NIH Promises to Improve Biomedical Research Training

Sally Rockey, NIH deputy director for extramural research


Reacting to a steep rise in the number of young biomedical scientists seeking scarce academic jobs, the National Institutes of Health (NIH) plans to launch programs to prepare scientists for nonacademic careers, move students through their Ph.D.s faster, and bolster the pay of postdocs.

NIH officials announced these plans yesterday at a meeting of NIH's Advisory Committee to the Director (ACD). It is a response to a report in June on the U.S. biomedical workforce co-chaired by Princeton University President Shirley Tilghman, who believes NIH is training too many young scientists in a system that she calls "dysfunctional." The report itself did not recommend imposing any ceiling on graduate training, however, and report co-chair Sally Rockey, NIH deputy director for extramural research, said yesterday that "we did not say we should reduce the numbers" of scientists being trained because that would require a modeling study. But NIH agrees that training needs to be improved, she noted.

Toward that end, Rockey said NIH is planning to take the following steps:

  • Launch a grants program for institutions to develop "innovative approaches" to complement traditional research training, for example, by preparing young scientists for careers in industry or science policy. NIH expects to fund up to 50 grants over the next 2 years in what Rockey called the "centerpiece" of NIH's efforts.
  • Require that every graduate student and postdoc supported by NIH work with their adviser on setting career goals through what is known as an Individual Development Plan. "We have to change the culture" so that investigators "are invested in the positive outcomes of their trainees," Rockey said.
  • "Encourage" institutions to limit NIH support for graduate students to 5 years, with exceptions for family leave or if a mentor dies.
  • Raise the starting postdoc stipend from $39,000 to $42,000 as soon as next year, and examine whether NIH needs a policy to improve postdoc benefits.
  • Increase funding for two types of grants aimed at giving young investigators independent labs quickly: "kangaroo" (K99/R00) grants that combine postdoctoral support with a research grant; and Early Independence Awards, which allow newly minted Ph.D.s to skip the postdoc and start their own lab.
  • Create a system to track the careers of all trainees, not just those on training grants. Graduate students will receive a unique NIH identification number now assigned only to investigators and postdocs. NIH will also "encourage" institutions to post online reports on the career outcomes of their trainees.

The one recommendation from June that NIH is not following is to shift some graduate and postdocs from research grants to training and fellowship grants, which offer a higher-quality experience. That would have been "extraordinarily complicated to do," Rockey said, partly because training and research are separate lines in NIH's budget and also because almost seven times as many postdocs are on research grants. However, Rockey says the steps that NIH is taking will accomplish the same goals.

Tilghman, who chaired a 1998 report that said the United States is training too many biomedical scientists, praised NIH's response to the June report: "This is the most attention manpower issues have received at NIH in my career," she told ScienceInsider. But she also told the advisory committee about her "cynicism" that institutions will fail to post the career outcomes of trainees on the Web and limit NIH support to 5 years because NIH is not requiring those steps. "Without the financial pressure to get people through in an expeditious length of time, I think we will be looking at data that looks just like this 10 years from now," she told the ACD panel.

NIH Director Francis Collins acknowledged Tilghman's concern that the plan lacks "real teeth." He and Rockey suggested that it may not be legally possible for the agency to force institutions to track down past graduate students. But it may be possible to make that a requirement for new trainees, he and Rockey said.