Background: although the precision of duplex scanning is of utmost importan
ce in vein-graft surveillance, it has not been properly assessed. This stud
y aims to analyse interobserver agreement on duplex scanning.
Methods: a blinded comparative trial of 69 infrainguinal vein bypass recons
tructions. Two consecutive duplex scans were performed by different examine
rs and duplex ultrasound machines on the same patient. The duplex examinati
ons were also compared with angiography, when available, and clinical follo
w-up.
Results: interobserver agreement in Kappa statistics was 0.69, signifying "
good" agreement between the examinations in detecting haemodynamically sign
ificant changes in the grafts. The sensitivity, specificity and accuracy fi
gures compared with a combination of angiography and follow-up data for the
two scans were 80%, 91%, 88% and 85%, 93%, 91%, respectively. The limits o
f agreement were, however, wide for Doppler-derived velocity characteristic
s.
Conclusion: duplex scanning is an accurate and reproducible method for dete
cting haemodynamically significant changes in infrainguinal vein grafts.