Biostatistics + Ethics Meetings | VICTR Operations Workgroup Meeting Report

Ethics + Biostatistics Group

Biostatistics Group

• Report out from VICTR Operations Workgroup meeting: VICTROpWorkgroup
• Report out from VICTR Scientific Review Committee meeting: VICTRSrc

VICTR Related Handouts

• Color.Clinics.Slide.pptx: Weekly VICTR funded clinics (color version)
• BW.Clinics.Slide.pptx: Weekly VICTR funded clinics (black ans white version)
• Biostats_N_Research.docx: 'Do I really need to involve a biostatistician in my research?'
  • Key components of material:
    • What is biostatistics?
      • Science of/set of tools for (1) the collection, summarization, & analysis of data from biological research studies; & (2) the presentation, explanation/interpretation of, & inference from the results.
    • Why is it important to understand (bio)statistics?
      • Understanding the literature – should be able to critique the science & the statistics of a study.
      • Designing research studies – much easier to plan for everything before the study is conducted than to try to fix things once the study is finished.
      • Producing manuscripts – reviewers very critical of whether the statistical analyses used were appropriate, what conclusions were drawn, & what inferences were made.
      • Acquiring grants – reviewers very critical of study design, sample size calculation/justification, & proposed statistical analyses.
    • What are some of the ways a biostatistician can help me?
      • Clarify the specific aim(s)/objective(s) & word them in such a way as to convey the statistical analysis used/to be used.
      • Determine the most appropriate study design, maximizing the efficiency & increasing the interpretability & generalizability in the process.
      • Ensure there are no flaws in the design that will affect the statistics & the ability to find anything or make inferences.
      • Use modern experimental designs that allow for early termination of experiments when sufficient evidence for effects is demonstrated, or extends experiments when results are equivocal.
      • Calculate & justify the sample size/power, incorporating any hypotheses, budgetary constraints, recruiting limitations, lost to follow-up, statistical needs, etc.
      • Assist with the selection/refining of the variables to collect, ensuring they relate to the study objective(s) & conform to the proposed statistical analysis.
      • Assist with the design & development of methods, operations & procedures to collect, edit, verify, & manage data, including the design of questionnaires & other data collection instruments.
      • Perform the planned statistical analysis, including data verification & data correction as necessary, using appropriate statistical analysis software & ensuring reproducible research.
      • Interpret & communicate the results in lay terms, explaining all statistical methods used.
      • Assist with the development of IRB & grant proposals, increasing their likelihood of being funded – includes providing formal write-up of sample size/power calculation/justification & statistical analysis plan, ensuring each specific aim/objective is addressed.
      • Assist with the writing & revision of abstracts, posters, slide presentations, & manuscripts for publication, including providing statistical methods & results sections, figures & tables – increasing the likelihood of acceptance & improving the quality of the result.
    • Where in the timeline of research can I benefit from working with a biostatistician?
      • ‘From inception to publication & beyond’... Get a biostatistician involved as soon as possible!
• BiostatSupport.docx: 'How do I get Biostatistics support for my research study?'
  • Key components of material:
    • (1)Check to see if your Department/Division has a collaboration with the Department of Biostatistics.
      • Specific %-effort of biostatisticians are funded by the collaborating department/division in a cost-matching plan with the Department of Biostatistics – ensures protected time for you & your research.
      • Current collaborations can be found at http://biostat.mc.vanderbilt.edu/CollaborationAssignments.
    • (2)If you don’t have a collaboration, take advantage of the Biostatistics support opportunities funded by the Vanderbilt Institute for Clinical and Translational Research (VICTR).1
      • a.Daily Biostatistics Clinics.
        • Monday through Friday from 12:00 – 1:00 PM.
        • See http://biostat.mc.vanderbilt.edu/Clinics for more details, including location & daily themes (eg, Tuesday focuses on Omics & Imaging Data, while Friday covers Basic & Animal Research).
        • Email notification of upcoming attendance helpful – contact biostat-clinic@list.vanderbilt.edu.
      • b.A Biostatistics “Voucher”.
        • Useful when the Biostatistics Clinics are not enough & you need more time with a Biostatistician.
        • Can request specific Biostatistics services including experimental or study design, sample size calculation, data analysis, writing a section of a grant proposal, and/or manuscript preparation.2
        • If approved, awarded $2,000, which covers ~20 hrs of a VICTR supported Biostatistician’s time.
        • Can be awarded >$2,000 (ie, when >20 hrs needed), but must pay for 50% of additional time.
        • Can be awarded up to two Vouchers in a 6-months time period (must be different projects).
        • Must submit a VICTR Resource Request.3
          • Log into StarBRITE (VICTR’s interactive web portal; https://www.mc.vanderbilt.edu/starbrite) with your VUnetID & password.
          • Click on the “Funding Support” tab.
          • Click on the “VICTR Resource Request” button.
          • Click “New Request” & complete the application process.
    • Also take advantage of the VICTR funded “Studios”.
      • A series of structured, dynamic sessions bringing together relevant research experts in a particular methodology to focus on a specific stage of research.
      • Consists of 2-6 experienced faculty, including a biostatistician, selected to participate in a guidance session based on specific areas of research & needs identified by the investigator.
      • Types of studios: Hypothesis Generation Studio; Design Studio; Implementation Studio; Analysis & Interpretation Studio; Translation Studio; Manuscript Studio; Grant Review Studio; & Community Review Board.
      • See the “Research Planning and Implementation” tab in StarBRITE (URL given above) for more information – specifically, click on the “Studios” link.
        • Includes a detailed description of the types of studios & a link to the studio request application.4