A critical evaluation of enzyme immunoassays for detection of antinuclear autoantibodies of defined specificities - I. Precision, sensitivity, and specificity
Em. Tan et al., A critical evaluation of enzyme immunoassays for detection of antinuclear autoantibodies of defined specificities - I. Precision, sensitivity, and specificity, ARTH RHEUM, 42(3), 1999, pp. 455-464
Objective. To determine the performance characteristics of enzyme-based imm
unoassay (EIA) kits for the detection of antinuclear and other autoantibodi
es of defined specificities.
Methods. Nine manufacturers of EIA kits to detect antibodies of defined spe
cificities participated in a study in which they received coded sera from t
he Centers for Disease Control and Prevention. These coded sera contained d
ifferent dilutions of antibody of one specificity mixed with sera containin
g antibodies of other specificities. The manufacturers were asked to use th
eir standard technology to determine antibody content and send the data to
a committee of the International Union of Immunological Societies for analy
sis. The data were analyzed for sensitivity and specificity in the detectio
n of anti-double-stranded DNA (anti-dsDNA), anti-single-stranded DNA, antih
istone, anti-Sm, anti-U1 RNP, anti-SSA/Ro, anti-SSB/La, anti-Scl-70 (DNA to
poisomerase I), anticentromere, and anti-Jo-1 antibodies. In addition, repl
icate samples were included in the coded sera to evaluate the precision of
each EIA method.
Results. Lack of sensitivity and specificity was most evident in the anti-d
sDNA and anti-Sm kits, although 2 kits for anti-dsDNA achieved acceptable s
ensitivity and specificity. Generally, anti-SSA/Ro, anti-SSB/La, anti-Scl-7
0, anticentromere, and anti-Jo-1 kits performed well. Many false-positive r
esults were obtained with a multiple myeloma serum containing cryoprecipita
tes, but multiple myeloma sera without cryoprecipitates presented no proble
m in the EIA system. Precision, based on evaluation of replicate samples, v
aried from very good to poor.
Conclusion. No single manufacturer was clearly superior to others in terms
of their products' overall sensitivity, specificity, and precision. Areas t
hat needed improvement were in kits for the detection of antibodies to dsDN
A and to Sm antigen. Some EIA kits achieved good sensitivity and specificit
y. Individual manufacturers were informed of the performance of their respe
ctive kits so they could take measures to correct perceived deficiencies an
d thus improve the reliability of a group of important diagnostic assays us
ed in the evaluation of systemic rheumatic diseases.