Chilling effect

Chilling effect

Reda SadkiGlobal health

We reached out to senior decision makers working in global health about the new Certificate peer learning programme for equity in research and practice.

Crickets.

One CEO wrote: “We aren’t currently in a position to enter into new strategic partnerships on the topic.”

The chilling effect is real.

Many organizations are retreating from publicly championing equity work—even those with deep commitments to fairness and inclusion.

But here’s the opportunity: While public discourse faces headwinds, meaningful work continues through trusted networks and communities of practice.

This is precisely when innovation in equity approaches accelerates—away from the spotlight but with profound impact.

The evidence is clear: health systems that neglect equity waste resources and deliver poorer outcomes.

When research excludes key populations or policies overlook certain communities, we all lose—through inefficiency, increased costs, and diminished impact.

This moment calls for courage from those who understand that equity is fundamental to effective health systems.

“The ultimate measure of a person is not where they stand in moments of comfort, but where they stand at times of challenge.” – Martin Luther King Jr.

If you’re still committed to this essential work, you’re not alone.

Question: How are you maintaining momentum on equity work during challenging times?

Image: The Geneva Learning Foundation Collection © 2025

MOOC completion rates in context

Online learning completion rates in context: Rethinking success in digital learning networks

Reda SadkiGlobal health, Learning

The comprehensive analysis of 221 Massive Open Online Courses (MOOCs) by Katy Jordan provides crucial insights for health professionals navigating the rapidly evolving landscape of digital learning. Her study, published in the International Review of Research in Open and Distributed Learning, examined completion rates across diverse platforms including Coursera, Open2Study, and others from 78 institutions. 

  • With median completion rates of just 12.6% (ranging from 0.7% to 52.1%), traditional metrics may suggest disappointment. Jordan’s multiple regression analysis revealed that while total enrollments have decreased over time, completion rates have actually increased. 
  • The data showed striking patterns in how participants engage, with the first and second weeks proving critical—after which the proportion of active students and those submitting assessments remains remarkably stable, with less than 3% difference between them. 
  • The research challenges common assumptions about “lurking” as a participation strategy and provides compelling evidence that course design factors significantly impact learning outcomes. 

These findings reveal important patterns that can transform how we approach professional learning in global health contexts.

Beyond traditional completion metrics

For global health epidemiologists accustomed to face-to-face training with financial incentives and dedicated time away from work, these completion rates might initially appear appalling. In traditional capacity building programswhere participants receive per diems, travel stipends, and paid time away from work. Outcomes such as “completion” are rarely measured. Instead, attendance remains the key metric. In fact, completion rates are often confused with attendance. From this perspective, even the highest MOOC completion rate of 52.1% could be interpreted as a dismal failure.

However, this interpretation fundamentally misunderstands the different dynamics at play in digital learning environments. Unlike traditional training where external incentives and protected time create artificial conditions for participation, MOOCs operate in the reality of participants’ everyday professional lives. They typically do not require participants to stop work in order to learn, for example. The fact that up to half of enrollees in some courses complete them despite competing priorities, no financial incentives, and no dedicated work time represents remarkable commitment rather than failure.

What drives completion?

The data reveals three significant factors affecting completion:

  1. Course length: Shorter courses consistently achieved higher completion rates
  2. Assessment type: Auto-grading showed better completion than peer assessment
  3. Start date: More recent courses demonstrated higher completion rates

The critical engagement period occurs within the first two weeks—after which participant behavior stabilizes. This insight aligns with what emerging networked learning approaches have demonstrated in practice.

Rather than judging digital learning by metrics designed for classroom settings, we must recognize that participation patterns reflect authentic integration with professional practice. The measure of success is not just how many complete the formal course, but how learning connects to real-world problem-solving and contributes to sustained professional networks.

Moving beyond MOOCs: Health learning networks

The Geneva Learning Foundation’s approach offers a distinctly different model from traditional MOOCs. While MOOCs typically deliver standardized content to individual learners who progress independently, the Foundation’s digital learning initiatives are fundamentally network-based and practice-oriented. Rather than focusing on content consumption, their approach creates structured environments where health professionals connect, collaborate, and co-create knowledge while addressing real challenges in their work.

These learning networks differ from MOOCs in several key ways:

  • Participants engage primarily with peers rather than pre-recorded content
  • Learning is organized around actual workplace challenges rather than abstract concepts
  • The experience builds sustainable professional relationships rather than one-time course completion
  • Assessment occurs through peer review and real-world application rather than quizzes or assignments
  • Structure is provided through facilitation and process rather than predetermined pathways

The Foundation’s experience with over 60,000 health professionals across 137 countries demonstrates that when learning is connected to practice through networked approaches, different metrics of success emerge:

  • Knowledge application: Practitioners implement solutions directly in their contexts
  • Network formation: Sustainable learning relationships develop beyond formal “courses”
  • Knowledge creation: Participants contribute to collective understanding
  • System impact: Changes cascade through health systems

Implications for global health training

For epidemiologists and health professionals designing learning initiatives, these findings suggest several strategic shifts:

  1. Modular design: Create shorter, more connected learning units rather than lengthy courses
  2. Real-world integration: Link learning directly to participants’ practice contexts
  3. Peer engagement: Provide structured opportunities for health workers to learn from each other
  4. Network building: Focus on creating sustainable learning communities rather than isolated training events

The future of professional learning

The research and practice point to a fundamental evolution in how we approach professional learning in global health. Rather than replicating traditional per diem-driven training models online, the most effective approaches harness the power of networks, enabling health professionals to learn continuously through structured peer interaction.

This perspective helps explain why seemingly low completion rates should not necessarily be viewed as failure. When digital learning is designed to create lasting networks of practice—where knowledge emerges through collaborative action—completion metrics capture only a fraction of the impact.

For health systems facing complex challenges that include climate change, pandemic response, and health workforce shortages, this networked approach to learning offers a promising path forward—one that transforms how knowledge is created, shared, and applied to improve health outcomes globally.

Reference

Jordan, K., 2015. Massive open online course completion rates revisited: Assessment, length and attrition. IRRODL 16. https://doi.org/10.19173/irrodl.v16i3.2112

Sculpture: The Geneva Learning Foundation Collection © 2025

What is complex learning

What is complex learning?

