Background: Clear, clinically relevant taxonomies of complex rehabilitation interventions are essential to enable objective evidence synthesis within systematic reviews. Definitions of current physical rehabilitation approaches are generally inadequate, lack universal international acceptance or interpretation, and are difficult to apply objectively when classifying interventions.
Objectives: We aimed to determine an internationally, clinically relevant taxonomy of physical rehabilitation approaches using language and terminology agreed by physiotherapists and stroke survivors, for use within a Cochrane review update.
Methods: We extracted descriptions of physical rehabilitation approaches systematically from the trials included in the existing review, and from foreign-language trials listed as ‘awaiting assessment’. We convened a stakeholder group (SG), comprising 13 purposely selected stroke survivors, carers and physiotherapists. Nominal group techniques were used to reach consensus decisions. The SG debated the treatment components described within the trials, reached consensus on key types of intervention components, agreed descriptions of components and determined categorisation for synthesis of evidence within the updated review.
Results: 84% of the SG agreed that the previously used categories were no longer clinically relevant. Twenty-seven intervention components were identified and described from the intervention components described within 31 trials. These were grouped into categories of: functional task training, musculoskeletal intervention (active), musculoskeletal intervention (passive), neurophysiological intervention, cardiopulmonary interventions, assistive devices and modalities; 100% of the SG agreed with these intervention component descriptions and categories.
Conclusions: A clinically relevant taxonomy of physical rehabilitation intervention components was agreed, and used to categorise evidence within a major update of a Cochrane systematic review of physical rehabilitation approaches for people after stroke. This taxonomy was used within subgroup analysis, facilitating exploration of effect of different treatment components.