Tom Lee Osborn
Research Subjects
Tom Lee Osborn has a total of 113 co-authors. Here are the top co-authors by number of shared publications:
John R. Weisz
Department of Psychology, Harvard University, Cambridge, USA
12 shared publications- Akash R. Wasil
University of Pennsylvania, Philadelphia
7 shared publications Christine Wasanga
7 shared publicationsElizabeth Roe
Department of Psychology, Harvard University, Cambridge, USA
7 shared publicationsJessica L. Schleider
Department of Psychology, Stony Brook University, Stony Brook, New York
7 shared publications- Katherine E. Venturo-Conerly
Department of Psychology, Harvard University
7 shared publications
Latest Publications
- Open Access
- 3 June 2026
- Published by JMIR Publications Inc. in JMIR Human Factors
- Vol. 13, e79107
Early Deployment of an Integrated Digital Platform (shamiriOS) for Scalable Youth Mental Health Service Delivery in Kenya: Development and Usability Study
Background: Scaling youth mental health services in low-resource settings requires digital infrastructure that supports not just clinical delivery but the full operational, supervisory, and engagement demands of community-based, task-shifted models. Existing platforms—whether commercial health systems, open-source medical records, or consumer-facing wellness apps—address fragments of this need, but none provide the integrated, offline-capable, and affordable architecture required for lay-provider delivery at scale. Objective: We introduce shamiriOS, an open-source, modular digital platform comprising 3 interlinked suites—the Shamiri Digital Hub (SDH) for operational management, Rafi for youth engagement, and the Shamiri Provider Platform (SPP) for clinical workflows—designed for scalable, stepped-care youth mental health delivery in Kenya. Our objectives were (1) to conduct an environmental scan of existing platforms and characterize their limitations, (2) to describe the user-centered design and development of shamiriOS, and (3) to report early deployment outcomes across centralized and decentralized settings between 2023 and 2024. Methods: We conducted a structured environmental scan of 6 case management platforms and 28 youth-facing mental health apps, assessing cost, usability, open-source availability, customizability, offline capability, and suitability for task-shifted delivery. Based on identified gaps, SDH was built as a browser-based operation platform, and Rafi was developed as a native mobile app (Android or iOS) with an offline-first architecture. The SPP was adapted from an existing electronic medical record system. Development followed a user-centered design process with community consultation, including cocreation workshops with 77 university-aged youths. Deployment was evaluated using use analytics, usability ratings, and Net Promoter Scores. Results: No reviewed platform met the combined requirements for stepped-care delivery. SDH was deployed across 11 sites serving 76,344 youths via 1195 lay providers and 111 clinical supervisors by Q1 2024. Staff reported high satisfaction (usability: mean 8.36/10, SD 1.49; Net Promoter Score: mean 8.63/10, SD 1.46). Rafi achieved 74.7% (n=3737) registration at Mount Kenya University, with 50.4% (508/1008) booking therapy sessions, but sustained engagement with self-guided features declined to near 0 by 9 months. Conclusions: shamiriOS demonstrates the feasibility of building modular, open-source digital infrastructure for scalable, task-shifted youth mental health delivery. Its component-based architecture is designed for adaptation to other contexts, though extension would require participatory redesign. The most significant obstacles to impact lie not in platform design but in implementation readiness, incentive alignment, and institutional integration. Future priorities include SPP deployment, artificial intelligence–assisted supervision features (shamiriAI), and strengthening sustained engagement.
- 3 June 2026
- Published by JMIR Publications Inc. in JMIR Human Factors
- Vol. 13, e79107
Open AccessEarly Deployment of an Integrated Digital Platform (shamiriOS) for Scalable Youth Mental Health Service Delivery in Kenya: Development and Usability Study
Background: Scaling youth mental health services in low-resource settings requires digital infrastructure that supports not just clinical delivery but the full operational, supervisory, and engagement demands of community-based, task-shifted models. Existing platforms—whether commercial health systems, open-source medical records, or consumer-facing wellness apps—address fragments of this need, but none provide the integrated, offline-capable, and affordable architecture required for lay-provider delivery at scale. Objective: We introduce shamiriOS, an open-source, modular digital platform comprising 3 interlinked suites—the Shamiri Digital Hub (SDH) for operational management, Rafi for youth engagement, and the Shamiri Provider Platform (SPP) for clinical workflows—designed for scalable, stepped-care youth mental health delivery in Kenya. Our objectives were (1) to conduct an environmental scan of existing platforms and characterize their limitations, (2) to describe the user-centered design and development of shamiriOS, and (3) to report early deployment outcomes across centralized and decentralized settings between 2023 and 2024. Methods: We conducted a structured environmental scan of 6 case management platforms and 28 youth-facing mental health apps, assessing cost, usability, open-source availability, customizability, offline capability, and suitability for task-shifted delivery. Based on identified gaps, SDH was built as a browser-based operation platform, and Rafi was developed as a native mobile app (Android or iOS) with an offline-first architecture. The SPP was adapted from an existing electronic medical record system. Development followed a user-centered design process with community consultation, including cocreation workshops with 77 university-aged youths. Deployment was evaluated using use analytics, usability ratings, and Net Promoter Scores. Results: No reviewed platform met the combined requirements for stepped-care delivery. SDH was deployed across 11 sites serving 76,344 youths via 1195 lay providers and 111 clinical supervisors by Q1 2024. Staff reported high satisfaction (usability: mean 8.36/10, SD 1.49; Net Promoter Score: mean 8.63/10, SD 1.46). Rafi achieved 74.7% (n=3737) registration at Mount Kenya University, with 50.4% (508/1008) booking therapy sessions, but sustained engagement with self-guided features declined to near 0 by 9 months. Conclusions: shamiriOS demonstrates the feasibility of building modular, open-source digital infrastructure for scalable, task-shifted youth mental health delivery. Its component-based architecture is designed for adaptation to other contexts, though extension would require participatory redesign. The most significant obstacles to impact lie not in platform design but in implementation readiness, incentive alignment, and institutional integration. Future priorities include SPP deployment, artificial intelligence–assisted supervision features (shamiriAI), and strengthening sustained engagement.
