1 Laboratoire de sciences judiciaires et de médecine légale, Department of Toxicology, 1701 Parthenais Street, Montréal, Québec, H2K 3S7, Canada; Concordia University, Department of Chemistry & Biochemistry, 7141 Sherbrooke Street West, Montréal, Québec, H4B 1R6, Canada. Electronic address: brigitte.desharnais@msp.gouv.qc.ca.
2 Laboratoire de sciences judiciaires et de médecine légale, Department of Toxicology, 1701 Parthenais Street, Montréal, Québec, H2K 3S7, Canada.
3 Concordia University, Department of Chemistry & Biochemistry, 7141 Sherbrooke Street West, Montréal, Québec, H4B 1R6, Canada.

A study of impaired driving rates in the province of Québec is currently planned following the legalization of recreational cannabis in Canada. Oral fluid (OF) samples are to be collected with a Quantisal® device and sent to the laboratory for analysis. In order to prepare for this project, a qualitative decision point analysis method monitoring for the presence of 97 drugs and metabolites in OF was developed and validated. This high throughput method uses incubation with a precipitation solvent (acetone:acetonitrile 30:70 v:v) to boost drug recovery from the collecting device and improve stability of benzodiazepines (e.g., a-hydroxyalprazolam, clonazepam, 7-aminoclonazepam, flunitrazepam, 7-aminoflunitrazepam, N-desmethylflunitrazepam, nitrazepam). The Quantisal® device has polyglycol in its stabilizing buffer, but timed use of the mass spectrometer waste valve proved sufficient to avoid the glycol interferences for nearly all analytes. Interferences from OF matrices and 140 potentially interfering compounds, carryover, ion ratios, stability, recovery, reproducibility, robustness, false positive rate, false negative rate, selectivity, sensitivity and reliability rates were tested in the validation process. Five of the targeted analytes (olanzapine, oxazepam, 7-aminoclonazepam, flunitrazepam and nitrazepam) did not meet the set validation criteria but will be monitored for identification purposes (no comparison to a cut-off level). Blind internal proficiency testing was performed, where six OF samples were tested and analytes were classified as "negative", "likely positive" or "positive" with success. The final validated OF qualitative decision point method covers 92 analytes, and the presence of 5 additional analytes is screened in this high throughput analysis.