Underrecognized Substance Exposure in Pediatric Emergency Departments: Insights from Universal LC–MS/MS Confirmation

Background Substance exposure among pediatric emergency department (ED) patients may be underrecognized because routine urine immunoassays have limited analytical sensitivity, and objective confirmatory toxicological data in pediatric ED populations remain limited.Methods This dual-phase observational study included patients aged 6–18 years who underwent toxicological assessment in two tertiary pediatric EDs in Central Anatolia, Türkiye. The retrospective phase (2018–2021) tested each sample by either immunoassay or LC–MS/MS, depending on time of collection, but not both. The prospective phase (January 2022–2023) enrolled clinically suspected cases, all tested by both methods regardless of screening result. Logistic regression evaluated factors associated with liquid chromatography-tandem mass spectrometry (LC–MS/MS)-confirmed positivity in the prospective cohort. Results: A total of 1039 patients were included (852 retrospective; 187 prospective). In the retrospective cohort, immunoassay screening (n = 570) identified positivity in 19.8% (113/570), while LC–MS/MS on a separate subset (n = 282) identified 32.6% (92/282); these derive from non-overlapping subsets, not paired testing. In the prospective cohort, LC–MS/MS positivity was 11.2% (21/187) by standard cutoff and 18.8% (35/187) with an exploratory limit of quantitation (LOQ)-supported benzodiazepine interpretation (after excluding 18 pseudoephedrine/ephedrine-only patients). Benzodiazepines and amphetamine-type stimulants were most frequently detected. Younger age, psychiatric illness, and daily screen exposure >3 h were independently associated with positivity. Immunoassay screening showed high specificity but limited sensitivity against LC–MS/MS. Conclusions: Routine immunoassay screening may substantially underestimate pediatric substance exposure in ED settings. LC–MS/MS provides broader and more sensitive toxicological detection, though differences also partly reflect broader analyte coverage rather than sensitivity alone.