4

Opportunities to Build Trust

It is not uncommon for guidance related to controversial issues to be questioned, especially when many stakeholders are involved. In such cases, the steps for how the guidance was developed are just as important to support the integrity of the process as the guidance itself. On the whole, this National Academies of Sciences, Engineering, and Medicine (the National Academies) committee believes the current selection process—identifying individuals and creating a diverse panel—can reasonably lead to the appointment of a fair and balanced Dietary Guidelines Advisory Committee (DGAC). However, this National Academies committee believes the selection process for DGAC members needs to be enhanced to optimize its integrity.

ASSESSMENT OF THE DGAC SELECTION PROCESS

The evidentiary base is limited for identifying characteristics of selection processes that most effectively address key concerns, such as diversity and balance. As such, to conduct an independent review of the DGAC selection process, this National Academies committee relied on assessments of good practices from other organizations (see Chapter 3) and its own collective expertise and experience. From an assessment of these sources emerged a set of values the committee believes could enhance the integrity of a selection process:

• Enhance transparency. The foremost important characteristic of an effective process for the selection of advisory committee members is transparency. To the extent practicable, each step ought to be described in as much detail as possible and be made available to the public for its understanding. This transparency can help reassure the public that no undue influences or untoward actions are being taken.
• Promote diversity of expertise and experience. A broad range of expertise and experience must be considered to create a balanced committee. Expertise has to align with the topic areas to be reviewed. Diversity with respect to nontechnical skills (e.g., ability to form consensus or develop compromise) also needs to be considered. Building on the first characteristic of transparency, involvement from a broad range of perspectives, including public involvement, is also critical to fostering diversity.
• Support a deliberative process. A deliberative process should be used that considers information from a wide variety of sources. Decision makers ought to freely exchange information with one another toward the goal of coming to agreement or consensus. To the extent possible, the public should be engaged as well.
• Manage biases and conflicts of interest. The biases of individual members should be balanced among a broad representation of perspectives. Actual and or perceived conflicts of interest—both financial and nonfinancial—should be eliminated to the extent possible or their effects be minimized.
• Adopt state-of-the-art processes and methods. As practicable, selection processes and actions ought to be based on the best available evidence for the broader purpose of managing bias and conflict of interest. They should be revised and improved on as new evidence arises.

This National Academies committee used these five values of an “ideal” selection process to assess the current DGAC selection process. The discussions in the following sections describe that assessment, recognizing that each characteristic does not necessarily apply to every step of the process.

Overall, this National Academies committee found that the DGAC selection process is a thoughtful process that works within the bounds of the relevant laws to serve the U.S. Department of Agriculture (USDA) and the U.S. Department of Health and Human Services (HHS), as well as the American public. However, as noted in Chapter 1, the lack of transparency in the current process could lead to the perception that the membership of the DGAC is inequitable, which affects its integrity and trustworthiness.

Perception can be just as important as the truth when earning the public’s trust. How the process is perceived cannot be controlled, but the process can be engineered to show that it is thoughtful, fair, and supports the task. The selection process has been focused on appointing a group of experts able to assess the evidence related to nutrition and health. Based on critiques of previous DGACs and the significant nature of their work, this National Academies committee believes a shift may be warranted such that the selection process also focuses on engendering trust from the public. These proposed additions to the existing process would address the perception that the selection process is currently “obscure and not transparent” (Willett, 2016).

The current DGAC selection process consists of nine steps beginning with the establishment of a charter and culminating in the swearing in of the members (see Figure 2-1). This National Academies committee’s assessment of the process focused on step 3: Conduct a review of nominations and propose a slate of candidates. The other steps reasonably meet the goals delineated by the committee.

The processes used to establish the charter follow the guidelines laid out to support implementation of the Federal Advisory Committee Act (FACA), such as publishing a notice in the Federal Register. The nominations process also involves receipt of input from a broad range of stakeholders, including the public, allowing candidates to be identified with various expertise and experience. The call for nominations is sent to various online mailing lists, but ought to be shared even more broadly throughout the nutrition and health policy communities, for example through publication in relevant journals and social media, to broaden the pool of candidates. There are relatively few drawbacks to expanding the call for nominations, especially when using fast and inexpensive technologies, but would have the probable advantages of improving diversity of expertise and experience.

The third step—conduct a review of nominations and propose a slate of candidates—is not as transparent as it could be, and not much information is publicly available about the basis for the selections that are made. Many of the processes used to complete these steps are currently described publicly at a high level. For example, the membership balance plan for the 2015 DGAC states that its members would be balanced across many dimensions, including diversity across geographic areas, academic institutions, gender, race, ethnicity, and disability. However, this National Academies committee’s review of previous DGACs found a preponderance of DGAC members from the northeast, while noting inclusion of several members from the midwest and south. However, geographic diversity will likely be an ongoing issue.

As this third step is perhaps the most subjective step in the selection

process, and is arguably the most critical, it ought to be made even clearer to the public. Discussions about how individual nominees were considered with respect to their qualifications against selection criteria could be disclosed to the public. How perspectives on the advisory committee are balanced could be shared publicly. The membership balance plan outlines the process and criteria against which candidates are screened for selection, but the high-level descriptions in the plan could provide more detail to foster trust and integrity in the process. Also unclear is how closely the membership balance plan is followed in the development of the DGAC and how well the plan is implemented.

Specific to consideration of biases and conflicts of interest, how USDA and HHS approach these concepts is not clear. A review of conflicts of interest is mentioned as part of the process for special government employees, but more specific information could describe how conflicts of interest are identified in selection of the DGAC and, if needed, how they will be managed. Understanding how biases and conflicts of interest are identified and managed through a deliberative process would provide more transparency to the public on vetting and selecting advisory committee members. This National Academies committee identified this third step as an opportunity to enhance transparency and inclusiveness, as well as minimize undue influences (see Figure 4-1). The remainder of this chapter lays out this committee’s suggestions and specific recommendations for modifying the DGAC selection process.

