M3 Practice Based Learning and Improvement Presentation Guide: Step 5: Validity

Xiong GL, Adams J. FRISBEE: Does the study fly in the face of evidence-based medicine? The Journal of Family Practice. 2007 Dec;6(12).  Available from: https://www.mdedge.com/psychiatry/article/62967/frisbee-does-study-fly-face-evidence-based-medicine

FRISBEE is great for randomized control trials. See one of the links below if you are using a systematic review. They websites have systematic review evaluation worksheets.

• NNT = number of patients needed to treat in order to prevent one bad outcome
• NNT = 1/ARR
• ARR = CER - EER
• Number Needed to Harm (NNH) is calculated the same way as NNT
• If a research study had non-statistically significant results, it is not appropriate to calculate NNT.
• Should be calculated for binary results (MIs versus no MIs).
• Students should state if is not appropriate to calculate NNT.
• NNT for a study is always for a specific time period.
• NNT does not consider baseline risk.
• NNT can be calculated from data, odds ratios, RRR, and prevalence.
  • Bandolier 
• It is not necessary to have a placebo. For example, you could be comparing to medications or treatments.
• More information - http://www.cebm.net/number-needed-to-treat-nnt/  and https://www.med.uottawa.ca/sim/data/Number_Needed_to_Treat_e.htm