The second charge to the committee was to "assess the process by which research funding decisions are made." This is probably the least-understood aspect of NIH operations and is closely tied to the equally complex budgeting process that determines (subject to OMB and congressional approval) how much funding each institute, program, and mechanism receives.

Each of the 21 separate institutes and centers has its own legislative history, mandate, and budget from Congress, and each differs from the others with respect to its goals, scope, absolute size, and allocation of funding between extramural and intramural research. This highly decentralized and complex structure for the administration and conduct of research has both strengths and weaknesses. Although this diversity has contributed to a research establishment of great vitality and research productivity, it has led to an administrative structure that prevents uniform oversight and planning, hinders coordination of programs that involve several or all of the institutes, and obscures public comprehension. For example, from the short descriptions that each institute and center provided, the committee saw substantial differences in their priority-setting processes, such as the extent to which and how they involve outside groups.

In recent years, the director of NIH has been exercising increased authority to coordinate NIH responses to opportunities and needs that involve multiple institutes. This chapter addresses the adequacy of the priority-setting processes and the analytical and planning capacities needed to coordinate an organization as decentralized as NIH.

The NIH booklet Setting Research Priorities has fairly full discussions of priority-setting criteria and their complexity, descriptions of the various mechanisms used to solicit advice, and lists of the types of individuals, groups, and organizations that are routinely consulted, but it contains little about the planning and budgeting processes that result in the annual budgets and the expanded and new initiatives that those budgets typically contain. The best description was probably provided by NIH Director Harold Varmus to the House appropriations subcommittee in 1997 (U.S. House of Representatives, 1997a:73-76).

In those hearings, the NIH director described priority setting as continuous and occurring at two pivotal levels: at the level of the institute directors and at the level of the director of NIH. The central element of the NIH planning process is the annual meeting of the NIH Leadership Forum, composed of the NIH director, the institute directors, and NIH administrators responsible for major trans-NIH research and management issues (e.g., women's health, rare diseases, and alternative medicine).

The actual dollar amount allocated to NIH is the result of a complex process of negotiations among OMB, DHHS, the NIH director, and Congress. Because of NIH's commitments to infrastructure and multiyear grants (its commitment base), each year it can realign only a fraction of its budget.¹ More specifically, the NIH director can set new priorities only with additional funds received beyond the commitment base.

Historically, an expansion of research in one area has resulted in a decrease in another area. Only in times of real growth have significant realignments been made. A further constraint on change comes from the fact that each of the institutes receives a separate appropriation from Congress, as do the Office of the Director, the Office of AIDS Research, and buildings and facilities. Budgets are put together from the bottom up (from the institutes and centers and through the NIH director) and are reconciled with budget limits and programmatic priorities imposed at higher levels within NIH, DHHS, and OMB in a long, complex, and interactive process. The institute budget proposals are negotiated with DHHS and OMB through the NIH director, who submits an NIH budget overview to the appropriations committees along with the individual institutes' budgets and, most importantly, works with the appropriations committees on how to distribute any budget increases. The director of NIH plays an important role in deciding how much each institute can ask for within the budget total imposed by DHHS and OMB and can, in this way, affect allocations.

Although Congress looks to the director for advice on how to allocate any budget increases that it might give to NIH, the director is limited in this exchange by the requirement to adhere to the budget and priorities set forth by the administration (via NIH negotiations with DHHS and OMB). Thus, the NIH director is unable to communicate directly to Congress, early in the budget process, his or her independent judgment regarding the best use of NIH funds in the coming year.² One outcome of this process is that the NIH director at times has been found to be in the position of defending a budget request that is lower than the congressional mark or that reflects priorities that differ from those of Congress and perhaps even those of NIH itself (for example, if a special research directive was issued by DHHS or the White House).³ Thus, the NIH director probably has the greatest, albeit limited, influence on NIH priority setting during the annual budget process. After the budget is passed, the appropriations are made directly to each institute, which then controls the use of the funds.

Aside from the budget process, which provides little opportunity for radical priority setting, the institutes have two broad opportunities for creating priorities: deciding (1) which grant applications to support through the peer-review process and⁴ (2) which research topics to authorize as part of their portfolio. The former opportunity is largely not within each institute's control, whereas the latter is decided through an iterative process involving the NIH director, the institute directors, and the various constituencies served by NIH and its research portfolio. The processes used to accomplish each of those goals are quite different.

