But the last mile in healthcare is not a place. It is a person.

It is the nurse trying to navigate a new workflow while managing clinical risk in real time. It is the doctor deciding whether to trust an alert or the patient sitting in front of them. It is the allied health professional moving between fragmented systems while trying to preserve continuity, safety, and trust.

Clinicians are expected to absorb policy, procurement, governance, compliance, and technology decisions and somehow translate them into care that still feels safe, human, and efficient. They are where strategy either becomes reality or quietly collapses under operational pressure.

That is why asking the clinician last may be one of the most expensive habits in healthcare transformation.

Why top-down healthcare transformation struggles

Healthcare still designs change in a sequence that feels rational from the boardroom but fragile at the bedside. Leaders define the problem, approve the investment, procure the platform, configure workflows, develop training, and set the go-live date. Only then are clinicians expected to adopt what has already become politically, financially, and operationally difficult to change.

We often call this engagement. In reality, it is frequently a consultation after commitment.

Research into healthcare transformation consistently shows that programmes shaped through participation and meaningful contribution outperform purely top-down approaches. That does not mean healthcare should abandon governance, prioritisation, or executive discipline. Health systems still need leaders willing to make difficult decisions about investment, sequencing, and trade-offs.

But leadership is not the same thing as designing from altitude.

In complex healthcare environments, the people closest to the work are often the best interpreters of reality. They understand where friction accumulates, where safety risks hide, and where a seemingly elegant workflow will quickly become a workaround.

Clinician adoption determines whether digital health succeeds.

This becomes even more important with digital health and AI.

A tool may perform well in a study and still fail in practice if it does not fit clinical workflows, earn trust, or support professional judgement in the clinical moment. Research into clinician adoption of AI found that perceived usefulness was the strongest predictor of whether clinicians intended to use a system. But adoption was also shaped by workflow integration, trust, autonomy, and perceived risk.

In simpler terms, clinicians adopt technology when it helps them care for patients without making them feel less safe, less efficient, or less clinically responsible.

Before exploring implementation strategies, consider what clinicians identify as essential ingredients for successful digital transformation.

The recommendations emphasise involving healthcare professionals throughout digitisation, building digital capability over time, and ensuring systems are interoperable, intuitive, and supported by strong organisational backing.

Beyond abstract recommendations, clinicians’ lived experience in digital hospitals reveals tensions that transformation dashboards rarely capture.

The findings highlight recurring concerns around documentation burden, inconsistent data quality, hybrid workflows, privacy concerns, and disruptions to clinician-patient interaction.

This is where many transformation programmes become dangerous. Dashboards may show successful deployment milestones, completed training, and rising login numbers, while clinicians’ lived experience is that work has become slower, cognitive load has increased, and workarounds have quietly become the true operating model.

This disconnect illustrates why healthcare transformation is rarely just a technology shift and why complexity is often underestimated.

The thematic analysis shows how digital transformation simultaneously affects workflows, communication, wellbeing, patient care, trust, interoperability, and organisational change. It reinforces why clinician adoption cannot be treated as a simple training or compliance exercise.

Addressing that gap requires more than better technology; it demands stronger leadership.

Clinician scepticism is not resistance.

Healthcare organisations often interpret clinician scepticism as resistance to change. In reality, scepticism can be a strategic asset. The sceptical clinician is frequently the person who can identify where a workflow will fail, where patient safety may be compromised, or where the promised efficiency has simply shifted hidden labour onto frontline staff.

Sometimes clinicians are not resisting the future. They are protecting the future from poor implementation.

If healthcare organisations truly want transformation, clinician engagement needs to begin before procurement decisions are set in stone.

Real engagement means involving clinicians early enough that their input can still change the design. It means recognising clinical judgement as one of the strongest safeguards against waste, harm, and failed adoption.

The last mile in healthcare has always been human. It is a workload, a professional judgement, and a patient interaction.

Ask the clinician first, because transformation fails if the person expected to deliver it quietly decides it does not work.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Hastings, B. J. (2025). Guidance for successful healthcare transformation: A systematic review of change management practices and outcomes. Australian Journal of Management, 50(3), 985–1005. 
• Mohamed, G. A. N., Ahmed, I. E., Mohamed, A., & Parambath, S. A. (2026). Factors influencing healthcare professional adoption of artificial intelligence: A mixed-methods systematic review and meta-analysis. InfoScience Trends, 3(11), 1–19. 
• Canfell, O. J., Woods, L., Meshkat, Y., Krivit, J., Gunashanhar, B., Slade, C., Burton-Jones, A., & Sullivan, C. (2024). The impact of digital hospitals on patient and clinician experience: Systematic review and qualitative evidence synthesis. Journal of Medical Internet Research, 26, e47715. 
• Harrison, R., Fischer, S., Walpola, R. L., Chauhan, A., Babalola, T., Mears, S., & Le-Dao, H. (2021). Where do models for change management, improvement and implementation meet? A systematic review of the applications of change management models in healthcare. Journal of Healthcare Leadership, 13, 85–108. 
• Wosny, M., Strasser, L. M., & Hastings, J. (2023). Experience of health care professionals using digital tools in the hospital: Qualitative systematic review. JMIR Human Factors, 10, e50357. 
• Borges do Nascimento, I. J., Abdulazeem, H., Vasanthan, L. T., Martinez, E. Z., Zucoloto, M. L., Østengaard, L., Azzopardi-Muscat, N., Zapata, T., & Novillo-Ortiz, D. (2023). Barriers and facilitators to utilizing digital health technologies by healthcare professionals. npj Digital Medicine, 6, 161. 

In healthcare, buying a system is not the same as adopting it. Adoption is not the same as sustained use. Yet organisations still measure success by contracts signed, systems launched, or early login numbers, instead of asking the harder question: has care actually improved?

That distinction matters because healthcare technology rarely fails solely on functionality. It fails when transformation is treated as a software rollout instead of a systems design challenge.

Why healthcare technology adoption is a systems problem.

Digital health exists inside one of the most operationally complex environments imaginable. Clinical care is unpredictable, time-pressured, fragmented, and heavily dependent on human judgement. Technology doesn’t replace that complexity. It inherits it.

The NASSS framework helps explain why so many healthcare programmes struggle to scale or sustain momentum. Rather than focusing solely on the technology, the framework considers the conditions shaping adoption, including clinical need and user buy-in, as well as organisational readiness and long-term adaptability.

The important insight is: digital health is never just a technology intervention. It is organisational change embedded inside operational pressure.

That changes the executive conversation entirely. Instead of asking whether a platform has enough features, leaders need to ask whether the organisation has the operational capacity, governance, and workflow maturity to absorb change at scale.

Technology only works if people change their behaviour.

The first test of any healthcare technology is whether it solves a problem painful enough to change behaviour.

If the only people who see the value are the vendor, programme team, or executive sponsor, adoption will remain fragile. Clinicians need to feel that the technology improves decisions, reduces friction, or makes patients safer. Patients need to see better access, clearer communication, or stronger continuity of care. Executives need evidence that operational and financial pressures are improving.

Without aligned value across those groups, usage becomes compliance rather than commitment.

This is where many implementations quietly stall. Health systems often assume access equals engagement. It doesn’t. Most clinicians already operate at cognitive overload. If a platform introduces extra clicks, duplicates documentation, or disrupts workflow, staff will find ways around it regardless of how strategically important the programme seemed during procurement.

Workflow friction is an infrastructure issue.

Healthcare organisations frequently frame adoption challenges as training problems. In reality, they are often infrastructure problems.

Poor interoperability, fragmented data flows, inconsistent governance, and disconnected workflows create operational drag long before the user interface becomes the issue. Clinicians don’t experience “digital strategy” as a concept. They experience it as friction during a consultation.

This is why digital transformation succeeds less through individual tools and more through coherent system design. Organisations that achieve long-term success tend to focus on shared infrastructure, integrated data foundations, and operational alignment rather than on isolated applications.

In other words, sustainable adoption is usually a consequence of better system architecture.

The real risk starts after launch.

Healthcare programmes rarely fail during implementation. They fail afterwards.

The energy, funding, and executive attention invested in procurement and go-live often dissipate once the implementation team leaves. Meanwhile, the difficult work of workflow redesign, governance, optimisation, and continuous adaptation begins.

Sustained success requires organisations to continuously absorb change. That means long-term ownership, operational discipline, ongoing funding, and the ability to adapt technology as clinical realities evolve.

Without that capability, health systems end up with portfolios of promising tools competing for attention but failing to compound into meaningful system-wide value.

The practical lesson for healthcare executives is simple: stop confusing inputs with outcomes.

Success is not whether technology has been purchased, implemented, or launched. Success is whether the behaviour has changed. Whether the workflow improved. Whether decisions became faster or safer. Whether the organisation operates differently a year later.

Bought is not adopted. Adopted is not sustained.

The real test of healthcare technology has never been whether it gets through the door. It’s whether the system around it is capable of turning technology into better healthcare.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Amiri, S., Mahmood, N., Mustafa, H., Javaid, S. F., & Khan, M. A. (2024). Occupational risk factors for burnout syndrome among healthcare professionals: A global systematic review and meta analysis. International Journal of Environmental Research and Public Health, 21(12), 1583. 
• lobayli, F., O’Connor, S., Holloway, A., & Cresswell, K. (2023). Electronic health record stress and burnout among clinicians in hospital settings: A systematic review. Digital Health, 9, 1 to 17. 
• Nagarajan, R., Ramachandran, P., Dilipkumar, R., & Kaur, P. (2024). Global estimate of burnout among the public health workforce: A systematic review and meta analysis. Human Resources for Health, 22, 30. 
• National Academies of Sciences, Engineering, and Medicine. (2019). Taking action against clinician burnout: A systems approach to professional well being. The National Academies Press. 
• Shanafelt, T. D., West, C. P., Sinsky, C., Trockel, M., Tutty, M., Wang, H., Carlasare, L. E., & Dyrbye, L. N. (2025). Changes in burnout and satisfaction with work life integration in physicians and the general US working population between 2011 and 2023. Mayo Clinic Proceedings, 100(7), 1142 to 1158. 

The vision for a Single Patient Record (SPR) in England is clear. 

The challenge now is how to move from vision to delivery in a way that is practical, achievable and builds on what already exists. 

Across the NHS, Shared Care Records have already made significant progress in connecting information across organisations within local systems. The next step is strengthening that capability, so it works consistently across boundaries. 

This is what we mean by SPR readiness. 

It’s not about building a new system. It’s about making existing systems work together more effectively so information can be shared, trusted and used wherever care is delivered. 

With increasing demand and pressure to improve productivity, how information flows across the system is becoming critical. Reducing friction in accessing and sharing information can make a real difference to both clinician experience and patient care. 

Based on our work in this space, here are five practical steps the NHS can take to help get started. 

1. Connect what already exists 

The starting point isn’t new systems. It’s making better use of the ones already in place. 

SPR readiness means linking existing systems, like Shared Care Records, into a trusted, federated network, rather than replacing them. 

2. Reduce information friction 

Too much time is still spent searching for information, duplicating work and re-documenting what already exists. 

Improving how information flows across systems helps reduce this friction, freeing up time and supporting better, more coordinated care. 

3. Invest in adoption and data quality 

Technology alone isn’t enough. 

Information needs to be integrated into day-to-day clinical workflows and trusted by those using it. That means focusing on data quality, usability and adoption, not just connectivity. 

4. Adopt standards and real-time exchange 

Consistent standards and more timely data sharing are essential to enable shared workflows and ensure updates are reflected across systems. 

Moving towards real-time exchange where it matters most helps improve coordination and decision-making. 

5. Strengthen governance and trust 

Clear information governance, auditability and patient confidence are critical. 

SPR readiness depends on creating the right conditions for safe, cross-boundary data sharing, with trust at the centre. 

From readiness to action 

None of this requires starting again. 

The opportunity now is to build on what’s already in place and take practical steps towards making information flow more consistently across the NHS. 

For a deeper look, our latest white paper, written in collaboration with Healthcare Innovation Consortium, explores how to get started with SPR readiness and build on existing capability across the NHS.  

Why burnout is a system issue, not an individual one.

Burnout is an occupational phenomenon caused by chronic workplace stress, expressed through exhaustion and reduced professional efficacy. In healthcare, it reflects how work is structured, demand intensity, workflow design, tool usability, and the level of control clinicians have over their work.

The National Academies describes burnout as a systems issue that affects frontline care, organisational design, and the wider environment. When job demands are consistently higher than available resources, burnout is likely to occur.

How we frame burnout matters. If we see it as an individual issue, the response is limited to well-being programs. If we see it as a system design problem, the focus shifts to improving staffing, workflows, incentives, and technology.

How technology contributes to burnout.

Digital health highlights how system design can either support or strain clinicians.

The authors of a 2023 systematic review developed a model to summarise their findings. The model provides an explicit answer to the systematic review’s two objectives. First, based on consensus across the reviewed studies, the association between EHR use and clinicians’ stress and burnout was found to be positive (indicated by a continuous, one-directional arrow). Second, the review identified several contributing factors that mediated or moderated the impact of EHR use on clinicians’ stress and burnout.

As shown in this chart, clinician stress rises alongside increased EHR use, with reported burnout rates reaching up to 65% in some countries.

The biggest factors are how easy systems are to use and the time spent in them, especially in busy hospital settings.

Technology is not neutral. If it is not well integrated, it adds clicks, duplicates documentation, and extends work into after-hours. This changes clinical workflows and increases mental strain.

Workplace conditions drive burnout

The broader evidence reinforces that burnout is driven by workplace conditions, not individual resilience.

The visual shows that workplace demands, such as toxic behaviour, role ambiguity, and workload, are seven times more predictive of burnout than positive enablers.

This graph highlights how administrative burden, particularly time spent in EHRs, is a leading contributor to burnout.

For many clinicians, administrative work is not just inefficient. It is the main source of stress.

Why well-being alone falls short

Burnout affects performance, quality, safety, staff retention, and costs. Still, many responses focus on individual support, such as counselling, coaching, and resilience training.

These interventions are helpful, but they are not enough. If a strategy ignores staffing, workflow problems, documentation burden, and poor system design, it only treats the symptoms, not the real causes.

Designing systems that reduce friction

Healthcare will always involve complexity and high-stakes decision-making. The goal is not to remove that complexity, but to eliminate avoidable friction.

This requires:

• Streamlined workflows that reduce duplication
• Technology that supports, rather than interrupts, clinical work
• Documentation aligned with patient value
• Greater clinician autonomy
• Leadership accountability for work conditions

Burnout is not unavoidable. It is a sign that the system is under too much strain.

