sign the consent form, and that verbal consent was obtained and witnessed.
Identification of an appropriate advocate for study participants to make study discontinuation decisions once study participants are no longer able to make decisions for themselves.	Assured the IRB that, at each study visit, participants would be briefly assessed for their ongoing understanding of and interest in continuing to participate in the study.
	In the consent form, and during the informed consent process, it was made explicit that:
	Participants may withdraw from the study at any time, and family members, treating clinicians, and other patient advocates also have that right. The willingness to continue in the study is assessed at each data collection time point. If the participant became too tired, confused, or otherwise unable or unwilling to speak with a research staff member, investigators would obtain information about him/her from a loved one/caregiver (proxy) who interacts enough to speak on the participant’s behalf. The approval or disapproval of this approach was explicitly documented on the consent form, with “yes/no” checkboxes.
Palliative care patients are “very sick” and thus may be compromised in their capacity to decide	Assured the IRB that study CRCs receive specialized training for consenting this population. We also emphasized that we are respecting patient autonomy by informing people of research opportunities and inviting them to participate, that it is, in fact, counter to autonomy to a priori decide if a cognitively intact individual is or is not interested in participating in research.
Participant burden-related issues
Practical concerns related to data collection: What if the patient cannot hold the phone?	We implemented the following options: 1) use of speaker phone; 2) in-person data collection visit; 3) speak with a loved one on the phone who would then ask the participant the questions and then relay the participant’s responses to the CRC.
Use of telephone for follow up data collection – concern that patients would feel abandoned	Participants appreciated not having to come to the campus/clinic for a face-to-face appointment as it is physically draining and often requires them to ask someone to take them to the study appointment. Furthermore, as a multi-site national trial, the protocol must be implemented uniformly across sites. Some of our participating sites were in rural areas, which made it prohibitive for those sites to require in-person data collection at each data collection time point.
For which items is proxy response scientifically acceptable? Does the proxy need to be consented?	All data that may be collected from the proxy were objective data. Subjective data (e.g., symptoms, depression, quality of life, anxiety) were collected from patient participants only. Should a study participant decline physically or cognitively to the point of not being willing or able to provide study responses, we only obtained the data elements identified a priori as acceptable to obtain from the proxy respondent.
	The consent form clarified that should the patient become unable or unwilling to be interviewed over time, we would ask a loved one (that they identify) to obtain objective data. The proxy, providing only objective data about the participant, would not require separate consent as a research subject because s/he would not meet the definition of a human subject in research.
Issues related to adverse events
Updating with Suspected Unexpected Serious Adverse Reaction (SUSAR) rules	In September 2011, Federal guidelines in language and reporting requirements for suspected unexpected serious adverse reaction (SUSAR), unanticipated problem involving risks to subjects or others (Suspected UPIRTSO) were modified. The Project Director worked directly and proactively with the IRB Panel Coordinator to amend the study protocol, assuring a streamlined approval process.

What Is Needed: A Shared Understanding of Ethical Conduct of Palliative Care Clinical Research

Ideally, IRBs and investigators would openly communicate concerns and perspectives to facilitate simultaneous protection of research participants and conduct of rigorous research that will inform and advance the field, ultimately enhancing care for people with serious illness. Progress in assuring effective communication between IRBs and palliative care researchers must begin with agreement on shared intentions. Both investigators and IRBs need to recognize that 1) palliative care research is a national priority, and 2) IRBs have an important role to play in ensuring the quality and safety of palliative care research, through the provision of external, unbiased, ethical review. Everyone – investigators, IRBs, clinicians, patients and caregivers – stands to benefit from streamlined processes, clear roles and definitions, and a mutual purpose to accomplish shared goals. The goal for all constituents should be to see ethically sound studies generate new knowledge that can meaningfully improve the care of, and outcomes for, people needing palliative care and their family caregivers. Research studies must be safe, ethical, and impactful. Burden on patients and caregivers should be minimized, and research studies should be completed in an efficient and streamlined manner. A concise statement of this shared purpose could serve as a preamble to resources, such as guidance documents or definitions of terms, designed to facilitate palliative care research.

