Expectations of Biostatistical Collaboration and Support
Collaboration with a biostatistician typically provides high-quality, reliable results that stand up to critical peer review. We identify and address issues that are not always obvious to investigators. Our goal is not only to ensure statistical rigor and appropriateness but also to make sure that the results easily understood by the investigator.
Prior to the start of the project
It is critical to involve our biostatisticians at the very beginning of your project. Issues relating to sample size and power, design efficiencies, nuisance variables, randomization, blinding, intervention strategies, time lines, data collection and management should be addressed in the planning stage.
During this initial meeting the background and scope of the project are queried. It is also very important that the type and level of statistical support be established: through a paid grant, an internal contract (MOU) or external contract, or direct hourly billing. Usually, there is not enough time during this initial meeting to work out all of the project details, but enough information should be presented so that a reasonable level of support can be estimated and the appropriate team of biostatisticians can be identified.
Please contact us if you are interested in setting up an initial meeting.
Expectations and responsibilities of the investigator
Often, investigators will have very clear ideas about the analysis of their project. However, we ask that all investigators coming for biostatistical support be willing to listen to alternatives. We may suggest more appropriate statistical designs to address the stated hypotheses. Collaboration is at its highest when there is good communication between the biostatistician and the investigator. The biostatistician needs an accurate description of the project and how the data was generated, addressing such problems as data management and collection, missing values, protocol deviations and endpoint definitions. For more details about our authorship policy, please see the “Manuscript and Abstract Preparation Policy” section below.
Any manuscripts or reports generated from this project should acknowledge the participation of the biostatistician as a co-author if they have made a significant contribution to the project. However, the biostatistician should not be associated with the research without their consent.
Role of the biostatistician after the data are collected
Although it is preferable to have our biostatisticians involved in projects from the beginning, many times the investigator will contact the biostatistician after the data have been collected. In these cases, we will suggest statistical methods based on how the project was designed and how the data were collected. Often, a more appropriate method of analysis will be recommended if the investigator has already run preliminary analyses. Our biostatisticians can also recommend methods that will be able to extract useful information from problematic data. We also analyze data to provide information on future efficient and less biased designs.
General Business Practices
The Center for Biostatistics is self-sustaining. Researchers, investigators and clients are provided statistical support when we are supported through grants as FTE collaborators, written into departmental contracts (MOUs), written into external contracts, or billed directly for hourly projects.
Grant Support: The majority of our funding comes from collaborative work on NIH grants. We have extensive experience in preparing P01s, R01s, and R21s.
Department Contracts (MOUs): Over the past few years, we have substantially increased our contracts with other departments (MOUs), promoting long-term collaborative relationships between center biostatisticians and investigators. MOUs increase efficiency through familiarity with special issues within a discipline, resulting in more responsive and comprehensive service to collaborators. Departments can provide salary support for a team of biostatisticians, and adjustments are made annually for increases or decreases in effort.
Hourly Projects: For small projects, hourly billing is available with the rate dependent on the level of statistical expertise required. Please contact us for current rate information.
External Contracts: Support is provided for external institutions or industry on a case-by-case contractual basis.
Grant Preparation Policy
Through collaboration with the Center for Biostatistics in the development of grants and protocols, investigators quickly learn to value the expertise provided by our biostatisticians. In general, the likelihood of funding is improved when a biostatistician has participated in the development of the proposal from the beginning. We do not allow a biostatistician’s name to be placed on a grant unless they have been involved in its development or reviewed the grant application.
Grant Support Provided
- Collaborating with the project investigators in the formulation of hypotheses and the design of experiments and clinical trials;
- Providing assistance in developing and implementing data collection tools and data management systems to allow investigators to effectively and efficiently manage and analyze their data;
- Conducting and directing the statistical analysis of data generated by the project investigators including both descriptive summary statistics and inferential methods;
- Coordinating the development of new statistical methodologies, when needed, to directly support research issues that may arise; and
- Providing clear presentations and discussions of results for publication.
Timing of Grant Support
Biostatisticians that are involved from the beginning need sufficient time to develop a strong grant for submission. The amount of time needed depends on the size of the grant and the agreed upon effort of the biostatistician:
- Small grants typically require at least one month of collaboration prior to submission.
- Large grants such as PPGs require at least three months of intense collaboration before the submission guideline.
Requests made closer in time to the deadline will be addressed only after work has been completed on grants that were submitted by investigators in a timely manner. Note that the earlier the support request, the more thorough and effective the level of biostatistical support will be, which in turn will increase the ability for your grant to impress critical reviewers.
Funding should be matched to the size, scope and complexity of the study design, data, and analysis. In general, we recommend 10% or more total effort per biostatistician per time period on a project. A 5% annual effort on a grant (2 hours/week) is insufficient to provide regular analytic support on a project and should only be used if the needs and expectations for a biostatistician are minimal (and this arrangement is agreeable to the statistician). We reserve the right to refuse to participate on a project if we feel the funding is insufficient to ensure proper and adequate analysis of the study data.
For large and/or complex projects (e.g. Program Project Awards or multi-center studies), typical total biostatistics FTE will be between 50–100%+. This frequently consists of multiple PhD and MS biostatisticians providing support through a biostatistical core. For individual R-level awards, typical total biostatistics FTE will range between 10%-50% for 1-2 biostatisticians, depending on the complexity and needs of the project.
Manuscript Preparation Policy
Sufficient time is needed for the biostatistician to write or review the statistical methods and the results section of a manuscript or abstract. Whether the biostatistician writes or reviews these sections in a manuscript is determined by the level of collaboration. Thus, communication between the project investigator and the biostatistician is essential in order to meet specific deadlines. Generally, two weeks are required for the biostatistician to write their section of the manuscript, however, four weeks may be needed for more complex analyses. Regardless, final versions of manuscripts and abstracts must be reviewed and approved before submission.
Occasionally there is concern about whether or not the biostatistician should be a co-author on the manuscript. It is our policy that the biostatistician (both PhD and MS) should be listed as a co-author if there has been a scientific or an intellectual contribution to the research and not be listed in the acknowledgments. The decision of authorship should not be influenced by payment for biostatistician time and effort. As stated by Perkins et al. (2016), “Financial considerations must be kept separate from authorship.” If this is going to be an issue with the project investigator, then an agreement with the biostatistician on authorship of any resulting manuscripts should be decided during the initial phases of the project.
The following recommendations for authorship come from the guidelines published by the International Committee of Medical Journal Editors (ICMJE):
“The ICMJE recommends that authorship be based on the following 4 criteria:
- Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND
- Drafting the work or revising it critically for important intellectual content; AND
- Final approval of the version to be published; AND
- Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.”