- Open Access
- 10 March 2026
- Published by JMIR Publications Inc. in JMIR AI
- Vol. 5, e95063
Automated Fidelity Monitoring of Lay-Delivered Mental Health Interventions Using Large Language Models: Development and Pilot Validation of shamiriAI (Preprint)
Background: Task-shifting can help close the mental health treatment gap in low- and middle-income countries, but its effectiveness depends on ongoing supervision, which is hard to scale. AI tools that process session recordings and generate structured fidelity feedback could offer a scalable alternative; yet, to our knowledge, none have been developed or validated for lay-delivered, multilingual, group-format interventions in low-resource settings. Objective: We developed and pilot-validated shamiriAI (ShamiriAI Institute), an automated fidelity-monitoring tool for lay-delivered mental health interventions, embedded within the Shamiri school-based program in Kenya. Methods: Across 6 secondary schools in Ngong Hub, Kajiado County, Kenya (May-September 2025), shamiriAI processed session audio from 47 lay providers through a 5-stage pipeline: ingestion, multilingual automatic speech recognition (ASR) with prosodic feature extraction, personally identifiable information scrubbing, large language model–based fidelity inference, and supervisor reports. The following two aims were assessed: (1) ASR performance on a held-out test set of manually transcribed sessions and (2) interrater reliability between shamiriAI and independent human supervisor ratings across 52 sessions (38 AI-augmented and 14 standard) on 6 domains (Required Contents, Specifics, Thoroughness, Clarity, Skill, and Purity; 1‐7 scale). Reliability used intraclass correlation coefficients, Bland-Altman analysis, adjacent-agreement rates, paired t tests with Holm-Bonferroni correction, and Gwet AC2 (ordinal weights) across 3 formulations of the human reference. Results: The ASR model achieved a character error rate of 0.19, word error rate of 0.34, and cosine semantic similarity of 0.77, indicating strong meaning preservation in code-switched speech. AI fidelity scores were systematically lower than the human composite overall (mean 5.14, SD 0.77 vs mean 5.93, SD 0.57 ); Δ=−0.79; d=−1.16; P <.001). Primary intraclass correlation coefficients ranged from −0.06 to 0.20 across the 6 domains, and AC2 sensitivity analyses (against each individual rater and the rounded composite) corroborated this dimension-level ordering. Three patterns emerged: large systematic underrating on holistic dimensions (Required Contents: d=−3.48; Clarity d=−1.56); bidirectional medium-effect bias on facilitation dimensions (Thoroughness d=−0.99; Skill d=+0.87); and substantial agreement on Specifics, Skill, and Purity (Gwet AC2 0.69‐0.76 against the rounded composite), relative to a human-human AC2 ceiling of 0.42‐0.60 estimated in the same dataset. Exploratory, underpowered subgroup and per-arm checks found no preliminary evidence of bias by lay-provider sex or age band or of arm-level differences. Conclusions: Within this 52-session pilot, shamiriAI shows technically feasible multilingual ASR and a coherent, dimension-dependent reliability profile. Specifics, Purity, and Skill already reach substantial agreement, while underperformance on holistic dimensions (Required Contents and Clarity) reflects diagnosable misalignments in rubric interpretation and prompt design, specifying a concrete agenda for shamiriAI (version 2; Shamiri Institute). Whether AI-augmented supervision improves provider skill or student mental health outcomes will be tested in a planned cluster-randomized noninferiority trial.
