Stroke Systems of Care 2.0: Moving towards Definability, Accountability and Equity

The Stroke Systems of Care (SSoC) model was introduced in the US in 2005, when a task force of the American Heart/American Stroke Association (AHA/ASA) was convened to make recommendations for the organization and operation of systems of care for stroke patients.¹ The ideal SSoC was defined as a coordinated system, marked by communication and collaboration, that promoted patient access to essential components of stroke care. Updates focusing on improving interactions within SSoC, care delivery across the continuum, and disparities followed.^2,3 Most recently consensus recommendations addressing prehospital destination strategies in SSoC have been published.⁴

Since 2005 we have witnessed tremendous successes in the advancement of SSoC, including the growth of telestroke and rapid reorganization of care systems in response to evolving evidence on reperfusion therapies. While AHA/ASA recommendations have served as critically important guides, the advancement of SSoC has occurred despite little direct funding or top-down directives. Moreover, measuring progress is challenged by the lack of an organizational and methodological framework to define and evaluate SSoC. As we continue to build upon this foundation toward developing a second generation SSoC, we need new approaches for measuring progress, ensuring that the benefits of SSoC are realized equitably across all communities.

How Far We Have Come

Although we lack a comprehensive way to document the growth and impact of SSoC, several key components have seen clear advances over the last 2 decades.

The growth of telestroke has helped improve access and reduce fragmentation of care leading to improved population access to acute neurological expertise.⁵ For patients, telestroke has improved access to vascular neurologists, transfer processes, and post-stroke outcomes.^6,7 For some hospitals, ensuring 24/7 specialist access telestroke has been vital for maintaining stroke center certification.

Over the last 10–15 years we have seen notable changes in transfer patterns in response to clinical guidelines on reperfusion.⁸ Transfers represent a key strategy to ensure access to revascularization therapy regardless of where patients live or first present. Early iterations of SSoC centered on the drip-and-ship paradigm enabled systems to rapidly adapt and reorganize to increase access to endovascular thrombectomy (EVT).

Evolving evidence for EVT has also led to changes in the prehospital approach to stroke patients, with ongoing research examining alternative pre-hospital stroke scales, new field-based technologies to identify large vessel occlusion, optimal triage and destination (bypass) strategies, personalized triage decision tools, and novel paradigms for EVT delivery (e.g., the flying interventional team).⁹

Although substantial strides have been made in improving access to EVT many holes remain, particularly for more geographically remote and vulnerable populations.

The Measurement Problem

While prior AHA/ASA SSoC statements provided justification for the initial development of SSoC,¹ and detailed recommendations to enhance evolving systems,^2,3 none proposed methods to objectively identify a given SSoC, or to monitor its functions and outcomes. Of course, this is understandable given that SSoC are highly heterogeneous, complex structures that involve multiple organizations with different interacting functions. However, to understand what populations are served by a given SSoC, determine what functions are working well, and identify what components need improvement, we need a methodological framework to define a SSoC and measure its operational functions and outcomes.

Measuring an interconnected, non-linear and dynamic network is not straightforward, however. A given SSoC may be defined by its geographic boundary (e.g., city, region, or state), administrative organization (e.g., hospital system, or centralized governance), functional organization (e.g., inter-connected hospital network), or the populations served (e.g., specific communities, underserved/minority populations). The overlapping and multi-level nature of SSoC means that these organizational structures are not mutually exclusive; a SSoC could be defined using all these dimensions. Depending on the context some organizational structures might be prioritized. For example, if the goal is to assign accountability, then the administrative organization -- whether defined by a statewide registry or a single integrated hospital system -- might be the most important.

SSoC are not the only complex systems that are difficult to define. Complex networks exist in health care, in the social, physical, and computer sciences, and society at large. There has been growing interest in using network science, particularly community detection methods to define network structure and functions. Here, network science may define SSoC functionally, for example through inter-hospital transfer of stroke patients.¹⁰ Network science methods have distinct advantages including the ability to identify the number and strength of relationships between multiple hospitals, such that a fully formed picture of the organization and function of the system can be defined (Figure).¹⁰

Figure.

