Manifesto
A Manifesto for an Open Source GP IT Platform
Opportunity
The decision by CSC to withdraw support for the iSoft primary care products is the most recent in a long line of commercial decisions that have not served the best interest of the NHS, GP practices and their patients.
Decisions like this are the inevitable consequence of a market in which companies and systems are bought and sold and in which decisions are taken in interests of shareholders and remote investors, not users, who as a result are left facing considerable cost and disruption.
There is another way – The Open Source way, which has increasingly demonstrated its ability to deliver software in ways which puts control in the hands of users without the risk of vendor lock-in, while simultaneously creating commercial opportunities and economic growth.
Openness
At a global level the rise of the Internet is the most powerful demonstration of the potential of this approach. In healthcare the success of products like VistA, OSCAR, OpenEHR and SMART Platform demonstrate the potential of the approach. In UK healthcare we have seen a rapid growth and engagement from the grass roots with initiatives including eHealthopensource, HANDI, NHS Hack Day, Digihealth, TheDigitalDoc and projects like OpenEyes, WardWare and the Leeds Open Portal.
Government too has embraced the open source approach in health with support for the initiatives above and with it own projects including the Open Data Platform and The NHS Code4Health initiative.
Integral to the Open Source approach is transparency and open governance driving up quality, ensuring clinical safety, security and sustainability with funders able to see where their money is being spent.
These developments make it the right time for a new initiative, which will ensure that never again are GPs and their patients held to ransom by supplier lock-in, or faced with an expensive and disruptive change. It also provides an opportunity to drive forward an open ecosystem facilitating not only the transformation of health and care but also growth opportunities for the UK life-sciences and IT sectors including those existing vendors with the vision to embrace this new way of working.
How will we do it?
We don’t simply want to replicate existing GP systems, but instead do something radically different: creating an IT ecosystem, initially for general practice, but easily expandable to other care settings. The key components of this will be:
-
An open electronic record repository. This would separate the storage of the record from the applications that manage it and handle the associated work-flows allowing the creation of multiple ‘front ends’ to meet different users needs and facilitating access for secondary uses under patient control.
- A set of open-standard interfaces (APIs) to enable third parties interact with the repository and associated services to provide apps and tools to support the process of care.
-
A set of open source components which would provide the core functionality needed to support a GP practice, which together provide a ‘front-end’ to support the delivery of health and care that puts patients and carers at the centre and all parties have the tools for effective and efficient care in partnership.
This open architecture will greatly reduce the investment required to build software creating an ecosystem of third-parties applications and services, adding value for patients, clinicians and carers in which users and enterprises of all sizes can participate creating a vibrant selection of free and commercial applications.
Philosophy
The overriding principle is that “no single component essential for the operation of the platform should depend on proprietary intellectual property capable of creating vendor lock-in”.
Initially, the focus will be on providing at least one open source implementation of all the key components, but it is expected that there could also be alternative implementations, under a variety of business models.
It is envisaged that core platform development will be carried out by a combination of professional developers working in small agile teams with domain experts and volunteers. This activity will need to be funded by those that otherwise will have to continue to fund the much more expensive and often ineffective proprietary NHS.
With the right support from key decision-makers, and leadership from the front-line of health and care, this approach will secure UK General Practice’s already enviable global leadership in the application of IT at the point of care, catalyse and facilitate the transformation of health care delivery, open-up health data under patient control for the benefit of the UK Life Sciences and create growth opportunities for UK industry.
Never again
-
could a supplier abandon a software platform, leaving its users high and dry and with a serious business continuity problem. With an open source system, we would still have the source code and a community capable of supporting it;
-
will NHS institutions be left with unmaintained and unmaintainable software, due to a supplier having ceased trading. With an open source system, we would still have the source code and a community capable of supporting it;
-
could a supplier unilaterally control the way a software platform develops in future. The user community will be able to control the direction(s) of travel;
-
will users have to make do with anything less than the version of the software that best meets their needs;
-
will users have to put up with ‘Well that’s just the way it works’ in response to functionality issues.
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.
This looks like an extraordinarily brave initiative. Its aims are admirable – I just hope that they are also achievable.
If only we’d got together & done this 10 years ago…
I would be interested to understand how to balance the needs for standards (As defined in the Health & Social Care Act 2012) against the drive for flexibility as propoesed here
well, I’m no techie, but I guess the parallel is, standards = clearly agreed definition of html, flexibility = use whatever browser you like?
except in this case the standard would be eg patient’s home phone number always live in a database table called PHONE-HOME. etc.
but don’t expect this government to support a ‘we’re all in this together lets cooperate’ approach, when we can have the ‘efficiencies’ of private providers as reliable tested and dependable as G4S. [ when I 'efficiencies', of course I mean maximum donations to party funds plus lucrative directorships when I leave parliament ]
“open” and “openness” in terms of standards, data, governance, and even open source are all rapidly becoming buzz-words in government and public sector circles, and we are hoping to be able to capitalise on that trend…. I can understand your cynicism too, but we are quietly optimistic.
