Silos on the rise
Despite recent mention of collaboration in e-health agendas, we still seem to be seeing old silos perpetuate or even new silos being erected in Australia. Why is this so? I think it is because we often don’t recognise a silo until we bump into it personally. Even the rush to embrace e-health is itself, in some ways, a manifestation of this symptom. Although it offers to improve information flow, it potentially disenfranchises those who are unwilling or unable to participate, possibly resulting in poorer healthcare for those on the wrong side of the digital divide [ apropos the maxim “first, do no harm”].
Some recent comments on this blog around Professor Enrico Coeira’s suggestion for a “co-production” approach to e-health has led me to delve deeper into silos, and the deeper I look the more silos I see.
I see silos not just in the organisational fabric, of which there are countless examples, but possibly even more so in the technical and informational aspects of e-health. I see information silos resulting from standards and terminology depressingly difficult to configure and deploy into clinical systems; or technology silos created by well-intentioned, thorough, pragmatic, but often poorly engineered initiatives such as Integrating the Healthcare Enterprise (IHE).
We all create silos, often unwittingly or through laziness. We build silos when we communicate through proprietary rather than open standards. A classic (although probably seen as trite by many of you) example is the thoughtless proliferation of Microsoft Office Word documents instead of PDF or HTML or ODF or plain text [ “oh, sorry, I just assumed you use Microsoft Windows”] ; or when we use proprietary channels of communication instead of open ones.
We do not break down silos by putting doors in them, and then handing out the keys to a few chosen colleagues! At a minimum, we need open doors and windows. That doesn’t mean completely throwing away control. But it does mean gaining and giving a lot more trust.
The so-called “e-health revolution” is primarily centred around information – and its better sharing – variants of the oft touted “Right information about the right person, to the right person at the right time”. Sure, the goal is better care resulting from that information, but it is the information itself, and its reusability, that is at ( or should be at) the centre of e-health. Without it, and without it in a reliable, consistent, reusable and safe form, there is no e-health, just noise. Information is potential gold. Just witness the insurance companies wishing to get hold of it on the one hand, and the passionate pleas for privacy protection from many consumers on the other.
Yet so often we pay far too little heed to the nature of this gold. We let sectional interests dictate what information should be collected. For medication management, I’ve witnessed designers of data sets preclude the capture of little nuggets such as “reason for prescribing”, because they [ one or two pharmacists ] don’t need it. I’ve encountered a senior clinician veto the recording of patients’ blood groups. In recent weeks we have seen the establishment of silos around pathology data, prescribing what information should be collected and what terminologies should be used for pathology test orders and results – satisfying someone’s dreaded, limited set of “use cases”.
I’m not trying to single out pharmacists or pathologists for special admonishment. I’m sure similar siloed approaches exist in most quarters. I simply wish to emphasise the point that clinical information is precious, it is the very essence of e-health. Its quantity, quality and reusability is paramount. It’s existence transcends the context in which it is created. Yet that context needs to be preserved alongside. This requires well designed, reusable, clinically safe information containers and patterns.
Clinical information is for the ultimate benefit of patients and consumers. We should continually be on the lookout for ways to maximise the potential of that information – for everyone’s health and wellbeing.
We need to be far less vertical in our thinking, stop erecting those damn silos and maybe then we can start mining the gold!
If not IHE then what interoperability standards should health organisations be looking at? From my research IHE is the only comprehensive health interoperability standard available.
Excellent question. I only wish I had an answer! I think we are a long way from a “comprehensive health interoperability standard”. I’m not convinced we need a single one, in any case. But I do think that clinicians, and thus all of us, have been let down in this area to date. Partly, because we talk about “health interoperability”, when I suspect what is really meant is “health information interoperability” ?
In any case, I would not agree that Integrating the Healthcare Enterprise (IHE) is an interoperability standard of any description.
I think I probably would acknowledge that IHE is perhaps the most comprehensive collection of s*******ds-based specifications for health information exchange, using the word standard very loosely. IHE do not, nor do they proclaim to produce standards.
IHE is an organisation that has a pragmatic approach to health information exchange, whereby it focuses independently on each specific health information exchange requirement and applies (at least usually) the predominant extant protocols and standards used in that domain to meet that specific exchange requirement. It doesn’t produce standards, it anoints, adopts and constrains them to solve a problem.
I think it does that really well. The participants roll up their sleeves and get on with it. And they test out each profile in a controlled test – almost a faint hint of clinical trial, there. IHE has produced a lot in a short time, is often industry led, and publishes its profiles openly and freely. There’s good rigour in their profiles. They do well at nailing down the ambiguities that plague many of the standards they adopt. All in all, it appears to be a good, successful, cooperative model. I’m impressed by their output. And quite, rightly, they are attracting plenty of interest.
However, I do not believe this is the ideal approach for information interoperability, particularly where shared electronic health records are concerned.
If you look at the rapidly expanding collection of profiles that IHE are producing, there is much heterogeneity, particularly in information structures and terminologies. They are, as I imply in my article, building silos. Silos that will be hard to destruct when the time comes to share information across the health domains covered by them. If we try to put all this information together, what will we get? We’ll get a dog’s breakfast. That’s what we’ll get. That’s not interoperability in my book!
With IHE, there is no overarching information architecture. There is no engineering. Perhaps there will be quick wins. Perhaps by adopting their profiles it will be easier for industry to send their data in a consistent fashion for each particular domain.
But ultimately, the test will be if the information can be successfully, reliably and safely reused, not whether (as is IHE’s prime focus) it can be transported once between A and B for a specific short term purpose.
I’ve generalised above when talking about IHE, because their products are a collection of separate and diverse things. It is hard to do justice to the arguments, let alone begin to try to answer your question in a short blog comment. I think its a great discussion to have, and to have broadly and openly ( and not just somewhere deep in the bowels of NEHTA!). That will give plenty of opportunity for those who wish to do so, to say things like “perfection is the enemy of the good”, or “at least with IHE we can move forward” or ….
It is only right and sensible that these and many other views can be aired and shared. That’s what we would have in a democracy. It’s just so much harder when we only have an authority.