Blueprint for a Way Forward

Standardized methods and resources are needed that 1) educate IRB members about palliative care populations and research, 2) educate palliative care researchers about the application of ethical principles to palliative care research and IRB review, and 3) support better communication and coordination between palliative care investigators and IRBs. By acknowledging and addressing differences in interpretation of key terms used in palliative care research and concepts under IRB jurisdiction, palliative care investigators and IRBs are likely to come to a common understanding. Potential misunderstandings can be addressed, and likely ameliorated, by creation and use of 1) a lexicon of key terms in palliative care research, 2) a taxonomy of key potential IRB concerns as related to palliative care research, and 3) a guidance document for developing ethically sound palliative care study protocols.

First, we suggest the development of a lexicon of key terms in palliative care research to serve palliative care researchers, and the discipline more generally. This document would contain definitions of key terms that are expert-derived, standardized, consensual, and explicit. Derivation of these definitions would require significant input from various stakeholders, expert review, vetting, and usability testing; a process for regular review and update should be established. A first step in this process would entail itemization of terms requiring definition (e.g., palliative care, terminal illness, end of life); broad input from the field should help ensure that this list is exhaustive. Further steps would be to: develop definitions through a formally articulated process; suggest alternative definitions or an alternate approach when a term cannot be well-defined (e.g., define a study population’s prognosis rather than propose to enroll patients “near death”); vet the definitions and test their usability, refining them accordingly; and disseminate the lexicon.

Second, we suggest the development of a taxonomy of key potential IRB concerns as they pertain to palliative care research. The statin trial illustrated several of these concepts, among them: ensuring adequate informed consent, sustained participant advocacy throughout the trial, ability of the participant to adhere to the study protocol, and identification of adverse events. Methods to identify other concepts for delineation in the taxonomy might include canvassing the field, interviewing IRB members who have previously reviewed complex palliative care studies, and soliciting experiences of palliative care investigators. Included concepts should encompass all key safety concerns for which IRB members assume oversight, and which might arise in palliative care clinical trials specifically. Differing from the lexicon, the taxonomy would, with respect to each concept, define the potential IRB perspective as it relates to the IRB ethical mandate, describe potential perspectives of palliative care investigators (that may or may not differ from those of the IRB), point out different perspectives, identify potential areas of impasse when understandings differ, and suggest strategies for moving forward.

Third, we suggest the development of a guidance document for palliative care protocols. Targeted to both investigators and IRBs, this document would distill best practices for ethical conduct of research in palliative care populations, solicited across diverse settings, scientists, clinicians, academicians, and ethicists; it should be extensively vetted and refined, and regularly reviewed and updated. A consensus-building conference, which should include representation of patients, caregivers, clinicians, industry, human subject protections specialists, and public and private funders of research, could provide a venue for initial development of the guidance document; a standing working group might perform regular review/update. Standardization of the content and format of protocols in this field could help familiarize IRB members with research in palliative care, ensure that investigators adequately address all important elements (thereby helping improve the quality of research in this discipline), and support appropriate use of the terms in the lexicon, as well as attention to concepts in the taxonomy, so as to avert misunderstandings and requests for explanation or further information.

The lexicon of key terms in palliative care research, taxonomy of key potential IRB concerns, and guidance document could serve, on a national basis, as resources to standardize research in palliative care to facilitate the efficient, safe, and ethical conduct of clinical trials. They could set the stage for coordination and productive rapport between IRBs and investigators, and could serve as a training tool for all involved, particularly for junior investigators or investigators new to palliative care research. Data regarding the conduct of palliative care clinical trials – e.g., time to initiation, time to completion, number of protocol modifications requested by an IRB – could help determine whether or not the tools proposed (lexicon, taxonomy, guidance document for protocols) enhance the process and may support development of an IRB training module in ethical review of palliative care research.

