The Accuracy of Subjective Assessment of Upper Limb Associated Reactions During Walking in People with Acquired Brain Injury

Abstract

Research Objectives To investigate the accuracy of subjective rating of upper limb associated reactions (ARs) during walking using the International Classification of Functioning, Disability and Health Framework's (ICF) Qualifiers Scale compared to criterion-reference three-dimensional motion analysis (3DMA). Design Cross-sectional observational study. Setting A major brain injury rehabilitation centre in Australia. Participants Forty-one participants with an ABI and hemiplegic upper limb AR during walking. Interventions N/A. Main Outcome Measures Participants underwent 3DMA at their self-selected walking speed to generate an AR outcome measure, quantifying upper limb kinematic deviation compared to healthy controls at each joint axis and whole upper limb. Three experienced neurological physiotherapists viewed video recordings of the participants’ walking trials and applied the ICF Qualifiers Scale to subjectively rate the global AR severity and severity at each upper limb joint on a 5-point scale. The comparative accuracy was calculated via quantifying relationships (Fisher's exact test), percentage agreement, sensitivity, and specificity. Results The ICF Qualifiers Scale had limited accuracy compared to 3DMA hemiplegic upper limb assessment. The wrist severity score was the only outcome demonstrating a significant relationship to the corresponding 3DMA outcome (p < 0.05). Four out of seven outcomes had ≤ 58% agreement in abnormality classification. None of the outcomes had high sensitivity and specificity. The sensitivity was relatively low for all outcomes (0.32-0.73) and specificity relatively high (>0.80) for four of the seven outcomes. Conclusions Given that the ICF Qualifiers scale tended to under classify abnormality with false negatives likely, it cannot be confidently implemented in clinical practice. There appears to be a disconnect between 3DMA and subjective rating of ARs. Evaluation of ARs during walking may require objective assessment. Author(s) Disclosures There are no disclosures to report. To investigate the accuracy of subjective rating of upper limb associated reactions (ARs) during walking using the International Classification of Functioning, Disability and Health Framework's (ICF) Qualifiers Scale compared to criterion-reference three-dimensional motion analysis (3DMA). Cross-sectional observational study. A major brain injury rehabilitation centre in Australia. Forty-one participants with an ABI and hemiplegic upper limb AR during walking. N/A. Participants underwent 3DMA at their self-selected walking speed to generate an AR outcome measure, quantifying upper limb kinematic deviation compared to healthy controls at each joint axis and whole upper limb. Three experienced neurological physiotherapists viewed video recordings of the participants’ walking trials and applied the ICF Qualifiers Scale to subjectively rate the global AR severity and severity at each upper limb joint on a 5-point scale. The comparative accuracy was calculated via quantifying relationships (Fisher's exact test), percentage agreement, sensitivity, and specificity. The ICF Qualifiers Scale had limited accuracy compared to 3DMA hemiplegic upper limb assessment. The wrist severity score was the only outcome demonstrating a significant relationship to the corresponding 3DMA outcome (p < 0.05). Four out of seven outcomes had ≤ 58% agreement in abnormality classification. None of the outcomes had high sensitivity and specificity. The sensitivity was relatively low for all outcomes (0.32-0.73) and specificity relatively high (>0.80) for four of the seven outcomes. Given that the ICF Qualifiers scale tended to under classify abnormality with false negatives likely, it cannot be confidently implemented in clinical practice. There appears to be a disconnect between 3DMA and subjective rating of ARs. Evaluation of ARs during walking may require objective assessment.