ENHANCE TRANSPARENCY DURING CANDIDATE REVIEW

“Conduct a review of nominations and propose a slate of candidates” (see step 3 in Figure 2-1) is inherently the most subjective step in the current DGAC selection process. The perception by some is that the entire selection process is influenced by political interests since the secretaries from both USDA and HHS are political appointments, and the fact that USDA also supports U.S. agriculture (Mozaffarian, 2016). However, this National Academies committee was unable to identify any hard evidence that the Dietary Guidelines for Americans (DGA) have been unduly influenced because of USDA’s potential conflict. Unfortunately, there are limited objective measures to judge a nominee’s qualifications and the overall balance of a committee. In such subjective matters, the public must trust in the veracity of the process to achieve dependable results. Full transparency in the form of publicly accessible deliberations about every nominee is not practical given the sensitivity around why someone is or is not considered a strong candidate for DGAC membership. For example, considerations of personal finances, past history and relationships, and personal beliefs are all included in these deliberations. In part

because of these constraints, there may be the perception that the departments are not fairly considering all qualified candidates and not including representation of a broad spectrum of perspectives (Gummalla, 2016; Maitin-Shepard, 2016). This National Academies committee deliberated three options for improving the openness of this step.

Candidate Review Option 1

The first option this National Academies committee considered was to continue the status quo, where USDA and HHS vet nominees and appoint members. However, options for how to improve the current selection process, while keeping the full process under USDA and HHS, would likely result in marginal improvements. A different tactic would be needed to resolve the questions raised about the equity and integrity of the selection and appointment of members in a meaningful way. This option would be budget neutral and would likely continue to result in a 6- to 7-month process.

Candidate Review Option 2

One way to reduce the bias and perceived influence of political interests in the DGAC is to remove the selection process entirely from USDA and HHS. This would involve finding a third party that could identify and review all candidates and develop and appoint a balanced panel of experts. This approach is taken by HHS for its Physician-Focused Payment Model Technical Advisory Committee, where, by statute, the Comptroller General of the United States appoints members (HHS, 2016).¹ Of note, FACA requires that unless called for by statute, presidential directive, or other establishment authority, the final authority to select members rests with the agency heads, which in this case is the secretaries.² Changing the appointment authority for the DGAC would either require legislative action or a presidential directive.

The third party would need to be an objective arbiter with a strong record of having both the theoretical knowledge and practical expertise of having assembled impartial, neutral committees. It could be a private, nonprofit, or government organization, but it should not be part of either of the sponsoring agencies. In this option, a third party would also need to have expertise in the areas affected by the DGA in order to properly balance perspectives.

___________________

¹ The U.S. Government Accountability Office serves in this role for a number of committees, but these are the only federal advisory committees to which it makes appointments (personal communication, M. Giffin, U.S. Government Accountability Office, September 21, 2016). The Medicaid and Children’s Health Insurance Program Payment and Access Commission, Medicare Payment Advisory Committee, Patient-Centered Outcomes Research Institute Governing Board, Patient-Centered Outcomes Research Institute Methodology Committee, and Physician-Focused Payment Model Technical Advisory Committee memberships are all appointed by the Comptroller General of the United States (GAO, 2016). Appointments to the Health Information Technology Policy Committee are also made by congressional leaders.

² Federal Advisory Committee Act of 1972, 41 C.F.R. § 102-3.130(a).

There is no assurance that another body would not be subject to outside influences, while also being close enough to the subject matter to appropriately balance perspectives and expertise. This option would potentially require additional resources and would likely add months to the already tight time frame for selecting members.

Candidate Review Option 3

Another option would be to have the review and final selection be separated into two steps, conducted by two different bodies. A third party could be tasked with ensuring that a wide range of viewpoints and expertise was considered for the panel and objectively narrow down the list of nominees. Specifically, it could consider candidates’ nomination packages and qualifications, identify other candidates as necessary, interview promising candidates, provide an initial cursory review of biases and conflicts of interest, and submit a slate of primary and alternate nominees for the secretaries’ consideration. With its experience in the field, USDA and HHS could still be responsible for the role of balancing expertise and perspectives for final selection. The departments could provide the third party with selection criteria for nominees to maintain continuity in the process. These criteria ought to be shared publicly as part of the description of balance to be filed in accordance with FACA.

Although it is uncommon for other organizations to conduct all the steps and not be the final appointing authority for a federal advisory committee, the National Vaccine Advisory Committee may provide a model for a more limited third party role in committee selection. In this case, the legislation states that the National Academy of Sciences (NAS) will review the small number of nominees received for each vacancy (HHS, 2015). Candidates’ subject-matter expertise are considered during this assessment, but a review of potential conflicts of interest resides with the federal government. It is anticipated that once a third party is identified and employed, the review process would take up to 4 months to complete. Additional time would be required for the departmental selections and approval processes.

A third party, as discussed in option 2, would need to be an organization without a political, economic, or ideological identity. Preferably, the organization would have a history of putting together well-respected, impartial panels of experts. Unlike option 2, since the third party would not be making final selections, it would not necessarily need to be expert in nutrition or dietary guidance, just expert in evaluating individuals’ expertise and experience. This would allow for a wider range of possible third-party organizations to be considered.

Many types of organizations could possibly serve as a third party

for narrowing down the candidate pool, including other government agencies, consulting companies, nonpartisan research and policy groups, or professional organizations. Options within the government ought to include agencies outside of the sponsoring departments, such as agencies of Congress. Private-sector options include nonprofit, not-for-profit, and for-profit organizations. Regardless of what type of organization is chosen, an examination of a private-sector organization’s funding sources and customer bases would help ensure impartiality. A third party would need to consider a wide variety of members, not just members of potential host organizations, a common practice identified for professional organizations, as discussed in Chapter 3. A thorough search for an appropriate external organization would be required.

There are advantages of using an unbiased third party with the characteristics delineated above. This National Academies committee’s opinion is that political bias—both the perception and reality—would be reduced by a third party since USDA and HHS would not be involved in narrowing the field of candidates. If the secretaries of USDA and HHS are selecting final nominees from a short list of equally well-qualified, nonconflicted candidates, there is a greater potential that the final DGAC will be neutral. This in fact has the potential to reduce bias but even stronger potential to improve perception. Having the secretaries continue to select the final membership would also remove the need for the external organization to have expertise related to the DGA, likely resulting in a broader pool of candidates.

