Comparison of the performance and reproducibility of various serological methods and diagnostic kits for the detection of rubella antibodies.

For rubella antibody detection performance and reproducibility of standard laboratory methods and newer assay systems have been evaluated. IgM antibody detection in sucrose density gradient ultracentrifugation (SDG) was compared with two commercial ELISAs and the non-commercial anti-my-hemadsorption immunosorbent technique. The two ELISAs proved to be more sensitive than the SDG even if done with a long incubation hemagglutination inhibition test (HAI) to increase sensitivity. Additional tests showed that the sensitivity of the SDG could be increased by using ELISA instead of HAI. The reproducibility of all tests was good. In 205 assays the ELISA enzygnost IgM showed a coefficient of variation (CV) with its reference serum of 6.3% for titers, while the CV for absorbance was 37.8%. With ELISA rubazyme M the same reference serum gave a CV of 13.7% in 149 subsequent assays. In 12 and 13 assays in the anti-my-hemadsorption immunosorbent technique the CV varied with the height of the reference serum used between 4.7 and 18.7%. Antibody detection using HAI was compared with IgG antibody detection using two commercial ELISAs and a commercial single radial hemolysis test (SRH). In 592 assays with a low positive control serum the HAI gave a CV of 6.0%. With the ELISA enzygnost a CV of 7.7% for the titers was obtained in 72 assays, but for absorbances the CV was 39.4%. The CV for the kit internal low positive control in 148 assays with the ELISA rubazyme was 26.8%. As with HAI the reproducibility of the SRH was good. The same control serum used in the HAI gave a CV of 6.2% in 418 subsequent assays. Because of the good reproducibility and sensitivity the SDG can now be replaced by newer techniques which are less expensive and time consuming. Interference with rheumatoid factor was only observed in the ELISA enzygnost if sera were not pretreated with latex adsorbents. The reproducibility of our HAI was comparable to that of the SRH while the results obtained with the two commercial ELISAs were less reproducible. Until it is known, that antibodies detected with the newer techniques are able to prevent re-infection, it would be unwise to reject the HAI completely for the determination of immunity status.