Functions Provided by Biostatistician Collaborators

When other departments, divisions, and centers take part in the biostatistics collaboration plan, they receive a number of benefits to further their research and educational missions:

• Priority work on grant proposals by PhD and MS biostatisticians
• Data analysis for existing non-grant-funded projects
• Study and experimental design for non-grant-funded projects
• Assistance with non-grant-funded manuscripts
• Assistance with journal clubs (methodologic review and in some cases co-leadership)
• Assistance with research conference (e.g., data analysis and pre-conference critique of residents, fellows, and junior faculty members' presentations)
• Teaching short courses in study design and analysis methodology for the group
• K award mentoring that is not funded directly by the grant award

There are many benefits of working with biostatisticians related to study design and data analysis as well as important benefits relating to report writing, data management, and implementing good reproducible research practices.

The ideal process for developing grant proposals is to assemble a research team more than three months before the planned submission date. The biostatisticians on the team will be of most benefit when they are involved before the study or experiment is designed. We have found that having the research team meet every two weeks or so results in the best quality proposal. Planning ahead in this way allows time for two major revisions to the proposal, significantly increasing the likelihood of funding.

In many cases communication among research team members is enhanced when one of the biostatisticians creates a web page to organize all documents, meeting announcements, and e-mail lists related to a proposal or a manuscript. Our collaborative web site http://biostat.mc.vanderbilt.edu has many tools for maintaining and sharing master copies of manuscripts and proposals and for tracking changes to the document.

Situation where the collaboration plan is particularly appropriate to use

There are many avenues to obtain biostatistics support at Vanderbilt including percent effort on grants, biostatistics clinics, VICTR vouchers, and collaboration plans. The collaboration plan is an ideal approach under many circumstances. Some of the major circumstances include

• When the collaboration plan biostatisticians already have detailed understanding of the research questions or underlying science of your analysis. This avoids having to reexplain a lot of details to another biostatistician
  • Similarly, if you plan to have multiple projects in the same research area, it will also be more efficient to work with the same biostatisticians over time. Collaboration plan biostatisticians often work with the same division or department for many years. However, long term research plans should including funding for biostatistics that is outside of the collaboration plan.
• Your project will lead to multiple papers or you have already used your VICTR resources
• Your project is small (e.g. < 10 hours) and you do not want to use your limited VICTR resources on such a project
• Your project will lead to a future grant application. A major goal of the collaboration plan is to improve grants, and this helps the biostatistician who will be involved on the grant get involved early. VICTR supports grant development but when using VICTR for this it is not possible to list a statistician's name on the grant proposal.
• Projects that are otherwise not able to fund biostatistics time. Projects stemming from K-awards commonly have this issue.