JMIR Publications

ABSTRACT

Background:

Scaling youth mental health services in low- and middle-income countries (LMICs) is hindered by fragmented infrastructure, scarce human resources, and a lack of affordable, adaptable digital tools. Existing platforms are often overly clinical, expensive, and ill-suited for community-based delivery that is common in these contexts.

Objective:

Here, we introduce shamiriOS, a modular open-source platform designed to support the scalable delivery of task-shifted youth mental health interventions in LMICs. Our objectives were threefold: 1) to review existing digital platforms for suitability in this context; 2) to describe the design and the development of shamiriOS; and 3) to evaluate early outcomes from its deployment in diverse community settings across Kenya.

Methods:

We conducted a semi-structured review of digital case management systems and youth-focused mental health applications, using criteria of cost, usability, open-source accessibility, customizability, and suitability for low-resource environments. Based on this review—and other identified gaps—we designed shamiriOS as a platform with three primary components. We developed two components: the Shamiri Digital Hub (SDH), backend operations platform for implementation staff (hub coordinators), and Rafi, a mobile-based companion app design for youth users. The third component, the Shamiri Provider Platform, was adapted from an existing third-party tool. Development followed an iterative, participatory process incorporating input from software engineers, implementation staff, clinical supervisors, and youth end users. Between 2023 and 2024, the platform was piloted in six secondary schools, community hubs, and a university wellness center. We assessed performance using platform usage analytics, user satisfaction surveys, Net Promoter Scores (NPS), and engagement rates.

Results:

Our analysis found that existing platforms lacked the modularity, offline functionality, and alignment with community-based stepped-care delivery models. The SDH was deployed across six school-based hubs serving 76,344 youths receiving care from 1,195 lay-providers and 111 clinical supervisors and therapists by 2024. Supervisors and hub coordinators (n=22) reported high satisfaction (usability M=8.36, NPS M=8.63) and noted improvements in session tracking, payment workflows, and triage coordination. Rafi was deployed at Mount Kenya University, reaching 3,737 students with 50.4% booking at least one session. However, sustained engagement with digital tools such as journaling and self-check-ins remained limited. Ongoing platform development efforts include the integration of the Shamiri Provider Platform and the development of shamiriAI to support precision mental health care and therapist capacity-building.

Conclusions:

The development and deployment of shamiriOS demonstrate the feasibility of building scalable, culturally grounded digital infrastructure to support youth mental health interventions in LMICs. With its modular architecture, open-source ethos, and integration of operational, clinical, and user-facing tools, shamiriOS serves as a model for future innovations in global mental health. Future work will focus on advancing AI-driven features, strengthening sustained engagement, and preparing the platform for broader global adaptation. Clinical Trial: N/A