Services
1.1 Free Services
Service is free for all Fred Hutch/University of Washington/Seattle Children's Cancer Consortium investigators for consultation on unfunded activities including advice on statistical support for grant applications (including power and sample size calculations), development of a Statistical Analysis Plans (SAPs) for a publication of study results, or short (1-2 hours) meetings to obtain advice on the use of statistical methods or interpretation of results. For support in preparing a grant application, the relevant criteria for receiving support from the BSR are that if sufficiently complex analysis are proposed as part of the proposal the grant will include FTE in the budget to support a biostatistician(s). The amount of FTE will be determined based on discussions between a member of the BSR and the investigator(s) using the Biostatistics Effort Guidelines (see Appendix). Note that if support for a statistician is not included on a grant application or other study plan, only very limited free support will be available as described above and subject to availability of staff supported by other funded projects. The use of the service is on a first come/first serve basis with priority given to junior investigators who have not yet received an R-level NIH grant (or equivalent from another institution). In addition, Fred Hutch/University of Washington/Seattle Children's Cancer Consortium investigators can contact the BSR for short term consultations on analyses they plan to conduct themselves, interpretation of results or database development. The support provided will generally consist of a few hours of a BSR member’s time.
1.2 Services Requiring Funding
When Fred Hutch/University of Washington/Seattle Children's Cancer Consortium investigators require more than a few hours of time from BSR members they should consult the Biostatistics Effort Guidelines (See Appendix) to determine how much support will be required to meet their needs. In general, a discussion between a BSR member and the investigator(s) who require statistical support will be held to determine the appropriate level of support. Funds can either be currently available with a project ready to go or requested in a grant budget for future support. Whenever possible, BSR members who receive funding from a grant will continue to support the project after the funding ends. The amount of time available for this will depend on the other BSR commitments but in general, approximately one month is our goal for wrapping up a project after funding ends. Requested revisions to manuscripts or similar activities will be supported after grant funding ends. In most cases, our masters level staff handles the requests with supervision as needed from a BSR faculty member. More complex requests are handled by BSR faculty members, or in some cases, our colleagues in the Biostatistics Department at the University of Washington (UW) or the Biostatistics Program at the Fred Hutch when complex, specialized approaches are required.
How to Contact the BSR for Support
Investigators are encouraged to use the Request for Consultation link on the BSR web page. Complete as much of the request as possible. This document can also serve as the beginning of a Statistical Analysis Plan.
If a collaboration with a Fred Hutch statistician is already ongoing you can discuss a request for additional services or services for a new project with the faculty statistician working on your projects. If the faculty statistician is not a member of the BSR they will contact the BSR directors with the request to be sure resources are available.
SRAs should not be contacted directly regarding additional work or new projects. Their supervisors will work with them to determine if they are available for additional work and/or a new project.
With any of these approaches, the BSR strives to contact the investigator who submitted the request within a week after receiving a request for consultation.
Biostatistics Support for Grant Submissions
Faculty BSR members are often asked for letters of support or Bio-sketches to be submitted with grant applications. We are happy to do this but require that when a letter of support or Bio-sketch is provided a minimum of 5 percent FTE must be requested as part of the grant for at least one BSR staff and/or faculty member. The amount of support will be discussed by a BSR faculty member, and the Investigator submitting the grant and will be based on the Biostatistics Effort Guidelines (See Appendix).
Questions and Answers About the Biostatistics Shared Resource
Any FH Cancer Consortium investigator can request support from the BSR. For free service junior investigators always have priority, although we are nearly always able to accommodate senior investigators when needed. Service is provided on a “first come, first served” basis.
Possibly. If you used the free service to develop the grant application, you may also need to write a BSR member into the budget for the grant if the proposed analyses are complex in nature. Whether support for a BSR member is required and the level of support will be determined through discussions with a faculty member in the BSR and based on the Biostatistics Effort Guidelines (See Appendix). In addition, see Sections 1.1 and 1.2 description of services above.
No. Investigators, students, and post-docs often conduct analyses proposed in a grant. However, when a BSR faculty member provides a letter of support or a Bio-sketch, it suggests that the analysis or other statistical work proposed in the grant requires statistical expertise. Thus, it is expected that a member of the BSR will receive support from the grant. If investigators, students, or post-docs are conducting most of the analysis the amount of effort required for BSR members will be reduced although the minimum of 5 percent FTE is required. See Biostatistics Effort Guideline (Appendix). Masters level statisticians do not provide Bio-sketches or Letters of Support.
Yes. Any Fred Hutch/University of Washington/Seattle Children's Cancer Consortium member or their post-docs and students can request support from the BSR.
Yes, but only if the amount of support requested is of the type described under free services in Section 1.1. The free service is specifically designed to support very short term consultations and unfunded research.