Reda SadkiGlobal health

Complex learning happens when people solve real problems instead of just memorizing facts.

Think about the difference between reading about how to ride a bicycle and actually learning to ride one.

You cannot learn to ride a bicycle just by reading about it – you need to practice, fall, adjust, and try again until your body understands how to balance.

Health challenges work the same way.

Reading about how to respond to a disease outbreak is very different from actually managing one.

Complex learning recognizes this difference.

5 key features of complex learning:

  1. Learning by doing: People learn best when they work on real problems they face in their jobs. Instead of just listening to experts, they actively try solutions, see what works, and adjust their approach.
  2. No single right answer: Complex learning deals with situations where there is no perfect solution that works everywhere. What works in one community might fail in another because of different resources, cultures, or systems.
  3. Adapting to local reality: Rather than following fixed steps, complex learning helps people adapt general principles to their specific situation. A rural clinic and an urban hospital might need different approaches even when dealing with the same disease.
  4. Connecting different types of knowledge: Complex learning brings together technical knowledge (facts and procedures) with practical wisdom (experience and judgment). Both are needed to solve real health challenges.
  5. Learning from mistakes: In complex learning, mistakes are valuable opportunities to learn, not failures to be hidden. When something doesn’t work, the question becomes “What can we learn from this?” rather than “Who is to blame?”

Why it matters for health work:

Most health challenges are complex problems. Disease outbreaks, vaccination campaigns, and health system improvements all require more than just technical knowledge. They require the ability to:

  • Adapt to changing situations
  • Work with limited resources
  • Coordinate with different groups
  • Solve unexpected problems
  • Learn from experience

Complex learning builds these abilities by engaging people with real challenges, supporting them as they try solutions, and helping them reflect on what they learn.

Unlike traditional training that assumes knowledge flows from experts to learners, complex learning recognizes that knowledge emerges through practice and experience. When health workers engage with complex learning, they don’t just know more – they become better problem-solvers capable of addressing the unique challenges in their communities.

What is networked learning

What is networked learning?

Reda SadkiGlobal health

Networked learning happens when people learn through connections with others facing similar challenges. Think about how market traders learn their business – not through formal classes, but by connecting with other traders, sharing tips, and learning from each other’s experiences. This natural way of learning through relationships is what networked learning tries to support.

5 key features of networked learning:

  1. Learning from peers: In networked learning, people learn as much or more from others doing similar work as they do from experts. A community health worker in one village might discover an effective way to increase vaccination rates that could help workers in other villages.
  2. Knowledge flows in all directions: Unlike traditional training where knowledge flows only from the top down, networked learning allows knowledge to move in all directions – from national programs to local clinics, between regions, and from local implementers up to policy makers.
  3. Connections create value: The relationships between people become valuable resources for solving problems. Having a network of colleagues to ask for advice or share experiences with helps everyone work more effectively.
  4. Crossing boundaries: Networked learning connects people who might not normally work together – like doctors, nurses, community health workers, and managers. These diverse connections bring together different perspectives and create new solutions.
  5. Building on existing relationships: People already learn from colleagues they trust. Networked learning strengthens these natural connections and creates new ones, expanding who people can learn from.

Why networked learning matters for health work:

Health systems are full of isolated practitioners who could benefit from each other’s knowledge:

  • A nurse who developed an effective patient education approach
  • A community health worker who found a way to reach remote households
  • A clinic manager who improved medicine supply systems
  • A doctor who adapted treatment guidelines for local conditions

Networked learning connects these isolated pockets of knowledge, allowing good ideas to spread and adapt across different contexts.

Unlike traditional training that pulls people away from their work for workshops, networked learning happens through ongoing connections that support everyday problem-solving. When health workers participate in networked learning, they gain access to a community of practice that continues to provide support long after formal training ends.

Networked learning doesn’t replace expertise, but it recognizes that valuable knowledge exists throughout the health system – not just at the top. By connecting this distributed knowledge, networked learning helps good practices spread more quickly and adapt more effectively to local needs.

Complex problems

What is a complex problem?

Reda SadkiGlobal health

What is a complex problem and what do we need to tackle it?

Problems can be simple or complex.

Simple problems have a clear first step, a known answer, and steps you can follow to get the answer.

Complex problems do not have a single right answer.

They have many possible answers or no answer at all.

What makes complex problems really hard is that they can change over time.

They have lots of different pieces that connect in unexpected ways.

When you try to solve them, one piece changes another piece, which changes another piece.

It is hard to see all the effects of your actions.

When you do something to help, later on the problem might get worse anyway.

You have to keep adapting your ideas.

To solve really hard problems, you need to be able to:

  • Think about all the puzzle pieces and how they fit, even when you don’t know what they all are.
  • Come up with plans and change them when parts of the problem change.
  • Think back on your problem solving to get better for next time.

The most important things are being flexible, watching how every change affects other things, and learning from experience.

Image: The Geneva Learning Foundation Collection © 2024

References

Buchanan, R., 1992. Wicked problems in design thinking. Design issues 5–21.

Camillus, J.C., 2008. Strategy as a wicked problem. Harvard business review 86, 98.

Joksimovic, S., Ifenthaler, D., Marrone, R., De Laat, M., Siemens, G., 2023. Opportunities of artificial intelligence for supporting complex problem-solving: Findings from a scoping review. Computers and Education: Artificial Intelligence 4, 100138. https://doi.org/10.1016/j.caeai.2023.100138

Rittel, H.W., Webber, M.M., 1973. Dilemmas in a general theory of planning. Policy sciences 4, 155–169.

Artificial intelligence, accountability, and authenticity knowledge production and power in global health crisis

Artificial intelligence, accountability, and authenticity: knowledge production and power in global health crisis

Reda SadkiGlobal health

I know and appreciate Joseph, a Kenyan health leader from Murang’a County, for years of diligent leadership and contributions as a Scholar of The Geneva Learning Foundation (TGLF). Recently, he began submitting AI-generated responses to Teach to Reach Questions that were meant to elicit narratives grounded in his personal experience.

Seemingly unrelated to this, OpenAI just announced plans for specialized AI agents—autonomous systems designed to perform complex cognitive tasks—with pricing ranging from $2,000 monthly for a “high-income knowledge worker” equivalent to $20,000 monthly for “PhD-level” research capabilities.

