The Department of Defense’s (DOD’s) Congressionally Directed Medical Research Programs (CDMRP) originated in 1992 in response to a congressional mandate to create and manage a research effort aimed at ending breast cancer. It has expanded to include major research support efforts addressing prostate cancer, ovarian cancer, and neurofibromatosis, and smaller scale and short-term initiatives on other health problems. CDMRP is a program administered by the U.S. Army that supports research by scientists in universities, industrial laboratories, federal and state government agencies, and other research institutions. The program focuses on basic and clinical research and training and has developed a reputation for consumer participation in priority setting and peer review, innovative award mechanisms, and support of cutting-edge research. Unlike most federal health research support mechanisms, CDMRP is appropriated on a year-to-year basis with program-specific funding levels.
Recently, Congress became concerned about funding increases for CDMRP, because of competing demands on the military budget, especially the wars in Iraq and Afghanistan. In 2004, at congressional request, DOD asked the Institute of Medicine (IOM) to report on the possibilities of augmenting program funding from nonfederal sources. IOM was asked to form a committee to identify mechanisms that could be used to leverage such funding, assess the impacts of alternative nonfederal sources and mechanisms of funding on CDMRP, and identify any legal or regulatory barriers to leveraging nonfederal funding. The Committee on Alternative Funding Strategies for DOD’s Peer Reviewed Biomedical Research Programs was not asked to evaluate CDMRP; recommend whether it should be
continued, restructured, or phased out; or look at the potential for co-funding research with other federal programs.
The report focuses on nonfederal funding sources and mechanisms and their potential effects on the program if adopted. The committee identified many potential nonfederal sources but concluded that the prospects for augmenting CDMRP funding from these sources are modest, at best, especially regarding “new” funds that would not otherwise be devoted to biomedical research. Voluntary collaborations can, however, leverage research results that cannot be achieved by individual funders working separately, for example, through the creation of synergies, critical mass, economies of scale, and other ways that make the whole greater than the sum of its parts.
The committee found numerous examples of federal agencies leveraging nonfederal funds. One common approach is to require cost sharing or matching by grantees; another is to encourage voluntary public-private collaborations between government and other donors or between university grantees and industry. However, substantial cost sharing or matching is costly for grantees to secure and document and for the granting agency to oversee, while successful public-private funding collaborations depend on an alignment of interests among the contributing parties to carry out a particular project or program, which most commonly occurs with clinical research or research directed at the development of products (e.g., diagnostics, drugs, vaccines, devices).
CDMRP has the authority to require cost sharing by grant recipients (which it exercises by requiring awardee institutions to provide the facilities and equipment for conducting research projects and the faculty for training programs), but the program lacks the authority to accept private funds and a foundation would need to be established to solicit private funds for collaborative projects. The committee developed recommendations for CDMRP to facilitate federal funding that emphasize voluntary collaborations in funding research and that focus on providing CDMRP with the authority to engage in jointly funded projects and programs while ensuring that the best features of the current program are not undermined.
POTENTIAL SOURCES OF NONFEDERAL FUNDS
Sources of potential additional funding for medical research supported by DOD include for-profit companies, venture capital firms, foundations and other philanthropies, and state governments. Because CDMRP already collaborates and co-funds projects with other federal agencies, most notably the National Institutes of Health (NIH), the committee was asked to focus on nonfederal sources.
Approximately 10 percent of pharmaceutical industry expenditures on research and development (R&D) goes to basic research. Thus, industry spends a great deal of money on biomedical research, but little on the types of activities
that the majority of CDMRP funding currently supports. First, industry is most interested in short-term research and development projects with commercial promise, such as the development of diagnostics, therapeutics, and devices and the creation of research tools and databases. Efforts to increase industry support of CDMRP programs could shift priorities toward such activities. Second, for-profit co-funding of awards to university scientists could introduce a number of issues and potential risks, including conflicts of interest for faculty who could benefit financially from their university research; increased secrecy and other restrictions on the dissemination of industrial research results; reduced faculty time commitments to the activities of the university; and the use of students in conducting privately funded research. Third, there is no avenue currently by which industry can contribute funding directly to CDMRP, although this problem could be remedied (see below).
