Modelling the probability and impact of false-positive serology for Borrelia burgdorferi sensu lato: A case study.
Authors: Houben Rosa M A C, Meersschaert Carole, Hendrickx Guy, Pitel Pierre-Hugues, Amory Hélène
Journal: Equine veterinary journal
Summary
# Editorial Summary: False-positive Lyme serology in horses—a Belgian perspective on diagnostic pitfalls Lyme borreliosis screening in horses relies heavily on serological testing, yet the diagnostic accuracy of these tests remains poorly understood in clinical practice. Houben and colleagues conducted a cross-sectional survey of 303 horses in southern Belgium, using ELISA to detect Borrelia burgdorferi antibodies and applying Bayesian modelling to distinguish true infection from false positives—a critical distinction that standard serology alone cannot make. Whilst the apparent seroprevalence was 22%, the true seroprevalence was estimated at only 11%, meaning approximately two-thirds of positive results were false positives and roughly one in five tested horses risked unnecessary antimicrobial treatment. The researchers calculated that around 5% of all antimicrobial use in Belgian equine practice may stem from treating horses that are actually seronegative, highlighting a substantial yet largely unrecognised source of inappropriate therapy. For practitioners, this work underscores the importance of contextualising ELISA results within the actual prevalence of infection in your region and reconsidering treatment decisions based on positive serology alone—particularly when clinical signs are absent or equivocal.
Read the full abstract on PubMed
Practical Takeaways
- •ELISA screening for Lyme borreliosis has poor diagnostic value; positive results should be interpreted cautiously and clinical signs considered before treatment, as approximately 67% of positive tests are false positives
- •Two out of every ten horses testing positive may not actually be infected; consider confirmatory testing or clinical correlation before starting antimicrobial therapy to avoid unnecessary treatment
- •Be aware that relying on serology alone for Lyme borreliosis diagnosis contributes significantly to antimicrobial overuse in equine practice; use serological results as one component of diagnosis rather than a definitive test
Key Findings
- •Apparent seroprevalence was 22% (95% CI 18-27%) but true seroprevalence was only 11% (95% credible interval 0.6-21%) in southern Belgian horses
- •Approximately two-thirds of positive ELISA samples are estimated to be false positives
- •One in five horses tested were potentially misdiagnosed as infected with Lyme borreliosis
- •Around 5% of antimicrobial use in Belgian equine practice may be attributed to treatment of false-positive serology results