This is happening at a time when traditional funding structures in global health, development, and humanitarian response face unprecedented volatility.

These developments intersect around fundamental questions of knowledge economics, authenticity, and power in global health contexts.

I want to explore three questions:

  • What happens when health professionals in resource-constrained settings experiment with AI technologies within accountability systems that often penalize innovation?
  • How might systems claiming to replicate human knowledge work transform the economics and ethics of knowledge production?
  • And how should we navigate the tensions between technological adoption and authentic knowledge creation?

Artificial intelligence within punitive accountability structures of global health

For years, Joseph had shared thoughtful, context-rich contributions based on his direct experiences. All of a sudden, he was submitting generic mush with all the trappings of bad generative AI content.

Should we interpret this as disengagement from peer learning?

Given his history of diligence and commitment, I could not dismiss his exploration of AI tools as diminished engagement. Instead, I understood it as an attempt to incorporate new capabilities into his professional repertoire. This was confirmed when I got to chat with him on a WhatsApp call.

Our current Teach to Reach Questions system has not yet incorporated the use of AI. Our “old” system did not provide any way for Joseph to communicate what he was exploring.

Hence, the quality limitations in AI-generated narratives highlight not ethical failings but a developmental process requiring support rather than judgment.

But what does this look like when situated within global health accountability structures?

Health workers frequently operate within highly punitive systems where performance evaluation directly impacts funding decisions. International donors maintain extensive surveillance of program implementation, creating environments where experimentation carries significant risk. When knowledge sharing becomes entangled with performance evaluation, the incentives for transparency about AI “co-working” (i.e., collaboration between human and AI in work) diminish dramatically.

Seen through this lens, the question becomes not whether to prohibit AI-generated contributions but how to create environments where practitioners can explore technological capabilities without fear that disclosure will lead to automatic devaluation of their knowledge, regardless of its substantive quality. This heavily depends on the learning culture, which remains largely ignored or dismissed in global health.

The transparency paradox: disclosure and devaluation of artificial intelligence in global health

This case illustrates what might be called the “transparency paradox”—when disclosure or recognition of AI contribution triggers automatic devaluation regardless of substantive quality. Current attitudes create a problematic binary: acknowledge AI assistance and have contributions dismissed regardless of quality, or withhold disclosure and risk accusations of misrepresentation or worse.

This paradox creates perverse incentives against transparency, particularly in contexts where knowledge production undergoes intensive evaluation linked to resource allocation. The global health sector’s evaluation systems often emphasize compliance over innovation, creating additional barriers to technological experimentation. When every submission potentially affects funding decisions, incentives for technological experimentation become entangled with accountability pressures.

This dynamic particularly affects practitioners in Global South contexts, who face more intense scrutiny while having less institutional protection for experimentation. The punitive nature of global health accountability systems deserves particular emphasis. Health workers operate within hierarchical structures where performance is consistently monitored by both national governments and international donors. Surveillance extends from quantitative indicators to qualitative assessments of knowledge and practice.

In environments where funding depends on demonstrating certain types of knowledge or outcomes, the incentive to leverage artificial intelligence in global health may conflict with values of authenticity and transparency. This surveillance culture creates uniquely challenging conditions for technological experimentation. When performance evaluation drives resource allocation decisions, health workers face considerable risk in acknowledging technological assistance—even as they face pressure to incorporate emerging technologies into their practice.

The economics of knowledge in global health contexts

OpenAI’s announced “agents” represent a substantial evolution beyond simple chatbots or language models. If they are able to deliver what they just announced, these specialized systems would autonomously perform complex tasks simulating the cognitive work of highly-skilled professionals. The most expensive tier, priced at $20,000 monthly, purportedly offers “PhD-level” research capabilities, working continuously without the limitations of human scheduling or attention.

These claims, while unproven, suggest a potential future where knowledge work economics fundamentally change. For global health organizations operating in Geneva, where even a basic intern position for a recent master’s degree graduate cost more than 200 times that of a ChatGPT subscription, the economic proposition of systems working 24/7 for potentially comparable costs merits careful examination.

However, the global health sector has historically operated with significant labor stratification, where personnel in Global North institutions command substantially higher compensation than those working in Global South contexts. Local health workers often provide critical knowledge at compensation rates far below those of international consultants or staff at Northern institutions. This creates a different economic equation than suggested by Geneva-based comparisons. Many organizations have long relied on substantially lower local labor costs, often justified through capacity-building narratives that mask underlying power asymmetries.

Given this history, the risk that artificial intelligence in global health would replace local knowledge workers might initially appear questionable. Furthermore, the sector has demonstrated considerable resistance to technological adoption, particularly when it might disrupt established operational patterns. However, this analysis overlooks how economic pressures interact with technological change during periods of significant disruption.

The recent decisions of many government to donors to suddenly and drastically cut funding and shut down programs illustrates how rapidly even established funding structures can collapse. In such environments, organizations face existential questions about maintaining operational capacity, potentially creating conditions where technological substitution becomes more attractive despite institutional resistance.

A new AI divide

ChatGPT and other generative AI tools were initially “geo-locked”, making them more difficult to access from outside Europe and North America.

Now, the stratified pricing structure of OpenAI’s announced agents raises profound equity concerns. With the most sophisticated capabilities reserved for those able to pay high costs for the most capable agents, we face the potential emergence of an “AI divide” that threatens to reinforce existing knowledge power imbalances.

This divide presents particular challenges for global health organizations working across diverse contexts. If advanced AI capabilities remain the exclusive province of Northern institutions while Southern partners operate with limited or no AI augmentation, how might this affect knowledge dynamics already characterized by significant inequities?

The AI divide extends beyond simple access to include quality differentials in available systems. Even as simple AI tools become widely available, sophisticated capabilities that genuinely enhance knowledge work may remain concentrated within well-resourced institutions. This could lead to a scenario where practitioners in resource-constrained settings use rudimentary AI tools that produce low-quality outputs, further reinforcing perceptions of capability gaps between North and South.

Confronting power dynamics in AI integration

Traditional knowledge systems in global health position expertise in academic and institutional centers, with information flowing outward to practitioners who implement standardized solutions. This existing structure reflects and reinforces global power imbalances. 

