The diagnostic accuracy of serological tests for Lyme borreliosis in Europe: a systematic review and meta-analysis

BACKGROUND: Interpretation of serological assays in Lyme borreliosis requires an understanding of the clinical indications and the limitations of the currently available tests. We therefore systematically reviewed the accuracy of serological tests for the diagnosis of Lyme borreliosis in Europe. METHODS: We searched EMBASE en MEDLINE and contacted experts. Studies evaluating the diagnostic accuracy of serological assays for Lyme borreliosis in Europe were eligible. Study selection and data-extraction were done by two authors independently. We assessed study quality using the QUADAS-2 checklist. We used a hierarchical summary ROC meta-regression method for the meta-analyses. Potential sources of heterogeneity were test-type, commercial or in-house, Ig-type, antigen type and study quality. These were added as covariates to the model, to assess their effect on test accuracy. RESULTS: Seventy-eight studies evaluating an Enzyme-Linked ImmunoSorbent assay (ELISA) or an immunoblot assay against a reference standard of clinical criteria were included. None of the studies had low risk of bias for all QUADAS-2 domains. Sensitivity was highly heterogeneous, with summary estimates: erythema migrans 50 % (95 % CI 40 % to 61 %); neuroborreliosis 77 % (95 % CI 67 % to 85 %); acrodermatitis chronica atrophicans 97 % (95 % CI 94 % to 99 %); unspecified Lyme borreliosis 73 % (95 % CI 53 % to 87 %). Specificity was around 95 % in studies with healthy controls, but around 80 % in cross-sectional studies. Two-tiered algorithms or antibody indices did not outperform single test approaches. CONCLUSIONS: The observed heterogeneity and risk of bias complicate the extrapolation of our results to clinical practice. The usefulness of the serological tests for Lyme disease depends on the pre-test probability and subsequent predictive values in the setting where the tests are being used. Future diagnostic accuracy studies should be prospectively planned cross-sectional studies, done in settings where the test will be used in practice.