Research without Consent: Community ... - Wiley Online Library

16 downloads 27294 Views 114KB Size Report
laypersons who are trained and equipped with auto- mated external defibrillators ... MD, Department of Emergency Medicine, Mount Sinai School of Medicine, 1 ...
1082

Richardson et al.

d

COMMUNITY VOICES STUDY

Research without Consent: Community Perspectives from the Community VOICES Study Lynne D. Richardson, MD, Ilene Wilets, PhD, Deborah Fish Ragin, PhD, Jennifer Holohan, MS, Margaret Smirnoff, RN, FNP, MPH, Rosamond Rhodes, PhD, Gary Winkel, PhD, Maggi Rodriguez, Edmund Ricci, PhD Abstract Objective: To explore community attitudes toward the federal regulations that allow investigators to conduct emergency research without obtaining informed consent from participants. Methods: Focus-group participants were recruited from residential sites in New York City that were enrolled in the Public Access Defibrillation (PAD) Trial. The PAD Trial, a National Institutes of Health–funded, randomized trial in which laypersons were trained to treat cardiac arrest, was granted an exception from informed consent under these rules. Community residents and those who had been trained as lay responders in the PAD Trial were asked about the ethical issues raised by the conduct of research without consent (RWC), their definition of community, and appropriate methods of community consultation regarding RWC studies. Grounded theory content analyses were conducted on focus-group data. Results: Seventeen (40%) men and 25 (60%) women from 15 buildings participated in six focus groups: four in English; two in Spanish.

Definitions of ‘‘community’’ varied widely among and across groups; no strategy for community consultation was consistently endorsed by the participants. There were significant support and occasional adamant opposition to conducting RWC; participants often recounted specific personal experiences as the basis for both positive and negative opinions. Individuals with negative attitudes toward RWC often voiced strong support for specific RWC scenarios. Conclusions: There is no consensus regarding the definition of ‘‘community’’ or appropriate methods of consultation. Community members’ attitudes toward RWC are often shaped by their personal experiences, and their general attitudes often differ from their reactions to specific RWC protocols. Key words: community consultation; emergency research; exception from informed consent; human subjects research; informed consent; waiver of informed consent. ACADEMIC EMERGENCY MEDICINE 2005; 12:1082–1090.

In 1996, the federal government finalized regulations that allow investigators to obtain a waiver of or exception from informed consent for emergency research when certain very specific criteria are met.1–3 The research participants must be unable to consent as a result of their medical conditions; the experimental intervention must be administered before it is feasible to obtain surrogate consent; the condition being studied must be life-threatening; and participation in the research must have the potential to directly benefit the participant. These regulations, hereinafter referred

to as the emergency exception from consent regulations, also require that a number of special protections be provided before the study begins, including ‘‘community consultation’’ and ‘‘public disclosure.’’ The specific form and extent of these activities are left to the discretion of the institutional review board (IRB) and the investigator conducting the study.4–7 Public access defibrillation (PAD) is a community strategy to treat out-of-hospital cardiac arrest using laypersons who are trained and equipped with automated external defibrillators (AEDs). New York City was one of 24 sites in the United States and Canada to participate in the PAD Trial—a multicenter, randomized, controlled trial, funded by the National Heart, Lung, and Blood Institute, to test the hypothesis that PAD significantly increases survival in out-of-hospital cardiac arrest.8,9 Because obtaining informed consent from the patient or a surrogate is untenable in the case of sudden cardiac arrest, the PAD Trial protocol was granted an exception from informed consent for emergency research under the 1996 regulations.10,11 The PAD Trial was conducted from 2001 to 2003. New York City (NYC) PAD recruited 61 community locations including residential buildings and museums for the trial; approximately 40,000 individuals lived or worked in the facilities recruited. Participating

From the Departments of Emergency Medicine (LDR, IW, DFR, JH, MR), Nursing (MS), and Medical Education (RR), Mount Sinai School of Medicine (GW), New York, NY; the Department of Psychology, Montclair State University (DFR), Montclair, NJ; the Department of Environmental Psychology, City University of New York (GW), New York, NY; and the Department of Sociology, University of Pittsburgh (ER), Pittsburgh, PA. Received March 1, 2005; revision received June 12, 2005; accepted June 13, 2005. Supported by Grant # 1R01HL073387-01 from the National Heart, Lung, and Blood Institute of the National Institutes of Health. Address for correspondence and reprints: Lynne D. Richardson, MD, Department of Emergency Medicine, Mount Sinai School of Medicine, 1 Gustave L. Levy Place, Box 1620, New York, NY 10029. Fax: 212-426-1946; e-mail: [email protected]. doi:10.1197/j.aem.2005.06.008

