Executive interactive model
Systems Thinking: From Insight to Systems Management
Public-health and healthcare outcomes emerge from the relationships among knowledge, people, incentives, structures, decisions, and feedback loops. This model turns that premise into tools leaders can use.
- Complex
- Adaptive
- Interdependent
- Learning-oriented
- Model version 2.0
What a system needs to function
Four interdependent domains from Leischow et al. (2008), now held inside two contemporary layers the peer review found missing. Select any part of the model.
Executive premise
Isolated programs rarely resolve system-level problems. Sustainable performance requires disciplined design of the relationships among data, people, workflow, governance, and learning.
Select a node, a layer, or a button below. Theory-of-change entries are the author’s extension, written to answer Section 7 of the peer review.
Domain 01 of the original framework
Build a common intelligence system.
Integrate formal data with what teams learn in real work. Make information timely, trusted, usable, and available where decisions are made.
Executive action
Define data governance, a shared operating picture, a learning agenda, and a structured route for frontline intelligence.
Theory of change
- Mechanism
- Timely, trusted information reaches the points where decisions are made.
- Intermediate outcome
- Faster, better-informed decisions with fewer blind spots between departments.
- Feedback loop
- Trusted data gets used; use exposes quality gaps; fixing them raises trust further (reinforcing).
- Measurable endpoint
- Decision latency, data completeness, and the share of frontline issues that receive a documented response.
- Where it can fail
- Data arrives after the decision is made, or frontline input goes unanswered and people stop offering it.
Domain 02 of the original framework
Design the relationships, not just the roster.
Connect stakeholders across disciplines and institutions through structures that carry real authority and real information.
Executive action
Define accountable partnership structures, shared decision rights, referral and communication pathways, and network health measures.
Theory of change
- Mechanism
- Explicit relationships and shared decision rights reduce handoff failure across boundaries.
- Intermediate outcome
- Fewer dropped referrals and faster decisions that involve more than one organization.
- Feedback loop
- Successful joint work builds trust; trust widens what partners are willing to decide together (reinforcing).
- Measurable endpoint
- Referral completion, network density, and time to a cross-organization decision.
- Where it can fail
- Partnerships exist on paper without decision rights, resources, or a named lead.
Domain 03 of the original framework
Test assumptions before committing resources.
Model interactions, feedback, and delays so leverage points and trade-offs surface before a major investment or policy change.
Executive action
Use causal maps, scenario modeling, and process simulation, each documented well enough to be challenged and revised.
Theory of change
- Mechanism
- Modeling exposes delays, feedback, and leverage before money or policy is committed.
- Intermediate outcome
- Fewer failed initiatives and trade-offs found while they are still cheap to address.
- Feedback loop
- Models updated with observed results forecast better next time (error correcting).
- Measurable endpoint
- Forecast versus actual performance and the number of initiatives redesigned or stopped before launch.
- Where it can fail
- The model is treated as the intervention, or its assumptions stay hidden from the people who must act on it.
Domain 04 of the original framework
Pair central direction with bounded local adaptation.
Replace rigid command-and-control patterns with structures that set direction centrally and let teams adapt tactics to real conditions.
Executive action
Pair centralized strategic governance with local adaptation rules, clear escalation pathways, and rapid-cycle improvement routines.
Theory of change
- Mechanism
- Central direction plus explicit rules for what local teams may change.
- Intermediate outcome
- Faster adaptation without losing the strategic aim.
- Feedback loop
- Local learning is escalated, and proven changes update enterprise standards (reinforcing).
- Measurable endpoint
- Cycle time from signal to adjustment and time to resolve an escalation.
- Where it can fail
- Adaptation becomes unbounded variation, or hierarchy blocks what teams learn from reaching leaders.
Contemporary layer (added in version 2.0)
Connect analysis to adoption, fidelity, and credible evaluation.
The peer review found the original framework strong but missing an implementation scaffold. This layer supplies it: owners, resources, fidelity, adaptation governance, and evaluation against a credible comparison.