The integration of AI within these systems could either exacerbate these inequities—by further concentrating knowledge production capabilities within well-resourced institutions—or potentially disrupt them by enabling more distributed knowledge creation processes.

Joseph’s journey demonstrates this tension. His adoption of AI tools might be viewed as an attempt to access capabilities otherwise reserved for those with greater institutional resources. The question becomes not whether to allow such adoption, but how to ensure it serves genuine knowledge democratization rather than simply producing more sophisticated simulations of participation.

These emerging dynamics require us to fundamentally rethink how knowledge is valued, created, and shared within global health networks. The transparency paradox, economic pressures, and emerging AI divide suggest that technological integration will not occur within neutral space but rather within contexts already characterized by significant power asymmetries.

Developing effective responses requires moving beyond simple prescriptions about AI adoption toward deeper analysis of how these technologies interact with existing power structures—and how they might be intentionally directed toward either reinforcing or transforming these structures.

My framework for Artificial Intelligence as co-worker to support networked learning and local action is intended to contribute to such efforts.

Illustration: The Geneva Learning Foundation Collection © 2025

References

Frehywot, S., Vovides, Y., 2024. Contextualizing algorithmic literacy framework for global health workforce education. AIH 0, 4903. https://doi.org/10.36922/aih.4903

Hazarika, I., 2020. Artificial intelligence: opportunities and implications for the health workforce. International Health 12, 241–245. https://doi.org/10.1093/inthealth/ihaa007

John, A., Newton-Lewis, T., Srinivasan, S., 2019. Means, Motives and Opportunity: determinants of community health worker performance. BMJ Glob Health 4, e001790. https://doi.org/10.1136/bmjgh-2019-001790

Newton-Lewis, T., Munar, W., Chanturidze, T., 2021. Performance management in complex adaptive systems: a conceptual framework for health systems. BMJ Glob Health 6, e005582. https://doi.org/10.1136/bmjgh-2021-005582

Newton-Lewis, T., Nanda, P., 2021. Problematic problem diagnostics: why digital health interventions for community health workers do not always achieve their desired impact. BMJ Glob Health 6, e005942. https://doi.org/10.1136/bmjgh-2021-005942

Artificial Intelligence and the health workforce: Perspectives from medical associations on AI in health (OECD Artificial Intelligence Papers No. 28), 2024. , OECD Artificial Intelligence Papers. https://doi.org/10.1787/9a31d8af-en

Sadki, R. (2025). A global health framework for Artificial Intelligence as co-worker to support networked learning and local action. Reda Sadki. https://doi.org/10.59350/gr56c-cdd51

Peer learning through Psychological First Aid: New ways to strengthen support for Ukrainian children

Peer learning for Psychological First Aid: New ways to strengthen support for Ukrainian children

Reda SadkiWriting

This article is based on Reda Sadki’s presentation at the ChildHub “Webinar on Psychological First Aid for Children; Supporting the Most Vulnerable” on 6 March 2025. Learn more about the Certificate peer learning programme on Psychological First Aid (PFA) in support of children affected by the humanitarian crisis in Ukraine. Get insights from professionals who support Ukrainian children.

“I understood that if we want to cry, we can cry,” reflected a practitioner in the Certificate peer learning programme on Psychological First Aid (PFA) in support of children affected by the humanitarian crisis in Ukraine – illustrating the kind of personal transformation that complements technical training.

During the ChildHub “Webinar on Psychological First Aid for Children; Supporting the Most Vulnerable”, the Geneva Learning Foundation’s Reda Sadki explained how peer learning provides value that traditional training alone cannot deliver. The EU-funded program on Psychological First Aid (PFA) for children demonstrates that practitioners gain five specific benefits:

First, peer learning reveals contextual wisdom missing from standardized guidance. While technical training provides general principles, practitioners encounter varied situations requiring adaptation. When Serhii Federov helped a frightened girl during rocket strikes by focusing on her teddy bear, he discovered an approach not found in manuals: “This exercise helped the girl switch her focus from the situation around her to caring for the bear.”

Second, practitioners document pattern recognition across diverse cases. Sadki shared how analysis of practitioner experiences revealed that “PFA extends beyond emergency situations into everyday environments” and “children often invent their own therapeutic activities when given space.” These insights help practitioners recognize which approaches work in specific contexts.

Third, peer learning validates experiential knowledge. One practitioner described how simple acknowledgment of feelings often produced visible relief in children, while another found that basic physical comforts had significant psychological impact. These observations, when shared and confirmed across multiple practitioners, build confidence in approaches that might otherwise seem too simple.

Fourth, the network provides real-time problem-solving for urgent challenges. During fortnightly PFA Connect sessions, practitioners discuss immediate issues like “supporting children under three years” or “recognizing severe reactions requiring referrals.” As Sadki explained, these sessions produce concise “key learning points” summarizing practical solutions practitioners can immediately apply.

Finally, peer learning builds professional identity and resilience. “There’s a lot of trust in our network,” Sadki quoted from a participant, demonstrating how sharing experiences reduces isolation and builds a supportive community where practitioners can acknowledge their own emotions and challenges.

The webinar highlighted how this approach creates measurable impact, with practitioners developing case studies that transform tacit knowledge into documented evidence and structured feedback that helps discover blind spots in their practice.

For practitioners interested in joining, Sadki outlined multiple entry points from microlearning modules completed in under an hour to more intensive peer learning exercises, all designed to strengthen support to children while building practitioners’ own professional capabilities.

This project is funded by the European Union. Its contents are the sole responsibility of TGLF, and do not necessarily reflect the views of the European Union.

Illustration: The Geneva Learning Foundation Collection © 2025

New ways to learn and lead HPV vaccination

AI podcast explores surprising insights from health workers about HPV vaccination

Reda SadkiGlobal health

This is an AI podcast featuring two hosts discussing an article by Reda Sadki titled “New Ways to Learn and Lead HPV Vaccination: Bridging Planning and Implementation Gaps.” The conversational format involves the AI hosts taking turns explaining key points and sharing insights about Sadki’s work on HPV vaccination strategies. While the conversation is AI-generated, everything is based on the published article and insights from the experiences of thousands of health workers participating in Teach to Reach.

