Comparing presumptive with direct-to-definitive drug testing in oral fluid vs. urine for a U.S. national sample of individuals misusing drugs.

Background: The aims are to compare the results of presumptive drug testing with confirmation of positives vs. direct-to-definitive drug testing, combined with investigation of urine vs. oral fluid as test matrices.

Methods: Paired oral fluid and urine specimens were collected voluntarily and anonymously from 1098 individuals applying for methadone treatment in 11 clinics across 7 U.S. states. All specimens were analyzed by immunoassay (IA) and liquid chromatography-tandem mass spectrometry (LC-MS-MS).

Results: Confirmed IA prevalences for urine were significantly higher than for oral fluid for 7 out of 10 drug classes - benzodiazepines, cannabis, cocaine, methadone, opiates, oxycodone and tramadol. Drug prevalences by direct-to-definitive LC-MS-MS were either the same or higher than prevalences by confirmed IA. Drug prevalences by LC-MS-MS were higher in urine for two drug classes (cocaine, methadone) and higher in oral fluid for two drug classes (buprenorphine, tramadol), but were equivalent in urine and oral fluid when averaged over all 10 drug classes. Certain drugs of special concern such as heroin and buprenorphine were more frequently detected in oral fluid than urine.

Conclusions: Urine analysis showed some technical advantage over oral fluid in sensitivity to several drug classes within a confirmed IA testing protocol, but this may be outweighed if there is reason to believe that tampering with urine specimens is a significant problem. Overall drug detection by direct-to-definitive testing was similar for oral fluid and urine, but one matrix may be preferable if there is a particular drug of clinical or epidemiological interest.