Converge Around the Person, not the Technology
Note — An edited version of this article was published on Digital Health on 31st May 2023
Just over a year ago Tim Ferris¹ announced plans to converge Electronic Patient Records (EPRs) in use across NHS Integrated Care System (ICS) boundaries. Since then, there has been much discussion and debate surrounding what is meant by “convergence” and how an ICS would go about implementing a “managed convergence” strategy.
In March this year, I spoke at the Digital Health ReWired conference where I presented some thoughts on this strategy and proposed a potential alternative approach to implementing it.
The perspective I presented is that convergence should be based around the person, not by simply throwing technology (such as an EPR) at the problem. I recognise there are many ways to deliver the same outcome — just look at the recent KLAS report for example, where there was extreme polarisation between different organisations using the same EPR. It’s not just about the technology. It is also about the adoption and change around it. But by designing outcome-based services that are centred around the person we stand a better chance of success.
This article expands on this idea further and the important role technology and data could have in enabling its delivery.
What problem are we trying to solve?
The current health and care landscape across England is wide and varied. There are many care settings including hospitals within acute, community, specialist, and mental health settings, primary care locations and social care provision. Within each care setting there are various care provider organisations, including statutory NHS Trusts, Primary Care Networks, General Practices, and community interest companies.
The EPRs deployed within and across these care settings are equally wide and varied. For example, in primary care there are a small group of vendors who dominate, which in turn are different to the slightly larger group of vendors who are mainly deployed across the acutes. Care provider organisations have historically had the freedom to procure their own systems, which has led to a diverse mix — both within and across different care settings.
The trouble is none of these systems talk to one another.
This results in a model where information capture and sharing between professionals is sub-optimal, leading to potential patient safety risks, potential costs of duplicate systems and duplicate data, and potential impact on staff having to use multiple different systems.
This is a significant problem for both professionals and service users, especially where the patient pathway has touch points across more than one care setting or indeed across more than one care provider organisation.
Enter Convergence
“The fact that two or more things, ideas, etc. become similar or come together”. Cambridge Dictionary
Given the debacle of the National Programme for IT (NPfIT) there was never going to be any attempt to converge around a national EPR, so the idea of converging at a regional level makes a lot of sense.
However, given the diversity of the NHS landscape, the problem convergence needs to solve is likely to vary across different parts of the country. In heavily urban areas like London for example a lot of people will present across the capital and across ICS boundaries, but in rural geographies there may be less breadth of interaction. Furthermore, some areas may have an economic challenge with a need to buy more for less. In other areas the population demographic may present a pathway challenge where there is a significant engagement across health and social care settings.
In his letter Ferris outlines how “Our primary focus is to achieve universal EPR coverage across all ICSs (i.e.to level up EPR provision)”. He goes on to say, “We are also encouraging ICSs to work towards the managed convergence of EPRs over time, to reduce the number of EPRs across acute care, community services, mental health, ambulance services, primary care, and social care”.
The proposed benefits include “to provide critical, real-time access to all health-related information for caregivers. It will also enable more simplified access for patients to their own data.”
Levelling up EPR provision is of course a good thing. Around 20%² of NHS Trusts today have no EPR, while approximately 60% are going through a process to meet the core level of digitisation. So, supporting these trusts in “levelling up” is a fantastic initiative.
I also think that converging EPRs over time to consolidate the footprint within individual care settings makes sense. This could help rationalise spending, optimise rollout, and reduce the overall number of siloed applications.
But the idea that in doing so will deliver the benefits of “real-time access to ALL health related information” or “simplified access for patients” is — in my opinion — a hugely unrealistic goal.
There are several reasons why I say this, including:
- The patient pathway will often span multiple care settings. I think it’s fair to say that while we may see a single EPR used by all providers within a single care setting, we will never see a single EPR across all care settings. So, even if an ICS does converge all EPRs across acute care, this does not solve a problem for professionals working in primary care, community services, mental health, social care — not to mention the private sector.
- Most ICS’s are at very different levels of maturity and rollout. If an ICS has only one EPR deployed, and the rest of the region is on paper then it makes sense to converge around that EPR. But where individual ICSs have EPRs from different vendors successfully deployed then can the ICS really justify the spend to consolidate into a single EPR?
