Current status of rubella testing: a report based on data from the College of American Pathologists' surveys, 1978-1980.

Data from national survey programs were used to define selected attributes of popular rubella methods. Of those participating in the College of American Pathologists' survey, 47% used passive hemagglutination (PHA); 45% used hemagglutination-inhibition (HI), 7% used indirect immunofluorescence, and less than 1% used enzyme immunosorbent assays for rubella testing. Enzyme assays were not evaluated. The other methods had sensitivities and specificities exceeding 94% and 88%, respectively. The predictive value of a negative test was used to identify differences among the various methods. The HI tests using human O cells had a 90-95% prediction rate for a true-negative result. Indirect immunofluorescence, PHA, and HI using kaolin had the lowest predictive values (68-72%) values for other methods were intermediate. Levels of reactivity also differed. The HI methods using human O cells yielded higher titers than did the reference method (1- to 3-day-old chick/heparin MnCl2), while the titers for HI methods using kaolin tended to be lower.