Mr. Chairman and Members of the Subcommittee, I am Harold Varmus, Director of the
National Institutes of Health. I am pleased to appear before you to discuss the research
funding process at the NIH. I want to thank you for the opportunity to discuss this
Congress, patient/health advocacy groups, and the scientific community have a
long-standing interest in how NIH sets priorities and allocates funds for medical
research. These constituencies are concerned about how the NIH accounts for its funding
decisions and the means by which the public can, and does, influence them.
A brief history
In 1997, I testified at two hearings on priority settingthe first in May, before the Subcommittee on Public
Health and Safety of the Senate Committee on Labor and Human Resources, and the second in
June, before the House Subcommittee on Labor, HHS, and Education, Committee on
Appropriations. At both of these hearings, the criteria and processes by which NIH
allocates research funds were examined and contrasted with the role of Congress in
authorizing and appropriating funds for medical research.
After these hearings and in response to public and Congressional interest in how NIH
sets priorities, I created the NIH Working Group on Priority Setting. This group,
consisting of 15 senior NIH staff, was charged with developing a document that would
clearly describe the principles and mechanisms by which NIH allocates its funds. In 1997,
NIH published Setting Research Priorities at the National Institutes of Health.
Although this booklet has been widely distributed and generally well-received www.nih.gov/news/ResPriority/priority.htm),
the public and members of Congress continued to express concern about the
priority setting process and the means by which the public can influence NIH
decision-making. In an effort to further address this issue, members of the Senate
Subcommittee proposed, through the FY 1998 Labor, HHS, Education Appropriations Act, that
the Institute of Medicine (IOM) conduct an independent study of decision-making at the NIH
and how resource allocation is influenced by Congress and the public.
The IOM Committee released its report, "Scientific
Opportunities and Public Needs: Improving Priority Setting at the National Institutes of
Health" in July of last year. The report
contains twelve helpful recommendations for improving priority setting and consideration
of public input at the NIH; ten of these recommendations were directed to NIH leadership,
while two of the recommendations were directed to Congress. More recently, the FY 1999
House Appropriations report encouraged the NIH to implement these recommendations and
requested a report from the NIH on the status of implementation; the report was submitted
to Congress in February of this year. (See Attachment)
What are NIH's Criteria for Allocation of
The allocation of funds to medical research is complex. Congress establishes the level
of available resources to NIH through separate appropriation accounts for each research
institute and center. Within these general parameters each institute and center must
decide which specific applications to fund and whether to emphasize certain research
topics within its authorized domain such as child health, cancer, cardiovascular disease,
diabetes, or infectious disease. These decisions are also constrained by the commitment
base, i.e., funding decisions made in previous years which limit the number of dollars
available for new grants or new initiatives. The net effect of these multiple processes
and decisions determines how much of the entire NIH budget is devoted to work in certain
scientific disciplines or on particular diseases.
There are five broad criteria that guide the planning and spending of the NIH budget;
these criteria were fully endorsed in Recommendation 1 of the IOM Report. First, the NIH
is committed to supporting work of the highest scientific caliber by ensuring rigorous
peer review. Second, the NIH must seize those opportunities that offer the best prospects
for new knowledge and for improving the prevention and treatment of disease. As important
as it is that we fund research on specific diseases, we must also fund research programs,
such as the Human Genome Project, that yield knowledge applicable to a broad range of
biological questions and clinical problems.
Third, because we cannot know in advance exactly when and where major discoveries will
occur, we also need to maintain a diverse research portfolio. For example, while we
continue to pursue advances in cell biology and genetics, we are also expanding our effort
in clinical research by initiating new training and career development programs for
clinical investigators; increasing funds to General Clinical Research Centers;
strengthening clinical research in the intramural program; expanding the number of
clinical trials; and developing a Clinical Trials Database to ensure that patients and
physicians know where and how to enroll in trials. Portfolio diversity is also evident in
our commitment to train, support, and encourage scientists in allied fields, such as
physics, engineering, chemistry, and computer science. This is accomplished by creating a
Bioengineering Consortium; by supporting instrumentation development, such as the
construction of new beam lines for structural biology; by developing interdisciplinary
training programs for drug development; and by attrActing
and training young computer
scientists into the growing field of bioinformatics.
