1. 2014;17;359-366KIMS, CEDIA, and HS-CEDIA Immunoassays Are Inadequately Sensitive for Detection of Benzodiazepines in Urine from Patients Treated for Chronic Pain
   Diagnostic Accuracy Report
   Alicia Darragh, MS, Marion L Snyder, PhD, Adam S Ptolemy, PhD, and Stacy Melanson, MD, PhD.

BACKGROUND: Patients treated for chronic pain may frequently undergo urine drug testing to monitor medication compliance and detect undisclosed prescribed or illicit drug use. Due to the increasing use and abuse of benzodiazepines, this class of medications is often included in drug screening panels. However, immunoassay-based methods lack the requisite sensitivity for detecting benzodiazepine use in this population primarily due to their poor cross-reactivity with several major urinary benzodiazepine metabolites. A High Sensitivity Cloned Enzyme Donor Immunoassay (HS-CEDIA), in which beta-glucuronidase is added to the reagent, has been shown to perform better than traditional assays, but its performance in patients treated for chronic pain is not well characterized.

OBJECTIVES: To determine the diagnostic accuracy of HS-CEDIA, as compared to the Cloned Enzyme Donor Immunoassay (CEDIA) and Kinetic Interaction of Microparticles in Solution (KIMS) screening immunoassays and liquid chromatography-tandem mass spectrometry (LC-MS/MS), for monitoring benzodiazepine use in patients treated for chronic pain.

STUDY DESIGN: A study of the diagnostic accuracy of urine benzodiazepine immunoassays.

SETTING: The study was conducted at an academic tertiary care hospital with a clinical laboratory that performs urine drug testing for monitoring medication compliance in pain management.

METHODS: A total of 299 urine specimens from patients treated for chronic pain were screened for the presence of benzodiazepines using the HS-CEDIA, CEDIA, and KIMS assays. The sensitivity and specificity of the screening assays were determined using the LC-MS/MS results as the reference method.

RESULTS: Of the 299 urine specimens tested, 141 (47%) confirmed positive for one or more of the benzodiazepines/metabolites by LC-MS/MS. All 3 screens were 100% specific with no false-positive results. The CEDIA and KIMS sensitivities were 55% (78/141) and 47% (66/141), respectively. Despite the relatively higher sensitivity of the HS-CEDIA screening assay (78%; 110/141), primarily due to increased detection of lorazepam, it still missed 22% (31/141) of benzodiazepine-positive urines. The KIMS, CEDIA, and HS-CEDIA assays yielded accuracies of 75%, 79%, and 90%, respectively, in comparison with LC-MS/MS.

LIMITATIONS: This study was limited by its single-site location and the modest size of the urine samples utilized.

CONCLUSIONS: While the HS-CEDIA provides higher sensitivity than the KIMS and CEDIA assays, it still missed an unacceptably high percentage of benzodiazepine-positive samples from patients treated for chronic pain. LC-MS/MS quantification with enzymatic sample pretreatment offers superior sensitivity and specificity for monitoring benzodiazepines in patients treated for chronic pain.