PROGNOSIS FOR ARM MOTOR RECOVERY AFTER STROKE: SYSTEMATIC REVIEW AND META-ANALYSIS

S. Salvalaggio¹, G. Castellini², S. Gianola², L. Cacciante¹, A. Lando¹, G. Pregnolato¹, G. Ossola³, A. Vedovato¹, M. Andò¹, S. Rutkowski⁴, A. Turolla⁵

¹San Camillo IRCCS, Venice, Italy, ²IRCCS Istituto Ortopedico Galeazzi, Milan, Italy, ³Centro di Fisiokinesiterapia, Treviso, Italy, ⁴Opole University of Technology, Opole, Poland, ⁵University of Bologna, Bologna, Italy

Background: Stroke is a leading cause of disability, and more than half of stroke survivors experience arm impairment in the long term. Content and dose of rehabilitation intervention represent boosting factors for motor recovery, but little is known regarding the association between rehabilitation and prediction of recovery after stroke, as well as whether rehabilitation may interfere with the predicted recovery potential.

Purpose: To systematically review and summarize current evidence on prognostic factors for the recovery of upper limb (UL) motor function in stroke patients, after rehabilitation intervention.

Methods: A systematic review was conducted in EMBASE, PubMed, Web of Science, Cochrane, CINAHL, Scopus databases, from inception to July 2021. Longitudinal observational studies were included if investigating at least one motor variable and its relationship to a measure of recovery of UL function, measured before and after a rehabilitation intervention. For each outcome measure considered, the minimally clinical important difference (MCID) was used to distinguish between responders (good recovery) and non-responders (poor recovery). Only studies with retrievable exact number of responders and non-responders were included. Interventions were classified according to rehabilitation modality (i.e. priming, augmenting, task-oriented) provided, thus information on hours delivered (dose) was extracted. The presence of predictive features was investigated for each study. Two independent reviewers completed data extraction and assessed study quality.

Results: 10748 studies were identified, then 7387 screened after duplicates removing. 608 full-texts were assessed for eligibility and 147 were included in the review (i.e. Priming = 35, Augmenting = 100, Task-oriented = 12). Overall, 1475 patients were included, among them 729 were responders (i.e. Priming = 189, Augmenting = 387, Task-oriented = 153) and 746 non-responders (Priming = 209, Augmenting = 471, Task-oriented = 66). 10 studies (i.e. Priming = 3, Augmenting = 6, Task-oriented = 1) found some predictive factors associated with UL motor recovery, such as presence of motor evoked potentials (MEP+), moderate level of residual motor function at baseline according to Fugl-Meyer Assessment Upper Extremity (FMA-UE) or Action Research Arm Test (ARAT) and higher change in the Fractional Anisotropy in the Cortico-Spinal Tract (CST).

Conclusions: Augmenting modalities are the most applied in the field of stroke rehabilitation, followed by priming and task-oriented. The number of responders was slightly lower than non-responders both overall and in the augmenting and priming groups, while for task-oriented intervention responders (n = 153) were more than non-responders (n = 66). Only 15% of the studies (i.e. 10/147) investigating the effect of rehabilitation intervention for UL after stroke also explored which features could predict motor recovery.

Implications: Preserved CST (both functional and structural) and high level of residual motor function are factors to be considered for prediction of UL motor recovery after stroke.Studies on UL rehabilitation after stroke may incorporate some predictive analysis to better identify patients more likely to recover by means of specific interventions.

Funding acknowledgements: The present work was unfunded

Keywords:
Stroke
Rehabilitation
Prognosis

Topics:
Neurology: stroke
Disability & rehabilitation
Neurology

Did this work require ethics approval? No

Reason: The present work didn’t require ethical approval since it is a literature review. However, it is registered in PROSPERO [CRD42021258188].