Skip to main content

Table 3 Summary estimates of sensitivity and specificity for all case-definitions, derived from a hierarchical summary ROC model. The results may be different from those in the main text, as here they are specified for immunoblots and ELISAs and for commercial and in-house tests separately, while in the main text the overall estimates are provided

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

Case definition Assay Design N (studies); N(2×2 tables); N(cases); N(controls) Sensitivity (95 % CI) Specificity (95 % CI) Heterogeneity Quality and Study Design
Erythema migrans In-house ELISA Case-control, Healthy controls 6, 10, 451, 658 0•41 (0•25 to 0•60) 0•97 (0•95 to 0•98) IgG lower sensitivity than IgM. Other sources of heterogeneity were not found. Study quality did not influence the accuracy
In-house IB 3, 3, 182, 380 0•52 (0•38 to 0•65) 0•98 (0•94 to 0•99)
Commercial ELISA 13, 32, 874, 2509 0•54 (0•44 to 0•65) 0•93 (0•90 to 0•95)
Commercial IB 3, 5, 161, 289 0•58 (0•49 to 0•67) 0•86 (0•75 to 0•93)
Two-tiered tests 2, 7, 125, 190 range 0•12 to 0•64 range 0•67 to 0•96
Lyme neuroborreliosis In-house ELISA Case-control, Healthy controls 6, 9, 277, 649 0•69 (0•60 to 0•76) 0•88 (0•72 to 0•97) IgM and IgG have similar sensitivity and specificity, IgG has a higher accuracy. Recombinant tests perform best. More recent studies perform better than earlier studies. If serology was not part of the reference standard, then specificity was lower.
(serum) In-house IB 5, 8, 253, 445 0•69 (0•57 to 0•80) 0•93 (0•86 to 0•97)
Commercial ELISA 11, 28, 484, 2920 0•81 (0•70 to 0•89) 0•94 (0•91 to 0•96)
Commercial IB 2, 4, 33, 286 0•81 (0•57 to 0•94) 0•92 (0•88 to 0•95)
  Two-tiered tests 1, 5, 15, 100 range 0•41 to 0•87 range 0•88 to 0•94
(csf) Any ELISA Case-control, Cross-reacting controls 6, 9, 385, 261 0•74 (0•38 to 0•93) 0•96 (0•85 to 0•99)
(serum + csf) Specific AI test 7, 10, 458, 380 0•86 (0•63 to 0•95) 0•94 (0•85 to 0•97)
Lyme neuroborreliosis Any ELISA or IB (in serum) Cross-sectional study 6, 12, 282, 412 0•78 (0•53 to 0•92) 0•78 (0•40 to 0•95) Sensitivity similar for IgG and IgM; specifcity higher for IgG. No other sources of heterogeneity.  
Specific AI test (in serum and CSF) 4, 4, 102, 118 0•79 (0•34 to 0•97) 0•96 (0•64 to 1•00)
Lyme arthritis All ELISA Case-control, Healthy controls 8, 26, 160, 1112 Median0•96
Interquartile range 0•93 to 1•00
Median 0•94
Interquartile range 0•91 to 0•97
IgM a much lower sensitivity than IgG. No other sources of heterogeneity. Study quality did not influence the accuracy
Acrodermatitis All ELISA Case-control, Healthy controls 10, 27, 256, 1415 0•97 (0•94 to 0•99) 0•95 (0•88 to 0•98) IgM a much lower sensitivity than IgG. No other sources of heterogeneity. Study quality did not influence the accuracy
Lyme borreliosis (unspecified) In-house ELISA Case-control, Healthy controls 4, 7, 115, 215 0•85 (0•71 to 0•93) 0•98 (0•93 to 0•99) Tests assessing both IgM and IgG have highest sensitivity; specificity not very variable. Recombinant tests and more recent studies perform worse. If serology was not part of the reference standard, then accuracy was lower.
In-house IB 2, 4, 98, 126 0•63 (0•33 to 0•86) 0•97 (0•93 to 0•99)
Commercial ELISA 10, 43, 658, 815 0•70 (0•52 to 0•83) 0•95 (0•89 to 0•98)
Commercial IB 1, 4, 26, 62 0•29 (0•07 to 0•68) 0•96 (0•90 to 0•98)
Lyme borreliosis (unspecified) Any ELISA or IB Cross-sectional study 5, 14, 226, 914 0•77 (0•48 to 0•93) 0•77 (0•46 to 0•93) IgM lowest sensitivity, but highest specificity; no other sources investigated.  
  1. Number of studies is for each combination of case definition and assay category. Thus the same study may appear more than once. ELISA Enzyme Immuno Assay, IB Immunoblot, AI Antibody Index, CSF Cerebrospinal Fluid