This alternative also has drawbacks like those delineated in option 2. There is no absolute guarantee that a third party will reduce bias; there is no evidence to say that a third party would not come up with the same exact committee of experts as assembled by the current process. Additionally, the secretaries of USDA and HHS remain the appointing authorities under this option. However, to the many critics of the process, a third party would ensure that USDA and HHS were more at an arm’s length from the selection of DGAC members. This committee believes that at the very least, this would improve public perception of a more objective process, as the mandate from Congress indicates that some subsets of the public do not trust the DGA. Another drawback is that this option would also likely have budgetary implications despite some savings in staff time, as well as lengthen the selection process. However, selection of a third party could begin before the charter is filed so as to leave the DGAC with as much time as possible to conduct its work.

Conclusion

This National Academies committee recognizes few objective measures exist to assess the effectiveness of a selection process. In considering the options, the committee could not find explicit evidence to suggest that the current process is biased, but also accepts that there is no explicit evidence to prove that it is not. Continuing the status quo would result in an unsatisfactory response to the fundamental issue of how the full list of nominees is narrowed down to the final committee and does not suggest option 1 be selected. The committee concludes that the current process needs to be more transparent than it currently is to assure the public that a wide set of viewpoints and expertise is being considered. However, the committee does not believe that the appointment authority needs to be removed from the secretaries of USDA and HHS, and therefore does not believe option 2 is the best approach.

This National Academies committee envisions that a hybrid approach be adopted as outlined in option 3 where a neutral, unbiased arbiter would be used as a first screen, still leaving responsibility for balancing the final advisory committee and appointing members with the secretaries. At a minimum, this National Academies committee believes divorcing the screening process from the appointment authority would improve perception that the DGAC process is more neutral than it currently is. The committee believes the proposed separation of the processes also has the potential to actually reduce bias. This separation would be feasible immediately, as only a change to the appointing authority would require action from Congress or the president. Given the lack of available evidence, it is this National Academies committee’s best judgment and opinion that the potential advantage of reducing bias and improving public perception would yield substantial benefits and believes the additional costs and time are important investments to make. No matter what option is chosen, the charter should be filed at the latest possible point to allow the DGAC as much time as possible to develop its report.

CRITERIA FOR SELECTING CANDIDATES

The criteria against which candidates are reviewed are discussed in this section. How composition is balanced and how expertise is considered are addressed in the next section. In the current process, candi-

dates are assessed “based upon their qualifications, level of expertise and knowledge, and ability to contribute to the work to be performed,” as well as diversity of “geographic areas, academic institutions, gender, race, ethnicity, and disability” (USDA/HHS, 2016a). Other factors are likely considered during the balancing process, but are not explicitly stated. Conflicts of interest and background checks are considered prior to appointment to the advisory committee.

Other factors are likely considered during the balancing process, but they are not explicitly stated. Additional criteria for considering candidates ought to include willingness to serve; minimal financial and nonfinancial conflicts of interest (see Boxes 3-2 and 3-3); biases that can be balanced with those of other members; and prior experience working on advisory committees or panels. Conflicts of interest and biases will be discussed in greater detail later in the chapter. Prior experience on panels is one important factor to evaluate the candidate’s interpersonal skills. A binary yes or no account of past service on a committee is often the only measure of past committee experience, but the quality of that participation is more difficult to assess and measure. The committee intends for this criterion to include an assessment of interpersonal skills such as ability to compromise and collaborate, and ability to express and reconcile divergent opinions. Such skills can best be gauged by seeking feedback regarding the candidate’s performance on prior consensus-developing activities. USDA and HHS could consider making all criteria available for public comment before adoption, if time allows. Receiving feedback on selection criteria could improve transparency, but would add more time to an already lengthy process to set up the advisory committee, and therefore would leave the DGAC with less time to complete its work.

Currently, information regarding candidates’ technical expertise is collected during the initial solicitation and consists of a statement of qualifications and a curriculum vitae or resume limited to 10 pages. The committee believes more directed information ought to be sought during the call for nominations for consideration during selection. This could include experience on relevant boards or panels, activities related to nutrition policy, consulting, appointment as chair or cochair of committees and study groups, elected positions, and relevant publications or presentations. One example for the types of information to include is the biosketch requested by the National Institutes of Health, which requests submission of a brief personal statement describing the candidate’s qualifications for the particular project; list of positions and honors; statement of the candidate’s most significant contributions to science; and additional information, such as research support and/or scholastic performance. The total length is limited to five pages (NIH, 2016).

The proposed chair, co-chairs, or vice chair would ideally be held to

a higher standard as the leadership of the advisory committee and would not necessarily need to be a subject-matter expert (IOM, 2011). Although difficult to achieve, to the extent possible, it would be preferable if these individuals were not biased in the specific areas being addressed. Most importantly, these members need to be free of financial conflicts of interest, and nonfinancial conflicts of interest if possible. The proposed chair would also preferably have served in a leadership role for other panels or committees.

Currently, as the advisory committee’s leader, the chair presides over meetings and serves as the liaison to the departments to help establish priorities. The chair also has a number of administrative duties, such as certifying accuracy of meeting minutes and requesting motions to vote. One important part of the role is to facilitate discussions. However, this can be difficult to do while also keeping focused on the substance of the deliberations. A facilitator could be used to guide the flow of discussions to ensure all members are heard. An effective facilitator would remain neutral, understand group dynamics, be flexible yet firm, know when to let the advisory committee work through an issue, and be sensitive to the politics at hand. Adding a facilitator would not be budget neutral, but could help the advisory committee complete its tasks in a more efficient manner. If a facilitator is used, facilitation skills would not need to be part of the criteria for selecting a chair. Otherwise, ability to facilitate a group ought to be a key criterion for choosing the leadership of the DGAC.

Skills need to be reviewed for the group as a whole, as well as individuals. The organization selecting candidates will need to review the collective expertise, experience, and perspectives before making final appointments. Through this deliberative process, the public can be assured that the advisory committee is objective and has the requisite expertise to complete its task.