ACAD EMERG MED

d

November 2005, Vol. 12, No. 11

d

www.aemj.org

1083

buildings were required to meet eligibility criteria regarding expected frequency of cardiac arrest; thus, the sites had a large number of persons more than 50 years of age.9,12 At each location, volunteers were trained to recognize and treat cardiac arrest, and a system was implemented to summon one of these trained individuals (referred to as ‘‘lay responders’’) whenever a cardiac arrest occurred on the premises.13 In half of the buildings, the lay responders were trained in cardiopulmonary resuscitation (CPR) only; in the other half, they were also trained to use AEDs. Thus anyone who lived in, worked at, or visited one of these facilities was a potential participant in the PAD Trial as a recipient of care; anyone who suffered a cardiac arrest while in one of these PAD buildings became a participant in the study. The Community VOICES (Views On Informed Consent in Emergency Situations) Study (funded by National Heart, Lung, and Blood Institute Grant 5R01HL073387) is a mixed-methods exploration of community attitudes toward the conduct of research when the medical condition being studied makes informed consent impossible. Community VOICES was set in the NYC residential buildings that participated in the PAD Trial. The first phase of the study was qualitative; focus groups were convened to assist in the development of a structured interview. In the second phase of the study, the interview will be administered at the PAD Trial buildings to collect empirical data for quantitative analysis. The specific aims of the Community VOICES Study were 1) to determine the attitudes of individuals who lived and worked in the PAD Trial community units toward a) the way ‘‘community’’ was defined and operationalized in the NYC PAD Trial, b) the processes of ‘‘community consultation’’ and ‘‘public disclosure’’ used by the NYC PAD Trial, and c) the relevant and appropriate definition of ‘‘community’’ for purposes of consultation and disclosure regarding research without consent (RWC) as required by the 1996 Emergency Research regulations; and 2) to determine the specific factors involved in judging RWC to be acceptable or unacceptable by community members. We report the findings of focus groups conducted with NYC PAD Trial community members regarding the conduct of research without obtaining consent, the specific provisions of the regulations that allow only life-threatening conditions to be studied, and only studies with the potential for direct benefit to participants to be conducted, the appropriate and relevant definition of community, and effective methods of communication for purposes of consultation and disclosure regarding such studies.

involving residents and lay responders from residential buildings that had participated in the NYC PAD Trial. Lay responders are individuals who were trained in CPR only or CPR1AED by the NYC PAD Trial staff. In most of the participating buildings, the lay responders were security, maintenance, or support staff that work in the NYC PAD building but live elsewhere; occasionally, the lay responders were building residents. Each focus group was scheduled to be 90 minutes long with six to eight participants. A purposive sampling strategy was used to identify and recruit focus-group participants who were representative of the NYC PAD buildings’ population with respect to age, gender, and race/ethnicity.12,13 Because 20% of the PAD units reported a large number of Spanish speakers, two of the six focus groups were conducted in Spanish. Building staff members who had served as liaisons for the PAD Trial identified interested prospective participants; phone calls were then made by Community VOICES Study staff inviting the prospective participants to the focus groups. Spanish-speaking prospective participants were contacted by bilingual Community VOICES Study staff. Separate focus-group sessions were held for residents and lay responders because it was hypothesized that their views about RWC might differ based on their level of involvement in the PAD Trial.14 Residents were involved in the PAD Trial only as potential patient participants. Lay responders had participated more actively: they were recruited by study staff, and gave informed consent to participate, before being trained as lay responders. They were organized to respond to emergencies within the building; they carried beepers, cell phones, or radios that allowed them to be summoned for this purpose; and they received regular communications from the NYC PAD Trial staff throughout the study. Focus groups were convened in a small meeting room at the Mount Sinai School of Medicine. An experienced facilitator, using established focus-group techniques to elicit opinions, conducted each focus group.15 One additional member of the study team was present as note taker. Focus groups for Spanish speakers were conducted by a bilingual facilitator. All focus groups were audiotaped and transcribed for data analysis. This study was reviewed and approved by the Mount Sinai School of Medicine IRB. Written informed consent was obtained from all focusgroup participants. Transportation to and from the focus group was provided by the Community VOICES Study. In addition, participants received a $40 NYC MetroCard or a phone card to compensate them for their time.

METHODS

Survey Content and Administration. The purpose of the focus groups was to assess the attitudes and opinions of individuals who live and work in the NYC PAD Trial community sites toward the relevant

Study Design and Population. From December 2003 through June 2004, six focus groups were conducted

1084 and appropriate definition of ‘‘community’’ for purposes of adhering to the emergency exception from consent research regulations, and to determine the specific factors involved in judging research without consent to be acceptable or unacceptable by community members. Four domains were identified to address the research questions: 1) informed consent for research, 2) concept of community, 3) research risks/benefits, and 4) research burdens. To explore the first two domains, a focus-group guide was developed that consisted of open-ended questions concerning the meaning of ‘‘community’’ and the acceptability of research without consent. These are summarized in Table 1. A set of scripted scenarios was developed to explore the third and fourth domains. In each scenario, participants were asked to evaluate a medical research study in which consent could not be obtained. In the scenarios, the level of risk of the proposed study intervention, the possible benefit of the intervention to the participant, the burden of the protocol for the participant, and the expected outcome of the medical condition being studied were systematically varied. For example, the PAD study had a high-risk, high-burden intervention (e.g., applying an electrical shock to the patient’s chest) with high potential benefit (may be lifesaving) in a condition with high mortality (the expected outcome is almost certain death). Treatment of critically ill or injured individuals often involves high-risk, high-burden interventions and so a PAD-like study was included in the scenarios. Since it was not possible to explore all

