ABSTRACT NUMBER: ESOC2026A74 PREDICTORS, SCREENING TOOLS, AND PROGNOSTIC MODELS FOR STROKE-ASSOCIATED ASPIRATION PNEUMONIA: A PRISMA-SCR–GUIDED SCOPING REVIEW WITH IMPLICATIONS FOR LOW-RESOURCE SETTINGS

Abstract Background and aims Stroke-associated aspiration pneumonia (SAP) is a frequent and preventable complication of acute stroke, contributing to higher mortality, prolonged hospitalization, and increased healthcare costs—particularly in low- and middle-income countries (LMICs) where specialist resources are scarce. We aimed to map global evidence on clinical predictors, bedside dysphagia screening tools, and prognostic models for SAP, and to identify critical research and implementation gaps relevant to resource-constrained settings.To map global evidence on clinical predictors, bedside dysphagia screening tools, and prognostic models for SAP, and to identify critical research and implementation gaps relevant to resource-constrained settings. Methods Following PRISMA-ScR guidelines, we searched PubMed, Embase, Scopus, Cochrane CENTRAL, and IndMed (inception–October 2025). English and non-English studies with available abstracts were included. Data were charted on predictors, screening accuracy, and model performance. Critical appraisal used PROBAST and QUADAS-2. Results Of 112 studies evaluated, most were observational cohorts from high-income regions. Dysphagia (adjusted OR 2.5–5.8), NIHSS ≥10, older age, and brainstem involvement consistently predicted SAP. The Gugging Swallowing Screen (GUSS) showed the highest diagnostic accuracy (sensitivity 90–97%, specificity 75–86%). The single-sip Water Swallow Test emerged as the most feasible initial screen in LMICs due to minimal equipment and training needs. Prognostic models A2DS2 and AIS-APS demonstrated fair-to-good discrimination (AUC 0.73–0.85) but lacked external validation in LMIC settings. Conclusions Standardizing SAP definitions, validating prediction tools in diverse populations, and integrating nurse-led, low-tech screening protocols are urgent priorities to reduce global SAP burden—especially in regions where resources limit access to specialist assessment. Conflict of interest Nothing to disclose