evolution, revolution and NEHTA
There is a wide spectrum of opinion in Australia regarding how e-health should be progressed. At one end of this spectrum we have the proponents of national-scale, top-down architected and developed infrastructure, such as exemplified by HealthConnect and by the NEHTA work program. At the other end, we have those who believe that such national-scale, government-run projects are doomed to fail, have been shown to fail around the world, take forever, and cost an order (or several orders) of magnitude more than organically evolving, bottom-up, industry-led alternatives. Where does the truth lie?
Many cite “the internet”, as an example of a successful, complex, evolving ecosystem that was not designed top-down by a government. However, some parts of it were. What is clear, is that the internet’s success is based on many soundly engineered components. These components are based on standards, modularity, clear separation of concerns, ease of implementability, relatively simple conformance requirements, and well tested implementations. This notion of soundly engineered components is one that is poorly understood or acknowledged in e-health, but one that needs to be if substantial progress is to be made.
Beyond sound engineering, there is another interesting facet to development of internet components – that of cost/benefit. With the evolution of the internet, it was the people who put in the investment who tended to reap the benefits. This is far from the case in e-health. Why would a pathology laboratory bother to improve the way information is electronically dispatched, if there is a substantial cost and little benefit? Once they have reaped the efficiency gains from getting their results out the metaphorical electronic door, why spend more? Why would a vendor of GP software invest in incorporating an elaborate, incomplete, complex SNOMED terminology into its product? Who benefits? Why would a hospital pay even more for a new Patient Administration System or Clinical Information System, than they currently have to pay, in order to have a raft of complex, bespoke interfaces to exchange data with external systems? Why would the Board pay for such extravagances at the expense of hospital beds? In all of these cases, the primary, and often sole beneficiary is the patient.
The internet grew remarkably slowly until reaching a critical mass of connectivity and information flow, where each individual or organisation participating in the information flow reaped a direct and usually immediate benefit. This is true whether we are looking up an on-line bus timetable, downloading a video, sending an email, or buying a new book. But with e-health, we are a long, long way from that. We may never have such an equivalent paradigm. So, if we are to herald in a new e-health world in the timeframe most people are hoping or (worse still) expecting(!), then an organic, industry-led, bottom-up approach will not be sufficient. Intervention will be required. Innovative investment approaches will be needed. Leadership will be required. And good engineering will be required. Wither NEHTA?!
Hi Eric, these are the dilemmas explored by Enrico Coiera in “Building a National Health IT System from the Middle Out,” Journal of the American Medical Informatics Association 16, no. 3 (May 2009): 271-273.
Though, as Greenhalgh pointed out at the ehealthcentral blog, “middle out” is becoming a cliché. The trouble is, what is the middle? The top and bottom are relatively easy to delineate, but in the “middle out” approach, who’s to say when there is too much top or too much bottom?
Rather than being top down, it seems that NeHTA is more middle out. The NPfIT is top down, and NeHTA is clearly not trying to emulate that approach. But of course, that’s the argument. Is NeHTA’s interoperability framework too top down? Or an absolute necessity to enable local innovation that is nationally applicable? As you point out, the internet and WWW were, in part, only made possible by respected bodies that ruled on disputes and determined overall architecture. I suppose the question is whether NeHTA is fulfilling this role effectively or not.
I know there are many who find NEHTA’s interoperability framework too academic. But coming from an engineering background I find it useful. I don’t believe it encompasses, nor is intended to encompass, a set of pre-requisites for organisations to adopt.
I do think it provides guidance on how to approach interoperability issues, and offers CIOs, departmental bureaucrats, advisors to GP network organisations, etc. a fresh perspective (even tool) to bring to their strategic planning and procurement processes and to the evaluation of new products and services for their organisations.
As for NEHTA’s role, I don’t think that has ever been clear to me or many others, despite what might appear in their Strategy document – and I was part of the organisation for the first couple of years!
What people expect of NEHTA, what they claim is their role, and what they deliver, often seem to me rather orthogonal. Would that it were otherwise.
The central idea of middle out for me is that e-health must be a *co-production* of industry, health service providers, and government. And this is the bit I just don’t see happening.
As Trish Greenhalgh so eloquently put, consultation is not a tick box. If you are an organisation in the e-health space and are ‘consulting others ‘ to make sure your strategy is right, then you are probably going top down (albeit at least with a recognition that others exist).
Middle out is about setting the agenda co-operatively from the start – consultation is not necessary because everyone is already in the room. These aren’t just words or cliches, they are a mind set.
Hi Enrico, perhaps the co-production is not happening because it seems that the sociological insights yourself and Greenhalgh talk about are often ignored. How could an ehealth mindset be developed given that cultural, political and business forces are normally pulling in different directions?
Perhaps users (consumers, clincians and administrators) could, somehow, be empowered to participate more effectively? And that we all start to favour small over big, incremental over big bang, modular over monolithic and interoperable over self-contained?
Perhaps there is a distinction to be made between willingness and ableness. The people I routinely encounter at meetings, various fora and conferences are – without exception – skilled, open, enthusiastic and oftentimes very passionate about working collaboratively and co-operatively. In other words I see a real desire to *co-produce* those sustainable, interoperable solutions (or even part solutions and make a small contribution to the larger endeavour); all of this on offer, coming from our health informatics practitioner community.
They know what needs to be done, and willingly admit – or emphatically insist – that achieving our ehealth goals won’t happen otherwise.
It appears not to be mindset problem, at least from this sector of the AU health community.
Rather, almost all these folk are encumbered by ‘governance mechanisms’ (for want of a better description): legal contracts, confidentiality agreements, Govt policies, jurisdictional or organisational agendas and priorities and budgets, proprietary interests or licensing obligations, IP concerns and other copyrights or patents etc. And importantly, procurement practices which insist upon competitive tendering for products and services (where the ‘winner’ takes all and we all ‘lose’). How many more large-scale, high-cost, single-supplier, infrastrutcure roll-outs do we need to see fail miserably before we recognise that it’s just not working and there are no silver bullets for us to buy off the shelf?
You’ll gather that the social strata I operate within has little or no influence over any those matters.
We’re in the middle, and we can’t get out!
So while I agree Enrico that a change in mindset is necessary, that in itself might have to come ‘from the top’ in order to empower the very people who might be able to help (as Brendon comments).
Do we think that’s likely?