[Update with additional Links ahead of NHS England/Intellect workshop on 8 July]
There is growing support for the production of a NHS version of the VistA the US open source EPR used by the US Veterans Administration and more recently by a number of health systems outside of the US.
I’m a tremendous fan of what the VA have achieved and believe we have lots to learn from VistA, but on balance I don’t think producing an “NHS VistA” is a good thing. Most of my reasoning for this is laid out in my earlier blog (if you have not read this read it first), in this I remain sitting on the fence and I now think it is time to come off firmly on the negative side and I want to explain why.
The NHS VA has achieved an amazing transformation from “basket case” to “best care anywhere” VistA was a necessary but far from sufficient driver of this transformation, which was down to truly inspirational leadership and a broader system redesign – Read Phillip Longman’s book “Best Care Anywhere”
The VA had reach rock bottom (loosing patients and eventually finding them dead in the hospital grounds). The NHS has it failings but is far from a “basket case” sitting as it does higher on the global healthcare quality/value league tables.
I’ve had a lot of experience of the localisations challenges associated with bring US software to the UK. These are massive and always wildly underestimated. While the core science of medicine is much the same across all western developed healthcare systems there is considerable variation in clinical practice, structure of services, terminology, units of measurement ,finance and administration that have a fundamental impact on the architecture and content models of an EPR
The core of VistA sit on 1970’s technology and those who have delved in the source code will confirm it is of Byzantium complexity Maybe not a problem if you just have to tweek it and can wrap it in a modern API to shield most developers from the spaghetti, but not an attractive option given the localisation challenge which will force substantial changes to the core.
My view is that rather than invest in producing an NHS VistA that our resources would be more effectively deployed on developing alternative open source approaches based on current thinking about health ecosystem architectures and contemporary tools and in particularly those that already have a foothold in the NHS.
What we need to take for the VA and VistA is not the VistA code, but the approach. In particular, we need to understand the importance of inspirational clinical leadership, the critical importance of getting frontline engagement in the development of healthcare software and the role of an open ecosystem in creating this engagement.
In my view an open-ecosystem has to take an open-source based approach to providing its key infrastructural components to avoid vendor lock-in through IPR ownership, but beyond this both proprietary and open-source components have a role to play and the choice should be left to end users and the patients they serve.
As well as my original blog on VistA other post to my blog have relevance to these discussions including:
For some more background see the following:
VistA Modernisation Report – Report commissioned by the VA on Modernisation options for VistA
A long report but this recommendation struck me:
“VistA is currently deployed to a small community of public, private and international users outside of the VA. However, because it is very difficult to operate and expensive to modify it has not had a much wider adoption. We recommend that VistA be used as a functional specification and be completely reengineered within the VistA 2.0 Open-source, Open-standards Ecosystem as recommended by this working group so that a much wider community can adopt and extend it more readily.” – Their emphasis not mine.
Not sure where this led but some work seems to have been commissioned with Harris Healthcare
Other sites where I have also made a contribution are worth a look