Illustration of Different Stroke System of Care Identified by Community Detection Methods*

* Footnote: Community detection methods, a tool form network science, can be used to identify different clusters of hospitals that are grouped (connected) by common organizational functions (e.g., patient transfers). Each hospital network can define a different stroke system of care.

Where SSoC Need to Go

As the AHA celebrates its 100^th anniversary, our understanding of factors that influence population health and patient outcomes continues to evolve. This includes the growing importance of social determinants of health (SDOH), ensuring equity in care delivery and outcomes for all patients, and understanding the effects of major shifts in the organization of US healthcare markets. These changes impact that environment in which SSoC exist, how we understand and define SSoC, and the metrics we use to evaluate them.

Adverse SDOH and inequities in the organization of SSoC directly impact patient outcomes. For example, patients’ access to EVT depends in large part on where they live, present for care, and whether they are transferred. Quality improvement program such as Get-With-The-Guidelines(GWTG)-Stroke have made important strides in improving care delivery and outcomes,¹¹ however hospitals that participate in GWTG-Stroke are disproportionately larger, academic, urban centers, and represent fewer rural patients. Reducing rural health disparities requires innovative approaches in these less-resourced sites, including efforts such as the AHA Rural Health Outcomes Accelerator program. Here again, approaches from network science may be useful. In California, for example, a network analysis identified that insurance-based disparities in access to stroke center care were explained by one particular cluster of hospitals and its transfer patterns.¹²

Rapid consolidation of the health care system, brought about by widespread hospital mergers and acquisitions, hospital closures, and new employment models for physicians (e.g., large health systems, corporations, private equity-backed staffing companies), has resulted in dramatic changes to where care is available.¹³ As yet, consolidation has been associated with few measurable benefits to patients, but with increased overall costs.¹⁴ Currently, we have little understanding of how changes to the health care market are impacting stroke care and the organization of SSoC in particular. As all SSoC function in this evolving health care market, we must monitor the impact of these changes on access, care delivery, and patient outcomes.

The concept of Learning Health Systems (LHS) may provide a foundation for monitoring and investing in SSoC. A LHS incorporates evidence, internal experience, and data analytics to improve care. While LHS strategies in stroke have been under-described to date,¹⁵ the concept is ripe for application to SSoC. Once we have an operational definition for identifying a particular SSoC, building out a stroke LHS would require comprehensive stroke centers (CSC) and other well-resourced hospitals (e.g., thrombectomy-capable) to expand their conceptualization of accountability, and to define the communities and populations the SSoC serves. Just as the accountable care organization model holds hospital systems accountable for population health, a defined SSoC should evolve to encompass a ‘learning stroke system of care with accountability.’ As we identify populations and communities served by the SSoC, we will be positioned to better evaluate them, and to consider system-level quality metrics including those addressing SDOH and health equity. Yet such changes will not happen on their own. A true systems-based approach must be implemented broadly, in a way that will require explicit investment and funding. Future funding of stroke care delivery – perhaps the “SSoC Act of 2026” – must consider how SSoC are defined, measured, and how data can be used to identify and address gaps in accountability and equity.

Conclusion

Stroke Systems of Care 2.0 will continue to evolve as we identify and apply systematic approaches to definitions and measurement. We need better measures of successes and failures with standardized approaches to capturing the performance of SSoC – a key opportunity for a new scientific statement from the AHA. The innovations that have advanced SSoC over the last 15 years have been tremendous. As SSoC evolve, we must drive toward improved measures of organizational structure and increase their accountability to the communities they serve.

Abbreviations

AHA/ASA

American Heart/American Stroke Association

CSC

comprehensive stroke center

EVT

endovascular thrombectomy

LHS

learning health system

SDOH

social determinants of health

SSoC

Stroke Systems of Care