A Commentary—Institutional Review Board Approval and Beyond

After reviewing the article by Dr. Phipps 2002, my initial responses are to congratulate her on both her convincing argument that health care providers do not need separate and specialized ethical standards to review end-of-life protocols and her comments concerning the very real ethical dilemmas that are faced by end-of-life researchers. Dr. Phipps' article makes a number of important points about why end-of-life research raises important, but not entirely distinct, ethical concerns. In my article, first I will briefly elaborate on reasons why I agree with Dr. Phipps and believe that it would be a mistake to develop separate guidelines for institutional review board (IRB) review of end-of-life research and comment on some of the special ethical issues she has highlighted. I will follow this with a discussion of why I believe it would be wise to think carefully about ways to educate end-of-life researchers about optimizing ethical practices and to prepare settings that provide end-of-life care to respond appropriately and ethically to requests to conduct research. Finally, I will propose that educators develop focused training programs and educational materials to prepare researchers for the ethical issues, and practical solutions to ethical and practical problems, that end-of-life researchers face, along with parallel efforts to educate and assist settings and providers involved in end-of-life care. I will briefly describe my own background, because our experiences certainly make a difference in our perspectives on these issues. I am a clinical psychologist, who has spent 20 years conducting research on stress and coping in older adults, particularly the experiences of ethnically diverse family caregivers of patients with Alzheimer's disease (Haley et al. 1995, Haley et al. 1996; Roth, Haley, Owen, Clay, and Goode 2001). I have been involved in a number of studies of family caregiving stress, bereavement, and end-of-life decision making in family caregivers (Allen-Burge and Haley 1997; Haley, LaMonde, Han, Narramore, and Schonwetter 2001; Owen, Goode, and Haley 2001) and have worked with colleagues at the University of South Florida (USF) to develop a Center for Hospice, Palliative Care, and End-of-Life Studies, in which investigators are conducting a variety of studies on end-of-life topics. We are especially fortunate to have two very large and successful hospice organizations, LifePath Hospice and Palliative Care and Hospice of the Florida Suncoast, as partners in this effort to build a center and to address barriers to research. Probably typical of most researchers who are moving into end-of-life research, none of my extensive coursework on research and ethics during graduate school, internship, and postdoctoral training even touched on end-of-life issues. As with most researchers, I have learned most of my lessons about special ethical problems in end-of-life research, and the associated practical problems, on the job. Through this work and that of my colleagues in our center, I have seen a number of examples of both superb proactive steps to deal with ethical and practical problems and difficulties that can occur without such forethought. The imposition of a separate IRB review protocol for end-of-life research would likely cause many more problems than it would solve. As Dr. Phipps has noted, current regulations already include a number of stipulations that address the major problems faced by end-of-life researchers. Such issues as informed consent and attention to vulnerable populations must be addressed by any application for IRB review. In addition, IRBs must already consider the risks and benefits of research projects, an area that can be particularly complex in end-of-life research (Casarett and Karlawish 2000). Imposition of a separate review system would create a bureaucratic nightmare. As writers such as Elizabeth Kubler-Ross 1975 and Ernest Becker 1973 noted some time ago, all of us are dying, but we commonly delude ourselves into thinking that we will live forever. By what standard would we consider a research protocol as requiring a special end-of-life review if special review standards were developed? Whereas some settings, like hospices, have patients who have been certified as having less than a 6-month life expectancy, most individuals do not die under hospice care. Hospitals are the leading site of death in the United States, but many doctors do not think of their patients as dying. Nursing homes are a growing site of death in the United States, with 19% of Americans dying in nursing homes, and many more older persons die after short stays in hospitals that follow nursing home placement (Moss 2000). Would studies in settings with high rates of death require special end-of-life review? Studies of older adults, particularly elders with chronic illness such as dementia, heart disease, and cancer, experience mortality of participants as a matter of course, even though end-of-life issues may not be the focus of study. Another concern might be whether studies of populations of people who are not terminally ill—such as family caregivers, hospice staff members, or even healthy people facing decision making about advance directives—would qualify as end-of-life research. Besides the problem of the "slippery slope" of which studies might require an end-of-life review, other difficulties are apparent. It appears that studies that might fit the category of "end of life" would likely raise different ethical concerns, for example, clinical trials of drugs that might yield complex trade-offs of quality and quantity of life, psychological studies of bereavement, and studies of college students reading vignettes about end-of-life scenarios. Finally, are there enough experts in end-of-life research to ensure that special guidelines were appropriately applied without unnecessary hassle and burden during review? Despite my agreement that end-of-life research is not entirely different from other types of research, there are important areas where there are special ethical problems and concerns in end-of-life research. Dr. Phipps has outlined a number of these problems, and I will touch on some of them. One topic discussed is that of informed consent. It is noted that informed consent should be required in end-of-life research, as in other areas. One important ethical issue that deserves more attention is the right of individuals to be somewhat uninformed. Researchers know that there are profound cultural differences in individuals' preferences to be told that they are dying (Koenig 1997). Investigators should be certain that informed consent does not improperly confront people with facts that they are psychologically unprepared to handle. The balance of providing helpful, but not unnecessarily stressful information, deserves careful attention. Another issue is that of vulnerability. As Dr. Phipps notes, individuals