Abstract
OBJECTIVE: To compare test characteristics of the Farr radioimmunoassay and an automated fluorescence immunoassay (ELIA dsDNA test) for the diagnosis of systemic lupus erythematosus (SLE). METHODS: A cross sectional study comprising 440 samples from 440 patients, sent to the laboratory over a three month period for anti-dsDNA testing. Chart review was performed, blinded for test results, to count for each patient the number of American College of Rheumatology criteria for the classification of SLE that were fulfilled. At least four criteria were met by 248 (56%) patients (SLE), one to three criteria by 77 (18%) (lupus-like disease, LLD), and no criterion by 115 (26%) (non-SLE/non-LLD). Results from serum samples from the non-SLE/non-LLD and SLE groups were used to calculate receiver operating characteristic curves. RESULTS: For the Farr assay, specificities of 95% and 99% corresponded to sensitivities of 72% and 56% respectively. For the ELIA dsDNA test these levels of specificity corresponded to sensitivities of 44% and 17% respectively. CONCLUSIONS: The Farr radioimmunoassay is superior to the ELIA dsDNA test for identifying patients with SLE.