* SOFTWARE WITH A FUTURE.
by Dr Adrian Midgley  akm@92tr.freeserve.co.uk

Future Health Bulletin - May 2002 - plain text version
 SECTION TWO: OPINION - OPEN SOURCE.


In the UK, organisations are required to retain their tax records for
six years. If during this period a supplier of an organisation's
financial software fails or is swallowed up by another company, the
difficulty of restoring these records is not huge: financial records are
by and large rolled up at the end of the year, and references to a
summary are usually all an organisation needs.

On average, IT companies last 5-10 years. By contrast, medical
records need to be usable in increasingly complex ways for over 100
years: the lifetime of the patient, plus the time in which their
offspring can sue for actions taken that might affect them, or require
information relating to genetic risk and the like.

The general practice in which I trained had a clear line of existence
three centuries long, and looks well-set to outlast the NHS. But that
practice is now already on its third IT supplier, having only had a
computer for 20 years. My current practice has records on its shelves
that predate the NHS - 60 years old or more - and they are as usable
now as they were then. The performance of the IT industry has in no
way approached that vital level of sustainability. My practice has had
three commercial IT suppliers: lost one and fired two.

The Open Source Healthcare Alliance (http://www.oshca.org) was
formed on the internet. Its first meeting brought together doctors,
businessmen, professional programmers and others in Rome in the
summer of 2000. The business of that meeting was to give form to
the loose alliance that had developed, establish a charter and then
move forward.

As is usual when one starts something like this, we found that there
were many people in different places who were actually doing the
same thing, and some had been doing it for years. We took this to
indicate that the moment was upon us for the change from
proprietary software with secret code, and hence 'lock-in' to vendors
who historically have lacked stability, to open source code
applications and systems that can survive the loss of a vendor.

The second meeting in September 2001 in London was a far larger
affair, with over 100 people present (see Future Health Bulletin,
September 2001). We secured sponsorship from the NHS
Information Authority - the body responsible for implementing IT
policy in the UK national health service - and ran an NHS session
alongside the international meeting (see
http://www.oshca.org/oshca2001.html).

The conference included a presentation about VistA
(http://www.va.gov/About_VA/ORGS/VHA/vista.htm), a well-
established medical record and hospital management system released
to the open source community by the US government.

VistA, which was developed for hospitals that care for US military
veterans, is now also used in the Indian Health Service, in parts of
Finland and Germany and in two Cairo hospitals. The requirements
for making VistA more widely available in the UK were discussed,
and the work here seems to be more in adapting programs to
administrative and legal regimes than the mechanical work of
translating labels on screens.

Private sector companies such as McDonalds and Starbucks
demonstrate that re-applying a formula that is known to work cuts
the risk of failure, and the history of health and IT shows that we
need technologies that are more reusable and fail less often. The
Open Source model of production and support of software is a
necessary condition for the existence of software systems upon
which health service organisations will be able to rely for stability.

NOTE: Adrian Midgley is a GP in Exeter with a 'high-tech small
practice' - see http://www.homefieldsurgery.nhs.uk