xml-solutions blog

…but nothing you can trust.

There is currently much debate about how best to deliver effective Electronic Health Records or EHRs. In particular this debate (at least in England) seems to be focussed on whether central or local systems are most likely to succeed. As my colleague Joseph Waller points out the argument is moot and is a distraction from the inevitable reality: one way or another integrated healthcare systems will become ubiquitous. As such it is almost irrelevant how the information is stored - perhaps we should move on to worry about what information is being stored. (more…)

As described in the recent blog by Dr. Michael Odling-Smee, there has been an immense amount of discussion in the press as to whether centralised or de-centralised IT systems are better (In his blog Mike refers to the commonly mis-used term ‘local’. Here I want to use the term de-centralised as a less emotive and more precise description of alternative approaches). Not wanting to make too high a claim for the progress of my brethren in the IT Industry, I nonetheless recently settled on God as a suitable analogy for conveying to the lay person the problems inherent in the question “Should national solutions for IT be centralised or distributed?”. For me, asking whether a modern IT system should be centralised is like asking whether God is better centralised or distributed, the question is an oversimplification which makes little sense. Furthermore, it is a very dangerous oversimplification. (more…)

It seems that large IT programmes for whatever reason seem to invariably involve very large monolithic releases that slowly deliver layers of functionality in a number of phases. There is nothing inherently wrong with this approach whilst delivering against an initial contract, but the nature of large programmes is that the behaviour during these long life-cycle releases becomes very heavily bureaucratic and the process, whilst a requirement for robustness during this period, becomes the very thing that restricts the programme’s ability to move forward post contract, or maybe even to manage change during this time. It is often the case that by the time usable user facing systems are built on top of the heavy business logic that constitutes the truly expensive bits, real world requirements for these will have changed since the original vision. Whether this is as a result of elaboration, or changes in the real world is irrelevant – the fact is, it happens. (more…)

Recently in the UK specialist and main stream media there has been much debate about the architecture of the English NHS National Programme for IT. Much of the negative press has focused on the centralised databases - in particular the Summary Care Record - that are hosted in a component of the national architecture known as the Spine. If we are to believe the press these central databases are responsible for a whole host of sins - cost, implementation delays, lack of competition, security lapses, patient consent confusion, privacy issues and many more. Instead of the central model, so the arguments go, we need local solutions. Furthermore - we are told - local solutions will stimulate innovation, reduce cost and speed up implementation. What most commentators fail to define is what local actually means. (more…)