Do Community Members Have an Effective Voice in the Ethical Deliberation of a Behavioral Institutional Review Board?

Abstract

Using concepts and methods from technical and professional communication and linguistics, the authors conducted an observational study of the voice of community members (CMs) in the deliberation of a behavioral institutional review board (IRB). In the discourse of deliberation, they found that CMs had an effective voice in constructing the compliance of individual research protocols under IRB review. But they also found that CMs had an ineffective voice in representing their African-American community, particularly in their efforts to advocate for more consideration of minority research sites and subjects and a fuller consideration of minority community attitudes.

Keywords

Institutional Review Board IRB nonscientist nonaffiliated NS–NA community members voice deliberation discourse analysis

In the United States, the National Research Act of 1974 coined the term Institutional Review Board (IRB) and authorized the first set of regulations governing the protection of human subjects in biomedical and behavioral research, Title 45—Public Welfare Part 46—Protection of Human Subjects (U.S. Department of Health, Education, and Welfare, 1974). These regulations specified that IRB membership could not consist entirely of “members of a single professional group” (i.e., scientists) or members who were “associated with the organization” (i.e., in situ researchers). Further, these regulations mandated consideration of “community attitudes” in the review of research protocols. At the time, researchers and universities hotly contested these regulations: In the comment period preceding their adoption, there were strident calls to eliminate or make optional such nonscientist (NS) and nonaffiliated (NA) members, strong objections to the vague charge of ascertaining community attitudes, and considerable anxiety that an IRB quorum could only be claimed when NA–NS members were present at a meeting, in effect “invest[ing] such persons with ‘absentee veto’ power.” Courageously, the Department of Health, Education, and Welfare regulators responded that the presence of “such persons” was essential to assure the independence and credibility of the IRB system.

The current U.S. Department of Health and Human Services (DHHS, 2009) Protection of Human Subjects regulations, known colloquially as 45 CFR § 46, have maintained the rationale and rules for IRB membership:

Each IRB shall have at least five members, with varying backgrounds to promote complete and adequate review of research activities commonly conducted by the institution. The IRB shall be sufficiently qualified through the experience and expertise of its members, and the diversity of the members, including consideration of race, gender, and cultural backgrounds and sensitivity to such issues as community attitudes, to promote respect for its advice and counsel in safeguarding the rights and welfare of human subjects. In addition to possessing the professional competence necessary to review specific research activities, the IRB shall be able to ascertain the acceptability of proposed research in terms of institutional commitments and regulations, applicable law, and standards of professional conduct and practice. The IRB shall therefore include persons knowledgeable in these areas. (46.107a)

Each IRB shall include at least one member whose primary concerns are in scientific areas and at least one member whose primary concerns are in nonscientific areas. (46.107c)

Each IRB shall include at least one member who is not otherwise affiliated with the institution and who is not part of the immediate family of a person who is affiliated with the institution. (46.107d)

But another term—community member (CM)—has long been in use in the IRB vernacular: After years of pointing out that CM was a nonregulatory term, current guidance from the U.S. Office of Human Research Protection (OHRP) finally endorsed a definition of a CM as “a single individual who fulfills both the NS and NA roles” (U.S. DHHS, 2010). The common practice nationwide has been to have one or two such CMs on each IRB (Anderson, 2006, p. 137).

A small literature has explored the role ambiguity and the related responsibilities of CMs on IRBs, making an enduring distinction between a layperson model and a community representative model (Veatch, 1975). In survey research, Porter (1986, 1987) found that the layperson model was a comfortable fit for the mostly biomedical IRBs of the time: Such CMs were praised for bringing a lay perspective to IRBs, and their primary responsibility was seen as reviewing consent forms for clarity. But Porter found that the community representative model was more problematically associated with representing community attitudes and requiring experts to be accountable to the public; specific responsibilities were not identified by Porter’s respondents (see also Allison, Abbott, & Wichman, 2008; Klitzman, 2012). In the most recent OHRP guidance on IRB membership, a rather stilted educational video, these now-reified distinctions remain. NS members were described quite favorably: “The value of a nonscientist’s role on the IRB is that they can review the research from the point of view of a layperson and look at the language in the informed consent document and the reading level.” NA members, however, were described more cautiously as ones who “serve an important role on the IRB as they can help represent the outside community,” hardly a description that licenses discursive power and consequential responsibility in IRB deliberation (U.S. DHHS, 2010).

A concern with the voice of CMs runs through the IRB literature. Porter (1986, 1987) identified having the courage to voice an opinion assertively and confidently as the top-mentioned survey category describing the ideal qualities of CMs. But Anderson (2006), among others, pointed to institutional and interactional barriers to CMs’ “abilities to make meaningful contributions to the ethical review process,” given that their “limited presence cannot realistically be expected to influence the [scientific and institutional] bias of the entire board” (pp. 136–137). In Anderson’s study, though, CM interviewees seemed more positive about their participation in deliberation; as one stated, “I feel like we [board members] all have the same job.…My voice is heard as equally as everyone else’s” (p. 148). Sengupta and Lo (2003) similarly found that the majority of their CM interviewees mentioned their goal was “to represent or give voice to the ‘community’ of human subjects” (p. 215) although they also found that CMs reported negative experiences interacting with faculty IRB members. Fitzgerald, Phillips, and Yule (2006) observed in an IRB meeting in New Zealand that Maori CMs “are often deferred to even if others did not seem to be in total agreement” (p. 383). And Stark (2012) described the deliberation of a CM’s suggestion for a change in a protocol’s research design: Faculty members on the IRB supported the argument of the principal investigator (PI) that the CM’s suggestion was not related to the goals of the study (pp. 66–70).

In a guidance report calling for increasing the number of CMs on IRBs, the U.S. DHHS (1998b) recommended there be “enough of them on a board so that their voices are more likely to be heard and their sense of belonging more likely to be enhanced” (p. 17). In a companion document (U.S. DHHS, 1998a), CMs themselves reported that “they brought a very different voice to the IRB’s deliberations, particularly when they raised nettlesome issues such as confidentiality or the PI’s efforts to reach out to diverse populations” (p. 18). Nevertheless, ambivalence and even suspicion about CM voice remain, even in the current IRB literature. Hearkening back to the comments on the first set of Protection of Human Subjects regulations (U.S. Department of Health, Education, and Welfare, 1974; see also Rothman, 1991), Amdur and Bankert (2011) warned that the voice of CMs “can represent a negative influence if they see their role as promoting a certain agenda,” even challenging the wisdom of the regulations: “Clearly there is a regulatory mandate that the IRB reflect community attitudes, but there may be times when these values run counter to the IRB” (p. 106).

In the IRB literature, no studies specifically examine CM voice, perhaps because the concept of voice is so difficult to define and operationalize in the science and social science literature. Technical and professional communication (TPC) researchers, however, are well suited to fill this gap because we have the interests and the infrastructure—our theoretical and methodological frameworks—to investigate IRBs as a unique academic workplace site of ethical deliberation and to pose the research question of whether CMs on this IRB have an effective voice in deliberation. To answer this question, we drew on the recorded and transcribed deliberation of a behavioral IRB over a 1-year period (12 meetings). For this study, we analyzed all deliberative exchanges with CM participation and developed an interactional definition of effective and ineffective voice in terms of the linguistic uptake of CMs’ concerns in deliberation.

In this article, we first describe the study site and participants and our data collection, coding, and analysis—a qualitative content analysis and a linguistic discourse analysis. Then, we present the findings of our analyses of CM participation and voice in the deliberation of this IRB, arguing that the interactional practices of deliberation could marginalize CMs’ efforts to function as community representatives and advocate for more consideration of minority research sites and subjects as well as minority community attitudes. In our discussion of these findings, however, we argue that a one-dimensional critique of these findings would be an oversimplification because the institutional reality of IRB deliberation requires extraordinarily close accountability to the scope of its authority. Given this context, we consider the implications of our findings at several levels: the interactional level of IRB meetings, the administrative level of institutional policy, and the federal level of regulations and guidance. We argue that the CM voice on IRBs could be strengthened by operationalizing a more robust principle of justice as originally suggested in the Belmont Report (National Commission, 1979), the philosophical document underlying the protection of human subjects in the United States. Finally, we acknowledge the limitations of this study and suggest directions for future research.

Method

Participants in this project served on the behavioral IRB at an urban public research university located in a city with a majority African-American population (> 80%). The project was approved by the university’s IRB in a process including a special review of the protocol by the assistant vice president for research compliance. All IRB members agreed to participate and signed consent forms.

The IRB members consisted of a chair, 10 faculty members, and two CMs. Both CMs (CM1 and CM2) were NS and NA: CM1 was a local minister, and CM2 was a retired teacher from a central city high school. The demographics of IRB members are presented in the Appendix.

