This community of practice emerges from the recognition that advancing health systems modelling requires more than technical solutions — it demands collaboration, shared learning, and sustained dialogue between modellers, policymakers, and health systems practitioners. Our vision is to create a dynamic ecosystem where expertise is shared, methodological innovations are disseminated, and practical experiences inform continuous improvement in how we approach health systems modelling.

We aim to bridge the gap between modelling capabilities and policy needs. Too often, sophisticated models remain academic exercises. They’re disconnected from policy realities. Meanwhile, critical decisions are made without systematic analysis of their likely system-wide impacts. This community seeks to foster meaningful collaboration where modellers understand policy constraints and timelines, while policymakers appreciate both the potential and limitations of modelling approaches.

The community will serve as a repository of collective wisdom — documenting what works, what doesn’t, and why. We’ll capture experiences from different countries and contexts. This helps us learn faster and avoid repeated mistakes. This includes not just technical aspects of model development but also crucial insights about engaging stakeholders, communicating results, and integrating modelling into policy processes.

Ultimately, we aspire to establish health systems modelling as a routine component of health policy and planning. Just as economic modelling informs fiscal policy and climate models guide environmental decisions, health systems models should become standard tools for designing, evaluating, and optimizing health systems interventions. Through this community of practice, we work toward a future where health systems investments are informed by rigorous analysis of their system-wide impacts, leading to more effective, efficient, and equitable health systems worldwide.

Health systems face an increasingly complex challenge: how to allocate limited resources to achieve maximum population health impact. Despite growing recognition of the importance of investing in health systems strengthening, stakeholders at both national and global levels cannot easily quantify the impact of these investments.

This inability to draw clear connections between health systems inputs — such as workforce training, supply chain improvements, or governance reforms — and tangible health outcomes puts health systems advocates at a disadvantage. Without evidence of clear returns on investment, health systems strengthening often loses out to more easily quantifiable, disease-specific interventions in the competition for limited resources.

The COVID-19 pandemic highlighted both the critical importance of resilient health systems and the potential of mathematical modelling to inform policy decisions. While models became central to designing and implementing public health measures during the pandemic, the application of similar approaches to broader health systems challenges remains underdeveloped.

This gap represents not just a technical challenge but a fundamental barrier to achieving universal health coverage and health equity. As countries work toward ambitious health goals with constrained budgets, the need for evidence-based approaches to prioritize and optimize health systems investments has never been more urgent.

Health systems modelling, as we conceptualize it for this community, goes beyond traditional disease-specific models to capture the complex, interconnected nature of health systems. At its core, a health systems model estimates the impact of interventions that strengthen health system building blocks — workforce, service delivery, information systems, medical products, financing, and governance — on population health outcomes. Crucially, these models must capture the horizontal effects of interventions across multiple disease programs, reflecting the reality that health systems interventions rarely affect just one aspect of healthcare delivery.

Consider, for example, an intervention to improve supply chain management. While a disease-specific model might examine how this affects antiretroviral drug availability for HIV treatment, a health systems model would capture how the same intervention simultaneously impacts vaccine cold chains, essential medicine availability for non-communicable diseases, and emergency obstetric supplies. This multi-disease, systems-wide perspective is what distinguishes health systems modelling from other forms of health economic or epidemiological modelling.

Health systems modelling presents distinctive challenges that set it apart from other modelling domains. The first and perhaps most fundamental challenge is complexity. Health systems are adaptive and context-dependent. Interventions can cascade through multiple levels and components. An intervention to decentralize decision-making, for instance, might affect governance structures, alter resource flows, change accountability relationships, and ultimately impact service delivery in ways that vary dramatically across different contexts.

Data poses another major challenge. While disease-specific models often rely on well-established epidemiological parameters and clinical trial data, health systems models require diverse data sources — from health facility assessments and routine information systems to household surveys and administrative records. Many countries lack comprehensive data on health systems functioning. What data does exist often comes from disparate sources that are difficult to integrate. Moreover, critical parameters, such as the effect of supervision on health worker performance or the relationship between management practices and facility efficiency, are often poorly quantified in the literature, and in any case are highly context-specific.

The challenge of achieving the right balance of model complexity presents a particular dilemma. Models must be sophisticated enough to capture the essential dynamics of health systems — the feedback loops, the interdependencies, the contextual variations — while remaining simple enough to be understood, trusted, and used by policymakers. Too simple? The model misses critical behaviors. Too complex? It becomes incomprehensible. This balance is not merely technical but fundamentally about ensuring models serve their intended purpose of informing real-world policy decisions.

Finally, there is the challenge of comparability and prioritization. Health systems interventions are inherently diverse — how does one compare the impact of training additional midwives against investing in digital health information systems or introducing performance-based financing? These interventions operate through different mechanisms, require different implementation capacities, and have effects that manifest over different time horizons. Creating models that can meaningfully compare such diverse interventions while accounting for their varying costs, feasibility, and contextual requirements remains a significant methodological frontier.