Comparison between the NASCET method and subjective visual impression in the evaluation of internal carotid artery stenosis

RESUMO

OBJECTIVE: To evaluate the accuracy of subjective visual impression (SVI) of an experienced neuro-radiologist in the measurement of the degree of internal carotid artery (IC) stenosis evaluated by digital angiography (DGA). METHOD: Ten symptomatic patients with internal carotid stenosis greater than 70 percent in a previous duplex scan were submitted to DGA. The degree of stenosis in both sides (symptomatic and asymptomatic) were evaluated by the same neuro-radiologist who gave his SVI and applied the NASCET method immediately after. Both methods were compared using the intraclass correlation coeficient (r) and its 95 percent confidence interval (95 percent ci). For each method, the sample (20 ICs) was also divided in surgical (stenosis between 70 and 99 percent) and non surgical ICs, using kappa concordance coeficient (k) to compare the results. RESULTS: The results comparing the 20 values obtained by each method are r = 0.90 (95 percent ci 0.77 -- 0.96). Dividing the sample in surgical and non surgical ICs, k = 0.857, p < 0.0001; sensitivity = 100 percent (39.6 percent - 100 percent); specificity = 93.8 percent (67.7 percent - 99.7 percent); positive predictive value = 80 percent (29.9 percent - 98.9 percent); negative predictive value = 100 percent (74,7 percent - 100 percent). CONCLUSION: The SVI may be used by at least some experienced neuroradiologists as a preliminary tool to evaluate the degree of IC stenosis with DGA, but a standardised and well established method should be routinely performed