The relative authority of the NIH director has long been a subject of debate and discussion, with many analysts calling for greater influence and centralization. In 1984 IOM wrote, "the Office of the Director is the logical locus for a central coordination activity at NIH" (Institute of Medicine, 1984). Historically, the authority of the director has been limited largely by the size and complexity of the NIH organization and by its budget process. More recently, in 1991, the director's authority was expanded in two specific areas: (1) the director was given the ability to transfer up to 1 percent of funds across institute lines, and (2) the director was given a discretionary fund with which to seed selected areas of research.⁵ In addition, the director has played a more active role in priority setting by requiring the institute directors to receive part of their budget increases through an initiative called the NIH Areas of Research Emphasis (described below).

With a unique NIH-wide perspective and ultimate responsibility, the NIH director looks at each institute's programs "with particular emphasis on needs and opportunities for crosscutting efforts" (U.S. House of Representatives, 1997a:73-76). The director accomplishes this through a system of regular senior staff meetings with deputy directors for intramural and extramural research and associate directors for science policy, legislative affairs, and communications and through feedback from interest groups and briefing sessions with institute directors. In addition, the Office of the Director has acquired a number of offices to advise the director of needs and opportunities and to coordinate NIH activities in various areas of concern, for example, rare diseases, alternative medicine, women's health, minority health, and behavioral and social research.

The NIH director oversees 15 offices with a combined budget of $201 million in FY 1998. Roughly one-fourth of this budget supports the operations of the Office of the Director; the remainder supports special functions, such as the Office of Women's Health and the Office of Alternative Medicine. An issue of considerable concern to the Office of the Director in recent years has been the congressional requirement that NIH hold its annual increase in research management and support levels to 1 percent. This has restricted the ability of NIH to engage in more planning and evaluation of programs since those activities are funded through that mechanism.

The current NIH director has used the transfer authority for the past 3 years as a part of the regular budget process to target specific research opportunities. The NIH director has indicated that he considers other factors as well when making these transfers, such as the effect on new and competing grants, success rates, overall effects on programs "assessed" for transfer funds, and the future impact on programs receiving funds (which will have to absorb the out-year costs).

To receive transfer authority funds, proposals are submitted by institute directors and are reviewed by ad hoc panels of outside reviewers using the criteria of (1) scientific merit and (2) relevance to a broad spectrum of the biomedical research community. In 1995, 1996, and 1997, the NIH director transferred $8.4 million, $24.8 million, and $33.8 million of a possible $110.5 million, respectively. Examples of accomplishments resulting from the receipt of transfer funds include construction of a map of the rat genome; establishment of a national resource for small-angle X-ray scattering and X-ray spectroscopy; establishment of an NIH consortium for research in neurodegenerative and brain disorders; collaborative efforts in programs related to bone, cartilage, and dental diseases; projects in the neurobiology and genetics of autism; and improved preparedness for rapid responses to emerging diseases.

In recent years, the director's power to lead NIH has been strengthened by authorizations for a director's discretionary fund (about $10 million a year). In submitting the NIH budget, the NIH director can identify areas of special emphasis that will receive incentive funding. This approach of providing incentive funding for a focused initiative has stimulated research and fostered collaborations among institutes. Sometimes the research areas are identified at levels above NIH. For example, within the area of the biology of brain disorders, an initiative spurred by the president's commitment to this area was begun in 1997 to encourage research on spinal cord injury. The initiative began with a major workshop that convened experts on spinal cord injury, along with leaders from other scientific areas, to foster new ideas and collaborations. Following the workshop, a program announcement highlighting important relevant research topics was issued. Other funds were used to support such efforts as clinical research training awards, research on an influenza vaccine, genomic research, and an African multilateral initiative on malaria.

In recent years the director has played a more active role in priority setting by requiring the institute directors to receive part of their budget increases through the NIH Areas of Research Emphasis. This initiative has been developed by the current NIH director as a way of planning for and prioritizing the use of funds resulting from the annual incremental budget increases. These Areas of Research Emphasis "provide a broad framework for building new initiatives to confront specific diseases or classes of diseases through laboratory and clinical research" (U.S. Department of Health and Human Services, 1998:5). They typically include research areas that are critical to the scientific agenda of more than one institute and that are fundamentally important to progress in understanding a wide variety of diseases.

The funds available for this initiative can be sizable, depending on the amount of the overall NIH budget increase. For example, the FY 1999 budget increase is more than $1 billion. Of that increase NIH plans to use 52 percent, or $598 million, to support the six Areas of Research Emphasis, including the biology of brain disorders, new approaches to pathogenesis, new preventive strategies against disease, new avenues for the development of therapeutics, genetic medicine, and computers and advanced instrumentation.