A fourth criterion, and one that has drawn particular attention, is public health need
as measured by the burden of disease. NIH gathers, analyzes, considers, and disseminates
data on all of the factors that describe burden of disease, including incidence,
prevalence, mortality, and morbidity, among others. These data are obtained from a variety
of Federal agencies, such as the CDC, AHCPR, HCFA, the U.S. Census Bureau and voluntary
health organizations. The NIH also funds longitudinal studies, short-term one time
studies, and recurring surveys on disease risk factors, epidemiology, etiology, and
natural history to ensure that we have all the necessary data to inform our
Fifth, NIH must build and maintain the necessary infrastructure for the conduct of
research. Productive science cannot be done without well-equipped laboratories,
well-trained scientists or modern and safe research facilities. To this end, funds must be
devoted to attrActing
, training and supporting young investigators and mid-career
investigators who serve an important role as mentors. NIH funds must also be available to
upgrade laboratories with state-of-the-art instrumentation, to construct and renovate
laboratory facilities, and for the purchase of expensive equipment.
Does the NIH plan science and, if so, how?
Because research, by definition, is the attempt to discover what is unknown, it is
unpredictable. And because it is unpredictable, there are genuine constraints on the
ability to plan science. History has repeatedly shown the benefits of allowing research to
be governed by the imagination and productivity of individual scientists, not by a formal
plan for alleviating specific diseases we do not yet fully understand.
While it is not possible to plan for specific research outcomes, it is, however,
possible to plan initiatives and set broad programmatic goals. Strategic planning has
always been carried out at the NIH, although the processes within the ICs have not always
been uniformly clear to the public. Some Institutes have formal planning processes and
publish the results of these deliberations, while ongoing planning processes in other
Institutes have been less visible. I have asked each IC to develop a 2-5 year strategic
plan, which includes input from scientists, patient advocates, and health care providers
with the goal of making these written plans available to the Administration, Congress, and
the public early in FY 2000.
There are many important yet competing factors that each IC must consider in planning
how, and by what mechanisms, its funds should be spent. For example, how many dollars
should be allocated to laboratory research vs. clinical research? to
investigator-initiated research vs. targeted disease-specific research? to research
project grants (RPGs) vs. contracts or centers? to intramural vs. extramural research? to
training vs. instrumentation or buildings and facilities? These decisions must be closely
tailored to the IC's overall research objectives and to the specific scientific
initiatives identified during the planning process.
How, and from whom, does NIH seek advice in setting priorities?
The factors that influence the planning and spending of budgets are multifaceted, so
opinions about them are solicited and provided from many quarters--the extramural
scientific community, patient advocacy groups, health care providers, Congress and the
Administration, as well as the NIH staff. In an effort to ensure that we hear from all of
those interested in, and affected by, medical research, we gather these opinions through
many means and over the course of each year.
The ICs have many established means for reviewing scientific progress in their areas of
responsibility, for developing long-range research objectives, and for formulating annual
budgetary plans and research initiatives in consultation with scientists and the public.
They use review groups composed of accomplished investigators (recently some have included
lay members) to evaluate grant applications for scientific merit. Each year many
conferences and workshops are organized to encourage scientists from diverse disciplines
and lay disease advocates to come together to examine and stimulate new areas of research.
IC Directors and NIH staff also frequently consult with members of other Federal agencies,
with the OMB and DHHS, and with Congressional members and staff on a variety of common
concerns. Some NIH ICs also engage the lay public by creating advisory groups like the NCI
Director's Consumer Liaison Group, while others,
such as NIDA and NIEHS, sponsor town meetings around the country to seek public input,
involving community leaders and groups, local schools, and state or local government
officials. In the past few years, the NIH has also made frequent use of extramural
advisory groups to assess trans-NIH activities (for example, the intramural research
program, the Clinical Center, gene therapy, clinical research, and AIDS research) and to
recommend budgetary and programmatic changes in those areas.
