• By enrolling in the percent effort-based collaboration plan
  • This plan allocates MS and PhD biostatistician percent efforts (typically in the 2.5:1 ratio) to work long-term with research groups, divisions, or departments. This approach makes it easy for the same statisticians to stay involved with a research group and facilitates the building of research teams and planning ahead for grant proposals, manuscripts, and abstract deadlines. It also helps our department plan its recruiting. Groups enrolling in this plan have 20% of the efforts of biostatisticians devoted to the research group paid by the Department of Biostatistics as long as the biostatisticians are doing work that is directly related to research development, teaching, or mentoring.
  • We track time spent working with other departments on this plan so the other groups can make adjustments in percent efforts they fund over time and so that research group leaders know how the resource is being used.
  • An excellent way to fund these efforts is for new faculty to include biostatistical support in their package.
• By fees for service (draft version)
  • Unlike divisions or research groups enrolling in our collaboration plan in which we provide at least a 20% discount in salaries and fringe benefits covered, the fee for service plan involves the other division or research group paying for all salary, fringe benefits, a 25% overhead, and an additional amount depending on how much advance notice is given our faculty and staff to work on the project. Unlike project work, for grant proposal assistance, we charge by the hour using the same rate for all PhD and MS biostatisticians.
    • For requests made more than 45 days before a deadline, the charge for grant proposal assistance for groups not taking part in our collaboration plan is $75 per hour.
    • For emergency requests with a deadline within 4 days of the request (e.g., a grant proposal needing to be completed within 4 days of the biostatistician receiving a draft of the proposal) the hourly charge is tripled.
    • For requests made between 45 days and 5 days of a deadline the additional charge factor varies from 1.0 to 2.5 according to the formula 2.6875 - 0.0375*d, where d is the number of days of advance notice.
    • The Department will provide the PI with an estimate of the expected cost in advance, and a "not to exceed" cost. We will notify the PI before working more hours than the "not to exceed" cost covers.
    • A grant proposal of average complexity with 45 days of advance notice requires about 50 hours of biostatistician time and costs the PI around $3750.
    • For multi-department grant proposals, each group not subscribing to the collaboration plan will be billed separately in proportion to the work done specifically for planning that group's design and analysis.

Biostatistics Clinics

• Monday: General (including diagnostic and prognostic research)
• Tuesday: Omics Data (data in which there are more measurements than subjects; operated with the Dept. of Biomedical Informatics with participation of the Mass Spectrometry Research Center)
• Wednesday: Surgery, Anesthesiology, Emergency and Critical Care Medicine
• Thursday: Clinical Research (non-surgical; co-sponsored by GCRC; Room A3210 MCN)
• Friday: Basic Research

All clinics are held noon-1:15pm and all except for Thursday are held in MCN D2221. For more details visit http://biostat.mc.vanderbilt.edu/Clinics.

The Clinics are sponsored by the Dept. of Biostatistics, Vanderbilt Institute for Clinical and Translational Research (CTSA), Dept. of Biomedical Informatics, Clinical Research Center, Section of Surgical Sciences, School of Nursing, and the Vanderbilt Kennedy Center for Research on Human Development.

Who to Contact

Otherwise, if you would like more information or to arrange a meeting to discuss biostatistical support, contact Diane Kolb, Department of Biostatistics, at diane.kolb@vanderbilt.edu or 343-2227.

If your research group is already covered by a Biostatistics Collaboration Plan, contact the lead statistician who is responsible for your group. Assignments are listed at CollaborationAssignments.