DGAC COMPOSITION

As described in Chapter 2, in accordance with FACA, agencies are required to consider candidates from all backgrounds who are “directly affected, interested, and qualified, as appropriate to the nature and functions of the advisory committee.” Advisory committees providing technical guidance should also include people with “demonstrated professional or personal qualifications and experience relevant to the functions and tasks to be performed.”³ The 2015 DGAC membership balance plan identifies a set of 17 specialty areas to be represented, ranging from osteoporosis to nutrition-related systematic review methodology. However,

___________________

³ Federal Advisory Committee Act of 1972, 41 C.F.R. § 102-3.60(b)(3).

the charter states that the final advisory committee is not limited to the categories listed in the balance plan. The membership balance plan does not anticipate how that distribution will be achieved across or within specialty categories. It calls for experts to have experience in one or more of the areas, so a one-to-one match is not needed between category of expertise and DGAC members. The plan also states that individuals for the 2015 DGAC are selected to represent viewpoints of the scientific evidence, not of any specific stakeholder groups.

After discussing various approaches, this National Academies committee concludes that the DGAC should represent a wide variety of perspectives so that any group of experts with a similar composition could be appointed and derive the same findings. During this deliberation, the complicated question arose of which should be developed first: the specific questions to be answered by the DGAC or the areas of expertise needed to address the charge? These are interdependent steps, but one must come before the other. The current process relies on the DGAC to develop priority topics for review rather than for an a priori process to identify which updates and reviews are most critically needed, thus influencing the expertise needed on the DGAC. This National Academies committee discussed the potential value of focusing on specific topic areas that need revision or updating in the DGA. This situation would allow for concentration of expertise in key priority topics rather than all the expertise needed for review of the complete DGA. This issue was discussed and debated because of the need to focus on recommendations for pregnant women and children from birth to 24 months in the 2020–2025 edition of the DGA. However, to meet the short timeline for this first report, this National Academies committee was not able to formulate a specific recommendation on this approach or more broadly about the advisory committee’s overall composition. The committee will address these issues in its second report when it can fully examine the charge to the DGAC and the overall DGA process.

The options below were discussed as potential strategies to address the 2020 DGAC composition, recognizing the limitations on the size of the advisory committee. Customarily, 13 to 17 members are appointed to the DGAC. Adding additional members may be able to help expand the number of viewpoints represented, but it might make it difficult to derive consensus and might potentially delay the delivery of the DGAC’s final report. Thus, it appears there are more areas of expertise that could help inform the DGAC than can reasonably be managed on an advisory committee. It may be worth taking a different approach to identifying members that allows for consideration of a greater mix of individuals with both particular expertise in focused areas of nutrition, as well as including broad, more general thinkers. Inclusion of other

types of federal advisory committee members (e.g., representative members, nonvoting members; see Box 2-2) could also be considered as a way to improve buy-in and expand committee membership as needed. If these types of members are included in future DGACs, it would be critical to balance their areas of expertise and have a wide range of viewpoints included. USDA and HHS could consider improving transparency by publishing the categories of expertise each member represents.

Other perspectives could be included through use of non-DGAC member subcommittees, consultants, and invited speakers. In each of these situations, the invited experts would not have a voting role on the DGAC. Instead, their roles would be limited to helping inform the DGAC’s deliberations and conclusions. Subcommittees could comprise outside experts and one or two DGAC members to make recommendations to the DGAC—not USDA or HHS—on broad questions requiring a number of areas of expertise. These subcommittees could be chaired by the one or two members of the DGAC. For more narrow topics not requiring a full subcommittee, or where the DGAC would like to supplement its expertise, consultants could be used. Because these individuals can influence decision making through the information provided to the DGAC for its deliberations, it is important that these participants provide balanced perspectives and that their conflicts of interest and biases are appropriately balanced and managed. The values identified by this National Academies committee to optimize the integrity of the process are important to consider for input from these non-DGAC members as well.

Multiple models for developing consensus were also discussed. For example, panels could consist of individuals knowledgeable in the topic area generally but not vested in the topic (e.g., through research or publicly stated opinions). Alternatively, it could comprise individuals with specialized expertise who are vested in and have published on the topic. New perspectives also could be considered, such as information management and systems science. These other disciplines can help bring innovations to the selection process. For example, an expert in new methods to manage information could bring about changes to the way candidates are reviewed, potentially minimizing the likelihood of bias and shortening the process for review. Another example includes modeling and simulation that could also be used to understand the potential long-term impacts of interventions that would not be possible to measure in the short-term.

ADDITIONAL PUBLIC COMMENT PERIODS

As discussed in Chapter 2, the current selection process seeks public comment through a call for nominations in the Federal Register and relevant online mailing lists (USDA/HHS, 2016a). Nominations are received

from stakeholder groups from varying viewpoints. This National Academies committee concludes that this portion of the nomination process is strong and encourages transparency. However, the only formal opportunity for public input during the selection process is in response to the call for nominations. Another public comment period, targeted at nominees being considered for appointment, would improve the transparency of the DGAC selection process. Two models were considered for receiving additional public comments based on a review of other processes.

Public Comment Period Option 1

In addition to a public call for nomination, some agencies request public comments on qualified candidates. In this model, individuals identified during the nomination period would be posted on a website and/or the Federal Register. The public would be invited to comment on the candidates in terms of the overall balance of the advisory committee with respect to expertise and management of bias. All feedback would be taken into consideration during final selections, in addition to other relevant information either submitted during the nomination process (e.g., statements of qualifications, curriculum vitaes) or collected by staff. Based on the totality of information available, the appointing authority would select a panel of members.

This model is used to assemble some federal advisory committees, among others. For example, the U.S. Environmental Protection Agency (EPA) and the U.S. Department of Labor (DOL) have both invited the public to submit comments on lists of candidates.

EPA posted candidates’ biographies for a 23-day comment period on the 72 nominees received for the Chemical Assessment Advisory Committee Augmented for Benzo[a]pyrene Review. Input received during that time was taken into consideration by the science advisory board’s staff office director (EPA, 2014). The DOL Advisory Board on Toxic Substances and Worker Health holds a 14-day period when the public can submit comments to the secretary of labor about the qualifications of 66 nominees (DOL, 2015).

One benefit of this approach is that the full list of candidates would be disclosed to the public, allowing input to be received on any nominee. Decisions made by the appointing authority would be informed by evidence from all sources, whether brought forth by the individuals themselves, staff, or the public. While this enhances transparency in one regard, it still leaves in question considerations about how balance is achieved on the panel. Insights into how balance was achieved would only be made available to the public upon announcement of the full committee.

