Performance of 45 laboratories participating in a proficiency testing program for Lyme disease serology.

OBJECTIVE--We show that significant interlaboratory and intralaboratory variations exist in Lyme disease proficiency testing. DESIGN--Six case-defined Lyme serum samples and three serum samples from individuals with no history of Lyme disease were randomized in four shipments and distributed to 45 participating laboratories. RESULTS--Interlaboratory and intralaboratory performances were highly variable. Approximately 4% to 21% of laboratories failed to identify correctly positive serum samples with titers of 512 or more using polyvalent serum or immunoglobulin G conjugates. With lower levels of anti-Borrelia burgdorferi antibody in the serum sample, approximately 55% of participating laboratories did not identify a case-defined serum. There was also a striking inability of many laboratories to reproduce their results on split samples from the same individual. In addition, 2% to 7% of laboratories identified serum samples from individuals with no known exposure to B burgdorferi as positive using polyvalent serum. The false positivity rate increased to 27% with the use of immunoglobulin G conjugate. CONCLUSIONS--Our results indicate that there is an urgent need for standardization of current testing methodologies. Until a national commitment is made, serological testing for Lyme disease will be of questionable value for the diagnosis of the disease.