Roadside Saliva THC Sensitivity Was Lower in the Study’s Oldest Group

An overall saliva-test sensitivity estimate did not describe every age subgroup in a large roadside survey analysis. The oldest group had lower estimated sensitivity, but the small number of blood-positive people makes a universal age rule unjustified.

The reference was chemical positivity

Researchers analyzed paired oral-fluid and whole-blood THC results from 4,596 drivers in the 2013 National Roadside Survey. Testing included laboratory screening and confirmation. Oral-fluid positivity was defined at 2 ng/mL, while blood positivity used the study’s detectable-THC criterion, corresponding to at least 1 ng/mL.

Across the dataset, oral-fluid testing had estimated sensitivity of 79.4% and specificity of 98.3% relative to blood positivity. These were comparisons between specimen classifications. They were not estimates of sensitivity or specificity for impaired driving.

The oldest subgroup was less precise

Among 717 drivers aged at least 55, only 22 were blood-positive. Estimated oral-fluid sensitivity in that group was 54.6%, with a 95% confidence interval of 32.2–75.6%. The interval communicates substantial uncertainty that the single percentage conceals.

The observation supports caution about applying the overall estimate to every subgroup. It does not establish that age itself caused the difference. Use patterns, timing, measured concentrations and participation selection could contribute, and this observational comparison did not isolate an age mechanism.

The sampling frame matters

Less than half of eligible drivers supplied both specimens, and consenting drivers differed from those who did not. The findings therefore require attention to potential selection bias. Changing either specimen’s decision threshold also changed the reported diagnostic performance.

Continuous saliva concentrations explained only about 29% of variation in log-transformed blood THC. Thus, even an association between positive classifications did not supply an accurate individual concentration conversion. For case review, retain the person’s actual sampling times and laboratory thresholds, and distinguish missed blood positivity from missed impairment.

Related: interpreting THC concentrations separately from impairment.

Primary research

Validity of oral fluid test for Delta-9-tetrahydrocannabinol in drivers using the 2013 National Roadside Survey Data. DOI: 10.1186/s40621-018-0134-2.