TABLE 1. Questions Asked of the Focus Group Informed consent d What do you think about medical research or medical experiments in general? d How would you feel about involving people in a medical research study without asking for their permission first? Would it be okay to get permission from a family member? Suppose there is no family member present and there isn’t time to contact anyone? d Suppose medical researchers wanted to include you in a study that might help you, but you were unconscious or too sick to give permission yourself? d Suppose that medical researchers wanted to include you in a study that would not help you, but might help other people in the future? Community d What is a community? What does community mean to you? d Which community or communities do you belong to? d One of the things researchers are required to do when they can’t get permission from individuals is to consult with the community before beginning the research study. Whom in your community should researchers talk to about this type of research study? d Once researchers are ready to start a study in which it is not going to be possible to get participants’ permission, how do you think researchers should let the community know about the proposed study?

Richardson et al.

d

COMMUNITY VOICES STUDY

TABLE 2. Content of the Scripted Scenarios Potential Expected Benefit Mortality to Patient None High None High

None None None High

Risk of Intervention

Burden of Intervention

None None Varied: none–high Varied: none–high Varied: none–high Varied: none–high High High

combinations of risk, benefit, burden, and mortality, we focused on those combinations that were most likely to illuminate the limits of consensus regarding the acceptability of research without consent. These are summarized in Table 2. The four English-speaking focus groups were conducted first; minor revisions to the focus group guide were made after the first two focus groups. The guide was then translated into Spanish by two bilingual members of the Community VOICES research team. An experienced Spanish-speaking translator, who was naı¨ve to the instrument, then back-translated the focus group guide into English; several minor discrepancies were reconciled. The two Spanish-speaking focus groups were conducted last using the Spanish version of the focus group guide. Data Analysis. All audiotapes of the focus-group sessions were transcribed; the tapes of the Spanishlanguage focus groups were transcribed in Spanish. Notes taken by the observer were used to assist with the transcriptions.16,17 Content analyses of the focusgroup transcriptions from the English-language focus groups were treated initially with a grounded theory approach.18–20 Text segments were manually sorted by individual investigators; emerging themes and categories were reviewed and discussed collectively; text segments were then resorted and assigned one or more thematic codes. The codes were then organized into conceptual categories by the research team. Another coding pass was then made through the data using the conceptual categories that had been developed inductively during the analysis. The Spanish transcriptions were analyzed by three bilingual (English–Spanish) reviewers using the conceptual categories developed during the analysis of the English-language focus groups as an a priori coding scheme. They then carefully searched for themes not raised in the English-language analyses.

RESULTS A total of 42 participants attended six focus groups: two groups with English-speaking PAD Trial lay responders, two with English-speaking PAD Trial building residents, one with Spanish-speaking PAD Trial lay responders, and one with Spanish-speaking PAD Trial

ACAD EMERG MED

d

November 2005, Vol. 12, No. 11

d

1085

www.aemj.org

TABLE 3. Demographics of the New York City Public Access Defibrillation (NYC PAD) Trial Buildings and Community VOICES Focus Groups Gender

NYC PAD population Focus-group participants

Race/Ethnicity

Age

Male

Female

White

African American

Hispanic

Other

,65 Years

$65 Years

35% 40%

65% 60%

43% 36%

29% 31%

23% 28%

4% 4%

37% 31%

63% 69%

building residents. Seventeen (40%) men and 25 (60%) women from 15 buildings participated. The gender, age, and race/ethnicity of the focus-group participants and of the NYC PAD population as a whole are summarized in Table 3. Informed Consent for Research. Participants were asked their opinions about medical research in general. Most had positive opinions about research: ‘‘it is important,’’ ‘‘we all benefit,’’ ‘‘necessary,’’ and ‘‘research is good.’’ Participants expressed appreciation for ‘‘improved treatments’’ and the ‘‘advancement of medicine.’’ Many were aware of examples of unethical research such as the ‘‘Tuskegee study,’’ ‘‘Hopkins,’’ and ‘‘LSD experiments by the military’’; though some thought ‘‘bad research is less common now’’ and ‘‘research ethics have improved.’’ Some participants voiced an understanding of the complexity of research: ‘‘research involves risks’’; ‘‘trial and error’’; ‘‘mistakes are unavoidable’’; ‘‘benefit to society versus the risk to the individual.’’ Others expressed ‘‘mixed feelings.’’ One participant stated that ‘‘research adds to health costs,’’ adding that much research was ‘‘high-cost, low-yield.’’ Another was concerned about the ‘‘conflict of the physician and researcher,’’ asking ‘‘Is their main responsibility to the research or to the patient?’’ When asked about involving people in a medical research study without asking for their permission first, there was often immediate and strong opposition: participation must be ‘‘voluntary,’’ ‘‘aware,’’ and ‘‘not forced,’’ with ‘‘no guinea pigs.’’ However, when presented with circumstances in which the person was too ill to consent, almost all participants supported the idea of getting permission from a family member or close friend, despite having previously spoken in opposition to the idea of research without consent. Only one participant opposed the concept of surrogate consent, adamantly asserting: ‘‘no one speaks for me.’’ Some participants thought that research subjects often lacked information and understanding about research: ‘‘patients never get complete, unbiased information,’’ ‘‘participants are not fully informed,’’ ‘‘randomized means you are not 100% informed,’’ and ‘‘participants don’t understand about placebos.’’ Others expressed frank distrust of researchers: ‘‘you have to read the fine print’’ and ‘‘private companies do it for profit.’’

