CT and clinical correlation of stroke diagnosis, pattern and clinical outcome among stroke patients visting Tikur Anbessa Hospital.

BACKGROUND: Stroke is the leading cause of morbidity and mortality and the most common neurological reason for hospitalizations world wide. Though stroke diagnosis is clinical, it is inaccurate in 10-15% of the cases and hence cross-sectional imaging, like CT and MRI, are necessary for accurate etiological diagnosis and advertent clinical management, and to rule out stroke mimics.

OBJECTIVE: To assess the CT and clinical pattern of stroke, their correlation, presentation, outcome, and risk factors of strokes in a teaching hospital setup.

METHODS: Retrospective chart review of clinically and CT diagnosed stroke patients diagnosed between January 2000 and March 2005 in Tikur Anbessa tertiary referral and teaching hospital.

RESULTS: Stroke accounted for 5% of all head CT indications done. Out of the eligible study population 55.7% were male with female to male ratio of 1.3:1. The mean age of the patients was 50.6 year (range 13-82). The main clinical presentation was hemi paresis (77.1%) and 20.8% were comatose with mean Glasgow coma scale of 5.7 +/- 2.8, stroke mortality was 21% and 31% had persistent neurological deficit. Clinical diagnosis was inaccurate in 30% of the patients, with low etiogical agreement between CT and clinical stroke subtypes diagnosis (Kappa = .334, 95CI .194-.474). On CT 54.8% patients had ischemic and 34.6% had hemorrhagic stroke diagnosis. The main risk factors were hypertension (52%) and Diabetes mellitus (26%). The mean duration of illness before CT diagnosis was 22 days (range 1 hr-360 days).

CONCLUSION: Stroke is not uncommon in our setting and associated with significant morbidity and mortality compounded by delayed diagnosis and possibly by less accurate etiological and clinical diagnosis. Therefore introduction and dissemination of CT service in public and private health institution should be encouraged.