Systems thinking in public health is a way of understanding health problems not as isolated issues with single causes, but as products of interconnected forces that influence and reinforce each other over time. Instead of asking “what is the cause of this disease?” a systems thinker asks “what web of social, biological, economic, and environmental factors keeps this problem in place?” The distinction matters because many chronic public health challenges, from obesity to antimicrobial resistance, have resisted decades of interventions aimed at one factor at a time. Systems thinking offers a framework for seeing why those single-target approaches often disappoint, and where smarter points of intervention might exist.
Why Single-Cause Thinking Keeps Falling Short
Public health earned its early wins by targeting specific causes: contaminated water, a particular microbe, a missing nutrient. That approach works brilliantly when you can draw a straight line from cause to effect. But the problems that dominate modern public health rarely sit on a straight line. Childhood obesity, for instance, involves food marketing, school policies, neighborhood design, family income, cultural norms around eating, agricultural subsidies, and the child’s own biology, all feeding into each other in loops that no single program can break.
Research on how policymakers design interventions has found that policies built on “rational-actor” models and housed in institutional silos tend to reinforce exactly this linear way of thinking. Policymakers acknowledged that blunt policy tools have multiple effects and that carrying out a “surgical strike” to change one outcome is rarely realistic, yet the silo structure of government keeps encouraging it.1PubMed Central. Understanding the unintended consequences of public health policies: the views of policymakers and evaluators Systems thinking emerged in public health as a direct response to this frustration: if the problem is a system, the response needs to account for how the system behaves.
The Building Blocks of a Systems Perspective
A literature review developing a framework for systems thinking in health identified six core characteristics the approach relies on: recognizing interconnections and system structure, identifying feedback, finding leverage points, understanding dynamic behavior over time, using mental models to explore possible solutions, and creating simulation models to test policies before deploying them.2PubMed Central. Development of the Systems Thinking for Health Actions framework: a literature review and a case study Those six ideas sound abstract, but they map onto very concrete things that happen in real health systems.
Feedback Loops
Feedback loops are the engine of systems behavior. A reinforcing loop amplifies change: more of something leads to even more of it. A balancing loop resists change: a push in one direction triggers a counterforce. In chronic disease modeling, for example, population growth feeds back into births (a reinforcing loop), while population and deaths form a balancing loop.3Preventing Chronic Diseases. Applications of System Dynamics Models in Chronic Disease Prevention: A Systematic Review You can see how this plays out in a real-world health problem: an industry that markets sugary foods successfully increases consumption, which raises social acceptability of those products, which makes marketing easier, which increases consumption further. A study in Chile mapped exactly this kind of reinforcing feedback, showing how food industries amplified their influence over purchasing decisions, brand acceptability, and political influence through interconnected loops.4Community Dental Health. Commercial Determinants and Sugar Consumption Patterns in Chile: A Systems Thinking Approach
Time Delays
Prevention measures almost never produce immediate results, and that delay is one of the most underappreciated features of health systems. Time delays in feedback loops cause instability, overshoot, and oscillation, and they undermine our ability to learn from experience because the effects of a policy show up long after the policy was enacted.3Preventing Chronic Diseases. Applications of System Dynamics Models in Chronic Disease Prevention: A Systematic Review This is one reason tobacco control policies took decades to show their full population-level impact, and why politicians often lose patience with public health strategies that work on a timescale longer than an election cycle.
Leverage Points
Perhaps the most practically useful concept in systems thinking is the leverage point: a place in the system where a small change can produce a large shift in outcomes. The idea was popularized by Donella Meadows, whose framework ranked different types of interventions by how deeply they alter system behavior. Researchers have adapted Meadows’s framework specifically for public health, creating what they call the “Public Health 12” framework that helps practitioners identify where leverage can be applied and where gaps exist in current efforts.5PubMed Central. The Public Health 12 framework: interpreting the ‘Meadows 12 places to act in a system’ for use in public health The logic is that not all interventions are equal. Changing a parameter (say, raising a tax) can help, but changing the goals or beliefs that drive the system’s structure is far more powerful, even if it is harder.
