A Breath THC Cutoff Can Be Analytical Without Being an Impairment Threshold

forensic toxicology working group members October 28, 2026 Forensic Toxicology

A breath assay’s reporting cutoff does not become an impairment threshold merely because the method accurately measures THC. A Cannabix collection-and-analysis study validated analytical procedures, not a concentration that identifies unsafe driving.

What was calibrated

The procedure collected a specified breath volume onto cartridges and used LC–MS/MS to separate delta-9 THC, delta-8 THC, CBD and cannabinol. Calibrators and quality controls were prepared by adding reference compounds to unused collection pods.

Cartridge calibration establishes the relationship between loaded analyte and laboratory response. It is distinct from proving how efficiently the complete device captures every person’s exhaled cannabinoid across different breathing patterns and conditions.

The human application had naturalistic dosing

Participants smoked their own cannabis ad libitum, with reported product potencies, and provided samples before and after use. Ambient-air collections were included. This demonstrates application to actual breath samples, but does not standardize absorbed dose.

The study’s analytical cutoff, including a 5-pg/L example, should be identified as a method-specific reporting decision. It was not derived by validating the boundary against driving outcomes in this experiment.

Split samples answer a different question

The collection system allows separately sealed primary and confirmatory samples. Preserving material for another analysis is valuable for review and chain of custody. It does not, by itself, validate the inference from a concentration to impairment.

Separate laboratory analysis can investigate identity and concentration while leaving unresolved whether the measured exposure caused a particular functional deficit.

A useful assessment separates the layers

Ask about collection volume, capture performance, cartridge recovery, blanks, calibration, analyte separation and repeat analysis. Then ask whether a separate study validates the proposed behavioral interpretation in the relevant population.

This paper supports an analytical breath-testing workflow and documents actual post-smoking samples. It does not establish a universal breath detection window, a legal limit or an individual driving-impairment classifier. Those additional claims require their own evidence.

For context, see our discussion of THC concentrations and the separate question of impairment.