Abstract
PURPOSE: The diagnosis of Lyme borreliosis (LB) can be challenging. The aim of this study was to investigate, describe and compare the actual use, application and documentation of LB serology in three neighbouring LB-endemic counties in Sweden. As part of this, we intended to study the concordance between laboratory reports and physicians' assessments regarding LB. METHODS: Three hundred patients sampled for LB serology in the counties of Jönköping, Kalmar and Östergötland, in 2016 were randomly selected for this study. Data was collected from the laboratory information technology systems of the departments of Clinical Microbiology in the three counties and from medical records. RESULTS: Suspected Lyme neuroborreliosis (LNB) was the most common indication for LB serology, and was found in a total of 188/300 (63%) patients: 75/100 in Jönköping, 66/100 in Östergötland and 47/100 in Kalmar. Cerebrospinal fluid examination was performed on a minority of patients in whom LNB was suspected, 34/188 (18%). LB serology was performed on sera from 15 patients with suspected erythema migrans. Sufficient information to enable an assessment of concordance between laboratory reports and medical records was available for 158/300 (53%) patients, while 94/158 (59%) were considered to have concordant records. CONCLUSIONS: LB serology is frequently performed on questionable indications contrary to guidelines, which limits the value and potential of the analysis. Notably, the use appears to be different in three neighbouring counties that follow the same national guidelines. Although new diagnostic technologies, may improve laboratory diagnostics in the future, there is still a need for interventions to enable a more rational use of LB serology.