Fujitsu's Cerner delays mount in the South

19 Oct 2006

Three more trusts have had to postpone Fujitsu implementations of Cerner Millennium in the South at short notice, as go-live dates for hospital systems slip further across the region. The delays mean that a key Connecting for Health delivery pledge will be missed.

The three latest sites to announce delays include Taunton and Somerset NHS Trust, which was due to have gone live the weekend of 7-8 October, and Mid Hants, an implementation covering two acute trusts and one primary care trust which was due to have gone live last weekend.

In addition, Worthing and Southlands Hospital has this week told EHI it will no longer implement as scheduled on 2 November and is “waiting for a new date”.

The latest delays make it impossible for the South’s local service provider (LSP), Fujitsu Alliance, to deliver the 12 installations of Cerner that it promised would occur by the end of October. To date, three trusts in the South have implemented Cerner’s Millennium software – two since June.

Mark Wark, Taunton and Somerset’s director of marketing and communications, told E-Health Insider: “We do not have a go live date that we are working towards at present - there are some technical situations that have to be sorted out to our satisfaction first.”

Mid Hants are understood to have had their Fujitsu implementation of Cerner Millennium pulled last Wednesday, three days before they were meant to go live.

Last-minute postponements of hospital trust Cerner implementations have become a hallmark of the Connecting for Health (CfH) programme in the South of England. As previously reported by EHI, Milton Keynes Hospital – which would have been by far the most complex implementation in the South - has twice had its go-live date put back at the eleventh hour.

Hospital PAS go-lives are extremely big projects. Last-minute postponements are costly and disruptive for NHS trusts who have had to extensively plan and prepare, clean data, train staff on new systems, schedule cover and arrange holidays around go live dates. Delays also create additional costs for suppliers.

In June Fujitsu told the House of Commons Public Accounts Committee that they would complete 12 initial Release Zero (R0) implementations of the Cerner Millennium system by the end of October.

One, Nuffield Orthopaedic NHS Trust (NOC), was already live and two more have occurred since: Weston Hospital NHS Trust and Buckinghamshire Hospitals NHS Trust. Nine more trusts were scheduled to have gone live by the end of October but for a variety of reasons have had to delay.

Millennium R0 was intended to be the version of Cerner Millennium already in use at Newham and Homerton hospitals in London, but has since required modification for use in the South. With the R0 releases running so late the subsequent R1 releases, which were to offer more clinical functionality, are getting pushed back.

EHI has been told by sources in the cluster that problems continue to exist with connections to the national Choose and Book system and with generating statutory reports from the software – a problem that caused NOC to be marked down on its performance by regulators at the Healthcare Commission.

Sources indicate that Mid and South Bucks is facing similar problems with reporting.

Problems are also said to exist within the workflow for patient management in Millennium. A well-informed source - who praised the clinical tools provided - told EHI of the difficulty of recording outpatient attendances. “Under Cerner it’s easy to miss out parts of the attendance and in terms of coding. You have to open a separate coding screen, select the patient and attendance and then encode, risking data being missed and taking twice as long as current PAS systems.”

The same source told EHI that delays were being compounded by the lack of technical environments to support the domain-based approach. In practice, this means a live site needs downtime when a new site goes live. “At Weston, for Bath to go live, Weston needs a significant period of downtime, as will any other domains who want other sites to go live,” the source explained.

They added that the generic training environment was also creating difficulties. “There is a generic training environment rather than one based on a local build, leading to generic rather than local process based training based on a local configuration.”

A frustrated LSP insider, however, stressed that that suppliers only want to provide good implementations - “which is why they work hard to support these difficult situations without receiving any additional funding from CfH even though it’s above what was contracted.”

Of the trusts that were meant to go live in October many of those contacted by EHI now appear to be operating a wait-and-see policy, waiting for news on other implementations and for software problems to be fixed before they go live.

Both Milton Keynes and Taunton and Somerset, have told EHI they are waiting for fixes at Mid and South Bucks. These delays are creating a knock-on effect pushing back later R0 implementations.

A spokesperson for Milton Keynes told EHI that the overall CfH project plan “requires a minimum of two weeks between each go-live date.”

Milton Keynes and Taunton and Somerset were due to have been followed by Hampshire, Surrey and Sussex Healthcare NHS Trust and Bath Royal United Hospital NHS Trust. The Medway NHS Trust told EHI that they were aiming to go live sometime in autumn/winter, but could not give a specific date.

Four trusts, Worthing and South West Hospitals NHS Trust and North Devon Healthcare NHS Trust told EHI at the end of September that they hoped to go live in November. Another two trusts, South Devon Healthcare NHS Trust and Yeovil District Hospital NHS Foundation Trust said they planned to go live in December. However, delays at the earlier sites now make these dates look extremely doubtful.

Links

Milton Keynes' Cerner implementation postponed

Reader's Comments

Speaking truth to power

19 Oct 06 18:11

As the rival "NPfIT is a massive success" and the "US systems can never be made to work in the UK" gangs renew battle on this thread may I offer an alternative view.

