The timeline followback (TLFB) is the most widely used calendar-based method for collecting retrospective estimates of drug use. The TLFB uses a calendar and other memory aids (e.g. birthdays, holidays, special events) to gather retrospective estimates of an individual’s daily substance use over a specified time period. Studies on the psychometric properties of the TLFB for the measurement of alcohol consumption have shown high temporal stability and correlations between self and collateral reports have been high. Studies examining illicit drugs provide support for using the TLFB to assess frequency of illicit drug use (including cannabis); however, no study has examined the reliability and validity of collecting cannabis quantity information for adults. Current administrations of the TLFB for assessing cannabis use only require individuals to nominate the days in which they have used cannabis. On the other hand, other assessment instruments require individuals to nominate both the frequency and quantity of cannabis used. The few tools that assess quantity of cannabis provide conflicting guidelines for quantity measurement. Therefore it is imperative that researchers develop a reliable method to assess cannabis quantity. By doing so, an improved understanding of hazardous use can be garnered. In addition, reliable assessment will improve between subject comparisons when evaluating treatment effectiveness.
- Improve cannabis quantity assessment
- Evaluate the test-retest and interrater reliability of the modified TLFB
- Evaluate the validity of the modified TLFB by examining relationships between TLFB data and data obtained from single-item assessment and collateral reports about cannabis use.
The TLFB study involved having current cannabis users attend two face-to-face interviews, 14 days apart, for provision of the TLFB interview for cannabis use assessment over the past 90 days. Collection of cannabis use data was facilitated by utilizing the cannabis substitute Marijuanilla for weight assessment. At their first visit, participants also undertook a clinical interview on their drug and alcohol use. Collateral informants were asked to come into the centre during this 14 day period to complete the TLFB interview regarding the participant’s cannabis use.
Inter-reliability was excellent, while test-retest reliability was good to excellent. Intra-class correlation coefficients between participant and collateral reports, while similar to previous research, were unacceptable. Quantity of cannabis use statistically significantly added to frequency of use in predicting cannabis problems and dependence severity, and explained more variance in cannabis problem severity than frequency of use. Concurrent and discriminant validity were established with single-item and positive impression management measures, respectively. In addition, Marijuanilla appeared similar to one specimen of street seized cannabis, but not to two others. Importantly, participants’ cravings to use cannabis did not increase as a result of using the cannabis substitute to report on their cannabis use.
Manuscript has been accepted at Drug and Alcohol Dependence.
The data suggested that utilising Marijuanilla to facilitate the reporting of grams of cannabis use may be reliable and valid.