[Drug tests on 198 drivers involved in an accident]

A prospective analytical study was performed in a large population of drivers implicated in traffic accidents to determine the significance of drug levels observed in blood, urine, saliva and sweat and which assays are best to perform in each sample.

Samples of blood (7.5 ml), urine (10-20 ml), saliva (salivette Sarstedt) and sweat (cosmetic pad spiked with water-isopropanol) were systematically collected in drivers implicated in non-fatal traffic accidents from March to November 1999. The samples were tested for pharmaceuticals (barbiturates, benzodiazepines, anti-depressants, neuroleptics, antiepileptics and antihistamines) and for drugs of abuse by hyphenated chromatographic methods (LC/DAD, GC/MS and LC/MS).

A total of 198 drivers (bicycle, motorbike, car, truck) were tested (age range 13-57 years, 82% males). Blood alcohol was positive in 27 cases (13.7%), ranging from 0.11 to 3.19 g/l (mean 1.49 g/l), being > 0.5 g/l in 21 cases. Cannabis was the most frequently observed illicit drug (9.6% of the cases). Its formal pharmacological effect could only be documented by blood testing using GC/MS. Even in the hospital setting, urine collection was difficult. In 16% of the cases, this fluid was missing. Parent compounds were excreted in both saliva and sweat. On-site devices devoted to urine and metabolites were inapplicable. Concentrations in sweat and saliva were very low, particularly for benzodiazepines and cannabis. There was also a risk of external contamination for sweat.

Saliva might be a good substitution fluid for blood for sample taking on the road side.