Abstract Urine drug screen (UDS) immunoassays remain essential tools in clinical toxicology due to their rapid turnaround time and operational efficiency. However, their susceptibility to cross-reactivity continues to pose challenges, particularly in complex patient populations receiving multiple medications. Verification studies performed prior to assay implementation often rely on limited sample sets and may not fully capture real-world assay performance. Ongoing post-implementation monitoring, especially through comparison with confirmatory assays such as liquid chromatography–tandem mass spectrometry (LC-MS/MS), is therefore critical for identifying unexpected assay behavior and maintaining analytical reliability. This study evaluated the real-world concordance of two methadone immunoassays—the Architect MULTIGENT Methadone assay and the Roche Methadone II (MDN2) assay—with an in-house LC-MS/MS method capable of detecting methadone and its primary metabolite, 2-ethylidene-1, 5-dimethyl-3, 3-diphenylpyrrolidine (EDDP), at a cutoff concentration of 100 ng/mL. A total of 6,753 Architect and 6,255 Roche methadone screening results were retrospectively reviewed. The Architect assay demonstrated a 2.8% positivity rate, with 94.8% of positive screens confirmed by LC-MS/MS, corresponding to a 5.2% false-positive rate. In contrast, the Roche assay produced a higher positivity rate of 3.6% but a substantially lower confirmation rate of 69.8%, yielding a 30.2% false-positive rate (χ2 = 106, P .05). Medication review of Roche false-positive cases revealed frequent use of quetiapine and vortioxetine, consistent with previously reported cross-reactivity. Additional medications historically associated with methadone immunoassay interference were also identified. These findings highlight significant performance differences between two widely used methadone immunoassays and underscore the importance of continuous, data-driven post-implementation surveillance. Confirmatory testing data such as LC-MS/MS results, can reveal clinically meaningful assay limitations, support accurate interpretation of unexpected results, and ultimately improve the reliability of UDS testing in diverse patient populations.
Jayson V. Pagaduan (Tue,) studied this question.