Finding, testing and demonstrating efficacy of new treatments for stroke recovery is a multifaceted challenge. We believe that to advance the field, neurorehabilitation trials need a conceptually rigorous starting framework. An essential first step is to agree on definitions of sensorimotor recovery and on measures consistent with these definitions. Such standardization would allow pooling of participant data across studies and institutions aiding meta-analyses of completed trials, more detailed exploration of recovery profiles of our patients and the generation of new hypotheses. Here, we present the results of a consensus meeting about measurement standards and patient characteristics that we suggest should be collected in all future stroke recovery trials. Recommendations are made considering time post stroke and are aligned with the international classification of functioning and disability. A strong case is made for addition of kinematic and kinetic movement quantification. Further work is being undertaken by our group to form consensus on clinical predictors and pre-stroke clinical data that should be collected, as well as recommendations for additional outcome measurement tools. To improve stroke recovery trials, we urge the research community to consider adopting our recommendations in their trial design.
Lack of a standardized approach to measurement in stroke recovery research hampers our ability to advance understanding of recovery mechanisms, devise better treatments and consolidate knowledge from a body of research using meta-analyses.1 As examples, examination of a recent Cochrane Overview of interventions to improve upper limb function after stroke identified 208 unique assessment tools from 243 trials2; another review found more than 100 measures of activities of daily living (ADLs).3 Furthermore, in most motor rehabilitation trials, measures are taken at arbitrary time points relative to stroke onset, e.g. time of admission to, or discharge from, rehabilitation rather than at standard time points aligned with underlying recovery processes.4
We must challenge the common assumption that most sensorimotor therapies are universally applicable and will achieve the same benefit for all people with stroke. The magnitude of change and likelihood of achieving clinically meaningful improvement in response to specific therapies will depend on age, stroke severity, and other factors including pre-existing comorbid conditions (e.g. diabetes, cognitive impairment, depression)5 and pre-stroke lifestyle factors (e.g. social engagement, exercise).6 The respective contributions of these factors have yet to be fully understood. Going forward, we need to identify the determinants that may help predict responders and non-responders to interventions.
The measurement working group of the Stroke Recovery and Rehabilitation Roundtable (SRRR)7 was established to develop recommendations for standardized assessment time points and measures to be included in all adult trials of sensorimotor recovery after stroke. Given the current lack of standards for data collection and heterogeneous reports in stroke recovery trials, our expert group also considered pre-stroke clinical, demographic and stroke-related data that should be collected to improve clinical prediction of recovery and characterization of patient cohorts.
The decision to focus on sensorimotor recovery reflects the volume of existing trials in this area, the range of outcomes currently in use across these trials, and the gap in current research that known international initiatives has not addressed (e.g. Core Outcome Measures in Effectiveness Trials Initiative (COMET), National Institute of Neurological Disorders and Stroke Common Data Elements (NINDS CDE), The International Consortium for Health Outcomes Measurement (ICHOM),8 Improving Research Outcome Measurement in Aphasia (ROMA)9 and Standardization of Measures in Arm Rehabilitation Trials after Stroke (SMART), Supplementary Table 1). Acknowledging that clinical measures cannot distinguish between true neurological repair (behavioral restitution) and use of compensatory strategies,10 a second objective was to consider whether we could recommend specific kinetic and/or kinematic outcomes that reflect quality of motor performance in order to better understand the neurophysiological changes that occur when patients improve.11,12 Our overall objective of the roundtable was to provide recommendations that, if applied, could improve the methodology of rehabilitation and recovery trials, help build our understanding of the trajectory of stroke recovery and aid discovery of new and more targeted treatments.