Tuesday, September 23, 2014

My Health Service background

I am really glad to be engaging with the emerging Fed-IP informatics body.

For the record here is a summary of my involvement with Health Informatics over the last 47 years....


I joined the NHS as a trainee programmer / analyst in 1967. My principal
qualifications were one year of a truncated Nuclear Physics degree and the highest score so
far in Wales in an aptitude test…..turned out not be a good indicator of programming skill,
but set me on a 40+ year technology people and buildings path.

In 1967 at the Welsh Hospital Board computers were used for payroll, accounting,
admin tasks - taking over from tabulators and sorters. But under pressure from very
enthusiastic clinicians and epidemiologists the emphasis shifted towards ‘scientific’
applications and I was extremely fortunate to be dropped straight in to development and
support of a wide range of non-admin applications. The key message here for today’s NHS is
that clinicians came knocking asking for help knowing that proper use of even 1960’s level
computer technology could make patients’ treatment better and management of clinical
services more effective; they were not disappointed.

After a 6 week course at ICT in London I was let loose on developing a computer based
Welsh National Cancer Register to take over from a card-sorter / tabulator produced analysis
of survival rates for each type of cancer. An online search today will turn up a 1985 paper
that suggests data was first collected in the mid 70s - my experience is that in the mid 1960s
dedicated researchers captured and published valuable reports albeit hindered by the
limitations of sorter / tabulators..

The internet seems to have assisted in the apparent loss of much early computing
evidence in the NHS. From the experimental hospital computer projects in the late 1960s
through laboratory automation and radiotherapy treatment planning visionary clinicians were
pushing the boundaries of computing across the UK. As well as the Welsh Cancer Registry I
was fortunate to be involved in many early excursions, and I will briefly stroll through the 10
years from 1967 to 1976.

1967 - The Welsh Hospital Board computer centre was in the grounds of the Velindre
Road entrance to the Psychiatric Hospital in Whitchurch, just north of Cardiff. (This pattern
of using spare land in the long stay mental hospitals was repeated, and in 1973/4 North West
Thames RHA opened their computer centre in the grounds of Friern psychiatric Hospital).
Another bit of the spare land at Velindre Road was occupied by the emerging Cancer
treatment centre first planned in 1947. Dr Ken Wong, (who was to become internationally
recognised for his work in computer security) was working at Velindre in 1968, and we met
for the first time when my manager allocated me to support a Fortran programme for 3D
treatment planning developed in Holland. Good time to learn Fortran!

1968 - Digico, the UK’s first mini computer company, established by Avo Hiiemae and
Keith Trickett, majored on medical applications. The Micro-16s was an NRDC backed
general purpose minicomputer that joined the Digiac, released in 1966 to process data from
mass spectrometers. See the Digico Working Group, a special interest group of the Computer
Conservation Society (CCS).
My work focussed on pathology laboratories, and we developed and installed systems to
capture and report on biochemical and haematology analysers at a number of UK hospitals
including Leicester and Kings College Hospital, Denmark Hill. The Kings College system
complemented the recently installed experimental computer project for real time online
medical records that was a world leader - Professor John Andersen was both the guiding force
of this first real-time online 24 hour per day on ward clinical and nursing records system and
the Chair of the first Medical Informatics conference, MedInfo 74 in Stockholm - those of us
who were there did not sleep much because during the day the best medical computing
practitioners presented alongside the creators of such computer languages as Algol, and at
night the first world computer chess competition was conducted between systems from US.
UK, Russia and others. (I recall Chess 4.0 from Northwestern played Kaissa from Russia in
the final marked by a massive shift in style when the programmers overseeing their computers
stopped being adversaries and started collaborating as they tried to work out why their
computer programs were making the moves they did. These were the 2 programs David Levy
played while on his way to winning his bet that computers could not beat him by 1978.)!!!

For a description of the Digico system see
J Clin Pathol. 1974 December; 27(12): 1005–1012.
Implementation of a British computer system for laboratory data handling
C. H. Gray, A. D. Hirst, P. J. N. Howorth, T. P. Locke, B. Mellor, and M. Walter

This system, and the paper, had a strong reference back to work at University college
Hospital by Flynn, supported by amongst others Michael Healey. Healey will return soon in
this history….

J Clin Pathol Suppl Coll Pathol. 1969; 3: 62–73.
Problems and benefits of using a computer for laboratory data processing.
F V Flynn

1969 - On 6th October 1969 a press conference held by the Medical Research Council
and the north-West Metropolitan Regional Hospital Board reported on progress of a
combined district hospital and clinical research centre at Northwick Park, Harrow. Michael
Healey was announced as head of the new Department of Computing and Statistics. From
that point on, until the hospital opened in 1971 work proceeded to develop and implement a
complete end to end patient records system for all inpatients and outpatients in the 815 bed
composite hospital.

At Northwick Park, as it was at Kings College and Barnet General (see 1973-1976) the
digitisation of the Master Patient Index was a key infrastructure platform from which to
develop consistent and integrated patient records. I recall long nights scanning index cards,
and complex software designed to cope with multiple spellings of the same name ( at Kings
Roger Chalke deployed Kansas City Police Dept vowel and consonant stripping techniques to
normalise and codify decades of imaginative spelling - a technique I subsequently used on a
project in Saudi Arabia, and essentially the same process used in genealogical data search).

