Proficiency testing of selected antigen and antibody tests for use in dogs and cats.

Abstract

OBJECTIVE: To determine the correlation of seroimmunologic test results between reference and nonreference laboratories. DESIGN: Retrospective data analysis. PROCEDURE: Serum samples obtained from naturally infected dogs and cats were distributed to reference and nonreference laboratories for seroimmunologic testing. Correlation of test results was evaluated by use of nonparametric analysis. RESULTS: Correlation coefficients were high between laboratory groups for samples tested for feline immunodeficiency virus antibodies, FeLV antigen, and toxoplasmosis antibodies in cats. Results for feline immunodeficiency virus antibody tests from reference laboratories were more likely to be positive than results from nonreference laboratories. Test results for feline infectious peritonitis antibodies, antinuclear antibodies, and Borrelia antibodies in cats were not significant. Coefficient correlations were significant for results of heart-worm antigen, Brucella antibodies, Toxoplasma antibodies, antinuclear antibodies, and rheumatoid factor in dogs. Results for Borrelia antibodies were not correlated between laboratory groups. CLINICAL IMPLICATIONS: Results were highly correlated between reference and nonreference laboratories for 8 of 14 seroimmunologic tests. Seroimmunologic tests for use in cats were less correlated as a group than those for use in dogs. Poor correlation of results between laboratories was attributed to variations in control agents, antigens, reagents, technical expertise, and cutoff values and end-point titers used for diagnosis.