PCEHR. How not to build an Information System

You would have thought the most obvious thing to do when building an Information System is to have at least some understanding of the information you want in it.

Not the PCEHR.

As I explained in my unsolicited submission to the PCEHR review team:

My opinion is that, in the case of the PCEHR, the root cause is a simplistic approach to the “problem” of Health Information. This problem has not been identified or analysed and its solution has not been defined. The PCEHR has been treated as an IT system not as a Health Information System. This is not unusual in large scale IT projects. The large costs are in the technology and the project and so they attract the attention of senior managers and project managers. To them information is just the stuff that goes into and comes out of the IT. There is no direct cost associated with information.

What senior managers and project managers fail to understand is that the value of the system lies in the information, how it is defined, managed and processed. There is no value in the technology, only cost.

It is worth examining the NEHTA document, High Level System Architecture, PCEHR, Final, v 1.35, November 2011. This is supposed to be a definitive description of the PCEHR system. Unfortunately it is silent on the topic of the Health Information that the system is supposed to be managing.

As an absolute minimum there should be an Information Architecture, Entity Relationship Diagrams and Data Flow Diagrams at both the conceptual and logical levels.

These documents should cover, not only the information within the PCEHR but the broader context including information in other systems and interface requirements including, but not limited to standards. There should also be discussions on information ownership, privacy, security, legal issues, data accuracy, data matching and a full description of the lifecycle of health information. Some, but nowhere near all, of these have been raised and discussed individually and from a technical perspective, but not in a comprehensive, holistic manner. Given that all these issues are inter-related, it is not possible to deal satisfactorily with them separately; they need to be considered holistically.

Without these artefacts the rest of the documentation is useless. The High Level System Architecture contains none of these, there are no references to other documents which might contain them and there is no evidence of any such documents on the NEHTA website or anywhere else.

As a highly experienced, professional system developer and an IT architect certified to international standards, my opinion of this document, and other architecture documents published by NEHTA, is that they are woefully inadequate and demonstrate a total lack of competence when it comes to understanding Health Information.

The lack of attention to Health Information means that an Information System has been created without an understanding of the information within that system. The consequences will (not might, but will) be significant rework as they try to correct for the failings in the fundamental design; errors in the system; a failure to meet the needs and requirements of users; and breaches of security and privacy.

This failure to understand what information problem the PCEHR is supposed to address is just one of many failings of this initiative; however it is the most important and is the one that will cause the most trouble, assuming that the PCEHR is not cancelled.

My full submission is here
PCEHRreviewsubmissionBRobertson-Dunn.pdf

And if they claim that they have analysed the information and have the documents to prove it, but not published them, then my reaction would be: Why have you have published a High Level System Architecture but without any description of the information? What did you think that would achieve?

Telling people about function but not data is totally meaningless.

Leave a Reply