Executive action
Attach an implementation plan and an evaluation design to every systems-informed initiative before launch.
Theory of change
- Mechanism
- Named owners, committed resources, and tracked fidelity turn a design into practice.
- Intermediate outcome
- Adoption at fidelity, with adaptations governed rather than improvised.
- Feedback loop
- Evaluation findings adjust the plan; the adjusted plan is evaluated again.
- Measurable endpoint
- Adoption, fidelity, and outcome change relative to a credible comparison condition.
- Where it can fail
- System activity is mistaken for system improvement.
Contemporary layer (added in version 2.0)
Let affected people shape the definition of success.
Patients, communities, frontline staff, and historically marginalized groups shape problem definition, success criteria, intervention selection, and the reading of trade-offs.
Executive action
Require inclusive stakeholder governance, transparent decision criteria, stratified outcome measures, and a documented review of who gains and who bears the burden.
Theory of change
- Mechanism
- The people affected by a system help define its problems and its success criteria.
- Intermediate outcome
- Interventions that fit the populations they serve, with legitimacy that sustains them.
- Feedback loop
- Stratified results reveal gaps; the gaps redirect effort and resources.
- Measurable endpoint
- Stratified outcomes, access by population, and whether measured gaps narrow.
- Where it can fail
- Communities and staff are treated only as data sources or end users.
The executive question
Not simply “What intervention should we adopt?” but “What system conditions are generating this outcome, who shapes them, what feedback loops sustain them, what trade-offs would a change create, and where is the highest-leverage point for action?”
What changed in version 2.0
- Two contemporary layers, implementation and evaluation, and equity and legitimacy, now surround the original four domains.
- Every domain carries a theory of change: mechanism, intermediate outcome, feedback loop, endpoint, and failure condition.
- Seven working tools: a problem screener, stage gates, a printable decision charter, an artifact card with an assumption register, a balanced measure builder, a roles matrix, and a maturity self-assessment.
- The peer-review scorecard now matches the published report, including the contemporary relevance rating.
Does this problem need a systems approach?
The peer review recommends systems thinking for problems that cross boundaries, have multiple causes, produce delayed effects, or recur despite local fixes. Answer four questions to find out which applies.
Question 1 of 4
Does the problem cross functional or organizational boundaries?
Several departments, disciplines, or organizations each hold part of the cause or the fix.
Question 2 of 4
Does it have multiple interacting causes?
No single owner or single change would resolve it.
Question 3 of 4
Are the effects delayed?
Actions taken now show up weeks or months later, or somewhere else.
Question 4 of 4
Has it recurred despite repeated local fixes?
Past projects improved it briefly, then it returned.
Answer the four questions to see a recommendation.
Recurrence despite local fixes is the strongest single signal that the cause sits between departments, not inside one.
How the score reads
0 to 2: a conventional, well-owned project is likely enough.
3 to 5: use a systems lens on a scoped basis. Complete the Map stage before committing resources.
6 to 8: run the full six-stage operating cycle.
These cut points are the author’s pilot triage heuristic. They have not been validated and should support judgment, not replace it.
How leaders turn complexity into action
Six stages from the peer review's systems-informed operating model. Each stage now carries gate criteria, so a team knows when it is ready to move on. The ring around each stage fills as its gates are met.
Stage 1 of 6. Start with the outcome
Frame the system-level problem.
- Leadership question
- What outcome requires a system-level response?
- Core deliverable
- A charter defining the population, boundary, objective, sponsor, and time horizon.
- Illustrative measures
- Outcome baseline; equity stratification; scope confirmation.
Gate criteria 0 of 4 met
Stage 2 of 6. See how the system produces the result
Map how the current system creates the outcome.
- Leadership question
- How does the current system create the outcome?
- Core deliverable
- Stakeholder-informed causal map, process map, and network map.
- Illustrative measures
- Delay points, handoff failures, network density, and key constraints.
Gate criteria 0 of 3 met
Stage 3 of 6. Find leverage before committing
Test leverage points and trade-offs.
