Been there, Done that, Doesn’t Work: Veterans Health Administration IT goes back in time

Dr. Tony ShannonIf you have an interest in the worlds of economics, healthcare or technology, here’s a story that’s emerged this month that is worth noting for the record books. In the US, amidst the chaos of the Trump administration, yet another mistake has been made this month. For the record, it is worth noting that the US Department of Veterans Affairs (VA) Veterans Health Administration (VHA) ended up with a contract for a large IT solution for the next 10 years worth about $10 billion as of May 2018. On the face of it that may appear to be unremarkable news: just another big expensive contract for an IT system. Yet there is a part sad/part silly dimension to it that is well worth flagging up at this point.

Some history: The Veterans Health Administration and its transformation

To explain the folly of this recent move, it is worth understanding a bit of background to the VHA.

Whilst the USA may be the richest country in the world and spend more money than most on its healthcare system (c 16/17% of its GDP), it has some of the most mediocre health care outcomes. One of the exceptions to their problematic healthcare system was the VHA which is one of the few publicly funded systems in the USA. The VHA is responsible for the healthcare of military veterans.

It wasn’t always that way. The story of the transformation of the VA from a dysfunctional healthcare system through its renewal as an efficient and effective beacon of healthcare in the US was nicely told in “Best Care Anywhere” by Phillip Longman.

Now I should say that I never worked in the VHA, so am commenting as an outsider, but my understanding is that one of the keys to this turn-around was the leadership by a fellow emergency physician: Dr. Ken Kizer. My understanding is that his leadership was key to the improvements that the VHA saw in the 1990s. Another key part of the transformation was the bottom-up IT innovation at the VHA, that created a comprehensive and extremely powerful Electronic Healthcare Record (EHR) known as Veterans Information Systems and Technology Architecture (VistA).

I won’t use this article to go into the pros and cons of VistA, but a mixture of clinical involvement, user-centered design, iterative development and a unique technology known as Mumps are key elements. If you have a technical interest you are likely to balk at the mention of Mumps, which you are most unlikely to have heard of. Yet don’t dismiss it so fast. From my understanding, and from what I’ve seen, its power stems from a unique combination of computer language and multidimensional database. Sadly, it is not a technology that is widely understood or appreciated.

Anyway, times change, leaders move on etc, and so some years ago the VHA ended up under pressure to make further improvements and refresh its technology. Such is the nature of change.

VHA IT Challenges: Leadership Steps

A range of options were explored for improving VistA in 2010 with a report known as the “VistA Modernization Report: Legacy to Leadership” and as a result, in my humble opinion, the VHA led the healthcare world when, in 2011, it made the brave decision to advocate an Open Source, Open Standards approach to the refactoring of VistA.. As part of this strategy, it established a separate organization to oversee the Open Source efforts: the Open Source Electronic Health Record Alliance (OSEHRA)

It will only be in the fullness of time that the significance of that leadership move will be understood and appreciated, but for a range of reasons, though the intent was noble and right, the execution of this strategy was problematic.

Why those Open Source VistA efforts and the efforts of OSEHRA were not successful remains unclear, but my own considered view is that, given the size, scale and style of the VistA codebase that was Open-Sourced, it was always going to be challenging to build an Open Source community around VistA .

VHA IT Challenges: Mistakes Made

Cynically, one could argue that a key reason for the failure was because the approach taken to open sourcing VistA was more akin to an Open Source “code dump”, i.e. Open Sourcing an existing solution, without ensuring that the technology can be easily picked up by newcomers etc. Indeed, this is one of the challenges faced when any legacy software is repackaged as Open Source.

Additionally, there appeared to be parallel refactoring efforts that, at least from where I was observing this story, were disconnected from building an Open Source community. I’m unclear whether this range of refactoring efforts – in particular, VistA Evolution and eHMP – were meant to translate to a Healthcare Open Platform for the VA and the wider world, but my perception was too much money being thrown at the problem, too many Powerpoint presentations of future architectures, and too few demonstrations of actual Open Source code deliverables to back the refactoring challenge up.

VHA IT Challenges: Opportunities missed

One opportunity that was surely missed, again in my humble opinion was some of the very smartest technology that emerged from the VistA scene, that of Enterprise Web Development by Rob Tweed, whose smart thinking recognised the power of M the database, whilst helping to migrate that technology towards the web age by forging very smart linkage with Javascript, the language of the web.

For reasons I cannot explain, the opportunity to exploit this innovative work was missed by those involved in the VistA refactoring, even though, somewhat ironically, Rob won OSEHRA’s first Innovators Award in 2014. All was not lost, however: I should point out the interesting exploration we made towards an NHS VistA by leveraging that smart stuff in the smallest possible way...

Another key blind spot the VHA had, and indeed the US still has, is in leveraging the power and potential of openEHR. The OpenEHR technology, with its focus on peer-reviewed, clinically-led specifications and standards, presents a new and powerful model for health IT, underpinned by the world’s leading research.

