Digital Mental Health Interventions in Crisis, Humanitarian, and Fragile Settings in LMICs: Delivery and Humanitarian Engagement

Introduction - Objectives: Digital mental health interventions can support continuity of care when conventional services are disrupted by humanitarian crises, conflicts, epidemics, or health-system fragility. Evidence on their use and adaptation in these settings remains limited. This study examined their characteristics, delivery, and humanitarian engagement in crisis or fragile settings in LMICs. Methodology: We conducted an exploratory secondary analysis of 110 studies from an international scoping review of digital mental health interventions in LMICs. Studies were selected using predefined eligibility criteria and a standardized extraction grid, with independent verification of extracted data. Variables included crisis and fragile settings, humanitarian actor involvement, delivery modalities, study characteristics, and context-specific adaptations. Descriptive analyses summarized frequencies and proportions. Associations were assessed using Fisher’s exact test with Monte Carlo simulation (100,000 replicates), with statistical significance set at p < 0.05. Thematic analysis explored context-specific characteristics and adaptations. Results and Discussion: Of 110 studies, 27 (24.5%; 95% CI: 16.8-33.7) were conducted in crisis or fragile settings, while 14 (12.7%; 95% CI: 7.1-20.4) explicitly involved humanitarian actors. Among these 27 studies, 15 (55.6%) addressed epidemic/pandemic settings, 11 (40.7%) humanitarian crises, and one (3.7%) systemic fragility. The proportion of studies conducted in crisis/fragility settings varied by region (p=0.006): 60.0% in the Middle East/MENA and 30.3% in Africa, compared with 13.3% in Asia and 11.5% in Latin America/Caribbean. No association was identified with period, population, implementation scale, or digital modality. Remote delivery predominated in crisis/fragility settings (24/27; 88.9%). Delivery modality differed by context type (p=0.031): all studies in epidemic/pandemic settings were delivered remotely, compared with 81.8% of studies in humanitarian crisis settings. Humanitarian actor involvement also differed by context (p<0.001), occurring in all humanitarian crisis studies compared with 13.3% of studies in epidemic/pandemic settings. Qualitatively, humanitarian crisis studies frequently involved conflict/displacement (81.8%) and refugees/migrants (63.6%). Adaptations explicitly facilitating access were rarely documented. Conclusions: Digital mental health interventions have an uneven presence in crisis and fragile settings in LMICs, with marked regional variation. Remote delivery predominates, while humanitarian engagement is concentrated in humanitarian crises. Better documentation and evaluation of context-specific adaptations are needed to support accessible mental health care during crises.