Commitments to open-source technology are enshrined in NHS policy but adoption is frustratingly slow. It is time to spread the word on the new way to collaborate on, co-fund and reuse digital technology, says Made Tech’s Hazel Jones
Open source refers to technology that can be modified and shared because its design is publicly accessible. Opening up digital products so that multiple NHS trusts can use them has potentially huge efficiency benefits for services across the UK. It’s a great approach for risk averse trusts as open-source digital tools have already been invested in, tried and tested and proven to work safely.
As regional organisations across both health and care face similar challenges and opportunities, an open-source approach offers the chance to pool resources and funding to build the same thing once and then share it. This is much more efficient than solving a shared problem many times – region by region. Sounds simple, doesn’t it? And it is simple. So why is there such a seemingly slow and low adoption rate?
Resources and tools
The good news is that the NHS and government have put in the work around open source. There’s lots of resources and tools out there.
The NHS Data Strategy includes two commitments on making code that we produce open and reusable. And the Future of Healthcare policy supports the need for interoperability and openness around standards, culture, technology and code.
The NHS Service Standard design guidance helps team build and run services that improve health outcomes. It also helps us improve experiences of health and care and the efficiency of the health service. We’ve now also got the Open source digital playbook that shares advice and case studies for those embracing open source within the NHS.
All these resources shine a light on the next step in this journey – how we spread awareness and help drive the adoption of open source.
Solutions already exist
We know that in any large organisation and particularly the NHS, priorities are set at the top and cascaded down. Within that, all trusts will have similar challenges that often crossover. And each will undoubtedly have their own solution to those challenges. Often trusts are already focused on current challenges and might not be aware that a relevant solution may already exist.
By continuing to work with organisations like the AHSN Network and FutureNHS who are driving innovation, technology providers can show, discuss and promote their open-source projects. And this is how we can help transform the health service.
Shout about the benefits
It doesn’t end there. We all need to shout about the benefits. Through open source, we’re helping build a future without the fragmented technology and platforms we still see all too often.
A project that highlights the potential of open source is NHS Gloucestershire’s digital support finder. On Your Mind Glos, aimed to get young people to the right support at the right time and improve their experience of accessing mental health support.
The dedicated website guides users through a series of questions to understand how they’re feeling and what support they might need. They’re then signposted to the most relevant service for their needs and given useful information about mental health along the way. It launched in schools alongside a programme of mental health awareness and has reached around 10,000 young people.
While it was developed for young people, other departments within NHS Gloucestershire and outside the organisation have shown interest in the tool. As the existing model has been tried, tested and proven safe to use, it could be adapted for different users, such as vulnerable adults in need of mental health support.
Future collaboration
Building a culture of working in the open will build essential collaborative relationships across hospitals and trusts. But we’ve still got to spread the message of the benefits of open source and how it works, from the top down.
When a digital product is funded with public money, it should be available for the whole NHS to reuse and adapt. Together we can build an eco-system of home-grown NHS and care technology that can be shared safely and freely built on without the dreaded vendor lock in.
Hazel Jones is head of health at Made Tech
Glad this comes up occasionally. We have a mixture of people who are happy spending other peoples money on licences, users who would freak out if you didnt give them Word, people who are scared of support arrangements for open source platforms such as databases or think they are technically inferior, and people who cant think imaginatively about contracts for support.
Having said that it’s not straightforward for a number of reasons. We need to know code is locked down, safe, supported for life of business case, backed up by contract terms, and marketed in such a way you’re not left with the only instance/version of it.
We continue to put money into monolithic products and contracts for extremely old software on extremely old platforms even older than the mainframes were when we laughed at them in the 90s.
Why is this? We seem to be actually blocking CIO ability to challenge these norms. Rather than not getting fired for buying IBM, nowadays you get fired if you don’t. There’s also that thing about trying to make none of it about the IT. Guess what, some of it is!
I remember once getting told by a vendor they did not offer their product on open source platforms because CIOs had said they wouldn’t buy it. In recent times we have moved all of our directory services onto AD because of this relationship with vendors, they just expect to find the AD. Now we cant even implement a simple system even on Linux without paying a Windows licence because it is looking for the domain. It’s probably gone too far.
We even shut down a small NHS Ubuntu project because it had the audacity to use the NHS logo.
