Tanzania:
Harnessing AI for Rapid Public Health Response: A Pilot Case Study of Mpox in Tanzania
To receive this newsletter, click here.
Harnessing AI for Rapid Public Health Response: A Pilot Case Study of Mpox in Tanzania
April 2025 — In August 2024, the World Health Organization (WHO) identified a regional public health threat due to an upsurge of Mpox cases in the Democratic Republic of Congo (DRC). By early September, cases had spread to neighbouring countries, raising concerns about potential outbreaks in Tanzania. Mpox, a viral zoonotic disease, presents with symptoms similar to smallpox and can spread through close contact with infected individuals or contaminated materials.
The Tanzanian Ministry of Health and President’s Office, Regional Administration and Local Government (PoRALG), supported by UNICEF, aimed to educate the public and prepare health care facilities for a potential Mpox outbreak. A significant challenge in such crises is the delay between data collection and the delivery of actionable insights to decision-makers. Traditional data analysis methods can be time-consuming, hindering timely public health responses.
To address this, UNICEF and Colectiv, a behavioural science organisation, initiated a rapid assessment pilot in Tanzania. The pilot involved in-depth interviews with public health officials across six regions to gather qualitative data on community preparedness. Colectiv used an AI-assisted process to accelerate qualitative analysis. A researcher led the analysis, while AI supported faster thematic coding and exploration. This approach enabled timely insights to inform public health campaigns and enhance Tanzania's response to the Mpox threat.
Strategy and Implementation
UNICEF conducted face-to-face interviews with public health officials across six regions in Tanzania to assess local awareness and readiness for a potential Mpox outbreak. The officials interviewed included health education coordinators, health officers, disease surveillance officers, environmental health officers, and vaccine coordinators. To address delays in qualitative data analysis, UNICEF collaborated with Colectiv to apply their AI-powered qualitative analysis tools to automate the coding of interview transcripts into thematic areas and make them searchable by keyword within four hours. The process involved manual transcription to ensure accuracy, followed by AI-assisted coding into themes such as public awareness, current activities, barriers to preparedness, and emergency response. The coded data was then made searchable by theme, demographics, and keywords for quick retrieval of relevant information.
Progress and Results
The pilot involved 26 interviews with public health officials across six regions in Tanzania. Health workers viewed Mpox as a significant threat due to confirmed cases in neighbouring countries, but the public was largely unconcerned and often confused Mpox with other diseases like chickenpox.
Health workers interviewed used various communication channels to raise public awareness about Mpox, including engaging influential community leaders, utilizing local media such as radio stations, churches, mosques, and media outlets, and leveraging government resources despite concerns about insufficient funding. Public health messages focused on transmission control strategies like encouraging regular handwashing, advising against close contact such as shaking hands, hugging, and kissing, promoting safe sex practices, and providing a toll-free number (199) for reporting suspected cases.
Several barriers to community preparedness were identified, including low public awareness, confusion between Mpox and other diseases, economic and social factors that conflicted with health advice, and resource limitations that highlighted the need for more training, equipment, and supplies. Preparations for an Mpox outbreak included training healthcare staff, establishing isolation rooms and handwashing facilities, setting up rapid response teams, and addressing shortages of protective equipment and medical supplies.
Public education efforts were broad but underfunded, focusing on transmission control strategies and using various communication channels. Barriers to preparedness included low public awareness, economic and social factors, and resource limitations. Health workers had mixed views on preparedness, with some feeling preparations were sufficient and others highlighting the need for more training and supplies. Rapid response teams and community engagement were seen as positive steps, but their effectiveness depended on available resources and trained personnel.
Lessons Learned
Rapid Analysis: AI tools significantly reduced the time required to analyze qualitative data, providing timely insights for public health decision-making.
Public Education: Effective public health messaging requires collaboration with trusted community leaders and media outlets. However, more funds and resources are needed to enhance these efforts.
Behavioural Change: Social and economic factors heavily influence the effectiveness of public health advice. Sustained efforts and targeted messaging are crucial for long-term behavioural change.
Resource Allocation: Adequate training, equipment, and supplies are essential for effective emergency response. Coordination and collaboration among stakeholders can help address these gaps.
Way Forward and Potential Application
AI-Powered Rapid Analysis: AI tools, such as those developed by Colectiv, offer humanitarian organizations a new way to rapidly analyze qualitative data across countries. This enables faster, more nuanced insights to guide timely, evidence-based SBC responses during public health emergencies.
Enhanced Communication Strategy: Develop a coherent communication strategy that spans both the public and health workforce, focusing on key behaviors and ensuring consistent messaging.
Targeted Education: Extend health education to high-risk groups, such as frequent travelers and individuals with higher numbers of sexual partners, and ensure comprehensive training for the health workforce.
Support Systems: Address social and economic barriers to case isolation by providing support for dependents of those in isolation and presenting isolation as a community good.
Resource Mobilization: Establish emergency funds and improve logistics to ensure a timely and adequate response to outbreaks. Collaborate with various stakeholders to combine efforts and resources.
For more information, please contact Awet Araya, SBC/RCCE Lead, UNICEF Tanzania at aaraya@unicef.org.