The Future of Healthcare Leadership

Healthcare leaders now need to build systems that do not rely on overworked staff to keep things running.

Successful organisations will see clinician capacity as a key asset. They will create workplaces where technology fits in smoothly, administrative tasks are reduced, and clinicians can focus on patient care.

Burnout does not happen because people lack resilience. It happens when systems depend on resilience to make up for their flaws.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Amiri, S., Mahmood, N., Mustafa, H., Javaid, S. F., & Khan, M. A. (2024). Occupational risk factors for burnout syndrome among healthcare professionals: A global systematic review and meta analysis. International Journal of Environmental Research and Public Health, 21(12), 1583. 
• lobayli, F., O’Connor, S., Holloway, A., & Cresswell, K. (2023). Electronic health record stress and burnout among clinicians in hospital settings: A systematic review. Digital Health, 9, 1 to 17. 
• Nagarajan, R., Ramachandran, P., Dilipkumar, R., & Kaur, P. (2024). Global estimate of burnout among the public health workforce: A systematic review and meta analysis. Human Resources for Health, 22, 30. 
• National Academies of Sciences, Engineering, and Medicine. (2019). Taking action against clinician burnout: A systems approach to professional well being. The National Academies Press. 
• Shanafelt, T. D., West, C. P., Sinsky, C., Trockel, M., Tutty, M., Wang, H., Carlasare, L. E., & Dyrbye, L. N. (2025). Changes in burnout and satisfaction with work life integration in physicians and the general US working population between 2011 and 2023. Mayo Clinic Proceedings, 100(7), 1142 to 1158. 

But there’s a catch: AI models are only one part of the system. The real constraint is whether healthcare organisations can access, trust, and use the data on which those models depend. Put simply, healthcare isn’t limited by AI. It’s limited by data, and data depends on interoperability.

The data problem behind healthcare AI

AI needs more than just volume; it needs quality, diversity, and context.

That includes:

• Clinical data, diagnostics, and medications
• Social and behavioural context
• Patient-generated data
• Longitudinal health journeys

The challenge? Most of this data is fragmented across systems, inconsistently coded, and often locked within organisational boundaries.

Even when digitised, it’s not always usable in real time. That’s why recent research consistently points to the same conclusion: AI in healthcare requires an interoperable data ecosystem to succeed.

Why fragmented data limits AI impact

A model trained on incomplete data can still look impressive, but it delivers narrow insights.

Without full patient context:

• Clinical decision support can miss critical signals.
• AI recommendations lose credibility.
• Outcomes become harder to trust.

Electronic health records (EHRs) improved digitisation, but they didn’t solve connectivity. In many cases, they reinforced silos instead of enabling connected care.

That’s why interoperability isn’t just technical plumbing. It’s strategic infrastructure.

How interoperability enables scalable AI

Interoperability determines whether data can move across systems, and whether it retains meaning when it does.

The evolution of interoperability

Understanding how we got here highlights why gaps still exist.

From early EMRs in the 1970s to modern standards like FHIR, progress has been steady. But as the graph shows, adoption hasn’t kept pace with the complexity of modern care environments, leaving critical gaps in scalable data exchange.

The layers that matter

Effective interoperability operates across multiple levels:

• Technical – systems can connect.
• Syntactic – data is structured consistently.
• Semantic – meaning is preserved.
• Organisational and legal – data is shared safely and appropriately.

Without alignment across these layers, data may move, but it won’t be usable.

Why interoperability is now a strategic priority

Healthcare leaders are already shifting their focus.

• 84.7% of decision-makers say interoperability is a higher priority than two years ago
• Key drivers include:
  • Improved operational efficiency
  • Better patient experience
  • Enabling value-based care

This isn’t about compliance anymore. Interoperability is becoming a core operational and commercial capability.

Rethinking AI investment in healthcare

Many organisations start with the wrong question:
“Which AI model should we buy?”

The better question is:
“Do we have the data foundation to make AI safe, effective, and scalable?”

That means:

• Ensuring data is accessible across care settings
• Standardising how data is captured and shared
• Embedding governance that builds trust

Because healthcare isn’t just a data problem, it’s a trust, safety, and equity problem.

• Patients need confidence that their data is used appropriately.
• Clinicians need confidence in AI recommendations.
• Regulators need assurance that models can be governed.

Without this, AI won’t scale.

From fragmented data to connected intelligence

The future of healthcare AI won’t be defined by the most advanced models. It will be defined by enterprise discipline.

The organisations that succeed will:

• Build strong, interoperable data foundations.
• Apply consistent standards and governance.
• Connect data across the full patient journey.

This is how healthcare moves from:

• Fragmented records to connected intelligence
• Passive data to an active decision-making asset

Closing the interoperability gap

Healthcare doesn’t have an AI shortage. It has an interoperability gap.

Until that gap is closed, even the most powerful AI models will remain constrained by the data beneath them.

The opportunity now is clear: build the foundations that allow AI to deliver real, measurable impact, safely and at scale.

Ready to Build a Stronger Data Foundation?

Interoperability isn’t just a technical upgrade; it’s the foundation for better outcomes, smarter systems, and scalable AI.

Explore how Orion Health enables connected, interoperable healthcare ecosystems.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Adegoke, K., Adegoke, A., Dawodu, D., Adekoya, A., Bayowa, A., Kayode, T., & Singh, M. (2025). Interoperability as a catalyst for digital health and therapeutics: A scoping review of emerging technologies and standards (2015–2025). International Journal of Environmental Research and Public Health, 22, 1535. 
• Hussein, R., Gyrard, A., Abedian, S., Gribbon, P., & Martínez, S. A. (2025). Interoperability framework of the European Health Data Space for the secondary use of data: Interactive European Interoperability Framework based standards compliance toolkit for AI driven projects. Journal of Medical Internet Research, 27, e69813. 
• Mandl, K. D., Gottlieb, D., & Mandel, J. C. (2024). Integration of AI in healthcare requires an interoperable digital data ecosystem. Nature Medicine, 30, 631–634. 
• Saberi, M. A., Mcheick, H., & Adda, M. (2025). From data silos to health records without borders: A systematic survey on patient centred data interoperability. Information, 16, 106. 
• Singh, M., Siek, K., Danks, D., Ghani, R., Griffin, H., LaMacchia, B., Lopresti, D., & Toscos, T. (2024). Enabling the AI revolution in healthcare. Computing Research Association.
• Snowflake & Hakkoda. (2026). The future of AI + interoperability in healthcare report: The role of interoperability in scaling AI in healthcare. Snowflake.
• Zahlan, A., Ranjan, R. P., & Hayes, D. (2023). Artificial intelligence innovation in healthcare: Literature review, exploratory analysis, and future research. Technology in Society, 74, 102321. 

For years, the focus has been on models. Which perform best, which vendors lead, and which pilots to run. That framing is now outdated. Models are rapidly improving and becoming more accessible.

The real constraint has shifted. Today, the limiting factor is the system around the model.

The AI pressure clock. When deployment outpaces governance.

The AI pressure clock highlights a growing risk.

When organisations face high disruption but lack resilience, they enter a state of institutional vulnerability. In this zone, the issue is not the technology itself, but the surrounding governance, workflows, and accountability structures.

Evidence from US hospitals makes this gap clear.

Predictive AI is already widely adopted. However, many organisations still lack formal processes for accuracy evaluation and bias assessment. Deployment is outpacing governance and increasing risk beneath the surface.

What mature AI governance looks like.

The HAIRA maturity model offers a clear benchmark for AI readiness.

At higher maturity levels, organisations move beyond tool deployment. They:

• Run multi-centre validation studies,
• Embed predictive monitoring into routine care,
• Operate governance at an executive level with real-time oversight.

Lower maturity environments remain reactive and vendor-led. Decisions lack structured evaluation, and monitoring is inconsistent.

This is not a gradual progression. It is a shift in operating model.

AI-decorated vs AI-native organisations.

The divide becomes clearer when comparing AI-decorated and AI-native organisations.

AI-decorated organisations focus on pilots and isolated tools, measuring success through activity.

AI-native organisations redesign workflows end-to-end. They treat data as a strategic asset and embed AI into decision-making at scale.

The impact is measurable:

• AI-decorated organisations struggle to sustain value.
• AI-native organisations gain speed, productivity, and stronger resource allocation.

For leadership, this reframes investment. Funding models alone will not deliver results if the system cannot support them.

Where investment should shift.

Sustainable AI adoption depends on system-level investment.

Governance, interoperability, validation frameworks, and clinical integration must be treated as core capabilities. This is where resilience is built.

Leadership also plays a critical role. Moving beyond pilots requires:

• Executive ownership,
• Clear accountability,
• Discipline to stop initiatives that do not scale.

Without this, organisations accumulate disconnected capabilities that fail to deliver enterprise value.

At a board level, the focus must shift from counting AI initiatives to assessing readiness. This includes governance strength, evaluation standards, and risk visibility.

The emerging divide in healthcare AI.

The divide is already forming.

Organisations that build the right foundations will not just adopt AI. They will reshape how care is delivered. Those that do not will remain stuck in pilot cycles, with growing exposure.

The question is not whether to invest in AI.
It is whether your organisation is ready to support it at scale.

Build the foundations for scalable AI.

AI success starts with the system, not the model.

From interoperability to governance and real-time data integration, strong digital foundations enable safe, scalable AI.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Frimpong, V. (2025). When institutions cannot keep up with artificial intelligence: Expiration theory and the risk of institutional invalidation. Administrative Sciences, 15(7), 263.
• Hussein, R., Zink, A., Ramadan, B., Howard, F. M., Hightower, M., Shah, S., & Beaulieu Jones, B. K. (2026). Advancing healthcare AI governance through a comprehensive maturity model: A systematic review. npj Digital Medicine.
• Hwang, Y. M., Ng, M. Y., Pillai, M., et al. (2026). The landscape of artificial intelligence implementation in US hospitals. Nature Medicine, 32(11), 99–112.
• Kelly, C. J., Karthikesalingam, A., Suleyman, M., Corrado, G., & King, D. (2019). Key challenges for delivering clinical impact with artificial intelligence. BMC Medicine, 17(1), 195.
• Mathews, G. (2024). AI native divide.
• Rajkomar, A., Dean, J., & Kohane, I. (2019). Machine learning in medicine. New England Journal of Medicine, 380(14), 1347–1358.
• Topol, E. (2019). Deep medicine: How artificial intelligence can make healthcare human again. Basic Books.

These infrastructures include triage algorithms, AI assistants, digital health records, and interoperable data platforms. They are becoming the operating system of modern health systems.

The shift in power: from providers to platforms

Power is shifting away from traditional providers to those who design and run digital infrastructure.

The key question is not whether digital transformation will happen, but whether public health systems can maintain oversight of the systems that control access, allocation, and innovation.

At the centre are large technology firms.  They are becoming systemic actors that shape how data is collected and monetised.

Public health systems generate vast amounts of data, but it is increasingly processed and monetised in ecosystems outside public control.

Governance gaps in the health data economy

Existing regulatory frameworks were not designed for a world in which data is continuously repurposed and used to train adaptive systems.

The General Data Protection Regulation (GDPR) attempts to balance protection with innovation. However, its broad interpretation of “scientific research” creates ambiguity. Activities undertaken by private actors can fall within the same category as public interest research, enabling access to data under more permissive conditions.

This creates a structural asymmetry. Public systems remain accountable to citizens, while private actors operate with greater flexibility in how they use and combine data. Over time, this erodes the public sector’s ability to control not only data but also the direction of innovation itself.

The new front door to healthcare

The more immediate risk lies in access.

Increasingly, patients are not entering the system through traditional channels such as general practice or national helplines. They are turning to digital platforms and AI assistants that provide instant, personalised guidance.

These tools handle personal health data, give triage advice, and help decide when and where people seek care. Triage is not neutral; it affects how demand is distributed, which in turn changes wait times, priorities, and outcomes.

When triage moves outside public infrastructure, control over these dynamics moves with it. Platforms optimise for user experience and engagement, not for system sustainability or equitable allocation.

Data capture, consent, and platform dynamics

At the same time, platforms are building longitudinal health records outside public systems, as patients upload data into convenient private environments.

Over time, this creates switching costs. As private records become more complete, patients are less likely to return to fragmented public systems, driving platform capture through gradual shifts in behaviour.

Open vs closed AI: control vs convenience

This dynamic extends to AI infrastructure. Closed models are easier and cheaper to deploy, but reduce control and increase dependence. Open or local models offer greater control and data sovereignty but require more capability and investment.

Governing digital infrastructure as a strategic asset

The strategic challenge is not whether to adopt digital infrastructure, but how to govern it.

Public systems must treat data infrastructure as a strategic asset, maintaining custodianship, setting interoperability standards, and controlling how data flows across the ecosystem. Governance also needs to shift from focusing on who uses data to how it is used, distinguishing between beneficial and extractive uses.

Strong institutional capability is essential. Governance bodies must have the authority to oversee data access, approve partnerships, and enforce accountability.

Control of access pathways is equally critical. Public systems need to shape how digital triage operates, ensuring decisions reflect clinical priorities rather than commercial incentives.

Finally, there is an economic dimension. When private actors derive value from publicly generated data, mechanisms must ensure that value is shared.

A defining choice for health leaders

The core issue is control.

Health systems have long been defined by their ability to allocate resources based on need and maintain public trust. These capabilities are now shaped by digital infrastructure.

If control of that infrastructure shifts, the foundations of public healthcare are at risk.

The choice is clear: passive adoption leads to loss of control, while active governance enables systems to harness digital capabilities while preserving equity, accountability, and long-term sustainability.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Marelli, L., Testa, G., & Van Hoyweghen, I. (2021). Big Tech platforms in health research: Re-purposing big data governance in light of the General Data Protection Regulation’s research exemption. Big Data & Society, 8(1), 1–14. 
• Rikap, C., & Lundvall, B.-Å. (2022). Big tech, knowledge predation and the implications for development. Innovation and Development, 12(3), 389–416. 
• Siira, E., Johansson, H., & Nygren, J. (2025). Mapping and summarizing the research on AI systems for automating medical history taking and triage: Scoping review. Journal of Medical Internet Research, 27, e53741. 
• Tony Blair Institute for Global Change. (2024). Preparing the NHS for the AI era: A digital health record for every citizen. 
• Tony Blair Institute for Global Change. (2026). Who controls access to NHS care in the age of Big Tech? 
• Williams, M., Karim, W., Gelman, J., & Raza, M. (2024). Ethical data acquisition for LLMs and AI algorithms in healthcare. npj Digital Medicine, 7, 377.

