Prediction of cerebral infarction and patient outcome in aneurysmal subarachnoid hemorrhage: comparison of new and established radiographic, clinical and combined scores

Background and purposeThere are numerous grading scales to describe the severity of aneurysmal subarachnoid hemorrhage (aSAH) and to predict outcome. Historically, outcome measures are heterogeneous and the comparability of grading scales is therefore limited. We designed this study to compare radiographic, clinical and combined grading systems in aSAH. MethodsData from 423 consecutive patients with aSAH were analyzed. Modified Fisher (mFish), Barrow Neurological Institute (BNI), Hunt and Hess (HH), World Federation of Neurosurgical Societies (WFNS), VASOGRADE (VG) and HAIR scores were calculated from clinical and radiographic data or the combination of both. Outcome measures included the development of new cerebral infarction (CI) and functional patient outcome assessed by the modified Rankin scale. ResultsCerebral infarction and unfavorable outcome were predicted by radiographic, clinical and combined measures (each with P0.001). Clinical (HH, WFNS) and combined (VG, HAIR) scores had superior predictive power for CI compared with mFish grading but not BNI [area under the curve (AUC)(mFish) 0.612, AUC(BNI) 0.616, AUC(WFNS) 0.672, AUC(HH) 0.673, AUC(VG) 0.674, AUC(HAIR) 0.638]. Predictive performances of clinical gradings (HH, WFNS) for patient outcome were superior to radiographic measures and of similar quality or better than combined systems (AUC(BNI) 0.628, AUC(mFish) 0.654, AUC(WFNS) 0.736, AUC(HH) 0.749, AUC(VG) 0.711, AUC(HAIR) 0.739). ConclusionsKnowledge of the merits and limitations of clinical, radiographic and combined scores is necessary in routine clinical practice. The new combined grading systems (HAIR, VG) showed no superiority compared with the established clinical measures (WFNS, HH) in predicting CI and unfavorable patient outcome.