The SWIFT Cast trial protocol: a randomized controlled evaluation of the efficacy of an ankle–foot cast on walking recovery early after stroke and the neural–biomechanical correlates of response

Valerie M. Pomeroy, Phillip Rowe, Jean-Claude Baron, Allan Clark, Richard Sealy, Ukadike C. Ugbolue, Ander Kerr

Research output: Contribution to journalArticle

8 Citations (Scopus)


An ankle–foot cast may enable people to repeatedly practice walking with a more normal movement pattern early after stroke.
To evaluate the clinical efficacy of using an ankle–foot cast [soft scotch ankle-foot (SWIFT) Cast] to enhance walking recovery and to find whether site of stroke lesion and/or baseline biomechanical characteristics predict response to a SWIFT Cast.
Randomized, controlled, observer-blind trial.
Participants (n = 120), 3–42 days after stroke with walking difficulty. All will receive conventional physical therapy. Those allocated to the experimental group will also receive a SWIFT Cast for up to six-weeks. During therapy sessions, the SWIFT Cast will be worn for retraining of walking as clinically appropriate. Outside therapy sessions, participants will initially wear the SWIFT Cast for the whole of their waking day, and this will be adjusted as clinically appropriate.
Measures will be undertaken before randomization, six-weeks thereafter and six-months after stroke. Primary outcome will be walking speed. Secondary outcomes will include the Functional Ambulation Category and efficiency of gait (e.g. step-time symmetry). Structural brain imaging using magnetic resonance imaging (standard fluid attenuated inversion recovery and T1-weighted high-resolution ‘volume’ spoiled gradient) will be undertaken at baseline. The clinical efficacy analysis will use analysis of covariance. The relationship between clinical response to therapy and biomechanical data will use correlation and multivariate regression techniques as required. For neuroimaging data, the relationship to clinical response to therapy will be computed using voxel-based lesion-symptom mapping. An interaction test across groups will identify which voxels are associated with different mean levels of treatment efficacy.
Original languageEnglish
Pages (from-to)86-93
Number of pages8
JournalInternational Journal of Stroke
Issue number1
Publication statusPublished - Jan 2012
Externally publishedYes



  • biomechanics
  • orthoses
  • physical therapy
  • rehabilitation
  • stroke
  • walking

Cite this