"It is disturbing that the sensitivities of the cannabis and cocaine tests were all quite low, although further testing of the cocaine tests is desirable due to the low prevalences and the low concentrations encountered in this study. There are several countries in Central and Southern Europe for which these two substance classes are of special interest. On the other hand, it seems the sensitivities of the devices are generally better for amphetamines, a frequently encountered drug class among the DUI drivers in the Nordic countries. The suitability of the device for the intended national DUI population should also be considered, for example, PCP is rarely, if ever, found in Europe, therefore at the current time utilising a PCP test is unnecessary. Since the on-site tests are relatively expensive the suitability of all the individual substance tests incorporated in the device should be considered.
"The evaluation showed that none of the evaluated tests is on a desirable level (>80% for sensitivity, specificity and accuracy) for all of the separate tests that they comprised. However, there were tests that performed already on a promising level for one or more substance classes. The DrugTest 5000 had the best overall results. The next best device was Rapid STAT, which performed at a similar level, except for the cocaine test which was somewhat less sensitive. Clearly the best device in terms of sensitivity for amphetamines was the DrugWipe 5+."

Source

Driving under the Influence of Drugs, Alcohol and Medicines (DRUID Project) 6th Framework Programme, "Analytical evaluation of oral fluid screening devices and preceding selection procedures," Deliverable 3.2.2 (Finland: National Institute for Health and Welfare, March 30, 2010), p. 95.
http://www.druid-project.eu/D…