Recommendations for Palliative Care Investigators

Pending development of the suggested resources, what can palliative care investigators do today to ensure clear communications with IRBs and expedient conduct of ethically sound palliative care research? Investigators can proactively frame the discussion when designing clinical study protocols, clearly defining their study population and terms utilized in the application, explicitly identifying areas of potential concern and proposing ways in which these concerns will be addressed. The inclusion of quality assurance processes, such as plans to measure patient safety, inconvenience, or data quality, into the conduct of the study also may be helpful. Palliative care investigators may need to pay particular attention to potential sources of coercion or undue influence in their methods of enrolling and engaging palliative care patients in research. A detailed description of the planned processes for working with, and protecting, this population may assuage the concerns of IRB reviewers. Clear, proactive, and precise communication between investigators and the IRB is essential to preventing misunderstandings before, during, and after the initial IRB review. IRBs and investigators must both be open to meeting and communicating outside of IRB panel meetings to broker understanding, and both investigators and IRBs should seek such communication as soon as concerns are raised.

Palliative care investigators should seek to understand, as fully as possible, areas of discomfort on the part of IRB members, while concurrently educating IRB members to better understand the palliative care population. It is critical for palliative care clinicians and investigators to become active on IRBs, providing expert review and insight, and offering ongoing education to IRB members. Additionally, understanding may be enhanced by creating a role for a palliative care patient advocate in IRB meetings, or in IRB/investigator discussions. The patient’s “voice” also may be strengthened by embedding, in the protocol of interventional studies, an inquiry into how participation in the current research impacts the patient and caregiver experience.

Above all, palliative care investigators must maintain integrity in their research methods, including the use of approaches that minimize bias and maximize generalizability of results. The quality of the evidence base in palliative care is at stake. Rigor and ethical considerations cannot be compromised for the sake of expediency, although study terms, funding plans, and budgets may need to be tailored to reflect realities faced by palliative care studies (e.g., time necessary for IRB review, realistic enrollment time frames).

Conclusions

Palliative care, a growing field that is gaining recognition for its vital role in the health care system, is actively working to develop its research infrastructure, methodology, and portfolio. Experiences in preparation for and conduct of the PCRC statin study illustrate the need for better coordination, more effective communication between, and education of palliative care investigators and IRBs about ethical conduct of research in the palliative care population.

Standardization presents one pathway toward instilling a high level of quality in future palliative care research studies. By establishing standards for the language and concepts used (lexicon of key terms in palliative care research and taxonomy of key potential IRB concerns, respectively), and for the design and conduct of studies (guidance document for protocol development), palliative care researchers could dramatically improve understanding between IRBs and palliative care investigators, thereby minimizing misinterpretations, resolving misunderstandings and disagreements that do arise, supporting communication during processes of inquiry, and facilitating study completion. These “deliverables” would elevate the quality of palliative care studies nationally, thereby helping build credibility for the discipline through development of a more robust evidence base. This latter benefit should not be underestimated. Palliative care occurs in a diversity of sites, ranging from large academic medical centers to nursing homes, inpatient hospices, and private homes. The small numbers of patients at any given site often require multisite recruitment to clinical studies; research capacity in many settings where palliative care patients receive care is underdeveloped. The creation of the proposed tools (lexicon, taxonomy, guidance document) could go far toward educating clinical personnel and those who are new to palliative care research, thereby enhancing research capacity and quality. Once developed, the resources’ impact, functionality and uptake should be examined and documented.

The National Institutes of Health, through an NINR cooperative agreement, has made substantial federal investment in the PCRC, including in the statin trial described above. The PCRC is well situated to collaborate with other palliative care research organizations, palliative care researchers, IRBs and other stakeholders to define, disseminate and implement methods to proactively address common ethical concerns related to conduct of palliative care clinical studies, facilitating efficient conduct of high quality palliative care research that adequately protects study participants.