Previously conducted prospective cohort studies have suggested that substance use begins in adolescence and peaks in young adulthood [1], and that regular cannabis use (CU) typically starts between the ages of 18 and 19 years [2]. In Canada, the 18-24-year age group has been found to have the highest percentage of cannabis users [3].

Recreational CU was legalized nationwide in Canada in 2018 [4]. Since then, the use of cannabis products has increased considerably, particularly among the 18-25-year age group [1,5]. It has been shown that increased availability of legal cannabis can lead to higher CU and associated riskier use behaviors [6]. Legalization has made it all the more imperative to develop and offer preventive interventions aimed at reducing CU-related harm and encouraging safer behaviors [7]. Young adults, in particular, stand to benefit from such initiatives, given the high prevalence of CU in the 18-25-year age group and the limited number of existing interventions designed specifically for them [8].

A new generation of interventions targeting CU through digital technologies and mobile health (mHealth) has drawn researchers’ attention over the past decade. Several systematic reviews and meta-analyses [9-14] have demonstrated the relative effectiveness of various digital interventions in reducing CU frequency, although the effect sizes have been relatively small. However, these reviews have primarily focused on web-based interventions.

Smartphone usage and social media are increasingly popular among young people [15], and young adults are open to digital health services, especially when delivered via mobile apps [16]. However, our own systematic review of the literature found that none of the digital CU-related interventions available through mobile apps had been evaluated using a randomized controlled trial (RCT) [17]. A scoping review based on 5 studies, which aimed to explore the technical and functional characteristics of mHealth apps, found that this intervention modality demonstrated adequate feasibility and acceptability [18]. The following CU-related mobile apps were covered in the review: APPT (Assess, Plan, Track, and Tips; 18-50-year age group; Australia) [19]; MiSARA, a substance abuse research assistant (16-24-year age group; United States) [20]; VoltEgySzer (Once Upon a High; 14-18-year age group; Hungary) [21]; Stop-cannabis (14-59-year age group; Switzerland) [22]; and MApp (Marijuana Smartphone App; 18-25-year age group; United States) [23]. Since this review, 2 additional mobile apps have been mentioned in the scientific literature: HAP-app, a mobile app intended to help individuals reduce or cease CU (Norway) [24]; and LooseLeaf, a mobile app for young people at clinically high risk for psychosis, designed to help monitor CU and cannabis-related experiences (14-30-year age group; Canada) [25]. In short, the current research on CU mobile apps remains at an early stage, and most published studies have targeted either specific clinical populations or broad segments of the general population.

An innovative and positive way of promoting lower-risk CU is through the use of protective behavioral strategies (PBSs). PBSs are approaches that can help mediate CU and reduce negative CU consequences [26-28]. They involve behaviors—enacted immediately before, during, or after using cannabis—that support better CU self-management [26]. The use of PBSs has been associated with reduced negative effects of CU and lower CU frequency and quantity in a population of young university students [26-28]. In a previous study involving Canadian university students who reported any CU in the past 30 days (n=211), we found that greater use of PBSs was related to lower CU frequency, and that daily cannabis users employed fewer marijuana PBSs on average than lower-frequency users did [28]. A recent scoping review showed that PBSs may be associated with reduced CU-related negative consequences [29].

While the incorporation of PBSs is linked to CU self-management and holds great promise for lower-risk CU interventions, there has been very little research on PBS-centered interventions beyond the realm of alcohol use [9], and only a few interventions to date have targeted CU specifically [30]. Preliminary testing of MApp, a smartphone app used as part of a brief in-person intervention, indicated that PBS use reduced CU and that young adults were inclined to use a mobile app to help moderate their CU [23]. More recently, Lewis et al [31] published an RCT study protocol to evaluate a brief PBS-focused web-based and SMS intervention, although the results have yet to be published.