Venture capital is invested in late-stage R&D, with the expectation of commercial products. Representatives of these sectors at the committee’s workshop said that they would not be likely to give CDMRP a blank check by transferring funds to it directly and that they would expect to play a significant role in determining where their research dollars were spent. At the same time, they indicated strong interest in collaborating when mutual interests are identified.
It is a common strategy of philanthropies, state governments, and universities to leverage federal funds to achieve their goals. However, foundations and public charities, such as the American Cancer Society, have limited resources compared with the federal government, and they focus on funding activities that are not well supported by the federal government, such as public health, or on activities that leverage federal funding, such as providing grants for exploratory research and new investigators. Some foundations collaborate with federal agencies on projects of mutual interest and even give funds to agencies to expand programs, which would be beneficial, but the amounts of money are modest relative to the appropriations for CDMRP.
Most states have economic development programs that are interested in funding medical research and biotechnology, often using dollars that are new and that, until recently, would not have been spent on medical research. But states, like industry, are generally most interested in funding research that has immediate commercial application, because the ultimate goal of the research is to build the state’s biotechnology industry and thus increase the number of jobs. Moreover, a principal thrust of most state economic development funds is to leverage multiple federal dollars for each state dollar, not the other way around. In addition, states must balance health research against other needs such as improved transit and the funding of manufacturing plants. In general, state research funds would be most available for biotechnology development and not innovative exploratory health research. Finally, strategies contingent on the availability of state funding would tend to direct federal funds to states that wish to mount research programs to parallel those of CDMRP and that already have the means to do so.
In general, nonfederal funding leveraged by CDMRP would not necessarily be “new” money that expands the overall amount of support for biomedical research. In most cases, the nonfederal funds would be redirected from other areas of biomedical research. For example, one effect of a cost-sharing requirement in CDMRP programs might be that charitable organizations will be called on to provide matching funds that they would have spent themselves on the same areas of research. Although the total amount spent on the CDMRP program would be increased, the overall amount of funding for biomedical research would not change. The principal difference would be the additional administrative effort required to account for funding streams and to secure agreement on co-funding arrangements.
In conclusion, there are many potential sources of nonfederal funding for CDMRP, although the amounts of additional funding they could contribute are likely to be limited because their priorities differ from CDMRP’s research agenda. Nonfederal funders are least likely to fund exploratory research on new ideas or to support new investigators, which constitute the majority of CDMRP support. Nonfederal donors are most likely to be interested in collaborating on applied research and development work on potential products with near-term commercial payoffs—e.g., diagnostics, therapeutics, and devices. While collaborations might not leverage much funding, they can leverage research results by creating synergies or otherwise enabling research that cannot be done by individual funders, and CDMRP should take advantage of opportunities to achieve this kind of leverage.
MECHANISMS FOR SECURING OTHER FUNDING
Federal agencies that support R&D use a variety of mechanisms to leverage other funding. Broadly considered, these mechanisms fall into two groups. One set of mechanisms relies on requirements that awardees pay a portion of the costs of the project from nonfederal sources, called cost sharing or matching. The other set of mechanisms consists of voluntary collaborations between federal agencies and nonfederal donors in supporting research of mutual interest.
Cost-sharing or matching requirements are commonly used by agencies in development projects for which industry involvement is desired and from which industry stands to gain. Cost sharing is not usually required for programs that support basic research conducted by individual investigators or small research teams. CDMRP requires a small amount of cost sharing by expecting grantees to provide the facilities and equipment needed in the proposed research or, in the case of training grants, by paying for the mentoring and other nonstipend costs. Currently, CDMRP does not document the monetary value or extent of this cost sharing.
