Design Your Translational Research Project With Patients In Mind

When designing a translational research project, be sure to target the end-user or end-market right from the first basic science experiments. This suggestion seems obvious. Yet time and again I see draft proposals that don’t accomplish this task, or pink sheets that complain a project misses the mark. Here are examples of how to achieve this goal from proposals I have worked on over the past few months:

*Enlist a physician as a consultant. Not a physician-scientist, but an M.D. who regularly treats the patient population you hope will benefit from your work. I recently worked on a polymer science grant that sought to create better cardiac stents—the Program Officer wisely suggested that a physician be added as consultant, given that the PIs were bench scientists. If you already have a physician on the proposal, make sure s/he emphasizes in the personal statement of the biosketch that s/he regularly sees the intended audience for the work being performed. Which leads me to the next point:

*Make sure you know what the patients want and need. According to one of my clients, researchers once thought that what quadriplegics wanted most was the ability to walk, therefore lots of money was devoted to that subject area. But when these patients are surveyed, it turns out that what they miss the most is the use of their hands. Before you design a bench project whose ultimate goal is to ease human suffering, make sure you know from what the patient suffers. If you can’t easily ascertain the main quality of life issues for that patient population, at least determine national healthcare costs that would be alleviated if your research findings informed clinical approaches.

*Keep the cost down. You could design the niftiest gizmo in the world that accomplishes precisely what you (and the patient population) want it to do. But if it is cost-prohibitive to produce, it will never see the light of day. Assure reviewers right from the design phase of your project that cost will be at the forefront of your mind as you proceed through your research. Whenever possible, point out where your device could interface with existing technology (ex- Bluetooth, cell phones, PCs). I work on a lot of NIH proposals where folks are trying to develop some sort of device, anything from activity monitors to robotic surgery gadgets. If you don’t confront the issue of cost from the very first design (yes, even in an R21) then you are almost certain to see that comment on your pink sheets.

*Bring on board a collaborator who has experience with regulatory approval. If you hope to move toward marketing a product, assure reviewers that someone on your team has some clue how to navigate the labyrinthine FDA approval process. If you have no such person, find one.

*Your goal is to help patients, not market a product. Remember that when writing an NIH proposal, you are asking for taxpayer dollars to fund your research, and therefore your work ought to benefit taxpayers. I am working on a diabetes grant for a for-profit company right now, and one suggestion I made was to state that the end goal is to help patients, not market their product. Perhaps it seems like a subtle language shift, but it will play better at review.

I could probably come up with other suggestions but I have finished my coffee and huevos rancheros and need to get back to work. So here’s the take-home message: Assure reviewers that from the very first experiment, your translational research project is aimed directly at the clinical end-users. More-than-the-usual concern prevails on the NIH campus right now that promising bench science is lying fallow, never being developed to the point where it will benefit patients (who, after all, paid for the research through their tax dollars). If you are on the bench side of the bench-to-bedside equation, make sure reviewers know that you are keenly aware of the clinical target and that you are not just a smart person in a white lab coat incapable of thinking beyond your cool idea.

%d bloggers like this: