Strengthening the education and continuous professional development of public health professionals is fundamental to achieving resilient, equitable, and responsive health systems. Recent global experiences including pandemics, protracted humanitarian crises, climate-related health risks, rapid urbanization, and technological transformation have highlighted both the critical role of the public health workforce and persistent gaps in training relevance, scale, quality, and sustainability. In many settings, educational models remain insufficiently aligned with real-world system needs, emerging competencies, and evolving career pathways.

While there is growing innovation in public health education, evidence and lessons learned are dispersed across disciplines and regions. A focused supplement in the Journal of Epidemiology and Global Health will provide a timely scholarly platform to consolidate high-quality research and practice-based insights.

The proposed Collection aims to advance knowledge in the education of public health professionals by:

Showcasing empirical evidence and innovative models for public health education and training

Examining how educational approaches align with health system priorities and workforce needs

Informing policy, institutional strategies, and investment in public health workforce development

Promoting equity, quality, and sustainability in public health education globally.

Scope and Thematic Areas

The supplement will invite original research articles, systematic or scoping reviews, and rigorously documented practice-based papers across the following thematic areas:

Competency-Based Public Health Education

• Core and advanced competency frameworks
• Alignment of curricula with population health needs and system priorities

Field-Based and Applied Learning Models

• Field Epidemiology Training Programs (FETPs) and similar applied training models
• Experiential learning, service-based education, and community engagement

Interprofessional and Multisectoral Education

• Collaborative training across health, social, environmental, and humanitarian sectors
• Preparing public health professionals for whole-of-government and whole-of-society approaches

Digital, Blended, and Distance Learning Innovations

• Online and hybrid training models
• Use of digital platforms, simulation, and emerging technologies in education

Education for Health Emergencies and Fragile Settings

• Workforce training for outbreak preparedness, humanitarian response, and conflict settings
• Adaptive education models in fragile and resource-constrained contexts

Equity, Ethics, and Inclusion in Public Health Training

• Gender, geographic, and socioeconomic equity in access to education
• Ethical dimensions of training, mentorship, and professional advancement

Leadership, Management, and Systems Thinking

• Training for public health leadership, governance, and policy engagement
• Building managerial and strategic competencies for system-level impact

Mentorship, Supervision, and Career Pathways

• Structured mentorship and supervision models
• Transition from training to practice and career progression

Monitoring, Evaluation, and Impact of Public Health Education

• Methods for assessing educational outcomes and workforce impact
• Long-term sustainability and return on investment of training programs

Partnerships, Financing, and Institutionalization

• Academic–government–partner collaborations
• Financing models and institutional integration of training programs

All submissions in this collection undergo the journal’s standard peer review process. Similarly, all manuscripts authored by a Guest Editor(s) will be handled by the Editor-in-Chief. As an open access publication, this journal levies an article processing fee (details here). We recognize that many key stakeholders may not have access to such resources and are committed to supporting participation in this issue wherever resources are a barrier. For more information about what support may be available, please visit OA funding and support, or email OAfundingpolicy@springernature.com or the Editor-in-Chief.

How AI shapes medicine and human health

Artificial intelligence (AI) in health now spans large language models, computer vision, multimodal foundation models, predictive analytics, and decision-support systems. ^{1 2} These tools can process clinical text, imaging, physiological signals, laboratory data, patient history, and public health information at scale, and they are being piloted to support tasks ranging from documentation and prescription renewal to diagnostic interpretation, risk stratification, surveillance, and health-system operations.³ As adoption accelerates, evidence on real-world performance, safety, equity, and downstream health effects remains limited.

Large language models are one important branch of AI in health, but they are only part of a broader landscape that also includes computer vision for medical imaging, machine learning for prediction and classification, multimodal models that combine text and images or signals, and AI tools embedded in clinical and public health workflows. Across all of these approaches, performance depends on data quality, representativeness, validation, and the context in which the system is deployed.

LLMs are systems designed to predict the most likely next word. The word “Large” implies the vastness of written data accessible electronically—including books, web data, and articles—used in these predictions. The word “model” implies that LLMs do not know facts as humans do; they operate on probability. When we type or ask a question (called a prompt), the model predicts the next most likely option, called a “token,” based on the patterns it learned during training.¹ This happens in a non-linear manner, meaning it processes words in relation to all other words in a sentence rather than one by one, helping it understand context.

In simple terms, the computing framework used in an LLM, the transformer, finds all possible routes on a map (neural networks) to reach a destination and then highlights the best options based on the logic it was trained to follow. The models are only as good as the data available to them and the training and fine-tuning they were subjected to.

For example, in Utah’s Prescription-Renewal Pilot Program³ launched in January 2026, LLMs are trained (the first 250 outputs require physician review for fine-tuning) and approved to perform an agentic function for 192 commonly prescribed chronic disease medicines that require frequent refills (a repetitive activity for physicians). Tethered to a deterministic logic engine, the system uses a multi-agent AI platform to handle the natural language interaction with the patient and data (patient records, patient-provided information on health status during the refill process, drug pharmacology, and common patient challenges).