The Action Scales Model, developed from practice, formalizes this idea by distinguishing four system levels: observable events and behaviors, the structures that produce those events, the goals the system works toward, and the deeply held beliefs underlying everything else. Changes at deeper levels tend to cascade outward, reshaping the superficial levels along the way.6PLoS ONE. Leverage point themes within Dutch municipalities’ healthy weight approaches: A qualitative study from a systems perspective
Tools That Make the Abstract Concrete
Systems thinking without tools stays philosophical. The field has developed several practical methods for turning systems ideas into something you can actually work with. The most commonly used tools, based on a review of published studies, are causal loop diagrams, system dynamics modeling, agent-based modeling, and concept mapping.2PubMed Central. Development of the Systems Thinking for Health Actions framework: a literature review and a case study
Causal loop diagrams are the simplest entry point. They are visual maps that show how variables in a system connect and influence each other, with arrows indicating direction and signs indicating whether the influence is reinforcing or balancing. You don’t need a computer to make one. Community groups, policymakers, and researchers can draw them together on a whiteboard. In the Caribbean, for example, researchers developed a causal loop diagram across multiple island nations showing nine reinforcing loops that linked schools, policy environments, commercial forces, community factors, and children’s personal experiences to rising childhood obesity rates.7PubMed Central. A systems thinking framework for understanding rising childhood obesity in the Caribbean
System dynamics models add mathematics to those diagrams. They simulate how variables change over time, allowing researchers to test “what if” scenarios before anyone implements a policy. The SimSmoke model, for instance, simulates the effects of tobacco control policies on smoking rates within a dynamic social system, helping policymakers compare the likely impact of different interventions like tax increases, clean-air laws, and advertising bans.8PubMed Central. Simulation modeling and tobacco control: creating more robust public health policies A version of that model has been used to examine the interrelationship between cigarette use and smokeless tobacco policies in the United States.9PubMed Central. The US SimSmoke tobacco control policy model of smokeless tobacco and cigarette use In Australia, a system dynamics model co-developed with local and national stakeholders was used to investigate the potential impacts of various tobacco control interventions on adult daily smoking prevalence in Queensland.10Tobacco Control. Policy options for endgame planning in tobacco control: a simulation modelling study
Agent-based models take a different angle. Instead of modeling aggregate flows, they simulate individual “agents” (people, organizations, pathogens) who interact with each other and their environment according to defined rules, then observe what collective patterns emerge from those interactions.11PubMed Central. Agent-Based Modeling in Public Health: Current Applications and Future Directions This bottom-up approach is especially useful for modeling the spread of infectious disease, where individual contacts and behaviors matter enormously.
Where Systems Thinking Is Being Applied
The range of public health topics that have adopted systems approaches is broader than most people realize. Obesity research has been at the forefront. The Foresight obesity map, developed by experts in the United Kingdom, became one of the best-known systems maps in public health, charting the complex web of factors that drive weight gain at a population level. Researchers have since compared that expert-developed map against community-created causal loop diagrams, finding that the perspectives overlap in some areas but diverge in others, which suggests that both expert and community views are needed to get a complete picture.12PubMed Central. Comparing complex perspectives on obesity drivers: action‐driven communities and evidence‐oriented experts
Food systems are a natural fit. One study created a systems map specifically for retailers and public health practitioners seeking to shift community-based organizations toward healthier food provision, identifying intervention points and key feedback loops that could drive unintended consequences of healthy food retail policies.13Food Policy. Mapping factors associated with a successful shift towards healthier food retail in community-based organisations: A systems approach The insight here is that well-meaning food policies can backfire if they don’t account for how the system will adapt around them.
Neglected tropical diseases provide another example. Researchers applied systems thinking to identify five leverage points for accelerating progress: clarifying realistic elimination goals, increasing support for interventions beyond drug delivery, reducing dependency on international donors, creating a less insular culture within the global neglected tropical disease community, and addressing health worker incentives.14PubMed Central. Applied systems thinking: a viable approach to identify leverage points for accelerating progress towards ending neglected tropical diseases Several of those leverage points have nothing to do with medicine and everything to do with organizational culture and economics, which is precisely the kind of insight that a disease-focused, linear approach would miss.
Antimicrobial resistance is gaining attention as a systems problem too. Researchers have argued that One Health systems-thinking approaches are crucial for understanding and predicting how human activities interact within environmental subsystems to drive the emergence and transmission of drug-resistant organisms.15PubMed. The need for One Health systems-thinking approaches to understand multiscale dissemination of antimicrobial resistance Antibiotic use in agriculture, waste management, water treatment, and clinical medicine all feed into each other. Tackling resistance in hospitals while ignoring antibiotic runoff from farms addresses only one part of the loop.