Public Comment Period Option 2

Another model this National Academies committee discussed was requesting input on a slate of provisional members. All necessary steps to finalize panel membership would be completed, but the appointment would be contingent upon consideration of public comments. Feedback would be invited for a reasonable period of time on the advisory committee’s proposed composition. Appointments would not be finalized until after public comments were received and considered by the appointing authority. If comments indicated a provisional member had an intractable conflict or revealed previously unknown biases that could not be balanced, the department could confer with the provisional member, and if needed, reconsider the appointment. An approach like this is mandated by Section 15 of FACA for committees appointed by the NAS and the National Academy of Public Administration as described in Chapter 3. If identification of another member was warranted, an alternate member identified in prior steps could be selected and subjected to the same process as other provisional members. A separate public comment period could be held to provide opportunity to review the entire revised committee’s composition. Use of multiple public comment periods would help minimize scrutiny as to why one person was originally listed as an alternate if the alternates are never turned to. It also reduces the burden on DGAC alternates by eliminating the need for these individuals to provide their entire history of potential conflicts of interest and reduces the possibility of unwarranted attacks.

This option promotes more complete public input than the current selection process by inviting feedback from all stakeholders on how the slate is balanced, as well as on potential individual members of the slate. Comments are encouraged to focus on both provisional members and the proposed committee as a whole. However, unlike the model discussed in option 1, the full list of candidates is not publicized; the public has to have confidence that provisional members were selected fairly.

Conclusion

Providing a reasonable amount of time for feedback is critical to a transparent process and the public should have an additional opportunity to comment after the initial solicitation of nominations (see step 7 of Figure 4-1). However, because allotting time for a public comment period does delay final appointments, the committee did not consider adding more than one opportunity for input.

This National Academies committee discussed the efficacy of the two options in the context of the DGAC. Over each of the past three cycles, the

departments considered between 150 and 200 candidates (USDA/HHS, 2016b). In the past, the DGAC’s lack of balance was criticized, and not necessarily for the qualifications of specific individuals (Maitin-Shepard, 2016). Considering the size of the candidate pool and the need to focus on the overall composition, the committee concluded that option 2 would more effectively address concerns raised by the public and encourage transparency.

All public comments about specific provisional members ought to be considered carefully. It is very likely that proposed membership would change as a result of the public comments and that a provisional member could be removed from or added to the slate. For the purpose of improving transparency, this National Academies committee did discuss posting public comments about each candidate, as well as explanations by USDA and HHS regarding why each primary and/or alternate candidate was or was not selected to serve on the DGAC. Candid information from the public about proposed members is critical for a deliberative process. However, posting such information would likely deter volunteers from wanting to serve on the DGAC. Although it is an honor to serve on such a prominent group, these individuals ultimately volunteer their time and expertise at their own will. Ad hominem attacks presented in public comments or in explanations by USDA or HHS could result in candidates being maligned in the public press and their reputations damaged, reducing the pool of qualified candidates willing to take part in such an activity. If USDA and HHS decide to share public comments made about each candidate, guidelines would need to be developed to protect candidates so that comments that are slanderous in nature would not be accepted for fear that they may be released, for example through a Freedom of Information Act request. In its review of other committee selection processes, this National Academies committee could not identify examples that make such information about candidates publicly available, and thus determined that the potential benefits were not substantial enough to warrant publication of all comments made about proposed members.

ADDRESSING BIASES AND CONFLICTS OF INTEREST

As discussed in Chapter 3, balancing biases and limiting conflicts of interest are both critical to establishing trust. This is particularly true in

nutrition research, where industry, political influences, and professional organizations are all seen to be major sources of perceived conflict. Due to the lack of high-quality prospective cohort studies that directly impact the DGA, it would be difficult to eliminate these influences.

A policy explicitly stating how an organization will identify and manage conflicts is a standard tool used to combat perceptions of undue influence. In medicine, these policies aim “to protect the integrity of professional judgment and to preserve public trust rather than to try to remediate problems with bias or mistrust after they occur” (IOM, 2009, p. 28). Organizations around the world—including the federal government, medical journals, and universities—have adopted policies concerning conflicts of interest (Boyd et al., 2004; Graham et al., 2015).⁴

The literature review described in Chapter 1 on conflict of interest examined these specifically in relation to guideline development and advisory committees. Significant variation was observed across advisory committees and guideline development groups in how information on conflicts of interest was defined, collected, and managed. The literature broadly supports the notion that disclosure and management of conflicts of interest is a necessary part of establishing public trust and increasing transparency in guideline development (Barrow and Conrad, 2006; Gessner et al., 2010; Rowe et al., 2009, 2013; Schünemann et al., 2006, 2015). However, policies for disclosure and management vary across groups.

Financial disclosures are not widely published in guidelines (Bindslev et al., 2013; Khalil et al., 2012; Tibau et al., 2015) despite broad agreement that disclosure is beneficial in reducing the risk of and appearance of bias resulting from conflicts of interest (Rowe et al., 2009). Across organizations that have already implemented disclosure practices, examples of disclosure methods identified include verbal, written on a detailed form, and written in response to broad questions. Examples of groups to whom disclosure was reported include a third-party reviewer or oversight committee, others in the guideline development group, and/or public disclosure.

The evidence base directly linking disclosed conflicts of interest to associated outcomes is limited, in part, because of a lack of standardization on outcomes to assess. Potential outcomes can be framed as both positive and negative. Positive outcomes could be some measure of trustworthiness, reduced bias, or actual health outcomes. Negative outcomes could be lower-quality recommendations as a result of not including experts or missed opportunities to improve health outcomes. Lack of public trust in the recommendations or perception of undue influence as

___________________

⁴ Standards of Ethical Conduct for Employees of the Executive Branch, 5 C.F.R. § 2635 (August 7, 1992).

a result of conflicts of interest is another negative outcome. Despite a lack of a formal body of evidence, Box 4-1 highlights some observed associations and the importance of disclosing and managing conflicts of interest.