AI can assist in patient care, public health surveillance, training of health care professionals, data monitoring, and policy development.^{2 4 5} Collective use of data can shape policy decisions in public health, including through rapid and live analysis of social media chats or disease report patterns to generate new hypotheses or public health responses.

Deployment of LLMs in healthcare is outpacing the evidence base for real-world performance, equity, and downstream population-health effects.

The challenge ahead

As it is new and rapidly evolving, there is a need to surface critical concerns and risks about AI in health.

Mainstreaming AI is a challenge across disciplines, geographies, and development barriers. Most existing literature treats clinical AI and population-health AI as separate conversations. In practice, a tool deployed at the bedside has immediate epidemiologic and policy consequences—who gets referred, who gets screened, which signals reach surveillance systems, and which populations bear the costs of miscalibration. Public health is yet to exploit the vast analytic power of AI in triaging huge amounts of healthcare delivery data and health-related social media discussions, or the utility of AI in healthcare delivery, including supply chains. There is an imminent need to avert a potential AI access or utilization gap between resource-rich and resource-poor countries.

There is a policy vacuum and regulatory gap regarding AI applications and licensing.

The lack of harmonization remains between the speed at which AI applications are evolving and the response from the regulatory framework.

What can researchers do?

The challenge to researchers is to disseminate information on the potential utility and impact on the healthcare delivery system, and how to steer policy proactively to avert potential drawbacks that are common to all new approaches.

This collection is designed to connect those conversations and to produce evidence useful to clinicians, epidemiologists, and policymakers simultaneously, with particular attention to equity, low- and middle-income country (LMIC) applicability, and unintended consequences. We welcome original research, systematic reviews, policy analyses, and commentaries on generative and non-generative AI across the following areas:

• Clinical and diagnostic domains — development, validation, and real-world evaluation of diagnostic, prognostic, and decision-support tools; LLMs in clinical workflows; computer vision in radiology, pathology, dermatology, ophthalmology, endoscopy, and surgery; multimodal models integrating text, imaging, laboratory, and waveform data; clinician–AI interaction and human oversight.
• Physiological signals, wearables, and remote monitoring — AI for ECG, EEG, ICU monitoring, mobile and home-based sensors, early warning systems, and longitudinal patient monitoring.
• Epidemiology and population health — surveillance, outbreak detection, and forecasting; AI-assisted causal inference and analysis of large observational datasets; applications to communicable and non-communicable disease epidemiology.
• Pharmacoepidemiology and real-world evidence of medication use. * Global health — model development and validation in low-resource settings; transportability and fairness across populations; national and regional deployment case studies, including program failures.
• Ethics, equity, governance, and policy — algorithmic bias and fairness audits; informed consent and transparency; regulatory frameworks and reporting standards (CONSORT-AI, TRIPOD+AI, DECIDE-AI); economic and health-system evaluations; workforce implications.
• Methodological and cross-cutting work — frameworks linking individual-level model performance to population-level outcomes; comparative effectiveness of AI-assisted versus standard pathways; studies of unintended consequences.
• Medical and health system research — development of prompts for large data use and highlighting utility to inform policy; innovations in surveillance systems; improving efficiency of scientific investigations, including methodology and survey development.

We believe the Collection will serve as inaugural contributions on this rapidly evolving interface of medicine, public health, biotechnology, and ethics, collectively enhancing our efforts to continuously improve human health and welfare.

References

1. Teo ZL, Thirunavukarasu AJ, Elangovan K, et al. Generative artificial intelligence in medicine. Nature Medicine 2025;31(10):3270-82. doi: 10.1038/s41591-025-03983-2

2. Rajpurkar P, Chen E, Banerjee O, et al. AI in health and medicine. Nature Medicine 2022;28(1):31-38. doi: 10.1038/s41591-021-01614-0

3. Gerke S, Parikh RB, Cohen IG. Utah’s Prescription-Renewal Pilot Program — Autonomous AI Managing Patient Care. New England Journal of Medicine 2026;394(16):1561-63. doi: doi:10.1056/NEJMp2601148

4. Panteli D, Adib K, Buttigieg S, et al. Artificial intelligence in public health: promises, challenges, and an agenda for policy makers and public health institutions. The Lancet Public Health 2025;10(5):e428-e32.

5. Rao VM, Hla M, Moor M, et al. Multimodal generative AI for medical image interpretation. Nature 2025;639(8056):888-96.

All submissions in this collection undergo the journal’s standard peer review process. Similarly, all manuscripts authored by a Guest Editor(s) will be handled by the Editor-in-Chief. As an open access publication, this journal levies an article processing fee (details here). We recognize that many key stakeholders may not have access to such resources and are committed to supporting participation in this issue wherever resources are a barrier. For more information about what support may be available, please visit OA funding and support, or email OAfundingpolicy@springernature.com or the Editor-in-Chief.