Along with these long-standing efforts to seek advice, NIH has undertaken several new
efforts which seek to build upon and improve both access to and communication from the
NIH. For example, within the Office of the Director, the Office of Communications is being
expanded and is now named the Office of Communications and Public Liaison to reflect its
public liaison functions. Each IC has an Office of Public Liaison which provides
information about an IC's research activities
and ensures that each Institute has a conduit through which public voices can be heard in
the Institute's deliberations on research directions and priorities. While the functions
of these offices are not new to communicate
with the NIH's many constituencies -- many of
them have recently been renamed so as to clearly identify them to the interested public.
The NIH also launched a new Web site to serve as a focal point for NIH public liaison
activities (www.nih.gov/welcome/publicliaison). In addition,
the new Director's Council of Public
Representatives, which met for the first time last month, provides another avenue for
greater public involvement in NIH's activities
Why is disease burden only a partial guide to spending NIH's research dollars?
In spite of NIH's extensive efforts to gather
and analyze data, information on disease burden is imperfect. There is no common or
accepted measure for disease burden. Morbidity, mortality, incidence, prevalence, the cost
of direct health care services or the cost of unreimbursed family care, and loss of work
productivity have all been touted as useful metrics for burden of disease. But each of
these factors is incomplete. The nature of burden varies from one condition to another.
Some diseases result in premature death while others result in diminished functioning.
Some terminal conditions require short-term costly health care, while others cause pain
and suffering over many years. To further explore the potential utilityand strengths and limitationsof disease-specific burden of illness, this summer we
are convening a small group of experts to identify data sources, review models for the use
of burden/cost of disease data, and explore how NIH might more effectively use this data.
Furthermore, estimates of spending by disease, while consistent from year to year for
any single disease, often do not allow meaningful comparisons across diseases. The
spending figures calculated for a specific disease are the result of a complex algorithm
of laboratory and clinical research efforts, which appear to be related to that disease.
In many cases, the most basic research on cellular function or gene expression may not be
clearly attributable to a specific disease. Nevertheless, findings from such research
often lead to real improvements in the prevention, diagnosis or treatment of that disease.
For example, recent progress in developing effective therapies for patients with AIDS was
based on much earlier cancer research on retroviruses found in chickens, mice, and other
animals. We now use drugs designed to inhibit the enzymes made by HIV's genes, diagnose
infection and follow the effects of therapy by measuring viral genomes in the blood, and
study resistance to treatment by detecting mutations in viral genes.
Calculations of spending by disease also ignore a very important element of resource
allocationthe importance of funding "enabling technologies."These are knowledge and technology platforms that
serve a broad range of scientific fields and disease-specific research. I already
mentioned one such program, the Human Genome Project; others include the Trans-NIH Mouse
Initiative and the Brain Molecular Anatomy Project. These programs are not easily assigned
to diseases and yet they are critical components of much, if not all, disease-specific
In sum, the complexities of assigning dollars to disease-specific research inevitably
lead to significant variations in the number of dollars spent on one disease as compared
to another. And because public health need is one of several criteria NIH uses to allocate
research funds, we can never expect a perfect correlation between disease specific funding
and disease-specific burden.
Can money alone drive scientific advance?
Advances in science are not a commodity and cannot be purchased by the simple
expenditure of dollars. Several important components of the research enterprise must be in
place for new dollars to yield real progress. The elements can be defined, although they
often are difficult to obtain. In the best case, public health need and scientific
opportunity co-exist with highly trained, creative investigators and modern laboratories
and research hospitals.
New scientific efforts are also driven by evidence that under-explored opportunities
exist and that they can attract talented investigatorsoften
newly trained scientists or scientists from other fields--who will then propose
meritorious projects. To this end, the NIH employs a variety of means to recruit new
talent to a scientific problem, including advertising an IC's interest in making funds
available to pursue a new scientific opportunity or a public health challenge through
program announcements, requests for applications, and requests for contract proposals;
inviting scientists from allied fields to workshops that highlight opportunities and needs
in an underserved field of medical research; and supporting training programs to encourage
new scientists to work in a designated area.
Mr. Chairman, I appreciate your providing a forum to present these views about an
important, contentious, and complex issue. I would be pleased to answer any questions you