- Leadership question
- Where are the leverage points and likely trade-offs?
- Core deliverable
- Intervention portfolio with explicit hypotheses and scenario tests.
- Illustrative measures
- Expected benefit, cost, risk, and unintended consequences.
Gate criteria 0 of 3 met
Stage 4 of 6. Govern the change
Implement with clear authority.
- Leadership question
- How will the change be implemented and governed?
- Core deliverable
- Implementation plan with decision rights, owners, resource commitments, and escalation mechanisms.
- Illustrative measures
- Adoption; fidelity; capacity; training; implementation barriers.
Gate criteria 0 of 4 met
Stage 5 of 6. Adjust on evidence
Learn and adapt on a fixed cadence.
- Leadership question
- How will the organization learn and adapt?
- Core deliverable
- Learning system with review cadence, feedback loops, and adjustment rules.
- Illustrative measures
- Process measures, balancing measures, and qualitative intelligence.
Gate criteria 0 of 3 met
Stage 6 of 6. Decide what happens next
Decide what to sustain, scale, redesign, or stop.
- Leadership question
- What must be sustained, scaled, redesigned, or stopped?
- Core deliverable
- Executive decision record and sustainability plan.
- Illustrative measures
- Outcome durability; distributional effects; ROI; deimplementation triggers.
Gate criteria 0 of 3 met
Record the decision
Examples of deimplementation triggers: a balancing measure worsens beyond an agreed threshold, a distributional gap widens, or the outcome fails to move after the agreed number of review cycles.
Questions, deliverables, and measures are from the published report. Gate criteria are the author’s version 2.0 extension.
Stage 1 decision charter
Frame the problem before anyone maps it. Complete the fields and the charter on the right updates as you type. It prints on one page and downloads as a Word document.
Entries stay in this browser only. Nothing is sent to a server.
Is this map a decision aid, or just a diagram?
The first executive discipline: a systems artifact earns its place only when it names a decision owner, a near-term action, a testable hypothesis, and a measure of progress. Document the model, then log the assumptions it rests on.
Artifact card
Diagram only, not yet a decision aid.
Missing: decision owner, near-term action, testable hypothesis, progress measure.
Documentation standard 0 of 7
Minimum reporting items the peer review recommends for any causal-loop diagram, system dynamics model, network analysis, or concept map.
Assumption register
Record what is known, what is inferred, and what remains uncertain, and what would cause the analysis to be updated.
Write the assumption before adding it.
| Assumption | Status | Basis | Update trigger | Remove |
|---|---|---|---|---|
| No assumptions logged yet. Add the first one above; most systems maps rest on at least three. | ||||
Build a balanced measurement set
The peer review calls for evaluation that covers process, outcome, balancing, and distributional measures. A set missing any quadrant cannot see part of what the change is doing.
Process
Is the change happening as designed?
Outcome
Is the target result moving?
Balancing
What else changed because of it?
Distributional
Who gained, and who bore the burden?
Coverage
Gaps
- Select or add measures to check coverage.
Who does what at each stage
Systems leadership needs both facilitation and accountable authority. Select a cell to cycle its assignment. The check on the right applies the rule behind pairing adaptation with accountability: every stage has exactly one accountable role.
| Role | 1. Frame | 2. Map | 3. Test | 4. Implement | 5. Learn | 6. Sustain |
|---|---|---|---|---|---|---|
| Senior sponsor | ||||||
| Operational owner | ||||||
| Data leader | ||||||
| Frontline teams | ||||||
| Community partners | ||||||
| Independent evaluator |
A AccountableR ResponsibleC ConsultedI Informed– Not involved
The starting matrix is an illustrative default by the author. Adapt it to local structure.
Governance check
How ready is your organization?
Rate eighteen statements across the four domains and the two contemporary layers. The profile shows where the system is strong and where it needs reinforcement first.
1 Not in place2 Beginning3 Partly4 Mostly5 Fully in place
Your profile
0 of 18 statements rated.