Because the federal regulations deliberately decentralized the review of research with human subjects to locally constituted IRBs (DHHS, 2009), the structure and practices of IRB meetings can vary considerably across institutions, as Stark (2012), Fitzgerald et al. (2006), and others have noted. For example, institutions vary in the number of IRBs they operate, from one IRB responsible for all protocols to multiple IRBs responsible for delineated sets of protocols, as in the behavioral IRB studied here, which was one of the five IRBs at the research university site (the other four were biomedical IRBs). In the Appendix, we provide a table describing the number and characteristics of the protocols reviewed by this IRB during the year of data collection, including the total number of new IRB submissions, the number of protocols forwarded to the full board (FB), the disciplinary fields of protocol PIs, the primary and secondary reviewers’ risk assessment of the protocols (no greater than minimal risk, greater than minimal risk, or the risks-to-benefits ratio could not be ascertained), and their recommendations for the subsequent vote of the FB (approve, approve with specific minor revisions, table if the risks-to-benefits ratio could not be ascertained, or disapprove). We also describe the typical phase structure in the review of protocols during IRB meetings—Phase 1, the primary and secondary review; Phase 2, the open deliberation of all members of the board; and Phase 3, the outcomes of deliberation (risk assessment and vote recommendation)—and present a table with definitions and examples of CM participation in each phase.

Data Collection

The behavioral IRB met once a month, with meetings averaging almost 90 minutes (mean 87.3, median 86.0, range 48–147). During the year of data collection, 12 IRB meetings were recorded (17.5 hours), transcribed, and anonymized.¹ The minutes of these meetings were also collected and anonymized. To protect the confidentiality of participants, we use acronyms for CMs (CM1 and CM2), the designation of Chair for the faculty member presiding over the meeting, and the abbreviation FAC for other faculty members on the IRB. To protect the confidentiality of the IRB and the institution, we refer to research protocols in general or disguised terms only (e.g., an arthritis study, a study of medication adherence, and a survey of high school students).

From the transcripts, we excerpted all 55 exchanges in which CMs spoke. These exchanges could be as short as a truncated question with a one-word answer (e.g., Chair: CM2, your recommendation? CM2: Second) or as long as a detailed, multiparty deliberation of the protocol at hand.

Data Analysis

Our primary methods of data analysis in this study were qualitative content analysis and linguistic discourse analysis. To analyze CM participation in the deliberation of the IRB, we followed Hsieh and Shannon (2005) to develop a semidirected content analysis of CMs’ ethical concerns: A directed content analysis derives its coding categories deductively from previous research or theory whereas a conventional content analysis develops its coding categories inductively from the data. For our analysis of the uptake of CM concerns in deliberation, we developed a discourse analysis following Huckin’s (1992) schema:

Select an initial corpus—here, the transcription of 12 IRB meetings.

Identify salient patterns—here, CM participation and the uptake of CM concerns in deliberation.

Determine “interestingness” for the field—here, the voice of CMs in ethical deliberation.

Select a study corpus—here, all 55 exchanges featuring CM participation.

Verify the patterns through coding.

Develop a functional/rhetorical analysis.

Huckin described a functional–rhetorical analysis as the search for “a full answer [to the research question]…in a plausible, if tentative, explanation, build[ing] the most faithful representation [we] can” (p. 92).

CM participation coding

To describe CM participation in the deliberation of this IRB, we first counted the number of FB protocols with and without CM participation. We also counted the number of exchanges with CM participation in the three phases of protocol review (primary and secondary review, open deliberation, and outcomes).

To explore the ethical concerns that CMs raised in deliberation, we conducted a semidirected content analysis. For the directed coding categories, we turned to the federal regulations defining the scope of IRB authority and accountability in the statement of the Criteria for IRB Approval of Research (Protection of Human Subjects, 46.111, 2009). Deliberation on IRBs is intended to apply all and only these criteria in the review of research. We combined and adapted four relevant regulatory categories of IRB approval for our coding schema. We also developed a fifth coding category based on a conventional content analysis of exchanges in which CMs raised nonregulatory concerns. Table 1 presents definitions and examples of our CM participation coding schema. As Table 1 shows, the definitions of the coding categories are broader than what their regulatory language indicates: For example, for Code 2, the regulations mention subject selection only, but we broadened this definition to include research sites; for Code 3, the regulations mention consent only, but we broadened this definition to include recruitment as well.

Table 1.

Coding Schema 1: Community Member (CM) Ethical Concerns.

Code	Definition	Examples
Code 1	Concerns about risks and benefits 46.111.a.1: “Risks to subjects are minimized by using procedures which are consistent with sound research design.” 46.111.a.2: “Risks to subjects are reasonable in relation to anticipated benefits, if any, to subjects, and the importance of the knowledge that may reasonably be expected to result.”	CM1: I also noticed that they said something to the effect that somehow they were going to use the internet. […] OK. At least when I read it, it didn’t seem to be so clear. It looked like they–It didn’t seem to be so clear how they were going–That’s why I was–because they mention the fact is that there has been some form of violence by means of utilization of the internet. (Excerpt 03.01)
Code 2	Concerns about research sites and research subjects 46.111.a.3: “Selection of subjects is equitable…[taking] into account the purposes of the research and the setting…[and] the special problems of research involving vulnerable populations.” 46.111.b: “When some or all of the subjects are likely to be vulnerable to coercion or undue influence, such as children, prisoners, pregnant women, mentally disabled persons, or economically or educationally disadvantaged persons, additional safeguards have been included in the study to protect the rights and welfare of these subjects.”	CM1: I feel as though my posture would be–is that I would like to protect the particular person whom this study is being directed at, because my posture would be then is that I think this is why this type of committee has been formed in the first place–to protect those particular persons whom we’re investigating. And so therefore if it raises–as we have the fundamental questions that have been raised, then I have to say that A, we have to protect those who they are doing the investigation on. We have to–at least from my position is that I would have to stand to protect those whom we are researching. (Excerpt 13.01)
Code 3	Concerns about recruitment and informed consent 46.111a.4: “Informed consent will be sought.” 46.111.a.5: “Informed consent will be appropriately documented.”	CM2: I don’t know if this is out of order, but they really need to look at the title of this [recruitment flyer]. (Excerpt 31.01) CM1: The only thing there is a question in my mind as relates whenever you say that the [provider] was going to keep the participant from a legal point of view, which raises up in my mind the client-[provider] confidentiality relationship. And the nature of the questions, I think, is very important. The fact is if he asks those type of questions that touches on that client-[provider] relationship, then that can be a violation. So I think that should be said in the text [of the consent form]. Chair: So in this case [...] the [client] has the option of simply being able to opt out at any time. (Excerpt 41.01)
Code 4	Concerns about privacy and confidentiality 46.111.a.7: “There are adequate provisions to protect the privacy of subjects and to maintain the confidentiality of data.”	CM2: Everybody who knows who is in this study because they rarely miss– FAC: Right. Right. But they don’t know which data goes with which person, and that’s the research assistant, who will not be doing any of the analysis of that. The master list needs to be destroyed as soon as the data is linked up. (Excerpt 39.10)
Code 5	Nonregulatory concerns about research protocols not explicitly tied to the 46.111 criteria above (e.g., concerns about research design: background and literature review, research questions and hypotheses, data collection, data analysis; characterizations of researchers; sloppiness in preparing and proofreading protocols).	CM2: My only problem was with the size of the group, that they were taking only ten people, and I just didn’t think that was enough. […] Chair: I also think that this is–this research is at a developmental stage because their [goals] are to develop the intervention and then to deliver it and obtain feedback on it to see what kinds of changes [clients] might make. So a small sample size is not unusual in that kind of development phase. (Excerpt 04.01)

We achieved consensus on our coding of each CM concern, resolving discrepancies in team meetings. Of the 55 exchanges with CM participation, 6 exchanges were excluded from the content analysis because they raised no ethical concerns, as in sequences consisting of simple agreement only (e.g., CM2: You [primary reviewer] covered everything). The total number of CM ethical concerns for the content analysis was thus 49.

CM uptake coding

To answer our research question—whether CMs have an effective voice in the deliberation of this IRB—we developed an interactional definition of the term voice as a CM raising a concern in the deliberation of a protocol. We then defined effective and ineffective voice in terms of the linguistic uptake (positive or negative, respectively) of each CM concern in deliberation. We defined uptake, a term from early speech act theory (Austin, 1962; Edmondson, 1981), as a response linking back to a previous CM concern. Positive uptake validated the CM’s concern, and negative uptake did not. To operationalize uptake as a measure of effective and ineffective voice, we used the following definitions. If a CM’s concern was deliberated and designated to be part of the IRB’s formal communication with the protocol’s PI, we coded it as a positive communicative uptake, an obvious case of validating a CM’s concern and a clear instance of effective voice. If a CM’s concern was discussed and resolved within the IRB deliberation without any need to communicate with the PI, we coded it as a positive interactional uptake because it validated the CM’s concern, counting it too as an instance of effective voice. If, however, a CM’s concern was rejected or ignored in deliberation, we coded it as a negative interactional uptake, an obvious instance of invalidating a CM’s concern and thus ineffective voice.

To examine CM uptake in deliberation, we used the coding schema in Table 2. For each CM concern, we achieved consensus on our coding of the uptake category, resolving discrepancies in team meetings. The total number of coded uptake sequences was 47. This number does not equal the number of CM concerns in the content analysis (49) because 2 of the 49 concerns in the content analysis were each part of a two-concern sequence offered without pause, and only the second concern in each sequence was deliberated, a pattern that is not unusual in oral discourse (see Sacks cited in Jefferson, 1992, pp. 300–301). To further examine CM uptake, we developed a discourse analysis of positive and negative uptake sequences in deliberation.

