On 30 September, Elias Zerhouni, director of the US National Institutes of Health (NIH), unveiled the 'NIH Roadmap', an ambitious plan designed to overhaul the biomedical research enterprise of the NIH and accelerate medical progress. The NIH may be long overdue for a major reorganization, but the Roadmap raises questions that the biomedical community should carefully consider before its implementation.

The Roadmap is the result of an effort to identify today's major scientific challenges and the means to address them within the collective capabilities of the NIH. From this assessment, based on the urgent need for new strategies to improve research progress, three major areas were singled out: pathways to scientific discovery, research teams and clinical research. Twenty-eight initiatives addressing these three themes will be launched in FY 2004.

The focus on “new pathways of discovery” recognizes the importance of updating the technological infrastructure of the NIH to keep up with advances in medical research. Under this guise, tools and facilities for chemical screening, molecular imaging, proteomics and metabolomics will be established, and computing and bioinformatics resources improved. New funding initiatives will foster technology-driven biology. Expansion and integration of reductionist approaches is aimed at a better understanding of complex biological systems.

New technological resources would be welcome additions to the NIH's arsenal, and improved accessibility of the public sector to advanced technologies is admirable, but the focus on the process of data accumulation rather than the scientific problem is an issue of concern to researchers in the biomedical community. Increasing information does not automatically enhance understanding, as should be clear from the completion of the Human Genome Project. Hypothesis-driven research is absolutely required to derive new biological insight and direct medical advances.

Under the Roadmap's second major theme, the “research teams of the future” are visualized as interdisciplinary efforts, and new partnerships between the public and private sectors will be encouraged. How these teams will be assembled and managed and what specific biological questions they will address remain to be determined.

Is it optimistic to presume that accomplished investigators will relinquish the recognition of their independent contribution for the greater good of the team effort? Will the Roadmap's proposal to award each key member of a team primary investigator status make for more or less collaboration, particularly in the absence of an assurance of remuneration, promotion or job security? While competition may not always be efficient, scientific research nevertheless benefits from its pressures. Team efforts may run the risk of diminishing creativity and innovation in favor of compartmentalizing tasks.

One of the questions Zerhouni poses in the Roadmap is, “Does the initiative position the NIH to do something that no other entity can or will do?” (Science 302, 63–72; 2003). The private sector is currently organized to fulfill very similar functions by pooling multidisciplinary talent to work towards a common goal. Within the confines of a company, scientific interaction is largely collaborative rather than competitive, and the company has a vested interest in ensuring that its facilities are productive and efficient. If the success of the Roadmap is predicated on the numbers of drugs taken to the clinic, the NIH should perhaps consider calculating its likely achievements on the basis of productivity in the private sector as a function of money, time and people invested.

The final theme of the Roadmap is the “reengineering of the clinical research enterprise”, a task confronted by major obstacles: insufficient infrastructure, prohibitive cost of clinical trials, lack of standardization or access to trial protocols, and a profusion of regulations imposed by research institutions, medical associations and the US Department of Health and Human Services. The cumulative effect of these obstacles may dissuade translation of important research findings or limit the information derived from a clinical trial. The Roadmap proposes increasing clinical workforce training, standardizing data accumulation and sharing, and establishing a clinical trials database for 'seamless' exchange of information. The database alone will be a tremendous boon to clinical researchers, although the cost of its creation and maintenance may prove substantial.

Implementation of the Roadmap's aims will cost $130 million in FY 2004, and $2.1 billion over the next five years. The financial burden will be borne by the budgets of the 27 institutes and centers of the NIH, rather than by new sources of funding. Proponents of the Roadmap assert there will be no adverse effect on extramural funding. But in view of the enormous infrastructure changes required to fulfill the Roadmap's goals, the cost of data management, the administration of new grants and the lack of discrete endpoints for completion of each initiative, it seems likely that the NIH budget for RO1 grants will ultimately be affected.

The efficiency of the NIH has recently been a major consideration for the US Congress and may dictate future budget decisions. At issue is the lack of a clear system for measuring the success of NIH initiatives. Coupled with a seeming lack of emphasis on disease-related questions, the NIH Roadmap may not mitigate the concerns of Congress. But the Roadmap should perhaps be viewed in the same vein as the science it champions: a work in progress that contributions from the biomedical research community can only improve.