- 10 March 2026
- Published by JMIR Publications Inc. in JMIR AI
- Vol. 5, e95063
Open AccessAutomated Fidelity Monitoring of Lay-Delivered Mental Health Interventions Using Large Language Models: Development and Pilot Validation of shamiriAI (Preprint)
Background: Task-shifting can help close the mental health treatment gap in low- and middle-income countries, but its effectiveness depends on ongoing supervision, which is hard to scale. AI tools that process session recordings and generate structured fidelity feedback could offer a scalable alternative; yet, to our knowledge, none have been developed or validated for lay-delivered, multilingual, group-format interventions in low-resource settings. Objective: We developed and pilot-validated shamiriAI (ShamiriAI Institute), an automated fidelity-monitoring tool for lay-delivered mental health interventions, embedded within the Shamiri school-based program in Kenya. Methods: Across 6 secondary schools in Ngong Hub, Kajiado County, Kenya (May-September 2025), shamiriAI processed session audio from 47 lay providers through a 5-stage pipeline: ingestion, multilingual automatic speech recognition (ASR) with prosodic feature extraction, personally identifiable information scrubbing, large language model–based fidelity inference, and supervisor reports. The following two aims were assessed: (1) ASR performance on a held-out test set of manually transcribed sessions and (2) interrater reliability between shamiriAI and independent human supervisor ratings across 52 sessions (38 AI-augmented and 14 standard) on 6 domains (Required Contents, Specifics, Thoroughness, Clarity, Skill, and Purity; 1‐7 scale). Reliability used intraclass correlation coefficients, Bland-Altman analysis, adjacent-agreement rates, paired t tests with Holm-Bonferroni correction, and Gwet AC2 (ordinal weights) across 3 formulations of the human reference. Results: The ASR model achieved a character error rate of 0.19, word error rate of 0.34, and cosine semantic similarity of 0.77, indicating strong meaning preservation in code-switched speech. AI fidelity scores were systematically lower than the human composite overall (mean 5.14, SD 0.77 vs mean 5.93, SD 0.57 ); Δ=−0.79; d=−1.16; P <.001). Primary intraclass correlation coefficients ranged from −0.06 to 0.20 across the 6 domains, and AC2 sensitivity analyses (against each individual rater and the rounded composite) corroborated this dimension-level ordering. Three patterns emerged: large systematic underrating on holistic dimensions (Required Contents: d=−3.48; Clarity d=−1.56); bidirectional medium-effect bias on facilitation dimensions (Thoroughness d=−0.99; Skill d=+0.87); and substantial agreement on Specifics, Skill, and Purity (Gwet AC2 0.69‐0.76 against the rounded composite), relative to a human-human AC2 ceiling of 0.42‐0.60 estimated in the same dataset. Exploratory, underpowered subgroup and per-arm checks found no preliminary evidence of bias by lay-provider sex or age band or of arm-level differences. Conclusions: Within this 52-session pilot, shamiriAI shows technically feasible multilingual ASR and a coherent, dimension-dependent reliability profile. Specifics, Purity, and Skill already reach substantial agreement, while underperformance on holistic dimensions (Required Contents and Clarity) reflects diagnosable misalignments in rubric interpretation and prompt design, specifying a concrete agenda for shamiriAI (version 2; Shamiri Institute). Whether AI-augmented supervision improves provider skill or student mental health outcomes will be tested in a planned cluster-randomized noninferiority trial.
- Open Access
- 10 March 2026
- Published by JMIR Publications Inc. in JMIR Preprints
Automated Fidelity Monitoring of Lay-Delivered Mental Health Interventions Using Large Language Models: Development and Pilot Validation of shamiriAI (Preprint)
BACKGROUND Task-shifting—the delivery of evidence-based mental health interventions by trained lay providers—has shown promise in closing the treatment gap in low- and middle-income countries. But the effectiveness of task-shifted interventions depends critically on ongoing supervision and monitoring, yet traditional supervision models are difficult to scale. Artificial intelligence (AI) tools capable of automatically processing session recordings and generating structured fidelity feedback for supervisors could offer a scalable alternative, but no such system has been developed or validated for lay-delivered interventions in multilingual, low-resource settings. OBJECTIVE We developed and pilot-validated shamiriAI, an automated fidelity monitoring tool for lay-delivered mental health interventions, embedded within the Shamiri school-based mental health program in Kenya. METHODS We conducted a pilot validation study across six secondary schools in Ngong Hub, Kajiado County, Kenya (May–September 2025). shamiriAI follows a five-stage pipeline: audio ingestion and preprocessing, multilingual automatic speech recognition (ASR) with prosodic feature extraction, personal identifiable information (PII) scrubbing, large language model (LLM)-based fidelity inference, and structured feedback report delivery to supervisors. We pursued two pilot aims: (1) ASR performance on a held-out test set of manually transcribed sessions; and (2) interrater reliability between shamiriAI-generated fidelity ratings and independent human supervisor ratings across 52 recorded sessions, spanning six domains (Required Contents, Specifics, Thoroughness, Clarity, Skill, Purity) rated on a 1–7 scale. Reliability was assessed using intraclass correlation coefficients (ICC), Bland-Altman analysis, adjacent agreement rates, paired-sample t-tests with Holm–Bonferroni correction, and Gwet's AC2. RESULTS The ASR model achieved a Character Error Rate of 0.19, Word Error Rate of 0.34, and cosine semantic similarity of 0.77, indicating strong meaning preservation despite surface-level transcription errors in code-switched speech. On fidelity ratings, AI scores were systematically lower than the human composite overall (M = 5.14, SD = 0.77 vs. M = 5.93, SD = 0.57; Δ = −0.79, 95% CI [−1.04, −0.53], d = −1.16, p < .001). Reliability varied markedly by dimension: ICCs ranged from −0.06 to 0.20 across all six domains. Three distinct patterns emerged: large systematic underrating on holistic interpretive dimensions (Required Contents d = −3.48; Clarity d = −1.56); a bidirectional medium-effect pattern on facilitation dimensions (Thoroughness d = −0.99; Skill d = +0.87); and no significant bias on structured detection dimensions (Specifics 78.8% adjacent agreement; Purity 73.1% adjacent agreement), where performance approached the human–human AC2 benchmark of 0.42–0.60. CONCLUSIONS shamiriAI demonstrates technically feasible multilingual ASR and a coherent, interpretable reliability profile. Underperformance was concentrated on holistic inferential dimensions — particularly Required Contents and Clarity — while structured detection tasks already approach operational utility. The underperformance pattern reflects diagnosable misalignments in rubric interpretation and prompt design, with clear engineering solutions identified. These findings provide the foundational validation evidence and dimension-specific diagnostics needed to guide the development of AI-augmented supervision for lay-delivered adolescent mental health programs in sub-Saharan Africa and other multilingual settings.
- 10 March 2026
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessAutomated Fidelity Monitoring of Lay-Delivered Mental Health Interventions Using Large Language Models: Development and Pilot Validation of shamiriAI (Preprint)
BACKGROUND Task-shifting—the delivery of evidence-based mental health interventions by trained lay providers—has shown promise in closing the treatment gap in low- and middle-income countries. But the effectiveness of task-shifted interventions depends critically on ongoing supervision and monitoring, yet traditional supervision models are difficult to scale. Artificial intelligence (AI) tools capable of automatically processing session recordings and generating structured fidelity feedback for supervisors could offer a scalable alternative, but no such system has been developed or validated for lay-delivered interventions in multilingual, low-resource settings. OBJECTIVE We developed and pilot-validated shamiriAI, an automated fidelity monitoring tool for lay-delivered mental health interventions, embedded within the Shamiri school-based mental health program in Kenya. METHODS We conducted a pilot validation study across six secondary schools in Ngong Hub, Kajiado County, Kenya (May–September 2025). shamiriAI follows a five-stage pipeline: audio ingestion and preprocessing, multilingual automatic speech recognition (ASR) with prosodic feature extraction, personal identifiable information (PII) scrubbing, large language model (LLM)-based fidelity inference, and structured feedback report delivery to supervisors. We pursued two pilot aims: (1) ASR performance on a held-out test set of manually transcribed sessions; and (2) interrater reliability between shamiriAI-generated fidelity ratings and independent human supervisor ratings across 52 recorded sessions, spanning six domains (Required Contents, Specifics, Thoroughness, Clarity, Skill, Purity) rated on a 1–7 scale. Reliability was assessed using intraclass correlation coefficients (ICC), Bland-Altman analysis, adjacent agreement rates, paired-sample t-tests with Holm–Bonferroni correction, and Gwet's AC2. RESULTS The ASR model achieved a Character Error Rate of 0.19, Word Error Rate of 0.34, and cosine semantic similarity of 0.77, indicating strong meaning preservation despite surface-level transcription errors in code-switched speech. On fidelity ratings, AI scores were systematically lower than the human composite overall (M = 5.14, SD = 0.77 vs. M = 5.93, SD = 0.57; Δ = −0.79, 95% CI [−1.04, −0.53], d = −1.16, p < .001). Reliability varied markedly by dimension: ICCs ranged from −0.06 to 0.20 across all six domains. Three distinct patterns emerged: large systematic underrating on holistic interpretive dimensions (Required Contents d = −3.48; Clarity d = −1.56); a bidirectional medium-effect pattern on facilitation dimensions (Thoroughness d = −0.99; Skill d = +0.87); and no significant bias on structured detection dimensions (Specifics 78.8% adjacent agreement; Purity 73.1% adjacent agreement), where performance approached the human–human AC2 benchmark of 0.42–0.60. CONCLUSIONS shamiriAI demonstrates technically feasible multilingual ASR and a coherent, interpretable reliability profile. Underperformance was concentrated on holistic inferential dimensions — particularly Required Contents and Clarity — while structured detection tasks already approach operational utility. The underperformance pattern reflects diagnosable misalignments in rubric interpretation and prompt design, with clear engineering solutions identified. These findings provide the foundational validation evidence and dimension-specific diagnostics needed to guide the development of AI-augmented supervision for lay-delivered adolescent mental health programs in sub-Saharan Africa and other multilingual settings.
- Open Access
- 6 February 2026
- Published by Cambridge University Press (CUP) in Cambridge Prisms: Global Mental Health
- Vol. 13, 1-30
Toward integrating clinical and non-clinical associates of suicidality to inform potential intervention points among youth in Nairobi metropolitan, Kenya
Suicide is a significant global public health concern, particularly among adolescents, with substantial implications for economies, societies and individuals’ mental well-being. Understanding its patterns and intention and psychosocial determinants in a given context can suggest potential intervention points. This population-based cross-sectional study aimed to document suicidal ideas, behaviors and intensity among youths aged 14 to 25 in the Nairobi metropolitan area and associated socio-economic position, demographic indicators and potential intervention points. A diverse sample of 1,972 participants was recruited from urban and peri-urban settings within the Nairobi metropolitan area. Data analysis included descriptive statistics, chi-square tests and logistic regression. Our findings confirm a high prevalence of suicidal ideas and behavior in the youth (19.9% and 3.6%, respectively), with very few significant differences between the urban and peri-urban areas. The severity of suicidal ideation and behavior reported methods and reasons, and the socio-demographic profile of participants, point to multiple potential intervention targets. These findings ought to be used to design, manage and evaluate suicide prevention programs.
- 6 February 2026
- Published by Cambridge University Press (CUP) in Cambridge Prisms: Global Mental Health
- Vol. 13, 1-30
Open AccessToward integrating clinical and non-clinical associates of suicidality to inform potential intervention points among youth in Nairobi metropolitan, Kenya
Suicide is a significant global public health concern, particularly among adolescents, with substantial implications for economies, societies and individuals’ mental well-being. Understanding its patterns and intention and psychosocial determinants in a given context can suggest potential intervention points. This population-based cross-sectional study aimed to document suicidal ideas, behaviors and intensity among youths aged 14 to 25 in the Nairobi metropolitan area and associated socio-economic position, demographic indicators and potential intervention points. A diverse sample of 1,972 participants was recruited from urban and peri-urban settings within the Nairobi metropolitan area. Data analysis included descriptive statistics, chi-square tests and logistic regression. Our findings confirm a high prevalence of suicidal ideas and behavior in the youth (19.9% and 3.6%, respectively), with very few significant differences between the urban and peri-urban areas. The severity of suicidal ideation and behavior reported methods and reasons, and the socio-demographic profile of participants, point to multiple potential intervention targets. These findings ought to be used to design, manage and evaluate suicide prevention programs.
- Open Access
- 20 December 2025
- Published by Springer Nature in BMC Psychology
- Vol. 14 (1)
Efficacy and acceptability of a blended intervention for emotion regulation (MAISHA) among young people in Kenya: study protocol for a cluster RCT
Background: Adolescence is a crucial developmental period characterized by increased challenges in emotion regulation, a key mechanism linked to mental health. In Kenya, one in every two adolescents and youth faces a mental health challenge, highlighting a need for early interventions targeting emotion regulation. This study will evaluate the efficacy and acceptability of MAISHA, a blended indicated mental health intervention co-designed with young people to strengthen emotion regulation among those experiencing related difficulties. Methods: In a parallel group, cluster-randomized 2 × 2 factorial controlled trial, 666 young people aged 13–25 will be randomly allocated to one of four 4-week study arms at a ratio of 1:1:1:1. (1) digital app blended with a face-to-face element; (2) digital app only; (3) face-to-face element only; or (4) care as usual (CAU). The digital app will build on an ecological momentary intervention. The face-to-face element will consist of four group sessions on growth mindset, values, and gratitude. The primary outcome is emotion regulation difficulties, assessed post-intervention using the short form of the Difficulties in Emotion Regulation Scale (DERS-SF), and comparing the blended intervention (digital and face-to-face) to CAU. Secondary measures include emotional reactivity, emotional distress, positive and negative affect, acceptability, desirability, safety, and usability. The trial will include an embedded micro-randomized trial and a process evaluation. Discussion: This trial advances adolescent global mental health by strengthening evidence on youth-focused, mechanism-informed, blended interventions while also highlighting participatory approaches in LMICs. Trial registration: The trial was registered in the Pan African Clinical Trials Registry on July 17th, 2025, as PACTR202507903096871 ( https://pactr.samrc.ac.za/TrialDisplay.aspx?TrialID=34896 ).