The Geneva Learning Foundation’s approach

Throughout the podcast, the hosts explore how the Geneva Learning Foundation (TGLF) has developed a five-step process to improve HPV vaccination implementation through their “Teach to Reach” program. This process involves:

  1. Gathering experiences from health workers worldwide
  2. Analyzing these experiences for patterns and innovative solutions
  3. Conducting deep dives into specific case studies
  4. Bringing national EPI planners into the conversation
  5. Synthesizing and sharing knowledge back with frontline workers

The hosts emphasize that this approach represents a shift from traditional top-down strategies to one that values the collective intelligence of over 16,000 global health workers who implement these programs.

Surprising findings

The AI hosts discuss several findings from peer learning that may seem counterintuitive, including:

  • Tribal communities often show less vaccine hesitancy than urban populations, potentially due to stronger community ties and trust in traditional leaders
  • Teachers sometimes have more influence than health workers when it comes to vaccination recommendations
  • Simple, clear communication is often more effective than complex strategies
  • Religious institutions can become powerful allies when approached respectfully
  • Male community leaders can be crucial advocates for what’s typically framed as a women’s health issue

Effective implementation strategies

The hosts highlight various successful implementation approaches mentioned in Sadki’s article:

  • Cancer survivors serving as powerful advocates
  • WhatsApp groups connecting community health workers for information sharing
  • Engaging schoolchildren as messengers to initiate family conversations
  • Integrating vaccination efforts with existing women’s groups
  • Community theater and traditional storytelling methods
  • Less formal settings often producing better results than clinical environments

System-level insights

The podcast discussion reveals that successful vaccination programs don’t necessarily require abundant resources. Instead, key factors include:

  • Strong leadership and clear vision
  • Commitment to continuous learning
  • Community mobilization and trust-building
  • Leveraging informal networks
  • Prioritizing social factors over technical ones
  • Local adaptation rather than standardization

The AI hosts conclude by reflecting on how these principles challenge global health epidemiologists to reconsider their roles—moving beyond data analysis to becoming facilitators who empower communities to develop their own solutions.

New ways to learn and lead HPV vaccination Bridging planning and implementation gaps

HPV vaccination: New learning and leadership to bridge the gap between planning and implementation

Reda SadkiGlobal health

This article is based on my presentation about HPV vaccination at the 2nd National Conference on Adult Immunization and Allied Medicine of the Indian Society for Adult Immunization (ISAI), Science City, Kolkata, on 15 February 2025.

The HPV vaccination implementation challenge

The global landscape of HPV vaccination and cervical cancer prevention reveals a mix of progress and persistent challenges. While 144 countries have introduced HPV vaccines nationally and vaccination has shown remarkable efficacy in reducing cervical cancer incidence, significant disparities persist, particularly in low- and middle-income countries.

Evidence suggests that challenges in implementing and sustaining HPV vaccination programs in developing countries are significantly influenced by gaps between planning at national level and execution at local levels. Multiple studies confirm this disconnect as a primary barrier to effective HPV vaccination programmes.

Traditional approaches to knowledge development in global health often rely on expert committee models characterized by hierarchical knowledge flows, formal meeting processes, and bounded timelines. While these approaches offer strengths like high academic rigor and systematic review, they frequently miss frontline insights, develop slowly, and produce static outputs that may be difficult to translate effectively into action.

How the peer learning network alternative can support HPV vaccination

At The Geneva Learning Foundation (TGLF), we have developed a complementary model—one that values the collective intelligence of frontline health workers and creates structured opportunities for their insights to inform policy and practice. This peer learning network model features:

  • Large, diverse networks with multi-directional knowledge flow
  • Open participation and flexible engagement
  • Direct field experience and implementation insights
  • Iterative development through experience sharing
  • Continuous refinement and living knowledge

This approach captures practical knowledge, enables rapid learning cycles, preserves context, and brings together multiple perspectives in a dynamic process that continuously updates as new information emerges.

HPV vaccination: the peer learning cycle in action

To address HPV vaccination challenges, we implemented a structured five-stage cycle that connected frontline experiences with policy decisions:

  1. Experience collection at scale: In June 2023, we engaged over 16,000 health professionals to share their HPV vaccination experiences through our Teach to Reach programme. This stage focused specifically on capturing frontline implementation challenges and solutions across diverse contexts.
  2. Synthesis and analysis: TGLF’s Insights Unit identified key themes, success patterns, and common challenges while highlighting local innovations and practical solutions that emerged from the field.
  3. Knowledge deepening: In October 2023, we conducted a second round of experience sharing that built upon earlier discussions at Teach to Reach. This stage featured more in-depth case studies and implementation stories, providing additional contexts and approaches to vaccination challenges.
  4. National-level review: In January 2024, we facilitated a consultation with national EPI (Expanded Programme on Immunization) planners from 31 countries. This created direct connections between field experience and national strategy, validating and enriching the collected insights.
  5. Knowledge mobilization: Finally, we synthesized the insights into practical guidance, ready for sharing back to frontline workers, and established a foundation for continued learning cycles.

This process uniquely values the practical wisdom that emerges from implementation experience. Rather than assuming solutions flow from the top down, we recognize that those doing the work often develop the most effective approaches to complex challenges.

Teach to Reach: Building a learning community for HPV vaccination

Our Teach to Reach programme serves as the hub for this peer learning approach. Since its inception, the community has grown steadily since January 2021 to reach over 24,000 members by December 2024. The participants reflect remarkable diversity.

