To perform an external validation of various scores that predict the development of symptomatic intracerebral hemorrhage (SICH) post recombinant tissue plasminogen activator (rtPA).

Thrombolytic therapy with rtPA has a risk of symptomatic intracerebral hemorrhage (SICH). Various computational scores have been developed that have considered a patient's variables at presentation and used them to predict the likelihood of the development of SICH with a variable degree of success. We aim to determine the validity of these scores in our cohort.

We performed a retrospective chart review of patients who received rtPA at University Hospital, Newark from 7/1/2014 to 3/31/2018. Patients who underwent mechanical thrombectomy and patients with neuroimaging negative for a stroke during the course of hospitalization were excluded. Documented variables were used to calculate the following scores for each patient: SPAN-100, DRAGON, CUCCHIARA, HAT, SEDAN, and SITS-ICH. SICH was defined as an intracranial hemorrhage that causes an increase in NIHSS score by ≥4 points within 7 days post rtPA. A receiver operating characteristic (ROC) curve was obtained by plotting sensitivity against false positive rate (1-specificity). An area under the curve (AUC) was calculated using a Riemann sum.

89 patients met our inclusion criteria, out of which 5 (5.62%) patients developed SICH. ROC analysis indicated that the AUC of the respective SICH scores were: HAT 0.769 (95% CI, 0.58-0.96; P 0.044), DRAGON 0.701 (95% CI, 0.50-0.91; P 0.132), SITS-ICH 0.655 (95% CI, 0.38-0.93; P 0.247), CUCCHIARA 0.705 (95% CI, 0.47-0.94; P 0.125), SPAN-100 0.576 (95% CI, 0.29-0.86; P 0.569) and SEDAN 0.617 (95% CI, 0.37-0.86; P 0.124).

Our study found that HAT score had the highest AUC and it was the most valid in predicting the development of SICH in our independent cohort. We conclude that the HAT score is the most accurate and reliable predictor of SICH development after rtPA.