Best conventional therapy versus modular impairment-oriented training for arm paresis after stroke: a single-blind, multicenter randomized controlled trial.

BACKGROUND The study investigated whether passive splinting or active motor training as either individualized best conventional therapy or as standardized impairment-oriented training (IOT) would be superior in promoting motor recovery in subacute stroke patients with mildly or severely paretic arms. METHODS A total of 148 anterior circulation ischemic stroke patients were randomly assigned to 45 minutes of additional daily arm therapy over 3 to 4 weeks as either (a) passive therapy with inflatable splints or active arm motor therapy as either (b) individualized best conventional therapy (CONV) or (c) standardized IOT, that is Arm BASIS training for severe paresis or Arm Ability training for mild paresis. MAIN OUTCOME MEASURES included the following: Fugl-Meyer arm motor score (severely paretic arms) and the TEMPA time scores (mildly affected arms). Pre-post (immediate effects) and pre-4 weeks follow-up analyses (long-term effects) were performed. RESULTS Overall improvements were documented (mean baseline and change scores efficacy: Fugl-Meyer, arm motor scores, 24.4, +9.1 points; TEMPA, 119, -26.6 seconds; P<.0001), but with no differential effects between splint therapy and the combined active motor rehabilitation groups. Both efficacy and effectiveness analyses indicated, however, bigger immediate motor improvements after IOT as compared with best conventional therapy (Fugl-Meyer, arm motor scores: IOT +12.3, CONV +9.2 points; TEMPA: IOT -31.1 seconds, CONV -20.5 seconds; P=.0363); for mildly affected patients long-term effects could also be substantiated. CONCLUSIONS Specificity of active training seemed more important for motor recovery than intensity (therapy time). The comprehensive modular IOT approach promoted motor recovery in patients with either severe or mild arm paresis.