The clinician and the vascular laboratory in the diagnosis of aortoiliac stenosis.

Two hundred and twenty-one lower limbs in 116 arteriopaths were examined for evidence of aortoiliac stenosis, by analyses of common femoral artery Doppler waveforms for pulsatility index (PI) Laplace delta, and principal components, pulse volume recordings (PVR) and thigh pressure measurements. Clinical assessment of the aortoiliac segment was also recorded in 124 limbs. The accuracy of prediction of aortoiliac stenosis, measured on biprojectional arteriograms, was calculated for the different methods, using multiple regression analysis. Considering all grades of stenosis together, clinical judgement was the best predictor of severity (R2 - the correlation coefficient squared = 0.66): addition of non-invasive tests increased the accuracy of prediction to R2 = 0.70. For stenoses greater than 50 per cent PVR amplitude was the best non-invasive test (R2 = 0.39), but was no better than clinical assessment (R2 = 0.40). However, for stenoses less than 50 per cent clinical judgement was poor (R2 = 0.13), and derived most benefit from addition of Laplace delta (R2 = 0.20) or PVR upstroke time (R2 = 0.26). In addition, common femoral PI and Laplace delta were compared in the arteriopaths, and in 146 limbs of normal control subjects, by calculation of sensitivity and specificity. PI was superior to Laplace delta for detecting severe stenoses and occlusions, but Laplace delta was more accurate in the detection of lesser degrees of disease. These results suggest that conventional clinical assessment is the best method for detecting aortoiliac stenoses greater than 50 per cent, but for less severe disease addition of vascular laboratory tests is helpful - especially PVR and Laplace waveform analysis.