Following A National Blueprint for the Single Patient Record, the next challenge is readiness, ensuring the right capabilities are in place to connect existing systems into a unified, trusted network of care.

Today, information still stops at organisational and regional boundaries. Clinicians spend valuable time searching for data, reconciling records, and repeating work, creating avoidable friction that impacts both productivity and patient outcomes.

That’s why Orion Health has partnered with Healthcare Innovation Consortium (HIC) to define the next step: preparing the NHS for a Single Patient Record (SPR).

Our latest white paper, From Blueprint to Readiness: Preparing for a Single Patient Record, outlines how to turn strategy into delivery. It explores:

• What “SPR readiness” means across technical, clinical, and governance capabilities
• How to connect existing systems through a federated, interoperable model
• Where readiness can unlock measurable productivity gains and reduce duplication
• Real-world use cases, including frailty and cancer pathways

The opportunity now is to move from blueprint to readiness, reducing information friction, improving care coordination, and enabling a more productive, connected NHS.

For years, vendors have delivered systems to support tasks such as documentation, scheduling, and analytics, while providers carried the burden of turning those tools into clinical or financial outcomes.

That model is now under pressure from three forces: financial constraint, workforce fragility, and AI that can execute work, not just assist it.

Why healthcare is shifting from tools to outcomes.

Health system leaders are prioritising cost control and care model transformation, with AI enabling more standardised workflows and scalable digital services.

Technology is no longer the product; it is an input into services that must demonstrate economic and clinical value.

Regulation, AI, and global pressures are accelerating this shift toward outcome-based models.

Execution is the constraint, not technology.

Despite investment, many organisations remain stuck in execution paralysis, where fragmented processes and misalignment prevent AI from delivering measurable value.

The advantage now lies with organisations that can operationalise intelligence into repeatable, outcome-producing workflows. As intelligence becomes abundant, tools commoditise, and value shifts to those delivering the work.

Routine, rules-based work, especially already outsourced functions, is most exposed to AI-driven execution.

AI is automating administrative work at scale.

The earliest impact is in administrative domains.

Medical coding and claims processing are rules-based, outcome-measured tasks, making them highly suited to AI-driven automation.

The gap between what AI can do and what it is already doing is rapidly narrowing, particularly in structured workflows.

This is not just efficiency. It is the substitution of manual work with automated, outcome-based services.

From reactive care to proactive, AI-driven models.

AI is also reshaping care delivery.

Predictive analytics, remote monitoring, and AI-driven follow-up are enabling a shift from reactive treatment to proactive, preventative care.

Value is moving from activity to measurable improvements in health outcomes, aligning with value-based care and whole health models.

From tools to orchestrated value streams.

Healthcare organisations must now orchestrate value streams.

AI needs to be embedded across clinical and administrative workflows, not layered on top.

Leading organisations are:

• linking AI to core value streams like diagnostics and patient flow,
• scaling beyond pilots,
• building integrated, end-to-end workflows.

At the same time, architectures are evolving toward modular, connected systems where multiple AI capabilities are coordinated across workflows.

In this model, the system that controls orchestration controls value.

The rise of outcome-based AI business models.

The economics of healthcare AI are shifting.

A new generation of companies combines service delivery with software economics, achieving software-like margins while delivering human-level outcomes. They are not selling licences. They are selling results.

This shifts the competitive dynamic:

• value accrues to those who own outcomes,
• data compounds advantage,
• AI improves through execution.

Over time, more of the value chain becomes automated.

What this means for healthcare leaders.

This is a strategic inflection point.

Organisations that deploy tools without owning outcomes risk disintermediation, while those that control both data and outcomes will capture disproportionate value.

This requires a shift:

• from product deployment to service orchestration,
• from features to measurable impact,
• from technology spend to value-based investment.

As AI embeds into care and decision-making, governance becomes critical, not just for compliance, but as a driver of trust and value.

Outcomes are becoming the only currency in healthcare.

The industry is still early in this transition. Adoption remains uneven, and many organisations have yet to scale or realise full value.

But the direction is clear.

Those who act early will design for outcomes, embed AI into workflows, and build feedback loops that continuously improve performance. Others will compete in an increasingly commoditised layer of the stack.

This shift is structural.

In healthcare, services are becoming the new software, because outcomes are becoming the only currency that matters.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Bek, J. (2026, March 5). Services: The new software. Sequoia Capital.
• Bessemer Venture Partners. (2026, January 22). State of health AI 2026.
• Boston Consulting Group & BCG X. (2025, December). How AI agents and tech will transform health care in 2026.
• Deloitte Center for Health Solutions. (2025, December 11). 2026 global health care outlook.
• Deloitte Center for Health Solutions. (2025, December 9). 2026 life sciences outlook.
• Guidehouse. (2026). Healthcare AI trends report.
• KPMG International. (2025). Intelligent healthcare: A blueprint for creating value through AI-driven transformation.
• Krishna, A., Friend, D., Gohad, N., & Reddy, P. (2025, November). The coming evolution of healthcare AI toward a modular architecture. McKinsey & Company.
• Massenkoff, M., & McCrory, P. (2026, March 5). Labor market impacts of AI: A new measure and early evidence.

When viewed this way, the conversation becomes less ideological and more operational.

Why governments engage the private sector

There are clear, practical reasons governments involve private providers, especially in systems under fiscal or workforce pressure.

Private providers can:

• Expand capacity quickly without major capital investment.
• Reduce waiting lists.
• Introduce operational flexibility.

In New Zealand, outsourcing elective procedures has helped treat more lower-risk patients and reduced pressure on public hospitals. Around the world, private providers often step in where public systems can’t reach or don’t have enough resources.

In these situations, involving private providers isn’t just a choice; it’s often a practical way to address system limitations.

Why healthcare doesn’t behave like a normal market

There’s a lot of uneven information, people usually can’t choose when they need care, and it’s hard to measure results compared to costs. These issues can skew incentives.

Because of this, private providers often focus more on saving money than on quality of care. This can cause:

• Selective intake of lower-risk, more profitable patients,
• Delivery models focused on cost control over long-term outcomes.

This behaviour isn’t accidental; it reflects the system’s design.

The shift in clinical complexity

When public and private sectors work side by side, a clear pattern appears: the complexity of cases shifts.

Private providers tend to deliver high-volume, lower-complexity care, while public systems retain higher-risk, resource-intensive patients.

This imbalance has structural consequences:

• Public systems become the default provider for complexity and cost.
• Workforce pressure increases as both sectors draw from the same talent pool.
• Training opportunities decline as simpler cases move out of the public system.

Over time, this erodes system capability.

The impact on quality and outcomes

The evidence on quality is mixed, but trending negative.

Longitudinal studies show that increased outsourcing is often associated with:

• Reduced staffing ratios
• In some cases, worse health outcomes

There is little consistent evidence that privatisation improves care quality. Some studies even link higher levels of outsourcing with more avoidable deaths.

At best, the impact is uncertain. At worst, it is detrimental.

Governance and system control challenges

Privatisation also introduces complexity at a system level.

Outsourcing can:

• Reduce accountability,
• Limit visibility into cost and performance,
• Fragment decision-making.

These problems make it harder for the system to use resources well and adapt when needs change.

The equity trade-off

The most significant impact is on equity.

Evidence consistently shows that privatisation can:

• Increase inequities in access and outcomes.
• Benefit lower-risk, higher-income populations.
• Underserve vulnerable groups.

In New Zealand, there are concerns that outsourcing may primarily benefit already advantaged populations, potentially widening inequities, particularly for Māori and rural communities.

If the system isn’t carefully planned, it just shifts benefits around instead of making things better overall.

Designing a system that works

The question isn’t whether the private sector should be involved; it already is.

The real question is: under what conditions does it improve system-level outcomes?

A disciplined approach to system design includes:

Define the role clearly.

Use private providers to absorb overflow demand and deliver standardised care, not to define care pathways or system architecture.

Align incentives to outcomes.

Shift from volume-based models to those that reward quality, continuity, and long-term outcomes.

Protect the public system’s capability.

Ensure the public system remains the anchor for:

• Complex care
• Workforce training
• Clinical capability development

Strengthen Transparency and Governance

Full visibility of cost, performance, and outcomes is essential to effective system management.

A question of design, not ideology

The private sector isn’t automatically a problem. If managed well, it can help the system run more efficiently. If not, it can fragment care and increase inequity.

The evidence doesn’t support wholesale privatisation. It supports disciplined, conditional participation.

For healthcare leaders, the challenge isn’t choosing sides. It’s designing systems where capital allocation and clinical outcomes are aligned.

Take the Next Step

Designing a more connected, equitable health system starts with better data use across organisational boundaries.

See how Orion Health’s Shared Care Record solutions can help everyone in the system work together, share information, and get better outcomes:

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Anaf, J., Freeman, T., & Baum, F. (2024). Privatisation of government services in Australia: What is known about health and equity impacts. Globalization and Health, 20(32).
• Baker, G. (2025, November 28). Entrenching outsourcing a hit on other providers and equity. New Zealand Doctor. 
• Goodair, B., & Reeves, A. (2024). The effect of health-care privatisation on the quality of care. The Lancet Public Health, 9(3), e199–e206.\
• McInerney, M., Hinchcliff, R., FitzGerald, G., & King, R. (2023). Private equity investment in private for-profit healthcare in Australia and New Zealand: A scoping review. Asia Pacific Journal of Health Management, 18(2), Article i2347. 
• Newton, K. (2025, October 9). Outsourcing easy cases to private sector risks two-tier health system, doctors warn. RNZ News. 
• Siddiqi, S., Aftab, W., Venkat Raman, A., Soucat, A., & Alwan, A. (2023). The role of the private sector in delivering essential packages of health services: Lessons from country experiences. BMJ Global Health, 8, e010742. 

Orion Health is supporting this next phase of Iowa’s statewide exchange with a standards-based interoperability platform designed to enable secure, scalable data exchange aligned to state and federal requirements. The platform will support clinical, public health, and community-based information sharing across diverse provider environments. 

State-level modernization efforts increasingly require infrastructure that performs reliably amid regulatory and operational complexity. Orion Health’s role is to provide the interoperability foundation that enables governance alignment, continuity of exchange, and long-term sustainability as healthcare ecosystems evolve. 

This engagement reflects Orion Health’s continued support of state and regional programs modernizing healthcare data infrastructure to strengthen coordination, accountability, and visibility across care settings. 

The full announcement from Converge Health Iowa can be read here.

Even with all this data, only a small part is regularly used to improve care. Many people see the potential of health data, but its real value is still not fully reached.

For more than two decades, interoperability has been positioned as the key to unlocking that value. Governments invested heavily in electronic health records (EHRs), standards bodies developed frameworks for exchanging information, and health systems built networks to connect organisations that historically operated in isolation.

These steps were important and have helped, but they haven’t fully solved the problem. Just being connected doesn’t mean the shared information is actually useful.

Interoperability is necessary, but not enough.

In many healthcare systems, the technical ability to exchange data exists. Yet clinicians still struggle to integrate external information into routine workflows.

Records may arrive in formats that require manual interpretation. Information may be incomplete or inconsistently structured. Clinical systems often lack workflows designed to incorporate data originating outside their organisation.

So, while systems may be technically connected, the real benefits are still limited.

Tracking progress in hospital interoperability.

Efforts are underway to measure progress more systematically. Researchers have recently proposed composite indices to assess how effectively hospitals exchange and use data across systems.

These measures are meant to show progress and highlight differences between organisations as policies change. Most importantly, they show that true interoperability is not just about being connected, but about actually using the shared data well.

Connectivity vs liquidity.

A new concept is now emerging within digital health: data liquidity.

In practical terms, data liquidity refers to the ability for health information to move securely, reliably, and meaningfully across organisations and workflows. Crucially, that data must also be interpretable and actionable without excessive friction.

While connectivity focuses on whether systems can technically exchange data, liquidity focuses on whether that data can be immediately understood and used within clinical workflows. A system may be connected, but if information arrives incomplete, poorly structured, or difficult to interpret, its value remains limited.

When health data becomes liquid rather than static, it enables:

• Real-time clinical decision-making
• Improved care coordination across providers
• Large-scale population health insights
• Structured inputs for AI-driven clinical tools

The difference between just being connected and having true data liquidity might seem small, but it matters a lot. Many healthcare systems can share information, but still struggle to use it in daily operations.

Policy momentum is turning interoperability into infrastructure.

Policies are changing, too. In many places, new rules focus more on standard ways to share data and push back against practices that limit access to health information.

National data exchange systems are being set up, patients expect more access, and interoperability is becoming a basic part of healthcare, not just a goal.

Once something becomes part of the basic infrastructure, the focus shifts. It’s no longer about whether systems can connect, but whether the whole system can work safely, reliably, and transparently at scale.

National frameworks such as TEFCA illustrate how interoperability is evolving into national infrastructure. The ASTP model highlights the foundational layers required to make that infrastructure work in practice.

As these national frameworks evolve, healthcare organisations increasingly rely on data exchange in everyday clinical practice. Governance, operational processes, and accountability must therefore mature alongside the technology enabling the exchange.

Why traditional healthcare IT models struggle with liquidity.

Many healthcare operating models were never designed for a world of fluid data exchange.

Electronic health records were originally developed to replace paper documentation and support billing processes within a single organisation. As a result, their architecture often reflects institutional priorities rather than the needs of a connected ecosystem.

Clinical workflow optimisation and external data integration were frequently secondary considerations rather than foundational design principles.

Today, many health systems have digital records, but still work as if data will mostly stay within their own walls. When data starts to move more freely, the limitations of this approach become clear:

• Information quality issues become harder to ignore.
• Fragmented workflows create operational bottlenecks.
• Governance questions arise around access, responsibility, and trust.

This dynamic suggests that the next phase of digital health transformation will be defined less by technology adoption and more by organisational adaptation.

Designing health systems for a shared data ecosystem.

Health systems that do well with liquid data won’t just be the ones with interoperable platforms.

Instead, the winners will be organisations that redesign clinical workflows, governance models, and incentive structures to operate effectively within a shared information ecosystem.

The benefits are significant:

• Clinicians gain access to a more complete patient history.
• Unnecessary test duplication can be reduced.
• Care teams can coordinate more effectively across settings.
• Population health insights become easier to generate.

