Summary
The serial use of non-invasive tests has been shown to be a safe method of managing
outpatients who are suspected of having lower limb deep venous thrombosis (DVT). Objective
testing has shown that the majority of these outpatients do not have venous thrombosis.
A rapid test to exclude DVT in these patients, without the need for expensive and
inconvenient serial non-invasive vascular testing, would have practical and economic
advantages.
Studies measuring the fibrin degradation product D-dimer using enzyme-linked immunoassays
(EIA) in patients with veno-graphically proven DVT suggest that it should be possible
to exclude this condition by the use of one of the rapid latex bead D-dimer tests.
We have examined 190 patients with suspected DVT using both a latex and an EIA D-dimer
assay. The latex D-dimer test used in this study was negative in 7 of the 36 proven
cases of DVT. This sensitivity of only 80% is not sufficient to allow this type of
assay, in its current form, to be used as an exclusion test for DVT. The same plasma
samples were tested with an EIA assay. This information was used to mathematically
model the effects of selecting a range of D-dimer discriminant cut off points for
the diagnosis of DVT. These results indicate that 62% of suspected clinically significant
DVT could have this diagnosis excluded, with a 98% sensitivity, if the rapid latex
or equivalent D-dimer test could be reformulated to measure less than 185 ng/ml of
D-dimer.