Although no systematic reviews were identified regarding management strategies on conflict of interest and their associated outcomes, several descriptive studies outlining existing conflict-of-interest policies and proposing best practices were identified (Boyd and Bero, 2006; Boyd et al., 2004, 2012; Mendelson et al., 2011). A thorough examination of conflict-

of-interest policies from six organizations,⁵ as well as relevant empirical data, found that significant variation existed on how a conflict of interest was defined, how and when it was disclosed, the minimum funding amount requiring report, the time period over which to report, how the disclosure was reviewed, and finally, how the conflict was managed. Still, several improvements for disclosure and management may be possible, as identified in the literature. The review found that all relevant individuals should disclose any potential conflicts of interest. If no conflicts exist, individuals should serve on the committee as planned, and should disclose anything that arises during their course of service that could be a potential conflict (Graham et al., 2015). If a conflict exists, depending on the type (financial or nonfinancial) and severity, these three management strategies may be employed:

1. The individual should not serve on the committee (Rowe et al., 2013).
2. The individual should serve on the committee in a limited capacity, but not participate in decision making or voting regarding the recommendation for which they have a conflict (Guyatt et al., 2010; Neumann et al., 2013).
3. The individual should serve on the committee as long as a counterviewpoint is represented for balance (Viswanathan et al., 2014).

This National Academies committee concludes conflicts of interest—whether actual or perceived cannot always be able to be eliminated entirely on a balanced panel of experts, as conflicts of interest are interpreted to include industrial, professional, political, and cultural influences. Although significant conflicts need to be avoided, some situations may exist where the requisite expertise cannot be found in individuals without conflicts of interest. In these instances, it is necessary to identify, disclose, and manage the influences in question. This National Academies committee concludes USDA and HHS need to develop an explicit policy to address biases and conflicts of interest to be shared with the public. This policy would apply to provisional members being considered for appointment. As discussed previously, individuals who provide input to the advisory committee as consultants or members of a subcommittee would also be assessed for their biases and conflicts of interest.

___________________

⁵ American College of Chest Physicians, American College of Physicians, American Medical Association, International Committee of Biomedical Journal Editors, Society for Critical Care Medicine, World Health Organization.

Identification

As discussed in Chapter 2, as a federal advisory committee, the current DGAC selection process uses the U.S. Office of Government Ethics’ (OGE’s) confidential financial disclosure process, including the OGE Form 450 described in Chapter 3. Conflicts of interest for the candidate are taken into consideration, as well as those of the candidate’s spouse, minor child, general partner, outside employer, and persons or organizations with whom the candidate is negotiating or has an arrangement for employment. This National Academies committee believes the OGE Form 450 adequately covers financial conflicts of interest, but it did not find any explicit, formal steps for DGAC candidates to disclose nonfinancial conflicts or biases. To enhance transparency, a form should be developed and used for the disclosure of nonfinancial conflicts of interest and biases. Since the OGE only addresses financial conflicts of interest, a form would have to be developed and approved by that office, such as the EPA Form 3110-48 discussed in Chapter 3. A nonfinancial conflict of interest and bias declaration form ought to be designed to capture appropriate professional experiences and relationships relevant to the task for which the individual is being considered. Examples include publications, speeches, testimony, or advisory roles where the individual discussed topics related to the task (see Box 3-3). Additionally, such a form could help identify affiliations or relationships with organizations that could benefit nonfinancially from the outcomes of the task. Once a bias or conflict has been judged to be present, ethics officers ought to try to determine the potential effect of the conflict.

This National Academies committee believes potential biases and conflicts of interest ought to be disclosed to three audiences: the appropriate ethics officers, other members of the specific activity, and to the public. Full disclosure of potential biases and conflicts should be shared with ethics officers via the above-described forms. In addition, disclosure should be made revealing potential sources of conflicts to others serving on the same activity. In the case of the DGAC, provisional members should share their biases and potential sources of conflicts—both financial and nonfinancial—with each other, preferably in a nonpublic administrative meeting. This would allow members to better understand the basis for each other’s positions during the advisory committee’s work. Finally, this National Academies committee proposes an abstraction of financial conflicts of interest deemed by ethics officials to be significant be shared with the public, including sources of funding, consultancies, and other relationships as appropriate. While it is unclear whether an exemption to the Privacy Act could be made to allow USDA and HHS to require provisional members to share the sources of potential conflicts with the public, this National Academies committee believes that in the spirit of transpar-

ency, general information about individual financial conflicts ought to be shared with the public. Provisional DGAC members ought to be willing to share this information publicly at a meeting or through their own means.

This National Academies committee suggests those who have had relationships with industry or advocates in the past 3 years can participate fairly on a panel if the nature of the relationship is incidental to the work of the panel, as it would allow for people with a greater diversity of expertise to potentially serve on the DGAC. Examples of lookback periods ranged from 6 months to 5 years. However, the committee could not identify literature that conclusively supported a specific length of time to consider. Without conclusive evidence about the effect of the length of a lookback period, the committee considered the process to update the DGA. Three years equals the amount of time from release of the last edition of the DGA to the formation of the next advisory committee. This National Academies committee weighed the potential risks and benefits of various lookback periods and used its collective judgment to agree on 3 years as a reasonable period. Appointment of a diverse DGAC would most clearly be tested if the proposed changes regarding bias and conflict of interest policies are made, and the DGAC composition does not change accordingly.

A detailed description of how biases and conflicts of interest will be identified and judged should be made publicly available as part of a policy on biases and conflicts of interest and ought to be monitored independently by someone not involved in managing the development of the DGA. As discussed in Chapter 3, organizations use a variety of structures to assess biases and conflicts of interest. For example, in some instances a steering committee is employed to evaluate and judge the effects of potential conflicts. The executive branch of the federal government relies on its ethics officers and the OGE to make determinations about potential financial conflicts. However, assessment of biases and nonfinancial conflicts are more difficult to make.

Management

Many tools exist to manage conflicts of interest. This National Academies committee recommends that a policy be shared with the public describing (1) a general plan for identifying and resolving biases and conflicts on the whole panel, and (2) plans for managing individuals’ specific conflicts, as needed.

Specific to conflicts of interest, exemptions could be applied in instances where the potential conflict is deemed too remote or inconsequential to significantly influence an individual’s judgment. Another tool to manage conflicts of interest is the granting of waivers that would allow

for varying ranges of participation in the given task depending on the significance and nature of the conflict. Individuals could also choose to sell stock or otherwise divest property deemed a conflict before participation. They could also choose to resign from a potentially disqualifying activity. These and other approaches are often used to mitigate the effect of activities deemed to be actual or perceived conflicts. Changes to the advisory committee’s structure could also be adopted to minimize the effect of any undue influence during its work. For example, members with potential conflicts could be excluded from the discussions or recommendations from specific subcommittees or writing groups. Other methods, such as limiting the number of individuals with waivers on the overall advisory committee or on subcommittees, could also be employed.