At the system level, liquid data also enables policymakers to respond more quickly to emerging public health challenges.

Why AI raises the stakes for interoperability.

Artificial intelligence is amplifying the importance of data liquidity.

AI-driven clinical decision support tools depend on structured, timely, and high-quality data. Fragmented information systems and inconsistent exchange, therefore, do not simply create inefficiencies, but they can directly affect the performance of algorithmic tools designed to support clinical judgement.

Yet even here, the most important questions remain organisational rather than technological.

Health system leaders must ask:

• Are our clinical workflows designed to incorporate external data?
• Do our governance models support cross-organisational data sharing?
• Do financial incentives encourage coordinated care rather than institutional optimisation?

The strategic question for healthcare leaders.

The infrastructure enabling interoperability is already emerging.

Technical standards are becoming more stable. National data exchange systems are being built. Policies are moving toward more openness and accountability in how health data is used.

So, the main challenge is changing.

It is no longer simply a matter of whether interoperability will arrive. Instead, healthcare leaders must consider how their organisations will function once data exchange becomes an expected part of everyday clinical operations.

Systems that treat interoperability as a checkbox may struggle to adapt. Those who view data liquidity as a strategic capability will be better positioned to navigate the transition.

The future of healthcare depends on liquid data.

The next wave of healthcare performance will not be determined solely by who has the most advanced digital platforms.

It will be determined by who can operate confidently in an environment where data moves fluidly across institutions and informs decisions throughout the care continuum.

Which leads to a question worth asking sooner rather than later:

If interoperability became enforceable tomorrow, would your operating model survive?

Unlocking the value of shared health data.

Achieving true data liquidity requires more than technical interoperability. It requires platforms that can connect clinical systems, normalise data across sources, and deliver meaningful insights directly into care workflows.

Orion Health’s Shared Care Record solutions are designed to help healthcare systems move beyond simple connectivity, enabling trusted data sharing across organisations and supporting coordinated, data-driven care.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Abernethy, A., Afsar, N., Anderson, B., Barfield, W., Bharel, M., Brown, J., Embí, P., Eschenlauer, A., Gordon, W., Gregurick, S., James, B., Jena, A., Lee, P., Maddox, T., Mandl, K., Parikh, R., Petersen-Lukenda, L., Sarich, T., Shaikh, A., Speyer, P., & Yale, K. (2026). Toward a national health digital and data architecture: Laying the foundation for digital transformation. National Academy of Medicine.
• Assistant Secretary for Technology Policy. (2026). Data liquidity, affordability, and access: The history and growth of TEFCA. U.S. Department of Health and Human Services.
• Finkley, K. (2026). Advancing interoperability through governance and sustainability: A qualitative comparative case study of state health information exchanges in the United States (Doctoral dissertation, Georgia Southern University).
• Global Digital Health Partnership. (2020). Advancing interoperability together globally: GDHP white paper on interoperability.
• Jordan, P. (2026, March 4). My view – Interoperability & consumer access: Is it time for regulation? Health Informatics New Zealand.
• Lane, P. (2026, March 10). Interoperability’s inflection point. Health Gorilla.
• Office of the National Coordinator for Health Information Technology. (2024). TEFCA awareness and planned participation among U.S. hospitals: 2023 (Health IT Data Brief No. 72).
• Pimenta, N., Chaves, A., Sousa, R., Abelha, A., & Peixoto, H. (2023). Interoperability of clinical data through FHIR: A review. Procedia Computer Science, 220, 856–861.
• Tuan, J. (2026, March 5). TEFCA is live: The practical playbook for getting patient data into your app. Topflight Apps.

EPRs sit at the centre of the government’s strategic shifts: from analogue to digital, from hospital to community, and from sickness to prevention. They digitise hospital records, enable secure information sharing within organisations, and provide the foundation for innovations such as ambient voice technology.

However, widespread coverage does not automatically translate into meaningful impact. The infrastructure is largely in place. The challenge now is delivering measurable returns across organisations, not just within them.

EPR coverage is high, but advanced use remains uneven

While roll-out has accelerated, the depth of use varies significantly.

The Health Foundation points out that while most trusts now have an EPR, not all are using the more advanced features. In the first year of the Digital Maturity Assessment, only a few trusts used tools like integrated prescribing or shared records with other hospitals.

This signals an important shift.

The main question now is not whether digital systems are in place, but whether they are actually helping to improve productivity, safety, and the quality of care.

Two decades of EPR policy: progress, setbacks, and hard lessons

This is not the NHS’s first attempt at digital transformation.

The National Audit Office (NAO) has repeatedly highlighted the complexity of large-scale digital change, shaped by legacy systems, integration challenges, and governance constraints. Earlier national programmes consumed substantial funding but struggled to deliver the full promised benefits.

The lesson is consistent: technology alone does not transform services. Sustainable improvement requires clinical engagement, behaviour change, workforce capability, and continuous optimisation.

The evolution of national ambition over the past two decades illustrates this clearly.

From the £6.2 billion National Programme for IT launched in 2002, which closed in 2011 after spending more than £10 billion, through to successive “paperless” targets in 2015, 2020 and beyond, digital strategy has been repeatedly reset.

By November 2023, 90% of NHS trusts had an EPR. Forecasts in May 2024 projected that 98% would have a suitable EPR by March 2026.

The infrastructure milestone is largely being reached. But history makes one point clear: installation is not transformation.

From digitisation to care improvement: The maturity journey.

To understand the optimisation challenge, it helps to examine how EPR value typically evolves over time.

Most organisations move through four broad stages:

1. Digitising health information – Replacing paper with electronic records.
2. Informing care – Making results, histories, and documentation accessible.
3. Shaping care – Embedding decision support, predictive tools, and operational insight.
4. Improving care quality – Applying advanced analytics and structured quality improvement.

Each stage builds capability. As staff use the system, workflows are refined, coding improves, and data becomes more reliable.

Yet many organisations remain between the first and second stages. The significant productivity gains policymakers seek typically emerge only in stages three and four, when data actively shapes care delivery and system performance.

And this is where a structural barrier becomes clear.

The Optimisation Gap: Where EPRs stop at organisational boundaries

One main reason for the value gap is that most EPR programmes do not go beyond the boundaries of each organisation.

Trusts have made genuine progress in optimising their EPRs locally. Workflows are configured, users trained, templates refined. But those improvements often do not follow the patient when they move between settings.

As a result, clinicians frequently encounter incomplete information. A recent medication change recorded by a GP, an allergy noted in primary care, or a specialist’s report from another trust may not be visible in the local EPR.

The consequence is duplicated effort, time spent chasing information, and unnecessary data re-entry. More importantly, clinical decisions may be made without the full context.

When data stops at organisational boundaries, optimisation stalls. Improving systems locally will not boost overall productivity if information stays scattered.

This “optimisation gap” is explored further in the EPR Network’s latest white paper, featuring contributions from Orion Health, which highlights why adoption, integration and cross-system data sharing are essential to unlocking real productivity gains. Read The Optimisation Gap: Why Having an EPR Isn’t the Same as Using It Well here.

Extending value through shared care records

That is why improvements need to go beyond single organisations.

Shared Care Records (ShCRs) help solve this problem by linking data from hospitals, primary care, community, and mental health services. Importantly, they do this without replacing local EPRs.

Instead of enforcing a single system, Shared Care Records create a longitudinal, cross-organisational view of the patient. They allow clinicians to see critical information regardless of where care was delivered.

The benefits are practical and immediate:

• Reduced duplication of tests and documentation
• Improved patient flow across settings
• Better multidisciplinary coordination
• A foundation for system-wide analytics and population health insight

Shared Care Records turn separate digital systems into a connected network. This means improvements made locally can benefit the whole system.

They also lay the groundwork for a true Single Patient Record, not as a single platform, but as a unified view across the care ecosystem.

The Frontline Productivity Programme: formalising the reset

NHS England’s upcoming Frontline Productivity Programme, launching in April 2026, reflects this strategic pivot.

It succeeds the Frontline Digitisation initiative and aligns with the NHS 10-Year Health Plan and statutory productivity requirements. The focus shifts from acquiring infrastructure to extracting measurable operational value.

Where Frontline Digitisation prioritised EPR deployment in secondary care, Frontline Productivity expands the scope to include:

• EPR optimisation
• Infrastructure and cyber security
• Change management
• Cross-sector systems integration

In other words, the programme acknowledges that productivity gains depend on how well systems work together, not just how well they function individually.

Why optimisation is harder and more important than roll-out

Deployment is a finite project. Optimisation is ongoing.

Trusts must redesign workflows, embed decision support, improve data quality, and build sustainable digital capability. This requires leadership, clinical engagement, and long-term commitment.

International examples show the scale of effort required. NYU Langone introduced its EPR in 2008 and spent 15 years developing advanced, data-driven capabilities, including predictive modelling and real-time dashboards.

An EPR that is live but underused can increase administrative burden rather than reduce it.

Real value depends on:

• First-time-right data entry
• Consistent pathway execution
• Strong clinical ownership
• Effective interoperability across services

Without sharing data between organisations, even the best EPRs cannot deliver all the productivity benefits.

The Era of Delivering Returns

The NHS is entering a more demanding phase of digital maturity.

The focus is shifting from just buying systems to actually getting results from them. Success will not be judged by how many trusts have an EPR. Instead, it will depend on how well these systems, connected across organisations, help frontline staff, improve patient care, and make the system stronger.

Having an EPR is only the beginning. To get the most value, organisations need to share data, coordinate care across different settings, involve clinicians, and treat information as a key resource.

If your organisation is moving from deployment to optimisation, now is the time to focus on interoperability and longitudinal patient views that extend beyond the hospital.

To learn how Orion Health’s Shared Care Record solutions help NHS regions get more value from their EPRs and connect care across boundaries, see our approach here:

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References:

• Cheshire, R. (2026, February 27). From digital acquisition to digital exploitation: Why NHS frontline productivity depends on adoption. FutureScot. 
• Healthcare Innovation Consortium (HIC). (2026). The optimisation gap: Why having an EPR isn’t the same as using it well. Electronic Patient Record (EPR) Network. Available at: https://hicdigital.co.uk/the-optimisation-gap-why-having-an-epr-isnt-the-same-as-using-it-well/ (Accessed: 11 March 2026).
• Lawrence, A., & Krelle, H. (2025, April 19). The NHS must get more out of the EPRs it has purchased. The Health Foundation. 
• Lovell, T. (2026, January 12). Details of NHS digital productivity programme leaked online. Digital Health. 
• National Audit Office. (2020). Digital transformation in the NHS (HC 317, Session 2019–2021). 
• NHS England. (n.d.). Digitising the frontline. 

However, adoption data shows there are still challenges.

Fourteen years after FHIR was introduced, 79% of countries have national implementation guides, but only 20% report using them widely (Firely, 2025). The gap is not purely technical. High coordination costs, loss of semantic meaning, and complex implementation still slow things down. Creating an implementation guide takes time, while real-world workflows keep changing.

Standards perform well in high-volume, repeatable use cases. But they are less flexible for specialised workflows like rare disease registries, complex referrals, or clinical trial matching.

In this setting, a new idea is emerging: Conversational Interoperability (COIN), also known as Language-First Interoperability (LFI).

What is Conversational Interoperability (COIN)?

Conversational Interoperability is an emerging approach in which autonomous agents use natural language to agree on data exchange needs before carrying out those exchanges via structured APIs.

Instead of defining every data element ahead of time, agents figure out the details in real time:

• What information is required?
• In what format should it be provided?
• How to assemble and route it across systems?

Agents can bridge structured FHIR data and unstructured clinical notes, dynamically resolving workflows.

Importantly, COIN does not reject standards. Instead, it offers a new way to use them. Agents may still rely on FHIR APIs, eligibility services and terminology standards once requirements are clarified through conversation.

The framing is important. COIN is not yet an institutional infrastructure. It is exploratory, shaped largely by advances in large language models.

From data exchange to intent fulfilment

Traditional interoperability asks:
Can System A send data to System B?

COIN, on the other hand, asks a different question:
Can System A understand user intent and orchestrate data across systems to fulfil it?

The diagram below presents a conceptual five-layer Clinical Intelligence Network (CIN) model. This framework builds on established interoperability and learning health system architectures, integrating semantic interoperability, data orchestration, governance, and AI-driven reasoning into a unified structure.

The 5 Layers of Clinical Intelligence Networks (CIN)

At its foundation sit data orchestration and semantic interoperability layers, enabled by standards such as FHIR, SNOMED and secure APIs. Above these are natural language understanding, governance and security controls, and reasoning engines that support decision-making and care pathways.

In this model, Conversational Interoperability functions as the interface, while Clinical Intelligence Networks represent the governed infrastructure that enables it.

This distinction is important. Without strong data foundations, identity assurance and audit mechanisms, conversational capability risks becoming surface-level automation rather than a trusted clinical infrastructure.

The question then becomes how this conceptual architecture performs when tested in practice.

Early experimentation: HL7 Connectathon demonstrations

At a recent HL7 Connectathon in the US, a small group implemented and tested COIN scenarios across multiple use cases.

Demonstrations included specialist referrals, clinical trial matching, rare disease registry reporting, prior authorisation simulations, and guideline-driven decision support.

There was also a demonstration of a cardiology referral agent that verified providers, checked insurance and scheduled appointments within a live EHR environment. When a structured JSON payload failed to parse, the agents went further and negotiated an alternative format, thus completing the workflow.

Other teams demonstrated bidirectional trial matching, in which agents acted as both client and server to dynamically negotiate eligibility criteria. The breadth of use cases prompted one participant to describe the moment as the birth of a new interoperability paradigm.

A first-of-its-kind pilot in the United States

One early example shows both promise and risk.

Utah has launched a pilot program that allows an AI system to autonomously renew certain prescription medications without a doctor’s direct involvement.

As shown above, the Utah pilot was intentionally designed with clear clinical guardrails.

The AI system is authorised to renew prescriptions only for a defined list of approximately 190 existing chronic medications. High-risk and controlled medicines are explicitly excluded from the programme.

If a renewal request is routine and no clinical red flags are detected, the AI can issue the refill directly to the pharmacy. However, if uncertainty arises or potential risk factors are identified, the case is escalated to a human clinician for review.

This structured escalation pathway keeps the system within a tightly bounded scope, positioning it as controlled automation rather than unrestricted autonomy.

Vendor data from 500 external cases indicated that AI treatment plans matched those of human clinicians in approximately 99.2% of cases (Politico, 2026).

