A Systems Thinking Perspective for Healthcare Professionals

A Familiar Scene

A patient receives the wrong medication dose and suffers a serious adverse event. An investigation is launched. The team pores over records, interviews staff, and eventually arrives at a conclusion: a nurse administered the wrong concentration after misreading a label. Case closed. Root cause identified.

But is it really that simple?

If you have spent any time working in a clinical environment, you know that the honest answer is almost always no. That nurse was working her fourteenth hour of an understaffed overnight shift. The medication labels had been flagged as confusingly similar by three staff members in the months before the event, with no formal response. The electronic medication administration system had a known alert fatigue problem that led most clinicians to dismiss pop-up warnings reflexively. Purchasing decisions made months earlier had introduced a new drug concentration without accompanying protocol updates. And a regulatory accreditation cycle had absorbed the quality team’s bandwidth for the previous quarter, delaying a planned medication safety audit.

The nurse did not cause this event. She was the last domino to fall in a chain that stretched far beyond her reach.

Healthcare adverse events are rarely the product of a single cause. They are emergent outcomes of interconnected systems operating under accumulated pressure.

This article explores why the search for a single root cause so often leads healthcare organizations astray, and how systems thinking tools like AcciMap can help teams see and address the fuller picture.

The Problem with “Root Cause”

Root cause analysis (RCA) has been a standard tool in healthcare quality improvement for decades. The Joint Commission requires RCA following sentinel events, and the method has undeniably helped shift attention away from purely blaming individuals toward examining system factors.

The landmark 1999 Institute of Medicine report To Err Is Human was a watershed moment in this shift. It estimated that between 44,000 and 98,000 Americans died annually from preventable medical errors and made the critical argument that most errors are systemic in nature and cannot be resolved at the level of individual healthcare providers (Kohn et al., 2000).

A study cited in that same IOM report found that 78% of all adverse drug events result from system failures, not individual mistakes (Eisenberg, 2000). This was a profound reframing. Yet despite this insight, investigations in practice have often struggled to fully escape the gravity of individual blame.

Research examining 300 patient safety incident investigations from 56 Australian health services found that nearly half of all identified contributing factors (47%) focused disproportionately on the people directly involved in the incident rather than on other system levels, and as a consequence, most recommendations were of medium or weak strength (Bowditch et al., 2025). Investigation practice has not yet caught up with the science.

The language of “root cause” is itself part of the problem. The term implies that a single originating failure can be located and corrected, like pulling a weed by its root. In genuinely complex systems, this is rarely how events unfold. Major healthcare incidents are not linear. They are what researchers call emergent outcomes: results that arise from the dynamic interaction of multiple contributing factors across multiple levels of a system, none of which would have been sufficient alone.

As one analysis of patient safety science observed, linear, unidimensional solutions applied to care settings have repeatedly fallen short of expectations, and the problem is a wicked one that resists simple fixes (Braithwaite et al., 2015).

Introducing AcciMap: Seeing the Whole System

One of the most powerful frameworks for moving beyond single-cause thinking is the AcciMap approach, originally developed by Danish engineer Jens Rasmussen in 1997 as part of a broader risk management strategy for complex sociotechnical systems (Rasmussen, 1997).

AcciMap graphically maps the contributing factors to an adverse event across multiple interconnected system levels, along with the relationships between them (Salmon et al., 2012). In healthcare, these levels typically include: government policy and budgeting; regulatory bodies and associations; organizational management; technical and operational management; frontline processes and conditions; and immediate outcomes.

The power of this approach is not just in what it maps, but in what it forces you to see. When every contributing factor is visually placed at its correct system level and connected to others by lines of influence, the picture that emerges is almost never a single chain. It is a web.

Accimaps model the dynamic interaction between multiple sociotechnical levels, including regulatory, organizational, and workplace factors, and account for the role these play in shaping the course of an event as it unfolds over time (Waterson et al., 2016).

Importantly, the framework recognizes that decisions made at the highest levels of a system, such as government funding priorities or regulatory frameworks, propagate downward through organizations and eventually shape the conditions frontline clinicians work in. A staffing ratio is not just a scheduling problem. It is the downstream consequence of budgeting decisions, workforce policies, and management choices that accumulated long before a patient was ever harmed.

The Six Levels Where Events Are Made

To understand how healthcare events emerge from interconnected systems, it helps to walk through the levels AcciMap illuminates. Each level contributes to the conditions that either protect or threaten patient safety.

