External validation of the SI2NCAL2C score for outcomes following cerebral venous thrombosis

Abstract

Objectives: Prognostication for cerebral venous thrombosis (CVT) remains difficult. We sought to validate the SI2NCAL2C score in an international cohort. Materials and methods: The SI2NCAL2C score was originally developed to predict poor outcome (modified Rankin Scale (mRS) 3-6) at 6 months, and mortality at 30 days and 1 year using data from the International CVT Consortium. The SI2NCAL2C score uses 9 variables: the absence of any female-sex-specific risk factors, intracerebral hemorrhage, central nervous system infection, focal neurological deficits, coma, age, lower level of hemoglobin, higher level of glucose, and cancer. The ACTION-CVT study was an international retrospective study that enrolled consecutive patients across 27 centers. The poor outcome score was validated using 90-day mRS due to lack of follow-up at the 6-month time-point in the ACTION-CVT cohort. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) and calibration plots. Missing data were imputed using the additive regression and predictive mean matching methods. Bootstrapping was performed with 1000 iterations. Results: Mortality data were available for 950 patients and poor outcome data were available for 587 of 1,025 patients enrolled in ACTION-CVT. Compared to the International CVT Consortium, the ACTION-CVT cohort was older, less often female, and with milder clinical presentation. Mortality was 2.5% by 30 days and 6.0% by one year. At 90-days, 16.7% had a poor outcome. The SI2NCAL2C score had an AUC of 0.74 [95% CI 0.69-0.79] for 90-day poor outcome, 0.72 [0.60-0.82] for mortality by 30 days, and 0.82 [0.76-0.88] for mortality by one year. Conclusions: The SI2NCAL2C score had acceptable to good performance in an international external validation cohort. The SI2NCAL2C score warrants additional validation studies in diverse populations and clinical implementation studies.