If models like this incorporated conversational negotiation across systems, the shift would move from decision support towards partial clinical execution. That would materially increase regulatory, safety and liability exposure.

These risks are real, not just theoretical. Research has already shown safety problems with conversational medical assistants when there are not enough safeguards (Bickmore et al., 2018). The WHO also stresses the need for strong governance and ethical oversight in AI for health (WHO, 2021).

Does conversational negotiation re-create integration complexity?

A common objection is that conversational negotiation risks reintroducing point-to-point complexity.

However, supporters argue that each agent represents a reusable capability. In theory, agents can converse with multiple complementary agents, enabling more linear scaling rather than quadratic integration growth.

If traditional standards solve the n² problem structurally, COIN may attempt to address it behaviourally.

That remains an open question.

Governance: The deciding factor

Technical feasibility alone will not determine whether COIN matures beyond experimentation.

Issues such as identity verification, authentication, consent, and auditability remain unresolved in this model. Mark Kramer has proposed adapting mandate-based delegation and cryptographically verifiable consent models to manage agent authority.

There is a clear tension here:

• Natural language negotiation introduces interpretive flexibility
• Governance demands deterministic boundaries

It is still unclear whether these two needs can work together at a large scale.

Framing COIN as a replacement for FHIR misunderstands the proposition.  Right now, conversational interoperability is viewed as something that works alongside existing standards. Standards lay the groundwork, and if COIN develops further, it could help with negotiation on top of those standards.

Strategic implications for health platforms

If conversational agents become a primary integration surface, differentiation may shift.

Platforms may compete not only on API maturity, but on:

• Agent intelligence and safety
• Identity assurance mechanisms
• Governance frameworks
• Traceability and audit controls

For enterprises, three practical questions emerge:

1. Where are coordination costs highest today?
2. Which workflows sit in the long tail where standards adoption is weakest?
3. Could conversational negotiation reduce integration friction without compromising compliance?

Administrative tasks and referral management could be good, lower-risk areas to test these ideas.

An emerging layer, not yet infrastructure

COIN, or Language-First Interoperability, appears to be in its formative phase. Conceptually, it emerged post-LLM advancements between 2022 and 2024. It is not widely deployed. Governance models are evolving. Evidence is early.

There are some demonstrations showing COIN can work, but there is no agreement yet among institutions.

The next layer of healthcare interoperability might not be a new technical standard. Instead, it could involve systems that agree on meaning before sharing structured data.

Whether this approach becomes part of the core infrastructure or stays experimental will depend more on trust, governance, and clinical responsibility than on technical ability.

This was a case for an emerging paradigm. This article presents early views on the technology and should not be relied on as legal or clinical advice. Readers should verify key claims with primary sources before acting

Explore the future of interoperability.

Interoperability is evolving. Whether through structured standards, intelligence platforms or emerging conversational models, the objective remains the same: coordinated, safe and scalable care.

See how Orion Health is creating secure, standards-based data foundations that are ready for the next wave of innovation.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Achiam, J., Adler, S., Agarwal, S., Ahmad, L., Akkaya, I., Aleman, F. L., et al. (2023). GPT-4 technical report. arXiv.
• Bickmore, T. W., Trinh, H., Olafsson, S., O’Leary, T. K., Asadi, R., Rickles, N. M., & Cruz, R. (2018). Patient and consumer safety risks when using conversational assistants for medical information: An observational study of Siri, Alexa, and Google Assistant. Journal of Medical Internet Research, 20(9), e11510.
• Firely. (2025). FHIR adoption survey 2025.
• Gosmar, D., Dahl, D. A., & Coin, E. (2024). Conversational AI multi agent interoperability, universal open APIs for agentic natural language multimodal communications. arXiv.
• HL7 International. (n.d.). FHIR overview. https://www.hl7.org/fhir/overview.html
• Kramer, M. (n.d.). Why conversational interoperability is essential for the future of healthcare. LinkedIn. https://www.linkedin.com/pulse/why-conversational-interoperability-essential-future-mark-kramer-retre/
• Mandel, J. C. (n.d.). Conversational interoperability takes shape: Read out from HL7. LinkedIn. https://www.linkedin.com/pulse/conversational-interoperability-takes-shape-read-out-from-mandel-md-379fc/
• Office of the National Coordinator for Health Information Technology. (2020). 21st Century Cures Act: Interoperability, information blocking, and the ONC Health IT Certification Program. Federal Register, 85, 25642–25961.
• Politico. (2026, January 6). Artificial intelligence is prescribing medications in Utah. https://www.politico.com/news/2026/01/06/artificial-intelligence-prescribing-medications-utah-00709122
• World Health Organization. (2021). Ethics and governance of artificial intelligence for health: WHO guidance.. “Germans Trust ‘Dr AI’ More Than Doctors Who Use It.” Medscape, February 18, 2026

In Germany, 45% of people say they consult an AI chatbot before seeking medical advice. 71% percent of people view AI in medicine positively. More than half say they understand chatbot responses better than conventional internet searches (Beneker, 2026).

On the surface, that looks like a vote of confidence in “Dr AI”.

And yet, when researchers tested how people perceive physicians who disclose they use AI, the results were sobering. In a large US study, doctors who described using AI were rated as less competent, less trustworthy, and less empathetic than those who did not mention it (Reis et al., 2025). Willingness to book an appointment fell when AI was explicitly referenced.

Patients seem ready to trust the machine itself, but they are cautious about doctors who use it.

AI adoption in healthcare is accelerating, but trust is lagging.

Health systems are embedding AI at pace.

In the United States, over a thousand AI-enabled medical devices have been cleared for clinical use. Across South East Asia, AI tools now power symptom checkers, mental health apps and clinical triage systems.

Yet trust is not keeping up with adoption.

In Singapore, 80% of residents report using AI, but trust drops sharply in sensitive domains such as mental health. Globally, nearly 60% of Americans report feeling uneasy about AI-assisted diagnosis (World Economic Forum, 2025).

This isn’t primarily a technical problem.

It’s a relational one.

AI risk and real-world incidents: what the data shows

Before we talk about trust, we need to talk about risk. The OECD’s AI Incidents and Hazards Monitor (AIM) tracks reported AI-related harms across sectors.

As shown in the graph, incidents have steadily increased across multiple industries, including healthcare, financial services and digital security.

Healthcare, drugs, and biotech are often mentioned in reported incidents. This does not always mean AI is unsafe. However, as more people use AI, the chances of risk and failure become more visible.

Trust cannot grow in the absence of accountability.

Industry differences in AI use and organisational support

AI uptake also varies significantly by sector.

Information Technology leads in both AI use (85%) and organisational support (76%). Healthcare and social assistance sit much lower on the scale, with 45% reporting AI use at work and 58% reporting organisational support.

This difference is important.

Healthcare is not simply another digital industry. It operates in high-stakes environments where human lives, ethical judgement and relational continuity are central.

Lower adoption rates may show caution instead of resistance.

Responsible AI and national readiness: Does governance build trust?

Trust in healthcare has never depended on patients’ understanding every mechanism at play.

As highlighted in the npj Health Systems perspective, trust involves vulnerability and uncertainty. Patients do not need to understand how paracetamol works at a molecular level to trust it. They rely on institutions, regulation, professional standards and the clinician who prescribes it.

AI changes how that trust works.

Many models work in ways that are hard to understand. Sometimes, explanations are only rough estimates instead of being truly clear. In one study about explainable AI in obstetric ultrasound, giving explanations did not always build trust, and sometimes even lowered performance. People felt more confident, but did not always rely on the system in the right way.

Governance, therefore, becomes critical.

The GIRAI ranks countries on responsible AI governance, human rights protections and institutional capability. The Netherlands, Germany and Ireland lead the index, while Australia and Canada score lower on specific governance dimensions.

The message is clear: responsible AI is not only about technical skills. It is also about having the right frameworks, taking action, and respecting human rights.

New Zealand released its first national AI strategy, titled Investing with Confidence, in July 2025. However, we do not yet have a dedicated national health AI strategy. Health AI policy is being developed through broader AI and health system programmes.

If trust is the foundation of healthcare, then the way we organise and communicate our approach to AI is just as important as the technology itself.

The clinician burden: when AI misfires

Trust is also shaped at the bedside.

A Scientific American report describes nurses pressured to act on sepsis alerts generated by algorithms that misfired, with overrides and accountability resting squarely on the human clinician.

When alerts cannot explain themselves, the clinician still carries the risk.

Trust in AI cannot come at the expense of trust in clinicians.

Why digital trust determines healthcare effectiveness

The World Economic Forum says that digital trust is needed for healthcare to work well. If patients feel uneasy, they may stop engaging, even if the care is accurate.

The New Zealand Medical Journal reinforces this: maintaining patient trust requires transparent governance, robust approvals and equitable deployment.

Overlay geopolitics, and the picture becomes more complex. The 2025 Government AI Readiness Index frames AI as a national capability, measured in compute power, infrastructure and policy maturity.

But healthcare is where the “people-first” principle is tested most strongly.

If AI is introduced primarily as a cost-saving device or as a way to reduce human interaction, patients will feel that.

If AI is presented as a way to support clinical judgement, make reasoning clearer, and give more time for conversation, it is much more likely to be accepted.

So, who do we trust more: AI or our doctor?

Most people do not want to choose. They want a doctor who uses good tools wisely.

They want assurance that AI systems are rigorously validated and equitably designed. They want clarity about accountability if something goes wrong. And they want to know that AI remains visibly subordinate to clinical judgement.

AI does not need to be loved. It needs to be demonstrably safe, responsibly governed and clearly aligned with patient wellbeing.

Trust will not be built through marketing claims or performance metrics alone. It will be built into consultation rooms, into governance frameworks, and into the everyday decisions of clinicians who know when to follow the prompt and when to override it.

In the end, the currency of healthcare is not data. It is trust.

And if we get that wrong, no amount of algorithmic accuracy will compensate.

Building trustworthy AI-enabled health systems

At Orion Health, we believe AI should be built on strong data, clear governance, and workflows led by clinicians, not take their place.

If you’re exploring how to embed AI safely within interoperable, standards-based health data platforms, learn more about our Amadeus AI platform.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Dobson, Rosie, Melanie Stowell, and Robyn Whittaker. “Maintaining Patient Trust as Artificial Intelligence’s Role in Healthcare Grows.” New Zealand Medical Journal 139, no. 1629 (February 13, 2026): 94–101. 
• European Commission. Regulation (EU) 2024/1689 of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act). 2024. 
• Hoffman, Karen M., et al. “Racial Bias in Pain Assessment and Treatment Recommendations, and False Beliefs about Biological Differences between Blacks and Whites.” Proceedings of the National Academy of Sciences 113, no. 16 (2016): 4296–4301.
• Nicolson, Angus, Elizabeth Bradburn, Yarin Gal, Aris T. Papageorghiou, and J. Alison Noble. “The Human Factor in Explainable Artificial Intelligence: Clinician Variability in Trust, Reliance, and Performance.” npj Digital Medicine 8 (2025): 658
• Obermeyer, Ziad, Brian Powers, Christine Vogeli, and Sendhil Mullainathan. “Dissecting Racial Bias in an Algorithm Used to Manage the Health of Populations.” Science 366, no. 6464 (2019): 447–453.
• Oxford Insights. Government AI Readiness Index 2025. Oxford: Oxford Insights, January 2026. 
• Reis, Moritz, Florian Reis, and Wilfried Kunde. “Public Perception of Physicians Who Use Artificial Intelligence.” JAMA Network Open 8, no. 7 (July 17, 2025): e2521643
• Sagona, Madeline, Tinglong Dai, Mario Macis, and Michael Darden. “Trust in AI-Assisted Health Systems and AI’s Trust in Humans.” npj Health Systems 2 (2025): 10. 
• Schellmann, Hilke. “AI Enters the Exam Room: When Alerts Misfire or Can’t Explain Themselves, Nurses Still Carry the Risk.” Scientific American, February 17, 2026
• World Economic Forum. “The Trust Gap: Why AI in Healthcare Must Feel Safe, Not Just Be Built Safe.” December 5, 2025
• Beneker, Christian. “Germans Trust ‘Dr AI’ More Than Doctors Who Use It.” Medscape, February 18, 2026

Rather than applying a one-size-fits-all model, PHM enables primary care to deliver targeted, equitable interventions. It shifts the system from passive, reactive treatment to proactive and anticipatory care.

On the ground, this means identifying those at highest risk, supporting self-management for people with chronic conditions, and ensuring complex patients receive coordinated case management. When done well, PHM strengthens both equity and system sustainability.

How the Kaiser pyramid illustrates risk-based care

A practical way to visualise PHM in action is through the Kaiser Pyramid. It shows how services can be balanced based on patient risk and complexity.

The pyramid shows:

• A small proportion of patients with complex needs require intensive case management.
• A larger proportion benefits from disease management and structured professional support.
• The majority are supported through health promotion, prevention and self-care.

Balancing self-care and professional care helps keep the system sustainable. When risk stratification and segmentation are applied effectively, resources are directed to those who need them most, reducing inefficient utilisation and avoidable emergency department visits.

Why continuity of care improves outcomes

Continuity of care is more than just a preference. It is closely linked to better health outcomes.

A systematic review of quantitative studies covering more than 15 million patients found that higher personal continuity between patients and general practitioner (GP) probably:

• Reduces premature mortality by 10–15%
• Lowers hospital admissions by 10–15%
• Reduces emergency department visits by 10–20%

These findings are supported by moderate certainty of evidence (Engström et al., 2025). Even small gains in personal continuity can significantly reduce healthcare use and improve access.

In general practice, continuity means patients feel their care matches their needs each time they visit. It is widely regarded as a key component of high-quality primary care.

Continuity builds trust and helps doctors get to know patients better, so care can be adjusted over time. But in today’s group practices, continuity does not always happen automatically.

Studies from Sweden and other global health systems show that simply registering with a GP does not always lead to better continuity of care at subsequent visits (Janlöv, Blume, & Glenngård, 2023). Systems need to support real relationships, not just assign names.

Primary care networks and the limits of structure without support

In the UK, Primary Care Networks (PCNs) were introduced to better integrate health and social care and to make primary care more sustainable. Early progress included governance arrangements and expanded multidisciplinary roles, such as pharmacy and social prescribing.

However, evaluations indicate:

• Limited management funding
• Heavy reliance on clinical directors
• Restricted backfill support

Without adequate infrastructure, expectations exceed capacity. For PCNs to thrive within integrated care systems, additional expertise in population health analysis and organisational development is required.

