Early Warning & Intelligence Node (EWIN)
The EWIN initiative is CGP’s flagship model for subnational epidemic intelligence- integrating AI-powered forecasting, 7-1-7 response monitoring, and community-based early warning. Anchored in the One Health approach, EWIN pilots will demonstrate how predictive analytics, trust-building communication, and digital surveillance tools can accelerate detection and save lives in vulnerable counties
7-1-7 Performance Improvement Pilot
The WHO-endorsed 7-1-7 framework aims to strengthen outbreak detection and response by ensuring signals are detected within 7 days, notified within 1 day, and responded to within 7 days. Many countries struggle to meet these timelines due to systemic gaps in data flow, workforce capacity, and coordination. We propose a pilot implementation of the 7-1-7 model in five counties in Kenya, using CGP’s digital tools and quality improvement (QI) methods to track and enhance performance.
- Establish baseline 7-1-7 performance indicators for priority diseases.
- Deploy CGP’s tracking dashboard and incident log tools.
- Conduct joint county-level QI meetings to identify bottlenecks.
- Document success factors and lessons for national policy recommendations.
- Retrospective analysis of recent outbreaks to establish baseline.
• Training county surveillance teams on 7-1-7 framework and reporting. - Dashboard customization and integration with MOH tools.
- Monthly QI cycles involving health, veterinary, and emergency response units.
- AAR and dissemination of findings at national TWG.
- Reduced time from detection to response for priority diseases.
- Improved multisectoral coordination at county level.
- Functional 7-1-7 dashboard generating real-time metrics.
- National uptake of pilot methodology for scale-up.
AI-Powered Epidemic Forecasting
Sub-Saharan Africa faces increasing risks of emerging and re-emerging infectious diseases. Traditional surveillance systems, while improving, are still largely reactive and struggle to integrate multisectoral, community-level, and environmental data. The rise of AI and machine learning (ML) offers an opportunity to transform disease forecasting and response planning at the subnational level. We seek to pilot an AI-powered epidemic forecasting platform that leverages IDSR, CBS, and climatic or environmental signals to anticipate outbreaks, generate risk alerts, and support decision-making by frontline health teams in Kenya.
- Build and pilot a machine learning-based early warning model using multi-source surveillance data.
- Integrate forecasts into a real-time visualization dashboard accessible to county and national actors.
- Train health officers, data managers, and decision-makers to interpret model outputs for preparedness planning.
- Generate evidence on the feasibility, acceptability, and scalability of AI-driven forecasting tools in Kenya.
- Data aggregation and cleaning from IDSR, CBS, weather, and GIS sources
- AI model development and back-testing for 2 disease categories (e.g., cholera, mpox, , kalar azar, dengue).
- Dashboard design and user interface testing.
- Capacity building workshops with MOH and surveillance stakeholders.
- Dissemination of results through policy briefs and national technical forums.
- A functional prototype of an AI-based epidemic forecasting tool.
- Improved lead time for outbreak alerts by 7–14 days.
- Increased local capacity to generate, interpret, and act on forecasts.
- High-level stakeholder buy-in for potential national scale-up.
Community-Led Risk Communication in marginalized areas
Communities in remote and underserved regions—including pastoralist, informal, and border areas—face systemic challenges in accessing timely, credible, and actionable health information. Structural barriers, cultural mistrust, and inadequate integration into formal health systems hamper early detection, signal reporting, and public trust during emergencies. This concept proposes a community-driven risk communication model that leverages trusted local structures and digital innovations to strengthen early warning, rumor tracking, and community trust.
- Enhance public trust and communication pathways in at-risk marginalized communities.
- Establish real-time rumor tracking and feedback loops to support early detection.
- Build local capacity of CHVs, youth leaders, and influencers as trusted communicators.
- Integrate grassroots reporting with county surveillance and RCCE platforms.
- Community engagement and participatory design workshops.
- Training of CHVs, elders, youth, and religious leaders on RCCE and signal reporting.
- Development of culturally contextualized IEC tools and mobile messaging formats.
- Establishment of WhatsApp and SMS-based rumor-tracking channels.
- Linkage with county disease surveillance and health promotion units.
- Improved early detection through grassroots signal and rumor alerts.
- Increased uptake of public health guidance in high-risk communities.
- Institutionalization of community feedback within RCCE systems.
- Strengthened community-state trust and resilience to misinformation
Climate-Sensitive Disease Surveillance & Early Warning Systems
Climate variability and extreme weather events are increasingly influencing the emergence and transmission of infectious diseases in East Africa. Floods, droughts, and changing vector habitats are contributing to recurrent outbreaks of cholera, Rift Valley fever, and other climate-sensitive diseases. CGP proposes a pilot to integrate meteorological, environmental, and health data to build early warning and response systems for climate-sensitive outbreaks.
- Identify and prioritize climate-sensitive diseases and risk indicators.
- Establish integrated data-sharing protocols between climate and health sectors.
- Develop GIS-based risk models and early warning dashboards.
- Strengthen county-level capacity to interpret and act on climate-health forecasts.
- Desk review and stakeholder consultations on climate-disease linkages.
- Data integration from Kenya Meteorological Department, NDMA, and MoH systems.
- Development and piloting of GIS tools for disease risk prediction.
- Capacity building for surveillance officers and county disaster committees.
- Dissemination of pilot results through workshops and policy briefs.
- Improved predictive capacity for disease outbreaks linked to climate variability.
- Functioning climate-sensitive surveillance system in two pilot counties.
- Strengthened cross-sectoral coordination and risk-informed planning.
- Scalable evidence for national and regional early warning strategies.
One Health Surveillance in Urban informal settlements
Urban informal settlements in sub-Saharan Africa are home to millions living in conditions that exacerbate disease transmission: high population density, poor sanitation, and close proximity between humans, animals, and waste. Despite the elevated risk of zoonotic and epidemic-prone diseases, these areas often fall outside the reach of formal surveillance systems. CGP proposes One Health surveillance model specifically tailored to urban informal settlements, integrating community-level reporting, animal health indicators, and environmental risk factors.
- Establish sentinel One Health surveillance sites in two high-risk informal settlements.
- Enhance event-based and syndromic surveillance by training human and animal health informants.
- Improve multisectoral data sharing and coordinated response at county level.
- Generate scalable evidence for integrating One Health surveillance into urban health systems.
- Community co-design workshops to identify key disease indicators and reporting channels.
- Recruitment and training of CHVs, veterinary paraprofessionals, and local health workers.
- Deployment of mobile surveillance tools and reporting apps.
- Routine joint data review meetings involving human, animal, and environmental health actors.
- Documentation and dissemination of lessons learned.
- Improved early detection and response to outbreaks in informal urban settlements.
- Strengthened multisectoral collaboration using a One Health approach.
- A tested and documented One Health surveillance model for replication.
- Increased community participation and ownership of local health systems.