StATS: I want to write a grant (created 2006-04-25).

I have been mulling over the idea of writing a research grant where I am the primary investigator. I have helped lots of other people write grants, but have never before taken the step of writing a grant myself.

I have a rough idea of the form that this grant would take, but I want to use this weblog to flesh out these ideas and articulate them more clearly. I suspect that most people develop their research ideas in secret out of fear that they may be "scooped" by a rival lab. But the ideas I am proposing are ones that I would just as soon see others adopt. If you are interested in collaborating with me, all the better, but feel free to take any ideas I propose and develop them yourselves if you prefer.

There is a crisis among Institutional Review Boards in that they are swamped with adverse event reports and they do not have the tools to place these reports in context or to act rationally on them. I believe that some simple statistical tools, such as control charts can be adapted to handle adverse event reporting. I would also add drill down capabilities to allow researchers to target specific subpopulations or study centers that might warrant concern.

The control charts I have in mind have a nice interpretation with respect to a concept recently advanced in the Evidence Based Medicine literature, the Number Needed to Harm (NNH). By monitoring the timing of events relative to accrual of patients, a chart can provide continual monitoring and updating of the NNH.

The software implementation of these tools could take many forms but my preference is to develop a library using an open source program like R. These tools would be available with a web interface or could be downloaded directly to the users computer for further customization.

 I have a few weblog entries that have touched on this issue of control charts and/or adverse event reporting:

I want to also allude to the possible application of these control charts to bioterrorism. This link is a bit more tenuous, but the basic concept is that everyone is developing extremely sophisticated data mining tools for detecting outbreaks that might be indicative of a bioterrorism attack. While these models are very valuable, a strong limitation is that they can only be used by trained experts. In contrast, a statistical control chart is easily understood and applied by novices and allows individual hospitals and other health care sites to monitor trends and provide an extra set of eyes for early warning of problems.

Part of me wants to keep the grant focused on IRBs and adverse event reporting. No one seems to be doing much in this area and there is a crying need for good tools.  In contrast, there are a lot of people competing for research money in bioterrorism. I suspect that the pool of money available in bioterrorism is quite large. Is the size of the pool large enough to offset my competitive disadvantage? It's hard to say this early in the process. Perhaps it might be best to vaguely allude to the possible applications of this work on tracking adverse events to other areas like bioterrorism.

When I get the chance, I will write more about this.

An important consideration is where should I submit a grant like this. I am leaning towards a small NIH grant like an R21 to support the initial work. A description of the R21 grant is at the NIH website:

The NIH describes the type of support in an R21 grant as follows:

The R21 mechanism is intended to encourage new, exploratory and developmental research projects by providing support for the early stages of their development. For example, such projects could assess the feasibility of a novel area of investigation or a new experimental system that has the potential to enhance health-related research. These studies may involve considerable risk but may lead to a breakthrough in a particular area, or to the development of novel techniques, agents, methodologies, models or applications that could have major impact on a field of biomedical, behavioral, or clinical research.

Applications for R21 awards should describe projects distinct from those supported through the traditional R01 mechanism. For example, long-term projects, or projects designed to increase knowledge in a well-established area will not be considered for R21 awards. Applications submitted under this mechanism should be exploratory and novel. These studies should break new ground or extend previous discoveries toward new directions or applications. Projects of limited cost or scope that use widely accepted approaches and methods are better suited for the R03 small grant mechanism (see R03 announcement citation).

Another possibility is the K25 grant, a mentored quantitative research development award, although I am not leaning in this direction. A description of the K25 award is at

The K25 grant is intended to

attract to NIH-relevant research those investigators whose quantitative science and engineering research has thus far not been focused primarily on questions of health and disease. The K25 award will provide support and “protected time” for a period of supervised study and research for productive professionals with quantitative (e.g., mathematics, statistics, economics, computer science, imaging science, informatics, physics, chemistry) and engineering backgrounds to integrate their expertise with NIH-relevant research.

The NIH has a program announcement (PAR-06-223) that looks like a possible fit If I propose an R21 grant. This announcement proposes collaboration with the National Centers for Biomedical Computing (NCBC). The program announcement points out that

The NIH NCBCs are devoted to all facets of biomedical computing, from basic research in computational science to providing the tools and resources (hardware, software, and staff) that biomedical and behavioral researchers need to do their work.

and it is the latter goal (providing tools and resources) that I am working on. Perhaps they are looking for something more computationally intensive than a control chart, though.

The FDA recently published some guidance on their perceptions about priorities for future research. I lost track of that report (it was somewhere on the web in PDF format), but here is a related report

that discusses the challenges on the path to development of new medical drugs. Another important document that I should review is the NIH Roadmap for Medical Research. Showing that my grant is consistent with one or more major goals of NIH should help my grant become more competitive.

Another important page on the NIH website is the Resources for New Investigators. I just finished a four hour class on grant writing, and I will post some of my notes from that class on this weblog when I have time.

This page was written by Steve Simon while working at Children's Mercy Hospital. Although I do not hold the copyright for this material, I am reproducing it here as a service, as it is no longer available on the Children's Mercy Hospital website. Need more information? I have a page with general help resources. You can also browse for pages similar to this one at Category: Adverse events in clinical trials.