This diversity of contexts and experiences creates a rich environment for learning. The programme demonstrates significant impact on participants’ professional capabilities—compared to global baselines, Teach to Reach participants show:

  • 45% stronger worldview change
  • 41% greater impact on professional practice
  • 49% higher professional influence

7 insights about HPV vaccination from peer learning at Teach to Reach

Through this process, we uncovered several important implementation insights:

1. Importance of connecting field experience to policy

  • Each stage deepened understanding of implementation challenges
  • We observed progression from tactical to strategic considerations
  • Growing recognition of systemic factors emerged
  • Evolution from individual to institutional solutions became apparent
  • Value of structured knowledge sharing across levels was demonstrated

2. Implementation learning

  • Success requires multi-stakeholder engagement
  • Sustained communication proves more effective than one-time campaigns
  • School systems provide critical implementation platforms
  • Community leadership is essential for acceptance
  • Integration with other services increases efficiency
  • Local adaptation is key to successful implementation

3. Unexpected implementation findings

  • Tribal communities often showed less vaccine hesitancy than urban areas
  • Teachers emerged as more influential than health workers in some contexts
  • Personal stories proved more persuasive than statistical evidence
  • Integration with COVID-19 vaccination improved HPV acceptance
  • Social media played both positive and negative roles
  • School-based programs sometimes reached out-of-school children

4. Counter-intuitive success factors

  • Less formal settings often produced better results
  • Simple communication strategies outperformed complex ones
  • Male community leaders became strong vaccination advocates
  • Religious institutions provided unexpected support
  • Health worker vaccination of own children became powerful tool
  • Community dialogue proved more effective than expert presentations

5. Unexpected challenges

  • Urban areas sometimes showed more resistance than rural areas
  • Education level did not correlate with vaccine acceptance
  • Health workers themselves sometimes showed hesitancy
  • Traditional media was less influential than anticipated
  • Formal authority figures were not always the most effective advocates
  • Technical knowledge proved less important than communication skills

6. Examples of novel solutions

  • Using cancer survivors as advocates
  • WhatsApp groups for community health workers
  • School children as messengers to families
  • Integration with existing women’s groups
  • Leveraging religious texts and teachings
  • Community theater and storytelling approaches

System-level surprises

  • Success was often independent of resource levels
  • Informal networks proved more important than formal ones
  • Bottom-up strategies were more effective than top-down approaches
  • Social factors were more influential than technical ones
  • Local adaptation was more important than standardization
  • Peer influence was more powerful than expert authority

In some cases, these findings challenge many conventional assumptions about HPV vaccination programmes. In all cases, they highlight the importance of local knowledge, social factors, and adaptation over standardized approaches based solely on technical expertise.

The power of health worker collective intelligence

Our approach demonstrates the value of health worker collective intelligence in improving performance:

  • High-quality data and situational intelligence from our network of 60,000+ health workers provides rapid insights
  • Field observations on changing disease patterns and resistance can be quickly collected
  • Climate change impacts can be tracked through frontline reports
  • The TGLF Insights Unit packages this intelligence into knowledge to inform practice and policy

This represents a fundamental shift from assuming expert committees have all the answers to recognizing the distributed expertise that exists throughout health systems.

Continuous learning: The key to improvement

In fact, previous TGLF research has demonstrated that continuous learning is often the “Achilles’ heel” in immunization programs. Common issues include:

  1. Relative lack of learning opportunities
  2. Limited ability to experiment and take risks
  3. Low tolerance for failure
  4. Focus on task completion at the expense of building capacity for future performance
  5. Lack of encouragement for learning tied to tangible organizational incentives

In 2020 and 2022, we conducted large-scale measurements of learning culture of more than 10,000 immunization professionals in low- and middle-income countries. The data showed that ‘learning culture’ (a measure of the capacity for change) correlated more strongly with perceived programme performance than individual motivation did. This challenges the common assumption that poor motivation is the root cause of poor performance.

These findings help zero in on six ways to strengthen continuous learning to drive HPV vaccination:

  1. Motivate health workers to believe strongly in the importance of what they do
  2. Give them practice dealing with difficult situations they might face
  3. Build mental resilience for facing obstacles
  4. Prompt them to enlist coworkers for support
  5. Help them engage their bosses to provide guidance, support, and resources
  6. Help them identify and overcome workplace obstacles

Impact and benefits of peer learning

This approach delivers multiple benefits:

  • Frontline workers gain broader perspective
  • National planners access grounded insights
  • Practical solutions spread more quickly
  • Policy decisions are informed by field experience
  • Continuous improvement cycle gets established

Key success factors include:

  • Scale that enables diverse input collection
  • Structure that supports quality knowledge creation
  • Regular rhythm that maintains engagement
  • Multiple levels of review that ensure relevance
  • Clear pathways from insight to action

How can we interpret these findings?

This model generates implementation-focused evidence that complements rather than competes with traditional epidemiological data. 

The findings emerge from a structured methodology that includes initial experience collection at scale, synthesis and analysis, knowledge deepening through case studies, national-level review by EPI planners from 31 countries, and systematic knowledge mobilization. This approach provides rigor and scale that elevate these observations beyond mere anecdotes.

For epidemiologists who become uncomfortable when evidence is not purely quantitative, it is important to understand that structured peer learning fills a critical gap in implementation science by capturing what quantitative studies often miss: the contextual factors and practical adaptations that determine programme success or failure in real-world settings.

When implementers report across different contexts that tribal communities show less vaccine hesitancy than urban areas, or that teachers emerge as more influential than health workers in specific settings, these patterns represent valuable implementation intelligence.

Such insights also help explain why interventions that appear effective in controlled studies often fail to deliver similar results when implemented at scale.

In fact, these findings address precisely what quantitative studies struggle to capture: why education level does not reliably predict vaccine acceptance; why some resource-constrained settings outperform better-resourced ones; how informal networks frequently prove more effective than formal structures; and which communication approaches actually drive behavior change in specific populations.

For programme planners, this knowledge bridges the gap between general guidance (“engage community leaders”) and actionable specifics (“male community leaders became particularly effective advocates when engaged through these specific approaches”). 

Accelerating HPV vaccination progress

To make significant progress on HPV vaccination as part of the Immunization Agenda 2030’s Strategic Priority 4 (life-course and integration), we encourage global health stakeholders to:

  1. Rethink how we learn
  2. Question how we engage with families and communities
  3. Focus on trust

By combining expert knowledge with the practical wisdom of thousands of implementers, we can develop more effective strategies for HPV vaccination that bridge the gap between planning and execution.

This peer learning network approach does not replace expertise—it enhances and grounds it in the realities of implementation.

It recognizes that the frontline health worker in a remote village may hold insights just as valuable as those of a technical expert in a capital city.

By creating structures that enable these insights to emerge and connect, we can accelerate progress on HPV vaccination and other public health challenges.

