Comments on: Special Report: Interoperability

By: Tim Benson

Tim Benson — Wed, 14 Dec 2022 09:36:59 +0000

In reply to Ian Swanson. The centre must decide on an interoperability language, let us call it X. There will always be at least two translations, which must be 100% correct. From system A to X and from X to system B. Both translations must be exactly right. This is why X is one of the few things that the centre really does need to decide on. X should be FHIR R4 profiles, because they are potentially fit for purpose and everyone else has adopted them. The new (2022) LOINC-SNOMED agreement means that as long as we use either of these it will work. Interoperability has important social components but if the wrong technology choices are made, it will never work.

By: Ian Swanson

Ian Swanson — Mon, 12 Dec 2022 11:17:37 +0000

This degree of standardisation is idealised madness. Given the current range of organisations involved – all of whom will have different legacy systems (or systems that have not been brought up to date to the latest standard as medicine and care evolve)
As such standards may need to be defined rigorously for a limited range of key information which everyone agrees to codify and maintain/convert to the current standard; and a greater range of information which can be organised into critical areas (latest view of citizens “about me”; latest holistic care plan summary; latest treatment or discharge summaries (for those who have these from legacy systems) ; end of life plan etc. Many of these may be expressed as documents with the key coding being the metadata for defining and organising them.

By: Tim Benson

Tim Benson — Mon, 12 Dec 2022 09:45:26 +0000

Interoperability is vital in any integrated health and care service. Progress during the past 25 years has been extemely disappointing. 25 years ago GP-Pathlab links were operational. The disaster was caused by the NHS declaring victory prematurely.
There is widespread confusion between minimum standards abd exact standards. Interoperability demands exact standards (exactly this way), while quality demands minimum standards (you can always do even better).
Interoperability will always fail if you do not get the technology right, which means both terms (codes) and format. Wordwide everyone is regarding FHIR Release 4 Profiles as the basis for format. The recent (2022) agreement between SNOMED and LOINC means that we can look forward to a consensus about the terms too.
Trish Greenhalgh’s NASSS (reasons for Non-adotion, Abandonment and failure to Sysstain, Scale-up and Spread digital health innovations) has seven dimensions (conditions, technology, value proposition, adopters, organisations, wider system and adaption over time). She points out that failure at any one can lead to disaster, failure at two or more means diaster is almost inevitable. In the NHS we have not even sorted the technology properly.
One of the reasons is the way that interoperability is paid for. We should pay for successful interoperability transactions (end-to-end), not for shelf-ware.