Participants had difficulty when asked to consider what should happen when there is no family member present and there is not time to contact anyone. Conceptually, participants were unwilling to imagine a situation in which there was no one to give consent for them. Participants found it very difficult to accept that a family member or close friend would not always be available. In all groups, the facilitator had to repeatedly refocus the participants on this issue. The other conceptual problem was the participants’ tendency to discuss circumstances involving individual medical treatment rather than decisions about research participation. It required constant reminders from the facilitator to get participants to consider the issue of conducting research without consent. Once participants were successfully engaged in a discussion about research without consent, a variety of opinions surfaced in every group. There was no observed difference between the opinions of the residents and the lay responders. Many thought that research without consent ‘‘is okay if it saves the patient’s life,’’ ‘‘if there is a high chance it will help the patient,’’ and ‘‘if conditions are serious, research is important.’’ Others were more altruistic: ‘‘research is okay even if it doesn’t help you, but it helps someone else.’’ Some were highly trusting: ‘‘Leave it up to the doctor’’ and ‘‘if it’s an extreme emergency, the doctor can decide.’’ One participant warned: ‘‘avoid frivolous and irrelevant research.’’ A common sentiment was that ‘‘it depends on the situation.’’ In almost every group, at least one person took the position that research without consent should not be allowed: ‘‘it is not ethical,’’ ‘‘not legal.’’ Many participants expressed confidence in the fact that ‘‘research is regulated’’ and ‘‘there is a group that approves research first.’’ In several groups, it was suggested that there be ‘‘a patient advocate’’ to help decide. Many participants’ viewpoints were influenced by their own experiences with medical care or research involving either themselves or members of their family. One participant, who believed that no medical research should be conducted without the express consent of the individual, relayed a story about a medical procedure done on his family member without the knowledge or consent of the patient or his next of kin. Another participant, who thought researchers should be trusted to do whatever they thought best, was a cancer survivor who credited the experimental treatment she had received with saving her life. In fact,

1086 almost all participants who offered reasons for their views relayed personal experiences; very few offered abstract philosophical arguments. Concept of Community. Participants were asked to define ‘‘community’’ and also to describe the community or communities to which they belonged. The most striking finding of this discussion was the lack of consensus about what constitutes a community. No group, English-speaking or Spanish-speaking, articulated a shared vision of ‘‘community.’’ Some participants acknowledged being part of multiple communities, while others did not identify with any community. Most frequently, people explained community geographically with reference to their neighborhood, senior residence, or tenant/homeowner association. Some defined their community as New York State, as the United States, or as ‘‘where people live and work.’’ Some participants explicitly rejected a geographic view of community in favor of a focus on people. One declared that, ‘‘Community is not the same as neighborhood.’’ Another said, ‘‘Community isn’t physical; it’s people.’’ Others commented that community is defined by people who are ‘‘concerned about my wellbeing’’ or ‘‘a group where you feel you belong’’ or ‘‘like a support group’’ or an ‘‘extended family.’’ For some, the people in their community were defined by age; their community consisted of other seniors or other working-age individuals. Community was defined also by activities and interests: ‘‘anything that people have in common, that’s a community.’’ Frequently cited within this category were religion, ‘‘church community,’’ ‘‘work’’ communities, and labor unions. One man proudly described his 20-year experience as a bartender in a local watering hole frequented by authors and poets, ‘‘You are the head of a community, a big shot, an integral, dynamic part of things.’’ Others identified community in terms of a sports interest, for example, ‘‘the karate community.’’ Some believed that a shared experience defined a community: ‘‘all New Yorkers are part of the 9/11 community.’’ Some participants expressed a more global and enduring view of community. One participant remarked, ‘‘Community to me is that there will be another generation.’’ This thought was echoed by another who explained that the community is ‘‘what will continue after you are no longer there.’’ Although some focus-group participants had welldefined job descriptions and professional identities, no one described community in terms of these roles (e.g., social workers, police officers, door attendants). It was also noted that no one identified ‘‘community’’ based on his or her racial or ethnic background. When asked who in the community should be consulted before beginning a research study, a lack of consensus among the participants was again the only consistent finding. Health care professionals and