During the course of an advisory committee’s service, other tools can be used to manage both biases and conflicts. A review highlighting any new biases or potential conflicts of interest could be discussed at the beginning of every meeting. When a potential bias is disclosed in relation to an item on the meeting agenda, individuals could be recused from discussions and voting that could be perceived as unduly influencing the work. Specific management plans could be tailored to an individual’s personal circumstances. These individualized management plans would explain what specifically was done to try to minimize the effect of any biases or conflicts. These plans ought to remain confidential, and would be reviewed annually by an ethics officer.

Consideration of the benefits and harms related to the management of conflicts of interest is important. Not all conflicts of interest unduly influence a person’s actions or decisions while serving on a committee. Although it is important to avoid the perception of conflicts of interest, there are also risks of being too stringent on conflict-of-interest disclosures, such as candidates not wanting to join as a result of an overly burdensome process. Recognizing these risks, how biases and conflicts of interest are managed is critical, and rigorous policies ought to be followed.

Transparent Process for Assessing Bias and Conflict of Interest

The current selection process for the DGAC requires members be screened for financial conflicts of interest and attend ethics training. However, the steps taken during this process are not necessarily clearly delineated or shared with the public. To enhance the process for assessing bias and conflicts of interest, this National Academies committee recommends that the secretaries of USDA and HHS certify that an independent review and assessment of biases and conflicts of interest was conducted. This certification would also include description of any management plans put in place prior to the commencement of the advisory committee’s work.

The committee suggests independent reviews be conducted by a federal ethics officer.

Certification provides a level of transparency to the public that biases and conflicts of interest were reviewed prior to the advisory committee’s work. A statement should also be made at the conclusion of the work to describe how biases and conflicts of interest were managed at a high level. Such a statement would give the public assurance that the biases and conflicts of interest were appropriately managed throughout the entire course of the advisory committee’s work, and that the advisory committee’s discussions were not unduly influenced.

CONCLUSION

The current selection process for the DGAC can be improved. To that end, this National Academies committee has provided recommendations and suggestions for the secretaries of USDA and HHS to consider to improve transparency and reduce bias. These are generally not radically new ideas, but they could all be implemented to help enhance the integrity of the current process by increasing transparency, identifying the potential for conflict of interest, and appropriately managing these conflicts to minimize bias. As part of an overall, comprehensive review of the process to update the DGA, additional findings and recommendations about the selection process may be made as part of this committee’s second report.

REFERENCES

Barrow, C. S., and J. W. Conrad. 2006. Assessing the reliability and credibility of industry science and scientists. Environmental Health Perspectives 114(2):153-155.

Bindslev, J. B., J. Schroll, P. C. Gotzsche, and A. Lundh. 2013. Underreporting of conflicts of interest in clinical practice guidelines: Cross sectional study. BMC Medical Ethics 14:19.

Boyd, E. A., and L. A. Bero. 2006. Improving the use of research evidence in guideline development: 4. Managing conflicts of interests. Health Research Policy and Systems 4.

Boyd, E. A., S. Lipton, and L. A. Bero. 2004. Implementation of financial disclosure policies to manage conflicts of interest. Health Affairs (Millwood) 23(2):206-214.

Boyd, E. A., E. A. Akl, M. Baumann, J. R. Curtis, M. J. Field, R. Jaeschke, M. Osborne, and H. J. Schunemann. 2012. Guideline funding and conflicts of interest: Article 4 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proceedings of the American Thoracic Society 9(5):234-242.

Campsall, P., K. Colizza, S. Straus, and H. T. Stelfox. 2016. Financial relationships between organizations that produce clinical practice Guidelines and the biomedical industry: A cross-sectional study. PLoS Medicine 13(5):e1002029.

Chartres, N., A. Fabbri, and L. A. Bero. 2016. Association of industry sponsorship with outcomes of nutrition studies: A systematic review and meta-analysis. JAMA Internal Medicine 176(12):1769-1777.

Choudhry, N. K., H. T. Stelfox, and A. S. Detsky. 2002. Relationships between authors of clinical practice guidelines and the pharmaceutical industry. Journal of the American Medical Association 287(5):612-617.

Cosgrove, L., H. J. Bursztajn, D. R. Erlich, E. E. Wheeler, and A. F. Shaughnessy. 2013. Conflicts of interest and the quality of recommendations in clinical guidelines. Journal of Evaluation in Clinical Practice 19(4):674-681.

DOL (U.S. Department of Labor). 2015. Advisory Board on Toxic Substances and Worker Health notice of comment period. Federal Register 80(199):62111-62113.

EPA (U.S. Environmental Protection Agency). 2014. Invitation for public comment on the list of candidates for the EPA science advisory board chemical assessment advisory committee augmented for benzo[a]pyrene review https://yosemite.epa.gov/sab/sabproduct.nsf/0/4DCFD0E5F45A8CAD85257B65005B17C8/$File/List%20of%20candidates%20with%20Biosketch%209-18-2014.pdf (accessed October 5, 2017).

GAO (U.S. Government Accountability Office). 2016. Health care advisory committees. http://www.gao.gov/about/hcac/index.html (accessed October 5, 2017).

George, J. N., S. K. Vesely, and S. H. Woolf. 2014. Conflicts of interest and clinical recommendations: Comparison of two concurrent clinical practice guidelines for primary immune thrombocytopenia developed by different methods. American Journal of Medical Quality 29(1):53-60.

Gessner, B. D., P. Duclos, D. DeRoeck, and E. A. S. Nelson. 2010. Informing decision makers: Experience and process of 15 national immunization technical advisory groups. Vaccine 28(Suppl 1):A1-A5.

Graham, T., P. Alderson, and T. Stokes. 2015. Managing conflicts of interest in the UK National Institute for Health and Care Excellence (NICE) clinical guidelines programme: Qualitative study. PLoS ONE 10(3):e0122313.

Gummalla, S. 2016 (unpublished). Comments presented at USDA Dietary Guidelines for Americans listening sessions: American Frozen Food Institute. Washington, DC, February 19, 2016.

