Early Recognition and Referral of Acute Stroke in Primary and Emergency Care: A Systematic Review
Source: PubMed Central Open Access, NCBI / U.S. National Library of Medicine
Introduction Early recognition and referral are critical to minimizing morbidity and mortality in acute stroke, but evaluation and referral processes differ worldwide. In this systematic review we examined the accuracy of recognition tools, referral patterns, outcomes, and factors affecting efficiency in primary and emergency care settings. Methods Following PRISMA 2020 guidelines, we searched PubMed, Scopus, Web of Science, and Cochrane Library for studies published January 2003–December 2025. Eligible studies included randomized controlled trials, cohort, case-control, cross-sectional, and large case series (> 30 patients) involving adults with acute ischemic or hemorrhagic stroke. Risk of bias was assessed using Cochrane Risk-of-Bias 2 (RoB) and RoB in non-standardized studies-I. We extracted data on diagnostic accuracy, referral pathways, outcomes, and systemic factors. Results We identified 206 papers, of which 33 studies met our inclusion criteria. Recognition tools such as Face, Arms, Speech, Time (FAST); Recognition of Stroke in the Emergency Room, the Cincinnati Prehospital Stroke Scale, and National Institutes of Health Stroke Scale showed good pooled sensitivity (79–95%) but variable specificity (52–84%). Newer technologies, including the PreHospital Ambulance Stroke Test, FAST-ED, and artificial intelligence (AI)-based models, showed promise but need validation. Referral strategies such as emergency medical services prenotification, dispatcher triage, and mobile s
Abstract
Introduction Early recognition and referral are critical to minimizing morbidity and mortality in acute stroke, but evaluation and referral processes differ worldwide. In this systematic review we examined the accuracy of recognition tools, referral patterns, outcomes, and factors affecting efficiency in primary and emergency care settings. Methods Following PRISMA 2020 guidelines, we searched PubMed, Scopus, Web of Science, and Cochrane Library for studies published January 2003–December 2025. Eligible studies included randomized controlled trials, cohort, case-control, cross-sectional, and large case series (> 30 patients) involving adults with acute ischemic or hemorrhagic stroke. Risk of bias was assessed using Cochrane Risk-of-Bias 2 (RoB) and RoB in non-standardized studies-I. We extracted data on diagnostic accuracy, referral pathways, outcomes, and systemic factors. Results We identified 206 papers, of which 33 studies met our inclusion criteria. Recognition tools such as Face, Arms, Speech, Time (FAST); Recognition of Stroke in the Emergency Room, the Cincinnati Prehospital Stroke Scale, and National Institutes of Health Stroke Scale showed good pooled sensitivity (79–95%) but variable specificity (52–84%). Newer technologies, including the PreHospital Ambulance Stroke Test, FAST-ED, and artificial intelligence (AI)-based models, showed promise but need validation. Referral strategies such as emergency medical services prenotification, dispatcher triage, and mobile stroke units reduced prehospital delays. Seven studies reported onset-to-door times 12–22 minutes faster and 7–12% increase in reperfusion eligibility. Increased referral efficiency was associated with a reduction in mortality of approximately 8–12% and improvements in functional independence of 10–15%, with persistent disparities reported in resource-limited settings. Conclusion Early recognition and referral improve outcomes in patients with acute stroke. Structured tools and system-level interventions reduce mortality, while AI and mobile stroke units show promise. Strengthening referral systems and adopting cost-effective triage strategies may support equitable implementation, particularly in low-resource settings, as addressing systemic and geographic barriers is critical for equitable stroke care.