One of the non-invasive techniques used in the diagnosis of patients with peripheral arterial occlusive disease is the measurement of the Ankle-Arm Systolic Pressure Ratio (“AA-Ratio”). The magnitude of different sources of measurement variation, relevant to clinical practice as well as multi-observer studies, was estimated from a random effects analysis of variance model. Thirty-five GPs and 24 practice assistants performed 503 AA-Ratio measurements on nine patients over a period of nine weeks, using pocket Doppler devices.

When the AA-Ratio is used for diagnostic purposes, a 95% prediction interval consisting of the measured value ± 15-20% AA-Ratio has to be taken into account. Furthermore, when the AA-Ratio is used in the follow-up of a patient with vascular disease, the difference between two subsequent measurements has to be at least 15.6-19.0% AA-Ratio to have an 80% certainty that this difference is not due to (intra-observer) measurement error.