A Systems Thinking Approach to Future Hospital Strategy: Insights from Singapore

Introduction: Context and the Need for a New Paradigm

The future design of acute care hospitals, particularly in rapidly aging city-states like Singapore, necessitates a sophisticated approach that moves beyond traditional planning methods. This complexity arises from several interacting factors, including the increasing burden of chronic disease, global efforts toward healthy longevity, and the persistent threat of crises such as pandemics. Building a new acute care hospital represents a significant investment in human, information, and physical assets, providing an opportunity to radically realign the supply of care with changing demands through a “future-back” perspective.

Globally, there is a recognized call to rethink hospital architecture, both internally and externally, emphasizing the integration of the hospital with the community. A central concept in this vision is the idea of “hospitals without walls,” where care is delivered in a patient-centric manner, integrating hospital services better with primary and community care. While the literature emphasizes the inherent complexity of Future Hospital design, published studies often concentrate on physical architecture, digitalization, and overall aims.

The authors of this study, Kumar et al. (2025), specifically address this gap by applying systems thinking to visualize and document the present and future dynamics of a hospital system, focusing on a new-generation public sector general hospital slated to be operational in Singapore by 2037. Singapore faces unique challenges, including a rapidly aging population and hospital overcrowding, which is symptomatic of a health system under increasing stress. The planning for this new hospital must optimize across multiple conflicting objectives, including the quadruple aim (population health, cost, patient experience, and provider satisfaction). Furthermore, the COVID-19 pandemic highlighted the critical need for agility—the ability to quickly ramp up and redeploy capacity in response to future crises.

This research aims to answer two fundamental questions regarding the Future Hospital system:

1. What are the variables that drive the present and future dynamics of the Future Hospital system?
2. How are these driving variables related?

Methodology: Systems Thinking and Collaborative Modeling

To unravel these complexities, the authors developed a series of visual logic maps in the form of Causal Loop Diagrams (CLDs), a proven systems thinking tool used to visualize and unpack complex health system behavior. CLDs capture how system variables interrelate and how external variables impact them.

The methodology involved a rigorous, collaborative group modeling process led by a cross-functional system modeling team, comprising key domain and opinion leaders from clinical, allied health, and support service domains, totaling 20 sessions (8 modeling discussions and 12 reviews) between September 2023 and February 2024. It is noted that patients were a limitation in this study as they were not involved at any stage.

Recognizing that overly dense CLDs can be difficult to interpret, the study system was decomposed into seven distinct, linked views or sectors. These seven interlocking CLDs serve as “boundary objects” to facilitate knowledge sharing across different Future Hospital stakeholders, including clinicians, architects, and operations planners. The seven views are:

• National View: Macro-level variables affecting the Study Hospital’s (SH) utilization.
• Public Healthcare (PH) Cluster View: Focusing on the organizational entity responsible for providing care through a network of institutions.
• SH Top-level View: Variables cutting across the four hospital subsystems.
• SH Outpatient, SH Emergency Department (ED), SH Inpatient, and SH Surgery & Procedures (S&P) Views: Functional breakdowns of key hospital services.

After developing the CLDs, the researchers applied network analysis to the combined graph of the seven views. Network analysis was used to generate additional insights on potential high-leverage points—the variables with the greatest criticality or mediating function within the network. The specific measure evaluated was betweenness centrality, which identifies nodes located on the shortest paths between other nodes.

Key Results: Dynamics, Interventions, and Agility

The modeling successfully generated seven comprehensive CLDs, identifying important feedback loops and 15 interventions for redesign.

Feedback Loops and System Dynamics

The analysis of the modeled services revealed that capacities are subject to positive feedback loops, reinforcing the need for alternative models of care to restrain ever-increasing demand.

For instance, the detailed SH Inpatient View (CLD 3.3), which uses inpatient bed capacity as a proxy for hospital size, illustrates the traits of a complex system with tight coupling, interactions, and dynamic behaviors over time. A critical finding in this view is the reinforcing loop, R3.3a, which reflects the core problem known as Roemer’s Law—the phenomenon that building out more beds or hiring more staff creates a vicious cycle where a bed built is a bed filled [26, 51–53].