• Benefits of Biostatistics to the Basic Scientist
• Collaboration Cost-Sharing Protected Time Plan
• Divisions and Research Groups Currently Covered by the Collaboration Plan
• Short Courses and Other Educational Material
• Checklist for Authors

Collaboration Protected Time and Cost-Sharing Model

Goals

• Integrate Biostatistics into research fabric of VU School of Medicine
• Develop long-term collaborative relationships; develop statistical scientists instead of statistical consultants
• Provide continuity that will allow biostatisticians to learn enough about biomedical research areas to be effective co-investigators
• Increase NIH grant funding by maintaining grant development capabilities
• Help organize research teams that plan in advance and submit better grant proposals that have a higher likelihood of funding
• Always have someone with available FTE who can be listed on a grant application
• Be able to respond to RFPs and RFAs with short lead times, and to training grant opportunities
• Hire new faculty and staff in proportion to anticipated grant funding
• Foster research in clinical departments, in ways other than just grant development
  • participate in meetings
  • work with fellows and residents
  • improve research methodology skills of faculty
  • help develop new clinical investigators

The Problem

• Easy to fund biostatisticians on grant proposals
• Difficult to hire ahead of funding
• Difficult to maintain enough unallocated percent efforts to collaborate on new ideas
• Moderately large clinical grant proposal requires more than 50 hours of combined MS + PhD statistician time
• Many PIs see grant proposals as needing only last-minute statistical consulting and may get a different consultant every time
• Biostatisticians do not have time to do their best work, especially with regard to study design
• Difficult to keep good biostatisticians (especially in a market with huge shortage of graduates)
• Charging by the hour for proposals will allow the department to fund efforts but will stifle the collaborative spirit
• Charging by the hour makes it very difficult to hire biostatisticians in advance of need since the amount of this funding is difficult to predict at the time when hiring budgets are finalized
• There are ways other than grant proposals that biostatisticians can help foster research in clinical departments

The Plan

• Replace consultation (other than for very small projects) with collaboration
• Support other departments through new long-term collaboration model and rarely through hourly-charged consultation
• Department of Biostatistics development fund will provide 1:4 cost matching to other Divisions and Departments
• Example: Suppose that a Division with 7 active researchers not supported by a biostatistics core facility such as the Cancer Biostatistics Division or the GCRC needs one full-time PhD and one full-time MS biostatistician. That Division would pay
  • 4/5 of the MS biostatistician salary + fringe
  • 4/5 of the PhD biostatistician's non-protected time salary + fringe (64% of tenure-track faculty salary, 4/5 of non-tenure-track faculty (Note: Tenure-track faculty in the Department of Biostatistics receive 20% protected time for methodologic research)
  • 1/2 of professional expenses for these personnel
• This would provide
  • priority work on grant proposals from this group
  • 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)
  • assistance with research conference (e.g., data analysis and pre-conference critique of fellows' presentations)
  • teaching short courses in experimental design and analysis methodology for the Division
  • K award mentoring
• Understandings:
  • As grant proposals arising from this arrangement are funded, grant funds will supplement rather than replace the collaborative funding arrangement
  • As grants are funded, either new personnel will be hired or identified by the Department of Biostatistics to work on the new projects, or personnel supported by the collaboration arrangement will move to the new projects and new personnel will assume positions funded by the cost-sharing arrangement
  • For groups not requiring full-time collaborators, existing personnel may assume both roles
  • Division or Department head will prioritize usage of the resource by her or his faculty
  • Division or Department investigators will provide yearly input to Department of Biostatistics for faculty and staff reviews, regarding quality of collaborations and research by biostatisticians
• Option for Divisions/Departments Not in Collaboration Plan
  • Consulting charged by the hour
  • Available on an ''available personnel'', first-come, first-serve basis if sufficient lead time given
  • Difficult to allocate percent efforts to consulting group
  • Becoming more difficult to charge for internal consulting not based on FTEs

How to Take Part in the Program

-- FrankHarrell - 07 Apr 2004, 28 Oct 2005

Biostatistics Collaboration Plan Assignments

*Note: If your department is not participating in a collaboration plan, you may wish to attend one of the Clinics offered through the Biostatistics Department. If your department is in a collaboration plan, you are always welcome at clinics too.

Benefits of Biostatistics to the Basic Scientist | Clinic

Example Questions Biostatistics Can Answer

• What does percent change really assume?
• Why did taking logarithms get rid of high outliers but create low outliers?
• What is the optimum transformation of my variables?
• What statistical method can validly deal with values below the detection limit?
• Do I need more mice or more serial measurements per mouse?
• What's the best way to analyze multiple measurements per mouse?