Matching fund requirements appear at first glance to be a good way to leverage funds but, for several reasons, they may not in fact increase the total amount of research. Indeed, they may add administrative costs and accounting
complexity without funding additional research. First, matching funds may be diverted from other research projects. Second, they are not free. The principal investigators and institutional research administrators must spend additional time lining up matching fund donors, planning the project, and preparing a more complex application. The time and effort that researchers and administrators spend searching for funding is time not spent conducting research and augmenting knowledge. In addition, some highly promising research may not be funded because of a lack of matching funds, and funding may be redistributed from new investigators to better known investigators who could attract other funding, wealthier institutions with greater means to provide the matching funds, or states with research programs able to match CDMRP funding. Finally, both the awardee institution and CDMRP would have to expand their staffs to account for and audit the matching funds to ensure that they are legitimate and not counted as matching on another federal grant.
Expectations that the program should leverage substantial funds through matching requirements would be likely to shift program priorities from basic research toward applied research and development, particularly if the additional funding came from industry, which is more interested in supporting the later stages of the R&D process. Industry and states have funded fairly basic research activities in some cases, however, such as biomarker development and stem cell research centers.
If well planned, voluntary collaborations between CDMRP and other organizations and agencies in funding research could, however, be mutually beneficial in terms of enlarging the base of research leading to better health outcomes, achieving a critical mass of resources needed for progress in a research area, and/ or promoting synergy among different sponsors with complementary knowledge, skills, techniques, and other resources. The funding arrangements involved in voluntary collaborations may include pooling funds for grants, but they also may include arrangements in which funding does not change hands, such as agreements to fund grants separately or to fund different but complementary parts of a project or for nonfederal donors to provide supplements to federal grantees.
Some federal agencies, such as NIH and the Centers for Disease Control and Prevention (CDC), have the authority to accept gifts for specified purposes and have foundations that can solicit private funds for their programs. If Congress wishes CDMRP to draw on foundations and private donors, it needs to create the authority to receive gifts to support research administered by CDMRP or in collaboration with CDMRP (see below).
ASSESSMENT OF ALTERNATIVE FUNDING SOURCES AND MECHANISMS FOR CDMRP
Despite initial skepticism in the scientific community about its location in DOD and the participation of consumers in peer review and priority setting,
CDMRP has shown that it has been an efficiently managed and scientifically productive effort and that it is a valuable component of the nation’s health research enterprise. Distinctive program features include its rigorous peer review of proposals for scientific merit and program relevance by outside reviewers that includes consumers; its inclusive priority setting process; its emphasis on exploratory high-risk/high-gain basic, translational, and clinical research projects and on research capacity building; and its holding of periodic national meetings to share results among the investigators and with the program’s constituencies.
Great care should be taken to ensure that changes in the program intended to leverage funding do not damage the features of the program that have made it efficient, driven by scientific priorities, and scientifically productive.
Currently, CDMRP has a relatively low administrative overhead of approximately 6 percent overall. Although increased cost-sharing requirements would impose costs chiefly on applicants and awardees, it also would require increased DOD staffing to ensure that those requirements are met. In addition, increased emphasis on raising outside funds for CDMRP programs would require additional staff to arrange and maintain relationships with the other funders. Likewise, establishing a foundation to plan collaborative programs and solicit funding from nonfederal sources would add to overhead costs.
CDMRP has used the two-tier review system recommended by the 1993 Institute of Medicine report to ensure that scientific quality and program relevance are the main determinants of the awards. Increased cost sharing would need to be carefully designed to ensure that the peer review system is not distorted, for example, by discouraging proposals from investigators representing institutions with little access to cost-sharing resources or by inciting a bidding war among applicants.
If CDMRP accepts funds from nonfederal sources—for example, donations from pharmaceutical or biotechnology companies to expand the pool of grants for a particular program of mutual interest—those funds should be distributed through the existing two-step peer review process to applicants who present the best proposals in terms of technical excellence and program relevance.
