Early computed tomography demonstration of cerebral infarction does not correlate with clinical outcome

The results of unenhanced computed tomography (CT) obtained in 105 consecutive patients admitted with cerebral infarction were compared with the patients' clinical outcome and their Glasgow Outcome Scores upon discharge. Specific analysis of the findings in patients with early admission CT (within 12 h of onset) was carried out. The size of the infarction was calculated from delayed CT images (small if less than 50 mm(3); moderate if 50-250 mm(3); large if greater than 250 mm(3)). Initial CT demonstration of the cerebral infarction was accomplished in 57 (54.3%) patients. Forty patients were studied early (<12h) and, of these, 20 (50%) showed abnormal admission CT. The size of the infarction correlated with its demonstration by admission CT. This correlation was found in all patients (p < 0.0001), as well as in those who had early CT (p = 0.015). Also, the size of the infarction correlated with the outcome of the patient (p < 0.001). The demonstration of the lesion by early CT, however, failed to correlate with the outcome of the patients. This finding, although at first surprising, is analogous to reports addressing the predictive value of other CT parameters.