Table 2.

Coding Schema 2: Community Member (CM) Uptake.

Code	Definition	Examples
Uptake 1	Positive communicative uptake: A CM’s contribution was deliberated and designated to be part of the formal IRB communication with the principal investigator (PI).	CM2: Did we address breaking up the compensation into steps so that it would not wait until the end? Chair: Yes, they must do that by university policy. […] And I’ll give them a little schedule. They need to give their laboratory incentive after the first visit. They need to give their incentives to fill out the [questionnaires] at the drop off and so on and so forth. They cannot keep the compensation until the full completion of the study. They might wish to adjust their amounts, and I can suggest that to the PI because sometimes they, you know, in order to get the good retention they offer more for the second part of the study than the first. But they must distribute it on an interim basis. (Excerpt 28.02)
Uptake 2	Positive interactional uptake: A CM’s contribution was deliberated and resolved without any need to communicate with the PI.	CM2: I don’t know if this is out of order, but they really need to look at the title of this [recruitment flyer]. Chair: [Internet Survey]. FAC: They gave a rationale for why that title is so nondescript. And that was to protect the participants so that if someone did find the survey on the computer or somehow was able to identify, it doesn’t identify the survey as going to [subjects]. They specifically say it’s a way of being vague on purpose so as to protect the confidentiality rights of participants. […] Chair: I agree with the rationale FAC provided. I think the fake title is appropriate in this circumstance. (Excerpt 31.01)
Uptake 3	Negative interactional uptake: A CM’s contribution was rejected for further deliberation.	CM2: Well, I came here interested in a list of all the places, they did not have any adequate means of informing parents of what adequate programs could be used of the things they already have in here that are not being–that could be done. And it could be done without the hoopla they have here that [at] that age could be done without the aura of values. I think that as soon as you say [behavior] it makes people curl up and want to go away [and] all you have to do is […] put it [out there] and there you go. Chair: Other comments? (Excerpt 37.01)

Findings

In presenting our findings, we describe the general features of the discourse of deliberation on this IRB and then present our analysis of CM participation and CM uptake in deliberation.

Discourse of Deliberation

IRB members, including CMs, often framed their contributions to deliberation using a variety of speech-act terms:

CM1:

Yes, because I was a little concerned when I saw that. (Excerpt 03.01)

FAC:

So our only concern really [is] the access to the list? I share [CM1’s] challenge in reading this in a comfortable zone. (Excerpt 35.02)

IRB members also framed their participation in terms of problems and questions:

CM2:

I agree with what you said, but my only problem was with the size of the group. (Excerpt 04.01)

CM1:

The only thing is there is a question in my mind. […] (Excerpt 41.01)

The most frequent framing terms were problem and concern, so we chose to characterize the overall discourse of deliberation on the IRB as concern-based because that is the broader term.

In IRB meetings, all members, including the chair, faculty members, and CMs, raised ethical concerns in the deliberation of research protocols. In deliberation, FAC often provided disciplinary expertise, and CMs offered both lay expertise and community representation. But it was the IRB chair who provided the regulatory and institutional expertise governing the discourse of deliberation, particularly with respect to the scope of IRB authority and accountability.

Although unique in academic and professional settings, IRB meetings share discourse features of academic review committees, hierarchical organizational meetings, and even classroom discourse. Many academic review committees deliberate on research and researchers, so deliberation on IRBs could resemble peer review in selection committees as described by Lamont (2009), with evaluations in terms of disciplinary norms, originality, design, significance, “interestingness,” and characterizations of researchers, both positive and negative. The discourse of IRB meetings also resembled hierarchical organizational meetings in that the chair directs and dominates the discourse authoritatively in speech acts that TPC researcher Bilbow (1998) characterized as “chair talk,” such as answering almost all questions and providing most direct responses to the deliberative concerns raised by IRB members. One effect of chair talk, Bilbow observed, was that “generally speaking, communication tended to flow through the chair rather than between one participant and another” (p. 174).

In this way, the discourse of deliberation on this IRB resembled classroom discourse: IRB members often raised their hands to speak, and many exchanges followed a reverse version of Mehan’s (1979) sequence of initiation, response, and evaluation. Thus, if an IRB member initiated a concern in deliberation, it was most often responded to and evaluated by the chair through uptake. Further, the level of participation by IRB members was variable, often resembling classroom discussion as described in Howard’s (2004) review of Karp and Yoels (1976).

The majority of contributors to class discussion will come from a very small number of students. Students come to accept and even expect this consolidation of responsibility for classroom participation often directing their attention to those few “talkers.”…Thus an instructor can leave a classroom thinking an effective discussion took place without recognizing that the discussion occurred primarily between the instructor and three to five students while the rest of the class primarily were onlookers. (p. 2)

The IRB chair observed for this study generally agreed with this description of the discourse of deliberation, acknowledging that she was dominant and noting that participation varied considerably: A few IRB members commented on almost every protocol being reviewed while others seemed to practice a pragmatic workload triage, participating mainly in the deliberation of those protocols for which they were the primary or secondary reviewers and answering disciplinary questions if directly asked by the chair or other IRB members (e.g., explaining the importance of inclusion–exclusion criteria in interventional studies, justifying standards for deception in social psychology, and providing a rationale for the use of focus groups on sensitive topics in education studies).

CM Participation Analysis: Constructing Compliance

In this section, we first bring forward three characteristics of the 41 protocols deliberated by the IRB during the year of data collection (see Appendix), report the number and phase of protocols with CM participation, and present our content analysis of the number and types of ethical concerns CMs raised in deliberation.

Protocol characteristics

The IRB chair forwarded 41 protocols to the FB based on the possibility of greater than minimal risk to research participants, 30 of which were from medicine and psychology. After the risk assessment of these 41 protocols, 33 were deemed to hold no more than minimal risk for participants; of these 33 protocols, 9 were approved outright while 24 were approved with specific minor revisions.

These findings point to three generalizations about this IRB’s review of protocols. First, behavioral research seemed to carry minimal risk, as evidenced by the small percentage of protocols forwarded to the FB (41 of 477, 8.6%) and by the final categorization of those FB protocols as posing no more than minimal risk to participants (33 of 41, 80.4%). Second, the majority of protocols forwarded to the FB were from either medicine or psychology (30 of 41, 73.2%), with fewer social science or education protocols (11 of 41, 26.8%). Third, the research protocols were generally approved either outright or with specific minor revisions (also 33 of 41, 80.4%).

These findings are consistent with the IRB literature. Although we found no empirical studies in the literature reporting the risk assessment of protocols by IRBs, OHRP guidance has long noted that “most behavioral research involves no physical intervention and no physical risk” and that the chief risk of behavioral research is “a risk of social harm (for example, harm to a subject’s reputation, which is sometimes a danger if confidentiality is not maintained)” (U.S. DHHS, 1993, chapter 5; see also Gunsalus et al., 2006). In calls to eliminate IRB review for social science studies, especially those using qualitative methods, IRB critics have claimed that these studies are always no more than minimal risk (see, e.g., Klitzman, 2015; Schneider, 2015; Schrag, 2010; see Hedgecoe, 2008, for a rejoinder). Regarding final IRB outcomes, Fitzgerald et al. (2006) reported that few or no protocols were approved outright by the IRBs they observed, with most returned to PIs for specific minor revisions (p. 389), another point of annoyance to IRB critics.

Fitzgerald et al. (2006) offered an interpretation of the frequency with which IRBs voted to approve with specific minor revisions: “There seems to be some need among committee members to make some comment or request some action,…[giving themselves] the sense that they are fulfilling their responsibilities as a committee” (p. 390), a sentiment echoed by IRB critics in harsher terms. While this observation characterizes the drive to critique in academic committees quite well (Lamont, 2009), we disagree that specific minor revisions to protocols are insignificant changes that have no regulatory motivation. Our explanation is that specific minor revisions were changes to bring a protocol into full compliance with the federal regulations and guidance as well as related institutional policies. In other words, specific minor revisions construct regulatory compliance.

The use of a master list in protocols is a good example of how compliance can be constructed by requesting specific minor revisions. Master lists are essential in most behavioral research: They provide the links between multiple categories of research data for analysis (independent and dependent variables for quantitative studies or thematic categories in qualitative studies). Master lists, however, also provide the means for potential social harm, making the loss of confidentiality (e.g., from stolen laptops to major breaches of an institution’s information technology system) a risk to participants. Such loss of confidentiality is a real risk in behavioral research studies that collect sensitive information from participants (U.S. DHHS, 1993, chapter 5). So one of the most frequent specific minor revisions requested by this IRB asked PIs to justify their need for a master list in their research design and to specify the disposal of the list in their confidentiality measures, as the following examples from the IRB minutes show:

Revise to indicate that the master list will be destroyed at the end of data collection when all data is linked. (Study 36) You indicate that a master list will be kept; however, with onetime data collection conducted by the PI, there seems to be no need to keep a master list to check for duplicate participants during a site visit [the PI’s justification in the protocol]. You are collecting sensitive information about […], so not keeping a master list would best protect participants’ confidentiality. (Study 11)

A review of the IRB minutes found that of the 41 FB protocols deliberated by this IRB, fully half (24 of 41, 58.5%) required specific minor revisions to bring the protocol into compliance for confidentiality, primarily by addressing the details and dispositions of master lists. Thus, in paying consistent attention to master lists, IRB members constructed compliance in consequential ways—protecting the PI in the case of an individual protocol audit and ensuring good confidentiality practices across the institution, a point that might be vetted in an institutional investigation or accreditation review. Thus, IRB requests for specific minor revisions were neither trivial nor unrelated to the scope of the IRB’s authority and accountability.

