Cannabis use is becoming increasingly accepted, with many countries legalising medicinal and recreational use (Nabil et al. Chem Rev Lett 7(5):827–828, [1]). A significant issue arising from this is how to reliably detect recent use that may pose a risk of impairment while avoiding detection of distant use. Oral fluid (OF) testing is one proposed method of detecting recent use (Hubbard et al. J Anal Toxicol 45(8):820–828, [23]; Wennberg et al. Heliyon 9(4):e14630, [24]). This is of particular relevance in the context of workplace testing in safety sensitive workplaces. This review attempts to provide further guidance and specifically assess probabilities of detection at ≥ 5 ng/mL(via mass-spectroscopy) following THC use. The ≥ 5 ng/mL limit has been used as this is the current cut-off value for reporting a positive oral fluid drug screen result in Australia under AS/NZS 4760:2019. Experimental studies providing individual participant mass-spectroscopy monitoring data for at least 6 h after use of THC have been examined and compared to the 5 ng/mL limit noting the sensitivity of detecting use at different time periods. Sixteen articles met the inclusion criteria. Studies were grouped into four categories: 1) Frequent users after inhaled use, 2) Occasional users after inhaled use, 3) Passive exposure (inhaled) and 4) Other methods of use. These 16 articles included 37 different research situations: 5 papers examined occasional smokers, 6 examined frequent or heavy smokers, 6 examining passive exposure, and 5 examining other methods of use. In the other methods of use group, 3 articles examined THC in baked products, 1 examined “light” cannabis, and 1 examined medicinal oil. Initial post exposure OF levels over all studies and subjects varied from 0 to 71747 ng/mL OF across 241 subjects. For all groups, THC was detectable (at OF levels ≥ 5 ng/mL) at 0–0.25 h after intake (There was a single negative result at the 0.17 h, however the next readings were 29 and 25, this appears to be most likely a measurement error). All groups except passive users were 100% positive at this early time point (passive smokers were 57% positive). At two hours post-use, 83% of frequent smokers, 38% of occasional smokers, 6% of passive users and 59% of edible users were positive. Maximum detection time (defined as 1 in 100 still positive using study data) was estimated using linear regression. Values were calculated by grouping data by study (12 groups) and by initial OF level (14 groups with group mean initial OF levels ranging from 18.6 [passive] to 13989 ng/mL [highest initial OF level frequent smokers]) for the different exposure methods. The maximum detection time ranged from 3.3 h (range 2.7 to 4.3 h) for the passive smokers to 162 h (range 34 to NQ (In the frequent smokers group there were very few tests able to be used for regression analysis resulting in very large confidence intervals with the upper limit not quantifiable. See text.)) for frequent users with the highest initial OF level. Regarding the number of oral fluid THC levels ≥ 5 ng/mL at different time points there was a high level of heterogeneity in the data both inter-study and within studies. The SD/(Initial OF level) for the studies averaged 1.28 (range 0.5 to 2.4). Levels were also high when this value was calculated at other time periods. Considerable heterogeneity existed among methods and doses, with no consistent study methods or dosages across studies. This is, however, consistent with the pattern of cannabis use in the general population. It is important to recognize this heterogeneity when interpreting these analyses and attempting to generalize to real-world situations. This study suggests that OF testing for THC shows consistent average detection times on a population level depending on initial dose, there are significant drawbacks on an individual level. In particular, there was variation among individuals with wide confidence intervals in the rate of decline in oral fluid levels. Even without incentives to mask test results, a large number of results showed a negative test shortly after use (well within the time period where one may be expected to be impaired). This study also identified several areas where further research is needed. In particular, more research is needed regarding recent and emerging patterns of THC use (for example; vaporised and medicinal products) as well as further studies and monitoring with more frequent and extended detection times, with greater numbers of test subjects particularly in heavy and regular users.
