Significance of Computed Tomography in the Diagnosis of Cerebrovascular Accidents

Abstract

Introduction: Cerebrovascular Accident (CVA) is defined as abrupt onset of a neurological deficit that is attributable to a focal vascular cause. CT scan is a widely available, affordable, non-invasive and relatively accurate investigation in patients with stroke and is important to identify stroke pathology and exclude mimics. Aim of this study is to establish the diagnostic significance of computed tomography in cerebrovascular accident and to differentiate between cerebral infarction and cerebral haemorrhage with CT for better management of CVA. Methods: A one year observational cross sectional study was conducted in 100 patients that presented at the department of radiodiagnosis from emergency or ward within the one year of study period with the clinical diagnosis of stroke, and had a brain CT scan done within one to fourteen days of onset. Results: A total of 100 patients were studied. 66 were male and 34 were female with a male/female ratio of 1.9:1. Maximum number of cases (39%) was in the age group of 61-80 yrs. Among 100 patients, 55 cases were clinically diagnosed as hemorrhagic stroke and 45 cases were clinically diagnosed with an infarct. Out of the 55 hemorrhagic cases, two cases were diagnosed as both hemorrhage and infarct by CT scan, one case had normal CT scan findings and one had subdural haemorrhage. These four cases were excluded while comparing the clinical diagnosis with CT scan finding. Among 51 clinically diagnosed cases of hemorrhagic stroke, 32(62.7%) cases were proved by CT scan as hemorrhagic stroke and among clinically diagnosed cases of infarct, 39(86.7%) cases were proved by CT scan as infarct which is statistically significant (p <0.001). A significant agreement between clinical and CT diagnosis was observed as indicated by kappa value of 0.49. Sensitivity, specificity, positive predictive value and negative predictive value of clinical findings as compared to CT in diagnosing hemorrhage were 84.2%, 67.2%, 62.8% and 86.7% respectively. The accuracy of clinical diagnosis is 74%. Conclusion: This study showed that CT scan is a useful diagnostic modality to identify stroke pathology and to exclude mimics.