How to write a grant proposal

Executive summary) The abstract is an important part of a study protocol because it is the first page that a reviewer reads. Reviewers Zlowodzki M, et al.: How to write a grant proposal significance chapter is to justify the study you are proposing. Describe how the result of your study will benefit society. You need to convince the granting agencies that it is worth their money.

A cademic success and promotion in medicine largely RCTs and subsequently a larger trial. identified a good idea, you need to clearly define the problem that needs to be addressed and formulate a After you decided to perceive with your study proposal, research question. Subsequently you need to ask yourself you need to determine how many study subjects you need, if that question is already answered [ Table 1]. A thorough how much money you need and who your collaborators literature review is therefore mandatory. If you have a truly will be. In order to be successful in obtaining a grant you good idea, you might find out that you are not the first one will need convincing data, which might require several having it. You do not want to spend a lot of time and energy preliminary studies and you will need to prove to the into a project only to find out later that there have been granting agency that you are capable of performing the already 17 trials and a meta-analysis performed and your study the way you propose it. The purpose of the research research question is answered.
plan is to describe what will be done, why it is important and how the study will be conducted. It is not only important to know how much was already published on that topic, but also what the quality of the ELEMENTS OF A STUDY PROTOCOL GETTING current evidence is. Rarely in medicine does a question have a definitive answer. If you are trying to compare two interventions for a certain disease, after performing a thorough literature search, you have to ask yourself the following questions: 1) Are there already multiple case series The key elements of the study protocol are the executive summary, specific aims, background and significance, preliminary results and research design and methods [ Table 2]. The research design and the methodology used in the process of planning and conducting the project should be described in detail. Prior work relevant to the proposed project should be included. Also if a pilot study was conducted, the results should be included.

Abstract (Executive summary)
The abstract is an important part of a study protocol because it is the first page that a reviewer reads. Reviewers Zlowodzki M, et al.: How to write a grant proposal significance chapter is to justify the study you are proposing. Describe how the result of your study will benefit society. You need to convince the granting agencies that it is worth their money.

Study design
In order to answer the question you need to choose an appropriate study design. The main clinical study designs of granting agencies may make their opinion based on the are interventional studies, observational studies and abstract alone. It may be difficult to overcome a bad first diagnostic studies -some overlaps may exist [ Table 3]. impression and conversely there may be a lot to gain with Which study design is most likely to answer the research a good first impression. The purpose of the abstract is to question, which one is most feasible and which one gives describe succinctly every key element of the proposed the highest quality results? The choice of the study design project. It is good to be complete but concise.
has a significant implication on the magnitude of the required funding. Ethical considerations also need to be Specific aims taken into account e.g. in some cases a certain study design The purpose of the specific study aims is to clearly describe might not be ethical. A clear description of the eligibility what research question the investigators are trying to criteria (inclusion / exclusion) is essential. Also describe answer by conducting the study. What is the problem to how outcomes will be measured during follow-up and what be addressed? The investigators need to describe why the the follow-up schedule will be like (frequency and duration). study is needed now. In detail, the hypothesis of the study and the primary and secondary goals should be stated.

Sample size calculation
Typically, the study question should be formulated to The sample size calculation is a crucial part of the study include the following: 1) the population to be studied, 2) protocol. The required sample size has major implications the intervention, 3) any comparison group to be studied on your required funding and the size of the team. Before (if relevant) and 4) the study outcomes. The study outcomes you can calculate the sample size you need to designate should be reported as the primary (main) outcome and the primary outcome. It is advantageous to choose an any secondary outcomes.
objective, reliable and highly validated outcome in order to limit bias. Ultimately, you should choose the clinically

