On Fri, 14 Nov 2003, Kantor, Gary wrote:
...
source: http://www.jamia.org/cgi/content/full/10/6/617
> ----------------------------------------------------------------
>
> We are well aware of the status of open-source electronic
> health records (EHRs),
Really, any evidence of this?
Which open-source EMRS did they review?
...
> Furthermore, none of these other countries have utilized an open-source
> approach-instead, they established standards for the EHR-for the
> underly-ing data types, and for a reference information model.
Beautiful, let's just replicate what U.K. and Australia did, nevermind
trying to learn from their experience.
> In countries such as England and Australia, a small number of vendors
> (typically 2 to 4) then established a major market share.
And ... so? What did we learn from the U.K. and Australian experience?
...
> Furthermore, even at a cost of $10,000 to 20,000 per provider, EHRs
> appear to be cost-saving. [1]
Do we employ wishful thinking to convince everyone to pay $10,000 to
20,000 for an EMRS? Or, should we simply enact a new federal law?
> Unfortunately, because of the structuring of our reimbursement system,
...
Here we go again.
For those from outside the U.S., it is quite fashionable for American
physicians to blame everything on "our reimbursement system".
There are many many reasons why EHR/EMRS are not popular / useful. In many
cases, even _free-as-in-beer_ is not enough incentive, let alone $10,000+
/ provider. :-)
> The open-source approach has worked best for tools, and there
> is no example that we are aware of in which something as
> complex as an EHR has succeeded.
The argument is plain wrong.
Open-source approach works well for software development. Open source
approach works very well for highly complex software projects.
Many examples of excellent complex free software exist.
Furthermore, it is very bizzare that "Complexity" is even cited as an
objection. Whoever wrote this reponse clearly has no clue about problem
solving or software engineering. You always decompose a complex problem
down to simpler, more easily solvable steps. Open source is inherently
more suited to distributed problem solving (via decomposition) than
closed-source approaches. It makes no sense to suggest that EHR is too
complex a problem for decomposition.
> Linux, for example, is an operating system.
And OIO, for example, is an infrastructure for building EHRs. What is the
point?
> Furthermore, there are concerns about versioning, interfaces,
> maintenance and improvement, and governance with decision support with
> an open-source approach.
These are legitimate "concerns". However, they are hardly arguments
against open source.
With open source, versioning, interface, maintenance, governance issues
are easily resolved. In contrast, with closed source, the same issues
often lead to costly litigation and orphaned EHR systems.
> Nonetheless, we believe that the open-source approach to
> EHRs is an attractive one, especially for small practices,
Clearly the authors did not review VistA or choose to classify VistA
under the "closed source" category.
I wonder why they said:
"We are well aware of the status of open-source electronic
health records (EHRs)"
> and, in particular, applaud the efforts of the American Academy of
> Family Practitioners.
Get a clue. AAFP's effort deserves no applause. Maybe this is not the
authors fault. Perhaps JAMIA is just too slow in their publishing workflow
to keep up with the latest in the medical informatic field.
> However, we suggest that parallel initiatives with both traditional and
> open source initiatives should be pursued.
Spoken like a politician. A statement that says nothing and offends no
one.
> For open-source approaches, studies and reports in the peer-reviewed
> literature are especially important.
I agree. Open source projects don't have any/competitive marketing budget.
Peer-reviewed publishing is an inexpensive way to communicate.
> Success with open source will require substantial investment,
True.
> and given the uncertain returns it is still not entirely clear how to
> make the model work
There have been quite a few documents, papers, and books that describe
HowTo make the open source development model work. Again, just because the
authors neglected to review these references and consequently unqualified
to address this issue, does not mean the issue is unclear.
> or who will make the investment
The immediate users of the system, their patients, and friends? Why is
this such a difficult concept to grasp?
...
> We believe that given the successes in the rest of the world with other
> approaches,
Success? Unqualified success? Partial success?
> the unproven nature of this one,
"Unproven" or "Unproven to the authors satisfaction due to ignorance"?
> and our capitalistic society,
vs. what kind of society? Implying that "Open source" does not work or
work-as-well in a capitalistic society?
> it would not be prudent to place all the nation's eggs in this basket.
A classic finish - how can anyone disagree with such a non-sense cliche?
> "Reprinted from J Am Med Inform Assoc. 2003;10:617. DAVID W. BATES,MD,MSC ,
> H. C. MULLINS ,MD, JOHN A. ZAPP,MD. Copyright (2003), with permission from
> the American Medical Informatics Association."
Thanks Gary, for obtaining permission and giving us an opportunity to
participate in this discussion.
Best regards,
Andrew
---
Andrew P. Ho, M.D.
Assistant Clinical Professor, Department of Psychiatry
Harbor-UCLA Medical Center
University of California, Los Angeles
