Except for total dry cannabis used, differences between the CEA and TLFB were non-significant

A self-report measure that can be completed more quickly may be more desirable when a participant is completing a host of other measures as well. In sum, measurement of cannabis use has generally not kept pace with shifts in patterns and modes of engagement and measures that do capture the diversity of ways in which cannabis is used are not designed to track changes over time. Therefore, we sought to develop the Cannabis Engagement Assessment , a self-report measure of use in the preceding month. To ensure content validity , individuals who use cannabis recreationally were engaged in qualitative interviews to examine current patterns and modes of cannabis use, amounts, and other factors that could be relevant for assessing cannabis use. The information obtained during the qualitative interviews, in conjunction with consideration of the literature, was used to guide development of the CEA. Types of cannabis products are divided into three major categories, consistent with how participants described the types of cannabis products: dried cannabis flower products, cannabis vertical farming concentrates, and edible products.

Visual depictions of both dried products and various concentrates, adapted with permission from Goodman, Leos-Toro, and Hammond , are also included to help participants estimate amounts used. As the measure is intended to be able to assess changes in cannabis use over time, we also aimed to determine whether information from the CEA could be used to provide an estimate of overall cannabis and THC used in the preceding 30 days. The project aimed to assess the psychometric properties of the Cannabis Engagement Assessment , a self-report measure of recent cannabis engagement . Qualitative interviews with individuals who use cannabis recreationally, in conjunction with the existing literature, guided the development of items on the CEA reflecting current use modes and patterns. Overall, the CEA showed good psychometric properties, supporting its utility of a measure of recreational cannabis use that can be used to track changes over time. These questions are consistent with the recently established guidelines for quantifying self-reported cannabis use . The iCann toolkit also recommends that all self-report questionnaires include an item that assesses when the participant most recently used cannabis.

Considering this recommendation, we have also added such a question to the CEA such that the CEA meets the expert-recommended universal guidelines for quantifying self-reported cannabis use. While the importance of assessing both frequency and quantity of cannabis is recognized , there is significant heterogeneity in the literature regarding the best operationalization of cannabis reduction . Given these ongoing issues, the CEA includes several indices of cannabis engagement that integrate both frequency and quantity.For concentrated cannabis products, approximately 60% of participants could not estimate the concentrate amount in mg or puffs/hits per session. For the participants who indicated the amount they used in puffs, we used a standard estimate of 5.2 mg/puff from Varlet et al.  to calculate the amount of mg. of cannabis product and subsequent THC estimates based on the THC concentration. While this is a rough estimate only, it is a first step in estimating overall amounts of THC consumed through recreational cannabis drying rack use, considering both quantity and frequency. It is important to note that other factors such as bioavailability may impact cannabis use and are not able to be estimated in a brief self-report measure. The CEA showed generally strong psychometric properties. Estimates of quantity and frequency obtained by the CEA showed strong associations with the same estimates obtained through timeline follow-back interviews, supporting its criterion validity.

On average, individuals reported 5.21 g more of dry cannabis product on the CEA than the TLFB. This is unsurprising as one of the variables used to calculate this estimate, amount used per session, showed poor temporal stability . Sessions of concentrate per day also showed low-test retest reliability . While some variation in patterns of cannabis use across weeks is to be expected, it is unlikely that the low test–retest reliability for concentrate sessions was due to a significant change in pattern of use during the week between time 1 and time 2; Test-retest reliability across other questions was good, suggesting that individuals did not markedly change their pattern of cannabis use. Rather, it appears that participants struggled to consistently estimate how much product they typically used in a single session in the previous month. To address this, we modified the questions on the final version of the CEA to ask how much dry product, on average, was used in a single day, rather than in a single session, and the average number of ‘hits’ per day for concentrated products.