Background and significance
most important outcome that is feasible. The purpose of the background and significance section is to lay out the rationale for the proposed research project The sample size calculation is different depending on the and to summarize currently available data in the literature type of the outcome; if you choose a categorical that is relevant to the project. If no systematic review or dichotomous outcome parameter (e.g. nonunion rate, meta-analysis was done on the topic, you should do one. infection rate) the sample size requirements are much Describe the magnitude of the problem to be addressed.
higher than if you choose a continuous outcome like a What is the patient population you are targeting? What is score (e.g., SF-36, DASH, SMFA, pain score). 2,3 In order the incidence of the problem? Is the problem likely to to perform a sample size calculation for dichotomous increase in the future (e.g. geriatric fractures)? You need to describe the historic management of the problem and whether or not there is any consensus on the current management of the problem. Are there any uncertainties about the treatment that need to be resolved? If you hypothesize that intervention A is better than intervention B you need to designate your primary outcome parameter and have some baseline data for a sample size calculation. Depending on the project, you might want to survey surgeons for their treatment preferences. Also consider surveying patients to find out about what outcome they consider to be impor tant. There might be some disagreements between the surgeons and patients perspectives. 1 The purpose of the background and outcomes, you must have an event rate (e.g., nonunion rate) for your gold standard treatment (e.g., treatment A) and you must hypothesize by how much treatment B is going to decrease or increase that event rate. For continuous outcomes you need to have a mean value for the gold standard treatment and hypothesize a difference for the alternative treatment. Using an alpha error rate of 0.05 (=accepting the probability of a false-positive result) and a beta error rate of 0.20 (=accepting the probability of a false-negative result), which corresponds to a power of 80% is a commonly accepted standard.
You can obtain baseline numbers either from a pilot study Zlowodzki M, et al.: How to write a grant proposal factors between groups. Additionally, you can use techniques like blocking and stratification in order to avoid random imbalances in small randomized trials. If you do not allocate treatment options randomly, you should account for imbalances in prognostic factors between groups, by matching the patients to the different treatment groups based on the known prognostic factors upon enrollment in your study or if that is not possible, account for it in the data analysis. However, the only way to balance unknown prognostic factors is randomization. or reports in the literature. Ideally the "hypothesized" differences should be in the magnitude of what you Blinding is another important technique for protecting consider clinically significant. You can calculate the sample against bias. Investigators should blind whoever they can: size by hand 4 or use one multiple tools to help with the the patient, the physician (not possible in surgical trials), sample size calculations 5 [ Table 4]. Be aware that the the outcome assessor and the data analyst. Lastly it is sample size calculation is based on assumptions; calculate helpful to choose an objective outcome measure like a the best-case and the worst-case scenario.
validated functional outcome scale. If the outcome parameter is subjective (e.g., union/nonunion), you should The justification of the estimated sample size should be consider to have an adjudication committee to assess the presented as a separate section in a grant proposal.
outcome. Investigators can present estimates of sample size varying across different mean differences between groups.

CONCLUSION
Alternative approaches are to present the study power across varying sample sizes and mean differences or the Grants are critical for success in academic medicine. The estimated mean differences of the outcome parameter key to a good grant is a good idea and the ability to "sell" across varying study power. 4 your idea to the reviewers of the granting agency. In order to "sell" your idea, good background research, the

Protecting against bias
appropriate study design and a well thought out Study results can be negatively affected by multiple types methodology are imperative. It is also important to of bias, mainly selection bias and measurement bias.
recognize that research is a team effort. Convincing the Investigators need to describe proposed methods for grant reviewer of your expertise is crucial; choosing protecting against bias. The most powerful techniques experienced team members therefore improves the chances for protecting against bias are 1) randomization, to obtain the desired grant. A successful pilot study and 2) concealment of randomization, 3) blinding and 4) the preliminary studies that serve as a justification for your choice of an objective outcome measure. study proposal can prove feasibility to the grant reviewers and be therefore a persuasive factor. You should propose If you are comparing the effect of multiple interventions an appropriate budget and a realistic timeline; otherwise on a specific outcome, the best method of protecting against failure is almost certain. Lastly, you should tailor their grant selection bias is random treatment allocation. application towards the granting agency's goals and use Randomization balances known and unknown prognostic the requested format for their application as that might differ from agency to agency. Targeting multiple government and industry-funded agencies increases the chance of getting funded.