First priority
Bands (below 2.0 Emerging, 2.0 to 2.9 Developing, 3.0 to 3.9 Established, 4.0 and above Leading) divide the scale evenly. They are unvalidated pilot bands for discussion, not a benchmark.
Worked case: emergency department boarding
The peer review lists emergency department boarding among the problems that recur despite repeated local fixes. Here is how the six stages apply to it.
This is an illustrative walk-through of the method. It contains no organizational data and makes no claim about effect sizes.
- 1
Frame
Define the outcome as the time from an admission decision to an inpatient bed. The boundary spans the emergency department, inpatient units, bed management, discharge planning, and post-acute partners. The chief operating officer sponsors. Boarding time will be stratified by age, payer, preferred language, and behavioral health status.
- 2
Map
The causal map shows a reinforcing loop: boarding crowds the department, crowding lengthens waits and walkouts, strain slows throughput, and slower throughput produces more boarding. Late-day discharges and post-acute placement delays feed inpatient occupancy from outside the department. The map also exposes a local-fix trap: hallway beds relieve crowding briefly but leave the inpatient constraint untouched.
Illustrative causal loop. Arrows marked + move in the same direction. Outside inputs such as late-day discharges and post-acute placement delays feed the loop through inpatient occupancy. - 3
Test
Candidate interventions are written as hypotheses, such as earlier discharge timing and coordinated post-acute placement. Scenario tests weigh expected benefit against cost, readmission risk, and inpatient staff workload before anything is funded.
- 4
Implement
Bed management receives explicit decision authority, escalation triggers are agreed in advance, and adaptation rules state what individual units may adjust and what stays standard across the enterprise.
- 5
Learn
A weekly review looks at process measures (discharge timing), the outcome (boarding time), balancing measures (readmissions, staff workload), and distributional measures (boarding time by group) together.
- 6
Sustain
At the decision point, leaders record whether to sustain, scale, redesign, or stop. A predefined deimplementation trigger applies if a balancing measure worsens beyond an agreed threshold.
Why the local fix failed
Individual components may perform well while the overall system fails. The department was not the only place the boarding problem lived.
The peer review behind the model
A narrative critical peer review of Leischow et al. (2008), informed by later systems-science, implementation, and evaluation literature. It is not a systematic review or meta-analysis.
Editorial recommendation
Accept with major revisions as a contemporary conceptual manuscript. Retain as foundational reading for executive education and systems-oriented public-health practice.
Key findings
Enduring conceptual contribution
Public health is treated correctly as a complex adaptive system with nonlinear causal pathways, distributed across sectors and subject to feedback, delays, and unintended consequences.
Necessary but insufficient operational model
The four domains are a credible organizing framework, but a contemporary operating model needs explicit decision rights, resource allocation, implementation strategy, performance measures, and adaptation rules.
Evidence and evaluation gap
Methods such as network analysis, concept mapping, and system dynamics are illustrated, with limited evidence linking the activity to population-level improvement or policy effects.
Contemporary revision requirement
A current version should integrate implementation science, equity-centered governance, transparent model documentation, and evaluation covering process, outcome, balancing, and distributional measures.
Critical appraisal scorecard
Ratings as published in the July 2026 report.
Importance of problem framing
The systems perspective remains central to population-health and healthcare transformation.
Conceptual integration
The four domains form a coherent organizing architecture for collaborative, adaptive work.
Empirical substantiation
Illustrative examples are useful, but direct evidence of outcomes is limited.
Method transparency
Models and analyses are referenced, but not fully documented.
Operational specificity
Provides strategic direction but not a complete implementation playbook.
Equity and power analysis
Requires explicit attention to legitimacy, representation, and distributional consequences.
Contemporary relevance
Later systems, evaluation, and implementation scholarship have reinforced the central thesis.
Five practices that make systems thinking operational
Practice 01. From the published report
Use the map to make a decision.
A causal-loop diagram, stakeholder map, or simulation is a decision aid, not the intervention itself. Require each systems artifact to identify a decision owner, a near-term action, a testable hypothesis, and an explicit measure of progress.