Conversely, the balancing loop, B3.3a, shows the potential of Hospital at Home (HaH) initiatives to rein in the runaway demand for physical inpatient beds.

Selected Interventions

The group modeling exercise identified key interventions categorized by whether they reduce patient inflow or increase throughput/reduce length of stay (LOS). These interventions are crucial for departing from the “business-as-usual” operational mode. Examples across the subsystems include:

• SH Top-level: Regionalization.
• Outpatient: Right-siting to teleconsultation or primary care, and operations efficiency advances to reduce waiting time.
• Emergency Department (ED): Right-siting to telehealth nurse or primary care, and preventive care in the ED.
• Inpatient: Right-siting of patients (e.g., to step-down care, ILTC, or HaH) and advances in lab diagnostics.
• Surgeries & Procedures (S&P): Pooling of resources, advancements in scheduling policies, and clinical advances allowing for more day surgeries.

Network Analysis: The Paramount Role of Agility

The network analysis, integrating all 157 variables from the seven CLD views, provided computational insight into the system’s structure. The result highlighted a crucial variable: Need for agility [3.0.15].

Need for agility (defined as the ability to quickly ramp up and redeploy capacity in response to crises) was found to be the variable with the highest betweenness centrality (0.377) in the combined network. This metric confirms its greatest brokerage and criticality within the system, demonstrating that the need for agility is paramount in Future Hospital plans.

Other variables with high betweenness centrality included fundamental drivers such as:

• Need for ED services [3.0.9].
• Need for Specialist Outpatient Centre (SOC) services [3.0.6].
• Resident population health service needs volume [1.10 s].
• Physical inpatient LOS [3.3.5].
• Proportion of elderly in resident population [1.7].

The diversity of these critical nodes—ranging from demand (service needs) and supply (capacities) to exogenous factors (crises) and strategic choices (design for flexibility)—illustrates the tightly coupled nature of the health system.

Conclusion: Looking Forward

This study addresses the scarcity of systems thinking insight in Future Hospital planning, which has traditionally been dominated by technical architecture perspectives. By making the resulting logic maps public, the researchers promote greater openness on Future Hospital initiatives.

The CLDs, despite their qualitative nature, are highly insightful for Future Hospital strategic planning exercises globally. The long planning horizon (until the year 2100), coupled with changing population structure and the threat of future crises, means accumulated statistical evidence alone is inadequate to guide policy. In this context, the soft systems (qualitative) CLD analysis is effective in documenting a shared understanding and deep domain knowledge, forming a living planning document.

The key takeaway is that the tight coupling and feedback loops inherent in the hospital system mean that actions taken in one area will have rippling effects, confirming that it is “not possible to do just one thing”. The identification of the Need for Agility as the network’s most central variable underscores the strategic importance of designing for flexibility and resource pooling to ensure the Future Hospital can withstand uncertainty and unforeseen crises.

The design of a Future Hospital is a rare chance to reimagine healthcare delivery while remaining anchored in present realities, and complex systems approaches, as demonstrated here, are perfectly suited to mapping this complex logic.

Analogy to solidify understanding: Thinking about future hospital design using systems thinking is like optimizing a highly integrated city rather than just designing a single building. If you only build more roads (increase bed capacity), you might temporarily ease congestion, but the core issue—the continuous positive feedback loop of people wanting to use the most convenient path (Roemer’s Law)—will quickly bring back gridlock. By using CLDs, the planners are essentially creating a detailed, dynamic map of the city’s traffic patterns, showing that the most critical intersections (high betweenness centrality nodes, like Need for Agility) must be managed with flexible, adaptable solutions (like dynamic lanes or shifting traffic away via right-siting) rather than just building an ever-larger highway.

APA Reference: Kumar, A., Lam, S. S. W., Chan, S. L., Xu, Y., Ge, Y., Gui, G. K. T., & Tan, H. K. (2025). Strategizing towards the future hospital: a systems thinking approach. Health Research Policy and Systems, 23(71). https://doi.org/10.1186/s12961-025-01333-9