These priority initiatives are identified through discussions involving many components of NIH leadership and its stakeholder communities, culminating in a dialogue between the NIH director and each institute or center director. The funds that are allocated through this process become part of the budget base of each institute that receives them. This process has thus become an innovative and influential means of setting research priorities across NIH.

In 1965 a presidentially appointed committee studied NIH and made a series of recommendations, one of which was for a policy council in the director's office:

A new advisory group should be established to assist the Office of the Director of NIH in the making of major plans and policies, especially those related to the allocation of NIH funds and resources. Referred to as the "Policy and Planning Council," the new unit should consist of experienced and distinguished scientists together with a suitable minority of outstanding non-scientists with wise understanding of and dedication to fulfillment of the Nation's needs (White House, 1965).⁶

The Advisory Committee to the Director (ACD) was established in 1966 to advise the director and the secretary of the U.S. Department of Health, Education, and Welfare (now DHHS) on NIH policy matters and to make recommendations regarding program development, resource allocation, policies, and regulations. The ACD is chaired by the secretary or the secretary's designee and is chartered to have 16 members: 11 researchers from the private sector and academic research communities and 5 representatives of the public. The President's Biomedical Research Panel recommended in 1976 that ACD be reorganized and strengthened as the NIH Advisory Board, and the IOM committee on the organization of NIH also recommended in 1984 greater use of ACD, stating:

The current Director's Advisory Committee should be converted to a stronger and more independent NIH Policy and Planning Council to provide for the Director a continuous evaluation of the research mission and function of NIH, and to advise the Director in the formulation of long-term plans and in setting research priorities (Institute of Medicine, 1984:5).

Each NIH director uses ACD in different ways and to various degrees. The NIH director can also use ACD to form special expert panels to provide advice on crosscutting issues. Recent examples include ad hoc panels on gene therapy, clinical research support, human embryo research, the NIH intramural research program, and the peer-review system. Although over the past three decades several committees have recommended that the NIH director use ACD in a more proactive and evaluative way, it is not clear that this has been the case.

Although the NIH director can influence priority setting through participation in the annual budget process, the use of blue ribbon panels to recommend new or reorganized programs, and the use of budgetary authority to launch crosscutting areas of research emphasis, the levels of funding for most of the budget, and therefore the research allocations, are determined at the institute level. NIH is a very large, complex, and decentralized enterprise in which most programming and budgeting activities come from the bottom up.

During its deliberations, the IOM committee received descriptions of each institute's priority-setting process. They reveal widely varying priority-setting processes. Some such variation is appropriate, because the institutes vary in their missions, histories, leadership, sizes, and complexities. Some institutes have more formalized planning and budgeting processes, which tend to be more easily understood by (i.e., transparent to) external constituencies. Some institutes routinely incorporate a wider range of external views, including the views of patients and their advocates.

Some institutes, such as the National Eye Institute, have a long-standing formal planning process for setting priorities. Others, such as NCI, have engaged in an elaborate set of activities to review its programs and progress and have expanded input from consumer advocates. For example, NCI relies on advisory committees (called working groups or review groups) in its priority-setting and planning process, with a mix of members including intramural and extramural researchers, NCI and other federal officials, consumer and patient advocates, and representatives of professional societies and organizations. Still others employ rigorous and well-structured internal and external planning reviews. For example, to review its priorities the National Institute of Allergy and Infectious Diseases (NIAID) engages in twice-annual retreats involving representatives of the scientific community and the public (see Box 3-1 for an example).

Each institute also has a national advisory council, the makeup of which is usually determined in the enabling language that created the institute. The national advisory councils review and approve all grants and provide policy advice to the institute director. In most cases, there are 18 members: 12 scientists, chosen for their research expertise, and 6 nonscientists (usually physicians and other professionals rather than representatives of patient groups or groups with special health problems).

Each institute also has a Board of Scientific Counselors, which provides advice on and reviews the intramural research program, and a Board of Scientific Advisors, which provides advice on the extramural programs. Institute directors often form working groups and ad hoc panels to help them address key issues in their research programs. Many institutes have offices for consumer issues. The institutes frequently engage in program reviews or progress reviews to determine if the institute is investing in the appropriate areas of research. NIH has 141 chartered advisory committees (see Appendix A) and makes use of numerous ad hoc advisory committees.

