Guidelines for Grant Preparation
Faculty and staff within Biostatistics can participate in grant development in numerous ways including:
- assisting with the development of a research team;
- assisting the investigators in defining and refining study aims and measures;
- developing study designs;
- writing statistical analysis plans; and
- computing the power or sample sizes necessary to achieve the study aims;
- (NOTE:We do not participate in proposals in which the investigator "only needs a sample size calculation" unless we have sufficient time to participate in the study design, because we almost never find that the sample size is as important as making improvements in the fundamental study design.)
Required Timelines for Grant Preparation
The strongest grant proposals are those in which the proposal team is assembled at least four months before submission. Holding team meetings every three weeks or more often and having at least two major revisions of the proposal are optimal. When biostatisticians have sufficient time to do their best jobs, the likelihood of funding is significantly enhanced.
With rare exception we cannot provide assistance with proposals without two or more months advance notice. There are exceptions for cases involving resubmissions, renewals, or submissions of proposals that are substantially similar to previous grants in which the biostatisticians fully participated. On rare occasions we can analyze small pilot datasets to inform power or sample size calculations when the design is very simple.
Use of Biostatisticians' Names on Proposal
We do not allow an investigator to use the name of a PhD or MS biostatistician on a grant or contract proposal unless the biostatistician has fully participated in the development of the proposal and can vouch for the study design and analysis plan.
It is Department policy that percent efforts of statisticians on grant proposals accurately reflect anticipated efforts. The minimum percent effort is generally 10% for clinical studies and 5% for basic science studies for faculty biostatisticians. The staff needs are negotiated on a project basis and approved by the Director. In certain circumstances biostatisticians can participate on grant applications as a consultant or be paid a fixed dollar amount or for a fixed number of hours on a grant proposal where the granting agency disallows funding percent efforts. On the average, a grant is best served by having both a PhD and an MS biostatistician on the team, with the ratio of MS:PhD percent efforts being about 2.5:1; thus, a clinical study involving a faculty statistician at 10% will have MS-level biostatistical support from our Center and/or Department at the 25% effort level.
Some factors that necessitate higher percent efforts are:
- many variables per subject or animal;
- repeated measurements per subject or animal;
- a significant number of subjects dropping out before the planned termination of their follow-up;
- other types of missing data;
- plans for sequential monitoring or interim analysis of treatment effects or safety;
- providing reports for data and safety monitoring committees;
- receiving and merging data from other sources;
- multi-site and studies; and
- outcome and/or covariates that only partially observable, among others.
Who to Contact
Unless the proposal is a renewal or resubmission (in which case you should contact the biostatisticians currently involved in the project), all requests for assistance on proposals must go to Project Registration. To determine if the proposal development is already assigned to a member and/or staff of the Center, contact Megan Tremblay, Project Manager, Department of Biostatistics & Data Science.