Jennifer Dearborn, MD

Strbian D, Michel P, Seiffge DJ, Saver JL, Numminen H, et al. Symptomatic Intracranial Hemorrhage After Stroke Thrombolysis:Comparison of Prediction Scores. Stroke. 2014

The most feared complication of thrombolysis is symptomatic intracranial hemorrhage, and there does not seem to be a reliable way to efficiently predict which patients will suffer this consequence. Several risk scores have been created, and Srbian et al. compare their performance in a multicenter cohort. They conclude, however that none of the scores had better than moderate performance, which make the scores clinical utility less apparent. 

The score with the highest predictive power was the SEDAN score. This score uses the following parameters to calculate risk of hemorrhage: age (>75 y), NIH Stroke Scale (>10 points), blood glucose (8.1-12.0 mmol/l 1pt; >12.0 2pt), early infarct signs, and hyperdense artery signs. This score had the highest area under the curve. All of the scores had similar components included, with some included age, ethnicity, or platelet count. The strength of this analysis is that it directly compares the scores in a large cohort of 3,102 patients. It is also important because it highlights the similarities of the scores side-by-side, so that one or two “best” approaches emerge.

Taken together, it seems that none of these scores have the capability to help the clinician discriminate about whether to give thrombolysis therapy or not. With an FDA approved treatment available to treat stroke, these scores to not seem adequate enough to alter that path of care if a patient meets other careful screening requirements. Scores such as the SEDAN score could however, be used to explain risk to patients for the clinician in the post-TPA management in the ICU. With interventional thrombolysis being a treatment under scrutiny currently, it is unlikely that a score will assist a clinician in deciding whether to pursue TPA or interventional management, as both carry a risk of hemorrhage. Another point is that the components of the score such as high NIH stroke scale or hyper dense artery, make it likely that patient will suffer a large ischemic stroke with a high rate of disability. These are the patients we may be willing to have a higher threshold of “acceptable risk” if there is likelihood of benefit with thrombolysis.