External validation of SEDAN-S score: A grading scale to predict hemorrhagic transformation after thrombolytic therapy in acute ischemic stroke

Disparities in stroke presentation, imaging access, and comorbidity patterns between rural and urban populations may affect the predictive accuracy of hemorrhagic transformation models. The SEDAN-S score, which includes clinical, laboratory, and radiographic parameters, was developed in a rural setting and showed high accuracy to predict hemorrhagic transformation (HT) in patients with acute ischemic stroke treated with thrombolytics. We aimed to externally validate the score in an urban population. A validated and reliable score applicable to both settings could guide safer thrombolytic use and improve individualized risk assessment regardless of geography. We reviewed patients with acute ischemic stroke treated with intravenous thrombolytics at The University of Oklahoma Health Sciences Center from January 2022 to October 2024. Demographics, clinical presentations, laboratory values, and neuroimaging were collected. HT found on neuroimaging within 24 hours after rtPA was reviewed, and the subgroup of symptomatic intracerebral hemorrhage (sICH) was noted. The cohort was divided in two groups, HT and NoHT. The two groups were compared by univariate analyses. Propensity-score matching (PSM) was applied to balance demographics and comorbidities between the two groups, and length of stay, mortality, and 30-day mRS were analyzed between the two matched groups. SEDAN and SEDAN-S scores were calculated, and area under the receiver operating characteristic curve (AUC) with its 95% confidence interval was computed for each model. P value was set at 0.05. Of 200 patients included in this study, 22 (11%) developed HT. The two groups were comparable in demographics and baseline comorbidities. HT group had higher median NIHSS (17 [10-18] vs. 10 [5-15], p<0.01), lower median ASPECT score (7 [6-9] vs. 9 [7-10], p<0.01) and higher rate of large vessel occlusion (68.2% vs. 45.5% p=0.045), compared to NoHT group. SEDAN score demonstrated significantly higher accuracy in predicting HT after rtPA (AUC=0.65, 95%CI:0.54-0.76) compared to SEDAN-S score (AUC=0.61, 95%CI:0.49-0.72), p=0.034. After PSM 1:1, only the subgroup of sICH remained independently associated with in-hospital mortality (OR: 8, 95%CI: 1.87-73.55, p=0.03) and lower odds of mRS0-2 at 30 days (OR:0.19, 95%CI: 0.065-0.555, p=0.002). In the urban setting, SEDAN score retains a better accuracy in predicting HT after thrombolytics compared to SEDAN-S score, highlighting inherent differences between urban and rural populations. Among HT, only the subgroup of sICH is associated with higher odds of in-hospital mortality and lower likelihood of favorable short-term outcome.