Discrimination of ischemic versus hemorrhagic stroke type by presenting symptoms or signs: a systematic review and meta-analysis

Background: Valid discrimination of ischemic and hemorrhagic stroke relies exclusively on neuroimaging. Novel biomarkers in acute stroke present an opportunity to reconsider combination diagnostic approaches with standardized clinical assessment. Our systematic review evaluates diagnostic properties of acute clinical features in determining primary stroke etiology. Methods: We conducted a systematic review and meta-analysis according to PRISMA guidelines. PubMed and EMBASE were searched from inception to 6^th March 2025. Eligibility criteria included cohort, cross-sectional, case-control, or randomised controlled trial; consecutive adults with an acute stroke confirmed by neuroimaging; and one or more acute stroke symptom(s) or sign(s) recorded by stroke subtype. A random-effects model was used to pool odds ratios. Results: A total of 60 studies, (n = 12,879,006; ischemic stroke = 10,814,474; hemorrhagic stroke = 2,064,532), were eligible. Mean age was 70.45 ± 14.36 years and 51.9 % were women. Clinical presenting symptoms/signs associated with significantly higher odds of hemorrhagic stroke (compared to ischemic stroke) included coma (odds ratio 8.81 [95 % confidence interval, 5.02–15.45]), neck stiffness (5.21 [2.22–12.21]), Glasgow Coma Scale ≤8 (4.37 [2.35–8.16]), vomiting (3.86 [2.71–5.49]), altered consciousness (3.55 [2.55–4.95]), headache (3.49 [2.63–4.64]), syncope (2.74 [1.96–3.83]), seizure (2.67 [1.72–4.15]), abnormal plantar response (1.94 [1.24–3.04]) and vertigo/dizziness (1.32 [1.04–1.68]). Clinical symptoms/signs associated with significantly lower odds of hemorrhagic stroke included morning onset (0.41 [0.32–0.54]), facial weakness (0.66 [0.46–0.94]), hemiplegia (0.68 [0.50–0.91]), and ataxia (0.73 [0.61–0.86]). Conclusion: Our review reports substantive differences in prevalence of stroke symptoms and signs between ischemic and hemorrhagic stroke subtypes.