Experimental Design

• Identifying sources of bias: biostatistics can assist in identifying sources of bias that may make results of experiments difficult to interpret, such as
  • litter effects: correlations of responses of animals from the same litter may reduce the effective sample size of the experiment
  • order effects: results may change over time due to subtle animal selection biases, experimenter fatigue, or refinements in measurement techniques
  • experimental condition effects: laboratory conditions (e.g., temperature) that are not constant over the duration of a long experiment may need to be accounted for in the design (through randomization) or in the analysis
  • optimizing measurements: sometimes optimizing measurements (e.g., changing pattern recognition criteria or image analysis parameters) may result in techniques that are too tailored to the current experiment
• Selecting an experimental design: taking into account the goals and limitations of the experiment to select an optimum design such as a parallel group concurrent control design vs. pre-post vs. crossover design; factorial designs to simultaneously study two or more experimental manipulations; randomized block designs to account for different background experimental conditions; choosing between more animals or more serial measurements per animal. Accounting for carryover effects in crossover designs.
• Estimating required sample size: computing an adequate sample size based on the experimental design chosen and the inherent between-animal variability of measurements. Sample size can be chosen to achieve a given sensitivity to detect an effect (power) or to achieve a given precision ("margin of error") of final effect estimates. Choosing an adequate sample size will make the experiment informative.
• Justifying a given sample size: when budgetary contraints alone dictate the sample size, one can compute the power or precision that is likely to result from the experiment. If the estimated precision is too low, the experimenter may decide to save resources for another time.
• Making optimum use of animals or human specimens: choosing an experimental design that results in sacrificing the minimum number of animals or acquiring the least amount of human blood or biopsies; setting up a factorial design to get two or more experiments out of one group of animals; determining whether control animals from an older experiment can be used for a new experiment or developing a statistical adjustment that may allow such recycling of old data.
• Developing sequential designs: allowing the ultimate sample size to be a main quantity to be estimated as the study unfolds. Results can be updated as more animals are studied, especially when prior data for estimating an adequate sample size are unavailable. In some cases, experiments may be terminated earlier than planned when results are definitive or further experimentation is deemed futile.
• Taking number of variables into account: safeguarding against analyzing a large number of variables from a small number of animals.

Data Analysis

• Choosing robust methods: avoid making difficult-to-test assumptions; using methods that do not assume the raw data to be normally distributed; using methods that are not greatly effected by "outliers" so that one is not tempted to remove such observations from the analysis. Account for censored or truncated data such as measurements below the lower limit of detectability of an assay.
• Using powerful methods: using analytic methods that get the most out of the data
• Computing proper P-values and confidence limits: these should take the experimental design into account and use the most accurate probability distributions
• Proper analysis of serial data: when each animal is measurement multiple times, the responses are correlated. This correlation pattern must be taken into account to achieve accurate P-values and confidence intervals. Ordinary techniques such as two-way analysis of variance are not appropriate in this situation. In the last five years there has been an explosion of statistical techniques for analyzing serial data.
• Analysis of gene expression data: this requires specialized techniques that often involve special multivariate dimensionality reduction and visualization techniques; attention to various components of error is needed.
• Dose-response characterizations: estimating entire dose-response curves when possible, to avoid multiple comparison problems that result from running a separate statistical test at each dose.
• Time-response characterizations: use flexible curve-fitting techniques while preserving statistical properties, to estimate an entire time-response profile when each animal is measured serially. As with dose-response analysis this avoids inflation of type I error that results when differences in experimental groups are tested separately at each time point.
• Statistical modeling: development of response models that account for multiple variables simultaneously (e.g., dose, time, laboratory conditions, multivariate regulation of cytokines, polymorphisms related to drug response); analysis of covariance taking important sources of animal heterogeneity into account to gain precision and power compared to ordinary unadjusted analyses such as ANOVA. Statistical models can also account for uncontrolled confounding.

