Performance characteristics of the Cozart RapiScan Oral Fluid Drug Testing System for opiates in comparison to ELISA and GC/MS following controlled codeine administration.

Sherri L Kacinko, Allan J Barnes, Insook Kim, Eric T Moolchan, Lisa Wilson, Gail A Cooper, Claire Reid, Dene Baldwin, Chris W Hand, Marilyn A Huestis

Index: Forensic Sci. Int. 141(1) , 41-8, (2004)

Full Text: HTML

Abstract

Oral fluid is an interesting alternative matrix for drug testing in many environments, including law enforcement, workplace drug testing, and drug treatment facilities. Performance characteristics of the FDA-cleared, qualitative, Cozart RapiScan Opiate Oral Fluid Drug Testing System (Opiate Cozart RapiScan System or Opiate CRS) were compared to the semi-quantitative Cozart Microplate EIA Opiate Oral Fluid Kit (Opiate ELISA) and to gas chromatography/mass spectrometry (GC/MS). The following oral fluid opiate cutoffs were evaluated: the GC/MS limit of quantification (LOQ) of 2.5 mg/l; 15 microg/l currently used for oral fluid testing in the United Kingdom (UK); 30 microg/l (Opiate CRS cutoff); and 40 microg/l, the proposed Substance Abuse and Mental Health Services Administration (SAMHSA) cutoff. Subjects provided informed consent to participate in this IRB-approved research and resided on the closed research ward throughout the study. Three oral codeine doses of 60 mg/70 kg were administered over a 7-day period. After a 3-week break, subjects received three doses of 120 mg/70 kg within 7 days. Oral fluid specimens (N = 1273) were analyzed for codeine (COD), norcodeine (NCOD), morphine (MOR) and normorphine (NMOR) by GC/MS with an LOQ of 2.5 microg/l for all analytes. MOR and NMOR were not detected in any sample; 26.5% of the specimens were positive for COD and 13.7% for NCOD. Opiate CRS uses a preset, qualitative cutoff of 10 microg/l; this is equivalent to 30 microg/l in undiluted oral fluid as the oral fluid collection process involves a 1:3 dilution with buffer. Sensitivity, specificity, and efficiency of Opiate CRS compared to Opiate ELISA were 98.6, 98.1, and 98.2% at a 30 microg/l cutoff and 99.0, 96.2, and 96.6% at a 40 microg/l cutoff. Compared to the much lower GC/MS LOQ of 2.5 microg/l, sensitivity, specificity and efficiency were 66.8, 99.3 and 90.7%. Increasing the GC/MS cutoff to the current UK level yielded performance characteristics of 81.5% (sensitivity), 99.3% (specificity), and 95.4% (efficiency). Using a GC/MS cutoff identical to the preset Opiate CRS cutoff yielded sensitivity, specificity, and efficiency of 88.5, 99.2, and 97.5%, respectively. At the proposed SAMSHA confirmation cutoff of 40 microg/l, sensitivity increased with little change in specificity and efficiency (91.3% sensitivity, 98.9% specificity, and 97.5% efficiency). Oral fluid is a suitable matrix for detecting drugs of abuse. Opiate CRS, with a 30 microg/l cutoff, is sufficiently sensitive, specific and efficient for oral fluid opiate analysis, performing similarly to Opiate ELISA at the same cutoff, and having performance characteristics >91% when compared to GC/MS at the proposed SAMHSA cutoff.