- 20 December 2025
- Published by Springer Nature in BMC Psychology
- Vol. 14 (1)
Open AccessEfficacy and acceptability of a blended intervention for emotion regulation (MAISHA) among young people in Kenya: study protocol for a cluster RCT
Background: Adolescence is a crucial developmental period characterized by increased challenges in emotion regulation, a key mechanism linked to mental health. In Kenya, one in every two adolescents and youth faces a mental health challenge, highlighting a need for early interventions targeting emotion regulation. This study will evaluate the efficacy and acceptability of MAISHA, a blended indicated mental health intervention co-designed with young people to strengthen emotion regulation among those experiencing related difficulties. Methods: In a parallel group, cluster-randomized 2 × 2 factorial controlled trial, 666 young people aged 13–25 will be randomly allocated to one of four 4-week study arms at a ratio of 1:1:1:1. (1) digital app blended with a face-to-face element; (2) digital app only; (3) face-to-face element only; or (4) care as usual (CAU). The digital app will build on an ecological momentary intervention. The face-to-face element will consist of four group sessions on growth mindset, values, and gratitude. The primary outcome is emotion regulation difficulties, assessed post-intervention using the short form of the Difficulties in Emotion Regulation Scale (DERS-SF), and comparing the blended intervention (digital and face-to-face) to CAU. Secondary measures include emotional reactivity, emotional distress, positive and negative affect, acceptability, desirability, safety, and usability. The trial will include an embedded micro-randomized trial and a process evaluation. Discussion: This trial advances adolescent global mental health by strengthening evidence on youth-focused, mechanism-informed, blended interventions while also highlighting participatory approaches in LMICs. Trial registration: The trial was registered in the Pan African Clinical Trials Registry on July 17th, 2025, as PACTR202507903096871 ( https://pactr.samrc.ac.za/TrialDisplay.aspx?TrialID=34896 ).
- Open Access
- 20 October 2025
- Published by Cambridge University Press (CUP) in Cambridge Prisms: Global Mental Health
- Vol. 12, 1-28
Implementation of a school-based risk management protocol within a task-shifted mental healthcare model
Adolescent mental health problems are prevalent in low- and middle-income countries, like Kenya, where access to care remains severely limited. Task-shifted, school-based interventions offer solutions but often lack structured protocols for managing risk, such as suicidality or abuse. The Shamiri Risk Management Protocol (Shamiri-RMP) was developed to address this gap through a tiered system for screening, classifying and responding to risk within a stepped-care mental health model. We conducted a mixed-methods implementation study across 149 public high schools in Kenya. Caseworker fidelity and risk classification accuracy were evaluated through a review of 222 student cases. The Consolidated Framework for Implementation Research guided the qualitative analysis of caseworker surveys to identify implementation barriers and facilitators. Of 76,855 students enrolled in the broader Shamiri program, 977 (1.27%) were referred for risk assessment, and 222 (0.28%) were enrolled in the Shamiri-RMP. Among them, 42.71% were low-risk, 35.68% moderate-risk and 21.61% high-risk. Risk reductions occurred in 60.47% of high-risk cases, 56.34% of moderate-risk cases and 51.76% of low-risk cases. Implementation facilitators included supervisory support (50.88% of caseworkers) and protocol clarity (80.70%), while barriers included referral gaps (5.26%) and confidentiality concerns (54.39%). Findings support the feasibility and scalability of the Shamiri-RMP in low-resource school settings.
- 20 October 2025
- Published by Cambridge University Press (CUP) in Cambridge Prisms: Global Mental Health
- Vol. 12, 1-28
Open AccessImplementation of a school-based risk management protocol within a task-shifted mental healthcare model
Adolescent mental health problems are prevalent in low- and middle-income countries, like Kenya, where access to care remains severely limited. Task-shifted, school-based interventions offer solutions but often lack structured protocols for managing risk, such as suicidality or abuse. The Shamiri Risk Management Protocol (Shamiri-RMP) was developed to address this gap through a tiered system for screening, classifying and responding to risk within a stepped-care mental health model. We conducted a mixed-methods implementation study across 149 public high schools in Kenya. Caseworker fidelity and risk classification accuracy were evaluated through a review of 222 student cases. The Consolidated Framework for Implementation Research guided the qualitative analysis of caseworker surveys to identify implementation barriers and facilitators. Of 76,855 students enrolled in the broader Shamiri program, 977 (1.27%) were referred for risk assessment, and 222 (0.28%) were enrolled in the Shamiri-RMP. Among them, 42.71% were low-risk, 35.68% moderate-risk and 21.61% high-risk. Risk reductions occurred in 60.47% of high-risk cases, 56.34% of moderate-risk cases and 51.76% of low-risk cases. Implementation facilitators included supervisory support (50.88% of caseworkers) and protocol clarity (80.70%), while barriers included referral gaps (5.26%) and confidentiality concerns (54.39%). Findings support the feasibility and scalability of the Shamiri-RMP in low-resource school settings.
