PCEHR Review

The PCEHR is being reviewed. In six weeks.

Dr David More, who runs the Australian Health IT blog, has been invited to make a submission. David asked for input, and I responded with the following:

David,

I’m glad to hear that the enquiry is at least prepared to listen to views from the outside.

I’ve been looking at the whole eHealth initiative as something that should have the goal of improving health outcomes. I’m not a health professional, however I have had significant experience in a wide variety of system developments and have seen many successful and too many unsuccessful projects.

The similarities between this project and the Human Services Access Card is frightening. The major difference being that the PCEHR was actually implemented and its success or failure will be very easy to measure.

My summary, written to cover what I think are the important points, not for readability.

First the issues:

An eHealth record is more than just IT, however the PCEHR has been implemented as an IT system to automate much of what is already in place. Little or no consideration has been given to how a Health record could change the way health is delivered. The opportunities to improve health processes has been missed. The PCEHR will most likely make future improvements in the delivery of healthcare more difficult because the PCEHR will need to be changed as well.

The measures of the PCEHR are project/IT related have nothing to do with health outcomes. If the PCEHR does not improve health outcomes or does not deliver the same health outcomes at reduced cost then what is it’s justification? There is no (published) justification of the PCEHR based upon health outcomes.

Not enough analysis has been undertaken regarding the data held in the PCEHR, specifically, its accuracy and the control over who can access it.. The access controls in the current system are primitive and are open to abuse. The dichotomy between patient privacy and the needs of health care professionals has not been properly addressed.

In addition, there are no clear processes or procedures for correcting inaccurate data or resolving issues and disputes between patients and health care professionals.

The legal responsibilities of the participants has not been addressed. Who is responsible if inaccurate data is used by a health care professional and the patient is injured or harmed?

The relationship between local and state based record systems is not clear. It has been stated that the data in the PCEHR should not be treated as up to date or complete. Most health professionals will use local systems as a matter of course and because more of their patients will have their data in those systems. What extra benefit is the PCEHR over local systems?

Second, recommendations.

This is not as easy as identifying the issues, because it is not clear from outside what the constraints and drivers are. There is a lot of politics and ego involved in eHealth and this project in particular.

The major lesson to be learned from this whole experience is that the public service learns very slowly, not at the individual level only at the organisational level. Public Servants rarely change, change only comes when new people replace the old.

Some recommendations – I may come up with some more in the next day or so.

Any health record system must be based upon trust. The PCEHR is in danger of losing trust such that a better approach may not gain broad acceptance for many years.

I recommend that the PCEHR be taken off line immediately in order not to risk losing more trust. So far no negative outcomes have occurred. IMHO, there is a high likelihood that some will. If they do it will set eHealth back significantly.

Secondly, the big bang approach has been a failure on many large projects, this is yet another one. As has been pointed out many times and by no less a person than the Public Service Commissioner Lynelle Briggs, health is a wicked problem.

I recommend that whoever is charged with sorting this mess out at least reads “Tackling wicked problems : A public policy perspective” available here http://www.apsc.gov.au/__data/assets/pdf_file/0005/6386/wickedproblems.pdf

I recommend a better approach with less risk and more opportunities to get it right. Start small and grow over time, using experiences gained from experiments and trial and error.

Whatever approach is decided on should focus on health processes and information, not technology

The approach should have health outcomes as the clearly specified goals and should detail how those outcomes are to be achieved.

The biggest question is “who should take this forward?”. It’s almost worth recommending that no one involved up to now should have anything to do with it in the future, but that’s a bit much to expect when you’re dealing with the federal government.

Notwithstanding that this is unlikely to happen, the second best thing is to make the project highly open and transparent. There are plenty of people who have far more experience developing and delivering such systems than could be employed by either NEHTA and/or Health.

Why don’t they try something really original and make public all project deliverables up to and including any RFT materials? That way they can elicit feedback from a wide range of, often very expensive, skilled resources. If it becomes unworkable, then that will be a sign that the problem is too difficult to resolve and the whole process can be reviewed and/or stopped.

In my experience, there are two major areas where things go wrong. Decisions are made in the wrong order and things are left out. Those who know what they are doing and have done it before can spot these things in minutes. Those who have never done it before won’t know and may never know. Getting outside help, if handled correctly, could result in a much better process.

And a final declaration of interest. I have no relationship with any government department, any vendor, any consultancy, any health organisation or any other commercial interest in the health industry. However, I am an Australian citizen, I’d like to see problems solved properly and I know we can do better than we’ve done so far.

Leave a Reply