A novel risk score to predict deep vein thrombosis after spontaneous intracerebral hemorrhage

BACKGROUND AND PURPOSE: Studies showed that patients with hemorrhagic stroke are at a higher risk of developing deep vein thrombosis (DVT) than those with ischemic stroke. We aimed to develop a risk score (intracerebral hemorrhage-associated deep vein thrombosis score, ICH-DVT) for predicting in-hospital DVT after ICH. METHODS: The ICH-DVT was developed based on the Beijing Registration of Intracerebral Hemorrhage, in which eligible patients were randomly divided into derivation (60%) and internal validation cohorts (40%). External validation was performed using the iMCAS study (In-hospital Medical Complication after Acute Stroke). Independent predictors of in-hospital DVT after ICH were obtained using multivariable logistic regression, and β-coefficients were used to generate a scoring system of the ICH-DVT. The area under the receiver operating characteristic curve (AUROC) and the Hosmer–Lemeshow goodness-of-fit test were used to assess model discrimination and calibration, respectively. RESULTS: The overall in-hospital DVT after ICH was 6.3%, 6.0%, and 5.7% in the derivation (n = 1,309), internal validation (n = 655), and external validation (n = 314) cohorts, respectively. A 31-point ICH-DVT was developed from the set of independent predictors including age, hematoma volume, subarachnoid extension, pneumonia, gastrointestinal bleeding, and length of hospitalization. The ICH-DVT showed good discrimination (AUROC) in the derivation (0.81; 95%CI = 0.79–0.83), internal validation (0.83, 95%CI = 0.80–0.86), and external validation (0.88; 95%CI = 0.84–0.92) cohorts. The ICH-DVT was well calibrated (Hosmer–Lemeshow test) in the derivation (P = 0.53), internal validation (P = 0.38), and external validation (P = 0.06) cohorts. CONCLUSION: The ICH-DVT is a valid grading scale for predicting in-hospital DVT after ICH. Further studies on the effect of the ICH-DVT on clinical outcomes after ICH are warranted.