Open Source On The Move.. In Health Care

[Dave Cooper]

What follows is a message regarding Yesterdays opening of a conference @
UCLA.
The conference deals with the fact that there is no Health Clinic database
systems that are worth... ( insert your term here )... Meanwhile overseas, a
number of Health systems have embraced Open Source and started a movement
that has just hit these shores... Currently the "best of class" is in
German. One aspect of the conference is to find enough folks in the US that
like it, and convert it to english....
  Current Health Care management software is VERY pricey, and as noted, no
so hot...
Any way... some interesting notes...

coop

Sent: Friday, November 15, 2002 7:41 AM
Subject: oshca 2002, day one


> oshca 2002, day one
> (third annual meeting of the open source health care alliance. the first
> meeting was in rome, the second was in london)
>
> over 40 people in attendance. about half are physicians with medical
> informatics angles to their careers; meanwhile, representing the u.s.
federal
> government i.t. staff, we have five from the v.a. and five from a
telemedicine
> project in hawaii; the remaining attendees are cio or i.t. staff.
basically
> this is a mix of hacker physicians, cio's, system administrators, and just
> plain hackers, some with formal government or academic positions, some in
> industry. most of the presentations are *not* powerpoint, and most laptops
are
> not running windows! england, france, canada, australia, new zealand,
germany
> and malaysia are represented. a sysadmin from uc davis med center is here.
we
> also have attendees from ibm, hewlett-packard, the us navy, and the ecc.
>
>
>
> dr. mike mccoy
> cio, ucla medical center
> - - - - - - - -
> our conference host is a physician, a hacker who programs in numerous
> languages, and cio of a 1,000 bed teaching hospital with 1,000,000 patient
> encounters/year, an i.t. staff of 400, an i.t. budget of $60 million/year
and a
> network with an average of 1,000 simultaneous users connected 24 x 7. he
has oc-
> 3's between buildings. he opened with a tale of being an intern at
stanford in
> the 1970s pushing a cart with a z80 microcomputer on his hospital rounds,
> operating a text based medical records program for charting which he wrote
from
> scratch, including the printer drivers, and he would paste the output (on
old
> greenbar paper) into the patient charts. as cio he delights in being "a
medical
> software salesman's worst nightmare." we tour his data center tomorrow.
>
> in his welcoming remarks, titled "a cio perspective on open source", he
> stressed that the strength of open source is not "free" software. he
> characterised the licensing costs of proprietary, closed source software
as
> less than 20% of the cost of software operations; his real beef with
> proprietary software is with "the complexity of vendor originated dongles,
> timebombs, and other nonsense. it's really about service, not software.
you
> can't steal service. and don't waste my time fussing with licenses or
metering."
>
> in his view software management is not about dollars, it is about "the
pathetic
> functionality of commercially available closed source software and the
> unbelievable rate of code abandonment." he is appalled by predatory
software
> vendors who acquire smaller companies merely for the customer base while
they
> sunset the old code and mandate an "upgrade" path; his example was a
digital
> radiology application in which ucla had invested $3 million. a new owner
> terminated the old code and forced them to "upgrade" from unix to windows
> servers. there was zero increase in functionality of the software in
> this "upgrade", and ucla went from supporting hundreds of radiology
terminals
> with a single small unix server, to supporting hundreds of radiology
terminals
> with a cluster of 20 large multiprocessor microsoft nt servers. he said
the
> only reason he "upgraded" is that he could not afford to spend the $5
million
> it would have taken to protect a simpler back end unix infrastructure for
> delivering that functionality. he is especially frustrated that millions
of
> dollars of perfectly functional code was simply terminated. this is the
long
> term business risk of purchasing closed source software: if closed source
> software is terminated, whether by market acquisition or simple business
> failure, the customers who depend upon that source code are out of luck.
>
> the i.t. consultancy cycle rewards long term inefficiency through a herd
> mentality. "no one ever got fired for buying cerner even though all known
> implementations are buggy and over budget."
>
> the core challenge of open source is to deliver functionality, quality and
> value. open source generally recognises that users are just as important
as
> developers.
>
> EMR as a vision is not possible yet because the small functional pieces
that
> are needed are missing or are incomplete. EMR still lacks an
implementation
> which understands the underlying abstraction of seamlessly streaming
records as
> they are needed by each type of user, by each type of records consumer in
the
> loop.
>
> he also reminded attendees that open source religiousity is inferior to
secular
> practicality.
> = = = = = = = =
>
>
>
> tim cook
> open paradigms programming, TORCH project leader
> - - - - - - - -
> tim was the leader of the old FreePM project that went down this year, a
victim
> of predatory venture capitalists who "acquired" the open source project
and
> redirected it to death in only a year of mismanagement. the problem was
that
> the funders approached the project as a software selling company seeking a
> quick return on investment rather than as a service oriented software
company
> with a long term goal of improving health care. "it's not about the money,
it's
> about the lasting results." tim quit the FreePM project and initiated a
new
> open source electronic records and practice management project called
TORCH. he
> closed his presentation with a slide quoting einstein, "if the idea at
first is
> not absurd, then there is probably no hope for it."
> = = = = = = = =
>
>
>
> dr. david chan,
> dept. of family practice, mcmaster university school of medicine,
hamilton,
> ontario, canada
> - - - - - - - -
> dr. chan is a practicing physician at a 27 doc family practice and a
