Tuesday, November 22, 2011

As We Leap Into The Unknown and Untested We Need To Be Very Careful! People May Die Otherwise!

This article appeared a day or so ago.

New Report Echoes Call for National EHR Safety Board

HDM Breaking News, November 21, 2011
A new report published in the Journal of Patient Safety advocates creation of an independent national board to monitor and improve the safety of electronic health records. Among other duties, the board would have the power to implement unannounced, randomly scheduled, on-site EHR safety inspections.
In February 2010, Dean Sittig, PhD, of the University of Texas Health Science Center; and David Classen, M.D., of the University of Utah School of Medicine advocated five ways to improve EHR safety in a commentary published in the Journal of the American Medical Association. A recent report from the Institute of Medicine mirrored two recommendations--mandatory reporting of safety issues and a national safety board.
Now, in the Journal of Patient Safety, Hardeep Singh, M.D., of Houston VA Health Services joins Sittig and Classen in a new report detailing an oversight process for EHRs. They note that the increasing scope and complexity of EHRs, combined with aggressive implementation timelines under the meaningful use incentive payments program, can create a potentially hazardous environment. "At present, it is unclear which single agency is responsible for EHR oversight."
An EHR safety oversight program should include mandatory, standards-based reporting of adverse events and near-misses, and data analysis, according to the new report, "Creating an Oversight Infrastructure for Electronic Health Record-Related Patient Safety Hazards." Both reporting and analyzing "should be overseen by a new independent board specifically charged with ensuring safety of EHRs nationally," the authors recommend.
The EHR safety board could be modeled after the National Transportation Safety Board. To support the new national board, institutional EHR safety committees, including a designated EHR patient safety officer, would investigate and report all known safety incidents in an organization and perform routine safety self-assessments.
The report is available for purchase here for $35.
More here:
The issue I believe we have here is that we a purporting to be going live with a PCEHR system just 8 months from now and having apparently done none of the things you need to assure any sense of genuine patient safety.
For reasons that utterly elude me we are running these multiple pilots of various parts of a conceptual system (the Wave 1 and 2 Sites) and then expecting that when those components that are working are somehow ‘fused’ into the PCEHR all will be well, it will all hang together seamlessly and Bob will be our uncle! All I can say is dream on!
It just won’t happen and to be doing this at a national level without some testing at a (small) regional level of the whole system is as Sir Humphrey would put it ‘courageous’!
Consider for a moment the risks associated with one minute aspect of the proposed system - the Consolidated View.
Here a whole range of information from diverse sources is planned to be assembled onto a single screen.
The number of ways all this could go wrong is frankly terrifying (think wrong data, old data, lost data etc., etc.). At the very least there needs to be robust testing and validation of not only the information flows but also the information quality. Even if you get that close to right you then have the issue of this view being intended to be used by both clinicians (who won’t be well trained) and consumers who will essentially be untrained. Where is the evidence that consumers will understand and then sensibly use whatever is provided for them on this screen? Answer it does not exist!
If ever there was a time to get into place mechanisms to properly analyse the utility and safety of the PCEHR System proposals in the hands of consumers (as well as clinicians) it is now.
Consider the apparently simple issue of displaying pathology results. Where is the research that demonstrates the ideal to present such information to consumers and is the same approach optimal for clinicians? Someone might let me know if it exists but we all know it doesn’t. Equally is the PCEHR Consolidated View going to be static no matter who is the user, or will it dynamically re-configure depending on user capability. I wonder who has researched how that is best addressed?
To not have answers to all this, and to not fully pilot a complete implementation at small scale for utility and safety, before batting on because of political deadlines is the height of incompetence and stupidity in my view. Not only are lives potentially at risk but so also is the whole viability of e-Health as a publicly acceptable idea.
Part of the non-existent governance mechanisms has the be a Safety Board to review what is happening on behalf of both consumers and clinicians an to have the power and responsibility to call time out when real risk is identified until it is properly and fully addressed.
The madness of all this just seems to roll on towards an inevitable and very sad outcome.


amcoz said...

Why don't 'we' simply carry an encrypted 'credit-card' style of device for our medical history. It gets updated every time you attend any sort of medical examination, consultation, including drug dispensary, hospital entry, and so on and the only way of accessing the data is by your PIN. What's the flaw in this simple example?

Anonymous said...

What if you lose it? Do you need a backup - how would you manage this? If you can copy the card, how is it better than a magstripe card? What if you forget the PIN? Who would man the helplines to do password resets? Who would pay for this? Who would control it? Smartcards have less than 1MB of storage on them. With only this tiny amount of storage, what would you put onto it? What happens when it is full? Who can delete stuff to make more room? What device would read the cards? Is this device already in *every* point of care you visit? If not, who would pay for the readers to be rolled out? What software can read the card? Is this software already in use at *every* place you get treated? If not, who will pay the vendors to implement it? How soon could this be done? Who would pay doctors for the extra time to read your card? Given that the information on the card might have unknown origin, would a clinician trust it? If they make a decision based on that information, who would be liable? Given that the device might be plugged into many computers, what viruses might it pick up from other computers?

Any of these reasons is enough to make the consumer-carried token a non-starter.

Welcome to the trainwreck of eHealth.

Anonymous said...

Well - for starters - we need to get a common design in place and that means we need to agree on what standard(s) will be used, then we need to get health providers (doctors, pharmacists, physiotherapists, and an army of others) all using the same medical record model system, then we need to ensure that the health service providers' computer systems which they use for recording their information is upgraded to a capability where it will be able to use, read and write to, the encrypted credit-card style device to which you refer, then we need to accommodate the numerous occasions when the card cannot be read or written to because it has been mislaid or damaged or the data corrupted, then we ......and that's only the beginning, not to mention the middle and the end. I'm sure there must be a flaw in there somewhere!

EA said...

In Victoria, the community mental health services do not communicate with each other, let alone with primary care systems or with HealthSmart.
Imagine the mess of 20 years of health records for someone with schizophrenia.
Has anyone heard about a rejuvenation of "coordinated care"?

amcoz said...

As I suspected, there's nothing to beat pen and paper. Good points, all of them.

Anonymous said...

Roxon's neck on the chopping block please.

Anonymous said...

amcoz, there a lots of things to beat pen and paper. The big problem is how many moving individual parts have to be updated with new capabilities AND then be made to work together. It's hard to do.