Anatomy of a PCEHR system

Well, two themes dominated the two day e-health summit in Melbourne this week:-

1. Infectious enthusiasm
2. vagueness of the PCEHR

These two themes actually run hand in glove. Both the vagueness of the PCEHR model and the rampant optimism are in large part due to ignorance. There was an endless supply of fresh consumer, bureaucrat and clinician faces, unjaundiced from the attempts to progress e-health both here and overseas over the past 10 or more years. The summit compere, Peter Couchman must have used the words “exciting”, “impressive” and their ilk dozens of times when singing the praises of presenters and presentations. But amongst the delegates, there was also a silent minority of the skeptical knowledgable and experienced.

It is interesting to ponder how much of the vagueness and the enthusiasm was deliberately manufactured for the event and how much was real. If more sits in the latter camp, then things are worse than even I imagined. But there was plenty of the former, due in large part to Mukesh Haikerwal’s formidable skill at harnessing clinician support and painting a picture of the long term benefits from the deployment and adoption of e-health, rather than focussing on the difficulties (impossibility?)   of implementing a national PCEHR system of any value by July 2012.

One got the feeling that Mukesh doesn’t much care if the July 2012 goal is not met; that the work will not go to waste, and that it is all about making progress towards the noble vision of a better e-health-enabled world that promises so much. In some respects, I agree with him. I certainly think that the process needs believers and pushers of his standing in order to gain significant amounts of investment. In some respects, it is like the stance of Senator Conroy and the National Broadband Network. Never mind the cost, just look in awe at the benefits!

But it sets up a Bushish divide – you’re either with us, or your against us. At the moment there are more on the “with us” side. Over time, that will inevitably change, as the complexities and realities of systems, standards, inertia, clinical workflow, testing, accreditation, authentication mechanisms, terminologies, procurement contracts, engagement, evaluations, etc. all hit. As the myriad chooks come home to roost and brood. This will make the governance even harder to get right, because the current process has shut out the skeptics and the naysayers and has disenfranchised a large number of interested stakeholders, whether they be supportive, antagonistic, skeptical or neutral.

Where I differ with the enthusiasts is in the assumption that the allocated pot of money is a) necessary, and b) sufficient for this PCEHR system. From the little information the public has been given, it is hard to know whether the $467million will buy us a Hornet, a Hercules or a Hindenburg. It is clear that the bulk of the honeypot has been “pre-allocated” to NEHTA to manage the completion of standards and infrastructure – whatever that might entail. And it is pretty clear that DoHA has no idea what that might entail.

From my perspective, the saddest aspect of the secretive approach to date is that we have reached a position, some 19 months out from a delivery date, where there is no published draft architecture, and no statement of the capabilities of the PCEHR system. Just what will the proposed system provide, and what will it make difficult to provide?

A fundamental principle of the PCEHR as envisioned by NEHTA is that information will be distributed in many repositories around the country, and that there will be an “indexing service” that has sufficient information to locate all the information for a given patient in those repositories accessible to systems conforming to some set of yet-to-be-published service interface standards.  There will be a national repository to contain important summaries, initially alerts, current conditions, medications, allergies. There is no detail about how this data will make its way into the national repository, nor what the “national” in “national repository” actually means. There is an implication medications information won’t initially be populated from NEHTA’s national electronic transfer of prescriptions from GP to pharmacy process. From whence else might it come? It appears that pathology and radiology test results are expected to be stored at pathology laboratory repositories, for access as and when requested.

The whole distributed repository architecture raises some serious issues, particularly if national infrastructure, comprising a simple indexing system and simple repository of a small cohort of summary types is proposed.

Firstly, it minimises the cost and capability of national infrastructure, i.e. centralised hardware and application capability. It places most of the burden of service orchestration onto all participating systems. This makes the cost of integration into existing clinical systems very significant and problematic. It places a significant onus of dealing with interoperability issues onto too many systems. It creates a problem of order N x N, rather than a problem of order N. For instance, every system that wants to access pathology test results will potentially have to deal with the nuances of every pathology lab result payload in the country. Such N x N problems have been found impossible to be resolved at the local and regional level and have led to and supported a market in integration engines in Australia and many other countries. How it could work on a national level in Australia is well beyond my imagination.

Secondly, it precludes the use of the PCEHR system for any query-based research. In most cases, all data for a given patient cohort would have to be downloaded into a research system before any queries could be run. Again, each research system would thus have to deal with all interfacing issues for all source systems. It would likely need access to the myriad terminology systems used by each information contributor.

In summary, the PCEHR as it is so vaguely described to date, appears to be a compromise that should not cost too much to initially establish, but will be a difficult and expensive endeavour  in which to participate for established clinical systems. It promises little for improved clinical care, at least in the short term. It offers even less for research, even in the long term.

My question, as with the NBN, is “could more be achieved for less money?“. If the answer is yes, then let’s get a better governance model together that can work to achieve a better outcome. If the answer is no, then let’s just wait for the enthusiasts and optimists to be proven right. But, I guess if you are already an optimist, you don’t even bother to ask the question.