Changing structures is not enough. PCNs also need enough resources, skills, and the right incentives.

The NHS neighbourhood health vision and place-based integration

The NHS neighbourhood health service vision highlights the value of local, integrated care models.

Neighbourhood health brings together preventive and personalised care delivered closer to home. It is shaped by community voices and organised to improve outcomes while reducing inequalities.

This approach aligns directly with PHM principles:

• Defined local populations
• Proactive risk identification
• Integrated services across health and social care
• A focus on prevention and equity

When neighbourhood-level integration is combined with data-driven population insight, systems can move from fragmented service delivery to coordinated, anticipatory care.

Aligning financial incentives with total population health

Structural change requires financial alignment. It has been argued that meaningful population health improvement will not occur until incentives are explicitly designed to support it.

Today’s biomedical model mostly rewards treating illness, rather than prevention and equity. A new system should encourage overall population health, including:

• Prevention
• Social determinants of health
• Equity-focused resource allocation

Ambitious “Big Hairy Audacious Goals” (BHAGs) have been proposed to focus attention on life expectancy, social spending, and broader determinants of wellbeing. These system-level goals aim to encourage shared accountability across sectors.

Population health management can support this shift. Through risk stratification and segmentation, PHM enables more informed and equity-sensitive resource allocation. This ensures that those with the greatest need receive appropriate support, while reducing avoidable demand on acute services.

Building a sustainable health system through integration

Health systems are unlikely to become sustainable through one-off actions.

It requires a coherent formula that integrates:

• Personal responsibility, supported by health literacy and self-management
• Continuity of care, embedded in relational and organisational structures
• Proactive population health management, addressing risk and social determinants
• Funding models that reward improved outcomes rather than activity alone

The evidence suggests that continuity improves outcomes and that population-based planning strengthens equity and efficiency. However, these components depend on aligned incentives and adequate infrastructure.

A sustainable health system needs more than just personal responsibility or funding changes. It emerges from integrating all three: personal responsibility, funding, and system design, while focusing on equity, prevention, and shared responsibility.

Turning population insight into action

Population health management is not just a policy ambition. It requires the ability to define populations, stratify risk, integrate data across care settings, and translate insight into coordinated action at the neighbourhood level.

Orion Health’s population health solutions enable health systems to connect data, identify risk earlier, and support proactive, integrated models of care. Explore how Orion Health can support your population health strategy and neighbourhood-level integration.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Anell, Anders, et al. 2023. Sweden: Health System Review 2023. Copenhagen: WHO Regional Office for Europe.
• Barker, I., A. Steventon, and S. R. Deeny. 2017. “Association between Continuity of Care in General Practice and Hospital Admissions for Ambulatory Care Sensitive Conditions.” BMJ 356: j84.
• Ellegård, Lina Maria, Anders Anell, and Gustav Kjellsson. 2024. “Enabling Patient–Physician Continuity in Swedish Primary Care: The Importance of a Named GP.” BJGP Open.
• Engström, Sven Göran, Malin André, Eva Arvidsson, Carl Johan Östgren, Margareta Troein, and Lars Borgquist. 2025. “Personal GP Continuity Improves Healthcare Outcomes in Primary Care Populations: A Systematic Review.” British Journal of General Practice.
• Hughes-Cromwick, Paul, and Sanne J. Magnan. 2024. “BHAGs for Aligning Incentives and Building a Learning System to Improve Total Population Health.” American Journal of Managed Care 30 (Spec. No. 13): SP1013–SP1023.
• Janlöv, Nina, Sofia Blume, and Anna Glenngård. 2023. Sweden: Health System Review. Copenhagen: WHO Regional Office for Europe.
• Maddox, Raglan, et al. 2023. “Ethical Publishing in Indigenous Contexts.” Tobacco Control.
• Smith, Judith A., Katherine Checkland, Manbinder Sidhu, Jonathan Hammond, and Sarah Parkinson. 2021. “Primary Care Networks: Are They Fit for the Future?” British Journal of General Practice 71 (714): 106–107.
• Thomas, Samantha, Joel Francis, Marita Hennessy, Kate Frazer, Charlotte Godziewski, Caitlin Douglass, Orkan Okan, and Mike Daube. 2024. “The Year in Review—Health Promotion International 2023.” Health Promotion International 39.
• World Health Organization Regional Office for Europe. 2023. Population Health Management in Primary Health Care: A Proactive Approach to Improve Health and Well-Being. Copenhagen: WHO Regional Office for Europe.

But in many health systems, acute services now carry a far heavier and more complex load than they were ever designed for. Emergency departments and hospitals have become the default destination not only for emergencies, but for unmet needs elsewhere in the system.

How acute care became the safety net for the health system

EDs are increasingly seeing cases of chronic disease exacerbations, mental health crises, medication issues, social instability and lack of access to timely primary or community care. This is not because acute care is the best place to manage these needs, but because it is often the only part of the system that is always available.

Research into acute, unscheduled care highlights that demand doesn’t start at the hospital door.  Social and individual factors, such as poverty, housing instability, health literacy, ageing populations, and multimorbidity, play a significant role in when and how people seek care.

When community resources are fragmented or difficult to navigate, the ED becomes the safety net by default.

The gap between short-term care and ongoing needs

This creates a fundamental mismatch. Acute services are optimised for episodic intervention, yet much of what now flows through them requires continuity.

Patients with chronic conditions often experience repeated acute episodes, not because their condition is unpredictable, but because care is poorly connected across settings. Discharge without effective follow-up, limited information sharing between providers, and weak transitions between hospital and community care all increase the likelihood of return visits. As a result, the acute system ends up managing recurrence rather than resolution.

What international comparisons show about health system design

International comparisons reinforce this pattern. Health systems with stronger primary care access and better service integration tend to see lower emergency department utilisation for lower-acuity needs, alongside more appropriate hospital admissions.

Where access to general practice or community services is constrained, acute activity rises, often without corresponding improvements in outcomes. The result is sustained pressure on capacity, staff burnout, and a poor patient experience as patients navigate a system that treats symptoms rather than causes.

Why acute care itself isn’t the problem

The problem isn’t that acute care is failing. In fact, it does its intended job very well.

The real issue is that acute care is being used to make up for gaps in other parts of the system. Over time, this changes both what is needed and how care is delivered. Focusing only on quick discharges or avoiding admissions can make things worse by valuing speed over coordination. Letting patients go quickly without solving their deeper health or social needs just moves the problem forward.

Rebalancing the system instead of stretching acute care further

The evidence clearly shows we need to rebalance the system, not keep stretching acute care.

Integrated care models emphasise teamwork and shared responsibility across different settings. Chronic care models and community coordination try to prevent avoidable acute episodes by meeting needs earlier and more completely. When patients get help managing their conditions over time, acute care can go back to its main job: handling crises.

Acute care as one part of a larger health system

This doesn’t make acute care less important. In fact, it highlights its value.

By seeing that emergency departments and hospitals are just one part of a bigger system, not the center, health systems can use acute care where it helps most. This means building up primary and community care, improving how information is shared, and making sure transitions treat an acute episode as part of a longer care journey, not just a one-time event.

Acute care will always be essential. But if it keeps carrying the burden of system gaps, it will stay under pressure, and patients will keep coming back when they don’t need to.

The real opportunity lies not in asking acute services to do more, but in enabling the rest of the system to do its part.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• American College of Emergency Physicians. 2017. The Acute Unscheduled Care Model: Enhancing Appropriate Admissions. Irving, TX: American College of Emergency Physicians.
• Duffy, Juliana, Peter Jones, Candace D. McNaughton, Vicki Ling, John Matelski, Renee Y. Hsia, Bruce Landon, and Peter Cram. 2023. “Emergency Department Utilization, Admissions, and Revisits in the United States, Canada, and New Zealand: A Retrospective Cross Sectional Analysis.” Academic Emergency Medicine 30 (9): 946–954.
• Global Burden of Disease 2019 Acute and Chronic Care Collaborators. 2025. “Characterising Acute and Chronic Care Needs: Insights from the Global Burden of Disease Study 2019.” Nature Communications 16: 4235.
• Philips Healthcare Transformation Services. 2016. Acute Unscheduled Care in Seven Developed Nations: A Cross Country Comparison. Amsterdam: Philips.
• World Health Organization Regional Office for Europe. 2016. Integrated Care Models: An Overview. Copenhagen: WHO Regional Office for Europe.
• World Health Organization Regional Office for Europe. 2016. Integrated Care Models: An Overview. Health Services Delivery Programme. Copenhagen: WHO.

In this video, Attendees at HINZ Digital Health Week 2025 reflect on where AI genuinely adds value in healthcare today, the risks it poses, and how health systems can take a pragmatic, people-first approach to adoption.

“I think AI in healthcare does have a lot of hype, as everywhere, but it also has a lot of promise and it also has a lot of risk.” – Ray Delany, Founder, CIO Studio

Understanding the risks of AI in healthcare

One of the biggest challenges with AI in healthcare is that the risks are still evolving. Unlike more established digital tools, AI systems can behave in unexpected ways, especially as they become more autonomous and adaptive.

“Part of the problem with AI, I think, is that we don’t really know what the risks are.” – Ray Delany, Founder, CIO Studio

This uncertainty makes careful decision-making essential. AI should never be adopted simply because it is new or fashionable; it must be the right solution for a clearly defined problem.

“We’ve got to be very careful going down the AI track and choosing it to make sure that it is the right solution for the problem.” – Gillian Robinson-Gibb, Founder, CEO at Hercules Health

Guidelines for responsible AI adoption in healthcare

Rather than blocking innovation outright, healthcare organisations need frameworks that guide responsible use. Clear guidelines help teams innovate safely, while still allowing flexibility as technology and understanding mature.

“We need guidelines instead of guardrails so that we’re making sure that we are guiding people through the journey and not putting big massive blockers in the way.” – Rowena Woolgar, Director at Arise Consulting.

This approach supports progress without losing sight of safety, ethics, and trust, all of which are critical in healthcare environments.

How AI can reduce administrative burden for clinicians

Where AI shows immediate promise is not in replacing clinicians, but in supporting them. Administrative burden remains one of the biggest contributors to burnout, and AI can help remove friction from everyday tasks.

“Where I see the opportunity is to focus on how we can free up clinician time to focus on the thing that they’re really good at, which is the soft stuff with patients. So, a mass of opportunity in terms of automating and lowering that kind of administration barrier.” – Harry Hawke, CEO, Managing Director at Webtools.

By starting with administrative and shared services functions, organisations can build confidence in AI before moving it closer to frontline care.

“If we’re getting good digital experiences using AI first with our admin and our shared services departments, then we’re building confidence with AI to then enable it to be in the frontline and actually interact with patients.” – Heather Phillips, Head of Corporate Affairs at Awanui Group.

Separating AI hype from real-world healthcare impact

Despite bold claims, AI has not yet delivered widespread, proven improvements in clinical outcomes. That doesn’t mean it won’t,  but it does mean expectations must be grounded in evidence.

“I think that AI has huge opportunities in some areas, particularly elements of back office work…But I think the flip side is that it’s going to be some incredible revolution. I haven’t quite seen the evidence that many of these tools are impacting clinical outcomes.” –  Dr Jono Hoogerbrug, General Practitioner and Clinical Informatics Director at Health New Zealand.

Healthcare leaders must balance optimism with realism, focusing on measurable value rather than hype.

The future of AI in healthcare: agentic AI and leadership

The next wave of AI, including agentic AI, could fundamentally reshape how healthcare systems operate. While the future isn’t fully clear, the scale of potential change is undeniable.

“Agentic AI is real, has a massive opportunity to completely change what the healthcare system is. I don’t know what that looks like. It’s both scary and exciting, and with the right leadership and the right guardrails, I think it can fundamentally change how Healthcare is delivered.” – Stella Ward, CEO at Digital Health Association (DHA)

A pragmatic path forward for AI in healthcare

AI is neither a silver bullet nor something to fear outright. Its success in healthcare will depend on strong leadership, thoughtful governance, and a relentless focus on people, both clinicians and patients. By starting where AI can deliver real, practical benefits today, particularly in administrative workflows and shared services, healthcare organisations can build trust, capability, and momentum for what comes next.

At Orion Health, we believe the future of AI in healthcare depends on strong leadership, open standards, and digital foundations that put people first. Learn more about how we support responsible, scalable innovation across health systems.

Watch What’s next for digital health in Aotearoa? to hear more perspectives from HiNZ 2025 on where digital health is heading and what health systems should be preparing for now.

While long-term interoperability initiatives such as APIs and network participation continue to evolve, public health teams cannot afford to wait. They need solutions that work securely and at scale across a wide range of providers and systems.

Direct Secure Messaging offers a practical, standards-based way to meet immediate public health reporting needs while supporting broader modernization goals.

What is Direct Secure Messaging?  

Direct Secure Messaging (DSM) is a standards-based method for securely exchanging clinical information between healthcare organizations. It enables encrypted, auditable messaging and document exchange between verified healthcare endpoints, ensuring sensitive data reaches the right recipient reliably.

For public health agencies and their networks, DSM offers a ready-now path to improve the timeliness, reliability, and reach of reporting workflows, ensuring data moves efficiently from the point of care to population health systems.

How public health agencies can use Direct Secure Messaging today.

Today, many organizations use Direct Secure Messaging to enable the timely submission of clinical information from providers to public health entities without requiring complex system integrations or real-time data queries.

Public health agencies commonly use Direct Secure Messaging to:

• Receive clinical reports and documents directly from providers.
• Support the timely submission of public health data without manual uploads.
• Enable communication with providers across different EHR systems and technical capabilities.
• Extend connectivity to smaller, rural, or resource-constrained organizations.

In addition, directory-based discovery also helps public health agencies identify and manage trusted Direct addresses, making it easier to communicate with the right providers across different organizations and technical environments.

In practice, DSM enables public health agencies to make immediate improvements to information exchange while remaining flexible as interoperability initiatives evolve.

Why Direct Secure Messaging works for public health reporting.

Public health reporting depends on timely, complete, and trustworthy information, especially when data must move across many organizations. Direct Secure Messaging works for public health reporting by enabling secure, auditable, and consistent information exchange across diverse providers and environments without requiring major system changes.

DSM is effective because it:

• Supports secure, standards-based exchange without requiring major system changes.
• Enables reliable communication across organizations of different sizes and capabilities.
• Provides clear visibility into message delivery and accountability.
• Reduces reliance on manual processes that introduce delay and risk.