CM participation

All IRB members, including CMs, were involved in constructing IRB compliance during deliberation, in major and minor ways (see Table 3). Our findings point to three generalizations about CM participation in the deliberation of this IRB. First, CMs participated in the majority of protocols reviewed by the IRB (23 of 41, 56.1%). Second, the IRB chair often selected CMs to review protocols: On this IRB, faculty members with relevant disciplinary expertise were assigned to be the primary reviewers, but CMs were assigned to be secondary reviewers for a good number of the FB protocols (17 of 41, 41.5%). Third, in 55 exchanges over 23 protocols, CMs participated actively in all phases of deliberation: Nearly half of the exchanges (25 of 55, 45.5%) came from open deliberation, indicating that CMs did not speak only when assigned as a reviewer for a protocol. Finally, CMs’ concerns in deliberation were substantive, as shown by the content analysis of 49 concerns ranging over the criteria for IRB approval of research, especially for recruitment and consent (13 of 49, 26.5%) and research sites and subjects (11 of 49, 22.4%). CMs also raised nonregulatory concerns about research protocols (11 of 49, 22.4%), particularly about research design: Of these 11 nonregulatory concerns, the majority were related to research design (8 of 11, 72.7%).

Table 3.

Community Member (CM) Participation.

Total number of full board (FB) protocols	41
Number of FB protocols with CM participation	23
Number and types of CM participation by phase	55
Phase 1—Review (all secondary)	17
Phase 2—Open deliberation	25
Phase 3—Outcomes (risk assessment and vote recommendation)	13
Number and types of CM ethical concerns	49
Code 1—Risks and benefits	8
Code 2—Research sites and research subjects	11
Code 3—Recruitment and consent	13
Code 4—Privacy and confidentiality	6
Code 5—Nonregulatory concerns about protocols (e.g., research design)	11

These findings are not consistent with the literature describing CM participation on IRBs. In the one empirical investigation of CM participation on biomedical IRBs based on actual data from IRB meetings, Candilis et al. (2012) and Lidz et al. (2012) found that CMs participated as secondary reviewers for only 11.8% (11 of 93) of the protocols reviewed, compared to the 41.5% (17 of 41) of CMs as secondary reviewers on this IRB. This could partially be a function of the differences between a biomedical IRB and a behavioral IRB because behavioral protocols do not involve assessing the risks of physical interventions in medicine; however, this practice did indicate a consistent value accorded to the participation of CMs in reviewing protocols submitted to this IRB.

With respect to the ethical concerns of CMs, Lidz et al. (2012) found that the most frequent categories of CMs’ concerns in deliberation were consent and confidentiality, with fewer comments on other categories such as risks and benefits or subject selection. From their interviews, they also learned that CMs defined their membership in the traditional terms of a layperson model (see Veatch, 1975), focusing on the review of the consent form: One interviewee commented, “I read it with a viewpoint that I’m representing the layperson. See if they can understand it” (p. 6). Further, Lidz et al. (2012) specifically pointed out that these CMs did not define themselves as community representatives:

In their interviews, the CMs did not discuss the community in which they lived or their responsibility to it. The same was true of their behavior in the meetings. They made no reference to the local community or its values. (p. 6)

Lidz et al.’s findings were not true of the CMs on the IRB observed for this study: CMs were frequent reviewers and active participants in all phases of deliberation, not limiting themselves to the review of consent forms. Although we too found that matters of recruitment and consent were the most frequent category of concern for the CMs (13 of 49, 26.5%), on this IRB their primary focus was not on changes to the consent form; it was on broader concerns such as research sites and subjects (11 of 49, 22.4%). Further, CMs raised nonregulatory concerns about protocols (also 11 of 49, 22.4%), indicating that CMs contributed broadly to the deliberation of all aspects of protocol review. The CMs on this IRB thus seemed to define their roles in terms not only of lay expertise but also of community representation (Veatch, 1975), which complicated the discourse of deliberation on this IRB considerably. As we argue, when CMs raised concerns that constructed IRB compliance within an expanded layperson role, their concerns fit easily under the scope of IRB authority and accountability. But when one CM repeatedly raised nonregulatory concerns within a community representative role, her concerns did not always fit easily under the scope of IRB authority and accountability.

CM Uptake Analysis: Representing the Community

In this section, we present a short discourse analysis of the three examples listed in Table 2 in order to describe each uptake category in more detail. We then report the uptake frequencies in terms of effective or ineffective CM voice and present our discourse analysis of the similarities and differences between positive and negative uptake sequences.

CM uptake examples

Each of the examples in Table 2 illustrates some of the discourse features and functions characterizing the linguistic uptake of CMs’ concerns in IRB deliberation. Uptake 1 is a straightforward example of a positive communicative uptake and effective voice because CM2’s concern was immediately accepted (Yes, they must do that by university policy) and then explicitly validated when the concern was designated to be included in the formal communication with the PI (I’ll give them a little schedule). This example also illustrates an interesting must–might distinction in the types of communication IRBs have with PIs, one that sometimes moves beyond the scope of IRB authority. To construct institutional compliance in this example, the PI must make specific modifications to the distribution of compensation. But the chair went beyond the scope of IRB authority to suggest the PI might wish to follow a practice other PIs have used to achieve good retention, perhaps to soften the sting of the required revision by calling on her own experience with this key issue for PIs conducting longitudinal research. The chair thus not only accepted and validated the CM’s concern but elaborated it as a means of enhancing the IRB’s relationship with the PI.

Uptake 2 is also an example of a positive communicative uptake and effective voice in that CM2’s concern was deliberated and resolved without any need to communicate with the PI. In open deliberation, CM2 raised a concern about a recruitment flyer to be posted to various Internet sites, and the primary reviewer (FAC) provided information from the full protocol to indicate that the PI had addressed this concern by providing a rationale for its vague title, relating it to confidentiality, one of the regulatory criteria for IRB approval of research (Protection of Human Subjects, 46.111, 2009). CM2’s concern was thereby validated as one that was under the scope of IRB authority and resolved by the provision of information not previously covered in the primary or secondary review. As in Uptake 1, the IRB chair completed the deliberation and disposition of this concern: Despite the soft features of the chair’s alignment (I agree) and opinion (I think), this was functionally a ruling that CM2’s concern had been resolved in deliberation.

Uptake 3 is an example of a negative interactional uptake and ineffective voice in that CM2’s concern was rejected for further deliberation, again with the IRB chair acting authoritatively but this time without any validation whatsoever of CM2’s concern. This protocol was a parenting study with a population of African-American families, investigating a somewhat stigmatized parental behavior. CM2 began by voicing a concern that she expects to contribute to the deliberation—an assurance that parents would receive good information about community resources to assist them in this area of family life. (This behavioral IRB often suggested that PIs include information and referrals in their consent forms as a way to support participants during and after the research.) But CM2 then moved to her major concern about the protocol—a strongly worded, and perhaps race-related, reservation about the research—characterizing the entire protocol as hoopla, criticizing its aura of values, and pointing to commonsense parenting. This characterization of the protocol, however, did not fit easily into the criteria of IRB approval of research. In the discourse of deliberation, the chair used her ability to control the phase structure of protocol deliberation to completely shut down the possibility of any continued deliberation of CM2’s nonregulatory concern (Other comments?).

CM uptake frequencies

Table 4 presents our findings on the number and types of linguistic uptakes of CMs’ concerns in deliberation. In sum, of the 47 CM concerns deliberated by the IRB, 34 (72.3%) were instances of effective CM voice (positive uptake) and 13 (27.7%) were instances of ineffective CM voice (negative uptake).²

Table 4.

Linguistic Uptake of Community Member Concerns.

Total number of uptake sequences	47
Number and type of uptake
Positive communicative uptake	13
Positive interactional uptake	21
Negative interactional uptake	13
Number and type of voice
Effective voice (positive uptake)	34
Ineffective voice (negative uptake)	13

Positive uptake sequences

In deliberation, a clear majority of CM concerns were validated with positive uptake (72.3%), warranting our generalization that the CMs did indeed have an effective voice in the deliberation of this IRB. But positive uptake was very much the IRB chair’s discursive act, as we just described. Any board member, including CMs, could raise an ethical concern, but deliberation ended authoritatively through the uptake of the IRB chair based on her role as the regulatory and institutional expert on the authority and accountability of this IRB. Of the 34 positive CM uptake sequences, the IRB chair concluded them 26 times (76.5%) compared to the 7 times that CMs closed the sequences themselves (20.6%) and the one time a faculty member did so (2.9%).