Acknowledgements

I wish to thank ISAI’s Dr Saurabh Kole and his colleagues for their kind invitation. I also wish to recognize and appreciate Charlotte Mbuh and Ian Jones for their invaluable contributions to the Foundation’s work on HPV vaccination, and Dr Satabdi Mitra for her tireless leadership and boundless commitment. Last but not least, I wish to thank the thousands of health workers who contributed their experiences before, during, and after successive Teach to Reach peer learning events. What little I know comes from their collective intelligence, action, and wisdom.

References

Dorji, T. et al. (2021) ‘Human papillomavirus vaccination uptake in low-and middle-income countries: a meta-analysis’, EClinicalMedicine, 34, p. 100836. Available at: https://doi.org/10.1016/j.eclinm.2021.100836.

Faye, W. et al. (2023) IA2030 Case study 18. Wasnam Faye. Vaccine angels – Give us the opportunity and we can perform miracles. The Geneva Learning Foundation. Immunization Agenda 2030 Case study 18. Available at: https://doi.org/10.5281/ZENODO.7785244.

Gonçalves, I.M.B. et al. (2020) ‘HPV Vaccination in Young Girls from Developing Countries: What Are the Barriers for Its Implementation? A Systematic Review’, Health, 12(06), pp. 671–693. Available at: https://doi.org/10.4236/health.2020.126050.

Jones, I. et al. (2024) Making connections at Teach to Reach 8 (IA2030 Listening and Learning Report 6). Available at: https://doi.org/10.5281/ZENODO.8398550.

Jones, I. et al. (2022) IA2030 Case Study 7. Motivation, learning culture and programme performance. The Geneva Learning Foundation. Available at: https://doi.org/10.5281/ZENODO.7004304.

Kutz, J.-M. et al. (2023) ‘Barriers and facilitators of HPV vaccination in sub-saharan Africa: a systematic review’, BMC Public Health, 23(1), p. 974. Available at: https://doi.org/10.1186/s12889-023-15842-1.

Moore, K. et al. (2022) Overcoming barriers to vaccine acceptance in the community: Key learning from the experiences of 734 frontline health workers. The Geneva Learning Foundation. Available at: https://doi.org/10.5281/ZENODO.6965355.

Umbelino-Walker, I. et al. (2024) ‘Towards a sustainable model for a digital learning network in support of the Immunization Agenda 2030 –a mixed methods study with a transdisciplinary component’, PLOS Global Public Health. Edited by M. Pentecost, 4(12), p. e0003855. Available at: https://doi.org/10.1371/journal.pgph.0003855.

Watkins, K.E. et al. (2022) ‘Accelerating problem-solving capacities of sub-national public health professionals: an evaluation of a digital immunization training intervention’, BMC Health Services Research, 22(1), p. 736. Available at: https://doi.org/10.1186/s12913-022-08138-4.

Wigle, J., Coast, E. and Watson-Jones, D. (2013) ‘Human papillomavirus (HPV) vaccine implementation in low and middle-income countries (LMICs): Health system experiences and prospects’, Vaccine, 31(37), pp. 3811–3817. Available at: https://doi.org/10.1016/j.vaccine.2013.06.016.

A global health framework for Artificial Intelligence as co-worker to support networked learning and local action

A global health framework for Artificial Intelligence as co-worker to support networked learning and local action

Reda SadkiGlobal health

The theme of International Education Day 2025, “AI and education: Preserving human agency in a world of automation,” invites critical examination of how artificial intelligence might enhance rather than replace human capabilities in learning and leadership. Global health education offers a compelling context for exploring this question, as mounting challenges from climate change to persistent inequities demand new approaches to building collective capability.

The promise of connected communities

Recent experiences like the Teach to Reach initiative demonstrate the potential of structured peer learning networks. The platform has connected over 60,000 health workers, primarily government workers from districts and facilities across 82 countries, including those serving in conflict zones, remote rural areas, and urban settlements. For example, their exchanges about climate change impacts on community health point the way toward more distributed forms of knowledge creation in global health. 

Analysis of these networks suggests possibilities for integrating artificial intelligence not merely as tools but as active partners in learning and action. However, realizing this potential requires careful attention to how AI capabilities might enhance rather than disrupt the human connections that drive current success.

Artificial Intelligence (AI) partnership could provide crucial support for tackling mounting challenges. More importantly, they could help pioneer new approaches to learning and action that genuinely serve community needs while advancing our understanding of how human and machine intelligence might work together in service of global health.

Understanding Artificial Intelligence (AI) as partner, not tool

The distinction between AI tools and AI partners merits careful examination. Early AI applications in global health primarily automate existing processes – analyzing data, delivering content, or providing recommendations. While valuable, this tool-based approach maintains clear separation between human and machine capabilities.

AI partnership suggests a different relationship, where artificial intelligence participates actively in learning networks alongside human practitioners. This could mean AI systems that:

  • Engage in dialogue with health workers about local observations
  • Help validate emerging insights through pattern analysis
  • Support adaptation of solutions across contexts
  • Facilitate connections between practitioners facing similar challenges

The key difference lies in moving from algorithmic recommendations to collaborative intelligence that combines human wisdom with machine capabilities.

A framework for AI partnership in global health

Analysis of current peer learning networks suggests several dimensions where AI partnership could enhance collective capabilities:

  • Knowledge creation: Current peer learning networks enable health workers to share observations and experiences across borders. AI partners could enrich this process by engaging in dialogue about patterns and connections, while preserving the central role of human judgment in validating insights.
  • Learning process: Teach to Reach demonstrates how structured peer learning accelerates knowledge sharing and adaptation. AI could participate in these networks by contributing additional perspectives, supporting rapid synthesis of experiences, and helping identify promising practices.
  • Local leadership: Health workers develop and implement solutions based on deep understanding of community needs. AI partnership could enhance decision-making by exploring options, modeling potential outcomes, and validating approaches while maintaining human agency.
  • Network formation: Digital platforms currently enable lateral connections between health workers across regions. AI could actively facilitate network development by identifying valuable connections and supporting knowledge flow across boundaries.
  • Implementation support: Peer review and structured feedback drive current learning-to-action cycles. AI partners could engage in ongoing dialogue about implementation challenges while preserving the essential role of human judgment in local contexts.
  • Evidence generation: Networks document experiences and outcomes through structured processes. AI collaboration could help develop and test hypotheses about effective practices while maintaining focus on locally-relevant evidence.