Richardson et al.

d

COMMUNITY VOICES STUDY

members of the clergy were most frequently suggested. Some participants suggested elected officials, and an equal number specifically rejected elected officials. There was a strong tendency to trust a deliberative group process rather than individuals, for example: church elders or clerical council rather than the pastor, community boards or precinct councils rather than an elected official, or senior groups or tenant associations rather than a board president. Several people voiced a wish to identify ‘‘true leaders rather than self-appointed ones.’’ To assure objectivity, several discussants cited the need for someone ‘‘with no vested interest in the study,’’ ‘‘someone not selected by the researchers,’’ or a community ‘‘ombudsman.’’ On the other hand, many participants believed that it would be necessary to have people who ‘‘understand the research’’ and who could judge the importance of the study from a medical point of view. No group spontaneously suggested that individuals with the disease being studied or their family members should be consulted. When asked specifically about this issue, some thought that it was essential that those affected by the disease and their families be consulted, but many participants believed that the broader community should be involved. In response to questions on how to notify people about a study that was going to be conducted in their community, the focus-group participants strongly favored low-tech methods of communication. They suggested announcements in local newspapers— particularly in small-circulation neighborhood-based papers, posting flyers on supermarket bulletin boards or in community rooms, or disseminating information through local community groups such as churches, synagogues, senior centers, and the YMCA. Few thought announcements on local public-access television would be effective, and only one participant suggested using e-mail or the Internet. Research Risks, Benefits, and Burdens. As a warmup to more complex scenarios, the participants were asked whether they would approve of a study that involved using a cotton swab to obtain sweat from all unconscious patients who arrived in a hospital emergency department. This would provide no benefit to the patient, but would pose no risk and essentially no burden. Almost all participants thought this would be acceptable if the researchers might learn something that would help future patients. A few participants had reservations about what the sweat would be used for, concerned that test results might be used in ways that violated the patient’s privacy. Only two of the 42 participants were adamantly opposed to even this most benign research protocol. In no case did the expected mortality of the patient’s condition enter into the participants’ decisions. In each of the next three scenarios, the focus-group participants were given a detailed description of a

ACAD EMERG MED

d

November 2005, Vol. 12, No. 11

d

www.aemj.org

specific research protocol detailing the patient’s condition and the expected outcome. The risks and burdens associated with the experimental intervention and the potential benefit to the patient, if any, were carefully described. In all cases, the patient’s medical condition had rendered him or her unconscious, no family member was available to give consent, and the intervention had to be performed immediately. The participants were told that the study could only be conducted without informed consent. At the end of each scenario, the participants were asked: ‘‘Do you think researchers should be allowed to conduct this study in your community?’’ In every group, the participants were initially reluctant to accept the idea that no surrogate was available. There was strong support for attempting to contact family members and, when possible, waiting for their arrival. Only when the facilitator reinforced the urgency and insisted that, despite the hospital staff’s best efforts, no family member could be contacted would the participants actually consider the scenario. In many cases, despite the detailed description provided, the participants sought additional, clarifying information about the scenario; the facilitator answered all questions by repeating the initial description without any additional data. Usually, with urging, the participants then gave their opinions about the acceptability of the scenario, but some participants gave conditional responses because of the absence of greater detail. As in the general discussion of informed consent, many participants’ viewpoints were influenced by their own experiences with medical research, or the experiences of family members. In the high-mortality, no-benefit, no-risk, no-burden scenario (Table 2), the focus-group participants were told that researchers were studying a condition in which ‘‘patients always arrive at the hospital in coma and they almost always die.’’ Participants were asked whether doctors should be permitted to use blood that had already been collected during the clinical care of the patient for research purposes. The facilitator then gradually increased the burden and risk of the research protocol: ‘‘Suppose blood had to be drawn specifically for the research protocol (one stick)?’’ ‘‘Suppose blood had to be drawn every 15 minutes for an hour (four sticks)?’’ ‘‘Suppose a small piece of tissue were needed (small incision with local anesthesia)?’’ ‘‘Suppose one-third of the patient’s liver had to be removed (major surgery with general anesthesia)?‘‘ A surprising number of participants voiced the opinion that doctors frequently used leftover blood for research without the patient’s knowledge or permission. ‘‘Doctors do that all the time!’’ ‘‘You don’t think they just throw it away, do you?’’ Nevertheless, most participants supported both the use of leftover blood and performing a single phlebotomy with no potential benefit to the patient. In some groups,

