From Hospital Building to Hospital System: How Structure, Organization, and Process Management Are Likely to Evolve in a Hybrid Era

Introduction

A terminology of "monoliths" for "High-Complexity Hubs" and "swarms" for "Distributed Care Network" is evocative rather than standard in health-services research, but it captures a real and increasingly well-documented transformation. The academic and policy literature more commonly describes this transition as a move toward centralization of high-complexity services alongside decentralization of appropriate care through integrated, people-centred networks, telemedicine, hospital-at-home models, virtual wards, and hub-and-spoke systems. In that sense, the future hospital is not disappearing; it is being reconstituted as a networked institution: centralized where rarity, risk, and capital intensity require concentration, and decentralized where continuity, accessibility, and lower-acuity care favor delivery closer to the patient.

In this essay I want to argue that hospitals will evolve along three interlocking dimensions. Structurally, the traditional hospital will persist, but increasingly as a (smaller) high-complexity regional hub embedded in a wider care network. Organizationally, hospitals will shift from self-contained institutions to coordinators of distributed service platforms with shared protocols, digital governance, and hybrid workforce models. In process-management terms, operations will move away from managing beds and departments within a single building toward managing patient pathways, flow, escalation, and quality across a geographically dispersed continuum of care.

The most defensible conclusion from current evidence is therefore not a choice between central “monolith” and decentralized “swarm,” but a hybrid architecture in which each compensates for the limitations of the other. The main weaknesses of centralized high-volume or highly specialized hospital care are access and equity problems, operational rigidity, network dependency, and the risk that very large centralized flows create coordination frictions of their own. The main weaknesses of distributed networks are usually governance and coordination failures.

1. Structural Evolution: From Standalone Hospital to Stratified Care Architecture

The large general hospital is unlikely to vanish completely because some services still benefit from concentration. Evidence on centralization shows that specialized, high-volume units can improve outcomes and operational efficiency by concentrating expertise, protocols, and capital-intensive infrastructure. Critical care literature makes the logic explicit: centralization supports specialist staffing, training, standardization, and lower variability in care, while examples such as centralized acute stroke pathways show that concentrating some time-critical services can reduce mortality and shorten acute stays. For these reasons, future hospitals will likely retain or deepen their role as high-complexity hubs for services such as intensive care, major surgery, advanced diagnostics, and other low-frequency, high-acuity interventions.

However, the structural future of these hubs is not simply “bigger hospitals.” Current WHO guidance emphasizes that future health infrastructure should be planned according to functional role and linkages across the system, not by bed counts alone. It also argues for facilities that are agile, climate-resilient, and flexibly designed, because rigid (monolithic) buildings quickly become bottlenecks when care models, technologies, and epidemiological pressures change. In practical terms, this means hubs are likely to be built or retrofitted as adaptable modular platforms (pavilions) in itself with reconfigurable clinical space, scalable utilities, and layouts designed around patient and process flow, infection control, and rapid operational repurposing rather than static departmental silos. 

This is where the “monolith” metaphor becomes misleading. The hub of the future will remain physically substantial, but it will be less territorially self-sufficient and more explicitly designed as a node in a regional care system. Structural centralization also has recognized costs: longer travel times, burdens on families, and potential weakening of local capabilities if too much expertise is withdrawn from peripheral sites. The literature on health-service centralization repeatedly warns that efficiency gains must be weighed against access losses. Therefore, the structural evolution of the hospital will not be toward unlimited concentration, but toward selective centralization for services where the volume-outcome and infrastructure arguments are strongest.

In parallel, the “swarm” side of the system is expanding the places in which hospital-level functions can occur. Hospital-at-home and virtual ward evidence suggests that, for selected patient groups and carefully designed programs, inpatient-level care can be delivered outside the conventional ward without increasing readmission risk and, in some studies, with equal or better patient experience. This does not abolish the hub; rather, it changes the hub’s structural role from being the universal site of care to being the backbone and escalation center of a distributed architecture. The building remains important, but it becomes one component of a larger service topology that includes homes, ambulatory sites, remote monitoring platforms, and local clinical nodes.

