NHS 10 Year Plan
The long term plan asks a lot: care shifted into communities, a neighbourhood health model, AI embedded at the point of clinical decision making. Every one of those ambitions has an integration problem sitting underneath it and usually several. Patient records that can’t follow the patient. Referrals that fracture at every organisational boundary. Waiting list data that exists in a dozen places and can’t be aggregated into one. None of these are clinical failures. They’re integration failures, and they quietly compound every inefficiency the NHS is trying to fix.
Any priority that touches more than one organisation, setting, or team has an integration event at its core. You either engineer it on purpose, or you inherit it as technical debt. There is no third option.
What that looks like in practice
- For a clinician, integration is the difference between a decision made on a complete picture and one made on fragments. Results, medications, allergies, and the care plan, in the same place at the same moment.
- For a patient, it’s care that travels with them. No re-telling the history. No referrals that disappear into a void. Discharges that actually connect to what happens next.
- For operations, it’s cleaner billing, less manual reconciliation, and audit trails that assemble themselves rather than being chased after the fact.
- For an ICS, it’s the difference between population health as an aspiration and population health as a workable dataset. The data almost always exists. Integration is what unlocks it.
Where integration breaks
Real NHS environments tend to fail in predictable ways, and each failure has a recognisable technical shape. Records fragmented across acute, primary, and community providers usually need a shared care record backbone on HL7 FHIR R4, with event driven synchronisation so systems don’t drift out of step. Referrals that stall at organisational boundaries need orchestrated workflow across EPRs, with routing, acknowledgement, retry, and audit built in rather than bolted on. Clinicians losing hours to manual re-entry need event driven task and notification routing, so the system does the context switching instead of the human. Population analytics blocked by incompatible source schemas need a canonical model and semantic normalisation, not a doomed effort to get every vendor to agree. And digital front doors that can’t reach the clinical systems behind them need an API first layer, OAuth 2.0, and SMART on FHIR so access is governed rather than improvised.
FHIR isn’t the finish line
The NHS has committed to FHIR as its interoperability standard, and that’s the right call. But FHIR compliance on its own doesn’t make you interoperable. It makes you technically compatible, which is a different thing. The work that actually matters is adopting NHS Digital’s FHIR profiles, registering with the API catalogue, implementing CIS2 authentication, and connecting properly to PDS, EPS, and the Summary Care Record through their published APIs.
And integration governance is clinical safety governance. DCB0129 and DCB0160 apply to middleware and to the interfaces it manages. A transformation that silently truncates a drug allergy field isn’t a technical defect. It’s a patient safety event. A retry that delivers the same prescription twice isn’t a bug. It’s a patient safety event. Organisations that build clinical risk management into their integration operating model from day one scale without incident. The ones that retrofit it usually do so in response to one.
Our Expertise
We’re a healthcare interoperability and NHS digital transformation consultant practice. Our work sits at the intersection of NHS 10 Year Plan delivery and the hard technical realities that determine whether those ambitions actually land. Most of our clients sit somewhere on a spectrum: a tangle of legacy systems that predate modern standards at one end, new platforms that still won’t talk to each other at the other, and in the middle the more common reality. A mix, where some integrations work and others quietly don’t.
We don’t lead with replacement. Rip and replace is expensive, disruptive, and usually unnecessary. We build around what you already run, to open standards (FHIR chief among them), with security, scalability, and long term maintainability designed in rather than layered on afterwards.
Integration programmes fail for organisational reasons at least as often as technical ones. Stakeholder alignment, change management, and clinical engagement aren’t soft nice to haves. They’re load bearing. We work closely with the people who’ll actually use what we build, because integration that doesn’t work for them doesn’t really work.
See It in Practice
A UK healthcare provider came to us with a heavily customised Microsoft NAV ERP holding years of operational and patient linked data. The system was stable and critical, but its integration options were narrow, every change carried clinical risk, and it couldn’t reliably exchange data with the organisation’s newer platforms. Replacement wasn’t viable.
We delivered NHS ready interoperability without a replacement programme. Data quality improved through automated normalisation at the point of exchange, and the integration layer we built is now a reusable foundation the organisation can extend as its digital estate evolves. The kind of investment the NHS 10 Year Plan demands, and the kind that rarely makes it onto a transformation slide.
Read the full case study → Legacy Healthcare ERP Integration on Microsoft Azure – Microservices Modernisation
