The National Institutes of Health recently released a proposal to cap researchers’ “Grant Support Index”. In effect, each grant type counts for a certain number of “GSI points”, with the R01 (the flagship independent-investigator award) counting for 7 points. The full scale is described in NIH’s GSI blog post linked above, and the rationale for it in this blog post from FOR’s Gary McDowell. Each named investigator would be limited to 21 points worth of NIH support, and could not accept any new award that would put him/her over 21 points. Public comments on the proposal soon to be solicited (presumably with a short time frame, since NIH hopes to deploy the GSI this year).  The overall cap idea is up for debate, but there’s one thing that is clearly bad: this proposal will make it harder to do truly collaborative work, and for researchers who do “team science” (methods development, computational modeling, or translational research) to get fair recognition for their efforts.

Why? Because of one critical flaw in the initial proposal: the weighting of collaborative grants. The most impactful science often requires a team approach, bringing investigators across disciplines together to tackle hard problems. On those grants, multiple scientists each act as named principal investigators (PIs), in recognition of their equal intellectual roles (even if one scientist tends to take on the majority of administrative burden as “contact PI”). In the current GSI proposal, those multi-PI grants “count” towards a scientist’s GSI point total as though they were a single-PI, lab-supporting lifeline. The negative impact is illustrated in two scenarios that are not at all hypothetical; they are moderately disguised examples of my peers and collaborators:

Dr. Patel is a recently-independent physician scientist, with her first R01 as PI (7 points). This leaves her with 50% of her salary uncovered, which she could make up through clinical work. However, during her prior K23 award, she became the departmental expert in a novel data analysis technique that is rapidly becoming the standard in her field. Two senior investigators approach her asking if she would be a collaborator on their R01s, at 10% effort each. Since this would increase her lab time and allow her to deepen her scientific expertise, Dr. Patel is of course interested. Unfortunately, she now faces an unpleasant choice. On either grant, her expertise is critical. If she is listed as Key Personnel, she will receive money, but outside evaluators may not recognize her absolutely essential role in the project. Her excellence will be lost behind that of her senior collaborators, and her negotiating position for last authorship on the resulting articles will be much weaker. Or, to demonstrate that value, Dr. Patel could request to be listed as a co-PI — at a cost of 6 GSI points each. This brings her to 19 points, where she would be unable to accept any further independent funding. In essence, Dr. Patel must put the future survival of her own lab at risk in order to gain reasonable credit for an otherwise beneficial collaboration.

Dr. Jones is a biomedical engineer who recently invented a new technique for manipulating physiologic systems in laboratory rodents. He is interested in the next steps, which include (A) demonstrating the relevance of this technique in disease models, (B) verifying that it works in primates, and (C) disseminating it to other labs. He approaches investigators with complementary expertise, and they obtain collaborative R01s to perform (A) and (B), giving Dr. Jones 12 points. For most of that work, he simply provides materials, and the work is done in others’ labs with his new ideas. He funds (C) with an independent R03, for 4 more points, a total of 16. Now Dr. Jones thinks about his next idea: an imaging technique for measuring the same systems in vivo. He needs four years to develop it, i.e., another R01. Unfortunately, if he were to get that, he would be at 23 GSI points, over the cap. Like Dr. Patel, he faces an unreasonable choice: he can develop his independent program, but in doing so, must give up credit he deserves for enabling another investigator’s research.

Neither Dr. Jones nor Dr. Patel is a “mega lab” — each has one solid idea that collaborators want to use, and would like to obtain a second grant that will help them bridge and broaden their support base. They are exemplars of modern science, simultaneously helping others advance while demonstrating their own intellectual merits. And yet, each is hurt by a policy that is meant to help creative young investigators, especially those trying to move beyond their first R01.

What might be a better path? Many are suggested in the comments of the original “Open Mike” blog post, but here’s my favorite among them: for multi-PI awards, split the points based on the percentage of the money that goes to each investigator’s lab. It’s relatively simple to calculate, it achieves the goal of balancing funding fairly among labs, and weights collaborations based on how much work is needed.

NIH has not yet opened a formal comment page, and for now, the best way is to add your opinion to the NIH Deputy Director’s blog post comment section. You could echo my preferred solution, or quite possibly you have a better idea. Either way: if you are a junior investigator who focuses on collaborative science, this could hurt you, and NIH needs to know that.

Contributor:

Alik Widge is an Assistant Professor of Psychiatry at Massachusetts General Hospital and Harvard Medical School. His team develops brain stimulation methods for severe mental illness, through combinations of mathematical modeling, real-time brain signal processing, and rodent/human translational research. Since graduate school, he has advocated for trainees and young scientists. He has served as an advisor, board member, or committee leader for numerous advocacy organizations, has trained scientists nationwide in legislative advocacy, and has authored both primary-data and opinion pieces in multiple domains of science policy.

Note, all views expressed are the personal opinions of the author and don’t represent an official stance from AFS.