Reporting and Graphics

• Statistical sections: writing statistical sections for peer-reviewed articles, to describe the experimental design and how the data were analyzed.
• Statistical reports: writing statistical reports and composing tables of summary statistics for investigators
• Statistical graphics: use many of the state-of-the-art graphical techniques for reporting experimental data that are described in The Elements of Graphing Data by Bill Cleveland, as well as using other high-information high-readability statistical graphics. Translate tables into more easily read graphics.

Data Management, Archiving, and Reproducible Analysis

• Data management: the biostatistics core can develop computerized data collection instruments with quality control checking; primary data can be quickly converted to analytic files for use by statistical packages. Gene chip data will be managed using relational database software that can efficiently handle very large databases.
• Data archiving and cataloging: upon request we can archive experimental data in perpetuity in formats that will be accessible even when software changes; data can be cataloged so as to be easily found in the future. This allows control data from previous studies to be searched. Gene chip data will be archived in an efficient storage format.
• Data security: access to unpublished data can be made secure.
• Program archiving: the biostatistics core conducts all statistical analyses by developing programs or scripts that can easily be re-run in the future (e.g., on similar new data or on corrected data). These scripts document exactly how analyses were done and allow analyses to be reproducible. These scripts are archived and cataloged.

• See Hackam DG, Redelmeier DA. Translation of research evidence from animals to humans. JAMA 296:1731-1732; 2006 for a review of basic science literature that documents systemic methodologic shortcomings. In a personal communication on 20Oct06 the authors reported that they found a few more biostatistical problems that could not make it into the JAMA article (for space constraints).
  • none of the articles contained a sample size calculation
  • none of the articles identified a primary outcome measure
  • none of the articles mentioned whether they tested assumptions or did distributional testing (though a few used non-parametric tests)
  • most articles had more than 30 endpoints (but few adjusted for multiplicity, as noted in the article)

Biostatistics Clinic for Basic Scientists: Fridays at noon (D2221 Med Ctr North, lunch provided)

Policies and Procedures for Grant Proposals

Biostatisticians in the Department of Biostatistics participate in grant development in various ways including:

• assisting with the formation and operation of a proposal development team
• assisting the investigators in refining study questions and measurement methods
• developing study and experimental designs
• writing statistical analysis plans
• computing precision, power, and sample sizes necessary to achieve a given precision of estimation or a given power

When an investigator claims to need only a sample size calculation, we usually find that their fundamental study design requires improvement. Unless we are provided sufficient time to participate in the study design, we do not work on proposals for which the investigator "only needs a sample size calculation."

Investigators Served

• explicitly covered under the auspices of the CTSA or Cancer Center Biostatistics Core Resources
• multi-department proposals of major importance to the School that are initiated by the School of Medicine (not by a department of the School)
• developed by a research group that is participating in our collaboration cost-sharing plan
• developed by other groups who we will charge by the hour (see ServicesOverview)
• logical extensions (e.g., competing renewal) of a grant that is already funding a biostatistician who will work on the new proposal
• research areas of special interest for a PhD faculty biostatistician

FOR INVESTIGATORS

Requirement of Advance Notice

Grant proposals require input from a number of experts. Teams assembled at least four months before submission generate the strongest grant proposals. Optimally, team meetings should meet every three weeks to generate at least two major proposal revisions. From the point of view of the biostatisticians, they will then have sufficient time to do their best work and thus, significantly enhance the likelihood of funding.

We usually require two or more months advance notice when asked to provide assistance with proposals. If not, we will provide assistance only if personnel are available and we will charge departments and divisions by the hour for this service. Exceptions will be granted for cases involving resubmissions, renewals, or submissions of proposals that are substantially similar to previous grants in which the biostatisticians have fully participated.

When there is insufficient time for one of our PhD or MS biostatisticians to work with you on a proposal, we will still provide free consultation during our daily Biostatistics Clinics. Although you will not be able to name the consulting statistician as a Key Personnel on the proposal, you will receive valuable advice on how to avoid major design or analysis pitfalls. On rare occasions, when the design is very simple, we can analyze simple pilot datasets or perform sample size calculations.

Use of Biostatisticians' Names on Proposal