2. Organizational Evolution: From Institution-Centred Bureaucracy to Network Governance

If structure answers where care happens, organization answers who governs it. Here the decisive shift is from the hospital as a bounded institution to the hospital as a network orchestrator. WHO’s people-centred care framework argues that health systems should deliver care “at the right time, in the right place, in the right way,” with services coordinated across the continuum rather than fragmented by institution or specialty. Organizationally, that implies governance arrangements that extend beyond the hospital campus: shared clinical standards across sites, integrated information systems, formal referral and escalation pathways, and management structures accountable for network performance rather than only facility performance.

The most plausible organizational template for this is a hub-and-spoke model. In critical care, telemedicine already enables specialist expertise to be extended from a central hub into local environments, allowing smaller sites to benefit from high-level support without replicating all the staffing and infrastructure of a tertiary center. A recent systematic review of Tele-ICU models found that hub-and-spoke and hybrid arrangements showed the most consistent benefits for mortality reduction, protocol adherence, communication, and specialist reach, although evidence on long-term sustainability and cost-effectiveness remains limited. This matters because it demonstrates that organizational centralization need not require all clinical activity to be physically centralized. Expertise can be centralized while delivery is selectively distributed.

The same principle is visible in hospital-at-home models that use a centralized command center (cockpit) to supervise geographically dispersed in-home care. In Mayo Clinic’s hybrid model across two regions, a single virtual command center coordinated remote clinicians, local in-home visits, and logistics over a broad catchment area. That is organizationally significant: it shows how one hospital can become less a place of admission and more a platform for dispatch, surveillance, escalation, and coordination. This kind of model redefines the hospital’s organizational core around command, standards, and specialist backup rather than around the exclusive physical co-location of patients and staff.

Such reorganization also changes workforce design. Distributed networks require more hybrid roles: remote monitoring teams, virtual hospitalists, community paramedics, hospital-at-home nurses, escalation coordinators, and clinicians who work across digital and physical interfaces. Yet the literature does not support the idea that technology simply substitutes for labor. On the contrary, both tele-ICU and home-hospital studies suggest that successful distributed models depend on careful workforce integration, explicit protocols, and strong human coordination. The organizational future is therefore not “AI replacing hospitals,” but hospitals redesigning labor so scarce expertise can supervise a wider geography while bedside and in-home care is delivered through mixed local teams.

This organizational evolution also introduces new governance burdens. OECD and WHO sources both stress that telemedicine and digital care require better measurement, clearer financing arrangements, interoperability, quality oversight, and attention to equity. Fragmentation between in-person and remote services is a recognized risk. Therefore, the distributed network cannot simply be a loose collection of apps, clinics, and home services; it must be governed as a single system with common safety rules, digital standards, and accountability for continuity of care.

3. Process Management Evolution: From Bed Management to Pathway Orchestration

The deepest managerial change may occur at the level of process. Traditional hospital operations are often organized around departments, wards, and bed occupancy. Emerging evidence instead points toward pathway-based and flow-based management, in which the relevant unit is no longer the ward but the patient journey across settings. A qualitative study of leading hospitals identified hospital-wide flow improvement as dependent on organizational alignment, coordination and transfer structures, digital and analytical tools, standards and routines, occupancy management, and external coordination beyond the hospital itself. This is precisely the type of process architecture required for a hybrid hub-and-network system.

Digital coordination centres and hospital command centers exemplify this shift. They are designed to integrate real-time data on demand, bed availability, transfers, and patient flow to support operational decisions across the institution and, potentially, beyond it. The current evidence base is still relatively early and heterogeneous, so their effectiveness should not be overstated; however, studies consistently describe them as instruments for improving situational awareness, reducing flow friction, and coordinating scarce capacity. In the hybrid future, such functions are likely to expand from managing internal throughput to managing network navigation: deciding which patients should remain at home, attend a local site, be transferred to a spoke facility, or escalate to the high-complexity hub.

