Lessons learned: challenges in recruiting and engaging people with heavy cannabis use for online interventions in Canada.

backgroundPeople who seek treatment often disengage between completing screening and starting treatment. Among those who begin treatment, many have low completion of program content and are lost to follow up. Currently, little is known about predictive factors of treatment initiation and engagement. The aims of this secondary analysis of the Canadian CANreduce program were to discuss the practicality of a randomized controlled trial for online heavy cannabis use treatment and to examine predictors of treatment initiation and engagement in the Canadian CANreduce program.

methodThe CANreduce intervention was pre-registered on clinicaltrials.gov for traceability (ID: NCT04965012). Statistical models were organized into three conceptual predictor groupings using baseline data: individual cannabis-specific factors, mental health and other substance use factors, and treatment belief factors. Binary logistic regressions examined which factors predicted treatment initiation in the CANreduce treatment program and multiple regression analyses examined which factors predicted percentage of the CANreduce program modules completed among participants who initiated treatment.

resultsDespite following the core elements of published treatment retention protocols, challenges in recruitment were evident. Of 928 people who created a profile on the website and began screeners, 86.3% (n = 801) completed screeners. Of the 801 who completed screeners, 31.3% (n = 251) were eligible for the program. Of those eligible and assigned to active treatment, 54.3% (n = 51) assigned to the therapist condition and 45.7% (n = 43) assigned to the research assistant condition initiated treatment. Treatment initiation predictors included higher cannabis use problems score, lower family history density, increased alcohol use frequency, and more positive attitudes towards treatment. Treatment engagement (percentage of program completed) predictors included increased social motives for cannabis use and more positive attitudes towards treatment. DISCUSSION: This secondary analysis shifts the focus from treatment efficacy to pre-treatment attrition and early engagement in an online cannabis intervention. Substantial recruitment and initiation challenges were observed, highlighting vulnerabilities in the treatment-seeking pathway before content is accessed. By identifying predictors of treatment initiation and module completion, this study offers practical “lessons learned” to inform recruitment and engagement strategies in future online cannabis trials.