- Open Access
- 16 June 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Early Deployment of an Integrated Digital Platform (shamiriOS) for Scalable Youth Mental Health Service Delivery in Kenya: Development and Usability Study (Preprint)
BACKGROUND Scaling youth mental health services in low-resource settings requires digital infrastructure that supports not just clinical delivery but the full operational, supervisory, and engagement demands of community-based, task-shifted models. Existing platforms—whether commercial health systems, open-source medical records, or consumer-facing wellness apps—address fragments of this need, but none provide the integrated, offline-capable, and affordable architecture required for lay-provider delivery at scale. OBJECTIVE We introduce shamiriOS, an open-source, modular digital platform comprising 3 interlinked suites—the Shamiri Digital Hub (SDH) for operational management, Rafi for youth engagement, and the Shamiri Provider Platform (SPP) for clinical workflows—designed for scalable, stepped-care youth mental health delivery in Kenya. Our objectives were (1) to conduct an environmental scan of existing platforms and characterize their limitations, (2) to describe the user-centered design and development of shamiriOS, and (3) to report early deployment outcomes across centralized and decentralized settings between 2023 and 2024. METHODS We conducted a structured environmental scan of 6 case management platforms and 28 youth-facing mental health apps, assessing cost, usability, open-source availability, customizability, offline capability, and suitability for task-shifted delivery. Based on identified gaps, SDH was built as a browser-based operation platform, and Rafi was developed as a native mobile app (Android or iOS) with an offline-first architecture. The SPP was adapted from an existing electronic medical record system. Development followed a user-centered design process with community consultation, including cocreation workshops with 77 university-aged youths. Deployment was evaluated using use analytics, usability ratings, and Net Promoter Scores. RESULTS No reviewed platform met the combined requirements for stepped-care delivery. SDH was deployed across 11 sites serving 76,344 youths via 1195 lay providers and 111 clinical supervisors by Q1 2024. Staff reported high satisfaction (usability: mean 8.36/10, SD 1.49; Net Promoter Score: mean 8.63/10, SD 1.46). Rafi achieved 74.7% (n=3737) registration at Mount Kenya University, with 50.4% (508/1008) booking therapy sessions, but sustained engagement with self-guided features declined to near 0 by 9 months. CONCLUSIONS shamiriOS demonstrates the feasibility of building modular, open-source digital infrastructure for scalable, task-shifted youth mental health delivery. Its component-based architecture is designed for adaptation to other contexts, though extension would require participatory redesign. The most significant obstacles to impact lie not in platform design but in implementation readiness, incentive alignment, and institutional integration. Future priorities include SPP deployment, artificial intelligence–assisted supervision features (shamiriAI), and strengthening sustained engagement.
- 16 June 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessEarly Deployment of an Integrated Digital Platform (shamiriOS) for Scalable Youth Mental Health Service Delivery in Kenya: Development and Usability Study (Preprint)
BACKGROUND Scaling youth mental health services in low-resource settings requires digital infrastructure that supports not just clinical delivery but the full operational, supervisory, and engagement demands of community-based, task-shifted models. Existing platforms—whether commercial health systems, open-source medical records, or consumer-facing wellness apps—address fragments of this need, but none provide the integrated, offline-capable, and affordable architecture required for lay-provider delivery at scale. OBJECTIVE We introduce shamiriOS, an open-source, modular digital platform comprising 3 interlinked suites—the Shamiri Digital Hub (SDH) for operational management, Rafi for youth engagement, and the Shamiri Provider Platform (SPP) for clinical workflows—designed for scalable, stepped-care youth mental health delivery in Kenya. Our objectives were (1) to conduct an environmental scan of existing platforms and characterize their limitations, (2) to describe the user-centered design and development of shamiriOS, and (3) to report early deployment outcomes across centralized and decentralized settings between 2023 and 2024. METHODS We conducted a structured environmental scan of 6 case management platforms and 28 youth-facing mental health apps, assessing cost, usability, open-source availability, customizability, offline capability, and suitability for task-shifted delivery. Based on identified gaps, SDH was built as a browser-based operation platform, and Rafi was developed as a native mobile app (Android or iOS) with an offline-first architecture. The SPP was adapted from an existing electronic medical record system. Development followed a user-centered design process with community consultation, including cocreation workshops with 77 university-aged youths. Deployment was evaluated using use analytics, usability ratings, and Net Promoter Scores. RESULTS No reviewed platform met the combined requirements for stepped-care delivery. SDH was deployed across 11 sites serving 76,344 youths via 1195 lay providers and 111 clinical supervisors by Q1 2024. Staff reported high satisfaction (usability: mean 8.36/10, SD 1.49; Net Promoter Score: mean 8.63/10, SD 1.46). Rafi achieved 74.7% (n=3737) registration at Mount Kenya University, with 50.4% (508/1008) booking therapy sessions, but sustained engagement with self-guided features declined to near 0 by 9 months. CONCLUSIONS shamiriOS demonstrates the feasibility of building modular, open-source digital infrastructure for scalable, task-shifted youth mental health delivery. Its component-based architecture is designed for adaptation to other contexts, though extension would require participatory redesign. The most significant obstacles to impact lie not in platform design but in implementation readiness, incentive alignment, and institutional integration. Future priorities include SPP deployment, artificial intelligence–assisted supervision features (shamiriAI), and strengthening sustained engagement.