There is a lot of open source in use in the NHS actually but a significant change of attitude is required if this is going to be supported in any way in the future. We probably need the next paradigm shift for any significant change
Ade
Open source is the only way! Open source is coming, even if at a snails pace. We just need to embrace it and educate those that are scared of it that open source is very very beneficial to the NHS. It can save millions (and to those that seem to say that it won’t, you need to show me the evidence!). I plan to use open source in all of my projects, are you?
If we could authenticate local access to local apps against the NHS.net, would mean NHS PLC only has to pay once for a license to authenticate the user.
I recognise the challenge and whilst to say it is simple, to implement is not. One step closer to shared networks and systems across ICS/ICB…
Mark
Open Source isn’t just about applications and operating systems.
It covers many more areas of software development including standards, frameworks, pathways, tooling, software languages, etc.
Expanding the definition still has the same issues but we do start to see some successful use of open source. For example: The Digital Playbook
We have several successful uses of IHE Query for Existing Patient Data (and INTEROPen/NHS England CareConnect API) frameworks. This also uses the open HL7 FHIR Standard.
UK Demonstrations of these standards use open source from AWS, openID, SMART, OAuth and HAPI (FHIR).
This has been implemented by Cerner, AllScripts, EPIC, Yorkshire and Humberside Care Record, etc.
It was also use in the NHS COVID Immunisation API’s and behind the vaccination certificate.
This change of definition isn’t going to save money on applications or system licenses but we do see examples of success. It does show we have groups already promoting open collaboration.
Maybe sharing/collaborating on software design, components, processes, standards, etc is a better way to approach open source. It it centred around the decision makers and can happen at many levels in an organisation (decision makers need enable this: allow open collaboration, promote it, etc)
From small acorns grow mighty oaks
Kevin,
Great post. I am working on a project that may tick off a lot you are talking about. HealthENV is designed to encourage networking and collaboration of digital health and care projects. See:
This is fab to read Kevin, thank you
I’m coming at this from a ‘if you are a digital delivery partner supporting NHS, Social Care or Local Gov teams, <we/you> need to be leading the way in terms of being open by default’; so that we are all tiny acorns making inroads within our sphere of influence. I guess I’m hoping if open source (and up-skilling as we go) becomes the standard way of working (for companies like us at Made Tech) - that we’ll be in a ‘more ready’ position with a wealth of open standard projects once the benefits are realised and adoption increases. I totally agree that Open Source isn’t just about applications and operating systems - I have learned from running this by CIO’s that there is a lot more to it - and established procurement processes aren’t really geared for Open Source either. But, if we can do our bit, and start contributing to this future, then hopefully we have played our part and been responsible in our standard practices. I believe.
Great post and I hear you Ade. I recognise your observations too.
Not to mention organisations and key individuals/technical leaders who insist on prejudicially referring to open source as “freeware”,
Yep, I do understand why some companies want to protect commercial IP. But companies that are brought in as delivery teams to develop within NHS (and Public Sector) could help to educate and lead on Open Source as standard. Hopefully one day it will be enough of a norm to not label it at all - it just is!
I’ve heard this too. I think it is around the narrow focus on application and OS.
I am aware of groups, developers and organisations that want to cooperate. It sort of accidentally get blocks by organisational structures. It would be useful to find ways to encourage peer to peer communication to get the ball moving.
I do speak to some of the Dr’s you mention and actually in one of the organisations listed. ← I also have a sandbox EPR and it seems our former boss did also …
Having helped introduce Marand (aka Better) into the UK with OpenEP, I along with others such as Ewan Davis tried to get some Trusts to buy into the model. It was a massive shock to Trusts to think that they were going to help buy into Open Source software and pay for development that may benefit other Trusts.
I was incredulous. Open source clinical systems have since floundered although the word ‘Open’ persists. Even OpenEyes demonstrated that the competitive tension for suppliers of developing an OS EPR (as per Moorfields) doesn’t work.
Sure COTS isn’t great. but unless the NHS procurement culture changes radically, OS will continue to be a marginal solution for niche applications.
Open Source is probably a lot more prevalent than most people give credit for.
There are lots of open source tools and libraries in use in all of the organisations that I deal with including my own. We have long had a strategy of using open source where feasible.
However, it is neither always feasible nor is it always more cost effective. And that is often overlooked.
Worse still, some organisations are using open source to avoid investing in tools and services and this is certainly very wrong. If your organisation relies on an open source tool, service or library, it should be investing in it. Not letting other people carry all the cost.
