|
Abstract:
|
As legalized recreational marijuana here in Colorado spreads throughout the U.S., marijuana use is on the rise and we are in the midst of an opioid epidemic. Thus, difficulties with accurately capturing characteristics of substance use (e.g. quantity, frequency, intoxication) are increasingly important and of high interest. This round table discussion will identify some of these problems and possible solutions with the goal of improving the characterization of addiction. Examples are: (i) Utilization and modeling of self-report data collected via Timeline Followback, the gold standard calendar method for assessing substance use. Typically, daily use information is grossly underutilized with data merely aggregated into summaries (e.g. proportion of days abstinent) that rely on questionable assumptions for missingness. (ii) Quantification of substance intoxication using metabolites in urine divided by creatinine level (e.g. for marijuana: THC-COOH/creatinine). The resulting “normalized” variables typically do not satisfy mathematical assumptions underlying ratios and can lead to spurious results. Suggested solutions for these issues will be discussed and other examples are welcome.
|