who lack legal competence cannot provide informed consent for research, and their decisions to participate in research should be made by a proxy. The interesting point is made that it is unclear what the standard should be for such consent; is it whether the dying individual would have wanted to participate? Another issue deserving additional attention is that mental capacity can wax and wane during terminal illness. Dementia and delirium are common and are often not detected in medical settings and in hospice (Lawlor, Fainsinger, and Bruera 2000). Does the researcher have an ethical obligation to assess cognitive impairment even though it is not noted clinically? For example, in an article from Dr. Susan McMillan's ongoing caregiver intervention trial based at USF (Allen, Haley, Small, and McMillan 2002), we have found that of 226 hospice cancer patients screened with the 10-item Short Portable Mental Status Questionnaire (SPMSQ; Pfeiffer 1975), more than 10% scored below 6 (indicating severe impairment). We chose to exclude such patients from our project because of concerns about their capacity to provide informed consent and their ability to provide valid self-reports. Even after excluding patients with severe impairment, 38.6% of these patients showed significant cognitive impairment on the SPMSQ (6–8 correct). The SPMSQ is a gross screening tool that certainly is an imprecise measure of cognition and capacity. Thus, another important question is how to best assess cognitive impairment and capacity in research protocols. Another excellent point made by Dr. Phipps is that research methods that are commonly considered to be low risk—such as interview and questionnaire studies that commonly qualify for exempt or expedited review—may have greater potential for complications in these populations. I agree completely with her point that we should go to special lengths to be sure that research assistants and students who are involved in data collection are well prepared to handle such complications as outpourings of grief. As I will describe later, we have begun systematic efforts to try to prevent such problems in our own setting. From my own experience in conducting research in this area, I am also convinced that we should carefully consider the idea of taking proactive steps to ensure that end-of-life researchers not only meet the adequate standards of human subject protection required by IRB review, but also avoid pitfalls that either place participants at risk or "poison the well" for other end-of-life researchers. Problems can occur even when projects do not pose real risks, but patients, families, or staff misperceive risks and benefits. I believe that attention to such issues may also serve to overcome some of the resistance to research by potential participants and "gatekeepers" that can make recruitment of participants difficult or impossible. In our end-of-life research center at USF, we have found that it is common for investigators new to end-of-life research to make easily preventable mistakes that can either cause ethical concerns or allow ethical issues to interfere with their projects. These issues can threaten the ability of investigators to complete projects that had been deemed meritorious enough to be funded (and with sufficient ethical protection to gain IRB approval) or can create problems that will carry over for other investigators. For example, investigators may attempt to use instruments or procedures that are too lengthy, burdensome, or repetitive, when alternative methods are possible (Casarett and Karlawish 2001). Some consent forms are technically informative, but do not engage potential participants in a way that encourages research participation. Investigators may not have made adequate plans for dealing with common problems, such as distress during interviews. All of these issues can be addressed by education and consultation with experienced end-of-life researchers. Another common problem we encounter is difficulty or inability to recruit for projects because of "gatekeeping" efforts by protective service providers who have ethical concerns. For example, in one study, a psychologist designed a project to study the issue of "terminal decline." In short, there is a sizable research literature showing that older adults in longitudinal studies who show drops in cognitive functioning are at increased risk for mortality (Small and Backman 1999). The investigator was interested in studying whether brief measures of cognitive functioning would be useful in predicting survival after admission to hospice. Hospice administrators encouraged the project because improved prediction of survival would be useful. Hospice staff members were educated about the project and were asked to identify patients who met the study's eligibility requirements and who seemed appropriate for the study. The patients would then be asked for consent to participate and receive a brief in-home cognitive assessment. Despite all of the encouragement and information offered by the investigator and the administration of the hospice, the project went several months without a single patient being nominated by the hospice social workers. An investigation revealed that part of the problem was that the hospice staff was concerned that the project would be overly intrusive for hospice patients to endure. The fact that the investigator was an "outsider" to the organization probably increased the perception that this might be an inappropriate burden for these patients. Staff members had essentially performed their own implicit risk/benefit analysis based largely on speculation about the project's effects on patients. A highly accomplished researcher directed this project with little specific experience in end-of-life, or specifically hospice, research. A hospice administrator who fully supported the project, but who also did not have a great deal of experience in the pragmatics of successful university/hospice collaborative research projects, had advised him. Discussion of this case and others in USF's monthly center meetings and with the hospice staff has led to practical suggestions to overcome the problem encountered in this study. A practical solution to this situation was to bypass general staff nominations and use a specific, motivated hospice staff member to identify and assess the patients. Of note, we have found that patients have been very receptive to the project, have enjoyed it, and have not found it intrusive. In several cases, patients approached over the objection of a hospice staff member consented to participate and experienced no ill effects from participation. In this case, exaggerated ethical concerns by the frontline staff functioned to remove patients' rights to even consider participation in the research. We believe that this case is part of a broader problem of overprotectiveness, suspiciousness, and perhaps paternalism that is commonly encountered among frontline