to a management mode of program operations. This shift will require strengthened management systems for planning, coordination, and evaluation in the NIH Director's Office.

NIH is a highly decentralized organization, a structure appropriate to a basic research enterprise. The agency comprises 13 research institutes, 5 research centers, 2 divisions, the National Library of Medicine, and the Office of the NIH Director, all of which conduct AIDS-related activities (Figure 2.2). Organizationally, NIH is further decentralized by the separation of the grants review process from extramural program management. All research grant applications receive a first-level review for scientific merit by independent disciplinary study sections of extramural experts, after which the applications are sent to the appropriate institutes for decisions on funding. Organizational analysts consider this interplay between the categorical institutes and the disciplinary study sections to be the genius of NIH's organization: “The primarily disease-based institutes enable Congress to understand, appreciate, and support the research accomplishments of the institutes, and also to express concerns and priorities about the need for further research. The study sections, on the other hand, cut across institute lines and ensure that appropriate scientific talent and ideas are brought to bear on the problems” (IOM/NAS, 1984:1). As a result, NIH has been able to sustain high-quality research, address public scientific concerns, as defined by Congress, and support basic biomedical research on which future advances will surely depend. It is evident, however, that this decentralized structure is not designed to respond as swiftly and efficiently as a more hierarchical organization to situations that require a coordinated, agency-wide response.

Within this highly decentralized organization, the NIH director and a small staff are expected to provide leadership in the interplay between the public and the scientific community. They must also coordinate programs that cross institute lines (but not micromanage the research programs of the institutes) and oversee common housekeeping functions for the institution as a whole. In the case of AIDS, each institute manages its AIDS activities as part of its overall program, involving the institute's staff, senior leadership, board of scientific counselors (which oversees the intramural program), and national advisory council (which discusses policy and program issues, reviews program concepts, and approves each grant). In most cases, AIDS activities are carried out within the institute's regular organizational structure; only NIAID, because of the large size of its AIDS program (53 percent of NIAID's budget in fiscal year 1990), has set up an organizationally separate AIDS activity. At the same time, AIDS-related efforts within the institutes are subject to an unprecedented (for NIH) degree of coordination and direction by the NIH Director's Office. The goals of this oversight are many: ensuring that information is shared and advances communicated to NIH's various publics, identifying and exploiting research opportunities, filling gaps in the research program, avoiding duplication of effort, and seeing that the overall program is adequate and balanced.

Historically, large new research programs on health problems of major public concern have usually been handled organizationally at NIH by creating a new institute or center. The committee considered and rejected the option of creating a national AIDS institute. The arguments for such an institute are that it would

• upgrade the status and visibility, and therefore the funding, of the research area;

• accelerate research progress by focusing research efforts in an integrated program (comprising, for example, basic, clinical, epidemiological, nursing, and behavioral research) and giving attention to such related activities as communications, training, and research resources; and

Front Matter (R1-R12)

Summary (1-18)

1 AIDS: The Research Challenge (19-34)

2 Managing the NIH AIDS Research Program (35-52)

3 Elements of the NIH AIDS Research Program (53-94)

4 Supporting the NIH AIDS Research Program (95-130)

Appendix A Study Issues (131-132)

Appendix B Persons Who Were Interviewed by Committee or Staff or Who Provided Written Materials (133-138)

Appendix C National Institutes of Health AIDS Program Advisory Committee (139-140)

Appendix D Abbreviations and Acronyms (141-142)