iampaulbrown

[iampaulbrown]

Interoperability in the NHS: A Pathway to Integrated Care

As healthcare systems evolve, the need for effective information sharing between various care settings, organisations, and geographies has become increasingly apparent. In the NHS, this need is crucial for optimising patient outcomes and ensuring high-quality care. The drive toward interoperability - the seamless exchange of information between different IT systems across healthcare - plays a pivotal role in the delivery of integrated and efficient care. Interoperability is not just about connecting systems but also about creating a future where patients, healthcare professionals, and organisations can work together to manage and improve healthcare delivery.

The current landscape of healthcare in England is witnessing the emergence of new models of care. Integrated care systems (ICS) have been established to improve coordination between health and social care services. These new care models require better collaboration between professionals and citizens to manage care effectively. At the core of this vision is the ability to share relevant information in real time across different services and organisations. To achieve this, the NHS is focused on ensuring that IT systems used in healthcare are interoperable, enabling seamless data exchange and, ultimately, better care coordination.

To support this transition, the NHS is focusing on several key strategic areas. One primary focus is working closely with various services to identify their specific business needs in relation to interoperability. This process helps inform the development of solutions tailored to meet those needs. In parallel, the NHS is developing priority use cases for interoperability, which serve to provide business justification for local investments and guide the development of supporting systems at a national level. By identifying these priority areas, the NHS ensures that local healthcare organisations can align their investments with national strategies, maximising the impact of their efforts to improve information sharing.

Supporting local organisations with tools and guidance is another important area of focus. As each region or organisation faces unique challenges, providing a set of clear guidelines and tools helps local healthcare providers develop effective solutions to their interoperability challenges. The NHS is also working on creating standards that can be adopted nationwide, which will ensure consistency and reliability in the exchange of health information. One critical aspect of this standardisation effort is the transition from paper-based transfers of care to electronic processes. Key examples include electronic discharge summaries for patients leaving inpatient care, mental health services, or accident and emergency (A&E) departments. Ensuring that these transfers are conducted efficiently, accurately, and electronically reduces delays in communication, improves patient safety, and ensures continuity of care.

To facilitate the sharing of patient information across systems, the NHS is developing standards that support systems capable of open access to patient information. This is being achieved through the development and adoption of CareConnect APIs, which provide a structured method of exchanging health records using nationally defined FHIR (Fast Healthcare Interoperability Resources) standards. CareConnect APIs allow clinicians in different care settings, such as A&E or outpatient clinics, to access and view a patient’s medical records from other regions or organisations. This interoperability between systems ensures that healthcare providers have access to comprehensive information when making clinical decisions, leading to better-informed care and improved patient outcomes.

Interoperability is further supported by collaborations between NHS Digital and INTEROPen, an industry-led group focused on fostering the adoption of interoperability standards - interopen.co.uk. These collaborations aim to accelerate the adoption of open interfaces and data-sharing protocols across the NHS, ensuring that healthcare providers can access, share, and interpret critical patient information in a standardised and secure manner.

The Chief Clinical Information Officer for health and care in England has outlined seven priority areas for the adoption of interoperability within the NHS. One of the most critical aspects is ensuring real-time access to the NHS Number, the unique patient identifier, across all healthcare services. This allows for better tracking of a patient’s health records and ensures that the NHS Number is associated with key care record elements, such as laboratory tests. Another priority area is ensuring that all medication information is interoperable and machine-readable across the NHS. This is important for improving the safety and efficiency of medication management, especially when patients are treated across different care settings.

Additional priority areas include creating a consistent way to identify and authenticate healthcare staff across services, developing unified standards for appointment bookings and scheduling, and implementing a standardised way to share basic clinical observations and pathology test results. The diagnostic coding system SNOMED CT is being widely implemented to replace older coding systems like Read codes, with the goal of creating a uniform clinical terminology across primary, secondary, acute, and mental health care services. This standardisation is essential for improving the quality and consistency of clinical data, enabling more effective data sharing and interpretation across the NHS.

The NHS Standard Contract has established clear requirements to ensure that major clinical information technology systems used by providers support the exchange of structured information through open interfaces. From April 2020, healthcare providers have been required to ensure that their IT systems are capable of sending and receiving CareConnect APIs to facilitate the transfer of care information. This includes the requirement to send discharge summaries within 24 hours of patient discharge to the relevant GP or referring healthcare provider. By adopting these standards, the NHS aims to create a more integrated healthcare system that ensures the continuity of care from one provider to another, reducing delays and improving patient outcomes.

In conclusion, the NHS’s commitment to interoperability is essential for the future of healthcare in England. By enabling seamless information sharing across healthcare settings, the NHS is not only improving the efficiency of care delivery but also ensuring that patients receive safer, more coordinated, and personalised care. Through collaboration with organisations like INTEROPen and the development of standards such as CareConnect APIs, the NHS is paving the way for a fully integrated and interoperable healthcare system that meets the evolving needs of patients and providers alike.

[iampaulbrown]

The State of EPR Adoption in NHS Trusts

The use of Electronic Patient Records (EPRs) in NHS Trusts across England is both complex and varied, reflecting the unique needs and preferences of different healthcare settings.

To get a clearer picture, I researched the EPR systems currently in use across all 214 NHS Trusts in England. Here are the key findings and what they mean for the future of healthcare IT in England.

Leading EPR Systems

Oracle Cerner - is the most widely adopted EPR system, used by 35 NHS Trusts. Cerner is especially prominent in acute trusts, where handling complex and high-volume patient data is critical. Cerner's widespread use highlights its reliability and the trust healthcare providers place in its ability to manage essential patient information efficiently.

The Access Group Rio - is the top choice for mental health trusts, with 28 trusts using its system. Rio's strength is in its tailored features for mental health services, making it the go-to for 24 out of 58 mental health trusts. This shows how crucial Rio is for managing patient care in this field.

TPP SystmOne - is also a major player, particularly in community and mental health trusts. It's used by 27 trusts, showing its versatility and effectiveness in different healthcare settings. Its integration capabilities and user-friendly interface make it a favourite for community healthcare providers.