Unintended Consequences and the “Fixes That Fail” Problem
One of the strongest arguments for systems thinking is its ability to anticipate side effects that conventional analysis misses. In systems language, a “fixes that fail” archetype describes a situation where a policy solves a visible problem in the short term but triggers adaptations elsewhere in the system that bring the original problem back, often worse than before. Researchers studying wildlife policy across China, the Democratic Republic of the Congo, and the Philippines used this archetype to analyze how well-intentioned bans on wildlife trade could displace activity into informal channels, potentially increasing rather than decreasing zoonotic disease risk.16PubMed Central. Wildlife policy, the food system and One Health: a complex systems analysis of unintended consequences for the prevention of emerging zoonoses in China, the Democratic Republic of the Congo and the Philippines
The commercial determinants of health represent a related area. Meadows’s framework has been used to analyze how unhealthy commodity industries, such as tobacco, alcohol, and ultra-processed food companies, influence public health policy in ways that are difficult to counter one regulation at a time.17PubMed Central. Systems Thinking as a Framework for Analyzing Commercial Determinants of Health The industry doesn’t just lobby against a specific tax. It builds political coalitions, funds sympathetic research, deploys corporate social responsibility campaigns, and shapes consumer preferences simultaneously, creating a system of influence with multiple reinforcing loops. A systems analysis makes that architecture visible.
Bringing Communities Into the Mapping Process
Systems maps drawn exclusively by researchers or bureaucrats reflect only their view of the problem. A growing body of work is bringing communities, particularly the people most affected by a health issue, into the mapping process. A scoping review of participatory systems mapping identified 57 studies that used the method for a variety of purposes, including informing policy and identifying leverage points. The review found that while policymakers and professionals were most often included, some studies described significant added value from including marginalized communities. The reported benefits related mostly to individual and group learning, though a limitation was that concrete actions did not always follow from the mapping exercises.18PubMed Central. The use of participatory systems mapping as a research method in the context of non-communicable diseases and risk factors: a scoping review
One program that tracked outcomes more closely found that community participants experienced increases in systems thinking capacity alongside increases in health equity thinking and action, particularly around social and structural determinants of health and a commitment to targeted actions.19PubMed Central. Changes in systems thinking and health equity considerations across four communities participating in Catalyzing Communities Participatory group model building, where community members and researchers collaborate to construct a simulation model together, has been proposed as a method for designing sustainable interventions for adolescent mental health, integrating diverse local understandings of complex processes to create contextually relevant policies.20PubMed Central. A systemic approach to identifying sustainable community-based interventions for improving adolescent mental health: a participatory group model building and design protocol
An analysis of public health recommendations around the social determinants of health found a shift over time towards what the authors called paradigmatic change and away from individual interventions. The analysis also revealed potentially powerful system intervention points that remain underused in public health action on social determinants.21PubMed Central. Systems change for the social determinants of health In other words, the field’s own recommendations are moving toward deeper system-level change, but practice has not caught up.
Hard Systems Versus Soft Systems
Not all systems thinking in public health works the same way. The field has two broad traditions that differ in what they assume about the world. Hard systems thinking treats the world as containing systems that can be engineered: you define the problem, build a model, optimize the variables, and find the best answer. Soft systems thinking, by contrast, starts from the assumption that the world is “problematical” rather than simply containing solvable systems. Its models are not meant to replicate reality but to help people explore and negotiate their different perspectives on a situation. Hard systems thinking ends when the right answer is identified. Soft systems thinking treats inquiry as never-ending.22PubMed Central. Systems thinking in, and for, public health: a call for a broader path
In practice, public health tends to gravitate toward the hard end of the spectrum, building quantitative models and running simulations. The soft tradition is less visible but arguably just as important, especially for problems where stakeholders disagree on what the problem even is. A childhood obesity initiative in which parents, school administrators, food retailers, and local government each see a completely different system is a soft-systems problem before it is a modeling problem. Skipping the messy negotiation of perspectives and jumping straight to a quantitative model risks building an elegant simulation of the wrong system.
The Data Challenge
Systems models are only as good as the data feeding them, and in public health the data situation is often messy. Researchers working on opioid systems modeling documented several common challenges: no single data source provides all the qualities needed, definitions and measurement methods vary across sources, available data often provide only approximations prone to errors and biases, and changes in data collection practices over time make it difficult to distinguish real trends from reporting artifacts, complicating model calibration.23PubMed Central. Data Needs in Opioid Systems Modeling: Challenges and Future Directions These are not unique to opioid modeling; the same issues surface whenever you try to build a quantitative systems model of a health problem that spans clinical records, survey data, economic statistics, and behavioral measures collected by different agencies using different definitions.