Currently, most funds appropriated to CDMRP support basic research, including a substantial number of exploratory grants. The program also supports new investigators and research programs in minority institutions. In addition, CDMRP funds research on the causes and prevention of the diseases these programs address. All of these are activities that industry is less likely to co-fund in favor of activities that support commercial development. Although foundations and disease charities do fund exploratory and basic research programs, and should be encouraged to engage in collaborations, their resources are relatively small when compared with those of industry or the federal government. The easiest way to attract nonfederal funding would be to change program priorities to emphasize the development and testing of drugs and vaccines and other efforts to develop commercial products; however, the advisability of such a shift is questionable given the amount of funding that industry and other federal agencies
already devote to such activities and in light of the need for additional basic—especially exploratory—research in understanding the diseases that CDMRP addresses. Such a shift likely would increase the concentration of work in areas that already enjoy ample funding at the expense of support of new ideas and new investigators, CDMRP’s signature area of strength.
As noted earlier, the risks of collaboration with for-profit firms include the imposition of secrecy on the scientific process, delays and bias in the reporting of research results, the shifting of research priorities toward near-term development rather than long-term research, and possibility of financial conflicts of interest for both research institutions and individual researchers. CDMRP would need to be aware of these risks, develop guidelines for industry collaborations, and be prepared to manage any conflicts.
CHANGES IN FEDERAL LAWS AND REGULATIONS REQUIRED BY ALTERNATIVE FUNDING SOURCES AND MECHANISMS
One mechanism for securing nonfederal funding for CDMRP to enlarge one or all of its grant programs would be through accepting voluntary contributions from foundations, companies, state governments, and other funders of medical research. As a model, the Secretary of Health and Human Services has statutory authority to accept conditional gifts for study, investigation, or research and other purposes, which has been delegated to the NIH director and the institute directors. CDC may also accept conditional gifts. Currently, according to the law which authorizes general gift funds, the recipient of contributions to the Army or DOD cannot be specified except for DOD hospitals, schools, and other health, education, and welfare activities. Other donations must be unconditional and appropriated by Congress before they can be used.
In order for CDMRP to accept nonfederal funds in support of a project or program, DOD would need statutory authorization to accept contributions from nonfederal donors for a specific purpose, for example, to fund research grants for a particular purpose. Based on NIH and CDC experience, however, the amounts of nonfederal funding would be very small compared with federal funding.
A related mechanism would be the establishment of a nonprofit foundation to solicit funds for CDMRP programs, because even if CDMRP were authorized to accept contributions, federal employees may not actively seek them to augment appropriated funds. To deal with this, some agencies, such as NIH and CDC, have foundations with staffs that seek donors for agency programs. For this to occur, Congress would need to charter a foundation similar to the ones it chartered for NIH and CDC or, possibly, expand the authority of the Henry M. Jackson Foundation for the Advancement of Military Medicine beyond its mission of supporting intramural research.
Grants and cooperative agreements for extramural research are subject to a number of conditions that may deter some commercial firms from doing business
with the federal government. The Defense Advanced Research Projects Agency and the military departments (including the Army) have special authority to enter into transactions other than contracts, grants, and cooperative agreements. Such “other transactions” are exempt from the usual controls and oversight mechanisms set forth in acquisition statutes and the Federal Acquisition Regulation and from laws applying only to contracts, grants, and cooperative agreements. Under this authority, DOD has established an assistance instrument called the Technology Investment Agreement (TIA) in which DOD partners with a company or consortium of companies which contributes half or more of the costs of the project. TIAs were created to increase the participation in defense R&D of for-profit firms that are reluctant to comply with traditional instruments whose requirements or procedures are considered too burdensome, intrusive, or costly. In a TIA, for example, DOD may negotiate less restrictive intellectual property rights than are required by the Bayh-Dole Act.
TIAs are most appropriate when CDMRP is trying to stimulate product or technology development with enough commercial promise that a firm or firms would be willing to pay half the costs. The Secretary of the Army would need to delegate the authorities to award and administer TIAs to CDMRP.