Applications across three global health challenges

This framework suggests new possibilities for addressing persistent challenges.

1. Immunization systems

Current global immunization goals face significant obstacles in reaching zero-dose children and strengthening routine services. AI partnership could enhance efforts by:

  • Supporting microplanning by mediating dialogue about local barriers
  • Facilitating rapid learning about successful engagement strategies
  • Enabling coordinated action across health system levels
  • Modeling potential impacts of different intervention approaches

2. Neglected Tropical Diseases (NTDs)

The fight against NTDs suffers from critical information gaps and weak coordination at local levels. Many communities, including health workers, lack basic knowledge about these diseases. AI partnership could help address these gaps through:

  • Facilitating knowledge flow between affected communities
  • Supporting coordination of control efforts
  • Enabling rapid validation of successful approaches
  • Strengthening surveillance and response networks

3. Climate change and health

Health workers’ observations of climate impacts on community health provide crucial early warning of emerging threats. AI partnership could enhance response capability by:

  • Engaging in dialogue about changing disease patterns
  • Supporting rapid sharing of adaptation strategies
  • Facilitating coordinated action across regions
  • Modeling potential impacts of interventions

Pandemic preparedness beyond early warning

The experience of digital health networks during recent disease outbreaks reveals both the power of distributed response capabilities and the potential for enhancement through AI partnership. When COVID-19 emerged, networks of health workers demonstrated remarkable ability to rapidly share insights and adapt practices. For example, the Geneva Learning Foundation’s COVID-19 Peer Hub connected over 6,000 frontline health professionals who collectively generated and implemented recovery strategies at rates seven times faster than isolated efforts.

This networked response capability suggests new possibilities for pandemic preparedness that combines human and machine intelligence. Heightened preparedness could emerge from the interaction between health workers, communities, and AI partners engaged in continuous learning and adaptation.

Current pandemic preparedness emphasizes early detection through formal surveillance. However, health workers in local communities often observe concerning patterns before these register in official systems.

AI partnership could enhance this distributed sensing capability while maintaining its grounding in local realities. Rather than simply analyzing reports, AI systems could engage in ongoing dialogue with health workers about their observations, helping to:

  • Explore possible patterns and connections
  • Test hypotheses about emerging threats
  • Model potential trajectories
  • Identify similar experiences across regions

The key lies in combining human judgment about local significance with AI capabilities for pattern recognition across larger scales.

The focus remains on accelerating locally-led learning rather than imposing standardized solutions.

Perhaps most importantly, AI partnership could enhance the collective intelligence that emerges when practitioners work together to implement solutions. Current networks enable health workers to share implementation experiences and adapt strategies to local contexts. Adding AI capabilities could support this through:

  • Ongoing dialogue about implementation challenges
  • Analysis of patterns in successful adaptation
  • Support for rapid testing of modifications
  • Facilitation of cross-context learning

Success requires maintaining human agency in implementation while leveraging machine capabilities to strengthen collective problem-solving.

This networked vision of pandemic preparedness, enhanced through AI partnership, represents a fundamental shift from current approaches. Rather than attempting to predict and control outbreaks through centralized systems, it suggests building distributed capabilities for continuous learning and adaptation. The experience of existing health worker networks provides a foundation for this transformation, while artificial intelligence offers new possibilities for strengthening collective response capabilities.

Investment for innovation

Realizing this vision requires strategic investment in:

  • Network development: Supporting growth of peer learning platforms that accelerate local action while maintaining focus on human connection.
  • AI partnership innovation: Developing systems designed to participate in learning networks while preserving human agency.
  • Implementation research: Studying how AI partnership affects collective capabilities and health outcomes.
  • Capacity strengthening: Building health worker capabilities to effectively collaborate with AI while maintaining critical judgment.

Looking forward

The transformation of global health learning requires moving beyond both conventional practices of technical assistance and simple automation. Experience with peer learning networks demonstrates what becomes possible when health workers connect to share knowledge and drive change.

Adding artificial intelligence as partners in these networks – rather than replacements for human connection – could enhance collective capabilities to protect community health. However, success requires careful attention to maintaining human agency while leveraging technology to strengthen rather than supplant local leadership.

7 key principles for AI partnership

  1. Maintain human agency in decision-making
  2. Support rather than replace local leadership
  3. Enhance collective intelligence
  4. Enable rapid learning and adaptation
  5. Preserve context sensitivity
  6. Facilitate knowledge flow across boundaries
  7. Build sustainable learning systems

Listen to an AI-generated podcast about this article

🤖 This podcast was generated by AI, discussing Reda Sadki’s 24 January 2025 article “A global health framework for Artificial Intelligence as co-worker to support networked learning and local action”. While the conversation is AI-generated, the framework and examples discussed are based on the published article.

Framework: AI partnership for learning and local action in global health

DimensionCurrent StateAI as ToolsAI as PartnersPotential Impact
Knowledge creationHealth workers share observations and experiences through peer networksAI analyzes patterns in shared dataAI engages in dialogue with health workers, asking questions, suggesting connections, validating insightsNew forms of collective intelligence combining human and machine capabilities
Learning processStructured peer learning through digital platforms and networksAI delivers content and analyzes performanceAI participates in peer learning networks, contributes insights, supports adaptationAccelerated learning through human-AI collaboration
Local leadershipHealth workers develop and implement solutions for community challengesAI provides recommendations based on data analysisAI works alongside local leaders to explore options, model scenarios, validate approachesEnhanced decision-making combining local wisdom with AI capabilities
Network formationLateral connections between health workers across regionsAI matches similar profiles or challengesAI actively facilitates network development, identifies valuable connectionsMore effective knowledge networks leveraging both human and machine intelligence
Implementation supportPeer review and structured feedback on action plansAI checks plans against best practicesAI engages in iterative dialogue about implementation challenges and solutionsImproved implementation through combined human-AI problem-solving
Evidence generationDocumentation of experiences and outcomes through structured processesAI analyzes implementation dataAI collaborates with health workers to develop and test hypotheses about what worksNew approaches to generating practice-based evidence

Image: The Geneva Learning Foundation Collection © 2024