In sum, although the evaluation of PBS use as an intervention target in the context of safe CU is still at an early stage, findings to date suggest that PBS use is a positive and innovative approach to promoting lower-risk CU [31]. Against this background, we used intervention mapping (IM) to develop a mobile app prototype designed to support CU self-management and reinforce PBS use, with the goal of promoting lower-risk CU among young adults in the context of legalized recreational CU in Canada.

This study aimed to develop Joint Effort, a mobile app prototype designed to support CU self-management and reinforce PBS use among young adults, and to assess its usability.

IM is a rigorous, theory- and evidence-based approach that emphasizes the role of research and theory in the intervention development process [32]. Our process was based on IM and a co-design approach. It included 6 steps: (1) conducting focus groups to identify the needs and preferences of young adults regarding CU interventions; (2) creating matrices of change objectives to select target behaviors and determinants; (3) selecting theory-based intervention methods and practical applications; (4) developing preliminary intervention content and structure; (5) conducting focus groups to validate the intervention structure and examples of tailored messages; and (6) transposing the intervention content into a mobile app prototype. The completion of one step guided the development of the next [32]. Various research phases and activities were embedded in the IM steps. These are described below.

Usability testing aimed to document the experience of using the mobile app prototype. This was conducted through semistructured interviews and the User Version of the Mobile Application Rating Scale (uMARS) [33]. A multimethod approach combining quantitative and qualitative data is generally recommended in usability testing to provide a more comprehensive understanding of user experiences [34]. The methods and results are reported in accordance with the Consolidated Criteria for Reporting Qualitative Research (COREQ) [35] (see Multimedia Appendix 1).

The target population for this study was selected based on 3 elements. First, substance use peaks in young adulthood [1], and the 18-24-year age group has the highest percentage of cannabis users [4]. Second, the legal age to purchase cannabis in Canada varies from 18 to 21 years, depending on provincial laws [36]. In Québec, where this study was conducted, it is illegal for anyone under 21 years to purchase cannabis. The lower age threshold for the study population was therefore set at 21 years. Third, most research on PBSs has been conducted with young adults and college or university students [26,37,38]. Given that the proof of concept for interventions based on PBSs has been demonstrated with this clientele, an academic setting was favored. This is why university students aged 21-24 years were chosen as a convenience sample to represent the young adult demographic. Recruitment was conducted at the Université de Montréal, a predominantly French-language university located in a metropolitan region (Montreal, Québec, Canada) in Eastern Canada.

All steps of the study were approved by the Centre Hospitalier de l'Université de Montréal Research Ethics Board (approval number 20.172) and the Comité d’éthique de la recherche en sciences et en santé (CERSES) of the Université de Montréal (approval number CERSES-20-114-D). All participants provided informed consent. All participants provided informed consent by signing a copy of the informed consent form. Compensation to participants was disclosed in previous sections. Privacy and confidentiality of participants’ data or identity were maintained during the study process and publication.

In May 2021, usability testing was conducted with 20 participants whose mean age was 21.8 (median 22) years. Most were women (14/20, 70%), born in Canada (15/20, 75%), and undergraduates (15/20, 75%). Additionally, most reported engaging in weekly CU in the past 30 days (14/20, 70%).

After the think-aloud portion of the interview, participants (n=20) were sent a link to complete the uMARS. Results per item and per subscale are presented in Table 5. The Joint Effort prototype received the following mean scores on the 4 subscales: 4.14/5.0 (SD 0.53) for Engagement, 4.60/5.0 (SD 0.47) for Functionality, 4.53/5.0 (SD 0.52) for Aesthetics, and 4.44/5.0 (SD 0.61) for Information. This yielded an overall app quality mean score of 4.43/5.0 (SD 0.53). The mean scores on the subjective items varied widely. The item “Would you recommend this app to people who might benefit from it?” received the highest score at 4.35/5.0 (SD 0.67), while “Would you pay for this app?” received the lowest at 1.65/5.0 (SD 1.09).

The objective of this study was to codevelop and conduct usability testing of a mobile app prototype, aimed at promoting CU self-management and reinforcing PBS use among university students.