System C and Dedalus Lorenzo - are also notable, with 21 and 14 trusts respectively. Their range of functionalities supports various healthcare processes, indicating strong trust and a solid foothold in the market.

Emerging and Niche Players

While major EPR systems lead the market, others like Epic (13 trusts), Altera Digital Health Sunrise (11 trusts), and MEDITECH Expanse (8 trusts) are gaining traction. Nervecentre Software and OneAdvanced Carenotes, used by 6 and 4 trusts respectively, are growing in importance too. These newer systems offer unique features and innovative approaches tailored to specific NHS needs.

Gaps and Custom Solutions

Interestingly, 14 trusts still don’t have any EPR system in place. This gap is a concern as it suggests potential inefficiencies in managing patient data, which can hinder coordinated and high-quality care.

Additionally, 10 trusts have opted for bespoke EPR systems. While these custom solutions might offer tailored functionalities, they could face integration and scalability challenges compared to more widely used systems. This bespoke approach reflects the unique requirements of certain trusts but also underscores the need for a balanced strategy that ensures compatibility and interoperability across the NHS.

Check out the full list of all 214 NHS Trusts and what EPR they are using here: https://6b.digital/insights/list-of-eprs-in-nhs-trusts-in-england.

[iampaulbrown]

Building a Positive Software Engineering Culture

In a fast-paced software development landscape, driving a healthy engineering culture is really important.

A positive engineering culture improves collaboration and innovation, it improves the quality of the delivered outcome, it plays a pivotal role in attracting and retaining top talent, and it makes work more fun and enjoyable for everyone involved.

Culture is not a ‘one and done’ activity, it is something a team needs to constantly be thinking about, working towards, and improving on.

Defining Culture

Culture is a set of shared values, goals, and principles that guide the behaviours, activities, priorities, and decisions of a group of people working towards a common objective. Culture shapes how a team operates, collaborates, and prioritises tasks, which includes goal-setting, technical practices, collaboration methods, decision-making processes, and the communication styles embraced by the team.

It is the responsibility of the team and management to define and drive the culture, and key to the success of this is the congruency between the defined values and the actual behaviours (particularly of management), as any disparities can lead to a breakdown in trust and morale.

Components of a Healthy Software Engineering Culture

A lot has been written about healthy engineering cultures, I think there are three essential components to focus on:

Personal Commitment: Developers commit to creating high-quality products by applying effective software engineering practices.

Organisational Commitment: Managers commit to providing an environment where software quality is a key success driver, supporting each team member in achieving this goal.

Continuous Improvement: All team members commit to continuously improving their work processes, thereby enhancing the quality of the products they create.

For these components to be successful they need to be driven with purposeful and measured activities and underpinned by shared values that are focused on collaboration, performance, and progression eg-

Defining Cultural Characteristics: Engage all team members in defining the principles, values, and attitudes that are important to the team’s culture. This shared understanding empowers everyone to contribute to shaping the culture.

Implementing a Metrics Program: Implement a metrics program to track project activities, providing insights into how work is done compared to how it should be done, to promote transparency and accountability.

Embedding Ongoing Cultural Evolution: Leaders should continuously reinforce agreed-upon cultural values. Periodic calendar reminders and alignment of actions with cultural values help sustain the positive evolution of culture over time.

With each activity leaders need to be present, bought-in, and engaged. Leaders are not just managers, even in flat structures, certain individuals will emerge as leaders in different areas, guiding a team towards common objectives and establishing and promoting values inline with the culture.

Conclusion

Building and sustaining a positive software engineering culture requires an ongoing commitment from both team members and leadership. By defining shared values and promoting continuous improvement, teams can create an environment where innovation thrives, and individuals feel safe and happy to contribute their best work.

[iampaulbrown]

Semantic Interoperability in Healthcare

In healthcare patient data is paramount and lives are at stake, efficient data sharing and communication are crucial. The concept of interoperability, specifically semantic interoperability, has emerged as a key driver in revolutionising healthcare by ensuring that information can flow seamlessly between different systems and healthcare providers.

There is a significance of semantic interoperability in healthcare in 2023, the challenges it addresses, and the positive impact it can have on patient care.

Understanding Semantic Interoperability

Semantic interoperability refers to the ability of different healthcare systems and organisations to understand, exchange, and use healthcare data in a way that preserves the meaning of the information being shared - it's about making sure that data from one source can be interpreted correctly by another, allowing healthcare providers to access and comprehend patient information, irrespective of the systems or languages used.

The Challenges of Interoperability in Healthcare

Diverse Data Sources: Healthcare data comes from a multitude of sources, including electronic health records (EHRs), wearables, imaging systems, and more. Ensuring that all these sources can communicate effectively is a major challenge.

Data Standardisation: Different healthcare systems may use various data formats and coding schemes. Semantic interoperability requires standardisation so that data can be understood universally.

Privacy and Security: Patient data is highly sensitive, and ensuring that it remains secure while being shared is a significant concern.

Data Volume: The sheer volume of healthcare data generated daily is overwhelming. Interoperable systems must handle large data sets efficiently.

The Benefits of Semantic Interoperability

Improved Patient Care: Semantic interoperability enables healthcare providers to access complete and accurate patient data in real-time, leading to more informed decisions and better patient outcomes.

Reducing Medical Errors: By ensuring that critical patient information is available and understood, interoperable systems reduce the likelihood of medical errors resulting from incomplete or misinterpreted data.

Enhanced Research and Analytics: Researchers and healthcare professionals can access large datasets more easily, which facilitates medical research and the discovery of patterns and trends.

Cost Reduction: By streamlining data sharing and reducing redundancy, semantic interoperability can lower administrative costs and improve overall operational efficiency in healthcare institutions.

Telemedicine and Remote Monitoring: The rise of telemedicine and remote patient monitoring relies heavily on interoperability. Patients can receive care and transmit data from their homes, increasing access and convenience.

