JMIR Publications

ABSTRACT

Background:

Young Black and African American males (YBM) are disproportionately affected by firearm homicides and injuries. In large city centers with multiple behavioral programs, minority youth see a 50-60% decrease in service utilization, especially after they reach age 17. Several cognitive and structural barriers to care are at play for YBM — who are both underserved and at the highest risk for violence victimization and perpetration. However, few evidence-based programs address these barriers to care. Current violence prevention interventions rarely prioritize primary prevention, are typically implemented in emergency settings (after violence has occurred), and do not offer continuous access to evidence-based crisis resources. Online/digital interventions are both pragmatic and effective therapeutic and barrier-reducing tools for preventing risk and involvement in risky behaviors, including co-occurring youth violence and substance use. We developed and adapted BrotherlyACT—a culturally-tailored, multicomponent smartphone app intended to reduce the risk and effects of co-occurring youth violence and substance use among at-risk young Black males aged 15-24. BrotherlyACT integrates life-skills coaching, safety planning tools, AI-powered talk therapy, and zip-code-enabled service connections.

Objective:

To evaluate the learnability, heuristics, and usability of BrotherlyACT with target users and mHealth experts.

Methods:

Using convergent mixed techniques, we extensively evaluated the high-fidelity BrotherlyACT using heuristic evaluation by our internal team of mHealth specialists and UI/UX professionals (n=8) and formative usability testing (UT) by end users (n=15). UT involved a think-aloud approach, cognitive walkthroughs, and usability assessment with the System Usability Scale (SUS). Overall, UT findings prompted prototype revisions and a heuristic evaluation (HE). In the HE, mHealth experts assessed the high-fidelity BrotherlyACT app using Nielsen's 10 principles of heuristic evaluation. Both forms of testing revealed user needs and design preferences during app development. Online and in-person testing was conducted via REDCap and Zoom.

Results:

Findings from the HE and UT evaluations highlighted the need for minor app revisions. The UT findings prompted prototype revision and HE. Usability problems were classified into eight thematic categories. Overall, the study identified minor usability problems as the app achieved an average SUS score of 79, corresponding to an A- grade, placing the app in the 85th percentile and reflecting its near-excellent usability. In contrast, HE by app testers identified only minor errors (e.g., with a score of 1.5, which was the highest violation on the severity ranking scale).

Conclusions:

The app received a positive overall rating from testers, who highlighted the importance of featuring Black actors in video content, incorporating mental health support resources, improving navigation with user-friendly interfaces, offering customizable dashboard options, and using 'real-person' graphics in videos. A pilot and feasibility study will next assess the app’s clinical efficacy in community health and emergency care settings. Clinical Trial: NCT06359990