Table 2

Sample estimates for potential candidate primary outcomes from ReTrain pilot RCT

Primary outcome measureSample size estimates*MCIDObserved SD rangeEffect size (MCID/SD)
Rivermead Mobility Index36–443.0†2.33–2.661.13–1.29
Timed Up and Go1438–26731.2 – 3.4‡15.69–21.390.06–0.22
Modified Patient-Specific Functional Scale16–2001.0 – 3.0‡1.58–1.940.52–1.7
PA (accelerometer)350–1458Not availableNot applicable0.2–0.45§
  • *Figures represent overall (two groups) sample size estimates required for a definitive trial. Sample sizes estimated for 1:1 allocation at 90% power and 5% alpha and assuming 20% attrition. Calculations are conservative showing range from best case scenario (largest MCID and smallest SD) to worst case scenario (smallest MCID and largest SD) of SDs observed in this trial and published MCIDs where available.

  • †MCID available from stroke research for the Rivermead Mobility Index (http://www.strokengine.ca/psycho/rmi_psycho/).

  • ‡MCIDs identified from other disease groups used as proxies as no published stroke MCIDs.47 48

  • §There are no MCID data available for PA (accelerometry) in stroke (or any other cardio vascular disease), we therefore applied sample size calculations undertaken for a relevant ongoing HTA NIHR trial, which estimated n=562 (effect size 0.3) or n=413 (effect size 0.35) (http://www.isrctn.com/ISRCTN15644451).

  • HTA NIHR, Health Technology Assessment, National Institute of Health Research; MCID, minimal clinical important difference; PA, physical activity; RCT, randomised controlled trial; ReTrain, rehabilitation training.