- Open Access
- 13 August 2023
- Published by Springer Nature in Trials
- Vol. 24 (1) , 1-14
Testing pathways to scale: study protocol for a three-arm randomized controlled trial of a centralized and a decentralized (“Train the Trainers”) dissemination of a mental health program for Kenyan adolescents
Background: Providing care in Kenya to all youth in need is difficult because of a shortage of professional providers and societal stigma. Previous trials of the Anansi model, which involves delivering low-touch mental health interventions through a tiered caregiving model (including lay-providers, supervisors, and clinical experts), have shown its effectiveness for reducing depression and anxiety symptoms in school-going Kenyan adolescents. In this trial, we aim to assess two different scale-up strategies by comparing centralized implementation (i.e., by the organization that designed the Anansi model) against implementation through an implementing partner. Methods: In this three-arm trial, 1600 adolescents aged 13 to 20 years will be randomized to receive the Shamiri intervention from either the Shamiri Institute or an implementation partner or to be placed in the treatment as usual (TAU) control group. The implementation partner will be trained and supplied with protocols to ensure that the same procedures are followed by both implementors. Implementation activities will run concurrently for both implementors. The Shamiri intervention will be delivered by trained lay providers to groups of 10–15 adolescents over four weekly sessions which will take place in secondary schools in Machakos and Makueni counties in Kenya. The TAU group will receive the usual care offered by their respective schools. Outcomes will be assessed at baseline, midpoint (2 weeks), endpoint (4 weeks), and 1 month follow-up. The analysis will be based on an intent-to-treat approach. Mixed effects models will be used to assess trajectories over time of the primary outcomes (anxiety and depressive symptoms, mental well-being, perceived social support, and academic performance) and secondary outcomes for the intervention groups and the control group. Effect sizes will be computed for the mean differences of the intervention and control arms at midpoint, endpoint, and follow-up. Discussion: This trial will provide insight into the comparative effectiveness of different strategies for scaling a school-based mental health care model. Findings will also indicate areas for improved efficiency of the model to enhance its replicability by other implementors. Trial registration: Pan African Clinical Trials Registry (PACTR) (ID: PACTR202305589854478, Approved: 02/05/2023).
- 13 August 2023
- Published by Springer Nature in Trials
- Vol. 24 (1) , 1-14
Open AccessTesting pathways to scale: study protocol for a three-arm randomized controlled trial of a centralized and a decentralized (“Train the Trainers”) dissemination of a mental health program for Kenyan adolescents
Background: Providing care in Kenya to all youth in need is difficult because of a shortage of professional providers and societal stigma. Previous trials of the Anansi model, which involves delivering low-touch mental health interventions through a tiered caregiving model (including lay-providers, supervisors, and clinical experts), have shown its effectiveness for reducing depression and anxiety symptoms in school-going Kenyan adolescents. In this trial, we aim to assess two different scale-up strategies by comparing centralized implementation (i.e., by the organization that designed the Anansi model) against implementation through an implementing partner. Methods: In this three-arm trial, 1600 adolescents aged 13 to 20 years will be randomized to receive the Shamiri intervention from either the Shamiri Institute or an implementation partner or to be placed in the treatment as usual (TAU) control group. The implementation partner will be trained and supplied with protocols to ensure that the same procedures are followed by both implementors. Implementation activities will run concurrently for both implementors. The Shamiri intervention will be delivered by trained lay providers to groups of 10–15 adolescents over four weekly sessions which will take place in secondary schools in Machakos and Makueni counties in Kenya. The TAU group will receive the usual care offered by their respective schools. Outcomes will be assessed at baseline, midpoint (2 weeks), endpoint (4 weeks), and 1 month follow-up. The analysis will be based on an intent-to-treat approach. Mixed effects models will be used to assess trajectories over time of the primary outcomes (anxiety and depressive symptoms, mental well-being, perceived social support, and academic performance) and secondary outcomes for the intervention groups and the control group. Effect sizes will be computed for the mean differences of the intervention and control arms at midpoint, endpoint, and follow-up. Discussion: This trial will provide insight into the comparative effectiveness of different strategies for scaling a school-based mental health care model. Findings will also indicate areas for improved efficiency of the model to enhance its replicability by other implementors. Trial registration: Pan African Clinical Trials Registry (PACTR) (ID: PACTR202305589854478, Approved: 02/05/2023).