For agencies managing reporting across hundreds or thousands of providers, this consistency is critical. DSM ensures information reliably moves from the point of care to public health systems, even across different technologies.

Direct Secure Messaging compared to other interoperability approaches.

Direct Secure Messaging is often discussed alongside APIs, health information networks, and emerging interoperability frameworks, which can create confusion about where it fits. Rather than replacing these approaches, DSM complements them by enabling trusted organizations to quickly send clinical information from one organization to another.

Many modern interoperability initiatives focus on query-based exchange, enabling systems to request and retrieve data on demand. While essential for long-term data access and analytics, these approaches often require significant technical investment, governance alignment, and time to implement across diverse provider environments. As a result, they may not always meet immediate reporting or coordination needs.

Unlike APIs and health information networks, which focus on query-based data access, Direct Secure Messaging supports push-based exchange, making it well suited for time-sensitive public health reporting without waiting for complex system integrations.

The table below clarifies how Direct Secure Messaging fits alongside other common interoperability approaches.

How Direct Secure Messaging bridges today’s public health needs and future interoperability.

As public health organizations work toward more advanced interoperability models, many face the challenge of meeting immediate communication needs while planning for future systems. Direct Secure Messaging provides a practical bridge by enabling secure, standards-based healthcare messaging that works alongside other technologies rather than competing with them.

By supporting secure healthcare communication today, DSM allows agencies to improve reporting and coordination without waiting for full-scale system transformations. This approach helps public health organizations make steady, meaningful progress by using direct messaging healthcare workflows to address current needs while building toward more connected, modern interoperability environments over time.

What to consider when adopting Direct Secure Messaging.

When adopting Direct Secure Messaging, public health agencies should begin by considering how DSM will fit within existing reporting and communication workflows.  

Key considerations include:

• Workflow alignment: Identifying how messages are sent, received, routed, and tracked across teams
• Governance and trust: Managing identities, verified endpoints, and access policies
• Operational visibility: Ensuring staff can monitor message delivery and activity
• Scalability: Supporting growing participation and evolving reporting requirements

Thinking ahead about how DSM aligns with these considerations can help public health organizations meet immediate needs while planning for growth.  

What to look for when choosing a Direct Secure Messaging platform.

When evaluating a Direct Secure Messaging platform, public health agencies should look for solutions that support secure healthcare communication without adding unnecessary complexity.

Important capabilities to look out for include:

• Standards-based, secure direct messaging that supports encrypted, auditable exchange.
• Support for verified healthcare identities and trusted networks, including participation in recognized trust frameworks (such as DirectTrust) where required.
• Clear visibility into message delivery, status, and audit trails to support accountability and compliance.
• Ability to exchange information with a broad range of healthcare partners, regardless of EHR or technical maturity.
• Efficient document handling, including one-click, in-platform viewing of common clinical formats such as C-CDA.
• Alignment with broader interoperability strategies, ensuring DSM complements APIs, networks, and future data exchange initiatives.

Platforms built for public health and network-scale exchange are better positioned to meet today’s reporting needs while supporting long-term modernization.

Orion Health Communicate is designed to support secure, standards‑based Direct Secure Messaging for healthcare organizations, enabling reliable communication across providers, public health agencies, and networks.

Moving forward with Direct Secure Messaging.

As public health agencies continue to modernize how information is shared and acted upon, Direct Secure Messaging offers a practical way to make progress. By supporting secure exchange across a wide range of organizations, DSM helps agencies strengthen reporting and coordination today while maintaining flexibility to adapt while interoperability technology evolves.

Direct Secure Messaging is particularly effective when public health agencies need to receive clinical information quickly, securely, and at scale, such as reports, documents, or files, without waiting for complex system integrations to be established.

Understanding where Direct Secure Messaging fits within a broader interoperability strategy allows public health leaders to take an effective approach to modernization. Rather than viewing DSM as a standalone solution, agencies can use it as part of a layered exchange model that addresses immediate needs and supports long-term goals. In doing so, DSM becomes a foundation for more connected, responsive public health systems across the nation.

Learn more about how Orion Health Communicate, available now on AWS Marketplace, supports Direct Secure Messaging for public health reporting and secure healthcare communication.

The evidence is increasingly clear: AI will not succeed in healthcare unless it is deeply and deliberately integrated into clinical practice.

The translational gap between performance and practice

A consistent theme across the literature is the gap between technical success and real-world clinical impact. Many AI models perform well on curated datasets or narrowly defined tasks, but their effectiveness drops sharply when deployed in live clinical environments.

Real-world care is messy. Patient populations are heterogeneous, data is incomplete, workflows evolve, and clinicians operate under constant time pressure. This translational gap is not simply a technical issue. It reflects a deeper misalignment between how AI systems are designed and how healthcare is actually delivered.

Why clinical decision making cannot be automated away

Clinical decision making (CDM) refers to the cognitive, professional, and contextual processes clinicians use to assess information, weigh risks and benefits, and make care decisions for individual patients. It combines clinical evidence, patient data, professional expertise, ethical judgement, and patient preferences to determine the most appropriate course of action in a given situation.

Importantly, CDM is not a linear or purely data-driven process. It is shaped by uncertainty, incomplete information, time pressure, and the realities of real-world care delivery. Clinicians continuously interpret signals, apply contextual judgement and professional experience, and adjust decisions based on evolving patient conditions and system constraints.

Evidence-based practice itself relies on integrating research evidence with clinical expertise. AI tools that attempt to automate decisions without supporting clinical reasoning often disrupt care rather than enhance it. Qualitative studies consistently show that such systems are abandoned when they fail to align with established routines or threaten professional accountability.

Standalone AI tools rarely scale in healthcare.

Another recurring finding is that standalone AI tools struggle to scale or sustain themselves in practice. Poor interoperability, particularly limited integration with electronic health record systems, creates friction and adds to clinician workload.

Lack of transparency and explainability undermines trust, especially in high-stakes decisions where responsibility ultimately remains with human professionals. Concerns about algorithmic bias, data quality, and patient privacy further complicate adoption, particularly when training data does not reflect real clinical populations.

What successful clinical AI looks like in practice.

Where AI adoption has been more successful, common characteristics emerge. These systems are designed as decision support tools rather than autonomous decision-makers.

They are embedded directly into clinical workflows, delivering relevant insights at the point of care. They are co-developed by clinicians, data scientists, and healthcare organisations, and evaluated not only on technical accuracy but on patient outcomes and quality of care.

Crucially, they are monitored and adapted over time, recognising that clinical environments and patient populations are dynamic.

A sociotechnical approach to AI in healthcare.

Recent frameworks emphasise the importance of a sociotechnical approach to AI. This perspective treats AI not as an isolated technology, but as part of a broader system that includes people, processes, infrastructure, regulation, and organisational culture.

From this viewpoint, clinical integration is not a final implementation step. It is a guiding principle throughout the entire lifecycle of an AI system. Real-world validation, continuous user feedback, and clear governance around responsibility and accountability are essential.

Why technical excellence alone is not enough.

Overemphasising technical performance can be counterproductive. AI systems optimised to outperform clinicians on narrow benchmarks may deliver limited real-world value if they do not address genuine clinical needs.

In contrast, modest tools that support consistency, reduce cognitive burden, or surface relevant information at the right moment often have a greater impact. Their success lies in being usable, acceptable, and trusted by clinicians.

From replacement to partnership: The future of Clinical AI

The future of AI in healthcare is not about replacing clinicians. It is about partnership.

AI has real potential to augment clinical expertise, improve safety, and support more sustainable healthcare systems. Realising this potential depends on moving beyond standalone tools and embracing clinical integration as the foundation for success.

Without integration, AI risks remaining trapped in a cycle of hype and disappointment. With it, AI can become a meaningful part of everyday care.

Moving from potential to practice

For healthcare organisations looking to move beyond pilots and proofs of concept, the priority is clear: invest in platforms and approaches that embed intelligence directly into clinical workflows, support clinical reasoning, and evolve alongside care delivery.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Abd-Alrazaq, Alaa, Barry Solaiman, Yosra Magdi Mekki, Dena Al-Thani, Faisal Farooq, Metab Alkubeyyer, Mohamed Ziyad Abubacker, et al. 2025. “Hype vs Reality in the Integration of Artificial Intelligence in Clinical Workflows.” JMIR Formative Research 9 (1): e70921.
• Alami, Hassane, Pascale Lehoux, Chrysanthi Papoutsi, Sara E. Shaw, Richard Fleet, and Jean-Paul Fortin. 2024. “Understanding the Integration of Artificial Intelligence in Healthcare Organisations and Systems through the NASSS Framework: A Qualitative Study in a Leading Canadian Academic Centre.” BMC Health Services Research 24: 701.
• Chustecki, Margaret. 2024. “Benefits and Risks of AI in Health Care: Narrative Review.” Interactive Journal of Medical Research 13: e53616. 
• .El Arab, Rabie Adel, Mohammad S. Abu-Mahfouz, Fuad H. Abuadas, Husam Alzghoul, Mohammed Almari, Ahmad Ghannam, and Mohamed Mahmoud Seweid. 2025. “Bridging the Gap: From AI Success in Clinical Trials to Real-World Healthcare Implementation—A Narrative Review.” Healthcare 13 (7): 701. 
• Karalis, Vangelis D. 2024. “The Integration of Artificial Intelligence into Clinical Practice.” Applied Biosciences 3 (1): 14–44. 
• Nilsen, Per, David Sundemo, Fredrik Heintz, Margit Neher, Jens Nygren, Petra Svedberg, and Lena Petersson. 2024. “Towards Evidence-Based Practice 2.0: Leveraging Artificial Intelligence in Healthcare.” Frontiers in Health Services 4: 1368030. 
• Sokol, Kacper, James Fackler, and Julia E. Vogt. 2025. “Artificial Intelligence Should Genuinely Support Clinical Reasoning and Decision Making to Bridge the Translational Gap.” npj Digital Medicine 8: 345. 

At the same time, failures in how health data is governed, shared, or protected can erode trust and place unsustainable burdens on clinicians and health systems. For this reason, health data stewardship is not merely a technical or administrative concern. It is fundamentally a question of responsibility: who is answerable for health data, and on whose behalf?

Why health data stewardship cannot be reduced to a single role.

The literature is clear that health data stewardship cannot be reduced to a single role or institution. Early interpretations of data stewardship focused on optimising data management for efficiency and interoperability, often framed through principles such as accessibility and reuse.

While this approach has delivered important advances, particularly in research data management, it is insufficient in healthcare. Health data is personal and relational. Stewardship must therefore extend beyond optimisation to encompass ethical, legal, and social responsibilities.

Stewardship as a shared responsibility across the health system.

Responsibility for health data is distributed across a complex ecosystem of actors, each with distinct but overlapping obligations.

The Role of Government and Public Authorities.

Governments and public authorities hold a foundational stewardship role because they set the legal, regulatory, and policy environment within which health data is collected, used, and shared. This includes defining privacy protections, security requirements, rules for secondary use, and accountability mechanisms in the event of errors.

International experience shows that where governance is fragmented or weakly enforced, health data initiatives struggle to earn and retain public trust. Conversely, coherent and transparent governance frameworks provide clarity and enable coordination across the system.

Health Organisations as Operational Stewards.

Health system organisations, including ministries of health, national agencies, hospitals, and primary care providers, carry direct stewardship responsibilities because they generate and operationalise health data.

Their obligations extend beyond technical security and regulatory compliance. They are responsible for ensuring data quality, accuracy, and appropriate access, and for embedding data practices that support safe, effective, and equitable care.

The experience of patient portals illustrates the consequences of failing to treat stewardship as both a social and technical responsibility. Poor design, limited interoperability, and uncontextualised release of results can increase patient anxiety, add to clinician workload, and undermine the patient–clinician relationship. These are not merely design flaws. They are stewardship failures.

Clinicians and the burden of digital data.

Clinicians occupy a distinctive position within the stewardship landscape. They are both producers and users of health data, and they remain professionally and ethically accountable for how data is interpreted and acted upon in care.

The expansion of digital communication channels and patient-facing records has increased expectations of constant availability and rapid response. Without appropriate safeguards, this contributes to burnout and moral distress. Stewardship that improves access while transferring unmanaged burdens onto clinicians cannot be considered responsible.

Technology vendors as stewards by design.

Technology vendors and digital health companies are also key stewards, particularly as private actors increasingly design and operate critical health data infrastructure.

Their responsibilities include building systems that are secure by design, interoperable, usable, and aligned with clinical and patient needs rather than regulatory minimums alone. Evidence from cybersecurity research shows that technical safeguards are insufficient if human factors, workflow, and training are neglected.

Vendors, therefore, share responsibility for mitigating risk and avoiding design choices that increase the likelihood of error or misuse.

Patients, communities, and the limits of individual responsibility.

Patients and the public are not passive beneficiaries of stewardship but central stakeholders. Modern governance frameworks increasingly recognise individuals’ rights to access, understand, and in some contexts influence how their data is used.

At the same time, placing full responsibility on individuals to manage complex data environments is neither realistic nor fair. Effective stewardship must balance individual rights with collective benefit, ensuring people are meaningfully informed and supported without being overburdened or exposed to harm. This balance is particularly important for secondary uses of data, such as research or system planning.

Stewardship, equity, and collective rights.

Indigenous communities and groups subject to historical or structural disadvantage raise additional stewardship responsibilities. Literature from Aotearoa New Zealand and internationally highlights that health data can affect not only individuals, but also communities and collective wellbeing.

Stewardship in these contexts requires recognition of collective interests and respect for principles of self-determination. Failure to do so risks reinforcing inequities rather than addressing them.

While not directly plug-and-play for healthcare, the DSO model illustrates governance architectures that prioritise the agency of data subjects and rights holders as data use becomes more complex.

Connecting the layers of responsibility.

Taken together, the evidence points to health data stewardship as a shared and layered responsibility, not a single office or role. Governments provide the mandate and guardrails. Health organisations operationalise stewardship in care delivery. Clinicians enact it in practice. Vendors embed it in technology. Patients and communities shape its legitimacy through trust and participation.

Effective stewardship depends not on concentrating responsibility in one place, but on clearly articulating how these responsibilities connect and overlap.

Stewardship in service of trust and better outcomes.

The purpose of health data stewardship is not data for its own sake. It is to support better health outcomes, more equitable systems, and sustained public trust.