The majority of CM concerns that met with positive uptake (29 of 34, 85.3%) were coded within the regulatory categories of IRB approval of research (see Table 3): risks and benefits (6), research sites and subjects (8), recruitment and consent (10), and privacy and confidentiality (5). Thus, positive uptake reflected a match between the CM concern and the scope of IRB authority and accountability. In other words, effective voice was highly related to constructing compliance. When CMs raised nonregulatory concerns about research design, such as the adequacy of the number of participants in a study or the relationship between the protocol study and its larger context, they received positive uptake only 5 times (5 of 34, 14.7%).

Negative uptake sequences

In contrast, the number of CM concerns with negative uptake was smaller but not insignificant (13 of 47, 27.7%). Negative uptake too was primarily the discursive act of the chair: Of the 13 negative uptake sequences, the chair concluded 10 (76.9%), and faculty members concluded only 3 times (23.1%). Also, in comparing positive and negative uptake sequences, we found a difference in the proportional distribution of the coding of ethical concerns (see Table 3). Like the majority of positive uptake sequences, the majority of negative uptake sequences (8 of 13, 61.5%) were concerns under the regulatory scope of IRB authority and accountability: risks and benefits (1), research sites and subjects (3), recruitment and consent (3), and privacy and confidentiality (1). But the single category containing the largest proportion of negative uptake sequences was nonregulatory concerns (5 of 13, 38.5%), a proportion considerably larger than that of nonregulatory positive uptake sequences (5 of 34, 14.7%). To be sure, the numbers for this comparison were small, but they suggested that negative uptake, or ineffective voice, was related to a mismatch between the CM concern and the scope of IRB authority and accountability. And the mismatched concerns in this study, we argue next, were related to representing the community.

The 13 negative uptake sequences were from seven different protocols. Four protocols had an exclusively African-American population, and six protocols focused on low-income families. In a focus not unusual for behavioral research generally (Seligman & Csikszentmihalyi, 2000), these protocols all investigated stigmatized behaviors, some highly stigmatized (e.g., substance abuse and attachment) and others less so (e.g., nutrition and exercise). Almost every one of the negative uptake sequences, however, involved CM2 (12 of 13).

Seven of these negative uptake sequences were fairly straightforward. In terms of ethical concerns (see Table 3), three negative uptake sequences involved recruitment and consent, and one concerned confidentiality. Three other negative uptake sequences concerned research design: Two were criticisms of the adequacy of protocol bibliographies, and one was concerned about the validity of stimuli being used in a protocol investigating victimization, which the chair acknowledged but definitively identified as outside the scope of the IRB:

CM2:

I think more than racism is the sexist part of it. And particularly the [parts] that they talk about is very sexist. And the females that are involved don’t know how to respond to it. And of course, my approaches are very different because I am the sister of eleven brothers. So that I would never take—but the average female would perhaps find herself in this victim—And it’s the victimization of the female and it’s a very sexist kind of thing that happens to take place in this [vignette from popular media]. […]

Chair:

I mean, I think he’s got a pretty narrow definition of racism in this selection, which seems to simply look at racially discordant settings. But again, that’s a design issue that he needs to discuss with his dissertation committee. (Excerpt 02.03)

Here, the IRB chair not only shut down further deliberation on the use of stimuli that could be seen as racist but also completely ignored CM2’s comments about sexism.

The six other negative uptake sequences all involved CM2 and were more highly charged invalidations of CM voice. Three of these sequences concerned research sites and subjects, two concerned nonregulatory research design, and one concerned risks and benefits.

In deliberation on a protocol having a population of young African-American elementary school students, CM2 seemed to call for direct or indirect benefits for minority participants:

CM2:

All of these issues take us back to the original thing that I said that I didn’t want to say, and that’s if it has nothing to do with benefiting the participants, it has to do with–it’s research.

FAC:

Well, most of the research we look at doesn’t benefit the participants.

CM2:

And it’s–

FAC:

And that line “There may be no direct benefits to you” is pretty standard in our Benefits [section on consent form].

Chair:

I think, [FAC], you are capturing something we see just in the process here. [FAC] saw that the deficiencies were sufficient enough, in her opinion, to warrant tabling. [CM2] saw it as specific minor revisions. So I think that reasonable people can disagree on this. (Excerpt 39.07)

This was one of the rare instances of IRB deliberation that came close to an argument, with a faculty member bluntly observing that direct benefits were neither required nor normative in behavioral research protocols. A new IRB chair attempted to avoid potential conflict by using the cliché reasonable people. He failed, however, and deliberation of this protocol remained contentious.

In deliberation of a research site and subjects for another protocol, CM2 raised a concern about minority subjects being offered fewer services based on their participation in an evaluation research protocol:

CM2:

But I have a few [concerns] because this is dealing with an at risk group of people because they’re talking about–

Chair:

Families at risk?

CM2:

Families at risk and mainly children at risk. […] But I would like to say that in that we’re dealing with a population, rather than talking about giving them less services because of the costs that’s involved. That was one of the first things they talked about, was the cost that was involved. […]

Chair:

But in terms of the participants, this amendment doesn’t place anyone at risk of receiving fewer services. It tests the intervention in a shorter period of time, but the people can ask for additional visits and go back into the [organization name] system. Now it has to be noted that [the organization] is not an open-ended social service provider. People don’t necessarily get everything that they ask for, but they are put back into the system, and then they’re treated within the system and the constraints of that organization. (Excerpt 27.01)

Here, the chair contradicted CM2’s concern about the possibility of less services by using condescending language (now it has to be noted), not allowing her a chance to articulate her perspective on the relationship between provider services and any benefits offered (or not offered) to participants in the research.

Finally, in two negative uptake sequences, CM2 raised nonregulatory concerns related to research design. We analyzed one of these concerns and its shutdown earlier (see Table 2, Excerpt 37.01): CM2 characterized a parenting research protocol as hoopla, which the chair simply ignored by asking an abrupt question directed to the rest of the board (Other comments?). The other sequence involved a protocol amendment requesting a change in the recruitment of low-income participants. In an exchange that the chair struggled to shut down, CM2 offered a dismissive characterization of protocol researchers as part of a sweeping request for them to start all over again as well as a call for the IRB to redeliberate the protocol:

CM2:

I think they need to start all over again. I really do. And I almost get the feeling that whoever is in charge thinks that we’re not–thinks that we’re not–that there’s a feeling that the people on this committee aren’t paying attention to what they are doing. And we are paying attention. Somehow or other, I would like to see them bring this to us with all of the things that we ask[ed] for when we sent it back to them the first time. Because they didn’t—you know the questions we sent back.

Chair:

Yeah.

CM2:

Do you have those questions?

Chair:

Yes. And I have–I went through them.

CM2:

Did they address them?

Chair:

Yes.

CM2:

They did address them.

Chair:

They did everything–they did–We [brought] up a lot of issues and let me try to remember to get all of them. We tabled that [original submission] and then the tabled response came back to the committee. And the committee either approved it with specific minor or it looks like they approved it [outright]. [looking at protocol file] Yep, it was approved upon response by both the reviewers and the committee. So the technical changes they’re making to the protocol are consistent with the protocol at large. That’s why I think, for example, that’s why we don’t have to look again at the [attachment task], which was a task we talked about a lot last time because we went through that and [they] made the changes [we] wanted. We approved it and they’re going to use the same task. So technically, what they’re proposing here is within the spirit of the protocol. As far as [other concern being deliberated] […] (Excerpt 06.02)

Despite the chair’s unsupportive acknowledgments of CM2’s concerns (a disrespectful yeah and one slightly elaborated and another unelaborated yes), CM2 persisted with her inquiries until the chair grudgingly provided a clichéd explanation of why it was not technically necessary to redeliberate (the spirit of the protocol). The negative uptake was fully accomplished with the chair’s abrupt transition into an uptake sequence about an unrelated concern (as far as […]), thereby assuring there would be no continued deliberation of CM2’s concern nor any response to her characterization of protocol researchers (whoever is in charge) as having a feeling that the people on this committee aren’t paying attention to what they are doing.

CM2 spoke often in the deliberation of this IRB (37 of 55 sequences, 67.3%), twice as much as CM1 (18 of 55 sequences, 32.7%), and she very much thought of her role as that of a community representative:

CM2:

I’ve served, really, to look out for those people. We are looking at the catchment of minority, low-income. (Excerpt 39.09)

But a recurring feature in her rejected concerns was her suspicion of researchers and research, not only in terms of IRB compliance, as in the previous example, but also in terms of researchers’ intentional selection of minority research sites and subjects, as shown in the following contentious deliberation of the protocol with African-American elementary school participants:

CM2:

One of the things that takes me back to that initial thing that I didn’t say, that I really don’t like to say, is that’s why a lot of the things in there, because the focus is on these certain kids. And I really–I suddenly know why we have [CM1], that you do need to protect the people who are in the study because there are some things that go past just because it’s research and they can get it by you. This is a minority area. This is a young area, and someone raised the issue why are they put[ting] it into these two [schools] closer to [the university]. The people who are in that area are more educated than people who live in my neighborhood, which is a poor neighborhood, where they don’t want to have it, and they would have to look out for the safety issues there, more than they have to look out–There are lots of innuendos that come in, nothing more than from a racial issue. And you have to be there. You have to be one of them in order to know what the problem is. I’m one of them. I know what the problem is.

[silence]

FAC:

I also want to point out that both letters of support indicate that there will be anonymity in this study. It suggests to me, and they’re both very brief, and they both basically say, yeah, whatever you think. (Excerpt 39.05)

In response to CM2’s direct and indirect expressions of suspicion and racism, deliberation on this protocol simply stopped, allowing silence to function as a negative uptake. After this pause, the primary reviewer completed the shutdown by changing the subject to an unrelated point about the letters of support provided by the school principals at the research sites for the study. In other words, once CM2 moved beyond constructing normative compliance to representing complex African-American community attitudes, she was silenced.

By voicing these concerns, CM2 brought up fraught historical and contemporary attitudes held by some members of the African-American community, using features from the African-American rhetorical tradition, including references to personal experience and cadences reminiscent of preaching (Heath, 1983; Rickford & Rickford, 2000). First was the charge of medical racism, invoking the history of medical experimentation on African-Americans without their knowledge or consent (Washington, 2006). Second was the expression of African-American suspicion toward institutions of medicine in general, suspicion that is well-founded, given the serious minority health disparities that have persisted to this day (HealthyPeople, 2020). In the deliberation of this IRB, however, CM2’s voice was ineffective in her attempts to advocate for her community (e.g., by calling for more consideration of minority research subjects and sites) and to grapple with the complexities of what it means for an IRB to consider community attitudes largely because the chair’s consistent use of negative uptake to shut down deliberation of nonregulatory concerns prevented meaningful engagement by the entire IRB.

Our uptake analysis revealed a distinct difference between CM1’s references to minority subjects and African-American community attitudes and those of CM2. When CM1 represented African-American community attitudes, which he did only twice, he situated his concern in terms of the significance of the research:

CM1:

[I]n looking at what she proposes to do is, to me, is a viable issue that needs to be addressed. Whether or not it should be addressed, as you say, [with] procedure this way is another thing. But when you really look at the substance of it, that’s a very viable issue that’s very important within the black community at this time.

Chair:

I agree. I think, as a topic it’s an important one. (Excerpt 23.05)

CM1’s representation of community attitudes received immediate positive uptake, easily fitting within the compliance-based authority and accountability of the IRB. But CM2’s representation of more complicated community attitudes, particularly in her expression of direct or indirect concerns about racism and suspicion of research and researchers, met with negative uptake—in part because her concerns were not easily related to constructing compliance under the scope of the IRB’s authority and accountability and in part because the faculty IRB members, including the chair, simply did not know how to move toward a respectful and meaningful engagement that would make CM2’s representation of community attitudes a matter of effective, and not ineffective, voice.

Discussion

The answer to our research question (i.e., Do CMs have an effective voice in the deliberation of this behavioral IRB?) is thus yes and no. Using CM participation as a measure of voice (see Table 3), we found that the CMs on this IRB participated in the deliberation of the majority of protocols and that they often participated as secondary reviewers—two practices that distinguish this IRB from other IRBs described in the literature (Candilis et al., 2012; Lidz et al., 2012; Stark, 2012). Further, these CM concerns in deliberation went significantly beyond Veatch’s (1975) description of the concerns of a layperson whose sole responsibility is to review consent forms: CMs on this IRB raised a wide variety of regulatory and nonregulatory ethical concerns.

Using CM uptake as a measure of effective and ineffective voice (see Table 4), we found that a large majority of the CMs’ concerns in deliberation were validated instances of effective voice when they contributed to the joint construction of the compliance of individual protocols according to the regulatory categories for the approval of research. In Veatch’s (1975) terms, however, it would not be accurate to say that the CMs on this IRB served in a full community representative role with meaningful uptake of their community-related concerns or their representation of complex community attitudes. From this perspective, our findings would seem to continue the long-standing findings of ambivalence and suspicion in the IRB literature—researchers suspicious of CMs, CMs suspicious of researchers—which began in the comments on the first regulations promulgated in the National Research Act of 1974 and continues to the present day (Amdur & Bankert, 2011; Fitzgerald, Phillips, & Yule, 2006; Stark, 2012).

Ending with a critique focused solely on the ineffective voice and interactional invalidation of CMs’ community-related concerns, however, would not fully describe or explain the discourse of deliberation on this IRB. IRBs, primarily through the chair, are accountable to all (and only) federal regulations and guidance regarding the protection of human subjects. The IRB chair is further accountable to institutional policies, practices, and consistency in the IRB review of protocols, no small task. The context of this close accountability underlies the chair’s sole focus on constructing compliance protocol by protocol in IRB meetings. From this perspective, nonregulatory concerns, such as CM2’s suggestion that a particular protocol should provide direct benefits to minority participants, seem problematic because there is no authoritative institutional or regulatory support to authorize such a change in a single protocol, much less all protocols. Similarly, the consideration of complex community attitudes, such as racism and suspicion, would also seem to have no place in the deliberation of individual protocol compliance.

We argue, however, that increasing and strengthening the voice of CMs in deliberation, as recommended in federal guidance (U. S. DHHS, 1998a, 1998b), is possible and that the research presented here points to feasible practices that IRB chairs can adopt in IRB meetings. Further, we argue that our research methods—particularly the discourse analysis of negative uptake in IRB deliberation—provide a discovery tool to identify and explore local community-related concerns and attitudes regarding institutional policy.

In IRB meetings, chairs can increase and strengthen CM participation by developing practices to proactively support CM voice in deliberation, such as assigning CMs as secondary reviewers in IRB meetings. The IRB chair observed for this study had such a practice in place, but the literature suggests that IRBs nationwide do not (Candilis et al., 2012; Lidz et al., 2012). Given the importance of the reviewers’ initial presentations of protocols, as argued by Fitzgerald et al. (2006; see Appendix), making CMs secondary reviewers more frequently is a concrete way to address the findings that CMs are not influential IRB members (Anderson, 2006; Sengupta & Lo, 2003; Stark, 2012).

IRB chairs can also increase and strengthen CM voice by implementing interactional practices to incorporate CMs more directly in the deliberation of protocols for which they were not assigned a reviewer role. Since IRB chairs control the phase structure of IRB meetings, including the transition from primary and secondary review to open deliberation, they could make a simple change in the structure of this transition in order to ensure CM participation for all protocols. On this IRB, the chair typically invited open deliberation with the question, Are there any other comments or questions about this protocol? This transition could be changed to contain two parts—a specific invitation for CMs to ask questions or express their concerns, if any, and then a general invitation to open deliberation:

Chair:

CM1 and CM2, do you have any questions or concerns about this protocol?

Chair:

Are there any other questions or concerns about this protocol?

This combination of assigning CMs as reviewers and changing the structure of the transition to open deliberation would thus ensure that CMs have a voice in the deliberation of all protocols deliberated at every IRB meeting.

Together these practices would operationalize the federal guidance recommendations so that CMs’ “voices are more likely to be heard and their sense of belonging more likely to be enhanced” (U.S. DHHS, 1998b, p. 17). These practices, however, would not necessarily guarantee that CM voice would be effective, particularly when CMs speak as community representatives, as our discourse analysis of negative uptake sequences has shown.

The discourse analysis of negative uptake in IRB deliberation identified important and complex community-related concerns and attitudes for this particular IRB: CM2’s concerns about the selection and use of minority subjects and sites of research, her advocacy for direct or indirect benefits for minority sites and subjects, and her representation of complicated attitudes of racism and suspicion regarding research in the African-American community. Although these are nonregulatory concerns that cannot be acted on in the deliberation of individual protocols in IRB meetings, a discourse analysis of negative uptake can serve as a discovery tool for identifying local community-related concerns and attitudes that might then be referred, explored, and (perhaps) addressed at the administrative level of policy development.

For example, in this study’s university site, the process of developing IRB policies is detailed and accredited: Any IRB member, including CMs, may forward policy drafts, questions, or suggestions to the university IRB administration. In practice, however, IRB staff members draft policies and vet them with respect to federal regulations and guidance as well as accreditation standards (Association for the Accreditation of Human Research Protection Programs, 2012). In this institution, the first step in developing IRB policy is an administrative review conducted by IRB chairs and administrators—specifically, the educational coordinator, the research compliance administrator, the associate director of the IRB, and the associate vice president for research integrity. There is no community representation at the policy level in this institution, which might well be the case on IRBs nationwide. Addressing this gap would seem to be an important opportunity for CMs to have a voice in identifying community-related concerns and exploring the nature and implications of community attitudes, possibly leading to institutional policies that could be systematically incorporated into the review and deliberation of all IRB protocols. In this university, for example, a policy-level discussion of concerns about minority sites and subjects could result in an elaborated set of questions on the form for IRB protocol submissions, which might prompt PIs to explain in detail their selection of minority sites and subjects and to consider the possibilities for direct or indirect benefits for minority participants and sites. In another example, a policy-level discussion of enduring community attitudes of racism and suspicion could result in an enhanced community outreach program for the IRB, one that features CMs speaking to the community as representatives of the IRB.

These practice and policy recommendations address locally salient issues for this particular IRB and its administration in a single university. Community-related concerns and attitudes for other IRBs and institutions would certainly vary, as would recommendations for responsive practices and policies—which is exactly the context set in place by the federal regulations requiring IRBs to be locally constituted. Identifying local issues through a discourse analysis of IRB deliberation, especially when deliberation is disrupted by negative uptake, would thus provide a method to discover salient community-related concerns and explore complex community attitudes at an administrative level within institutions. But developing and implementing meaningful community-related policies governing the review of all protocols at an institution can happen only if there is robust regulatory guidance at the federal level.

