Summary
A multicenter study involving 13 laboratories was designed to compare a common procedure
for screening lupus anticoagulants (LA) to the different practices currently in use
in these laboratories. The common procedure combined 3 phospholipid- dependent assays,
including mixing studies and a phospholipid neutralizing test. Due to the heterogeneity
of LA expression, an abnormal result in at least one of the tests was sufficient to
classify a sample as positive for LA. Consecutive samples referred for LA diagnosis
were evaluated in parallel by each participant and the data found using the common
procedure were analyzed independently according to mutually agreed cut-offs and criteria
for sample classification. Within a period of 3 months, 535 samples were included,
of which 147 were judged LA positive, 29 undetermined and 359 negative by the respective
laboratories using their current practice. When using the common procedure, 149 plasmas
were said to be positive, 38 undetermined and 348 negative. Absolute concordance occurred
for 81% of the specimen population and absolute discordance (positive versus negative)
for 7%. The level of agreement between the common procedure and the current practices,
assessed by kappa indexes, indicated noticeable variations in the rates of detection
from laboratory to laboratory. Among the different tests used in the common procedure,
regular APTT was the least sensitive (about 50% detection) but none of the other tests
alone recognized more than 73% of specimens from the LA positive population. This
yield increased to about 90% with any combination of 2 sensitive tests. The common
procedure should not be considered as a methodological gold standard but as an acceptable
basis for the collaborative works required for the improvement and standardization
of laboratory practices for screening LA.
