Abstract 76: Exploring Efficacy and Safety of Very Early Mobilization Within 24 Hours of Stroke Onset Versus Usual Stroke Unit Care (A Very Early Rehabilitation Trial, AVERT): Pre-specified Subgroup Analysis

Julie Bernhardt,Peter Langhorne,Richard I Lindley,Amanda G Thrift,Fiona Ellery,Janice Collier,Leonid Churilov,Marjory Moodie,Helen Dewey,Geoffrey A Donnan,
DOI: https://doi.org/10.1161/str.47.suppl_1.76
IF: 10.17
2016-02-01
Stroke
Abstract:Introduction: Potential benefits of starting out of bed activity early after stroke are tempered with concerns of harm. In this pre-specified subgroup analysis, we examine the potential influence of patient and process factors on outcome at 3 months. Methods: This parallel group, single blind, randomized controlled trial was conducted in 56 acute stroke units, in five countries. Eligible participants were admitted within 24 hours of stroke onset, (ischemic or hemorrhagic) meeting physiological safety criteria. Patients were randomized to very early mobilization (VEM), delivered by physiotherapists and nurses, commenced < 24 hours or to usual care (UC). Modified Rankin Scale (mRS) at 3 months was the primary outcome. Efficacy and safety analyses (serious adverse events [SAEs], deaths) and additional exploratory analysis of age, stroke severity, stroke type, treatment with rtPA, time to first mobilization and geographical region (AUS/NZ vs Asia vs UK), followed a pre-published statistical analysis plan. Results: 2104 patients were enrolled; 1054 randomized to VEM and 1050 to UC and 2083 (99.0%) were followed to 3 months. 525 (50.2%) patients in UC versus 479 (46.2%) in VEM had a good outcome (mRS 0-2; adjusted odds ratio [OR] 0.73, 95%CI 0.59-0.9, p=0.003). 72 patients in UC died (6.9%) versus 88 in VEM (8.4%) (OR 1.34, 95%CI 0.93-1.93. p=0.194). There was no difference in the proportion of patients (UC 209[19.9%] versus VEM 201[19.1%]) experiencing a SAE (IRR 0.88, 95%CI 0.72-1.07). Over the course of the trial, UC shifted to begin a median of 28 minutes earlier each year (95%CI 11.3, 44.6, p<0.001). The efficacy subgroup analysis showed no significant treatment-by-strata interactions, but a more favorable outcome for the UC intervention in patients with intra-cerebral hemorrhage. Death subgroup analysis also favored UC intervention, with no significant treatment-by-strata interactions. Subgroup analysis of SAEs revealed a single significant age-by-treatment interaction (p=0.031), and more favorable outcomes for the VEM in those aged 65-80 years and treated in AUS/NZ. Conclusion: These findings will change the practice of very early stroke rehabilitation.
peripheral vascular disease,clinical neurology
What problem does this paper attempt to address?