Related Experiment Video
Updated: Aug 16, 2025
07:35
Author Spotlight: Rehabilitation of Stroke Patients With a Digital Occupational Training System
Published on: December 29, 2023
1.3K
Early prediction of upper limb functioning after stroke using clinical bedside assessments: a prospective
Margit Alt Murphy1,2, Ahmad Al-Shallawi3, Katharina S Sunnerhagen4
1Department of Clinical Neuroscience, Rehabilitation Medicine, Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden. margit.alt-murphy@neuro.gu.se.
Scientific Reports
|December 21, 2022
Summary
Predicting stroke recovery early is crucial. Simple bedside tests at 3 days post-stroke can accurately predict upper limb function at 3 months, aiding treatment and clinical trial decisions.
Area of Science:
- Neurology
- Rehabilitation Medicine
- Clinical Prediction Modeling
Background:
- Accurate prediction of stroke recovery is essential for effective treatment planning and patient selection in clinical trials.
- Early identification of recovery potential can optimize rehabilitation strategies and resource allocation.
- Current assessment methods may be time-consuming or complex, necessitating simpler alternatives.
Purpose of the Study:
- To develop and validate a prediction algorithm for upper limb capacity at 3 months post-stroke.
- To utilize simple, early clinical bedside measures for predicting functional recovery.
- To differentiate between poor, limited, and good recovery outcomes.
Main Methods:
- Secondary analysis of the Stroke Arm Longitudinal Study at Gothenburg University (SALGOT) dataset.
- Inclusion of 94 adult stroke patients with upper limb impairment.
- Cluster analysis to define outcome strata based on 3-month Action Research Arm Test (ARAT) scores.
- Development of prediction models using logistic regression and random forest on training (70%) and testing (30%) sets.
Main Results:
- A final algorithm incorporating three bedside tests (grasp, grip strength, shoulder abduction/elevation) at 3 days post-stroke achieved 86-94% accuracy.
- The model effectively differentiated between poor, limited, and good functional recovery outcomes.
- Additional measures at 4 weeks and stroke type (hemorrhagic) refined underestimated classifications.
Conclusions:
- Simple bedside assessments performed early post-stroke can reliably predict 3-month upper limb capacity.
- The developed algorithm offers a practical tool for routine clinical practice, therapy decisions, and clinical trial patient selection.
- External validation of the prediction model is recommended for broader applicability.
