"DrugTest 5000 screening results were evaluated against Quantisal confirmation data to determine TP [True Positive], TN [True Negative], FP [False Positive], FN [False Negative], diagnostic sensitivity and specificity, and efficiency at various cutoffs (Tables 1 and 2). When compared to THC alone, the diagnostic sensitivity and specificity and efficiency were 86.2%–90.7%, 75.0%–77.8%, and 84.8%– 87.9% at the 5-μg/L cutoff and 75.9%–92.7%, 76.0%–100.0%, and 78.8%– 86.4% at the 10-μg/L DrugTest 5000 cutoffs. Overall, the DrugTest 5000 performed better with the 5-μg/L screening cutoff, with diagnostic sensitivity and efficiency above the DRUID-recommended 80%. There were few FP and FN tests, and when they occurred, concentrations were at or near the confirmation cut-off. A limitation of this study was the inclusion of a small number of TN samples, only 6 –12 with the 5-μg/L DrugTest 5000 and 1- and 2-μg/L confirmation cutoffs, to adequately evaluate diagnostic specificity. On the basis of previous reports, more TN samples were expected over the 22-h collection period. Detection rates were highest and windows of detection were longest when we confirmed for THC alone (Fig. 1 and 2). However, the recent report of THC concentrations in OF following 3 h of passive exposure to cannabis smoke advocate for the inclusion of THCCOOH in confirmation criteria, because this analyte is not present in cannabis smoke and was not found in any OF [Oral Fluid] samples following passive exposure (18)."

Source

Nathalie A. Desrosiers, et al., "On-Site Test for Cannabinoids in Oral Fluid," Clinical Chemistry, Oct. 2012, 58(10):1418-25.
http://www.ncbi.nlm.nih.gov/p…
http://www.clinchem.org/conte…