David Dubins, Ph.D., B.A.Sc.
Associate Professor, Teaching Stream
Leslie Dan Faculty of Pharmacy
144 College Street
Toronto, Ontario
M5S 3M2
Lab Website: http://individual.utoronto.ca/ddubins/pharmlab

My Tiny Contribution to Clinical Research: FARTSSIE.

Hi, I used to work in clinical research, specifically in clinical trial design and analysis. I love working things out on my own. It's often the best way to understand a phenomena. Sample size calculation is one of those finicky niche areas that require precise calculations using expensive statistical packages. I thought I would develop a free tool to estimate sample sizes, largely based on equations from the following reference:

Julious, SA. "Tutorial in Biostatistics - Sample sizes for clinical trials with Normal data."
Statist. Med. (2004); 23:1921–1986.

I've created an Excel spreadsheet that uses the equations in a Microsoft Excel workbook. The Excel workbook is called "Free Analysis Research Tool for Sample Size Iterative Estimation".

Had I known this spreadsheet would travel, I would have given it a less "cheeky" name. To download the latest version of FARTSSIE, click the following link:

Download the latest version of FARTSSIE.

I now use this spreadsheet in my bioequivalence electives at the Leslie Dan Faculty of Pharmacy, University of Toronto.

Visit Dr. Russel Lenth's Website, the person who made this software possible.

A much better, more powerful, and open source sample size program is PowerTOST, avalilable here: PowerTOST

  • Version 2.9: Release Date 27-Jan-23
    • Two worksheets were added to estimate the combined dropout rates and buffer the sample size accordingly: "Dropouts - 2 rates", and "Dropouts - 3 rates"
  • Version 2.8: Release Date 15-Mar-22
    • TotalCV calculation added to Parallel BE sheet. Prompt for CV updated to Total CV. Fixed Error 13 bug on ISV calculator for OSX
  • Version 2.7: Release Date 10-Mar-21
    • Corrected PowerTOST suggested code, with thanks to Helmut Schütz. Moved FARTSSIE to github.
  • Version 2.6: Release Date 01-Feb-21
    • Fixed variable definitions for Mac, and added graphs for Parallel Bioequivalence worksheet.
  • Version 2.5: Release Date 13-Oct-20
    • Removed all Reference Scaling - Redirect this method to PowerTOST
  • Version 2.4: Release Date 14-Mar-19
    • Corrected rendering of d on superiority and non-inferiority trials
  • Version 2.3: Release Date 1-Feb-17
    • Added Lehr's Formula to Superiority, Parallel worksheet
  • Version 2.2: Release Date 1-Feb-17
    • Added ±d values to graphs, adjusted BE limits for replicate designs
  • Version 2.1: Release Date 16-Jan-17
    • Added distribution graphs to most spreadsheets
  • Version 2.0: Release Date 11-Jan-17
    • TPD HIghly Variable Drug BE limit calculation added
  • Version 1.9:
  • Version 1.8:
    • Fixed precision of ISV and ratio for BE sheets
    • Updated to .xlsm format
  • Version 1.7:
  • Version 1.6:
    • Alternate approaches for expansion of BE limits also included:
  • Version 1.5:
    • "Calculate Power" button added to Bioequivalence worksheets
  • Version 1.4:
    • Now compatible with Excel 2007
    • No extra library to install
    • Opt-out warning at 1001 subjects
    • Bioequivalence warning if expected ratio isn't within BE limits
  • FARTSSIE calculates sample sizes for the following clinical trial designs:
    • Superiority Trials: Parallel and 2-way Crossover
    • Clinical Equivalence Trials: Parallel and 2-way Crossover
    • Non-Inferiority Trials: Parallel and 2-way Crossover
    • Bioequivalence Trials: Crossover, Replicate, and Parallel
    • Trials to a Given Precision: Parallel and Crossover
    • Difference Between Proportions (for smaller trials, up to 170 subjects)
    • Variance Correction (baseline correction, post-dose measures correction, or both)
  • A "boss button" on most trial designs which will work backwards from the sample size your boss (or client) wants. How terrible is that?!
  • No annoying advertisements, registration options, or guarantees. 

I hope you find this tool useful. Did it work? Is there something you liked or disliked about it? Do you have a suggestion that could make it better? Your feedback is appreciated.

   Email Me
Send an Email to David Dubins

Last Updated: 13-Oct-20