These authors use the Get With The Guidelines registry as their derivation and validation group to develop a prediction rule for sICH in the setting of thrombolytic use for acute ischemic stroke with in 3 hours of symptom onset. Subsequently, they "externally validate" their rule by applying it retrospectively to the NINDS data set. At the end of it, they come up with a not-so-handy point scale with six clinical features and twenty discrete elements, but basically, these things are bad:
- Older patients
- More severe strokes
- Higher systolic blood pressure
- Elevated blood glucose
- Asian ethnicity
- Male gender
The C-statistic was .71 on their derivation cohort, .70 on their validation cohort, and .68 on the NINDS cohort - which is more or less just OK. In practical, clinical terms, their tool more or less discriminates between folks who are at 1-3% risk of ICH, and then 6-10% risk of ICH. And, I think it's extremely valuable when discussing risks with our patients to not use blanket generalities, and attempt to tailor the discussion to the individual.
"Risk Score for Intracranial Hemorrhage in Patients With Acute Ischemic Stroke Treated
With Intravenous Tissue-Type Plasminogen Activator"