The Role of Standards in Semantic Interoperability

Standards, such as HL7 (Health Level Seven) and FHIR (Fast Healthcare Interoperability Resources), play a pivotal role in achieving semantic interoperability. These standards provide a common language and framework for structuring and exchanging healthcare data, making it possible for diverse systems to understand and use the information.

It is important though to consider that standards can be extended and customised, so that the standard itself becomes non-standard, so alignment to national standards (eg in the UK, FHIR UK Core) is essential.

Challenges in Achieving Semantic Interoperability

While the benefits of semantic interoperability in healthcare are evident, there are several challenges:

Legacy Systems: Many healthcare organisations still rely on legacy systems that were not designed with interoperability in mind. Upgrading and integrating these systems can be complex and costly.

Data Governance: Ensuring that data is accurate, complete, and up-to-date is critical for semantic interoperability. Data governance models must be established and maintained.

Regulatory and Legal Hurdles: Compliance with regulations like GDPR (General Data Protection Regulation) can add complexities to data sharing.

Conclusion

Semantic interoperability through technologies like HL7 FHIR and OpenEHR is transforming healthcare by breaking down data silos, facilitating more effective and efficient patient care, and promoting innovation in medical research. While challenges remain, the benefits of seamless data sharing, improved patient outcomes, and reduced costs make the pursuit of semantic interoperability a vital goal for healthcare organisations worldwide. As technology continues to advance, it is clear that the path to a more interconnected and data-driven healthcare system is on the horizon, with federated architectures and semantic interoperability at its core.

[iampaulbrown]

Understanding NHS Enterprise Architecture Principles

The NHS Architecture Principles for digital services serve as a guide for technical architects and service designers, ensuring that the digital infrastructure supporting any new NHS digital service is not only effective but also sustainable, interoperable, and user-centric. There are 10 key NHS Architecture Principles that need to be considered with each new implementation:

1. Delivering Sustainability: The first principle underscores the importance of delivering digital services sustainably, considering both the robustness of the system and its environmental impact. Digital services need to weather the storms of change while remaining responsive. An intriguing aspect of this principle is its focus on 'embodied emissions', which delves into the ecological and social impact of IT equipment's production and disposal. Architects are tasked with minimising data retention, optimising code efficiency, and limiting architectural obsolescence. Also, the preference for low-energy devices like tablets and mobiles and the judicious application of artificial intelligence reflect the NHS's commitment to a greener future.

2. Browser-Based Services: The NHS recognises the power of modern web technologies, advocating for all digital services to be browser-based and built on open web standards. This approach not only enhances user flexibility but also aligns with the mobile-first mindset. A strategic move away from Internet Explorer 11 towards modern browsers is encouraged, ensuring seamless accessibility across a variety of devices. The shift to browser-based tools also unlocks continuous security updates and improvements, providing a strong defence against cyber threats.

3. Internet First: Embracing the Internet as the default means of accessing information is the essence of the third principle. By adopting internet standards and protocols, the NHS maximises compatibility, enabling a wide range of technologies and digital services to cater to diverse user needs. This principle not only promotes decentralisation but also highlights the importance of data accessibility, echoing the freedom of accessing email from anywhere. The emphasis on public internet accessibility aligns with the NHS's commitment to digital transformation.

4. Public Cloud First: The fourth principle reflects the NHS's readiness to embrace the advantages of cloud computing. By moving digital services to the public cloud whenever feasible, the NHS accelerates deployment timelines and reduces emissions. This cloud-first approach allows the NHS to harness the scalability, agility, and cost-efficiency offered by cloud services. However, it also acknowledges that in some cases, specific service level characteristics or cybersecurity concerns may warrant deviations from this cloud-first strategy.

5. Data Layer with Registers and APIs: The fifth principle orchestrates a symphony of data management, focusing on efficient storage and accessibility through open APIs. Storing data once, and making it available through standardised interfaces, not only reduces costs but also ensures data consistency and security. By breaking down silos and promoting interoperability, this principle sets the stage for smarter and more agile healthcare services.

6. Adopting Cyber Security Standards: The sixth principle mandates adherence to cybersecurity standards, compelling architects to keep software, networks, and systems up to date. This dedication to security underpins public trust and patient confidentiality, ensuring that health and care services remain safeguarded.

7. Using Platforms: The seventh principle promotes collaboration and efficiency by encouraging the use of existing platforms. Reusing common infrastructure not only saves time and resources but also contributes to sustainability efforts. The adoption of this principle not only reduces architectural debt but also streamlines operational processes, ultimately benefiting patients and healthcare providers.

8. Putting User Needs First: User-centred design is essentially the eighth principle. Designing services around user needs enhances usability, inclusivity, and cost-effectiveness. By complying with legal requirements, conducting user needs analysis, and adhering to UX design principles, the NHS ensures that digital services cater to a wide spectrum of users, contributing to a more accessible and equitable healthcare landscape.

9. Interoperability with Open Data and Technology Standards: The ninth principle shines a light on the significance of open data and technology standards. Embracing open standards not only facilitates interoperability between different systems but also promotes innovation by allowing seamless integration of emerging technologies. This interoperability lays the foundation for a flexible and future-ready healthcare infrastructure.

10. Reuse Before Buy/Build: The final principle encourages the reuse of existing solutions or considering off-the-shelf products before resorting to new builds underscores the NHS's commitment to cost-effective and environmentally conscious practices. By minimizing redundancy and maximizing resource utilization, this principle contributes to a more streamlined and agile digital architecture.

Overall the NHS Architecture Principles for digital services encapsulate a holistic approach to designing and implementing a robust, user-centric, and sustainable healthcare ecosystem. These principles reflect the NHS's commitment to innovation, inclusivity, and environmental responsibility, reshaping the way healthcare services are conceptualised and delivered. By following these guiding principles, architects and service designers can play a pivotal role in revolutionising digital healthcare.

[iampaulbrown]

Achieving Digital Excellence in Integrated Care Systems

In the wake of the COVID-19 pandemic, the NHS in England has undergone a significant digital transformation that would typically take years to achieve. This digital acceleration has paved the way for a more resilient and patient-centric healthcare system and it has been guided by NHS England's What Good Looks Like (WGLL) framework - a roadmap designed to shape the future of healthcare delivery

Understanding the WGLL Framework

The WGLL framework is a strategic initiative aimed at advancing digital transformation and elevating healthcare services across ICSs. This initiative draws on local experiences and good practices to provide clear guidance for health and care leaders. By digitising, connecting, and transforming services safely and securely, the WGLL framework aims to improve the outcomes, experiences, and safety of citizens.

Seven Key Success Measures

1. Well Led: The foundation of any successful transformation lies in strong leadership. ICSs and organizations are expected to have clear strategies for digital transformation. Leaders across the spectrum must collectively own and drive the digital transformation journey, ensuring that citizens and frontline perspectives remain central. This entails building digital and data expertise into leadership and governance structures, aligning with ICS strategies, and fostering a culture of innovation.

2. Ensure Smart Foundations: The reliability and sustainability of digital infrastructure are paramount. ICSs are encouraged to invest in multidisciplinary teams with expertise spanning clinical, operational, informatics, design, and technical domains. Compliance with sustainability objectives like net zero carbon and technology standards, coupled with robust cyber security measures and secure data management practices, are key to laying a strong foundation.

3. Safe Practice: Maintaining high standards of safe care in the digital realm is a non-negotiable. Organisations within ICSs should establish robust cyber security plans inline with DSPT, DTAC, and Cyber Essentials, including risk management and mitigation strategies. Adequate resourcing of cyber security and clinical safety functions, along with adherence to relevant frameworks, is vital to ensure patient safety and data integrity.

4. Support People: Empowering the workforce with digital literacy is integral to realising the full potential of digital health technology. ICSs must promote a digital-first approach and provide continuous professional development opportunities. Intuitive, user-friendly systems, flexible working arrangements, and 24/7 digital support services contribute to a motivated and efficient workforce.

5. Empower Citizens: Citizen-centricity is the cornerstone of modern healthcare. ICSs should co-design strategies for citizen engagement and digital services, ensuring inclusivity and accessibility. The use of national digital tools (like the NHS.uk, NHS App, NHS Login), coupled with effective communication channels, enables citizens to actively participate in their healthcare journey.

6. Improve Care: Digital technology has the power to reshape care pathways, reduce variation, and enhance patient outcomes. ICSs should leverage data and digital solutions to redesign care, promote safety, and facilitate remote consultations and monitoring. Collaborative care planning and integration of digital tools drive multidisciplinary and patient-centered care.

7. Healthy Populations: Leveraging data insights can lead to better population health management. ICSs are encouraged to harness data to inform care planning and decision-making, develop innovative care models, and drive digital and data-driven population health initiatives. Collaborations with academia, industry, and partners foster continuous innovation.

What does this mean for digital health innovators?

Digital health innovators need to fully align with the principles of the WGLL framework. From developing secure and reliable digital infrastructure to creating intuitive user interfaces and facilitating data-driven insights, digital health innovators need to be equipped to support every aspect of the WGLL journey.

The "What Good Looks Like" framework sets a clear direction for the future of healthcare in Integrated Care Systems, and provides NHS organisations with a clear path to procuring and managing digital health innovation across healthcare organisations through prioritising interoperable, patient-centred, and safe health and care.

[iampaulbrown]

Sharing Data Across Different Healthcare Settings

In recent years, collaboration between healthcare providers has become a priority. The NHS, underpinned by the Long Term Plan, emphasises the need for local health and care organisations to work together operationally and strategically. One crucial aspect of this collaboration is sharing data effectively across different care settings. Interoperability, the ability to make IT systems work together seamlessly, remains a top priority for NHS IT leaders. However, many healthcare providers still struggle with data sharing, leading to fragmented and inconsistent records. This post aims to provide an overview of sharing data and achieving interoperability in the health and care landscape.

The interoperability landscape

Access to accurate and consistent data is vital for decision-making, improving patient experiences, creating better care pathways, and understanding population health. While sustainability and transformation plans (STPs) and integrated care systems (ICSs) outline the strategic direction for organisational collaboration, the challenge lies in effectively and securely sharing data between stakeholders. Digital transformation projects have been initiated to improve data sharing, but the progress across the NHS varies. Clinicians often face challenges accessing patient data from multiple systems, leading to inefficiencies and delays in care delivery.

Data sharing offers numerous opportunities for improving collaboration in health and care services. Sharing data between hospital departments, such as radiology and pathology results, can significantly enhance the patient experience and reduce costs by avoiding duplicate tests. Data sharing between hospitals, trusts, and clinical commissioning groups (CCGs) has historically been challenging due to duplicate systems, but efforts are being made to consolidate testing and streamline commissioning. Urgent care situations require real-time access to patient information to ensure timely and appropriate care. Sharing data between primary and secondary care is becoming easier with interoperability solutions like Medical Interoperability Gateway (MIG) and GP Connect. Primary Care Networks (PCNs) and local partnerships also benefit from data sharing to enable coordinated and integrated care.

Prioritising data

When embarking on an interoperability project, it is crucial to involve the users and understand their needs. Prioritisation is essential to avoid overwhelming data lakes or repositories that hinder meaningful data extraction. Integration of data from various sources should be done strategically and with minimal disruption. A user-led design approach ensures adoption and protects the investment in interoperable records. Clinical efficiency is a priority, and data should be entered once by clinicians and shared across systems. Access to patient data should be seamless, and digital workflows and care plans should replace paper-based processes.

Information governance and security are crucial for safe and effective data sharing. Organisations need to establish data sharing agreements and ensure compliance with security requirements. Sensitive data, such as mental and sexual health information, requires special consideration, and access should be restricted based on role-based access control. Patient consent should be recorded and easily visible to clinicians. Balancing information governance with integration expertise is key to enabling secure collaboration.

Strong partnerships are essential for effective collaboration. Organisations should foster collaboration within their trust and with external stakeholders in the health and care community. Tactical and strategic cooperation supports rapid scaling of data sharing in critical situations. When selecting suppliers for interoperability solutions, organisations should seek partners with the right skills, experience, and understanding of their goals. Integration experts should work closely with organisations throughout the process, from requirements and architecture to deployment and future enhancements.

Technology to underpin interoperable transformation

Improving access to data should focus on improving existing processes rather than imposing unnecessary change or cost. An effective interoperability strategy avoids disrupting underlying software and familiar workflows. Integration solutions should seamlessly work with existing systems and enhance day-to-day operations. The goal is to improve efficiency and patient outcomes without unnecessary IT changes.

Conclusion

Sharing data and achieving interoperability across different care settings is essential for effective collaboration and delivering high-quality patient care. While challenges exist, such as varying IT systems and information governance requirements, there are steps that organisations can take to facilitate data sharing:

Emphasise user needs: Involve clinicians and other healthcare professionals in the design and implementation of interoperability solutions. Understanding their requirements and workflows will ensure that the technology meets their needs and is adopted effectively.

Prioritise data integration: Instead of overwhelming systems with vast amounts of data, prioritise the integration of key data sources that are most relevant for decision-making and care delivery. This approach prevents information overload and allows for meaningful data extraction.

Ensure information governance and security: Establish robust data sharing agreements and implement appropriate security measures to protect patient information. Adhere to regulatory requirements and consider role-based access control to limit access to sensitive data.

Foster partnerships: Collaboration is crucial for successful data sharing. Develop strong partnerships within your organisation, as well as with external stakeholders, to support interoperability initiatives. Seek out experienced integration partners who understand your goals and can provide the necessary expertise.

Leverage existing technology: Focus on leveraging and enhancing existing systems and workflows rather than implementing disruptive changes. Interoperability solutions should seamlessly integrate with current IT infrastructure to minimise disruption and ensure user acceptance.

By following these guidelines, healthcare organisations can improve data sharing across care settings, leading to better collaboration, enhanced patient experiences, and more informed decision-making. As technology advances and standards evolve, the healthcare landscape will continue to progress towards seamless data interoperability, enabling comprehensive and integrated care for patients.

[iampaulbrown]

Primary Care Interoperability Deep Dive

Interoperability in primary care can be tricky to navigate, particularly because there are lots of clinical systems in play and lots of paths to take to interface.

In this article I am going to look into interoperability with the three ‘principle’ clinical IT systems used by GPs in the UK Vision, EMIS and SystmOne, exploring routes to interface, standards and assessments, network access and testing.

Routes to interface

There are two common routes to take when considering interfacing with primary care systems, either the NHS Digital API Platform or a direct Vendor integration.

The NHS Digital API platform is a ‘front door’ for health and care APIs, primarily for the NHS in England. There are 100+ free APIs available to digital innovators through the platform, with over 50% following modern consistent API patterns like REST, FHIR, OAuth with clear OAS documentation.

Direct to vendor integration is currently only available via the EMIS Partner Program. The EMIS partner programme offers a few additional APIs, as well as technical and commercial support from EMIS. It is a paid programme and we have taken this route a number of times at 6B when the NHS Digital API Platform capabilities do not fit the project requirements.

Standards and Assessments

There are several standards and assessments to be aware of when embarking on a primary care integration project, particularly SCAL and DTAC.

SCAL

Connecting to a few NHS Digital APIs requires completing the NHS Digital self-certification tool, known as the Supplier Conformance Assessment List (SCAL). SCAL is a technical document detailing your organisation, your integration requirements, approach to information governance, clinical safety, functional testing.

The document is worked on collaboratively with NHS Digital and (if relevant) API providers

The SCAL must be kept up to date with any changes to your product (e.g. version or technical details)

You are accountable for reviewing the content of the completed SCAL

The latest version of the SCAL can be obtained from NHS Digital

If you have not worked with NHS Digital before, there is also a requirement to register to use the NHS Digital National Service Desk (ssd.nationalservicedesk@nhs.net) to log incidents

The SCAL is slightly different for each NHS Digital API following this assurance process

DTAC

The Digital Technology Assessment Criteria for health and social care (DTAC) gives staff, patients and citizens confidence that the digital health tools they use meet our clinical safety, data protection, technical security, interoperability and usability and accessibility standards.

Your product should expose any API or integration channels for other consumers, unless there is ‘legitimate rationale’ not to do so. API’s must follow Government Digital Services Open API Best Practice, be well documented, and be freely available and that third parties have reasonable access to connect.

If your product is a wearable or device or if it integrates with them the developer must evidence compliance with ISO/IEEE 10073.

If your product integrates with EHRs, you must use industry standards for secure interoperability (e.g. OAuth 2.0, TLS 1.2).

If you use NHS number to identify patient record data you must use NHS Login unless there is ‘legitimate rationale’ not to do so.

Network access for NHS Digital APIs

Generally newer NHS Digital APIs are available on the internet and do not require a HSCN connection. Older (and commonly used APIs to interface with EHRs) do require a HSCN connection. Check the API network requirements on the API catalogue.

HSCN-only APIs

NHS Digital are working on making all their APIs available on the internet in line with their Internet First policy.

When HSCN is a requirement there are lots of suppliers that offer HSCN connectivity services.

Internet-only APIs

Some internet-only APIs require the end user to be a strongly authenticated healthcare worker, eg via NHS Smartcard

Smartcard requires a HSCN connection, but you can also do strong authentication via NHS CIS2

Testing NHS Digital APIs

For most NHS Digital APIs NHS Digital provide sandbox environments to test your integration before going live.

Sandbox APIs are open-access, don't usually require authorisation and are usually stateless - they return hard-coded responses

Not all APIs have sandbox environments available

Some sandbox environments have limited or pooled synthetic data, and some sandboxes allow you to test using your own data

Quite often testing evidence is required before applying to go live with an integration

For some interfaces, and if your software is using smartcards for user authentication you will need an ODS code, HSCN connection and to request physical smartcards for testing

NHS Digital do not provide environments for performance testing, it is worth noting some APIs (and all RESTful APIs) have rate limits to protect them from overuse

If you have a primary care integration on your roadmap and you are unsure on anything please reach out to me or 6B.

[iampaulbrown]

Interoperability in Healthcare

Interoperability is the ability of health information systems to work together within and across different organisations to advance the effective delivery of healthcare for individuals and communities. 

Interoperability in healthcare is basically about getting healthcare systems to talk to one another:

Disparate software exchanging data between a range of sources including hospitals, GP’s, pharmacies, laboratories and clinics.

Health data exchange architectures, application interfaces and standards enabling data to be accessed and shared appropriately and securely across the complete spectrum of care.

There are many health and social care systems used across healthcare settings today. Bi-directional connectivity between systems in real-time, is essential to support health and care professionals in the delivery of safer, improved patient care.

Why is interoperability in healthcare important?

Health data has always been challenging to share because it’s sensitive and requires a high level of privacy and security, and the inability to access it can cause significant harm. But a lack of interoperability can result in an incomplete understanding of an individual’s or population’s health needs, which can lead to poorer outcomes and higher costs.

Interoperability is increasingly critical:

Populations around the world are ageing and people are living longer

The use of electronic health records (EHRs) has increased significantly

With better interoperability we can transform healthcare:

Clinicians achieve a more complete view of patients to drive better care models, better patient safety and better experiences

NHS and Government develops a better understanding of data, including trends, utilisation, demand

Life science organisations can drive faster, more informed research

Why Interface?

The main reason why any organisation should consider interfacing is to improve patient outcomes and quality of care. 

Interoperability gives clinicians a complete and consistent view of patient information and enables them to make better-informed decisions, faster. Patient experience also improves with this as they no longer have to repeat the same information to different healthcare professionals involved in their care.

Ultimately, interfacing enables access to information that informs an individual’s full, longitudinal health story, enabling; clinicians to deliver better care; patients to become active participants in their care plans; and health IT developers to leverage evidence to create systems that improve care delivery.

It is also important to note that Point 17 in the NHS England Service Standard requires a digital health service to be interoperable.

Opportunities to interface

With these benefits in mind, taking a use-case driven approach can frame how information sharing may inform care or business practices and can help frame the potential benefits and drive interoperability forward. The following examples provide a snapshot at how interoperable exchange may occur:

Care delivery data

Consent management

Lab results

Patient intake

Provider alerts

Query services

Record locator services/master patient indexes

Referrals

Secure messaging

Transitions of care

What routes can I take to interface?

Direct vendor integration

Full access to APIs and functionality with reduced throttling

No third parties dependencies in architecture

Direct support from the vendor

Integrations are usually pay to play and can be expensive

NHS Digital API Platform integration

APIs are mostly free to use

API functionality and access is accelerating, with a strong roadmap of upcoming features

Support from NHS Digital. NHSD API management mission is to “make integration easier”, NHSE’s policies are aligned, see Open API policy

There are limitations with some APIs

Integration via a third party proxy

Off the shelf product which can be implemented quickly

A fully managed service which includes implementation, support and maintenance

Third parties are pay to play and can be expensive

No ownership of the technology, reliance on a third party

Useful Interoperability resources

Healthcare Information and Management Systems Society (HIMSS)

The Global Consortium for eHealth Interoperability

Health Level Seven HL7, HL7 FHIR, HL7 UK

Systematized Nomenclature of Medicine (SNOMED) International

Clinical Data Interchange Standards Consortium (CDISC), SNOMED CT (on NHS Digital website)

NHS England Open API Policy

NHS Digital guidance to building healthcare software

NHS Digital API Platform, NHS Digital API catalogue, NHS Digital API Management Product Backlog

[iampaulbrown]

Opportunities for Artificial Intelligence in healthcare

The opportunity for AI in healthcare is an area I am both optimistic and passionate about.

Healthcare has always faced significant challenges, and AI solutions have the potential to do more to help healthcare. So we need to focus on it, the potential benefits to patients, healthcare professionals and the overall health ecosystem are enormous, now more-so than ever:

Healthcare systems are under enormous strain. The World Health Organisation (WHO) estimates an ageing population, increase in diversity of needs and increase in conditions.

Covid-19 has had a significant impact on how healthcare systems operate. Longer waiting lists, a stretched healthcare workforce, and a backlog of procedures.

Mistakes and missed diagnoses occur during delivery of healthcare. Approximately 5% of outpatients are misdiagnosed and there are an estimated 237 million medication errors in England every year.

The volume of healthcare data is accelerating. There is a huge and ever increasing amount of health data being generated daily by both patients, clinicians and via wearables such as smartwatches.

There has never been a better time to focus on AI in healthcare, and we need to accelerate the adoption of good AI in healthcare. To achieve this we first need to understand the significant underlying AI technologies and components that are particularly relevant to healthcare:

Natural Language Processing. NLP enables the processing of human language, from understanding unstructured data in a EHR to enabling speech recognition tools and services.

Computer vision. Perhaps the most common AI utilised in healthcare currently, Computer Vision enables the processing and analysis of visual images. Use cases include classification of images, eg size and shape of lesions or melanoma detection.

Federated learning and synthetic data. Federated learning enables decentralisation of training data, essentially large-scale shared anonymised data sets for training data models. Synthetic data is data generated via statistical or generative models, which is useful when 'real world' data is not available.

There are core processes through the planning, prevention, delivery and management of healthcare where these AI technologies can add value. In particular the focus for new opportunities for AI in healthcare must focus on better patient outcomes and in particular enhancing the delivery of care, for example:

Identifying personalised treatment

Planning personalised patient communications

Identifying high risk groups

Preventing avoidable errors and adverse drug effects

Enhancing investigations and diagnosis of disease

Supporting self management

Supporting triage

Solutions to support remote monitoring

Solutions to support remote care delivery

Scaling of AI adoption in healthcare, and in particular recognising new opportunities for AI in these key areas will be an evolution of healthcare delivery in the coming years, and more successful implementations of AI in healthcare will build confidence, momentum and develop the enablers needed to overcome challenges, including:

Access and availability of curated training data. Increasing interoperability and access to data across healthcare systems.

Developing confidence and trust in the wider public. The emphasis must be placed on the expected patient outcomes and benefits, and reduction of health inequality.

Increased national investment in healthcare AI. Additional investment at a national level must be made to ensure AI projects and tools are successful, and to ensure the right people are engaged to support AI projects.

AI projects are big and challenging, especially in healthcare. The key message to overcome challenges and discover new opportunities for AI in healthcare is to work together. I believe through collaboration and focus AI can help improve how we deliver healthcare, and most importantly improve the lives of patients and their families.

[iampaulbrown]

Web3 and decentralisation

Web3 technology provides increased data security, scalability, and privacy for users, whilst combating the influence of large technology companies. Web3 is (in the most part) over 10 years old, but it is still early days in many ways.

The general ambition for web3 is to ‘decentralise everything’. Web1 (initially) was fully decentralised and web2 centralised everything into platforms, so the hope is web3 will decentralise everything again.

There is good reason why centralised platforms emerged in web2 though:

Web1 wrongly believed everyone would be both a publisher and consumer of content and infrastructure. Each person would have their own web server, their own website, their own mail server and email, but quickly it was realised that is not what people actually want. People don’t want to run their own infrastructure. Even techies don’t want to run their own servers at this point.

A decentralised protocol moves more slowly than a centralised platform. After 30+ years, email is still unencrypted; meanwhile WhatsApp went from unencrypted to e2ee in 1 year. Lots of services are still trying to standardise sharing a video reliably over IRC; meanwhile, Slack lets you create custom reaction emoji based on your face.

In web1 and web2 we found when a service is truly decentralised it becomes very difficult to change. It is a problem we see time and time again for technology, the rest of the ecosystem moves quickly and if a service can’t change fast it will fail.

So we know that when technology does not change quickly, it is a problem, and historically taking a stand-still protocol, centralising it, and then iterating quickly worked well to move things forward.

This might suggest that decentralisation is not actually the most pressing objective for web3.

We know people do not want to run their own servers, and prefer simplicity over complexity, but equally we believe in the ambition of decentralising services and we want to distribute trust (but without having to distribute infrastructure).

I think there are two fundamental truths we need to consider whilst we venture into web3:

To move fast infrastructure may still be centralised, but using decentralised protocols (like cryptography) and teams to distribute trust.

Software projects require a lot of centralised human resources currently, improving our relationship with web3 technology requires distributing trust and distributing teams to make software easier to create.

[iampaulbrown]

Interview: A week in my life

[iampaulbrown]

The hard thing about hard work

If you want to achieve something great, you have to work hard.

But working hard for a sustained period of time is complicated, it is a dynamic balance between putting in a lot of hours and maintaining effectiveness - put in as many hours each day as you can without harming the quality of the result.

The trick to working hard for a sustained time is really to set clearly defined goals, and hold yourself accountable to them. You have to be honest with yourself, you have to not procrastinate (which is a form of lying to yourself), you must not get distracted, and you must not give up when things go wrong.

Once you master holding yourself accountable you have to understand what your limit is, how much hard work can you do before you become ineffective. That limit varies depending on the type of work and the person. My limit for the harder types of work like programming is about five hours a day, whereas in my day-to-day role, I can work all the time.

The only way to find a limit is by crossing it. But you have to be careful, there is nothing admirable about working too hard and getting poor results. You must always remain conscious of the quality of your output, sometimes stopping and changing course completely can be the most productive use of your time.

Finding the limit of working hard is an ongoing process, not something you do just once. Both the difficulty of the work and your ability to do it can vary day to day, so you need to be constantly judging both how hard you're working and how well you're doing.

The hard thing about hard work is the balance between working hard and working smart, but in truth to achieve something great you have to do both.

[iampaulbrown]

Building a fundamentally different company

Too many businesses place too much control with the employer, and not enough with employees. This is especially true when it comes to managing fundamentals, such as people's time.

But if we're part of something we really believe in, we should be trusted to manage our energy and our time.

To do our best work, we need to balance work and rest, and sometimes, the most effective thing to do is to take a break. You can come back energised and having solved problems you couldn't solve by working through them.

More flexibility means more energy which results in a happier team and better results. This is why at 6B we offer full flexibility - flexible hours, flexible location, and unlimited holidays.

Unlimited holidays rocks the boat, but it is the right thing to do. If we are building a company, let's aim to build the best company, and that means fighting for wild ideas that will improve the status quo.

We've been gradually transforming 6B into a fundamentally different company. We are creating an environment where team members are self-directed, motivated and have freedom to take control - allowing people to bring their full selves to work.

[iampaulbrown]

Embracing change, coding less

Recently I have had a challenging realisation, as a CEO you have essentially chosen to never become an expert of anything.

What about development?

It’s weird. I haven’t coded much over the last six months. I have been really busy, but I have not been as hands-on with code as I used to be.

A lot of my life I have been largely involved in development. So it feels odd that in just a year, I can be so distant from it. And that’s exactly how I need to feel. That’s what needs to happen for 6B, and it’s what will help me grow the most, personally.

Repeatedly firing myself

If you’re a founding CEO, I believe that you are probably doing your company a disservice if you don’t fire yourself from your skill position.

But beyond firing yourself from your skill position, you need to keep firing yourself every time you become too busy in one area. For the company to grow and move forwards it needs specialists and the CEO has to remain more of a generalist, with an understanding of each component but focused on the strategy, vision and the future.

Feeling lost, and getting used to it

Being an expert of nothing is draining, and something I never anticipated. There is a lot to do, and you don’t really know how to do any of it. On top of that, you’re supposed to be the leader, to know everything. You’re meant to be the expert that everyone can look to. They're counting on you.

