IJCNMH ARCpublishing

Issue: Issue 6 (2019) – Supplement 1


Carotid stenosis grading: inter-observer agreement using CTA-NASCET

David Alves Berhanu, Carla Guerreiro, Francisco Raposo, Inês Carneiro, Paulo Sequeira, Rita Sousa, and Graça Sá
Background: Carotid artery stenosis is estimated to be responsible for approximately 10-15% of ischemic strokes. Patients with severe symptomatic stenosis benefit from revascularization therapy with endarterectomy or angioplasty and stenting. Revascularization is highly beneficial in patients with greater than 70% stenosis and may be beneficial in certain subgroups of patients with moderate stenosis of 50-69%. Measurement of internal carotid stenosis (ICA) to identify such patients can be made with computed tomography angiography (CTA) using the North American Symptomatic Carotid Endarterectomy Trial (NASCET) method. 
Objectives: Evaluate the reproducibility of the NASCET method between observers in patients with different grades of carotid stenosis and the possible impact of such variability on referrals to revascularization 
Methods: Thirty-seven cervical ICA measurements were obtained from a group of randomly selected patients who presented with acute stroke at a tertiary care hospital. CTA images were obtained using a 64-slice scanner and ICA stenosis was assessed using the NASCET method. Three neuroradiologists performed blinded measurements of the internal carotid arteries. We assessed inter-observer variability using the Cohen's kappa coefficient and Bland-Altman plots. 
Results: Significant agreement differences were found in the measurements obtained by the 3 observers. The kappa statistics (k<0,40) revealed a week-moderate inter-observer agreement. Correspondingly, differences were found in the categorisation of percentage of carotid stenosis, and correspondingly in which patients meet the criteria for referral to revascularization. 
Final Considerations: These findings suggest that the inter-observer variability of carotid stenosis grading using the NASCET method is not neglectable. The week-moderate agreement found suggests that strategies to increase accuracy of carotid stenosis measurement and evaluation using complementary methods may be necessary for correct referral of patients for revascularization.

From the Lisbon Stroke Summit, Lisbon, Portugal. 5–6 April 2019.

International Journal of Clinical Neurosciences and Mental Health 2019; 6(Suppl. 1):P29
Icon_pdf Download PDF