Purpose: The venous filling index (VFI) was introduced as a noninvasiv
e measure of venous valvular reflux. Because it is quantitative, ident
ifying increasingly severe reflux should be possible. The purpose of t
his study is to evaluate the sensitivity and predictive value of the V
FI as a predictor of phlebographically demonstrated ''critical'' venou
s reflux. Methods: Thirty-one limbs with suspected venous insufficienc
y underwent both descending phlebography and air plethysmography. Nine
Limbs had deep venous reflux ending above the knee on descending phle
bography (group 1), and 22 limbs had ''critical'' deep reflux to the b
elow-knee level (group 2). The VFI, ejection fraction, and residual vo
lume fraction were calculated, and a tourniquet was used to distinguis
h superficial from deep venous reflux. Results: Three of six group 2 l
imbs with a VFI less than 7 had obliteration of their iliac veins. Tou
rniquet application improved the VFI in 13 limbs, all of which showed
either phlebographic or venous duplex evidence of greater saphenous in
competence. Conclusions: A VFI greater than 7 showed a 73% sensitivity
and 100% positive predictive value of identifying ''critical venous r
eflux.'' The VFI may underestimate the degree of reflux in patients wi
th proximal venous obstruction. Improvement of venous hemodynamics wit
h tourniquet application suggests that venous reflux might be improved
in some patients by correcting superficial venous insufficiency.