1087

serious objections were raised at the point of drawing four tubes of blood, other groups balked at the biopsy, and one group found everything, including the partial hepatectomy, to be acceptable if it might help future patients, commenting that current knowledge about medical cures was due, in part, to people’s willingness to put themselves at risk for the benefit of others. In the no-mortality, no-benefit, no-/low-risk, no-/ low-burden scenario, similar escalation of the risk and burden of the research protocol was given in the setting of a patient who was expected to survive, with no prospect of benefit to the patient. Although the focus-group participants were much more conservative in the level of risk they were willing to endorse than they had been with the high-mortality condition, many participants did find low-risk interventions to be acceptable. In the high-mortality, high-potential-benefit, highrisk, high-burden scenario, the focus group responses were highly varied. Many felt very strongly that research that increased a patient’s chances of dying should never be undertaken without that person’s express consent. Others believed that if the experimental procedure was the patient’s ‘‘only chance,’’ and the doctors thought it might help, the risky experimental intervention should be administered. Many participants declared themselves unqualified to weigh the risks and benefits involved in reviewing such a protocol. At the end of the first focus group, which consisted of PAD Trial lay responders, one of the participants asked why they had been selected to discuss this topic. The facilitator explained that the PAD Trial, in which they had all actively participated, had been a high-mortality, high-risk, high-potential-benefit, high-burden interventional study conducted without informed consent, just like the final scenario that had been presented. All of the participants were profoundly surprised; none of them had made the connection between the PAD Trial and the issues they had been discussing. The participant who had most vociferously opposed the idea of research without consent protested: ‘‘PAD wasn’t research, we were helping people!’’ The participants were then reminded that the PAD Trial had been a research study; the buildings had been randomized, and the individuals who collapsed in cardiac arrest had been involved in research without their consent. While acknowledging these facts, every participant, including those who had previously condemned research without consent as unethical, expressed strong support for the PAD Trial. The fact that the purpose of the PAD Trial was beneficent meant to the participants that it could not be ‘‘research,’’ which was, in their minds, suspect in its purpose. So interesting was this observation that at the end of each subsequent focus group, the facilitator pointed out that the PAD Trial was similar to the final

1088

Richardson et al.

scenario. In every group, the complete dissociation between the PAD Trial and the discussed scenarios observed in the first group was repeated.

DISCUSSION Since their release in 1996, the federal regulations allowing research without consent in certain emergency circumstances have been the subject of heated, ongoing debate among investigators, ethicists, and regulators. While some investigators have performed community consultation and provided public disclosure under the 1996 regulations,4,5,11,21–26 the number of studies that have sought an emergency exception from consent under the 1996 rules is lower than had been expected, perhaps because the provisions regarding community consultation and public notification are so problematic.5,27–30 Despite the inclusion of ‘‘the community’’ in the protective provisions of the regulations, very little input from the community has been sought in the ongoing debate. In 1997, Smithline and Gerstle reported a small survey of ED patients’ attitudes toward research conducted with an exception from informed consent.31 Their data suggest that there is significant discordance between the regulations as written and the attitudes of potential research participants. In a more recent study of ED patients and visitors, McClure et al. documented the ambivalence toward this issue that we observed in our study: 88% of their respondents said that people should not be enrolled in research when consent cannot be obtained, but 70% said they would not object to being entered into such a study.32 To the best of our knowledge, no previous study has systematically explored the attitudes of community members toward all of the provisions of the 1996 regulations. The Community VOICES focus groups were conducted to learn how community members thought about and talked about research without consent to guide the development of a structured interview to address these issues. It is clear from our focus-group findings that community members’ responses to questions regarding research without consent are driven to a great extent by the way the question is asked. General questions about the ethics of research without consent yield largely negative attitudes. However, specific questions about particular research scenarios yield very different answers, even from individuals who expressed opposition to the concept of research without consent. For most community members, this is not an easy topic to consider. They misunderstand the question being asked; they overlook important elements of the issue; they become confused by complex hypothetical scenarios. Some terms have strong negative connotations for certain individuals, independent of the specific facts presented. All of these factors were evident in our focus groups.

d

COMMUNITY VOICES STUDY

The regulations refer to ‘‘the community’’ as though it were a tangible entity that can be directly contacted. Our focus-group participants, however, were unable to agree upon specific, tangible community entities or representatives. On one hand, the absence of consensus about the definition of community may make the lack of specificity of the regulations the best approach. On the other hand, this lack of consensus about the definition of community and effective methods of communication makes it uncertain as to which choices made by investigators and IRBs will be supported by community members. The obvious reluctance of community members to delegate decision making regarding research without consent presents an important challenge to the research community to develop successful methods of communicating with communities. Investigators often conceptualize the relevant ‘‘community’’ to be consulted as the population of eligible subjects from which research participants will be drawn. While many of our focus-group participants believed that those who are most directly involved in the research as participants, or as family members or providers of participants, should be consulted, there was no support for the idea of limiting the definition of ‘‘community’’ to a cohort of eligible subjects. When asked to describe the communities to which they belong, the participants did not mention gender as a broad community affiliation (all women, all men), nor did they identify community by race/ethnicity, i.e., the African American community or the Jewish community. Perhaps the heterogeneous nature of the focus groups with respect to race and religion repressed such suggestions to avoid giving offense or violating social taboos. However, neither of the Spanish-language focus groups mentioned the Latino community as a distinct community. In fact, our data did not reveal any consistency of views on the definition of community or on informed consent based on the participants’ age, gender, race, ethnicity, language, status as a PAD lay responder, or assignment of building (CPR only vs. CPR1AED) in the PAD Trial. The only consistent predictor of participants’ views that was observed during the focus groups was personal experiences, whether good or bad, with researchers or health care professionals. Our data do not appear to support the limitations on research without consent imposed by the regulations. Most participants did not believe that the condition being studied had to be life-threatening or that the intervention being tested had to have potential benefit to the patient. Many expressed the opinion that low-risk, low-burden research without consent that held no benefit for the research participant was permissible if it might help someone else. Some were even more altruistic, endorsing higher research risks and burdens as acceptable. However some participants thought that the first obligation of the medical and research team was to maximize the patient’s