hospice staff members. It is unclear to what extent these attitudes may be related to lack of knowledge about the benefits and ethical protections that accompany well-designed research or negative experiences with research that does not provide benefits and appropriate ethical protections. We believe that, ultimately, the staff will need extensive education on research ethics and information on research projects to be more willing partners in our research. One of the areas requiring greater education and dialogue with service providers is the importance of acknowledging that there is a real need for high-quality research on end-of-life issues to guide public policy and clinical practice. The Institute of Medicine (IOM) report "Improving Care at the End of Life" (Field and Cassel 1997) makes several important points about the ethical benefits of end-of-life research. First, lack of research on effective palliative care therapies may be a factor serving to decrease resources devoted to end-of-life care in an era in which evidence-based practice is increasingly emphasized (Bartels, Haley, and Dumas 2002). Second, as the IOM report states, "to deprive palliative care patients of the opportunity to make their own decisions about participation deprives them of the special meaning of doing something to benefit future generations. Altruism is an important value for some people faced with irreversible illness" (Field and Cassel 1997, p. 252). In our center meetings, we have discussed a number of other practical approaches that can both make ethical research practices likely and prevent ethical concerns from becoming issues. For example, we have developed written protocols for how researchers should handle problems such as discovery of elder abuse or suicide risk or encounters with clinical emergencies. To address the concerns of the hospice staff that clinicians might be unaware of research participation by their patients, LifePath Hospice and Palliative Care developed a policy in collaboration with our faculty requiring brief chart notes indicating participation in research (but no details that violate participants' confidentiality). We have also worked to standardize ways of recruiting research participants that avoid some of the initial fears of participants. We have found that recruitment letters and phone calls that mention not only the university, but also the hospice are preferred. Nametags that identify the researcher as working through the hospice should be worn. Both of these steps serve to clarify that the research has the support of the hospice. LifePath Hospice and Palliative Care has now developed its own detailed policies on research (LifePath Hospice 2001). These include requirements for a Bioethics Committee review of projects (independent of our university IRB), orientation of investigators and interviewers who will work with LifePath patients, as well as mechanisms for informing the staff about the nature of projects and the participation of patients under their care. Investigators are also expected to report the findings of their research to the hospice staff, avoiding the problem of "end-of-life safaris" mentioned by Dr. Phipps. In some cases, staff members have used research findings to improve clinical care, for example, improvement of pain management protocols and adding standardized measures for depression to intake assessments. The availability of such policies and procedures in other hospices and nursing homes would prevent many of the problems investigators experience in approaching these important settings for end-of-life care. We have also worked out an arrangement in which the hospice has allowed us to recruit research assistants from their trained hospice staff, making it relatively easy to put experienced interviewers, who are familiar with hospice policies and procedures, to work on our funded projects. This helps us avoid the ethical problems that can occur with sending relatively inexperienced interviewers into this highly charged environment. Interestingly, after a number of years of involvement in research at LifePath Hospice and widespread knowledge of the thorough review and coordination required of investigators conducting research in this setting, we believe that we may be seeing some change in the "culture" of the hospice in regard to research. For example, in a recent project focused on bereavement, clinical staff members were offered the opportunity to "red flag" any family members they thought might be inappropriate for participation in the project because of crisis or vulnerability; only about 5% of bereaved caregivers were so identified. Because staff members have come to expect useful clinical feedback from our research findings and are aware that ethical issues are receiving careful attention, building trust and cooperation is more readily accomplished. In summary, I believe that three important steps can be taken to directly improve the ethical practice and indirectly improve the feasibility and quality of end-of-life research. One obvious step is to develop specialized educational programs and materials for end-of-life researchers, with examples of forms and procedures that optimize ethical and effective research practice. Experienced end-of-life researchers have likely found effective ways to write consent forms, handle emotional distress by participant, and use short forms of instruments to decrease respondent burden; such materials could be readily assembled and disseminated to new investigators undertaking end-of-life studies. A second step that we should consider is developing model policies and guidelines for settings such as nursing homes and hospices, that may ease access to investigators and also protect participants. Ideally, we should avoid situations in which organizations afraid of ethical put barriers to research course, the ethical concerns may other issues, such as an to or where organizations do not all possible steps to the of participants. A as important, is frontline staff and caregivers about the benefits of research, and that they receive feedback about projects so that they can research to practice. for IRBs will not be effective in the ethical practice of research at the of The need is for experienced researchers who have already learned the difficult lessons of how to ethically conduct research in end-of-life settings to develop materials that can be readily disseminated to settings such as hospices and nursing homes, frontline providers that may serve as for projects, and investigators who should as as possible from the mistakes of The of have a to projects in this area of education about research ethics of which should in this not only in ethical but also in the practical details that make ethics and research work should be a major We should focus not only on the issues, but also on that are and readily so that research in this area can The is to Dr. Dr. Small, and Dr. Susan who comments on a of this