python
> programmer, and the lead developer on the OSCAR project. OSCAR development
was
> funded by a grant from the province of ontario to mcmaster university. it
was
> originally designed for provincial medical practice management, but has
since
> been adapted to practices beyond ontario, including u.s. practice
management
> installations. OSCAR stands for "open source clinical application
resource." it
> runs on the zope webserver platform (written in python), which uses tomcat
and
> mysql in the back end. OSCAR is released under the GPL. it is available
for
> download at sourceforge. it also runs as an asp model from a data center
in
> texas at the rate of $200/month/provider. for backups the asp installation
> model recommends a small on site server that backs up the asp records once
a
> night and burns a cd backup for local off site storage.
>
> http://oscarhome.org
> tar files to /root
> follow instructions on install.txt
>
> icd9 coding is native in OSCAR. it is easy to add other code libraries as
> needed.
>
> his 27 doc clinic runs their entire practice management and EMR OSCAR
> installation on one small server (single processor, 1 gb RAM, RAID 2,
redhat
> linux) with a shadow box that copies files nightly and burns a backup cd
for
> offsite storage. for user terminals he buys the $300 lindows PCs from
wal-mart,
> and strips them of their operating system, replacing then with generic
redhat.
>
> dr. chan's undergraduate degree is electrical engineering. he's also a
runner
> in training for his first marathon. we met this morning at 6:00 am to do a
> couple laps of the ucla campus, and on saturday mike mccoy, who heard we
are
> runners training for marathons, said he would arrange for a ucla van to
take us
> to santa monica at 5:30 am so we can run on the beach!
> = = = = = = = =
>
>
>
> dr. molly sheah
> PrimaCare pilot project, kuala lumpur, malaysia
> - - - - - - - -
> this is a pilot project funded by the government of malaysia. dr. sheah is
the
> medical director of the project. she brought along her 16 year old php
> programming wizard who has written their code base. they have 50 solo
> practitioners running their private practices with the beta version of
> PrimaCare. each install runs locally on a single pc, with redhat 7.2,
apache
> 1.3.36, and mysql 3.23.32. each pc burns a cd every night for backup. it
has
> scheduling, patient billing, and EMR.
> = = = = = = = =
>
>
>
> cason lee
> team leader, UCLA medical center i.t. quality assurance
> - - - - - - - -
> live demonstration of ucla's homebrewed "tivoli/openview/unicenter
killer", all
> written in perl and python. he said the cost of writing these modules was
less
> than $100,000 in staff time, and the cost of buying the same functionality
from
> proprietary software was "in the millions". he said his team analyses
every
> help desk ticket to determine if there is a monitoring system they can put
into
> place that would have enabled i.t. to have anticipated and prevented the
help
> desk ticket from occuring. they continuously monitor the load on every
> processor on every server, including the virtual unix servers running on
their
> old mainframes, streaming this data to live consoles with event specific
color
> coded displays. also, they created naming conventions for all devices on
the
> UCLA network so the name identifies the first responder support team in
case of
> a failure in that component.
>
>
>
> dr. horst herb
> founder and project leader, GNUmed
> - - - - - - - -
> dr. herb is a private practice physician in australia, and is global
> coordinator of the GNUmed software development project. it is platform
> independent, independent of language or national health care system,
> independent of third party dependencies (except proprietary national code
bases
> needed during localisation), and is a purely volunteer open source project
> independent of all funding constraints. it is written in python with a
> postgreSQL back end. it has a two tier architecture, with all business
logic
> migrated from the middle layer to the back end. it is still in beta and
looking
> for more active developers to get to the project to the first release
candidate.
> = = = = = = = =
>
>
>
> dr. adrian midgley
> - - - - - - - -
> in a presentation laced with the driest possible british humor, dr.
midgley
> skewered various NHS (u.k. "national health service") software follies,
praised
> the VistA project (usa v.a. software, written in MUMPS, runs on open
source),
> and discussed a variety of developments in the field of open source health
care
> software since the 2001 OSHCA meeting in london. he maintains that the
> development issue for open source is now simply software engineering, not
> invention.
>
> he referenced the QuinetiQ study, published last year. he also mentioned
SQL-
> ledger for free accounting software.
>
> dr. midgley is a private practice physician in england.
> = = = = = = = =
>
>
>
> dr. andrew ho,
> psychiatrist, UCLA-Harbor medical center
> - - - - - - - -
> dr. ho, who said "unlike our wealthy cousins here at UCLA medical school,
> harbor is a v.a. facility with no software budget." dr. ho coordinates the
> development of OIO (open infrastructure for outcomes), an EMR and provider
> scheduling system. it runs on zope, postgreSQL, apache and linux. it can
be
> configured to allow patient browsing of records. it's model is metadata
> disguised as user created and managed web forms, with workflow tools to
manage
> the sequencing of steps in the forms. this gives each physician infinite
> control over their forms for each type of chart. there is also fine
granualr
> control for allowing patient access to viewable fields. unfortunately,
this
> extreme field level design freedom may also compromise it's utility in a
> combined EMR and data warehouse environment.
>
> "until computer based charting systems are better than paper based
charting
> systems, they will not be adopted."
> = = = = = = = =
>
> reporting from kinko's in westwood,
>
> wr
> --
>
>
> will ross
> director of communication services
> alliance for rural community health
> 776 s. state street, suite 102-b
> ukiah, california 95482
> [desk] 707.462.1477 ext 105
> [cell] 707.272.7255
> [fax] 707.462.1503
> www.ruralcommunityhealth.org
>