The committee determined that it would be possible to leverage nonfederal funds, but only to a limited extent for certain types of research that are not the main focus of the CDMRP program. This includes later-stage research with potential near-term applications that are likely to be commercially viable, a type of research that already is supported by major medical research funders. The possibility of obtaining co-funding from industry, the largest nonfederal funder by far, is least likely for early-stage exploratory research, which has been CDMRP’s most important contribution. It would be desirable for CDMRP to pursue the leveraging of outside funds for later-stage research, but only to the extent it does not unduly shift the program’s priorities as set by its advisory panels. Leveraging would be facilitated if Congress granted the authority and means for CDMRP to solicit and use outside funds for extramural awards. Finally, the risks inherent in public-private collaborations must be addressed and managed.
The findings and recommendations that follow are not likely to result in a significant influx of new outside funding into CDMRP, primarily because of the program’s high-risk focus on innovation, a focus that is properly the function of government, not the private sector.
Recommendation 1: Facilitate Collaboration When Appropriate
Findings. The majority of biomedical research funding comes from industry and is not readily accessible to a program such as CDMRP. Experience with cost
sharing in other federal basic research programs shows that requiring recipients to provide significant percentages of the cost of projects to augment federal funds imposes additional expenses on both the recipient and the funder. This requirement also can have unintended—and often unwanted—consequences (such as discouraging the submission of outstanding proposals from researchers at institutions with limited means), and may not substantially increase the total amount of funding in an area of research (or may redirect it from other important uses). Cost sharing is most appropriate when the co-funder receives a tangible benefit, and this is much more likely to happen in later-stage research or infrastructure projects than in basic and exploratory research. Yet CDMRP’s greatest strength has been its support of new ideas and new investigators, where cost sharing beyond the provision of facilities and equipment is least justified.
The experience of federal R&D funding agencies with voluntary public-private collaborations generally has been positive, although collaborations must be individually negotiated, which can add significant costs to a project and increase the time it takes for the research to begin. Appendix A includes a number of examples in which foundations, companies, and state governments have partnered with federal agencies in a research project or program.
Recommendation 1. CDMRP should facilitate collaborative arrangements for funding research when collaboration would be beneficial and appropriate—for example, when it would achieve greater results through synergy or economies of scale or critical mass—but CDMRP should not expect such arrangements to augment significantly overall program funding.
Opportunities for collaboration with other sponsors of biomedical R&D should be encouraged, not to stretch program funds, but rather to achieve program goals that could not be met otherwise. For example, increased funding might allow attainment of a critical mass in an area that no single funder could achieve by itself, development of a shared infrastructure, or the creation of a synergistic effect through the interactions of the different collaborators.
CDMRP should experiment with award mechanisms that facilitate collaborative R&D arrangements among academic institutions, industry, philanthropies, state governments, and/or other supporters of research, as has been done in the Breast Cancer Research Program with Collaborative-Clinical Translational Research Awards and Biotechnology Clinical Partnerships. In addition, CDMRP should, through its inclusive planning process, develop programs that outside funders would be willing to help fund (see Recommendation 2, below, for a private foundation to be set up to solicit and transfer such funds). An alternative would be for CDMRP to approach nonfederal funders to explore the possibility that they might fund projects that receive high scores but that cannot be funded by CDMRP. This reliance on the CDMRP application and review process would save the nonfederal funders administrative costs.
Recommendation 2: Provide DOD with Gift Authority
Findings. Some federal agencies, such as NIH and CDC, have authority to accept gifts for specified purposes and have foundations that can solicit private funds for their programs. Under current law and regulations, however, the Army is only allowed to accept private donations in its Army General Gift Fund for certain purposes (e.g., to benefit a school, hospital, library, museum, cemetery, or similar Army institution or organization), which do not include augmenting the funding of an extramural grant program such as CDMRP. Unlike some of these other federal research agencies, neither CDMRP nor its parent organization, the U.S. Army Medical Research and Materiel Command (USAMRMC), has the authority to receive outside funds to augment its budget for extramural awards. Even if it had such authority, it could not be used actively, as it is not legal for federal employees to seek funds from private sources.
The Henry M. Jackson Foundation for the Advancement of Military Medicine was established to be the recipient of funding for medical research and education projects from other federal and nonfederal sources, but only on behalf of the faculty of the Uniformed Services University of the Health Sciences, researchers at Walter Reed Army Institute of Research, and other intramural DOD researchers. The Jackson Foundation does not fund extramural research. Nonfederal funders also may contribute funds to the Defense Cooperation Account (DCA), but Congress must appropriate these funds and authorize their use for a specific purpose, and donors are asked not to designate the intended use of their contributions to DCA. If Congress wishes CDMRP to draw on foundations and private donors, it needs to create the authority to receive gifts to support research administered by CDMRP or in collaboration with CDMRP.
Recommendation 2. Congress should provide CDMRP with authority to:
a. receive gifts and donations from individuals, companies, foundations, and other organizations for the support of research grants and contracts awarded by CDMRP, and
b. charter a nonprofit foundation with authority to solicit and transfer nonfederal funds for the support of research grants and contracts awarded by CDMRP.
Gift authority might be granted to the Secretary of Defense, the Service Secretaries, or the Commander of USAMRMC, as long as it is delegated to the CDMRP program. A nonprofit foundation with the mission of assisting CDMRP or USAMRMC could be modeled after the Foundation for the National Institutes of Health or the CDC Foundation. Alternatively, the mission of the Henry M. Jackson Foundation could be expanded to include fundraising for CDMRP or USAMRMC extramural research programs. Based on the experience of similar foundations for other agencies, however, expectations of substantial donations to
such a foundation should be modest. Congress also would have to provide for the additional costs of establishing and maintaining the foundation.
Recommendation 3: Limit Cost-Sharing/Matching Requirements
Findings. Cost sharing and matching requirements do not always advance research goals. Usually, cost-shared funds do not expand the total funding provided for research. Rather, they involve the rechanneling of existing funding streams. In addition, cost-sharing requirements can impose additional administrative costs on both recipients and funders. However, cost sharing makes sense where the extra costs are more than offset by additional benefits resulting from the partnership, such as when research results have foreseeable commercial applications.
Recommendation 3. CDMRP should not impose cost-sharing or matching fund requirements beyond those currently required, except when a tangible benefit to the award recipient is anticipated beyond the immediate term or scope of CDMRP-supported activity (for example, funding of instruments and facilities).
Care should be taken to see that cost sharing does not divert funds from other desirable activities, such as other research projects that would have been funded by those dollars. CDMRP should not let expectations of increased nonfederal funding shift the program’s scientific priorities away from its focus on innovative exploratory research, research into disease prevention and causation, and epidemiological studies.
Recommendation 4: Issue Guidelines for Collaboration
Findings. Research on university-industry and government-industry partnerships and similar collaborations has identified a number of potential benefits and costs. The benefits generally take the form of induced private investment in developing research results into commercial goods and services, but these benefits do not necessarily require co-funding or formal matching requirements. The costs can include the imposition of secrecy on the scientific process, the occurrence of delays and bias in the reporting of research results, the shifting of research priorities toward near-term development rather than long-term research, and the possibility of financial conflicts of interest for both research institutions and individual researchers.
Recommendation 4. DOD should issue guidelines for collaboration with the private sector, paying special attention to the potential impact of research collaborations with nonfederal funders on (a) program costs; (b) the integrity of the peer review process; (c) program priorities;
(d) perceived and actual conflicts of interest; (e) openness in scientific communication, and (f) other issues that may arise in federal-private co-funding arrangements.
Other research funding agencies such as NIH and CDC have issued guidelines, which could serve as models for DOD guidelines. They focus on such issues as potential conflicts of interest that must be identified and addressed; intellectual property rights; the timely publication of research results; and the maintenance of academic freedom, and they contain suggestions of ways to avoid or manage them.