Absolutely not! Because it certainly is not always the best or most appropriate choice, Nor is it always the most cost effective for the public purse.
Unalloyed enthusiasm aside (and I am an open source developer outside of my professional role), blind faith in any approach is dangerous and misguided. We need to look at the requirements and find the best tools, services and libraries. Where those are open source, that is great, where they are not, that is fine.
And, as mentioned above - where will open source projects get their survival resources from? If we, as the NHS, consume masses of open source without investing, they will all collapse. That cannot be a good or sustainable approach.
In case anyone thinks that I am anti-open source. Nothing could be further from the truth. However, I believe we should be providing the best possible outcomes for the NHS. As such, open source has a part to play, as does open data. But a far more important part is played by “Open Standards”.
I do have a use case which I would love to treat as open source.
In one of my roles (primary care supplier) we are looking at how to automate GMS1 Patient Registration with a GP Surgery.
Currently the registration forms (in our application) is 100% digital however like many others to get this data into a GP system is manual copy and paste task. This adds in time and risk of transcription errors.
We know that many suppliers are also doing this. Some are using screen scrapping (the 90’s term)/robotic process automation (modern term), GMS1 converted to editable PDF for emailing, etc.
It looks like quite a common problem.
We could go down a RPA route but expect that to add a bit of cost (and I know some suppliers and Information Governance have taken issue with this in the past). I would prefer to use an API.
If we approach NHS England (https://developer.community.nhs.uk/t/is-there-an-api-we-could-use-to-register-a-patient), TPP, EMIS (in progress), etc individually we will get our requests rejected.
My musings on an open solution can be found here
https://simplifier.net/guide/Virtually-Healthcare-FHIR-API/Home/Frameworks/Patient-Registration.page.md?version=current
It does also make use of open standards. I’ve gone down a FHIR route but this is mainly to show the gist of the solution (if you don’t like FHIR or are non technical - ignore it).
I would prefer to go down open standards route (and FHIR) as I’m aware of other use cases on the horizon (in another role I have).
Did this ever get published / housed in the restructure?
The NHSX GitHub is still live. There are ideas to merge with NHSE GitHub but it isn’t a priority.
Hi Kevin, Not sure that you would get a request rejected by NHSE - and if you did, you could always reach out to me and I’ll find out why and do my best to get it treated with more respect.
NHSE has hundreds of open source repositories and is a proponent of the approach. We also publish blueprints created by people from all over the NHS as examples of good and best practice. So you should find a pretty open door here.
It’s more around directing the question to the correct group. If you’ve picked the wrong group you may not get the answer you need.
For example: our use case is of this type
But we need EMIS+TPP API’s to allow us to proceed (as we don’t want to buy a RPA system, it’s probably cheaper to leave as a manual system at present).
p.s. I think we actually found the right (NHS Digital) team (in Nov).
It’s quite laughable that people are having to use screen scraping (the new term for RPA) for something that so obviously needs an API. Good luck with it, but I feel people now see the technical work necessary to do proper interfaces as a bit old hat
Ade
That is certainly true! Even before the merger there were far too many different teams and initiatives to keep track of.
However, this is one of the things we need to look at as part of the “new” NHSE. Having clearer entry points for standards, guidance, policies, etc. Already in discussion with a number of groups about this.
Can’t argue with that! Even so though, there isn’t much point in letting perfection get in the way of practicality. RPA is a fancy term for getting an application to do the actions of a person. So where systems were deployed without proper consideration of future automation, it is a practical and cost-effective tool to tide things over until changes can be made.
Past fragmentation of efforts have often resulted in systems that cannot be integrated and automated. One of the possible (but certainly not guaranteed) benefits of the merged NHS England is the renewed central Architecture functions where we are finally at a point where most people understand that the “centre” should not be building national monoliths but rather helping to guide and share standards and encourage appropriate integration and automation capabilities.
The fairly new Data Standards team and portal along with the Blueprints service are good examples of these.
Press release from NHS England
Click on the
automate the inputting of informationSorry, forgot to remind you about the NHSE RPA service that is already in use in several Trusts. Reach out to the Enterprise Apps team if you haven’t already. Alternatively, I can put you in touch with them.
That has already set up all of the infrastructure required and there are options for self- and centrally- hosted services. Should reduce the cost massively I would think.