Prevalence of Detectable Biotin in Five US Emergency Department Patient Cohorts

Background: The objective of this study was to estimate the prevalence of biotin supplementation in United States emergency department patients using a multi-site, geographically distributed sampling model. Methods: Biotin was measured using an Abbott ARCHITECT Biotin research use only assay in 7118 emergency department patient serum or plasma samples from five US medical centers. Samples with biotin >= 10 ng/mL underwent additional LC-MS/MS confirmatory testing for biotin and its primary metabolites. The overall and site-specific prevalence of detectable biotin was determined using the screening assay while biotin speciation (i.e., prevalence of detectable metabolites) was determined using LC-MS/MS. Results: Of 7118 samples screened, 291 (4.1%) had biotin >= 10 ng/mL and were considered positive. Across five medical centers, the fraction of positive samples ranged from 2.0% to 5.4%. The maximum biotin concentration observed was 355 ng/mL. Of the 285 positive screens that underwent additional LC-MS/MS testing, 89 (31%) showed detectable biotin, bisnorbiotin, and/or biotin sulfoxide. Biotin, bisnorbiotin, and biotinsulfoxide were detected in 82/89 (92.1%), 61/89 (68.5%), and 18/89 (20.2%) samples, respectively; biotin was detected in the absence of either metabolite in 18/89 (20.2%) samples. Conclusions: Using a screening assay, 4.1% of emergency department patient samples were found to be potentially susceptible to interference from biotin. Confirmatory testing showed detectable biotin and/or biotin metabolites in 31% of positive screens (1.3% overall). The prevalence of biotin >= 10 ng/mL varied 2-3-fold across US emergency department patient cohorts. Biotin metabolites were observed in 80% of samples confirmed to have detectable biotin species by LC-MS/MS, suggesting that rigorous assessments of assay susceptibility to biotin interference, often performed using in vitro studies, should consider the potential role of biotin metabolites present in vivo.