In summary, the institutes have varied systems for receiving advice, planning, and setting priorities. However, there is tremendous variability in the levels of diversity and activism of these activities and in the specificities, scopes, and time frames of the plans developed. For example, some institutes appear to adopt plans developed by a proactive staff with the endorsement of advisory groups, whereas others follow closely the recommendations of external advisory groups.

Priority setting is decentralized at NIH, which is appropriate for a research organization in which those closest to a problem are in the best position to decide on approaches and in which expertise is highly specialized. The priority-setting processes also vary from institute to institute and from area to area within institutes. Some such variation is appropriate, because the institutes vary in their missions, histories, leadership, sizes, and complexities. The committee did find that some institutes and programs have priority-setting processes that incorporate a broader range of inputs and views, including those of nonresearchers and nonclinicians.

More recently, NIH has been making decisions on priorities and funding allocations that are more centralized than in the past; that is, NIH is looking across traditionally independent institutes and centers and focusing on certain crosscutting needs and opportunities where joint or unified action is desirable. This trend stems from the growing realization that common biological processes underlie diseases that were previously seen as different or that important diseases and other health problems are more complex than was previously thought, affect more organs and processes than was previously realized, and happen to be addressed in more than one institute.

The committee concluded that the Office of the Director of NIH needs an increased capacity to analyze such crosscutting needs and opportunities and to interact with the public (the latter process-related issue is addressed separately in Chapter 4). Improvement requires a more central role for the NIH director and more uniformity in the data and analyses presented to the Office of the Director.

Recommendation 5. In exercising the overall authority to oversee and coordinate the priority-setting process, the NIH director should receive from the directors of all of the institutes and centers multiyear strategic plans, including budget scenarios, in a standard format on an annual basis.

In any organization, change toward centralization raises concerns about accountability. As the authority of the director is strengthened, greater accountability of the Office of the Director could be achieved through a strengthened ACD, one that is more actively engaged in the NIH priority-setting process.

Recommendation 6. The director of NIH should increase the involvement of the Advisory Committee to the Director in the priority-setting process. The diversity of the committee's membership should be increased, particularly with respect to its public members.

The charter of the current ACD gives it this responsibility. However, it meets infrequently and provides advice on matters that do not appear to be sufficient to fulfill this responsibility. Furthermore, although 5 of the 16 ACD slots are for nonscientists with "wise understanding of and dedication to fulfillment of the Nation's needs" (White House, 1965), a review of ACD membership reveals that these reserved slots are not being filled with enough public members consisting of individuals from organized voluntary groups, individuals active in advocating for those with specific diseases or medical conditions, and members of organizations and individuals who represent population groups with special or disproportionate health problems. Not having more public members from these groups is a missed opportunity and has resulted in the perception of some groups that NIH does not encourage public input at the highest levels. The committee recommends that the director of NIH work more actively with the secretary of DHHS in appointing public members to ACD and in providing guidance to ACD on how it can work more effectively in an advisory capacity.

NOTES

¹In addition, it is important to note that NIH has several mandated activities requiring collaboration with other federal agencies, such as the Centers for Disease Control and Prevention, the U.S. Food and Drug Administration, the Environmental Protection Agency, the U.S. Department of Defense, and NSF.

²In contrast, by law, the NCI director can bypass DHHS and OMB in the budget process and go directly to Congress with a budget request that matches directly the professional judgment of NCI and its leadership regarding research needs and priorities. NCI is the only institute with this authority.

³For this and other reasons, past reports have recommended that the NIH director report at least to a higher level—to the secretary of DHHS rather than the assistant secretary for health or to the president directly—if not directly to Congress. For example, in 1984 IOM recommended that "the Secretary of Health and Human Services should delegate to the Director of NIH the authority, direction, and control over NIH that the position does not now possess" (Institute of Medicine, 1984).

⁴Peer review is the system used to help decide research funding among competing applications. Although it is not explicitly a system of priority setting, it does rank proposals within a research area in order of scientific impact or significance, which de facto determines the de facto priorities among problems and approaches in that area. The proposals are referred to an institute for funding, and the institute's national advisory council may fund lower-ranked proposals that would have greater relevance to the institute's mission. The NIH peer-review system is under extensive review and is not addressed further in this report.

⁵Both authorities were recommended in a 1984 IOM report. In addition, the discretionary fund was recommended in a 1988 IOM analysis, Report of a Study: A Healthy NIH Intramural Program: Structural Change or Administrative Remedies? (Institute of Medicine, 1988).

⁶The report went on to say that the council should not only be called on by the director of NIH "but it should be encouraged by the congressional committees concerned with NIH to participate in the annual budgetary hearings and in the formulation of the NIH programs."