Radioimmunoassay (RIA) tests for anti-hepatitis A virus (HAV) IgM were carried out on 728 sera: 382 were tested by both a method using an anti-μ serum bound to a solid phase and a method involving preliminary separation of IgM by sucrose density gradient (SDG) centrifugation, 354 by the anti-μ method alone and two by the SDG method alone. Similar proportions of sera were found to be positive by each method (42·5%, 41·7%), but equivocal results were commoner by the SDG method (4·7% compared with 1·5%). There were 21 (5·5%) discrepant results from the sera tested by both methods, 20 of which could have been due to the higher sensitivity of the anti-μ method.
The SDG method generally gave unequivocal results on sera collected within six weeks of the onset of jaundice. Separation of the IgM fraction by re-orientation centrifugation was quick, but otherwise offered no special advantage over separation on a swing-out rotor. The use of 2 mercaptoethanol (2 ME) reduction to assess the purity of the IgM fraction increased confidence in the specificity of the test. It led, however, to the exclusion of 16 reactive sera (4·2%), all of which were found to be positive in the anti-μ test.
The anti-μ method gave better discrimination between positive and negative sera than the SDG method and detected IgM both earlier and later in infection. The results of tests designed to check the specificity of the anti-μ procedure were satisfactory. As it is potentially cheaper and easier to perform, the anti-μ method seems, in all respects, to be superior to the SDG method.