Canine parvovirus serology: a collaborative assay.

Fifteen laboratories were supplied with coded samples of canine sera for testing for the presence of antibodies against canine parvoviruses. One of these sera had been designated as a potential British standard canine parvovirus antiserum. Most of these laboratories were either providing a canine parvovirus serology service, or represented pharmaceutical companies which manufacture canine parvovirus vaccines for the United Kingdom market. No attempt was made to influence the test methods used. Thirteen of the laboratories used a haemagglutination inhibition test, three an enzyme-linked immunosorbent assay (ELISA), and two performed serum neutralisation tests. Three laboratories used two different techniques. Adequate analysis was possible only with the results of the haemagglutination inhibition tests. The variability of the results between laboratories could be partly controlled by the use of the standard serum. Much of the residual variability was associated with particular laboratories. The results from the vaccine manufacturers tended to be less variable than those from the diagnostic laboratories.