In the criteria for IRB approval of research, federal regulations are already in place for systematic consideration of minority sites and subjects within a vulnerable populations framework:

Selection of subjects is equitable. In making this assessment the IRB should take into account the purposes of the research and the setting in which the research will be conducted and should be particularly cognizant of the special problems of research involving vulnerable populations, such as children, prisoners, pregnant women, mentally disabled persons, or economically or educationally disadvantaged persons. (Protection of Human Subjects, 46.111.3, 2009)

But this regulation is fraught with undefined terms and uncertain implications. What does it mean to be “particularly cognizant”? What are these “special problems”? Who would qualify as “economically or educationally disadvantaged persons”? Should an entire ethnic group—such as African-Americans—be considered a vulnerable population? If so, is a standard of direct benefits from participating in research an appropriate expectation? Should a standard of indirect benefits be adopted, defined in terms of contributions to a vulnerable community? Is there a better way to approach equity other than by considering vulnerable populations and protectionism, which involves designating certain individuals and classes as having diminished capacity for autonomous informed consent?

Further, and even more fraught with uncertainties, how should community attitudes be addressed? How would multiple and complex variation within and across communities be represented? (E.g., in the city of this IRB, the largest minority community is African-American; however, there are large and diverse Latino and Middle Eastern communities as well, with potentially different attitudes toward medical and behavioral research.) What would be the effects on the research design of behavioral studies, especially for unfunded, minimal risk studies? How would a consideration of community attitudes be implemented by researchers as a meaningful effort—and not as another IRB burden? Without ongoing federal guidance that can be applied consistently to develop institutional policies and govern IRB deliberation, the marginalization of CM concerns might continue, as described here and in the IRB literature as well.

At the moment, federal guidance focuses primarily on individual subjects and vulnerable populations with respect to informed consent; guidance is silent on other community-based concerns and attitudes. But a final argument supporting the need to develop federal guidance covering a broader consideration of community-related concerns and attitudes comes from the Belmont Report (National Commission, 1979). When the National Research Act of 1974 first established regulations for the IRB review of research, it also established the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The National Commission was charged to “identify the basic ethical principles that should underlie the conduct of biomedical and behavioral research involving human subjects.” Of the three famous principles of the Belmont Report—autonomy, beneficence, and justice—justice is first defined philosophically (“An injustice occurs when some benefit to which a person is entitled is denied without good reason or when some burden is imposed unduly”) and then discussed historically (“For example, during the 19th and early 20th centuries the burdens of serving as research subjects fell largely upon poor ward patients, while the benefits of improved medical care flowed primarily to private patients”; National Commission, 1979, n.p.). In the Belmont Report, the Tuskegee research study is specifically referenced as an example of the continued history of injustice in medical research.

The single exemplification of justice in the Belmont Report, however, was subject selection within a vulnerable populations framework. The Belmont Report acknowledges the wider social causes of injustice—“social, racial, sexual, and cultural biases institutionalized in society”—but seems to release investigators and institutions from any responsibility to engage with social justice beyond attention to subject selection:

Even if individual researchers are treating their research subjects fairly, and even if IRBs are taking care to assure that subjects are selected fairly within a particular institution, unjust social patterns may nevertheless appear in the overall distribution of the burdens and benefits of research. Although individual institutions or investigators may not be able to resolve a problem that is pervasive in their social setting, they can consider distributive justice in selecting research subjects. (National Commission, 1979, n.p.)

But federal guidance, perhaps due to the overwhelming complexities of distributive justice, continues to operationalize justice solely in terms of subject selection and vulnerable populations without referencing root causes or exploring distributive justice in any detail. Distributive justice, however, has a long and strong connection to voice in deliberation: For example, Lamont (2009), following Habermas (1984), argued that just academic deliberation “necessitates that each participant be given full liberty to express her opinion without any reprisal, and that each be provided opportunity for full and equal voice” (p. 117).

The absence of social justice in IRB review of research has been challenged in the literature, especially in the last decade. In an oral history interview, Beauchamp (2004), one of the bioethicists on the National Commission, critiqued the narrow focus of justice: “I think justice is severely downplayed. There’s very little discussion of justice. The main application has to do with the selection of subjects.…So, justice becomes a distant third consideration [after autonomy and beneficence]” (p. 11). Another member of the Commission, the legal bioethicist Patricia King (2004), also observed that justice is “the forgotten principle” of the Belmont Report (p. 12). In “Justice Beyond Belmont,” King (2005) argued for reconceptualizing justice at the level of population groups: “A reasonable extension of the distributive principles of the Belmont Report requires a concern for the well-being of disadvantaged or powerless groups as well as safeguards against the abuse of particular persons” (p. 145). King saw the move to group well-being as crucial:

The objective is for groups [women, African-Americans, and others] to play a more central role, not just in the deliberations of IRBs but in the entire research process, including a role in setting the research agenda and determining what counts as beneficial research. (p. 145)

King also suggested a method for working toward this kind of social justice: “Attending to groups may thus require that we investigate and understand the particularities of the groups’ experiences in order to learn from them” (p. 143). She did not, however, provide any details of how such a method would be developed and implemented.

We contend that analyzing the effective and, especially, the ineffective voice of CMs in IRB deliberation is a method of answering King’s call for learning the particularities of community concerns and coming to a better understanding of complex community attitudes under an enhanced principle of social justice for IRB review and deliberation and for institutional policy development. A synthesis of such institutional policy development could then become the basis for federal guidance aimed at achieving more socially just deliberation on IRBs, thereby fulfilling the letter and spirit of the membership regulations:

The IRB shall be sufficiently qualified through the experience and expertise of its members, and the diversity of members, including consideration of race, gender, and cultural backgrounds and sensitivity to such issues as community attitudes, to promote respect for its advice and counsel in safeguarding the rights and welfare of human subjects. (Protection of Human Subjects, 46.107, 2009)

A robust principle of social justice supporting CM voice in deliberation, institutional policy, and federal regulations and guidance would begin to address the means toward this end of considering the welfare—defined as well-being—of research participants and their communities.

Conclusion

This discourse analysis of the deliberation on a behavioral IRB offers an interactional perspective on Veatch’s (1975) distinction between the roles of layperson and community representative that CMs play on IRBs, and it has implications for institutional practices and policies as well as federal regulations and guidance. But this study has two significant limitations.

First, this study was limited by its single-site design, which allowed only a local version of our research question. But the study was based on a unique, longitudinal corpus of actual ethical deliberation on IRBs, and it is the first observational study of ethical deliberation on a behavioral IRB. It is also the first study to offer a method for discovering community-based concerns and attitudes within the dual context of IRBs (federally regulated, locally constituted).

Second, this study lacked a comparative analysis of participation and uptake of CM concerns compared to those of faculty IRB members. This is the first study to offer an interactional perspective on the voice of CMs, looking at both their concerns and the uptake of those concerns by the FB. Unfortunately, we do not have the resources necessary to expand this coding for a comparative analysis across the entire data set of 12 meetings; however, we expect to design issue-oriented studies that incorporate data from all IRB members in future research.

Looking forward, this study is the first to use concepts and methods from TPC in an observational study of effective and ineffective voice in ethical deliberation, defining and developing a concept as old as ancient rhetoric and as new as the formation of contemporary writing studies (Elbow, 2007). More specifically, this study suggests that TPC researchers have valuable theoretical and methodological frameworks to productively investigate the discourse of deliberation on IRBs and develop the implications of such research. We imagine both specific and general directions for TPC researchers to investigate IRBs within the context of federal regulations and guidance as well as institutional policy and practices. For example, because the authority of the IRB to require changes in research design is the most hotly contested aspect of IRB review, we are now working on a study examining how IRB members actually deliberate research design; in this issue-oriented study, we will be able to compare the contributions of both CMs and FAC members. Other TPC research on IRBs could include a study of the discourse of IRBs as a documentary reality (Smith, 1974), a study of IRB reviewer forms as part of the checkbox culture of medicine (Gawande, 2009), a forest–trees study of IRB accreditation (Association for the Accreditation of Human Research Protection Programs, 2012), and more.

Footnotes

Appendix

Acknowledgement

Work toward this article began in a 2015 summer graduate seminar, including the authors as well as Deanna Laurette, Christopher Susak, and Nicole Varty. Deanna Laurette participated in the collaborative writing sessions leading to the initial submission sent to JBTC in August, 2016. Only the listed authors participated in the collaborative revising sessions leading to further manuscript submissions in November, 2016, and February, 2017.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Abbott

Grady

(2011). A systematic review of the empirical literature evaluating IRBs: What we know and what we still need to learn. Journal of Empirical Research on Human Research Ethics, 6, 3–19. Retrieved from http://doi.org/10.1525/jer.2011.6.1.3

Allison

R. D.

Abbott

L. J.

Wichman

(2008). Nonscientist institutional review board members at the National Institutes of Health. IRB: Ethics & Human Research, 30, 8–13.

Amdur

R. J.

Bankert

E. A.

(2011). Institutional review board: Member handbook. Sudbury, MA: Jones and Bartlett.