ACAD EMERG MED

d

November 2005, Vol. 12, No. 11

d

1089

www.aemj.org

chances of survival. Thus, research scenarios that offered no chance of benefit for the patient received less support than those in which the patient might benefit from the intervention. Many focus-group participants were prepared to allow risky experimental treatments if the research held even the slightest prospect of improving the patients’ chances of survival. Research without consent was regarded as acceptable in conditions that were not life-threatening, although risky interventions received much less support. The universal approval of the PAD Trial expressed by the participants did not seem to be an attempt at self-justification by individuals who had, after all, been involved in the PAD Trial, but rather a sincere endorsement of the PAD study. This illustrates the complexities of thinking abstractly about principles for guiding the ethical conduct of emergency research, and the difficulties in obtaining accurate information from community members about their opinions and beliefs. The use of certain words elicited disapproval, while a description of identical facts using different terms was met with approval. The Community VOICES research team is currently attempting to empirically test the hypotheses generated by these focus-group data, by using a structured interview.

LIMITATIONS This study gathered qualitative focus-group data, which were analyzed inductively to generate hypotheses. This is a technique designed to provide understanding and insight about areas of uncertainty and to allow people to speak in their own voices. Focus groups are regarded as a state-of-the-art tool in the development of valid and reliable survey instruments.33 It would not be prudent to form conclusions from these data without further empirical testing. The focus-group participants were selected to be representative of the population of individuals who lived or worked in certain high-rise buildings in NYC. While the participant group was racially and ethnically diverse, persons over the age of 65 years were disproportionately represented among the participants, because of the eligibility requirements of the PAD Trial. Transferability cannot be assumed. It is possible that other populations would yield different qualitative data providing different insights and generating alternate hypotheses. Our initial design called for separate focus groups for individuals from buildings that had been in the CPR-only arm of the PAD Trial and those from CPR1AED buildings. It is usually best to form focus groups that are homogeneous with respect to important characteristics,14 and we believed that the building assignment in the original trial might affect attitudes, just as we speculated that individuals’ status as PAD lay responders might affect their views.

This design feature, segregation by building assignment, was abandoned after the first focus group for both conceptual and logistic reasons. It became apparent that individuals’ experiences with the PAD Trial did not predict their views toward research without consent, and having so many strict criteria for focusgroup assignment made scheduling difficult. Nevertheless, the separation of lay responders and residents was maintained for all groups.

CONCLUSIONS Focus groups of laypersons revealed that there is no consensus regarding the definition of ‘‘community’’ or appropriate methods of consultation or notification regarding research without consent. Definitions of community varied widely among and across focus groups; no strategy for communicating with the community was consistently endorsed by the participants. There was significant support for, and occasional adamant opposition to, conducting research without consent. No consistent patterns were observed based on the participants’ age, gender, race, ethnicity, language, status as a PAD lay responder, or assignment of building (CPR only vs. CPR1AED) in the PAD Trial. The focus-group participants often recounted specific personal experiences as the basis for both positive and negative opinions. Individuals with negative attitudes toward the idea of research without consent often voiced strong support for specific research scenarios. The authors thank Laudy Rodriguez, Heidi Fieselmann, and Tara O’Reggio for their invaluable assistance with instrument development, data collection, and content analysis; and Marlene Taveras for her assistance with the manuscript.

References 1. Protection of Human Subjects; Informed Consent and Waiver of Informed Consent Requirements in Certain Emergency Research: Final Rules (21 CFR Part 50.24 and 45 CFR Part 46.101). Fed Reg. 1996; 61(192):51497–531. 2. Biros MH, Runge JW, Lewis RJ, Doherty C. Emergency medicine and the development of the Food and Drug Administration’s final rule on informed consent and waiver of informed consent in emergency research circumstances. Acad Emerg Med. 1998; 5:359–68. 3. Adams JG, Wegener J. ‘‘Acting without asking’’: an ethical analysis of the Food and Drug Administration waiver of informed consent for emergency research. Ann Emerg Med. 1999; 33:218–23. 4. Biros MH, Fish SS, Lewis RJ. Implementing the Food and Drug Administration’s final rule for waiver of informed consent in certain emergency research circumstances. Acad Emerg Med. 1999; 6:1272–82. 5. Kremers MS, Whisnant DR, Lowder LS, Gregg L. Initial experience using the Food and Drug Administration guidelines for emergency research without consent. Ann Emerg Med. 1999; 33:224–9. 6. Davis AM. Exception from informed consent for emergency research: drawing on existing skills and experience. IRB. 1998; 20:1–8. 7. Santora TA, Cowell V, Trooskin SZ. Working through the public disclosure process mandated by use of 21 CFR 50.24

1090

8.

