COMPARISON OF THE TOE-BRACHqIAL INDEX WITH THE ANKLE-BRACHIAL INDEX. IS THERE A PSEUDO-NORMAL ANKLE-BRACHIAL INDEX?
Author Block: M Matangi, A Thakrar, DW Armstrong, C Tobin, D Brouillard
PURPOSE: The current recommendation is to perform the toe-brachial index (TBI) in elderly patients (P), or in P with diabetes where the ankle-brachial index (ABI) is normal or high but there are symptoms suspicious for PAD. The purpose of our investigation was to compare the toe-brachial index with the ankle-brachial index in a large unselected cohort of P referred for peripheral arterial testing.
METHODS: PADfile, the peripheral arterial disease module of CARDIOfile, our cardiology database was searched for all P who had peripheral arterial testing. Only P who had both the ABI and TBI were included. An abnormal ABI was defined as <0.91 and an abnormal TBI <0.72. The sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV) and accuracy were calculated using the ABI as the gold standard. We also looked at the prevalence of common risk factors between groups.
RESULTS: 2,316 lower limbs had both their ABI and TBI recorded. The correlation coefficient between the TBI and ABI was 0.64. The major findings are found in Table 1 and Image 1.
CONCLUSIONS: The accuracy of the TBI is relatively good. There is a very low false negative rate. By comparison, the false positive rate is much higher. What is of interest is that the false positive group has the highest average age and the second highest rates of smoking and diabetes, the smoking and diabetic rates only exceeded by the true positive group. This brings up the possibility that in the false positive group some of the apparently normal ABIs with an abnormal TBI may in fact be pseudo-normal ABIs due to incompressibility secondary to peripheral arterial calcification, the prevalence of which is known to be higher in the elderly and diabetics.
|.||ABI <0.91||ABI >0.90||.|
|TBI <0.72||352||388||PPV =47.6%|
|TBI >0.71||36||1542||NPV =97.7%|
|.||Sensitivity =90.7%||Specificity =79.9%||Accuracy =81.8%|
Author Disclosure Information: M. Matangi, None; A. Thakrar, None; D.W. Armstrong, None; C. Tobin, None; D. Brouillard, None.