Validation of shunt dependency prediction scores after aneurysmal spontaneous subarachnoid hemorrhage.

ABSTRACT

BACKGROUND: Currently available scores for predicting shunt dependency after aneurysmal spontaneous subarachnoid hemorrhage (aSAH) are limited and not widely accepted. The key purpose of this study was to validate a recently created score for shunt dependency in aSAH (SDASH) in an independent population of aSAH patients. We compared this new SDASH score based on a combination of the Hunt and Hess grade, Barrow Neurological Institute (BNI) score, and the presence or not of acute hydrocephalus with other published predictive scores. METHODS: The SDASH score, Hijdra score, BNI grading system, chronic hydrocephalus ensuing from SAH score (CHESS), Graeb score, and modified Graeb score (mGS) were calculated for a cohort of aSAH patients. Logistic regression analysis was used to determine the reliability of the SDASH score, and the area under the curve (AUC) of the receiver operating characteristics (ROC) curve was used to assess the discriminative ability of the model. RESULTS: In 214 patients with aSAH, 40 (18.7%) developed shunt-dependent hydrocephalus (SDHC). The AUC for the SDASH score was 0.816. The SDASH score reliably predicted SDHC in aSAH (odds ratio 2.93, 95% CI 1.99-4.31; p < 0.001) with no statistically significant differences being found between the SDASH score and the CHESS score (AUC 0.816), radiological-based Graeb score (AUC 0.742), or modified Graeb score (AUC 0.741). However, the Hijdra score (AUC 0.673) and BNI grading system (AUC 0.616) showed lower predictive values than the SDASH score. CONCLUSIONS: Our findings support the ability of the SDASH score to predict shunt dependency after SAH in a population independent to that used to develop the score. The SDASH score may aid in the early management of hydrocephalus in aSAH, and it does not differ greatly from other predictive scores.