1969 - Tim de Dombal published with Hartley and Sleeman “A computer-assisted
system for learning clinical diagnosis” in the Lancet, January 18th pp145-149, this research
lead to diagnosis of acute abdominal pain being assisted by computer. clinicians were quick to
catch on to how computer technology could encode, store and share expertise among a wide
group of junior doctors, enhancing the initial diagnosis of acute abdominal pain and
hastening appropriate and effective treatment. Dr. Charles Hodes at Borehamwood published
“The Computer in General Practice” in The Practitioner, and continued to champion the
cause of computer screening procedures while I was in charge of medical computing at
North West Thames Region from 1973 to 1976.

1970 - Whipton near Exeter became the first GP practice to go paperless.

Preece JF,
Gillings DB,
Lippmann EO,
Pearson NG
. An on-line record maintenance and retrieval system in general practice. Int J Biomed
Comput 1970; 1: 329–337.Medline

The experimental computer project at Exeter focussed on community links, and by
1975 Ottery St Mary was integrated with the hospital systems.
Why general practitioners use computers and hospital doctors do not—Part 1:
incentives. Tim Benson BMJ 2002; 325 doi: http://dx.doi.org/10.1136/bmj.325.7372.1086
(Published 9 November 2002)

1971 - BUPA the private health insurer ran a diagnostic and multi-phasic health testing
facility in London. I was bowled over by the use of Microfilm players programmed to ask
medical history questions in the language of the patient, and to deliver comprehensive history
in the language of the clinician. The concept of multi-phasic testing, designed to catch
conditions early and pre-empt reactive and expensive interventions struck a chord with the
team at Kings College but the DHSS were firmly of the view that the NHS was there to treat
patients when they presented with symptoms, not to manage their lifestyles to avoid costly and
complex treatment later on in the development of conditions. The DHSS were also not up
for using a general measure of public utility (the cost to the nation of an hour extra waiting in
outpatients) as a justification for the roll- out of an outpatient management system that
WITH the savings to the nation would pay for itself in weeks; so much so that the

implementation was pulled just before go-live in a stark example of how IT was becoming
seen as a cost not a clinical benefit by the powers that be - this mindset blighted clinical IT for
decades.

1972 - I was asked by research clinicians to develop a quality control program to ensure
radio-immunoassay machines were working within tight ranges of accuracy. The integration
of quality control into automated processes in analysis was about to become a standard
practice, and I was again engaged with the implementation of similar processes at North
London Blood Transfusion working with BASIC on a Wang 2200, the forerunner of the
microcomputer.

1973 - 1976 I was appointed to lead Medical Computing for the North West Thames
region. For 3 years we had a fantastic opportunity to investigate, develop and roll out
practical, groundbreaking clinical solutions in teaching hospitals, district general hospitals and
in public health. 3 projects stand out for me over this period.

1. Barnet General Hospital. With a strong clinical drive Barnet hosted a team that I was
fortunate to lead, made up of analysts, organisation and methods specialists and clinical staff
at the hospital. We documented everything, producing a comprehensive description of how
and why everything worked at Barnet, which was used to develop a specification for a
distributed processing network of Cogar System 4 microcomputers connected in a dual failsafe
ring network supplying optical character recognition for unique patient id that did not
require barcodes with their “I am not a number” negative connotation……how come
barcodes are STILL the norm?? The Cogar System 4 is acknowledged as a forerunner of the
personal computer and was supplied to us by Singer Business Machines as the Singer 1500,
later rebranded the ICL 1500 when ICL acquired Singer.

2. Working with the Regional Scientific Officer (Harold Glass) we developed a
standardised pathology laboratory automation and management system based on Data
General Nova computers and rolled out in a number of large and medium scale pathology
laboratories. Harold Glass was later cited

Harold Glass

"The Impact of PACS on Hospital Information and Practice", Proc. SPIE 1093,
Medical Imaging III: PACS System Design and Evaluation, 354 (May 25, 1989); doi:
10.1117/12.953348; http://dx.doi.org/10.1117/12.953348

3. I was seconded to the DHSS to investigate the total systematisation of the Supplies
Vocabulary throughout the hospital service. Using techniques pioneered in the UK armed
forces we worked with all the regional supplies officers ton develop a common codification to
both reduce inventory complexity and improve ordering and management of clinical supplies
across England. the re emendations were accepted and North West Thames region became
the processing centre for supplies catalogue management.

Over the last 27 years I have returned to engage with the NHS on a number of
occasions, latterly working in Somerset with Taunton and Somerset Foundation Trust and
North Somerset Community Partnership. It is sometimes difficult to see how the fantastic
range of innovative and effective developments made by many dedicated clinical and IT
specialists in the 1960s and 1970s have in the main been lost, and the wheel re-invented many
times at escalating cost and complexity. It is encouraging to look at General Practice where
the early lessons were learned, and where there is now almost universal application of
comprehensive, compatible and cost-effective systems.

Let us hope that the recent resurgence of clinical leadership in Information Technology
will be allowed to deliver the systems and processes identified as valuable and deliverable 40
years ago……..otherwise the imminent ‘internet of things’ revolution and the ipad linked
personal vital-signs systems will grow in a vacuum to everyone’s detriment.

No comments: