Thoughts on the Digital Health Record

Introduction

It is the 21st century, we can access almost every aspect of our personal lives online, but yet we are still unable to access our health information through a single consolidated platform in England.

Conway’s Law outlines the size of the challenge for the NHS.?

“[O]rganizations which design systems (in the broad sense used here) are constrained to produce designs which are copies of the communication structures of these organizations.”

This in simple terms means that an organisations ability to manage applications and data is relative to the organisations structure that hosts these resources.? In real simple terms the structure of the NHS is the major blocker to having a single consolidated health record.? For clarity, the NHS is 7 regions, 42 ICS, 230 NHS Trusts and 6500 GP practices.

This should not be a problem, distributed data is everywhere, and other industries have managed to tackle the open portability challenge.? The Open Banking consortia that have worked together for years on the standards needed to port finance data from one service to another.? This reduces vendor lock-in, enables innovation and improves customer service.? Although the datasets are much broader, the concept is the same.

As they say, ‘we are where we are’ and we won’t ever have a national Electronic Health Record like Northern Ireland so how do we solve this problem?? In order to move towards a consolidated record we must enforce event driven architecture.? This will allow us to leverage well-known software and data patterns so that any digital health record system is always up to date whilst retaining the distribution of data.

The scope of version one of the Digital Health Record will be critical, a focus must be on the core facets of the record, standardise these and deploy the standards to keep them up to date.? The methods I define in this thought paper provide a proposed plan, and allow for onboarding of care settings which adopt the required standards.? A critical mass of vendors and health systems will need to be onboarded to prove the concept of an event driven Digital Health Record.

Architecture, Data and Information Standards are key to making this work, without standards the Digital Health Record will become a mess, and we must prevent that from happening.?

With a successful Digital Health Record we will be able to:

·?????? Manage our appointments in one place using a simple integration mechanism.

·?????? See all of our Results, and understand them using AI tools.

·?????? Be seen and treated from anywhere with a clear summary of information flowing between care settings – to the Digital Health Record ‘Hub’.

·?????? Contribute towards our care plans and take an active role in our health and care.

·?????? Seamlessly transfer our care from one provider to another and ingest elements of the record into local systems.

·?????? Integrate accredited health trackers and enable ‘just in time’ healthcare.

·?????? Leverage the Federated Data Platform to detect deterioration and enable proactive, rather than reactive care.

·?????? Create new markets, providing an accredited marketplace, that can access the Digital Health Record, enhancing the management of our care.

·?????? Provide a place for integrated screening management for all conditions from a single platform.

·?????? Link with large datasets such as genomics and imaging.

Leveraging events will unlock our future health system and move the responsibility of our care to the patient.? An event model reduces technical complexity and provides a platform for the future.? With the new Labour government, and the policy commitment the time is now, but we must get this right.

Where are we?

The NHS has made leaps and bounds on the digitalisation of primary care through the National Programme for IT.? This has led to every GP practice having a digital record for the patient, deployed by two primary care providers (EMIS and TPP).? Secondary care is catching up through the Frontline Digitalisation Programme, although the maturity and consistency of use is still patchy.?

Our primary challenge is that these systems were not designed to flow to a national Digital Health Record and do not provide the modern publication architectures needed to adopt event driven healthcare.? Due to funding, we also have not fully standardised the digitalisation for Mental Health, Community and Social Care (if it even exists at all).? The NHS Number is not universally applied, and there are known data quality issues in the SPINE that need to be fixed.

Standards must be standardised we must “Standardise the Standards” which is an oxymoron.? At this time, we use SNOMED, ICD10/11, OPCS, National Directories, and due to financial pressures, we have dropped investment on the Unified Test List (UTL) and we do not have an agreed national data or information architecture.? Many moons ago InterOPEN were active and were pushing for a ratified NHS set of FHIR profiles.? These profiles, although published, have not been updated for some time and it is unclear what measures we are using to ensure they are being adopted.

Well managed standards are required to develop a Digital Health Record and must be enforced and mandated.? The use of OpenEHR is almost completely missing from our national thinking and architecture – but yet it is the only standard that is clinically led.? If we are to be serious about providing solutions that are fit for clinical practice, they must capture and publish the information that has been universally agreed as useful to the clinician.

The Messaging Exchange for Health and Social Care (MESH) has been upgraded and is now fit for the future, built on AWS.? Disappointingly it is still providing the glue for the transfer of legacy messaging standards using the MESH mailbox architecture from the 1990’s.? Anecdotally the reason why EPIC does not support the Transfer of Care standards is because the solution was originally a CDA wrapped FHIR message.?

Although CDA is longer required, we have not funded any transfer of care (TOC) work for some time. ?Most recently development efforts have moved to Pharmacy First and other pressing programmes, which are developing new standards and approaches for transferring information between pharmacy and primary care.? All of this could be much simpler if we just pushed the data to a single place and ingested it at the point of care.

Nationally we have procured the Federated Data Platform which should not host a Digital Health Record.? The Palantir system is a data operating system that places your data into ontologies, which are then accessible through subtly bespoked open-source methods, published through modernised visualisation and tooling.? There is no doubting Palantir is a great, modern product but it was not designed to support real-time data transfer into a Digital Health Platform – and for architectural reasons never should be.

Most of the national services are now aging and were built at a time when cloud technology didn’t exist.? All of these services are application centric, not service centric, and have been adjusted to support modern ways of working.? Our national services are long overdue an Audit, Architectural and Cost review.? And a clear, consolidated architectural plan is needed.? Conway’s Law plays it part as to why this has not been possible, event post the merger of digital functions.

What would I do?

I think the problem is too big to deliver in the centre.? The Shared Care Record programme demonstrated that if you divide a problem, you can solve it locally.? Taking this principle, I would break down the requirements into regionalised deliverables whilst ensuring the governance enables the regional leads to work together to define the standards needed to consolidate the record ‘virtually’ in the centre.? Each region should be commissioned to deliver the real-time Digital Health Record, ensuring that each regional platform can integrate with the NHS App.

Taking this approach also does now ‘throw the baby out with the bathwater’ and will enable regions to potentially evolve Shared Care Records into regional Digital Health Records – whilst adopting the appropriate standards.? It is vitally important that each region agrees the boundaries and clear requirements.? These requirements will define the core outcome for the Digital Health Record and will ensure that the platform delivers the goals / outcomes defined in the introduction.

In order to achieve this, the governance for the programme will need to deliver and agree following;

·?????? A set of primary use-cases for the DHR and the order they will be delivered, and targets for delivery to prevent another National Programme for IT.

·?????? The information governance to onboard a patient and what security controls are needed to protect the data.

·?????? An agreed Minimum Viable Record data architecture for a patient for the Minimum Viable Product (Alpha).

·?????? The event interoperability standards that vendors must develop to update the records in (near) real time.

·?????? The integration platform and architecture to ensure that the record always persisted, standardised and ordered in the Digital Health Record.

·?????? A set of agreed terminology and information standards to populate the centralised record so that an ‘apple is always an apple’.

·?????? Agreement as to what national services are required to integrate with the Digital Health Record with existing, legacy national services (SPINE, eRS, ePS etc.).

·?????? The agreed API implementations needed to integrate with the NHS App.

To onboard the patients, I would work with the existing IM1 feeds and industry partners to ensure that once a patient is onboarded that any changes to their record will flow to the Digital Health Record as events happen.?? Overtime new events and activities can be embedded into the Digital Health Record, whilst working with partners to enable bi-directional flows back into industry platforms.? This approach will ensure that the patient can take control of facets of their record, reducing administrative overheads in the service, whilst moving the NHS to a Digital First model.

I am hopeful that the current government will focus on this programme, ensure it creates a model for the future, whilst leveraging the appropriate standards.? We must not continue to create national services that are not aligned to the international community, and it is important this problem is owned locally.?

The current NHS architecture is unsafe, now is the time to fix it.