It’s pretty hard at times. There are days when I wonder what it is I even do anymore. Everything used to be so tangible - I would write a line of code, and it would do something for me. The feedback loop for a developer is really tight. These days, there are these fluffy things like culture (and it’s so important), and I have to direct strategy and hire new people. I have to manage much of the team, and I have to learn most of this as I go.

Every day I’m an expert of nothing. And just when I finally start to feel like I know how this role works, and the activities I need to do? That’s exactly the point when I need to hire someone to replace myself, so I can move onto the next thing (and learn how to do it).

I’m starting to find a peace and comfort in this place now. I like change, so I quite like the challenge. It is a real privilege to be able to experience it.

[iampaulbrown]

Four predictions for AI and ML

Recently we have seen the acceleration of technology and science despite the devastation that COVID has had on businesses and industry.

Artificial intelligence (AI) and machine learning (ML) have pushed science faster than ever before. From commercial airlines using machine learning to predict what passengers will order for lunch to ML in the boardroom to better understand how to restart business and reduce risk in a continuously volatile market, and more.

The Race To Dominate In-Cloud AI, CSPs and Hardware

The introduction of cloud computing and the exponential advances that AI researchers have made over the past five to 10 years have changed the playing field, pitting the chip giants' against their biggest customers—CSPs—in the race for cloud AI dominance.

The examples are everywhere – including Google Cloud TPU or Amazon’s many tools in this space. We’ve moved past the cloud wars, and into the cloud AI wars.

Healthcare Will Be Increasingly Reliant On AI

Healthcare is perhaps the industry where AI can add the most value over the next year (and maybe even the next decade). The challenges our global healthcare systems have faced due to Covid-19 have spotlighted this need, and also the opportunity.

Despite regulatory caution, AI will continue to revolutionise healthcare – from reducing administrative complexity to scanning X-rays and MRIs for more immediate results. This increase will be especially beneficial in communities with strained healthcare systems and that lack the resources and staff of well-funded counterparts.

Chip Deals Will Continue To Take Centre Stage

We’ll continue to see big chip deals this year, despite the fact that some governments have objected to such moves in the past.

Governments trying to previously stop similar deals have seemed to be more focused on political reasons than truly being driven by anti-compete reasoning and I don’t believe this will hinder the big players. However, we may see some middle and smaller sized companies avoid doing big deals because they don’t want to cross regulators.

AI Transformation Is The New Digital Transformation

The industries best equipped to embrace digital transformation were those who were digital-first: internet companies, telecom, etc. What industries are likely to embrace AI first? Those with data-driven decision making as central to how they operate: think internet companies, pharmaceuticals, finance, and others.

I expect to see AI introduced to the long tail of industries in myriad ways – airlines will leverage predictive analytics to optimise food consumption, and therefore waste, while hospitality and healthcare companies will look to automate more mundane tasks like data collection and categorisation, freeing up skilled workers for more complex tasks. Across all these industries, what we’ll really see is AI in the boardroom – top executives will embrace new technologies like never before, leading even the most stagnant of industries to begin modernising.

All in all, I am excited to see how AI and ML will be used this year to address a wide variety of global challenges and opportunities across industries.

[iampaulbrown]

Supervised vs unsupervised machine learning

Supervised and unsupervised learning are the two different types of tasks in machine learning. In short supervised learning is done when we have prior knowledge of what the output values for our samples should be, and therefore, the goal of supervised learning is to learn a function that best approximates the relationship between input and output observable in the data. Unsupervised learning, on the other hand, does not have labeled outputs, so its goal is to infer the natural structure present within a set of data points.

Supervised Learning

Supervised learning is typically done in the context of classification, when we want to map input to output labels, or regression, when we want to map input to a continuous output. Common algorithms in supervised learning include logistic regression, naive bayes, support vector machines, artificial neural networks, and random forests. In both regression and classification, the goal is to find specific relationships or structure in the input data that allow us to effectively produce correct output data. 

When conducting supervised learning, the main considerations are model complexity, and the bias-variance tradeoff. Note that both of these are interrelated.

Model complexity refers to the complexity of the function you are attempting to learn - similar to the degree of a polynomial. The proper level of model complexity is generally determined by the nature of your training data. If you have a small amount of data, or if your data is not uniformly spread throughout different possible scenarios, you should opt for a low-complexity model. This is because a high-complexity model will overfit if used on a small number of data points. 

The bias-variance tradeoff also relates to model generalisation. In any model, there is a balance between bias, which is the constant error term, and variance, which is the amount by which the error may vary between different training sets. So, high bias and low variance would be a model that is consistently wrong 20% of the time, whereas a low bias and high variance model would be a model that can be wrong anywhere from 5%-50% of the time, depending on the data used to train it.

Unsupervised Learning

The most common tasks within unsupervised learning are clustering, representation learning, and density estimation. In all of these cases, we wish to learn the inherent structure of our data without using explicitly-provided labels. Some common algorithms include k-means clustering, principal component analysis, and auto-encoders. Since no labels are provided, there is no specific way to compare model performance in most unsupervised learning methods.

Two common use-cases for unsupervised learning are exploratory analysis and dimensionality reduction.

Unsupervised learning is very useful in exploratory analysis because it can automatically identify structure in data. For example, if an analyst were trying to segment consumers, unsupervised clustering methods would be a great starting point for their analysis. In situations where it is either impossible or impractical for a human to propose trends in the data, unsupervised learning can provide initial insights that can then be used to test individual hypotheses.

Dimensionality reduction, which refers to the methods used to represent data using less columns or features, can be accomplished through unsupervised methods. In representation learning, we wish to learn relationships between individual features, allowing us to represent our data using the latent features that interrelate our initial features. This sparse latent structure is often represented using far fewer features than we started with, so it can make further data processing much less intensive, and can eliminate redundant features.

So in summary:

Supervised: All data is labeled and the algorithms learn to predict the output from the input data.

Unsupervised: All data is unlabelled and the algorithms learn to inherent structure from the input data.