T he introduction of computed tomography (CT) to clinical practice has had a great impact on our knowledge of cerebrovascular disorders, and cerebral CT has become the most commonly used primary radiologic investigation for stroke.1 " 3 Cerebral CT has shown that the prognosis of intracerebral hemorrhage (ICH) is not as poor as was supposed when small hemorrhages were often undiagnosed or misdiagnosed as ischemic events,and it has changed the order of diagnostic procedures for stroke. Furthermore, in differentiating ischemic infarcts from hemorrhagic lesions, 1 -36 ' 7 cerebral CT has proved to be of crucial importance for therapeutic considerations, particularly anticoagulant treatment.Despite a vast number of CT studies devoted to cerebrovascular disorders, the available knowledge on clinical and CT correlations in stroke is based on highly selected patient samples 8 -14 and little is known of the findings in routine neurologic patients. To elucidate these correlations, we prospectively studied all 1,191 neurologic patients referred for cerebral CT during 1 year in a neurological department. The relation between clinical and CT findings was assessed at two phases, first, after the neurologic eval- Received November 16, 1988; accepted July 11, 1990. uation made on admission and following the first CT investigation and, second, after the final diagnosis was made based on all available information.
Subjects and MethodsDuring the year under survey, 1,191 neurologic patients were referred for cerebral CT. This hospitalbased patient sample comprised 572 men and 619 women aged 15-87 (mean 46) years. Cerebrovascular etiology of central nervous system (CNS) symptoms and signs was suggested clinically in 386 patients (32.4%). The patient sample included all those in whom ICH, subarachnoid hemorrhage (SAH), subdural hematoma, or cerebral infarction were suspected and virtually all those with suggested transient ischemic attack (TIA). Generally accepted diagnostic criteria 15 were used.The CT study was ordered only after recommendation by a neurologist who defined and registered the clinical data and suspected localization of the lesion. The radiologist determined the CT diagnosis. The final diagnosis was based on all the clinical, radiologic, electroencephalographic, neurosurgical, neuropathologic, and ancillary information accumulated. In addition to the 386 patients in whom CT was performed because of a suspected cerebrovascular disorder, 805 patients in whom cerebrovascular disease had not been initially suspected but in whom CT and/or the ancillary investigations revealed cerebrovascular pathology were also analyzed.