|
By Adam Landsman, DPM, PhD, FACFAS
Good grant proposals begin with good research questions. Although this may sound trivial, many investigators waste time and energy on developing a project that is poorly conceptualized from the beginning. Good research questions have the following characteristics:
- The topic is interesting and probably plays a significant role in a much broader project.
- The question can be answered. Many questions cannot be answered in a reasonable period of time and cannot be performed on a small budget.
- The question is unique, and has not been tested time and again.
Once the question has been identified, go to the literature to determine if the question has already been answered. The investigator should evaluate the literature to locate the pitfalls, and assess how the existing literature can be advanced with the work proposed. The literature review also plays a critical role in the development of the protocol, so spend time in this area, and try to become an expert on the topic.
The hypothesis stems from the research question and should be written in a way that will naturally lead to an experimental design. For example, if your hypothesis states that surgical procedure A is "better" than surgical procedure B, you must define what "better" means. Does “better” mean faster healing time, less pain, fewer complications, lower cost, lower infection rate, or cosmetic appearance? The hypothesis should state that in which you are specifically interested. For example, if we were to have a hypothesis stating procedure A is more stable under physiologic loads than procedure B, we know what we are going to measure. Once you have written the hypothesis, you are ready to write the proposal.
When I write a grant proposal, I start with the introduction, which should review the pertinent literature. It is designed to set the stage for the work being proposed and should not contain a large amount of unrelated details. Ideally, there will be some pilot study data to go along with the introduction. The data are critical for explaining the motivation for the study and/or for illustrating that the proposed methodology is valid. The pilot study plays such a critical role in the development of the proposal that some agencies will even reject a proposal just because it lacks pilot data. Although not a formal policy, the National Institutes of Health will typically consider a proposal without pilot data unacceptable (un-fundable).
The introduction should naturally lead up to the hypothesis. I like to have the hypothesis stated formally so there is no ambiguity about the nature and goals of the proposal. A null hypothesis is usually preferred from a statistical point of view (i.e., there is no difference between the stability of procedures A and B, rather than procedure A is more stable than B).
After the hypothesis, I give a highly detailed description of the methodology. Greater detail lets the reviewer know that this project can actually be done. If you have used this methodology in previous studies and have published the results, it is usually beneficial to list that reference here. In cases where several methods can be used, you should give some justification for the method selected (and also justification for not selecting other widely accepted methods). Where appropriate, consider using line drawings or photographs to clearly illustrate the method to be employed. If you use questionnaires for collecting clinical data, you may want to include a draft copy of such a document in an appendix.
The methodology section should also include some discussion of the techniques you plan to use for analyzing the data. Talk about sample sizes, the "power" of the study, and any other pertinent information. It is not sufficient to simply say a statistician will be consulted, a common and serious error. What would happen, for example, if a statistician was consulted after the project was approved and recommended 500 samples, rather than 20? The costs of running a study with 20 samples would be less than a study requiring 500 samples, resulting in an under-funded, potentially useless study. Furthermore, justification for sample size is also a budgetary issue. You must establish that you have asked for the correct amount of money (not too much or too little).
A discussion section usually follows the methodology and gives the reviewer some idea about the potential value of the study. It may also discuss potential pitfalls in the work and predictions about studies following this one. This is a good place to address any controversial issues related to the methods or type of experiments.
Other items typically included in a grant proposal are time lines, budget, references, and appendices. The time line is your best estimate of the time required to assemble any experimental apparatus or data collection tools, collection of the data, analysis of the data, and preparation of a manuscript for publication. Be realistic in your expectations here, and be sure to keep within the time frame mandated by the grant program. If your grant proposal is time consuming and/or large in complexity, consider breaking it into two or more smaller grants.
Budgets are always a touchy item and usually quite restrictive. You may have a great idea, with well-planned methods, but you still need to keep in mind several important considerations. If you ask for items the grant program specifically excludes, such as overhead or administrative assistance, you can be assured the proposal will be rejected. In my experience, granting agencies will never come back to you and offer approval of the grant if you delete one item or another; they will just reject the proposal, with a letter of explanation. There is usually no gray area here; it either follows their rules or it does not.
Some level of budget padding may also take place, in anticipation of cost overruns or supply price escalation. Be reasonable: if you appear to abuse the system, it will surely work against you. In cases where salary offset is offered, there typically are hourly wage limits. If you are a highly successful surgeon who can earn $1,000/hour, you should not expect to be reimbursed at the same level when doing research.
It is a great idea to include matching funds in your grant, when available. This indicates to the reviewer that they will get a lot more bang for their buck, and it shows that a second party has already shown confidence in your proposal. However, if you require more funding than a particular program allows and do not indicate how these other funds will be secured, it will reflect very badly on your proposal.
Include a list of references and a copy of your curriculum vitae in the proposal appendix. There may also be other forms required for each individual grant program, which you can also place in the appendix.
Finally, remember that most grant proposals are not funded. For private organizations, approximately one out of four is funded, and at the federal level, about one out of ten. However, you will probably find that your success rate increases with practice. Before you submit a grant, ask a third party to review it and be sure that your proposal does all of the following:
- It describes the work you want to do.
- It shows that the work is significant.
- It shows that the work can be done.
- It conveys a sense that the investigator is prepared.
- It represents you and your institution in an ethical and professional manner.
Source: ACFAS Bulletin, May/June 2000
|