Sir

Frederick Sachs argues that the number of US National Institutes of Health (NIH) grants should be limited to two per investigator in order to fund additional young investigators (Nature 388, 222; 1997). Although there may be merits to limiting further funding to investigators with a large number of grants or resources, the suggested limit would be too restrictive.

Such sums would support only a relatively small laboratory. Clearly the best qualified senior scientists can and should be able to support significantly larger and more productive research laboratories. Unfortunately, Sachs fails to provide proof of his assertion that shifting funds from larger laboratories to a greater number of smaller laboratories would provide more major discoveries.

His discussion on Nobel prizes is irrelevant. Very few scientists will receive such an accolade and often it will be for a body of work spanning their career rather than a single discovery. This type of award system cannot recognize the excellent and critical research performed by thousands of other researchers. Indeed, major discoveries are usually built on a framework of preceding supportive work that comprises most of the funding from NIH and other agencies and which accounts for the majority of scientific literature.

The criterion of performance per existing research dollar by an investigator should be established and used to guide future funding decisions. The research dollars would include all funding available to an investigator and his or her laboratory, including research grants, endowments, awards to graduate and postdoctoral students, industrial support and donations.

Of course, any judgement of academic performance (number and impact factor of publications, for example) does include an arbitrary component, but the suggested formula is more quantifiable than existing procedures. For example, a junior investigator with one grant totalling $100,000 and one article in Naturehas performed 2.5 times better than a senior investigator with grants totalling $1 million and four papers in this journal, and should thus receive priority for an increase in funding over the senior investigator. The application by granting agencies of this formula for high-risk grants could also be useful in an era when ‘safe science’ grants often score better than they necessarily should.