Government Policy and Resource Allocation

At the uppermost level of the sociotechnical system sit government bodies that set healthcare funding, staffing regulations, scope-of-practice laws, and broader system priorities. These decisions shape the resource environment in which every organization operates (Svedung & Rasmussen, 2002).

When governments impose funding freezes, cap workforce growth, or fail to fund infrastructure upgrades, the effects do not stay at the policy level. They flow into hospital budgets, staffing models, equipment procurement, and ultimately the environment in which care is delivered. A clinical decision made under resource pressure is not simply a frontline choice; it is a downstream expression of upstream constraints.

Regulatory Bodies and Accreditation Pressures

Regulatory frameworks set the tone for legal protections, reporting mandates, and liability boundaries. When they promote non-punitive policies, they empower leadership to implement just culture principles. When they are primarily compliance-driven, they can inadvertently distort organizational priorities (Harwood et al., 2025).

Accreditation cycles, for example, frequently absorb significant organizational energy. When quality improvement teams spend months preparing documentation for an external review, proactive safety work can stall. This is not a failing of any individual manager; it is a structural tension built into the regulatory environment. AcciMap makes this tension visible rather than invisible.

Regulatory frameworks also influence what gets investigated and how. The Joint Commission’s sentinel event reporting requirements have driven the adoption of root cause analysis, but the structure of those requirements can inadvertently reinforce single-cause thinking when teams feel pressure to identify a discrete correctable factor.

Organizational Culture and Leadership Decisions

Effective leadership is necessary in medicine to foster an organizational climate that promotes patient safety. The Joint Commission defines safety culture as the collection of beliefs, values, attitudes, perceptions, competencies, and patterns of behavior that determine an organization’s commitment to quality and patient safety (Karkera Kanjia, 2024).

Leaders who promote a positive organizational climate contribute to higher job satisfaction among employees, decreased burnout, fewer medical errors, and an overall improved culture of safety (Karkera Kanjia, 2024).

But organizational culture does not exist in isolation from the decisions leaders make day to day. When an organization systematically prioritizes throughput over reflection time, when incident reports go without timely feedback, when raising concerns is structurally discouraged even if never explicitly forbidden, a culture of risk accumulates quietly. AcciMap helps organizations trace the line between leadership decisions and frontline conditions, making visible what operational pressures have obscured.

Research examining the relationship between organizational culture and adverse events found that workgroup-level factors including civility, employee engagement, leadership, psychological safety, and resources all contribute meaningfully to safety outcomes, suggesting that culture is not just a backdrop to events but an active causal ingredient (George et al., 2020).

Technical and Operational Management

The middle layers of the system are where strategic decisions meet operational reality. This is the level of departmental managers, charge nurses, shift supervisors, and clinical leads. It is also where policy meets practice, and where misalignments between the two become embedded in routine work.

Every action in a system causes a reaction elsewhere in that system, and these reactions can lead to unintended consequences, sometimes long after the original action and not always attributed to them (Illing, 2017). Operational decisions about how shifts are structured, how handovers are conducted, or which technology systems are prioritized for maintenance create conditions that either support or undermine the work of frontline staff.

When an operational manager approves a workaround to cope with understaffing, they are making a reasonable response to a constrained situation. But that workaround, normalized over time, becomes the default practice. Months later, when a harm event occurs, investigators who focus only on the frontline encounter a workaround without understanding the organizational chain that created it.

Frontline Conditions and Clinician Work

Maximizing safety in modern healthcare requires a systems perspective in which safety is an emergent property of the system and healthcare workers and patients are parts of the same basic system (Sorensen et al., 2015).

Frontline clinicians work at what is often called the sharp end of care. They are the last line of defense before a patient is harmed, and because of this, investigations often stop with them. But sharp-end failures are almost always the expression of blunt-end conditions: resource constraints, unclear protocols, alert fatigue, poor communication infrastructure, and accumulated organizational decisions that created the environment in which the clinician was working.

Active errors occur at the sharp end of healthcare, where an action directly causes an error in real time. Latent errors occur at the blunt end and do not directly cause the error, but arise upstream of the event through factors such as building layout, organizational processes, human resources, and equipment failure (Gwinnutt & Gwinnutt, 2017).

This distinction is critical. A system that focuses exclusively on active errors at the sharp end will never address the latent conditions accumulating at the blunt end. AcciMap is designed precisely to make those latent conditions visible and traceable.

What AcciMap Reveals That Root Cause Analysis Misses

Consider a hypothetical but realistic scenario: a patient suffers harm from a delayed time-critical medication in an inpatient ward. A traditional RCA identifies that a nurse did not administer the medication within the prescribed window. Recommendation: re-education of nursing staff on medication timing protocols.

