Sunday, March 29, 2015

This All Strongly Suggests The Information Management Controls And Governance Of The PCEHR Are Both Unsafe and Poor.

This amusing - at one level - but more worrying at a more fundamental level - article appeared during the week.
First we had:

GPs labelled ‘meat inspectors’ in e-health gaffe

26th Mar 2015
AN EMBARRASSING fault that labelled PCEHR users “Meat Inspectors” has raised questions of transparency and accountability in the eHealth system.
The issue is understood to have affected a number of GPs who uploaded shared health summaries to the PCEHR over a period of around a fortnight in February. 
The Department of Health today blamed an unidentified medical software vendor.
But e-health experts told MO they believe a government gaffe is responsible for the peculiar fault.
Medical Observer can reveal that representatives of the medical software company Zedmed met in February with a GP who raised the issue.
In an email obtained by MO, a Zedmed representative tells that GP a digital library provided by the government-run National eHealth Transition Authority was responsible.
“NEHTA provides Zedmed with libraries containing occupations that are linked to IDs, in this case the ID given was incorrectly linked as it was pointing to a ‘Meat Inspector’ instead of a ‘General Practitioner’,” the email states. 
NEHTA has since supplied an updated library to fix the problem, which has been fixed in a Zedmed update, the email states. 
The email, sent in late February, also said it was not possible to fix those records which already had the “Meat Inspector” tag on them — something the department disputed when contacted by MO
“Healthcare providers are able to change the incorrect tag by uploading a new document once they have updated their software,” a spokesman said.
While the incident is said to be an embarrassment for the department, it is not thought to be a clinical safety issue. 
But one expert questioned the rigour of the process by which the government approves and authorises software changes. 
Another said it went to the transparency around the PCEHR rollout, which has resulted in the government binding stakeholders to secrecy through non-disclosure agreements. 
Both experts declined to be named.
More here:
There was also some detailed coverage provided here:

The $1 billion e-health record that calls doctors meat inspectors

  • March 26, 2015
  • Sue Dunlevy
CONFIDENCE in the government’s troubled $1 billion e-health record is under further question after a GP found the system was identifying his job as a “meat inspector”.
Former AMA president Dr Mukesh Haikerwal who helped the government design the system before resigning in despair said it was the latest evidence the system wasn’t working.
“If the system is allowed to get roles so wrong, how do we know what it is doing to our health information?” he asks.
Australian Medical Association GP spokesman Dr Brian Morton said he too was disheartened when he tried to access a patient’s e-health record recently.
“I went to upload information on the shared health summary and it said there was no space to put a record,” he said.
The e-health record only has space for 1,000 files but is easily clogged because it records every medical visit and Medicare claim made by the patient.
This sick patient who has renal disease and has frequent visits to hospital had 1,047 document on her record, he said.
 “There should be clinical information on the record not all these useless Medicare interactions,” he said.
The Department of Health said the “meat inspector” classification was “a known issue with a 3rd party software vendor product that connects to the PCEHR system”.
“The software vendor has advised that this issue has been fixed within their product.” the Health Department spokeswoman said.
The system operator has notified the small number of affected healthcare providers of the issue, and provided advice on rectifying the incorrect tagging on PCEHR documents that had already been created, the spokeswoman said.
Fewer than one in ten Australians (2.1 million people) currently has an e-health record even though they were launched in 2012.
And doctors have uploaded just 41,998 shared health summaries onto these records, which means most of the more than 2 million e-health records are empty.
More here:
The number of issues these two articles are raise are legion.
First, just why did NEHTA not come forward when it became aware of the issue and announce how it was being fixed and just what the ‘root cause’ of the error in the reference table occurring? Was this a deliberate bit of nonsense or does this all reveal some systematic maintenance issues or something else? It is pretty clear the whole thing should never have happened.
Second why did it take a high profile clinician to notice the problem before it became public?
Third just where were the breaks in quality control of the information and how can we all be assured this will not happen again?
Fourth just what other, and how many other, issues with the PCEHR have been identified, quietly fixed and not disclosed? How can we all be sure that clinically relevant issues are being properly addressed and disclosed?
Fifth just why would a health record that is meant to be lifelong be limited to 1000 information elements?
Overall what we see in all this are symptoms of secrecy and mismanagement which are reflective of the very poor governance that has bedevilled the whole project since the get-go. It has been a long running saga of failed delivery and lack of quality, consultation  and transparency.
The time for the full audit of the whole system is surely upon us!


Anonymous said...

So if this is a result of a digital library supplied by NEHTA, which would,mean that it is part of the implemeters toolkits, found under the ehealth reference platform, which there is absolutly nothing wrong with that.

My concern is that the reference platform thing claims to be backed by a lot of expertise technically, it also claims that data is clinically validated.

Q1, whynwould this not have been picked up by testing, surely their use of schematron or other technical testing would have identified this

2. Is this a Freudian slip by those in charge at NeHTA and displays what they jokingly think?

3. Is this a clinically valid swipe at consumers?

Anonymous said...

Whatever the actual error is, it is just another example of why people mistrust the PCEHR system. Watch the blame happening here: is it DOHA, NEHTA, or Zedmed's fault? Somewhere in this 'system', there should be some validation checks that reject things like 'meat inspectors'. The system as a whole has to be accountable as a trusted architecture that can deliver accurate information. Not have pockets that a blaming other pockets for poor quality data.
We have already seen a fair proportion of Medicare sourced data which was incorrect. Now we have meat inspectors. How can we trust that it medication or allergy data will be correct?
And I thought that vendors had to pass strict conformance tests by Nehta with their software before it could be 'connected'. You would think that they may have noticed the 'meat inspector' problem during testing. What else have they missed?

Anonymous said...

Anon March 29, 2015 10:46. Working for a software house, we have gone through NEHTA CCA, once connected to the PCEHR, we no longer need to include this step in our release plans for major releases. I am told soon you won't even need nehta at all the market will do self assertation

Anonymous said...

I doubt that NEHTA have anything at all to do with this.

DOHA have outsourced everything to Accenture, who run the thing as a managed service, so maybe it's Accenture who is a fault.

Anonymous said...

I beg to differ, the PCEHR as a system is not at fualt here, its just a dumb document store. The article states the error came from a digital package via NeHHTA, i am not sure NEHTA should or is proving anything other than sample code and guidance.

I am intrigued as to how the term meat inspector would creep into software designed for clinical use

This by no means detracts from the dismal management and non-leadership of the architecture disciplines in NEHTA

Anonymous said...

Doesn't anyone test anything these days? Including Accenture, as proxy system operator.

In fact, here's a question.... who's responsible for the system?

And yes, the PCEHR is a dumb system, in more ways than one.

Anonymous said...

yes of course it gets tested- in production once it has gone live!

Who is responsible for the system? No one and everyone. Depends on how you define the system. Is it the dumb document repository or is it the gp software?

Perhaps we should blame the meat inspectors?