This implies that hospital process management will become increasingly triage-intensive and escalation-sensitive. Instead of treating admission to the main hospital as the default, systems will need robust front-door sorting mechanisms that allocate patients to the most appropriate care setting from the outset. In distributed acute care models, that sorting is supported by telemedicine, remote vital-sign monitoring, and structured criteria for escalation back to hospital. Reviews of inpatient-level care at home suggest that many such models can achieve outcomes comparable to inpatient care for selected populations, but the keyword is selected: these are not universal substitutes for hospital admission. Process redesign therefore depends on increasingly precise patient selection, surveillance, and rescue pathways (e.g. escalation pathway, clinical deterioration pathway, transfer protocol, and rapid escalation process).

Artificial intelligence (AI) will likely enter this process layer first as an augmentative operational tool. OECD analysis identifies hospital operations, administrative automation, and resource allocation as major areas in which AI may create value. The most credible near-term applications are forecasting surges, predicting admissions or deterioration, optimizing scheduling, and highlighting discharge or transfer risks. That said, the present evidence base supports AI as decision support rather than autonomous process control. Human oversight (human-in-the-loop, HITL), data governance, and safety assurance remain necessary, especially when digital tools are used to govern admission, discharge, and remote monitoring decisions.

A further process-management implication concerns finance and incentives. Distributed care models are often clinically promising but operationally fragile when payment systems still reward facility-based volume. OECD work on telemedicine highlights persistent weaknesses in financing and integration, while recent economic analysis of an all-virtual acute-care model found that savings may accrue differently to hospitals and payers depending on reimbursement design. In other words, hospitals may be organizationally capable of becoming distributed networks before health financing systems are ready to support them. Process management in the future hospital will therefore include not only logistics and quality control, but also the ability to align reimbursement, contracting, and performance measurement with cross-setting care pathways.

Note: Human-in-the-loop (HITL) is an artificial intelligence (AI) design pattern that embeds human oversight, feedback, and expertise directly into the machine learning (ML) lifecycle. Instead of fully delegating tasks to autonomous systems, HITL ensures that humans are involved in training, validation, or real-time operations to enhance precision, safety, and accountability.

4. Limits, Risks, and Conditions of Success

A balanced answer must also state where the evidence is strong and where it remains limited. The evidence for selective centralization is strongest in particular service lines and circumstances, not as a blanket rule for all hospital care. Likewise, the evidence for distributed acute care is strongest for selected diagnoses, carefully designed hospital-at-home programs, and systems with strong escalation capacity. Command-center evidence is still developing, and cost evidence on centralization remains methodologically uneven. The hybrid model is therefore better understood as a direction of travel than as a finished blueprint.

The principal risks are also clear. Excessive centralization can worsen geographic inequity and weaken local services. Excessive decentralization can fragment care, overload patients and families, and create safety risks if remote pathways are poorly governed. Digital dependency introduces risks around interoperability, data quality, privacy, exclusion, and uneven digital literacy. For that reason, WHO guidance repeatedly frames telehealth and integrated care not as merely technical innovations but as governance problems involving quality assurance, patient safety, and equitable access.

Conclusion

Hospitals are not moving toward extinction, nor toward a simple future of ever-larger centralized campuses. They are evolving into clinically stratified, digitally mediated care systems. The high-complexity hospital will endure as a regional hub for rare, risky, and capital-intensive care; at the same time, a growing share of routine, chronic, follow-up, and selected acute care will be delivered through distributed networks of home-based, ambulatory, virtual, and community services. This hybrid future will reshape hospital structure into hub-and-network architectures, organization into cross-site governance platforms, and process management into real-time orchestration of pathways across settings rather than throughput within one building.

The most accurate conclusion, then, is not that the hospital becomes either a “monolith” or a “swarm.” It becomes both: a centralized locus of high-complexity capability and a decentralized service ecosystem designed around people rather than premises. The future hospital is, in effect, less a building than a managed network of care, with the building retained for what only the building can do. 

Bibliography

Åhlin, P., Almström, P., & Wänström, C. (2023). Solutions for improved hospital-wide patient flows: a qualitative interview study of leading healthcare providers. BMC Health Services Research, 23(1), 1–17.