Focus groups were conducted to identify the needs and preferences of young adults regarding CU intervention. The data collected during this phase informed the development of the intervention content. A second round of focus groups was then held to validate the structure of the intervention and examples of tailored messages it might convey. Subsequently, the intervention content was transposed into digital form through iterative software development, and the mobile app prototype was created. A third validation phase was carried out using think-aloud interviews and a short questionnaire to document the user experience.

Usability findings suggest that participants appreciated the app. The prototype received an overall app quality score of 4.43/5. The mean scores on the uMARS subscales were relatively high for Engagement (4.14/5), Functionality (4.60/5), Information (4.44/5), and Aesthetics (4.53/5). Interestingly, the item that received the lowest rating was “Would you pay for this app?” (1.65/5.0, SD 1.09). This suggests that, despite the app being perceived positively, participants were not willing to pay a subscription fee. A systematic review has shown that offering health-related apps for free or at a low cost can positively influence their uptake and engagement [50].

The results of the think-aloud interviews (qualitative component) aligned with the uMARS findings. Participants found navigation smooth and intuitive, which made it easy for them to find their way around, and the content conducive to prompting behavior change. A systematic review and thematic synthesis (n=35 studies) of mHealth interventions identified reinforcement, communication, navigation, credibility, message presentation, and interface aesthetics as key design features to consider for improving user engagement [51]. Given the strong uMARS scores obtained for these components, it is reasonable to expect that the Joint Effort app may successfully elicit user engagement.

However, comparing our uMARS results with those of other studies is challenging. Although this tool has been used elsewhere, most mobile apps evaluated with the uMARS differ substantially from Joint Effort. To date, in the field of cannabis and other substance use research, the use of uMARS has primarily been mentioned in protocols [52-54]. Only Santesteban-Echarri et al [55] reported using the uMARS to evaluate a cannabis-related mobile app for youth at high risk for psychosis. They found a good overall score (3.75/5) and high subscale ratings for esthetics (4.48/5), information (4.32/5), and functionality (4.29/5) [55]. In comparison, the Joint Effort prototype received a higher overall app quality score (4.43/5). The slightly higher scores obtained in our study may be attributed to the IM and co-design approach used, along with multiple validation rounds. The target population was involved from the outset (ie, needs analysis), and potential end users were engaged throughout the process. However, these comparisons should be interpreted with caution, as subjectivity is inherent when using a scale such as the uMARS. For instance, aesthetic preferences may vary across cultural groups. Moreover, the uMARS was completed after a single use of the app prototype, rather than after extended use of its final version.

The aim of the mobile app prototype developed is to promote PBS use, which involves setting a goal, engaging in a change process, and implementing an action plan (SMART, take action). Our qualitative results suggest that participants appreciated the opportunity to engage at their own pace, set their own goals, and initiate the behavior change they desired, if any. In a rapid review of the literature (n=43 studies), Monarque et al [56] concluded that the uptake of digital interventions among youth depended on the incorporation of harm reduction principles and skills training. These approaches were favored in the development of Joint Effort, as participants were supported in self-managing their CU by reinforcing PBS use, rather than being encouraged to quit altogether.

Different approaches were used to support CU self-management and reinforce PBS use in developing the app, including personalized feedback, anticipated outcomes, positive reinforcement, goal setting, self-observation, and activation of intention. In a systematic review examining the efficacy of behavior change smartphone apps (n=27 studies), Schoeppe et al [57] emphasized common strategies frequently used in successful app-based interventions, such as goal setting, self-monitoring, and performance feedback. These findings were later reinforced in the systematic review by Milne-Ives et al [58], which identified 6 common behavior change techniques associated with user engagement in mHealth apps: goal setting, self-monitoring of behavior, feedback on behavior, prompts/cues, rewards, and social support. In light of the above, the Joint Effort mobile app prototype was designed using the most effective strategies to promote engagement and intervention success. These strategies will be evaluated in subsequent research phases, with a focus on both experiential and behavioral engagement [59].

