THE SENSITIVITY, SPECIFICITY, PREDICTIVE VALUES AND ACCURACY OF THE ANKLE-BRACHIAL INDEX AS A SCREENING TOOL FOR GENERALIZED ATHEROSCLEROSIS
Author Block: M Matangi, D Armstrong, A Johri, D Brouillard
INTRODUCTION: It is our impression that when comparing the usefulness of the common carotid intimal medial thickness (IMT) to the ankle-brachial index (ABI) that the IMT is a far more useful investigation. The purpose of our analysis was to determine the clinical usefulness of the ABI to detect clinically important atherosclerosis as determined by common carotid IMT measurement.
METHODS: 213 patients referred to our outpatient facility for assessment of cardiac symptoms were then referred for carotid duplex imaging and peripheral physiologic leg testing. Carotid studies were interpreted using standardized Intersocietal Commission for Accreditation of Vascular Laboratories, velocity and imaging criteria. We defined an abnormal ABI as 0.99mm. An ABI of >0.94 was considered normal and a normal IMT as <1.00mm. All data were entered into CARDIOfile, a cardiology database, for later retrieval and analysis. Statistical calculations were based on the common carotid IMT being the gold standard for the presence or absence of atherosclerosis.
RESULTS: Table 1 shows the data results and the statistical calculations.
CONCLUSIONS: The ABI has a low sensitivity (25.4%), low negative predictive value (23.2%) and low overall accuracy (39.0%) when compared to the common carotid IMT for the detection of clinically important atherosclerosis. The specificity of the ABI is high (97.5%) as is the positive predictive value (97.8%). Clinically, the ABI is really only useful when abnormal. A normal ABI should not dissuade one for looking for evidence of atherosclerosis in other areas.
|IMT >0.99||IMT <1.00||Total|
|ABI <0.95||44||1||45||PPV 97.8%|
|ABI >0.94||129||39||168||NPV 23.2%|
|Sensitivity 25.4%||Specificity 97.5%|