When stewardship is narrowly defined as compliance or efficiency, systems fracture, and people suffer. When it is understood as a collective ethical responsibility, health data becomes a shared asset that can be used confidently and wisely in the service of both individuals and society.

The challenge now is not whether health data matters, but whether our stewardship models are fit for the responsibility they carry.

Authored by Tom Varghese, Global Product Marketing & Growth Manager at Orion Health.

References

• Irizarry, Taya, Annette DeVito Dabbs, and Christine R. Curran. “Patient Portals and Patient Engagement: A State of the Science Review.” Journal of Medical Internet Research 17, no. 6 (2015)
• Johnson, Adam M., Andrew S. Brimhall, Erica T. Johnson, Jennifer Hodgson, Katharine Didericksen, Joseph Pye, G. J. Corey Harmon, and Kerry B. Sewell. “A Systematic Review of the Effectiveness of Patient Education through Patient Portals.” JAMIA Open 6, no. 1 (2023)
• Morley, Jessica, Lisa Murphy, Abhishek Mishra, Indra Joshi, and Kassandra Karpathakis. “Governing Data and Artificial Intelligence for Health Care: Developing an International Understanding.” JMIR Formative Research 6, no. 1 (2022)
• Nemec Zlatolas, Lili, Tatjana Welzer, and Lenka Lhotska. “Data Breaches in Healthcare: Security Mechanisms for Attack Mitigation.” Cluster Computing 27 (2024): 8639–8654.
• Stillman, Michael. “Death by Patient Portal.” JAMA 330, no. 3 (2023): 223–224. 
• Wendelborn, Christian, Michael Anger, and Christoph Schickhardt. “What Is Data Stewardship? Towards a Comprehensive Understanding.” Journal of Biomedical Informatics 140 (2023)
• Gue, D’Arcy. “How Patient Portals Are Failing Healthcare & Patients.” Medsphere Systems Corporation, March 21, 2019.

That context is clearly reflected in Black Book Market Research’s State of Australian Healthcare IT 2026 report, an independent assessment based on feedback from more than 450 Australian clinicians, digital health leaders, and executives. The findings point to a market at an inflection point: EMRs are firmly embedded as essential infrastructure, but confidence in their ability to support long-term transformation is far from uniform.

Across EMR systems in Australia and New Zealand, national digital health strategies are also lifting the bar. Interoperability, shared records, data use and consumer access are no longer future ambitions; they are current requirements. As a result, EMR decisions are becoming less about feature checklists or vendor scale and more about how well a platform aligns with local models of care, system maturity, and long-term strategic direction.

Against this backdrop, a common question is emerging: are EMRs simply helping health systems run yesterday’s workflows more efficiently, or are they truly equipped to support how care needs to be delivered next?

This article outlines the key priorities shaping EMR systems in Australia and New Zealand, based on independent market research and regional digital health strategy.

1. EMRs must enable new models of care, not just digitise old ones.
2. Clinical usability and medication safety remain the foundation.
3. Interoperability is now core health infrastructure.
4. Data readiness and AI must deliver real clinical value.
5. Transformation is an ongoing journey, not a go-live event.
6. Resilience, cloud and partnership value are rising fast.

EMRs must enable new models of care, not just digitise old ones.

Across Australia and New Zealand, EMRs are being reassessed through a sharper strategic lens. The real value of a modern EMR is no longer found in digitising clinical documentation alone, but in its ability to enable new models of care that health systems urgently need.

As care delivery increasingly extends beyond hospital walls into virtual, community and home-based settings, EMRs are expected to act as connective tissue, supporting seamless information flow across clinicians, care settings, organisations and regions. This shift reflects growing pressure on health systems to improve access, safety and sustainability while managing workforce constraints and rising demand.

While EMRs have delivered measurable gains in patient safety and access to information, their contribution to genuine care model transformation has been uneven. The next phase of digital health maturity in Australia and New Zealand demands platforms designed for evolution, supporting interoperability, data-driven decision-making and continuous improvement, rather than optimising static acute workflows.

Clinical usability and medication safety remain the foundation.

Despite growing strategic expectations, Australian and New Zealand evidence is unequivocal: clinical usability remains non-negotiable. Clinician workflow fit and medication safety consistently outrank all other EMR selection criteria.

Transformation cannot succeed if EMRs increase cognitive load or disrupt real-world clinical practice. High adoption is not achieved through mandate alone; it requires systems that make daily work easier, safer and more intuitive. Sustained clinician engagement, training and post-go-live optimisation remain critical to long-term success.

This aligns with findings from global clinician experience research, reinforcing that usability and medicines management underpin every other digital ambition.

Interoperability is now core health infrastructure.

Interoperability has moved decisively from aspiration to baseline requirement.

In Australia, seamless integration with My Health Record and national digital services is now central to EMR evaluation. In New Zealand, national shared record initiatives are driving similar expectations, focusing on reducing fragmentation and improving continuity of care across regions.

EMRs must support open, standards-based data exchange to enable interdisciplinary care. Standards like HL7, FHIR and SNOMED CT are no longer technical nice-to-haves. Instead, they are foundational enablers for collaborative decision-making and integrated service planning at scale.

Data readiness and AI must deliver real clinical value.

Health systems across Australia and New Zealand are moving beyond retrospective reporting towards real-time insight, predictive analytics and AI-enabled decision support.

EMRs are now expected to provide a robust data foundation to support use cases such as early clinical deterioration detection, patient flow optimisation and readmission risk prediction. Generative AI is already showing promise in summarising complex patient records and reducing administrative burden.

However, the priority in Australia and New Zealand remains pragmatic. Health systems are less interested in experimentation than in demonstrable improvements to safety, flow and outcomes. This places a premium on EMRs architected for continuous evolution, where data, analytics and AI are embedded by design rather than added as bolt-ons.

Transformation is an ongoing journey, not a go-live event.

A recurring theme across both Australian research and real-world experience is that EMR value is not delivered at go-live. It emerges over time through continuous measurement and adaptation.

Too often, adoption data and usage patterns are underutilised, despite offering powerful insight into what is working and where intervention is required. Leading organisations are treating EMR deployment as a long-term capability journey, supported by ongoing benefit measurement.

This approach recognises that care models, clinical expectations and digital maturity will continue to evolve, and that EMRs must evolve alongside them.

Resilience, cloud and partnership value are rising fast.

Cybersecurity, reliability and business continuity have become board-level concerns. Cloud and managed service models are gaining acceptance across the region, provided that sovereignty and security requirements are met.

Equally important is partnership behaviour. EMR programmes span decades, and Australian evidence shows an increasing sensitivity to vendor responsiveness, transparency and whole-of-life value. Health systems are increasingly wary of platforms that promise transformation without delivery capability or long-term alignment to national and organisational digital health strategies.

Choosing the right EMR is about strategic fit, not size or features.

There is no single “best” EMR for Australia or New Zealand. What matters is strategic fit, the right balance of clinical usability, interoperability, data capability, resilience and partnership value for each health system’s context and ambition.

The EMRs most likely to succeed over the next decade will be those that move beyond transactional digitisation and actively enable integrated, patient-centred models of care. The real question is no longer which EMR is the most powerful, but which is best able to evolve alongside healthcare itself.

What this means for EMR strategy in Australia and New Zealand

Across Australia and New Zealand, EMRs are no longer judged by scale or feature depth alone. Success increasingly depends on strategic fit: how well a platform supports clinicians, enables interoperability, uses data responsibly, and evolves alongside changing models of care. The EMRs that will deliver lasting value over the next decade will be those designed not just to digitise healthcare, but to adapt with it.

Explore how Orion Health supports strategic-fit EMR approaches

As health systems across Australia and New Zealand rethink what they need from EMRs, many are looking beyond monolithic platforms to solutions that strengthen interoperability, shared records and care coordination across settings.

Orion Health works with public and private healthcare organisations to enable connected, patient-centred models of care, supporting interoperability at scale, region-wide shared care records, and digital foundations tailored to private healthcare environments.

• Explore interoperability solutions to enable connected care across organisations and regions.
• Discover shared care record solutions designed to support continuity, collaboration and national digital health strategies.
• Learn how Orion Health supports private healthcare providers with flexible, interoperable digital solutions.

Whether you’re evolving an existing EMR environment or building a more connected digital ecosystem, Orion Health helps health systems choose solutions that can adapt as care continues to change.

References

• Black Book Market Research LLC. Healthcare IT 2026: State of Australian Healthcare IT. January 2026.
• Jones, D. “EMR solutions evolve to help transform healthcare.” Health Services Daily, April 2025.
• Deloitte. 2025 Global Health Care Outlook. 2024.
• KLAS Research. Arch Collaborative Global EHR Satisfaction Report. 2024.

From general practice and policymaking to digital strategy and patient advocacy, their reflections point to a new phase for digital health in Aotearoa, one grounded in interoperability, shared standards, and a clearer focus on what really matters.

Fix the foundation: data and interoperability.

If there’s one theme that cuts across every perspective, it’s this: we can’t build better systems on broken foundations. Data needs to flow seamlessly and securely across the ecosystem; only then can innovation truly take root.

“Some of the big opportunities for technology and Health New Zealand lie in the data and interoperability space… making sure we’re using modern APIs and have high-quality data.”
—  Dr. Jonathan Hoogerbrug, General Practitioner (GP) and Clinical Informatics Director at Health New Zealand

“There are huge opportunities in actually getting the basics right. If something works, it works, and it can be scaled across multiple environments.”
— Stella Ward, CEO at Digital Health Association (DHA)

Modern, standards-based integration isn’t a “nice to have”; it’s the essential enabler of every improvement that follows.

Focus beats noise: do fewer things, but do them better.

In a sector often distracted by shiny tools and pilot projects, several voices called for a mindset shift: less is more.

“My one wish for digital health is that we take a ‘less is more’ attitude and focus on one thing at a time.”
— Ray Delany, Founder of CIO Studio.

The message is clear: it’s time to stop reinventing the wheel and start scaling what already works. The opportunity is not in novelty, but in execution and alignment.

Design around the patient: a unified experience.

Despite major advancements, the digital experience for patients remains fragmented. Healthcare journeys still involve repeated questions, disconnected systems, and limited visibility across care providers.

“The biggest opportunity is enabling the patient to digitally manage multiple providers through a single kind of journey in a single interface.”
— Harry Hawke, CEO & Managing Director, Webtools

“A continuous health record… accessible at any point of the health system, at any age or stage of life… in live time.”
— Gillian Robinson-Gibb, Founder and CEO at Hercules Health.

Designing for patients means more than giving them a login. It means connecting the dots, so care can follow them, not the other way around.

Let more voices lead the way.

True transformation means listening to a broader range of voices, not just those in formal leadership roles. It means bringing lived experience and frontline expertise into every conversation.

“I have two wishes, the first is to make sure that the patient and whānau stories, their pūrākau, their voice, drives and amplify what we do. And the second is, It wouldn’t just be the medical voice. It would also be allied health, scientific and technical, and the nursing and midwifery voice.”
— Rowena Woolgar, Director at Arise Consulting

This is about creating inclusive systems, shaped by the communities they serve, and the teams who deliver care every day.

A call for cohesion: collaborate and execute.

Despite the challenges, optimism remains high. The sector knows what needs to change; the next step is to align and act.

“My wish for the sector is that we actually work together and stop being so siloed.”
— Heather Phillips, Head of Corporate Affairs, Awanui Group

“To collaborate. To communicate. To be clear about our vision… and to execute on that vision.”
— Stella Ward

There’s no shortage of insight or innovation in Aotearoa. The opportunity now is to connect through shared goals, standards, and responsibility.

Missed the reflections from HINZ 2025? Discover what clinicians and leaders had to say about 25 years of transformation in Aotearoa’s health system:

Digital health then and now: A clinician’s view

Over the years, the impact of digital technologies on clinical workflows has been profound. From the introduction of the internet and electronic health records to the adoption of e-prescribing and AI scribes, healthcare professionals have had to constantly adapt to new tools and ways of working.

Stella Ward, CEO at Digital Health Association (DHA), reflects:

“The internet was an amazing piece of technology… it enabled so much in terms of patient experience, and also research, search, and research at your fingertips.”

But the most enduring lesson isn’t about any specific technology. It’s about change itself.

“Technology comes and goes. But the main constant is people and how people react and interact with technology.” – Ray Delany, Founder of CIO Studio.

COVID-19 as a digital catalyst

The pandemic fast-tracked the adoption of key technologies, including e-prescribing, remote care, and AI-assisted documentation, among them. What might have taken years to implement happened in months.

This period showed what’s possible when urgency meets digital readiness. It also demonstrated the need for systems that are intuitive, adaptable, and integrated, bringing together fragmented information to support continuity of care.

What is good digital care in New Zealand?

Good digital health is invisible. Throughout the video, a powerful theme emerges: digital care is most effective when patients barely notice it.

“Good digital care for the patient is something they’re not aware of… where they’re still getting that human connection… At the end of the day, digital has to become invisible” – Gillian Robinson-Gibb, Founder and CEO at Hercules Health.

What does this look like in practice?

• A system that supports, not replaces, the patient-clinician relationship.
• Digital experiences that are seamless, equitable, and intuitive.
• Technologies that are invisible to the patient, yet invaluable to the clinician.

It reflects a growing vision across the health sector, where digital maturity isn’t just about technology, but about empowering people, simplifying care, and turning data into meaningful action.

The future of digital health: invisible, intuitive, integrated

As Orion Health continues to support health systems across New Zealand and around the world, we’re helping shift digital from front-of-mind to behind-the-scenes, ensuring clinicians can focus on care, not clicking.

Solutions like the Digital Care Record and Digital Front Door are already enabling this future by:

• Aggregating patient data into a single view for care teams
• Empowering patients with access to their care plans, health data, and tasks
• Reducing administrative burden with digitised forms, messaging, and workflows

Whether it’s enabling cancer diagnosis through digital pathology or streamlining admissions for surgical care, the goal is the same: make digital health feel human.

A relationship-centred future

Perhaps the most powerful insight, and one summarised by Dr. Jonathan Hoogerbrug, General Practitioner (GP) and Clinical Informatics Director at Health New Zealand, is:

“Good digital care looks like a relationship between patients, whānau, and healthcare providers.”

After 25 years of digital evolution, that vision remains unchanged. Technology should enable stronger relationships, not replace them. And as we look ahead, the healthcare experience of tomorrow will be built on the same values that have always mattered: trust, connection, and compassion.

Looking ahead? Check out our companion blog from HINZ 2025 on what’s next for digital health in Aotearoa, from interoperability to patient-led design: