Factors associated with neurological outcome and lesion progression in traumatic subarachnoid hemorrhage patients.

OBJECTIVE: Traumatic subarachnoid hemorrhage (tSAH) is a frequent finding after closed-head injuries, and its presence is a powerful factor associated with poor outcome. The exact mechanism linking tSAH and an adverse outcome is poorly understood. The aim of this study was to identify the factors that may predict outcomes and changes in the computed tomographic (CT) scans of lesions in a selected population of tSAH patients. METHODS: We evaluated 141 patients admitted consecutively from January 1, 1997, to January 31, 1999, with a CT diagnosis of tSAH. The admission and "worst" CT scans were recorded. CT scan changes were reported as "significant CT progression" (changes in the Marshall classification) or "any CT progression." The amount of subarachnoid blood was recorded using a modified Fisher classification. Outcome was assessed at 6 months after injury with the Glasgow Outcome Scale. RESULTS: Twenty-eight patients (19.9%) had an unfavorable Glasgow Outcome Scale outcome. In the univariate analysis, prognosis was significantly related to age, admission Glasgow Coma Scale score, Marshall CT classification score at admission and on the worst CT scan, amount of tSAH, and volume of the associated brain contusions. From multivariate analysis, the only factors independently related to outcome were the Glasgow Coma Scale score (P < 0.01) and size of the tSAH at admission (P < 0.001). Thirty-four patients (24.1%) had significant CT lesion progression, and 66 patients (46.8%) had some lesion progression. Patients having significant progression of the lesion had a higher risk of an unfavorable outcome (32 versus 10%; P = 0.004). Unadjusted factors predicting CT progression were the Glasgow Coma Scale score at admission, the Marshall classification at admission, the amount of subarachnoid blood, and the presence or volume of associated brain contusions at admission. Independent factors associated with significant CT progression were the amount of tSAH (P < 0.001) and the presence or volume of brain contusions at admission (P < 0.001). CONCLUSION: The outcome of patients with tSAH at admission is related in a logistic regression analysis to the admission Glasgow Coma Scale score and to the amount of subarachnoid blood. These patients also have a significant risk of CT progression. The amount of subarachnoid blood and the presence of associated parenchymal damage are powerful independent factors associated with CT progression, thus linking poor outcomes and CT changes.