9.

10.

11.

12.

13.

14.

15.

16. 17. 18.

19. 20.

(exception to informed consent): guidelines for success. J Trauma. 1998; 45:907–13. Hallstrom AP, Ornato JP, Weisfeldt M, et al. Public Access Defibrillation Trial Investigators. Public-access defibrillation and survival after out-of-hospital cardiac arrest. N Engl J Med. 2004; 351:637–46. Ornato JP, McBurnie MA, Nichol G, et al. The Public Access Defibrillation (PAD) Trial: study design and rationale. Resuscitation. 2003; 56:135–47. Hsieh M, Dailey MW, Callaway CW. Surrogate consent by family members for out-of-hospital cardiac arrest research. Acad Emerg Med. 2001; 8:851–3. Mosesso VN Jr, Brown LH, Greene HL, et al. Conducting research using the emergency exception from informed consent: the Public Access Defibrillation (PAD) Trial experience. Resuscitation. 2004; 61:29–36. Richardson LD, Gunnels MD, Groh WJ, et al. Community implementation of public access defibrillation: experience from the PAD Trial. Acad Emerg Med. 2005; 12:688–97. Ragin DF, Holohan JA, Ricci EM, Grant C, Richardson LD. Shocking a community into action: a social marketing approach to cardiac arrests. J Health Soc Pol. 2005; 20(2):49–70. Krueger RA, Casey MA. Participants in a focus group. In: Focus Groups: A Practical Guide for Applied Research, 3rd Edition. Thousand Oaks, CA: Sage, 2000. Krueger RA, Casey MA. Moderating skills. In: Focus Groups: A Practical Guide for Applied Research, 3rd Edition. Thousand Oaks, CA: Sage, 2000. Morgan D. Focus Groups as Qualitative Research. Thousand Oaks, CA: Sage, 1997. Patton MQ. Enhancing the quality and credibility of qualitative analysis. Health Serv Res. 1999; 34:1189–208. Miles MB, Huberman AM. Qualitative Data Analysis: An Expanded Sourcebook, 2nd Edition. Thousand Oaks, CA: Sage, 1994. Corbin J, Strauss A. Grounded theory research: procedures, canons and evaluation criteria. Qual Sociol. 1990; 13:3–21. Strauss AL, Corbin J. Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory, 2nd edition. Thousand Oaks, CA: Sage, 1998.

Richardson et al.

d

COMMUNITY VOICES STUDY

21. Shah AN, Sugarman J. Protecting human subjects under the waiver of informed consent for emergency research: experiences with efforts to inform the community. Ann Emerg Med. 2003; 41:72–8. 22. Lewis RJ, Berry DA, Cryer H 3rd. Monitoring a clinical trial conducted under the Food and Drug Administration regulations allowing a waiver of prospective informed consent: the Diaspirin Cross-linked Hemoglobin Traumatic Hemorrhagic Shock Efficacy trial. Ann Emerg Med. 2001; 38:397–404. 23. Erb T, Sugarman J. Ethical issues on informed consent and recruitment for clinical trials. Anesthesiology. 2000; 92:1851–2. 24. Marwick C. Assessment of exception to informed consent. JAMA. 1997; 278:1392–3. 25. Clifton GL, Knudson P, McDonald M. Waiver of consent in studies of acute brain injury. J Neurotrauma. 2002; 19:1121–6. 26. Biros MH. Research without consent: current status, 2003. Ann Emerg Med. 2003; 42:550–64. 27. Baren JM, Anicetti JP, Ledesma S, Biros MH, Mahabee-Gittens M, Lewis RJ. An approach to community consultation prior to initiating an emergency research study incorporating a waiver of informed consent. Acad Emerg Med. 1999; 6: 1210–5. 28. Kowey P, Ornato J. Resuscitation research and emergency waiver of informed consent. Resuscitation. 2000; 47:307–10. 29. Fish SS. Implementing the final rule. Acad Emerg Med. 1999; 6:1188–9. 30. Richardson LD. The ethics of research without consent. Mt Sinai J Med. 2005; 72:242–9. 31. Smithline HA, Gerstle ML. Waiver of informed consent: a survey of emergency medicine patients. Am J Emerg Med. 1997; 16:90–1. 32. McClure KB, DeIorio NM, Gunnels M, Ochsner MJ, Biros MH, Schmidt TA. Attitudes of emergency department patients and visitors regarding emergency exception from informed consent in resuscitation research, community consultation and public notification. Acad Emerg Med. 2003; 10:352–9. 33. Sofaer S. Qualitative methods: what are they and why use them? Health Serv Res. 1999; 34:1101–8.