Notes for Biostatistics Clinic Faculty and Staff

  1. Wear your ID tag.
  2. Start each clinic by going around the room with self-introductions (name, department, do they have data or questions?).
  3. If too many investigators attend a clinic to be helped in 1.25 hours, split the investigators and clinic faculty and staff into two rooms (the other room reserved is D2221)
  4. Ask if anyone brought data or has questions and develop a plan to cover these in 1 hour.
  5. If someone brings messy data, show them quickly how to clean it and suggest they bring it back to a future clinic.
  6. Make the session interactive and demonstrate how to analyze data when possible. The hard part is to analyze data and make it interesting enough for others to watch at the same time.
  7. Have a backup lecture to give in case no one brings data or questions.
  8. The clinics run from noon to 1:15p and individual conversations can go until 1:30p.
  9. Print a copy of ClinicSignin (click Printable from that page) and have attendees fill out the form. Notify any attendees other than faculty or med/surg fellows whose faculty mentors are not in attendance that we would like their mentors to come at least once.
  10. Keep in mind that there are other clinics that you can refer researchers to, if you run out of time or their projects are better suited for a different clinic day.
  11. Remind those attending clinics that the drinks are stored in the fridge.

Notes for the Biostat-Clinic Email

  1. The email goes to all statisticians within the department.
  2. Frank will be saving all correspondence for bookkeeping reasons.
  3. Please remember to hit Reply All if you respond so we all know somebody responded.
  4. To make more efficient use of clinic time, view their data beforehand and ensure the data is de-identified and in an appropriate format.
  5. After screening the data, it can be deposited on to the ClinicsData page, a secure page that can only be accessed by people in our department.
  6. It is recommended to paste the following in your initial response:
For information on how to appropriately format data for analyses at a clinic,
see http://biostat.mc.vanderbilt.edu/DataTransmissionProcedures.
For details about the clinics see http://biostat.mc.vanderbilt.edu/Clinics. 

Use of twiki

  • Make an entry for each investigator or medical student presenting a problem, with name, department, division, name of mentor, title, date. See ClinicGeneral under Notes, Data, and Analyses.
  • Taking notes of advice given to investigators will help them when they return to their office. See GenClinicAnalyses for a template.
  • Put the following at the top of any Notes, Data, and Analyses page:
<editsections/>
%TOC%
  • Datasets may be attached to the Notes, Data, and Analyses section in advance of the clinic for easy finding during the clinic.
  • R and other code developed to analyze the data can be inserted under the bullet for that consultation. For example
<highlight>
d <- read.csv("file")
x <- a + b
</highlight>
renders as

d <- read.csv("file")
x <- a + b
and can be copied into a command window.
Attendance Sheet
Ideas from DanielByrne

References

1


[1] Reena Deutsch, Shelley Hurwitz, Janine Janosky, and Robert Oster. The role of education in biostatistical consulting. Stat Med, 26:709-720, 2007.

Topic revision: r12 - 16 Jan 2009 - 16:51:24 - FrankHarrell
 
Register | Log In
Copyright © 2009 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback