Safety and effectiveness of the Walk 'n Watch structured, progressive exercise protocol delivered by physical therapists for inpatient stroke rehabilitation in Canada: a phase 3, multisite, pragmatic, stepped-wedge, cluster-randomised controlled trial

Abstract

Background: Although clinical guidelines support high repetitions of walking after stroke, practice is slow to change. We undertook an implementation trial to enable entire stroke units to use the Walk 'n Watch structured, progressive exercise protocol. Our objective was to evaluate the effectiveness of the Walk 'n Watch implementation package on patient outcomes after 4 weeks in an inpatient stroke rehabilitation setting.

Methods: This pragmatic phase 3, stepped-wedge, cluster-randomised controlled trial took place in 12 sites (inpatient stroke rehabilitation units) across seven Canadian provinces. Each site was randomly allocated (1:1:1:1) to one of four transition sequences with three sites in each sequence. Sites changed practice from usual care to Walk 'n Watch at different times according to their allocation. All front-line physical therapists were trained to deliver the Walk 'n Watch protocol. Walk 'n Watch required completion of a minimum of 30 min of walking-related activities per session, which progressively increased in intensity based on heart rate and step count monitors. Progressions were prescribed on the basis of a screening 6-minute walk test (6MWT) done by the front-line physical therapist as part of the protocol. The primary endpoint was walking endurance (6MWT) at 4 weeks after randomisation. Masked assessors completed evaluations at baseline and 4 weeks later. The primary analysis used a linear mixed-effects model adjusted for unit size, stratum, calendar time, age, sex, and baseline 6MWT. This trial is registered with ClinicalTrials.gov, NCT04238260, and is complete.

Findings: Between June 4, 2021, and March 1, 2024, we enrolled 12 sites with 314 participants, of whom eight were deemed ineligible after enrolment and 306 were included in the primary analysis (162 in the usual care group, 144 in the Walk 'n Watch group). Participants had a mean age of 68 years (SD 13), a mean time since stroke of 29 days (17), and a baseline 6MWT of 152 m (106). 188 (61%) were male and 118 (39%) were female. The mean 6MWT in the usual care group was 137·1 m (100·9) at baseline and 223·6 m (130·4) after 4 weeks. The mean 6MWT in the Walk 'n Watch group was 163·6 m (112·7) at baseline and 297·2 m (133·2) after 4 weeks. The 6MWT improvement was 43·6 m (95% CI 12·7-76·1) greater in the Walk 'n Watch group than in the usual care group. No serious adverse events occurred during a Walk 'n Watch session. Nine serious adverse events were reported and required admission to acute care (four were recorded in the usual care group and five in the Walk 'n Watch group).

Interpretation: The Walk 'n Watch protocol resulted in a clinically meaningful improvement in walking endurance in patients with subacute stroke in a real-world setting. The protocol can be readily implemented into practice with minimal additional resources. Further research is needed to identify characteristics of patients who might benefit the most from Walk 'n Watch.

Funding: Canadian Institutes of Health Research, Canada Brain Research Fund, Michael Smith Health Research BC, Fonds de recherche du Québec-Santé, Canada Research Program, and Heart and Stroke Foundation of Canada.