Practice 02. From the published report
Write down what you know, and what you only believe.
Systems work is especially vulnerable to unexamined assumptions about boundaries, causality, data reliability, and stakeholder incentives. Governance forums should require teams to document what is known, what is inferred, what remains uncertain, and how the analysis will be updated.
Practice 03. From the published report
Adapt tactics, never the aim.
Adaptive implementation does not mean unbounded local variation. Define non-negotiable outcomes, decision rights, performance thresholds, escalation pathways, and the conditions under which local teams may alter tactics.
Practice 04. From the published report
Put equity on the scorecard.
Equity belongs in the operating scorecard, not only in an introductory statement. Review stratified outcomes, access, patient and workforce experience, and distributional consequences alongside utilization, quality, finance, and implementation metrics.
Practice 05. Version 2.0 extension, drawn from Stage 5 of the operating model
Make learning a standing meeting, not an event.
Schedule reviews that examine process, outcome, balancing, and distributional measures together, and agree in advance which results trigger an adjustment, an escalation, or a stop.
References
- Leischow, S. J., Best, A., Trochim, W. M., Clark, P. I., Gallagher, R. S., Marcus, S. E., & Matthews, E. (2008). Systems thinking to improve the public’s health. American Journal of Preventive Medicine, 35(2 Suppl.), S196-S203. https://doi.org/10.1016/j.amepre.2008.05.014
- Carey, G., Malbon, E., Carey, N., Joyce, A., Crammond, B., & Carey, A. (2015). Systems science and systems thinking for public health: A systematic review of the field. BMJ Open, 5(12), e009002. https://doi.org/10.1136/bmjopen-2015-009002
- Peters, D. H. (2014). The application of systems thinking in health: Why use systems thinking? Health Research Policy and Systems, 12, 51. https://doi.org/10.1186/1478-4505-12-51
- Bagnall, A.-M., Radley, D., Jones, R., Gately, P., Nobles, J., Van Dijk, M., Blackshaw, J., Montel, S., & Sahota, P. (2019). Whole systems approaches to obesity and other complex public health challenges: A systematic review. BMC Public Health, 19, 8. https://doi.org/10.1186/s12889-018-6274-z
- McGill, E., Er, V., Penney, T., Egan, M., White, M., Meier, P., Whitehead, M., Lock, K., Anderson de Cuevas, R., Smith, R., Savona, N., Rutter, H., Marks, D., de Vocht, F., Cummins, S., Popay, J., & Petticrew, M. (2021). Evaluation of public health interventions from a complex systems perspective: A research methods review. Social Science & Medicine, 272, 113697. https://doi.org/10.1016/j.socscimed.2021.113697
- Rutter, H., Savona, N., Glonti, K., Bibby, J., Cummins, S., Finegood, D. T., Greaves, F., Harper, L., Hawe, P., Moore, L., Petticrew, M., Rehfuess, E., Shiell, A., Thomas, J., & White, M. (2017). The need for a complex systems model of evidence for public health. The Lancet, 390(10112), 2602-2604. https://doi.org/10.1016/S0140-6736(17)31267-9
- Whelan, J., Fraser, P., Bolton, K. A., Love, P., Strugnell, C., Boelsen-Robinson, T., Blake, M. R., Martin, E., Allender, S., & Bell, C. (2023). Combining systems thinking approaches and implementation science constructs within community-based prevention: A systematic review. Health Research Policy and Systems, 21, 85. https://doi.org/10.1186/s12961-023-01023-4
- Northridge, M. E., & Metcalf, S. S. (2016). Enhancing implementation science by applying the best principles of systems science. Health Research Policy and Systems, 14, 74. https://doi.org/10.1186/s12961-016-0146-8
- Gadsby, E. W., & Wilding, H. (2024). Systems thinking in, and for, public health: A call for a broader path. Health Promotion International, 39(4), daae086. https://doi.org/10.1093/heapro/daae086