Statisticians should be involved early in the grant submission process and should be included in both the development of the study design, analyses, and sample size determination as well as in budgetary decisions about biostatistician FTE, computer and software purchases, and training and travel. The following guidelines are intended as a starting point for budget discussions with an investigator and a faculty member from the BSR as well as biostatistician-track researchers who may desire support from a Masters level statisticians. These percentages should be increased for junior biostatisticians and for graduate students. Decisions about the amount of funding required for statistical support will be finalized in discussions by the Biostatistics Shared Resource member assigned to a request and the investigator making the request. The following are intended as a starting point for these discussions.
Definitions:
Biostatistician I SRA I or II, typically with less than 1 year of experience
Biostatistician II SRA III - V, typically with 1-5 years of experience
Biostatistician III SRA IV-V with 5 or more years of experience or a faculty Biostatistician.
Examples of Types of Support Provided at Different Funding Levels
Extremely Limited
5 Percent: Biostatistician II/III.
In general, not appropriate for Biostatistician I
- Consult with PI about choice of statistical methods to use, but not enough time to carry out analyses
- Supervise a level I/II biostatistician
- Meeting attendance will be infrequent and restricted to discussion of statistical issues.
This percentage is often too low to be useful. An important exception occurs when Biostatistician IIII supervises a Biostatistician I or II. This FTE level is too low to guarantee regular meeting attendance, depending on the frequency of meetings and/or one-to-one discussions.
Limited
10 Percent Biostatistician II/III and 15 Percent for Biostatistician I
- Involvement through consultation with the PI about choice of statistical, limited standard statistical analysis and/or database development and maintenance. Advice on study design including power and sample size calculations.
- Preparation of regular reports
- Co-author on papers
This percentage is realistic in many cases where a study investigator or colleague is planning to conduct the primary study analysis with advice from the Biostatistician but is too low if the statistical analysis is at all complicated. Meeting attendance may be infrequent and restricted to discussion of statistical issues.
Standard*
20 Percent Biostatistician II/III, 25-40 Percent
- Routine study design and analysis, e.g analysis carried out using standard statistical techniques
- Involvement in study design (including power and sample size calculation). Study implementation including collaboration on the development of Case Report Forms, advise on data collection including collaboration with database experts on database development including development of a data dictionary.
- Maintenance of an already prepared database.
- Active participation in manuscript publications including opportunities for first authored papers.
*This level is standard for complete standard statistical support.
Extensive
25 Percent or more Biostatistician II/III, 40 Percent or more for Biostatistician I with supervision from Biostatistician III
- High involvement in the development and implementation of the research project, which may take many forms, including:
- development of and/or use and interpretation of new statistical methods
- development and implementation of complex study design
- analysis and coordination of multi-center projects
- development, maintenance, QC and development of a data dictionary of an analysis database.
- Active participation in publications, with opportunity for first authored papers
- Project-related travel and external presentations.
- Not appropriate for Biostatistician I without oversight by a Biostatistician III.
Additional Considerations
- In a multi-disciplinary research center, investigator-biostatistician responsibilities are two-fold. One is to support the research of other investigators in the center. The other is to stay connected to statistical research through their own work and to stay up-to-date on emerging statistical methods and techniques. The latter is essential for funding and growth in the center. In some cases, new methods will need to be learned or developed when more standard available methods are not appropriate for a given analysis.
- Any changes in percent support made during proposal writing or after research has been funded should be made jointly between the investigators providing support and the biostatisticians.
- The guideline percentages can be variable over the lifetime of the grant. For multi-year projects, it might be reasonable to support biostatisticians at a higher level during the design phase, a lower level during data collection periods after design issues have been settled and before analysis and analysis have begun, returning to a higher-level once analysis database development and statistical analysis begin. This can be accomplished by providing an average level of support over the entire duration of the grant or specifying different levels during different years of the grant.
- In some cases, investigators have a “portfolio” of related grants. The sum of effort for statistical support from these grants can be combined to reach an appropriate percentage for the type of support desired as described above.
- As a rule, Senior Investigators with established research programs who request effort from a BSR Statistician must make a commitment for funding for at least six months. Exceptions can be made based on discussions with a BSR faculty member. Junior or Early Stage Investigators can make shorter term commitments for providing support.
Guidelines for Authorship on Papers
Authorship on papers should be discussed early. Usually a statistician working 10 percent or more on a project will have contributed enough to the research to be a co-author on papers. In this situation, the statistician is usually listed as the second, or in some cases the third author. Since different fields have different guidelines for authorship these will be taken into consideration during discussions. Authorship may be merited for supervising statisticians who guide the analyses carried out by more junior Biostatisticians.