Background: Ability to predict risk of postthrombolysis intracerebral hemorrhage (ICH) is currently limited.
Methods: Using data from the Multicenter Recombinant Tissue Plasminogen Activator Stroke Survey Group, we developed a score to predict this risk. One point was assigned for the presence of each of 4 variables: age older than 60 years, baseline National Institutes of Health Stroke Scale score greater than 10, glucose greater than 8.325 mmol/L, and platelet count less than 150,000/mm(3).
Results: Rate of any ICH increased with higher scores: 0 points, 2.6%; 1 point, 9.7%; 2 points, 15.1%; and greater than or equal to 3 points, 37.9%. The model had reasonable discriminatory capability (C-statistic 0.69). A similar pattern was seen with symptomatic and asymptomatic ICH separately, and with radiographically defined parenchymal hemorrhage.
Conclusion: A simple risk score may be useful for predicting postthrombolysis ICH.