Test-retest reliability of mgTHC consumption in the self-administered Cannabis Exposure Index (CEI)

Cannabis legalization and adult use are expanding across the U.S., requiring greater knowledge of cannabis risks and benefits. Cannabis use measures traditionally assessed frequency but omitted quantity, an essential element of consumption that is complicated by increasingly varied cannabis products and use patterns. To address this, we developed the self-administered CEI (Cannabis Exposure Inventory) to determine milligrams of THC used per using day (mgTHC/using day). In this study, we examined the test-retest reliability of the CEI mgTHC measure and key component items (i.e., products, routes of administration).

Participants were recruited through social media (Facebook and Instagram ads) and Qualtrics Research panels. Eligible participants (n = 511) completed initial and retest CEI surveys. Chance-corrected agreement between initial and retest surveys on mean mgTHC/using day was indicated with Intraclass Correlation Coefficients (ICCs); kappa (k) indicated reliability of key dichotomous component variables.

Overall, ICC for mean mgTHC/using day= 0.77, indicating substantial reliability. In demographic subgroups, ICCs were 0.54 ('other' race/ethnicity) to 0.86 (Hispanic). ICCs for mean mgTHC/using day among those who used for medical-only, recreational-only and medical-plus-recreational reasons were 0.72, 0.69 and 0.77, respectively; ICCs for those in non-legalized, medical-only and medical-plus-recreational states were 0.70, 0.92 and 0.75, respectively. Binary measures generally exhibited substantial reliability (mean k, last 30 days=0.74; last 7 days=0.73).

Findings support the CEI mgTHC measure as a reliable instrument for quantifying cannabis use, addressing a critical gap in cannabis measurement. This measure offers a promising approach to provide urgently-needed information on potential harms and benefits of THC exposure.