This is part of why some applications of systems thinking remain at the qualitative diagram stage rather than progressing to full simulation models. A causal loop diagram can capture stakeholder knowledge about how a system works even when the data for quantitative modeling do not exist. That has real value, but it also means the predictions remain more like informed hypotheses than testable forecasts.
Evaluating Whether Systems Approaches Actually Work
Evaluating the impact of systems-oriented public health interventions poses its own set of difficulties. Traditional randomized controlled trials assume you can isolate an intervention and measure its effect in a controlled setting. Systems interventions, by design, are meant to work through multiple pathways simultaneously, adapt to local context, and trigger feedback effects that ripple out over time. That makes the classic trial design a poor fit.
A range of alternatives has been proposed, including cluster randomized trials, stepped wedge designs, interrupted time series, multiple baseline designs, and controlled pre-post studies.24PubMed. Evaluation of systems-oriented public health interventions: alternative research designs Contribution analysis, a theory-based approach that shifts from asking “did this intervention cause the outcome?” to “how did this intervention contribute to the outcome in this context?”, has been advanced as another option for complex health interventions.25International Journal of Integrated Care. Evaluating Complex Health Interventions through Contribution Analysis: A Scoping Review Guidance has also been developed for using complex systems models specifically in economic evaluations: quantitative, dynamic, non-linear models that incorporate feedback and interactions among elements in order to capture emergent outcomes and estimate health and economic consequences for informing policy.26PubMed. Guidance on the use of complex systems models for economic evaluations of public health interventions
The honest assessment is that evaluation methods for systems approaches are still maturing. The field knows the traditional toolkit is insufficient but hasn’t yet converged on a universally accepted replacement. That gap is a genuine limitation, not just an academic concern: funders want evidence that an approach works before investing in it, and if the evaluation methods aren’t settled, making the case for systems-level spending gets harder.
Training the Public Health Workforce
You can’t apply systems thinking if you were never taught it, and most public health professionals were trained in a linear, risk-factor-oriented paradigm. Efforts to change that are underway. The Region 2 Public Health Training Center developed online practice-based training modules that integrate systems thinking strategies into public health policymaking skills.27Journal of Public Health Management and Practice. Incorporating Systems Thinking Approaches Into Practice-Based Training to Strengthen Policy-Making Skills At the postgraduate level, a scoping review of curricular approaches found a range of methods being used to teach systems thinking in public health programs, providing what the authors described as a descriptive menu of approaches that could inform curriculum design and faculty development.28PubMed Central. Teaching Systems Thinking in Postgraduate Public Health: A Scoping Review of Curricular Approaches
The challenge isn’t just curriculum. Systems thinking requires comfort with ambiguity, iterative learning, and stakeholder negotiation, skills that sit uneasily in institutions built around evidence hierarchies and standardized protocols. A practitioner who has spent a career designing programs around randomized trial evidence may find the soft-systems end of the spectrum, where there is no single right answer and inquiry never ends, genuinely disorienting. That cultural shift is probably the biggest barrier of all, and it won’t be solved by adding a module to a training program.
Resilience and Pandemic Response
The COVID-19 pandemic offered a global stress test for systems thinking. Researchers studying resilience argued that times of crisis offer a rare opportunity to understand the mechanisms underpinning the resilience of complex adaptive systems.29Ecology and Society. Resilience in the times of COVID: what the response to the COVID pandemic teaches us about resilience principles The pandemic exposed exactly the kind of feedback loops and unintended consequences that systems thinkers have long warned about: lockdowns reduced viral transmission but disrupted food systems, mental health services, and childhood vaccination programs. Travel bans slowed importation of cases but strained supply chains for medical equipment. Each intervention sent ripples through interconnected systems in ways that linear planning had not anticipated.
The pandemic also demonstrated the value of simulation modeling. Countries that used dynamic models to project how different combinations of interventions would play out, rather than relying on single-variable forecasts, generally made better decisions about timing and sequencing of restrictions. Whether or not the word “systems thinking” was used, the underlying logic was the same: model the whole system, watch for feedback, and expect time delays between action and result.