Newcastle Abandons NPfIT posted:
Thursday September 18, 2008
At last! Someone still working in the NHS has the
guts to tell it how it is.
The news that Newcastle NHS Trust has given up
on NpfIT is a breath of fresh air in the suffocatingly horrible
debâcle that is NpfIT. Could it actually now be, that shamed
by the enormity of their error, our public servants (by whatever
unaccountable organisation they are employed) will now accept that
they are totally unfitted to the task of commissioning mega-scale
and mission-critical computer systems? Such systems cannot be built
to order (especially when those orders are constantly countermanded
by detached and unrealistic bureaucrats) and be expected to work,
because the requirements of their design are way too complex. Like
all complex but successful entities, huge IT systems have to evolve
from simple and successful building blocks and combine organically.
Trial and error, natural selection, order from chaos, call it what
you will, only an evolving organic approach can hope to grow a system
where every working sub-system is dependent on working sub-sub-systems
almost ad infinitum.
Could this complexity have been reduced? Probably
not to any great extent. The tortuous route of a beetle crossing
a ploughed field is not a function of the beetle’s navigation
skills, it is directly related to the complexity of the terrain;
and make no mistake, the complexity of an individual’s medical
records, and the structure needed to store, retrieve and analyse
it, is often very complex terrain indeed.
Now years behind schedule and billions over budget,
NpfIT gets ever closer to an horribly expensive and embarrassing
demise. From the outset, despite protestations from the higher echelons
of both government and the NHS itself, many independent commentators
have been saying that the whole project was doomed to failure at
conception. No doubt many senior insiders also knew this but anyone
who dared say so was ignored or sidelined. NHS administrators, fed
on and mindlessly regurgitating sound-bites and acronyms, quickly
become deluded by their own ‘expertise’. The contractors
also had their doubts, but knew they would make a killing anyway.
In any case, there are now just too many voices singing the same
song for any objective individual or organisation not to accept
that cancellation is now, at the very least, a distinct possibility.
Unfortunately, NpfIT is the latest and greatest
failure in a long and dishonourable line of huge government IT projects
managed by the over-promoted idiots who never seem to learn anything
from the last disaster.
Having resigned, after 5 years in an NHS IT department, totally
disillusioned with the whole set-up, I can no longer sit back and
keep my thoughts to myself. Although you may not totally agree with
them, many probably won’t come as a surprise; but they do
represent a fairly typical view from an ex-insider. I accept that
office, canteen and dinner-party gossip, do not constitute a scientific
survey, but combined with first-hand experience they can offer valuable
Let’s have a look at what went wrong –
it’s all pretty basic and has the all-pervasive smell of inevitability.
The whole project was horribly over-ambitious
and over-specified. Quite simply, it was never a realistic expectation
that a single computer system could be designed from scratch to
handle the wildly diverse data in the healthcare records of 60
million patients - especially across such a sub-divided and chaotic
organisation as the NHS.
There are Primary Care Trusts (the GP bit), Acute Trusts (the
hospital bit), Mental Health Trusts, Ambulance Trusts, Foundation
Trusts, national and regional authorities, NHS Direct, NHS Walk-in
Centres – literally thousands of interlinked, overlapping
and heavily interdependent organisations. Paradoxically, many
are largely autonomous, with their own goals and agenda, sometimes
at each other’s throats. All are fighting for a greater
share of the available finances. Money is power and within the
NHS is ruthlessly used by largely unaccountable management for
parochial gain rather than for the benefit of the organisation
as a whole.
Between them, these organisations are
the NHS. They employ a staff totalling 1.3 million, encompassing
the clinical (doctors, dentists, opticians, nurses, therapists,
pharmacists, radiographers etc.), the administrative (accountants,
managers at all levels, clerical staff etc.) and a whole raft
of support staff such as porters, cleaners, caterers, maintenance
staff, security people, engineers, IT people – the list
goes on and on.
Unfortunately many of these employee groups also have conflicting
interests, often for the most worthy of reasons. For instance,
clinical staff want to treat people free from administrative interference,
administrative departments often have conflicting interests (finance
and almost any department you care to mention for example) and
the whole sorry mess is subject to interference and targets from
pettifogging bureaucrats both external (government) and internal
(a central NHS management that long ago lost touch with the grass-roots
even if they had ever mown a lawn in the first place).
Of course, across the NHS are thousands of isolated
and/or non-compatible computer systems; often many in a single
organisation. But at least all these systems work after a fashion.
Like all complex systems that exhibit stability (a swarm of bees,
a human, a successful business), the existing computer systems
have largely grown organically – most of the parts work
fairly well in isolation and have gradually evolved within the
NHS structure so they largely do their jobs, albeit sometimes
not as well as everyone would wish. It is therefore understandable
that a single integrated IT system should be such an attractive
Unfortunately, NHS management is, on the whole,
woefully inadequate – too distant, isolated, unaccountable,
inexperienced, unaware and plain stupid to specify a new blue-sky
system. That is not to say that management per se is
a waste of resources. It is (or should be) a vital, perfectly
legitimate and respectable means to an end – but administration
is a service, a foundation, a framework; its sole function is
to support the clinical staff deliver healthcare. It should, within
the financial restraints, provide an optimum environment in which
the clinical staff can operate to the best effect. But all too
often NHS management serves only itself. NHS managers discourage
or even fear individual initiative or lateral thinking, especially
when an underling suggests or produces a simple and, in hindsight,
obvious solution to a longstanding problem. In the NHS, career
progress is for those who toe the party line; it is foolhardy
to stick your head over the parapet.
These people (not of course the underlings at
grass-roots level – the very people who actually understand
the requirements of, and operate the existing system) were asked
to tell the system designers what they wanted. Unsurprisingly
the conflicting requirements were complex and often contradictory.
Those mediating the conflicts didn’t really understand the
issues so everyone ended up dissatisfied with the final specification.
Changes to the specification inevitably followed adding to the
overall confusion. Those in charge could not be seen to be leading
a failing project so they talked it up. The supremely arrogant
phrase ‘it will succeed because it cannot be allowed to
fail’ became the stock answer to those who could see what
was coming and dared to speak up. Many good people jumped ship
(and continue to do so) and the remaining became more disillusioned
The problem is not, and never was, the technical
requirements – the system is founded on large-scale industrial-strength
databases and communication systems; these are tried, tested and
reliable technologies. No, the problem is ever changing goals,
the need to keep things going in the changeover period and horrendously
over-ambitious expectations of what can be achieved with a ‘big
I don’t pretend that this post is
comprehensive or totally accurate or represents anything other than
my slant on matters. However, I do believe it to be a rough approximation
of the actual situation.
Please feel free
to comment by email