Extended use of cranial CT in the evaluation of patients with stroke and transient ischaemic attacks.

The hypothesis was explored that patterns of ischaemic brain damage shown by CT may be of greater value in estimating the relevance of angiographic stenosis in the carotico-vertebral arteries in some patients than presenting clinical features. Five angiographic and six CT abnormalities were defined and charted independently in a blinded manner in 312 patients whom clinical features and subsequent management were known. Charts were combined for statistical analysis. Statistically significant associations were found between two types of CT lesion and angiographic abnormalities, but these were present in only 18% of cases. The distribution of clinical features did not differ significantly within this subset compared to the whole population, from which it was concluded that the hypothesis was upheld. Furthermore significantly more patients in this subset were subjected to carotid endarterectomy, suggesting that management decisions may have been modulated by CT findings in some patients.