Whilst I agree with the principle and am keen to promote competition and openness to force EMIS and SystmOne to listen to their users we should not forget that the market was designed by our government and they are partially responsible, along with those GPs who voted for the last GP contract, for creating the environment that the multinationals exploited. Small and openness are beautiful good luck with your campaign you have my personal support.
Chris Frith
GP Hereford, Taxpayer, EMIS Share Buyer and Chair of EMIS National User Group
Thanks for the support Chris. I view the GPSoC contract model as a significant cause of the contraction and failure of the primary care market. Only opening up, by developing reference implementations of the requirements for GP computing (EPS, QOF, GP2GP) will we begin to see innovation in this space. Without disruption and innovation in this sector the 2015 target of record access will come and go, the desire of the NHSCB for a platform/portal/apps store will not materialise and an opportunity for digital excellence in healthcare will be lost.
I agree, the main requirements for NHS GP systems, above and beyond normal record keeping, are EPS, QOF, GP2GP etc. Without a system that supports these the openGPSoC will be dead in the water.
The development of accredited open implementations of these requirements would be a fantastic step forward.
An admirable cause that has an ambitious goal. Once you look past the stack of opensource tools and eager volunteers you can not hide from the fact that this will take significant development, clinical and commercial skills to pull off. Commercial companies heave been here before with plentiful supplies or resources and have failed despite deploying large dedicated development teams, in the end the functionality demanded by GPSoC and the change adverse customers resulted in defeat and withdrawal.
So how does openGPSoC proposes to succeed where others have failed?.. As welcome as open source software is, its only one small aspect of what would be required.
“Commercial companies heave been here before with plentiful supplies or resources and have failed despite deploying large dedicated development teams”
I would say “Commercial companies heave been here before with plentiful supplies or resources and have failed BECAUSE OF deploying large dedicated development teams”
Anyone who has been around the NHS for long enough knows the inertia that is created around any new project, getting bogged down in managerial bureauocracy, vested interests, the tendering process, and overlarge organisations. I’m not saying we are immune to all of the above, but the open source way does reduce some of it.
If we were writing a commercial platform, we wouldn’t be able to use any open source components (unless they had a fairly liberal open source license). We will be able to use any available open source componenents and standards. And, if we do fail (which is certainly possible) – any progress we DO make will be available for others to build on in the future (instead of being buried at a crossroads at midnight as in the case of commercial failed software)
I not aware that anyone has made a serious attempt to enter the market since TPP did so successfully more than 20 years ago (unless you count early attempts by IDX and ISOFT at the beginning of the NPfIT – which for many reason you should not) So while you are right about the magnitude of the task you are wrong to say others have tried and failed.
The fact that no one has tried for over 20 years so just how moribund the market is. To a significant extent this is because of the barriers to entry created by onerous and ineffective accreditation requirements which serve no one except the accreditation and testing industry they have created and which generate massive costs for suppliers and ultimately the NHS.
OpenGPSoC believes that the current policy drive to open data and open systems creates an opportunity to get Government to remove the barriers to innovation they have created with transparent development and testing under open governance providing a better way of ensuring systems are safe and meet users requirements.
Amongst the reasons for this initiative is the fact that the existing GPSoC Contract needs renegotiating. The prospect of a viable, open source (or NHS owned, if the NHS wanted to fork the project) alternative would provide options which would encourage stringent evaluation of the proposals of the existing suppliers. It may be that that turns out to be a significant benefit of this initiative.
Wonder how this fits in with the revised medical devices directive, as technically GP systems should be CE marked.
Perfect.
This is the only possible solution to the kind of software market failure, that has been going on for years in healthcare, as I pointed out last year at EHI Live : http://www.ehi.co.uk/news/acute-care/7291/trusts-%27at-risk%27-not-using-open-source
This kind of initiative is unlikely to deliver overnight, but over the next few years, people will wonder how NHS IT stumbled on without it.
Even if the initiative doesn’t deliver a platform in the near future( but i hope it does!), it will deliver more openness across healthcare IT – which is what we all want.
Malcolm
Thanks Malcolm, it’s great to have really enthusiastic support such as yours. I’m sure we will need it as the project progresses. It’s been very gratifying to see how many people WANT this project to happen, even those who are mindful of the great difficulty and problems.
Marcus
I’m looking forward to the first build already
Can we do a system for Dentists too ? There is a lot of overlap in the two. Someone started an Open Source system for dentists years ago and I think it has a small footprint in the US but no adopters in the UK AFAIK.
Hi Grant, there is a project called open Dental which you might want to take a look at: http://www.valuedecision.com/openerp-development/ This project was demonstrated at EHI last year.