Evidence-based diagnosis of Lyme disease

The aim of this study was to make an evidence-based comparison of four commercial enzyme immunoassays (EIAs) (Serion Classics, Sigma Diagnostics, Cambridge Biotech and ICN Diagnostics) and an in-house enzyme immunoassay (EIA) in order to select the most appropriate screening assay for diagnosis of Lyme disease. Borrelia burgdorferi sensu stricto cultured in BSK-H medium was used to develop the in-house assay. Escherichia coli antigen (0.9 mg/ml) was included in the serum diluent to reduce non-specific background. Comparison of the number of tests needed to diagnose (i.e. to indicate a positive result) and the cost per positive diagnosis for the five assays was made using a panel of 176 Western blot-characterised sera. The Cambridge Biotech and Sigma assays had the highest sensitivity but poorer specificity, whereas the Serion and ICN assays had highest specificity but poorer sensitivity. The in-house assay had average sensitivity and specificity, the number of tests needed to diagnose being 2.32 compared to 1.92 for Serion, 2.17 for ICN, 2.5 for Sigma and 2.7 for Cambridge Biotech. In a diagnostic protocol that uses an EIA as screening test, with confirmation by Western blot, a good balance of sensitivity and specificity is essential. The in-house assay was the most cost-effective (lowest cost per positive diagnosis), and is probably the best option for specialist laboratories in Europe.