The oral fluid THC test can flag recent THC use in a driver with high accuracy. But it cannot say how much THC sits in the blood. That is the verdict of a Columbia University study published in Injury Epidemiology in 2018. Meanwhile, Australia’s food regulator has explained in a letter what the law covers when THC turns up in food. The two documents come from different worlds, yet both deal with how THC gets checked.
How the Oral Fluid THC Test Performed on the Road
The research team, led by Huiyan Jin and Qixuan Chen, used data from the 2013 National RoadsideSurvey of Alcohol and Drug Use by Drivers in the United States. First, they picked out 4,596 drivers who gave both an oral fluid sample and a whole blood sample. Overall, 8.9 per cent tested positive for THC in oral fluid and 9.4 per cent in blood.
Then came the accuracy check. Using the blood result as the benchmark, the oral fluid THC test showed a sensitivity of 79.4 per cent and a specificity of 98.3 per cent. In plain terms, it caught about four in five drivers with THC in their blood. It also rarely flagged a driver who had none. The authors called it a highly valid method for detecting the presence of THC in the blood.
Odds ratios tell the same story. After adjusting for alcohol, time since last use and driver details, a positive oral fluid result went with about eleven times the odds of a positive blood result.
Why Roadside THC Screening Cannot Replace a Blood Result
However, the study drew a hard line at concentration. Oral fluid THC explained only about 29 per cent of the variation in blood THC. A 10 per cent rise in oral fluid THC matched just a 2.4 per cent rise in blood THC. The authors advised against working out blood levels from an oral fluid result.
That matters because blood is the standard for legal limits. The paper describes THC in blood as a marker of recent use, within roughly one to eight hours. It also notes that 20 US states have per se limits for marijuana impaired driving. Four use 5 nanograms per millilitre, and the rest apply zero tolerance. So a screen can show presence, but it cannot settle a threshold.
The test also had weak spots. Sensitivity fell to 54.6 per cent among drivers older than 55. Specificity dropped to 66.7 per cent among drivers who had used marijuana within the past 24 hours. In addition, the authors warned of possible bias. Fewer than half of eligible drivers agreed to give both samples.
Younger Drivers and Recent Use Stand Out
The figures also show who is most likely to test positive. Among drivers aged 16 to 20, 15.9 per cent had THC in their blood. For drivers aged 55 and over, the rate was 3.1 per cent. Among drivers who reported using marijuana in the past 24 hours, 75.6 per cent tested positive.
The paper adds that marijuana detection among drivers in fatal crashes tripled from 4.2 per cent to 12.2 per cent between 1999 and 2010. Because THC in blood points to recent use, those numbers underline why early conversations with young people matter.
What Australia’s Food Regulator Told Drug Free Australia
Food Standards Australia New Zealand (FSANZ) wrote on 11 September 2026 to Herschel Baker, International Liaison Director at Drug Free Australia. The letter replies to his 4 August correspondence about hemp derived THC products. Matthew O’Mullane, General Manager of the Food Safety Branch, signed it.
The regulator explained that Standard 1.4.4 of the Food Standards Code lists cannabis as a prohibited plant for food unless an exemption applies. One exemption covers specified foods from low THC hemp seeds. Here, the Code defines low THC hemp as Cannabis sativa with no more than 1 per cent delta 9 THC in the leaves and flowering heads.
Mr Baker had raised concerns about delta 8 THC, delta 10 THC, tetrahydrocannabiphorol, THCA and other variants. In reply, FSANZ said the standard sets maximum levels for CBD and total THC. Total THC means delta 9 THC plus delta 9 THCA. Foods must come from hemp seeds, and they may contain only cannabinoids that occur naturally in or on those seeds. FSANZ built these rules on a risk assessment under Proposal P1042.
Enforcement sits with state and territory food agencies. FSANZ suggested that anyone worried about a product raise it with the local agency. It also said it does not set standards for medicinal cannabis or other non food cannabis derived products. The Therapeutic Goods Administration and state and territory authorities regulate those. Still, the letter does not announce any change to the Code.
Two Documents, One Question of Where THC Gets Checked
Each system answers a different question. The Code limits what THC food may carry, and state agencies enforce it. By contrast, the oral fluid THC test asks whether THC is present in a driver at all. Neither document links hemp foods to positive roadside results. The study also covers drivers in the United States only.
For now, the letter leaves the Code as it stands, and the research leaves the blood test as the benchmark. Until that changes, the oral fluid THC test remains a screen.
(Source: WRD News)