Basic-Research Training for Physician-Scientists

woman named Dr. Bola Famakin

I read with fascination Edward Korn’s letter about “Early Graduate Programs at NIH” in the September-October 2012 issue of the NIH Catalyst ( His letter rekindled my interest in advocating for a structured basic-research-training program for physicians, like myself, who have completed medical training and want to become independent investigators engaged in basic research.

There is a precedent at NIH for such programs for physicians: the Associate Training Program (ATP), created in 1953. In 1956, the NIH Scientific Advisory committee proposed the formation of a “two-year [research associate] training program for people who have their M.D. degree and intend to go into medical research as a career.” These research associates were trained in the fundamentals of basic research and research methodology. This instruction was through specialized didactic courses approved by the advisory committee and taught by experienced scientists.

The significance of the ATP program was highlighted in a 2011 study (in Academic Medicine), which showed that physician-scientists who participated in it have had more successful academic careers than physician-scientists who did not participate. It is interesting to note that nine of those physicians who trained in the ATP program were later awarded Nobel prizes for their contributions to biomedical research. Unfortunately, the ATP ended in 1992, coinciding with the national decline in the number of clinical investigators (Acad Med 86:502–508, 2011; Science 338:1033–1034, 2012).

The ATP program's remarkable success is reason enough to call for establishing a similar program at NIH for physician-scientists. Today’s avenues for physicians to obtain structured, advanced basic-research training are either too cumbersome or impractical. The typical M.D.-Ph.D. program involves continuous training for 16–20 years (college; medical school; graduate school; residency; fellowship training) and is not attractive to many physicians who are interested in pursuing basic science research. Although NIH’s Graduate Partnerships Program (GPP), in which NIH partners with participating medical and graduate schools, offers the M.D.-Ph.D. Partnership Training Program (, it is impractical for practicing physicians. The program is full-time and inflexible, and it does not allow physicians to see patients on a part-time basis so they can maintain their clinical skills. Even the more flexible GPP—the NIH-Oxford Cambridge Scholars Program—is problematic because its students must spend significant amounts of time outside the country.

I propose, therefore, that NIH offer an advanced, structured, basic-research training program that would be open to board-certified, eligible physicians who have completed residency and/or fellowship training. The elements would include:
     • A competitive four- to five year-Ph.D. research fellowship that would be offered by the NIH and jointly administered by local degree-granting universities;
     • Coursework developed with the help of Ph.D.-granting local universities to bridge the gap between medical training and bench research;
     • Participating physicians being able to see patients, part-time, to help foster the development of research questions from bench-to-bedside and from bedside-to-bench;
     • Pay for participating physicians that would be commensurate with the pay for physicians who are just entering the workforce;
     • Participating physicians being awarded Ph.D. degrees, by the partner institutions, upon successful completion of the required coursework and research thesis.

The NIH is the right organization to once again lead the development of strong, innovative, structured research-training programs to prepare the next generation of physician-scientists who will tackle complex medical questions in biomedical and translational research.

Response to Famakin Commentary

Dr. Famakin’s proposal to provide coursework and formal research training for physicians who have completed their clinical training is intended to allow active clinicians to retain their clinical skills while they obtain a Ph.D. Training for physician-researchers is certainly a timely topic as we try to improve the research-training experience for clinically trained scientists at the NIH.

Her proposal is for NIH to not confer degrees but to take advantage of existing university programs through the Graduate Partnerships Program; students would do their thesis work at NIH and basic-science coursework at a university that can confer degrees. NIH would pay. Paying students the proposed salary would be a reasonable investment because of the applicants’ clinical credentials and their participation in patient-care rotations. It is easy to imagine a program in which clinical fellows would have the opportunity to get their Ph.D.s and become board eligible in a subspecialty. This possibility would be attractive to some early-career investigators and should attract the very best to NIH.

Extramural academic centers already provide such opportunities through support provided by NIH K12 awards. A more time-efficient alternative, which would more closely mimic the highly successful ATP program, would be to provide research coursework and rigorous research training in laboratories at the NIH without the need for the formal requirements (and the four-to-five year commitment) of a Ph.D. program. Some ICs, such as NCI, already provide this type of alternative for clinical fellows.

Our recent experience suggests, however, that few fully trained M.D.s are interested in a sustained laboratory experience—whether it is structured as described by Dr. Famakin or is more flexible as I have suggested. Such a program would have to be made more attractive with competitive salaries and loan-repayment opportunities.