Reliability of Clinical Scoring Systems in the Diagnosis of Stroke Types in a Tertiary Care Center in South India

Introduction Differentiation of stroke subtype is an essential first step in its management. In low-resource settings where access to neuroimaging is delayed, clinical scoring systems such as the Siriraj and Allen scores have been proposed as bedside tools to differentiate ischemic from hemorrhagic stroke. The primary objective of this study was to determine the diagnostic accuracy of the Siriraj and Allen scoring systems in differentiating ischemic and hemorrhagic stroke by comparison with computed tomography (CT) imaging. Methods A cross-sectional study was conducted among 156 patients admitted with acute stroke in a tertiary care center in South India. Clinical data were collected from patients within 24 hours of their admission to calculate the Siriraj Stroke Score and Allen Stroke Score. All patients underwent non-contrast CT brain imaging, interpreted by radiologists who were blinded to the scores. Sensitivity, specificity, and predictive values of both the scores for detecting hemorrhagic and ischemic stroke were calculated. Results Among the 156 patients (mean age: 54.1 years), 112 (71.8%) had ischemic stroke and 44 (28.2%) had hemorrhagic stroke. The Siriraj and Allen scores yielded equivocal results in 17.3% and 28.8% of cases, respectively. The Siriraj score demonstrated a sensitivity of 38.7% (95%CI, 23.8-56.2%) and a specificity of 96.9% (95%CI, 91.4-99.0%) for detecting hemorrhagic stroke. The corresponding positive and negative predictive values were 80.0% (95%CI, 54.8-92.9%) and 83.3% (95%CI, 75.5-89.1%), respectively. The Allen score showed a sensitivity of 35.7% (95%CI, 20.7-54.2%) and specificity of 88.0% (95%CI, 79.3-93.4%), with positive and negative predictive values of 50.0% (95%CI, 29.9-70.1%) and 80.2% (95%CI, 70.8-87.0%), respectively. Conclusions We found that the Siriraj and Allen stroke scores have poor diagnostic accuracy and yield a large proportion of equivocal results. We conclude that they cannot replace imaging in the differentiation of stroke subtype in the South Indian population.