An AcciMap analysis of the same event might reveal the following across system levels:

• Government level: Persistent underfunding of hospital pharmacy services has reduced overnight pharmacist coverage, creating delays in medication preparation.
• Regulatory level: Reporting requirements have focused the quality team’s attention on documentation compliance rather than proactive safety work in recent months.
• Organizational level: A recent cost-cutting decision removed ward-based pharmacy technicians, shifting medication tracking responsibilities to nursing staff without updated training.
• Operational level: The electronic medication record system does not generate timely alerts for time-critical drugs on the overnight shift, a known issue raised in three previous incident reports.
• Frontline level: The nurse responsible was covering two additional patients due to an unplanned absence and had received no specific orientation to time-critical medication protocols under the new staffing model.

No single item on that list caused the event. Each contributed a condition that made harm more likely. A recommendation targeting only nurse education would address approximately one-fifth of the problem while leaving the other four-fifths intact, waiting to produce the next incident.

AcciMap analyses that graphically represent contributing factors across all levels of the system have been found to achieve high levels of validity for placing contributory factors at the correct healthcare system level. Critically, this accuracy depended on structured training in systems thinking, underscoring the need for education and support in using such methods effectively (Hulme et al., 2022).

The Challenge of Applying Systems Thinking in Practice

Embracing a systems approach is not without its difficulties. There are real and legitimate reasons why root cause analysis, for all its limitations, remains the dominant method in healthcare investigation.

First, systems thinking takes time and expertise. Mapping the contributing factors across six levels of a complex sociotechnical system is not a task that can be completed in an afternoon. It requires skilled facilitators, diverse perspectives, and access to information that spans organizational and regulatory levels. Many healthcare organizations are resource-constrained and conduct investigations under significant time pressure.

The need for further work to support the use of state-of-the-art safety science methods in practice is well recognized. This includes the provision of education in systems thinking, methods training programs, development of further support materials, and modification of methods to improve their usability (Hulme et al., 2022).

Second, there is an institutional momentum behind existing methods. Root cause analysis has decades of regulatory endorsement, established curricula, and familiar terminology. Shifting an organization’s investigation culture toward systems thinking requires sustained leadership commitment and investment, neither of which is guaranteed when budgets are tight.

Third, accountability is genuinely important. A systems perspective should not become a mechanism for avoiding accountability when it is warranted. The goal is not to excuse harmful practice but to ensure that accountability is placed at the correct level of the system, and that improvement efforts address the actual contributors to harm rather than the most visible ones.

Embracing systems thinking is almost synonymous with embracing the complexity of healthcare and appreciating that incidents may not follow a linear causation process (Walsh et al., 2023).

What This Means for Healthcare Workers

For frontline clinicians, charge nurses, department managers, and quality leaders, the shift to systems thinking carries practical implications.

Incident Reporting Becomes More Valuable

When staff understand that incidents are emergent outcomes of interconnected conditions rather than individual failures, the value of near-miss reporting increases. Every report is a data point about system conditions, not just an account of a personal mistake. Organizations that build genuine psychological safety around reporting create the information flows needed to identify systemic risk before events occur.

Investigation Design Changes

Teams that use AcciMap or similar methods begin their analysis not with “what did the frontline worker do?” but with “what conditions across all system levels made this outcome possible?” This requires inviting voices from multiple levels of the organization into the investigation, including managers, administrators, and in some cases external stakeholders.

Recommendations Become More Effective

Research has consistently found that investigations focused primarily on frontline contributors generate weak prevention strategies. When recommendations address higher system levels, including organizational processes, resource allocation, and leadership decisions, they are more likely to produce durable change (Bowditch et al., 2025).

The Language We Use Matters

Moving away from terms like “root cause” and “human error” toward language like “contributing factors,” “system conditions,” and “emergent outcomes” is more than semantic hygiene. It shapes how teams think, what they look for, and what they recommend. Language carries assumptions, and in the investigation of harm, those assumptions have consequences.

Conclusion: Rethinking What We Are Looking For

The search for a root cause is understandable. It offers resolution, a clear target, and the reassuring sense that a problem has been contained. But in complex adaptive systems like healthcare organizations, this resolution is often an illusion. The next event is already forming in the gaps between system levels, in the decisions that were never connected, in the pressures that were never surfaced.

AcciMap and the broader systems thinking framework do not offer the comfort of a single answer. What they offer instead is a more accurate picture of how events actually occur, and therefore a more realistic foundation for preventing them.