Bhattarai, N., McMeekin, P., Price, C., & Vale, L. (2016). Economic evaluations on centralisation of specialised healthcare services: a systematic review of methods. BMJ Open, 6(5), e011214.

Collado, E., Luiso, D., Ariza-Sole, A., Lorente, V., Sánchez-Salado, J. C., Moreno, R., ... & Comin-Colet, J. (2021). Hospitalization-related economic impact of patients with cardiogenic shock in a high-complexity reference centre. European Heart Journal. Acute Cardiovascular Care, 10(1), 50-53.

Elrod, J. K., & Fortenberry Jr, J. L. (2017). The hub-and-spoke organization design: an avenue for serving patients well. BMC health services research, 17(Suppl 1), 457.

Franklin, B. J., Gandara, E., Bustamante, J., et al. (2022). Use of Hospital Capacity Command Centers to Improve Patient Flow and Safety: A Scoping Review. Journal of Patient Safety, 18(6), e912–e921.

Gaillard, G., & Russinoff, I. (2023). Hospital at home: A change in the course of care. Journal of the American Association of Nurse Practitioners, 35(3), 179-182.

Leong, M. Q., Lim, C. W., & Lai, Y. F. (2021). Comparison of Hospital-at-Home models: a systematic review of reviews. BMJ Open, 11(1), e043285.

Levine, D. M., Desai, M. P., Findeisen, S. M., et al. (2025). Hospital-Level Care at Home for Adults Living in Rural Settings: A Randomized Clinical Trial. JAMA Network Open, 8(12), e2545712.

Morris, S., Hunter, R. M., Ramsay, A. I. G., et al. (2019). Impact and sustainability of centralising acute stroke services in English metropolitan areas: retrospective analysis of hospital episode statistics and stroke national audit data. BMJ, 364, l1.

OECD. (2025). Leading practices for the future of telemedicine: Implementing telemedicine post-pandemic (OECD Health Working Papers No. 173). OECD Publishing.

OECD. (2026). AI in healthcare. In Progress in Implementing the European Union Coordinated Plan on Artificial Intelligence, Volume 2. OECD Publishing.

Ostermann, M., & Vincent, J.-L. (2019). How much centralization of critical care services in the era of telemedicine? Critical Care, 23(1), 423.

Paulson, M. R., Shulman, E. P., Dunn, A. N., et al. (2023). Implementation of a virtual and in-person hybrid hospital-at-home model in two geographically separate regions utilizing a single command center: a descriptive cohort study. BMC Health Services Research, 23(1), 139.

Qtait, M., et al. (2026). Implementation and Impact of Tele-Intensive Care Unit (Tele-ICU) Models on Critical Care Outcomes: A Systematic Review. Nursing in Critical Care, 31(2), e70401.

Shaefi, S., O'Gara, B., Kociol, R. D., Joynt, K., Mueller, A., Nizamuddin, J., ... & Shahul, S. (2015). Effect of cardiogenic shock hospital volume on mortality in patients with cardiogenic shock. Journal of the American Heart Association, 4(1), e001462.

Shi, C., Dumville, J., Rubinstein, F., et al. (2024). Inpatient-level care at home delivered by virtual wards and hospital at home: a systematic review and meta-analysis of complex interventions and their components. BMC Medicine, 22, 145.

Spellberg, B., Lynch, C., Yee, H. F., & Banerjee, J. (2025). Health Economic Analysis of an All-Virtual, At-Home Acute Care Model. JAMA Network Open, 8(6), e2517114.

World Health Organization. (2016). Framework on integrated, people-centred health services. World Health Organization.

World Health Organization. (2023). Strategic health infrastructure investments to support universal health coverage. World Health Organization.

World Health Organization Regional Office for Europe. (2024). Telehealth quality of care tool. WHO/Europe.

World Health Organization Regional Office for Europe. (2024). Support tool to strengthen telemedicine: Resource for assessment, strategy development, and strengthening of telemedicine services. WHO/Europe.

Van Osta, Peter. An essay concerning a new healthcare.