ABSTRACT NUMBER: ESOC2026A1644 PREDICTIVE VALUE OF EARLY CLINICAL TRIAGE SCALES FOR REHABILITATION OUTCOMES AFTER ISCHEMIC AND HEMORRHAGIC STROKE

Abstract Background and aims Early triage tools, such as the AlphaFIM and Orpington Prognostic Scales (OPS) are commonly used to guide inpatient stroke rehabilitation referrals. The AlphaFIM, standardized for Day-3, has not been well validated in intracerebral hemorrhage (ICH), a population with greater initial disability and often delayed recovery. We examined predictive value of Day-3 AlphaFIM and serial OPS for rehabilitation and long-term outcomes in ischemic stroke (IS) and ICH survivors. Methods This single-centre, prospective pilot study enrolled adults with first-ever IS or spontaneous ICH with moderate–severe disability. Day-3 AlphaFIM was obtained as standard care. OPS was assessed serially at baseline, days 7, 10, 14, and 30, or until transfer to inpatient rehabilitation. Rehabilitation outcomes (FIM gain, FIM efficiency) and longitudinal outcomes (modified Rankin Scale) at 3, 6 and 12 months were collected. Results Sixty-five participants (47 IS, 18 ICH) were recruited; 48 (74%) were admitted to inpatient rehabilitation. Day-3 AlphaFIM showed weak correlations with FIM gain and efficiency in both stroke subtypes. In contrast, the last available OPS demonstrated moderate correlations with FIM gain, suggesting larger gains among more impaired patients. Both scales were directionally associated with long-term mRS, with stronger correlations in ICH. Conclusions Day-3 AlphaFIM demonstrated limited predictive value for rehabilitation gains. Serial OPS, particularly when performed later in the acute phase, showed more consistent associations with functional improvement and may represent a feasible alternative scale for delayed recovery. These findings support further study of optimal timing of rehabilitation triage assessments, particularly for hemorrhagic stroke survivors. Conflict of interest Sebastian Fridman: nothing to disclose. Lauren Mai: nothing to disclose.