Guyatt, G., E. A. Akl, J. Hirsh, C. Kearon, M. Crowther, D. Gutterman, S. Z. Lewis, I. Nathanson, R. Jaeschke, and H. Schünemann. 2010. The vexing problem of guidelines and conflict of interest: A potential solution. Annals of Internal Medicine 152(11):738-741.

HHS (U.S. Department of Health and Human Services). 2015. National Vaccine Advisory Committee charter. https://www.hhs.gov/nvpo/nvac/charter/index.html (accessed October 5, 2017).

HHS. 2016. Physician-Focused Payment Model Technical Advisory Committee charter. https://aspe.hhs.gov/charter-physician-focused-payment-model-technical-advisory-committee (accessed October 5, 2017).

IOM (Institute of Medicine). 2009. Conflict of interest in medical research, education, and practice. Washington, DC: The National Academies Press.

IOM. 2011. Clinical practice guidelines we can trust. Washington, DC: The National Academies Press.

Khalil, B., K. Aung, and I. A. Mansi. 2012. Reporting potential conflicts of interest among authors of professional medical societies’ guidelines. Southern Medical Journal 105(8):411-415.

Lesser, L. I., C. B. Ebbeling, M. Goozner, D. Wypij, and D. S. Ludwig. 2007. Relationship between funding source and conclusion among nutrition-related scientific articles. PLoS Medicine 4(1):e5.

Maitin-Shepard, M. 2016 (unpublished). Comments presented at USDA Dietary Guidelines for Americans listening sessions: American Cancer Society/American Cancer Society Cancer Action Network. Washington, DC, February 19, 2016.

Mendelson, T. B., M. Meltzer, E. G. Campbell, A. L. Caplan, and J. N. Kirkpatrick. 2011. Conflicts of interest in cardiovascular clinical practice guidelines. Archives of Internal Medicine 171(6):577-584.

Mozaffarian, D. 2016 (unpublished). Comments presented at USDA Dietary Guidelines for Americans listening sessions. Washington, DC, February 19, 2016.

Mühlhauser, I., and G. Meyer. 2013. Evidence base in guideline generation in diabetes. Diabetologia 56(6):1201-1209.

Neumann, I., R. Karl, A. Rajpal, E. A. Akl, and G. H. Guyatt. 2013. Experiences with a novel policy for managing conflicts of interest of guideline developers: A descriptive qualitative study. Chest 144(2):398-404.

NIH (National Institutes of Health). 2016. Biosketch format pages, instructions and samples. http://grants.nih.gov/grants/forms/biosketch-instructions-Forms-D.docx (accessed October 5, 2017).

Norris, S. L., H. K. Holmer, L. A. Ogden, and B. U. Burda. 2011. Conflict of interest in clinical practice guideline development: A systematic review. PLoS ONE 6(10):e25153.

Norris, S. L., H. K. Holmer, L. A. Ogden, B. U. Burda, and R. Fu. 2013. Conflicts of interest among authors of clinical practice guidelines for glycemic control in type 2 diabetes mellitus. PLoS ONE 8(10):e75284.

Pham-Kanter, G. 2014. Revisiting financial conflicts of interest in FDA advisory committees. The Milbank Quarterly 92(3):446-470.

Ridker, P., and J. Torres. 2006. Reported outcomes in major cardiovascular clinical trials funded by for-profit and not-for-profit organizations: 2000-2005. Journal of the American Medical Association 295(19):2270-2274.

Rowe, S., N. Alexander, F. M. Clydesdale, R. S. Applebaum, S. Atkinson, R. M. Black, J. T. Dwyer, E. Hentges, N. A. Higley, M. Lefevre, J. R. Lupton, S. A. Miller, D. L. Tancredi, C. M. Weaver, C. E. Woteki, and E. Wedral. 2009. Funding food science and nutrition research: Financial conflicts and scientific integrity. American Journal of Clinical Nutrition 89(5):1285-1291.

Rowe, S., N. Alexander, C. M. Weaver, J. T. Dwyer, C. Drew, R. S. Applebaum, S. Atkinson, F. M. Clydesdale, E. Hentges, N. A. Higley, and M. E. Westring. 2013. How experts are chosen to inform public policy: Can the process be improved? Health Policy 112(3):172-178.

Schünemann, H. J., A. Fretheim, and A. D. Oxman. 2006. Improving the use of research evidence in guideline development: 1. Guidelines for guidelines. Health Research Policy and Systems 4(13).

Schünemann, H. J., L. A. Al-Ansary, F. Forland, S. Kersten, J. Komulainen, I. B. Kopp, F. Macbeth, S. M. Phillips, C. Robbins, P. Van der Wees, and A. Qaseem. 2015. Guidelines international network: Principles for disclosure of interests and management of conflicts in Guidelines. Annals of Internal Medicine 163(7):548-553.

Tibau, A., P. L. Bedard, A. Srikanthan, J. L. Ethier, F. E. Vera-Badillo, A. J. Templeton, A. Ocana, B. Seruga, A. Barnadas, and E. Amir. 2015. Author financial conflicts of interest, industry funding, and clinical practice guidelines for anticancer drugs. Journal of Clinical Oncology 33(1):100-106.

USDA/HHS (U.S. Department of Agriculture/U.S. Department of Health and Human Services). 2016a (unpublished). Dietary Guidelines for Americans: Process brief, sections 1–3. Prepared for the Committee to Review the Process to Update the Dietary Guidelines for Americans.

USDA/HHS. 2016b (unpublished). HMD follow-up questions for USDA Response to Committee to Review the Process to Update the Dietary Guidelines for Americans.

Viswanathan, M., T. S. Carey, S. E. Belinson, E. Berliner, S. M. Chang, E. Graham, J.-M. Guise, S. Ip, M. A. Maglione, D. C. McCrory, M. McPheeters, S. J. Newberry, P. Sista, and C. M. White. 2014. A proposed approach may help systematic reviews retain needed expertise while minimizing bias from nonfinancial conflicts of interest. Journal of Clinical Epidemiology 67(11):1229-1238.

Willett, W. 2016 (unpublished). Invited perspectives: What two recommendations would you make to USDA/HHS related to the Dietary Guidelines Advisory Committee selection process. Presentation to the Committee to Review the Process to Update the Dietary Guidelines for Americans.