[Inter-observer variability in the interpretation of arteriographies of the carotid arteries and lower limbs].

This study attempts to measure inter-observer variability in reading 14 carotid angiograms and 8 aortograms. 24 surgeons and radiographs with training in arterial diseases participated in the study from different hospitals. The results were analysed with the Kappa statistical method which incorporate a correction for the extend of agreement expected by chance alone, in the evaluation of the reliability of diagnostic procedures. From a topographic point of view, the interpretation of the proximal arterial segments is more reliable than the distal segments, with the exception of the aorto-iliac bifurcation and the profunda femoral arteries. Inter-observer agreement is the most important for the occlusive lesions and the identification of stenosis is of poor reliability.