- A single EPR will never cover all the data within an organisation. Even the biggest adopters of the “mega-suite” EPRs still have an eco-system of other applications and services. For example, one of the largest US health systems who use a market leading EPR claim to spend around 50% of their IT budget on applications and innovations not covered by the EPR.
- This is a technology first delivery approach with a pre-determined view that it will solve the problem. As outlined by the NHS Service Manual people should be at the heart of everything you do and services should be designed for the outcome. In my experience, projects that don’t do this typically fail.
Furthermore, the idea that convergence is aligned to individual vendors introduces a whole new set of risks. For example, all vendors have their own roadmaps. Innovative health systems cannot rely on vendors to deliver what they need. Given the diversity of the NHS there are unique needs that will need tailored solutions to address.
In practical terms, I believe a realistic goal is to achieve some consolidation of EPRs within each of the various care settings. But to solve the problem and deliver the outcome we must go further.
Can we not converge around current healthcare standards then? That’s the point of interoperability, right?
Unfortunately, not.
The classic definition of interoperability is something along the lines of
“the ability of computer systems or software to exchange and make use of information”³
This is precisely what the NHS (and most other global healthcare systems) have been struggling with for years through the adoption of different healthcare standards, including HL7 v2, v3, CDA and most recently various iterations of FHIR. This approach typically involves standardisation at a technical (or syntactic) level, with the clinical content (or semantics) being mapped or translated when sent between systems.
Adoption of these standards is still required and will be an important part of the overall landscape, but unless we agree upon and adopt at a semantic level then stitching together an entire ICS in this manner will not scale.
What about Shared Care Records?
Shared Care Records are great, and they deliver a huge number of benefits, but they are typically a read-only view of what has happened to the patient up to a point in time. To deliver the benefits outlined above and to support regional transformation, an architecture is required that will allow service users and professionals from all care settings to have (governed and secure) real-time read and write access to data.
There is only one way to solve this across an entire ICS.
Converge around person-centred data
With this approach we see individual applications being separated from the data, moving towards the development of a single longitudinal record⁴ for an individual, that is integrated into the existing application landscape. This approach allows persistent data to follow the person — wherever they may be — and not to be stuck inside individual provider systems. In doing so, you provide the foundation for different regions to transform in a way that is appropriate for them.
Key to this approach is a commonly agreed set of information models that are adopted consistently across all ICS’s. If adopting a standard such as openEHR there are a rich library of mature models already available that could support this. This of course is only part of the answer and would need to sit alongside an empowered group of clinical informaticians who would implement and govern the process around these models.
As outlined in the EY Connected Health Cloud paper (see this link) these repositories can sit alongside existing proprietary data stores or as fully independent data stores that natively use the adopted information models.
To ensure early and incremental delivery the entire process should be based around pathway-based use cases. This would ensure ICSs don’t get overwhelmed with trying to get all their data in order in one go. Again, OneLondon is a good example, where they adopted a similar approach and went live with their first pathway in only seven months.
Such a model is illustrated in the diagram below, in which applications across multiple care settings interact with an independent set of person-centred data repositories. Note how some EPR’s have been consolidated within some of the care settings.
In conclusion, managed convergence at an ICS level will require a hybrid approach. Simply rolling out an EPR from an individual vendor may be part of the answer but by itself is not enough. Adopting standards such as FHIR and combining this with a longitudinal data layer will allow data to flow more easily in support of pathways that span multiple organisations and care settings.
Consolidate EPRs within care settings. Comply with standards. Converge around the person.
===
Contributions
Thanks to the following people who have contributed towards this article.
- Matt Cox
- Jack Williams
- Tomaź Gornik
References
1 — In May 2023 it was announced that Tim Ferris will leave his position. While his replacement may have a different delivery strategy the problems to be solved remain the same.
3 — Google search, first result.
4 — This does not imply the data needs to be held in a single, centralised data store. As outlined in the EY paper “Connected Health Cloud” (see this link) this approach enables data to be held at several different levels — including an organisation, ICS, region or national level. The data could even be held at personal level, as outlined in this paper from the Apperta Foundation