Indeed it is safe to say the vast majority of those involved in health IT in the US is completely oblivious to the international efforts focused around OpenEHR that are, slowly but surely, building an Open Platform that is truly fit for 21st Century Healthcare.

VHA and the DOD: Interoperability dodged

The story, however, becomes even messier and unsatisfactory, courtesy of, the US Department of Defense (DoD) who have close links to the VHA. DoD is responsible for the healthcare of active US military service personnel, but that responsibility switches to the VA when they become Veterans. The linkages and transition of their health records between the DoD and the VA have long been recognized as a key requirement. The DoD’s solution to modernising their health IT was to take totally different approach: they awarded a large contract for a proprietary EHR in about 2015. Whilst both the VHA and DoD had shared some heritage in their earlier HealthIT (the DoDs earlier Composite Health Care System (CHCS) had its roots in VistA ) over the years they had diverged, no doubt for political differences of opinion as well as technical ones.

You won’t be surprised to hear that, for many years, there has been major political pressure to tackle the complex challenge of interoperability (within both the VHA - between the VistA instances spread across the many VA sites within the USA - within the DoD and of course between the VA and DOD). In recent years, all parties have struggled to achieve this. To be fair, this is in common with most other healthcare sectors across the planet.

So, when the VHA IT refactoring efforts started to struggle, combined with the continuing interoperability problems between the VA and DoD, a groupthink started to emerge: one that I have also seen in many other quarters in the past.

Rather than tackle the global interoperability challenge that healthcare faces by understanding, exploring and building a next generation open platform for healthcare, pressure came to bear on the VA to move away from VistA, and instead, replace it with another proprietary IT solution. Not surprisingly, the perceived wisdom was to use the same IT vendor chosen for the DoD.

Now, this post is nothing to do with the particular proprietary IT vendor involved in these deals. However, many of us that have been around for a while have seen poorly informed customers purchase mediocre proprietary technology for years. Most of these vendors over-promise but under-deliver on interoperability, and an increasing number of us know that no one commercial organisation is going to own the platform that transforms healthcare. The reality is that a more open, common-based approach to the platform challenge is needed.

Lessons from Health IT : Steps back in time

So my considered view is that now, in May 2018, the good work of the VHA as a true leader in the field of both healthcare and health IT for many years, is likely to take a step back in time.

The term “step back in time” is deliberate as this folly (polite term) is very reminiscent of the approach used during the failed NHS National Programme for IT of England (where I spent 5 years) of 2002- 2016. Safe to say that the initiative was a failure, huge sums of money were lost, much time was wasted and the NHS is still trying to shake off the chains of the resulting proprietary health IT market. If you look at this latest move by the VHA towards IT, the NHS in England has been there, done that, it doesn’t work.
Thankfully, the NHS is only now, years later, waking up to smarter moves towards an Open Platform approach.

Which brings me quickly to what I and others have meant by an Open Platform in healthcare (a term misused by some in recent years, who sense a paradigm shift in the health IT market is coming and are starting to jump on the bandwagon). My own work and related writings have been towards a blend of open source and open standards for some time, towards what I’ve called The Open Platform that will transform 21st Century Healthcare. To bring this up to date, please take a close look here at the “Defining an Open Platform” paper from the Apperta Foundation (Declaration: I’m a Director of the non profit Ripple Foundation as well as the non-profit Apperta Foundation. Both organisations are working towards an Open Platform in healthcare. I also contributed to that paper).

So I wish the VHA luck with their new proprietary EHR contract, they are going to need it, especially if they wish to avoid the troubles the DoD now find themselves in with their latest EHR efforts. What I honestly expect now is several years of effort by the VA to try to make that contract work before it ultimately grinds to a halt. I can predict from my experience with the NHS National Programme for IT, that the report on why this VHA move to a proprietary EHR ultimately failed will detail a mix of people, process and technology issues. My guess is that this report won’t be written and published until about 2023…

Health IT: the world moves on to an open platform

Beyond this profound mistake, I look forward, in time (hopefully not too long) for the VHA to emerge from this costly experiment and make a move towards the Open Platform table.

The good news is that thanks to the leadership of the VA back in 2011, an increasing number of us are now actively engaged in Open Platform cooperativism around the planet., The push towards an Open Platform in healthcare is well underway, but, sadly, the VA will not be a participant.

So here endeth this story, a mix of sad and silly folly by the VHA in the US, to the tune of some $10 billion.

Or maybe not… :o)

PS To give a taste of our Open Platform future, take a look at this open source and open platform work that has emerged out of the lessons learned by the NHS .., working with a tiny fraction of the funds that the VHA has been playing with and are now planning to spend.

PPS To get a sense of how the Open Platform push will grow, check out our 1% Open Platform push, the related responses we got, and this link to a world leading event.. open.coop.2018.