In recent years, the interest in the use of oral fluid as biological matrix has increased significantly, particularly for detecting driving under the influence of drugs (DUID). In this study, the relationship between the oral fluid and blood concentrations of drugs of abuse in drivers suspected of DUID is Nevertheless the data reflect the variability of the OF/B ratios in drivers under the influence of drugs.The wide range of the ratios, however, does not allow reliable calculation of the blood concentrations from oral fluid concentrations.

The authors present a global overview on the issue of drugs and driving covering four major areas: (1) Epidemiology and Prevalence--which reviews epidemiological research, summarizes available information, discusses the methodological shortcomings of extant studies, and makes recommendations for future research to better define prevalence and epidemiology; (2) Effects of Medicinal and Illegal Drugs on Driving Performance--focuses on the six classes of drugs most often found in impaired and injured drivers, draws conclusions regarding the risk of these drugs to traffic safety and discusses the need for additional research; (3) Toxicological Issues--discusses ways to identify drug users via behavioral testing and analytical techniques, reviews the approaches used by different countries, screening and confirmation techniques, alternative specimens (e.g., urine, oral fluid, sweat), and how rapid roadside testing could be coupled with behavioral and laboratory testing in an effective approach to identifying and prosecuting drugged drivers; (4) Driving Under the Influence of Drugs [DUID] Laws--provides an overview of DUID laws in the United States and Europe, discusses the basic tenets of these laws, the various types of DUID statutes, the reasons why many existing laws hinder the prosecution of drugged drivers and the rationale for developing per se legislation as a strategy to more effectively manage the drugged driver problem.

Blood, urine, oral fluid (by spitting or with a Salivette®), and sweat samples (by wiping the forehead with a fleece moistened with isopropanol) were obtained from 180 drivers who failed the field sobriety tests at police roadblocks. With quantitative GC-MS, the positive predictive value of oral fluid was 98, 92, and 90% for amphetamines, cocaine, and cannabis respectively. The prevalence of opiate positives was low. The SAMHSA cut-off values for oral fluid testing at the workplace, proved their usefulness in this study. The positive predictive value of sweat wipe analysis with GC-MS was over 90% for cocaine and amphetamines and 80% for cannabis. The accuracy of Drugwipe® was assessed by comparing the electronic read-out values obtained on-site after wiping the tongue and the forehead, with the corresponding GC-MS results in plasma, oral fluid, and sweat. The accuracy was always less than 90% except for the amphetamine-group in sweat.

Estimating the detection time of a drug in urine is complex because of many different influencing factors and the lack of experimental data. Detection times vary depending on dose and route of administration, metabolism and characteristics of the screening and confirmation assays. Using a cut-off value of 1000 ng/mL, urinary samples can be positive for amphetamine for up to 5 days after intake of the drug. At the lower 300 ng/mL cut-off, amphetamine will be detectable one day longer. Very few data are available for designer amphetamines. After smoking one marijuana cigarette, THCCOOH (9-carboxy-delta 9 tetrahydrocannabinol) is detectable (using a screening cut-off of 50 ng/mL) for 2-4 days. More frequent use will be detectable for almost 1 month, exceptionally 3 months. Immunoassays to detect cocaine are targeted against the metabolite benzoylecgonine and use a cut-off of 300 ng/mL. An intravenous dose of 20 mg cocaine can be detected for 1.5 days. Street doses (administered via different routes) are detectable up to 1 week, and extremely high doses up to 3 weeks. Heroin rapidly metabolizes to 6-acetylmorphine and morphine. Immunoassays for heroin are calibrated with morphine but important cross-reactivity occurs and positive results must be confirmed by GC-MS. Experimental data for total morphine using a cut-off of 300 ng/mL suggest a detection time of 1 to 1.5 days for relatively low doses of heroin (3-12 mg) administered via i.v., IN or i.m. route.

Although administered as a short-acting hypnotic for sleeping disorders, flunitrazepam, often in combination with alcohol or other drugs, was one of the most frequently abused benzodiazepines over the last 10 years. It has been reported in cases of driving under the influence, and its use is associated with marked psychomotor impairment. Studies over the last five years have investigated the use of oral fluid as an alternative matrix to blood and urine, especially when non-intrusive and quick sampling procedures are important (e.g., screening for drugs of abuse at the roadside and screening and confirmatory workplace drug testing). In this study, Rohypnol (flunitrazepam) was administered to four healthy volunteers, and oral fluid samples were collected by spitting into a polypropylene tube at fixed times between 0 and 6 h after the intake of a tablet of 1 mg. A specific and very sensitive method was developed, both for flunitrazepam and for its main metabolite 7-aminoflunitrazepam, based on solid-phase extraction of the oral fluid samples, stored at +4 degrees C, and gas chromatographic-mass spectrometric analyses using negative chemical ionization with methane as the ionization gas. The heptadeuterated parent compound and metabolite were used as internal standards. The respective limits of detection and quantitation were 0.05 microg/L and 0.1 microg/L for flunitrazepam, and 0.1 and 0.15 microg/L for 7-aminoflunitrazepam. The parent drug could only be detected when the analyses were performed within 12-24 h after collection of the oral fluid samples or when 2% of NaF was added to the collection tubes. The stability of flunitrazepam in oral fluid was poor, even at +4 degrees C, when no NaF was added to the sample. In any case, concentrations remained below 1 microg/L. The metabolite was detected in slightly higher concentrations, with or without the presence of NaF, reaching a maximum of 1-3 microg/L within 2-4 h after administration. In all cases the drug was detectable, but at extremely low concentrations, for 6 h after intake of a normal dose of Rohypnol and it will be an analytical challenge to come up with a sufficiently sensitive onsite test for low-dose benzodiazepines in oral fluid.