We performed a retrospective study on patients who had a positive screening antibody test result for antibody to Borrelia burgdorferi to determine the clinical indicators used by physicians to order this test. Eighty-two evaluable patients who were screen positive (indirect enzyme-linked immunosorbent assay) between August 1991 and March 1993 were included. Additional tests, isotype-specific capture immunoglobulin enzyme immunoassay and Western blot (immunoblot) analysis (immunoglobulin G), were performed on positive samples. Of 82 patients with a positive screening test result, 54 (66%) had no serologic evidence of
Lyme disease on the basis of additional testing (positive predictive value, 34%). Only 28 of 82 patients (34%) had clinical indicators suggestive of
Lyme disease. Antibody screening tests may provide misleading information if they are not accompanied by more specific assays. Inappropriate testing of patients without indications of
Lyme disease is frequently performed, and the ordering practices of physicians should be reassessed.