An unexpected element emerged from the qualitative interviews regarding the nonjudgmental approach used. Participants reported not feeling judged or stigmatized and expressed that their behavior felt normalized. This is particularly important given the ongoing stigmatization of CU, both in jurisdictions where cannabis is legal and where it remains illegal [4,60]. In terms of intervention access, several barriers have been documented, including limited availability, confidentiality concerns, and stigma [61]. The intervention was developed with the intention of serving as a tool for self-directed prevention, offering simple and achievable strategies that require no reliance on external resources. This could address a specific need among young adults who use cannabis and are interested in changing their consumption habits. The advantages of interventions delivered via mobile apps are well documented in the literature. For example, mobile apps provide an economical and easily accessible means of delivering low-intensity interventions for mental health–related issues [62]. Anonymity, asynchrony, and easy, immediate access (without the need to leave home) were the advantages most appreciated by participants in this study.

The principal strength of this study lies in the rigorous, theory-based process used to cocreate the mobile app prototype. IM is a robust framework for developing theory- and evidence-based interventions [32]. Multiple methods were used to conduct an extensive needs analysis, including a systematic review [17], an online CU study [28,63], and focus groups. The overall IM process involved the continuous integration and analysis of 3 types of knowledge: empirical (previous studies and needs analysis), experiential (input from key informants), and theoretical (behavioral determinants). The iterative codevelopment process included multiple phases of consultation with potential end users of the mobile app and several validation loops to ensure the intervention aligned with user preferences and needs. Finally, the study was conducted by a well-rounded, thoughtfully assembled interdisciplinary team of researchers with diverse expertise, including cannabis, information and communication technologies, health promotion, youth mental health, knowledge transfer, and mixed methods research (both quantitative and qualitative), working in collaboration with potential end users.

A key limitation of the study is the overwhelmingly positive user reactions and reviews received during the think-aloud interviews. Participants did not mention any aspects they disliked or suggested elements for improvement at that time. This may be attributed to the methodology used: given the stage of the design process, participants interacted with the prototype for the first and only time during the interview. As such, the results reflected their initial general impressions. A more extended and in-depth use over time may have elicited more nuanced feedback.

Another limitation is the predominance of female participants across the focus groups conducted to identify user needs (11/13, 85%), validate the intervention content (7/11, 64%), and during the usability testing phase (14/20, 70%). Given that negative CU consequences are disproportionately concentrated in the young adult male population [7,29], it is reasonable to assume that a higher proportion of males with different viewpoints might have led to a slightly different intervention. In sum, these findings highlight the challenges of developing effective mobile app interventions for CU that account for gender differences across various clinical correlates (eg, readiness to change, sources of motivation for behavior change) that may influence young adults in their decision to take steps toward better managing their CU.

In the context of the recent legalization of cannabis and its widespread consumption, strategies to promote safer use and reduce harm need to extend beyond traditional abstinence-based approaches. mHealth interventions appear promising in this regard, particularly in addressing challenges related to health care access and stigmatization. Following this study, and given the promising results from usability testing, we proceeded with the further development of the Joint Effort mobile app. While the work described in this manuscript yielded positive signals from potential end users, further evaluation is needed to confirm its acceptability, feasibility, and ultimately, its efficacy in terms of CU outcomes. A pilot randomized trial [64] was conducted to assess the intervention’s acceptability (user uptake, user engagement, user-participant profiles, and intervention appreciation) and to document the feasibility of the study process (online recruitment rate, adherence to online data collection methods, and attrition rate). Based on the logic model of change (Multimedia Appendix 3), an RCT is ongoing to evaluate the intervention’s efficacy [65] on PBSs as the primary outcome, with frequency of CU as a secondary outcome, and intention to take action on CU as a mediator outcome. If proven efficacious, Joint Effort could help diversify the available tools to improve CU outcomes.