When healthcare organizations commit to this level of analysis, they make a different kind of investment: not in the tidy fiction of root causes, but in the harder, more honest work of understanding how their systems function as a whole. For healthcare workers who have long known that harm rarely has a simple origin, that investment is long overdue.

References

Bowditch, L., Molloy, C., King, B., Abedi, M., Jackson, S., Bierbaum, M., Yu, Y., Raggett, L., Salmon, P., Braithwaite, J., Westbrook, J. I., Clay-Williams, R., Lingam, R., Middleton, S., Magrabi, F., Mumford, V., & Hibbert, P. (2025). Do patient safety incident investigations align with systems thinking? An analysis of contributing factors and recommendations. BMJ Quality & Safety. https://doi.org/10.1136/bmjqs-2025-019063

Braithwaite, J., Wears, R. L., & Hollnagel, E. (2015). False dawns and new horizons in patient safety research and practice. International Journal for Quality in Health Care, 27(5), 371–374. https://doi.org/10.1093/intqhc/mzv069

Eisenberg, J. (2000, February 22). Statement on medical errors [Testimony before the Senate Committee on Health, Education, Labor and Pensions]. Agency for Healthcare Research and Quality.

George, J., Elwy, A. R., Charns, M. P., Maguire, E. M., Baker, E., & Burgess, J. F. (2020). Exploring the association between organizational culture and large-scale adverse events: Evidence from the Veterans Health Administration. Joint Commission Journal on Quality and Patient Safety, 46(6), 322–332. https://doi.org/10.1016/j.jcjq.2020.02.003

Gwinnutt, M., & Gwinnutt, C. (2017). Error and root cause analysis. BJA Education, 17(9), 323–333. https://doi.org/10.1093/bjaed/mkx010

Harwood, R., Doyle, C., & Morrice, A. (2025). Adverse event reporting and patient safety: The role of a just culture. Frontiers in Health Services, 5, 1581516. https://doi.org/10.3389/frhs.2025.1581516

Hulme, A., Wynne, R., Salmon, P. M., Walker, G. H., & Waterson, P. (2022). Toward the translation of systems thinking methods in patient safety practice: Assessing the validity of Net-HARMS and AcciMap. Safety Science, 156, 105897. https://doi.org/10.1016/j.ssci.2022.105897

Illing, S. (2017). Using systems thinking in patient safety: A case study on medicines management. Nursing Management, 24(5), 35–40. https://doi.org/10.7748/nm.2017.e1563

Karkera Kanjia, M. (2024). Effective leadership and patient safety culture. Anesthesia Patient Safety Foundation Newsletter. https://www.apsf.org/article/effective-leadership-and-patient-safety-culture/

Kohn, L. T., Corrigan, J. M., & Donaldson, M. S. (Eds.). (2000). To err is human: Building a safer health system. National Academies Press.

Rasmussen, J. (1997). Risk management in a dynamic society: A modelling problem. Safety Science, 27(2–3), 183–213. https://doi.org/10.1016/S0925-7535(97)00052-0

Salmon, P. M., Cornelissen, M., & Trotter, M. J. (2012). Systems-based accident analysis methods: A comparison of AcciMap, HFACS, and STAMP. Safety Science, 50(4), 1158–1170. https://doi.org/10.1016/j.ssci.2011.11.009

Sorensen, G., Sparer, E., Williams, J. A., Gundersen, D., Boden, L. I., Dennerlein, J. T., Hashimoto, D., Hopcia, K., Katz, J. N., Kenwood, C., Stoddard, A. M., Sugimoto, H., Tveito, T. H., Tworney-Guichard, J., & Wagner, G. R. (2015). Frontline hospital workers and the worker safety/patient safety nexus. Journal of Occupational and Environmental Medicine, 57(9), 953–958. https://doi.org/10.1097/JOM.0000000000000513

Svedung, I., & Rasmussen, J. (2002). Graphic representation of accident scenarios: Mapping system structure and the causation of accidents. Safety Science, 40(5), 397–417. https://doi.org/10.1016/S0925-7535(00)00036-9

Walsh, C., Lydon, S., & O’Connor, P. (2023). Thematic reviews of patient safety incidents as a tool for systems thinking: A quality improvement report. BMJ Open Quality, 12(2), e002073. https://doi.org/10.1136/bmjoq-2022-002073

Waterson, P., Jenkins, D. P., Salmon, P. M., & Underwood, P. (2016). “Remixing Rasmussen”: The evolution of AcciMaps within systemic accident analysis. Applied Ergonomics, 59, 483–503. https://doi.org/10.1016/j.apergo.2016.09.004