Anderson

E. E.

(2006). A qualitative study of non-affiliated, non-scientist institutional review board members. Accountability in Research, 13, 135–155. Retrieved from http://doi.org/10.1080/08989620600654027

APA Style Blog. (2015). Using italics for technical (or key) terms. Retrieved from http://blog.apastyle.org/apastyle/2015/04/using-italics-for-technical-or-key-terms.html

Association for the Accreditation of Human Research Protection Programs. (2012). Home page. Retrieved from http://www.aahrpp.org/

Austin

J. L.

(1962). How to do things with words. Cambridge, MA: Harvard University Press.

Beauchamp

T. L.

(2004, 9 22). Oral history of the Belmont Report and the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research/Interviewer: B. A. Schwetz. Retrieved from https://bioethics.georgetown.edu/wordpress/wp-content/uploads/cnn4/InterviewBeauchamp.pdf

Bilbow

G. T.

(1998). Look who’s talking: An analysis of “chair-talk” in business meetings. Journal of Business and Technical Communication, 12, 157–197. Retrieved from http://doi.org/10.1177/1050651998012002001

10.

Candilis

P. J.

Lidz

C. W.

Appelbaum

P. S.

Arnold

R. M.

Gardner

Garverich

… Simon

L. J

. (2012). The silent majority: Who speaks at IRB meetings? Institutional Review Board, 34, 15–20.

11.

Edmondson

(1981). Spoken discourse: A model for analysis. New York, NY: Longman.

12.

Elbow

(2007). Voice in writing again: Embracing contraries. College English, 70, 168–188. Retrieved from http://scholarworks.umass.edu/cgi/viewcontent.cgi?article=1006&context=eng_faculty_pubs

13.

Fitzgerald

M. H.

Phillips

P. A.

Yule

. (2006). The research ethics review process and ethics review narratives. Ethics & Behavior, 16, 377–395. Retrieved from http://doi.org/10.1207/s15327019eb1604_7

14.

Gawande

(2009). The checklist manifesto: How to get things right. New York, NY: Picador.

15.

Gunsalus

C. K.

Bruner

E. M.

Burbules

Dash

L. D.

Finkin

Goldberg

… Pratt

M. G

. (2006). The Illinois white paper—Improving the system for protecting human subjects: Counteracting IRB mission creep (SSRN Scholarly Paper No. ID 902995). Rochester, NY: Social Science Research Network. Retrieved from http://papers.ssrn.com/abstract=902995

16.

Habermas

(1984). The theory of communicative action ( Thompson

Held

, Trans.). Cambridge, MA: MIT Press.

17.

HealthyPeople. (2020). Disparities. Retrieved from https://www.healthypeople.gov/2020/about/foundation-health-measures/disparities

18.

Heath

S. B.

(1983). Ways with words: Language, life, and work in communities and classrooms. New York, NY: Cambridge.

19.

Hedgecoe

(2008). Research ethics review and the sociological research relationship. Sociology, 42, 873–886. Retrieved from http://doi.org/10.1177/0038038508094567

20.

Howard

(2004). What does research tell us about classroom discussion? In Howard

(Ed.), Discussion in the college classroom: Applications for sociology instruction (pp. 2–8). Washington, DC: American Sociological Association.

21.

Hsieh

H. F.

Shannon

(2005). Three approaches to qualitative content analysis. Qualitative Health Research, 15, 1277–1288. Retrieved from https://journals-sagepub-com.web.bisu.edu.cn/doi/pdf/10.1177/1049732305276687

22.

Huckin

T. N.

(1992). Context-sensitive text analysis. In Kirsch

Sullivan

P. A.

(Eds.), Methods and methodology in composition research (pp. 84–104). Carbondale: Southern Illinois University Press.

23.

Jefferson

(Ed.). (1992). Harvey Sacks: Lectures on conversation. Oxford, England: Blackwell.

24.

Karp

Yoels

(1976). The college classroom: Some observations on the meaning of student participation. Sociology and Social Research, 60, 421–439.

25.

King

P. A.

(2004, 9 9). Oral history of the Belmont Report and the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research/Interviewer: L. B. Walters. Retrieved from https://bioethics.georgetown.edu/wordpress/wp-content/uploads/cnn4/InterviewKing.pdf

26.

King

P. A.

(2005). Justice beyond Belmont. In Childress

J. F.

Meslin

E. M.

Shapiro

H. T.

(Eds.), Belmont revisited: Ethical principles for research with human subjects (pp. 136–147). Washington, DC: Georgetown University Press.

27.

Klitzman

(2012). Institutional review board community members: Who are they, what do they do, and whom do they represent? Academic Medicine: Journal of the Association of American Medical Colleges, 87, 975–981. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3549463/

28.

Klitzman

(2015). The ethics police: The struggle to make human research safe. New York, NY: Oxford University Press.

29.

Lamont

(2009). How professors think: Inside the curious world of academic judgment. Cambridge, MA: Harvard University Press.

30.

Lidz

C. W.

Simon

L. J.

Seligowski

A. V.

Myers

Gardner

Candilis

P. J.

… Appelbaum

P. S

. (2012). The participation of community members on medical institutional review boards. Journal of Empirical Research on Human Research Ethics, 7, 1–6. Retrieved from http://doi.org/10.1525/jer.2012.7.1.1

31.

Mehan

(1979). Learning lessons: Social organization in the classroom. Cambridge, MA: Harvard University Press.

32.

National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. (1979, 4 18). The Belmont Report: Ethical principles and guidelines for the protection of human subjects of research. Retrieved from http://www.hhs.gov/ohrp/humansubjects/guidance/belmont.html

33.

National Research Act of 1974, 42 U.S.C. 289 § 474. Institutional review boards; Ethics guidance program. Retrieved from https://www.gpo.gov/fdsys/pkg/STATUTE-88/pdf/STATUTE-88-Pg342.pdf

34.

Porter

J. P.

(1986). What are the ideal characteristics of unaffiliated/nonscientist IRB members? IRB: Ethics & Human Research, 8, 1–6. Retrieved from http://doi.org/10.2307/3564370

35.

Porter

J. P.

(1987). How unaffiliated/nonscientist members of institutional review boards see their roles. IRB: Ethics & Human Research, 9, 1–6. Retrieved from http://doi.org/10.2307/3563622

36.

Rickford

J. R.

Rickford

R. J.

(2000). Spoken soul: The story of Black English. Hoboken, NJ: Wiley.

37.

Rothman

D. J.

(1991). Strangers at the bedside: A history of how law and bioethics transformed medical decision making. New York, NY: Basic Books.

38.

Schneider

(2015). The censor’s hand: The misregulation of human-subject research. Cambridge, MA: MIT Press.

39.

Schrag

Z. M.

(2010). Ethical imperialism: Institutional review boards and the social sciences, 1965-2009. Baltimore, MD: Johns Hopkins University Press.

40.

Seligman

M. E. P.

Csikszentmihalyi

(2000). Positive psychology: An introduction. American Psychologist, 55, 5–14. Retrieved from http://doi.org/10.1037/0003-066X.55.1.5

41.

Sengupta

(2003). The roles and experiences of nonaffiliated and non-scientist members of institutional review boards. Academic Medicine: Journal of the Association of American Medical Colleges, 78, 212–218.

42.

Smith

D. E.

(1974). The social construction of documentary reality. SOIN Sociological Inquiry, 44, 257–268.

43.

Stark

L. J. M.

(2012). Behind closed doors: IRBs and the making of ethical research. Chicago, IL: University of Chicago Press.

44.

U.S. Department of Health, Education, and Welfare. (1974, 5 30). Title 45—Public welfare part 46—Protection of human subjects. Retrieved from http://wayback.archive-it.org/4657/20150930181944/http://archive.hhs.gov/ohrp/documents/19740530.pdf

45.

U.S. Department of Health and Human Services. (1993). OHRP institutional review board guidebook. Retrieved from http://wayback.archive-it.org/org-745/20150930181805/http://www.hhs.gov/ohrp/archive/irb/irb_guidebook.htm

46.

U.S. Department of Health and Human Services. (1998a, 6). Institutional review boards: Promising approaches. Retrieved from http://www.oig.hhs.gov/oei/reports/oei-01-97-00191.pdf

47.

U.S. Department of Health and Human Services. (1998b, 6). Institutional review boards: A time for reform. Retrieved from http://www.oig.hhs.gov/oei/reports/oei-01-97-00193.pdf

48.

U.S. Department of Health and Human Services. (2009). Title 45—Public Welfare CFR 46 Protection of Human Subjects. Retrieved from https://www.hhs.gov/ohrp/regulations-and-policy/regulations/45-cfr-46/index.html

49.

U.S. Department of Health and Human Services. (2010). OHRP educational videos: IRB membership. Retrieved from https://www.youtube.com/watch?v=GHtIbdLkSwU&list=PL5965CB14C2506914&index=7

50.

Veatch

R. M.

(1975). Human experimentation committees: Professional or representative? Hastings Center Report, 5, 31–40. Retrieved from http://doi.org/10.2307/3561227

51.

Washington

H. A.

(2006). Medical apartheid: The dark history of medical experimentation on Black Americans from colonial times to the present. New York, NY: Doubleday.