Abstract
Computed tomography (CT) mixed density of traumatic extra-axial hemorrhages (TEH) or the 'swirl sign' has been reported to correlate with active bleeding found at craniotomy and poor outcome. This study was done to test the hypothesis that mixed density of TEH detected by third or fourth generation CT correlated with the type of bleeding or clinical outcome. All cases of TEH operated at Detroit Receiving Hospital from 1991-1997 were reviewed for type of bleeding (active vs. not active; arterial vs. venous); Glasgow Coma Scale (severe 1-8, not severe 9-15); and Glasgow Outcome Score (1-3 poor; 4,5 good). CT density (CTD) of 51 cases with specific written documentation of bleeding type were then independently reviewed (SKS and MHR) and classified into TEH with mixed or high density. Data was analyzed using corrected Chi Square analysis, Fischer's Exact Test and Pearson's Correlation (SPSS 6.0). The Pearson Chi Square probability for correlation follows: CTD vs. active bleeding, 0.21; CTD vs. arterial, 0.41; CTD vs. severity, 0.57; and CTD vs. outcome, 0.81. No other statistical analysis identified a significant correlation, thus the null hypothesis could not be rejected. CT mixed density was not found to be correlated by more than chance with bleeding type, injury severity or outcome. Surgeon inaccuracy in documentation of bleeding type and use of later generation CT may account for the discrepancy between this and previous studies. Nevertheless, we conclude clinical exam and other published CT criteria are better indicators of injury severity and outcome. [Neurol Res 2002; 24: 125-128]