NHS IT implementation timetables have always been driven more by hope than realism. Delays are not unique to Cerner, Fujitsu or indeed NPfIT.

However what happened to testing and milestones in project management? Even if these were short circuited, did people on the ground really not know earlier than a week or days in advance that these go-lives were not possible?

The answer is it takes courage to be the first to say "this isn't working". Some of us have been there, done that (and consequently been told to clear our desks) years before NHS CfH was dreamed of.

The sole differences here are scale and Politically dictated deliverables and deadlines. In order for the message to be heard and understood many powerful people need to risk their positions in the comfort zone. Do not hold your breaths :-(

All too rushed

19 Oct 06 21:31

I think the problems arose from its all being too rushed and thinking that all hospitals are the same. Users did not get to see a near final version until "training" on a generic system. Trainers did not know answers to questions about local use. No end user testing of whole processes. Underlying architectute unsound, so whole construction wobbly etc. Supplier does not understand local issues in detail, assumes all hospitals the same like a supermarket chain, supplies one system, which fits one hospital, which had huge time and infrastructure investment. Surprises when it does not work elsewhere, especially if elsewhere has a good PAS. Still could be saved by a pause of say 6 months, whilst supplier and client work out waht they want locally.

Mid Hampshire CRS deployment

20 Oct 06 04:40

The deployment in Mid Hampshire is not to Mid Hampshire Primary Care Trust but to the Mid Hampshire Deployment Family of three Trusts: Winchester and Eastleigh Healthcare, Hampshire Partnership and Hampshire Primary Care. The CRS release does virtually nothing for the PCT and it's the two other Trusts which are running the deployment project. It would help to correct this article. (Also, I'm not surprised Mid Hants PCT failed to respond to you as it ceased to exist on 1 October!)

Fit for purpose?

mary.hawking@nhs.net

20 Oct 06 08:30

I'm not in the secondary care sector - but there do seem to be some fairly fundamental problems here. Coding (ICD and OPCS) is vital for billing in the new world of PBR and PBC - and yet :-"Problems are also said to exist within the workflow for patient management in Millennium. A well informed source - who praised the clinical tools provided - told EHI of the difficulty of recording outpatient attendances, “under Cerner it’s easy to miss out parts of the attendance and in terms of coding, you have to open as separate coding screen, select the patient and attendance and then encode, risking data being missed and taking twice as long as current PAS systems”." *When* clinical functions are included, how will Coding (Read and/or SNOMED-CT) be organised for the clinical record?

"EHI has been told by sources in the cluster that problems continue to exist with connections to the national Choose and Book system and with generating statutory reports from the software – a problem that caused NOC to be marked down on its performance by regulators at the Healthcare Commission. .". Isn't connection to spine functions one of the main reasons for having NPfIT in the first place?

If these problems do exist (and are hard to fix without rewriting the whole thing) , *is* the product "fit for purpose"?

I have profound sympathy for the Trusts involved in deficits due, in part, one must suppose to expenses , both direct and opportunity lost,

Divergence between spin and reality

20 Oct 06 09:06

The ongoing reality gap between Granger, Fujitsu and the NHS coal face is continuing to undermine the NPfIt programme. At some point the Southern Cluster and Fujitsu need to own up that they got it wrong in signing up to deploy Cerner as is. Whilst it has huge potential - it is still not fit for purpose in supporting basic PAS type processes. This combined with supplier product expertise spread to thinly across to many deployments, lack of underlying technical environments and political/commercial driven implementation plans will result in further delays and an increasingly hard to bridge credibilty gap.

If we were dealing with a marriage - you would suggest some form of councelling to bring the two main parties together in the interest of the Children - who in this case are local NHS organisations and who are getting all the fall out from this failing and increasingly dishonest relationship.

multi-factoral problems

20 Oct 06 09:47

When the Detailed Implementation Plan was first shared, most people I talked to suggested that it was very optimistic to have such an aggressive roll-out strategy, with little room for slippage or regrouping. This is repeated for R0, R1 and R2.

Such is the nature of the rigid standardisation, the un-anglicised software (and that is not just changing the labels, but the underlying logical model), and the political pressure, that many would suggest this outcome was always inevitable.

Is the system underneath all this sound ? Some of the Clinicians directly involved are saying it is, others have walked away in disgust. But it is not organised to help UK administrative staff do their jobs efficiently, and these are the key users for R0.

Is the model of delivery right ? The ten domains for the South are an artificial construct, based around IT limitations, and organisational boundaries that are already history, not around clinical groupings. The robustness of the data centre/N3 delivery is unproven, and the spine is proving its weakness with further choose and book problems this week.

The resilience of clinical services is not enhanced by putting all the hospitals across 1/5 of the country at the mercy of one IT configuration. In my view the patient benefits are marginal for the few exceptions, and the risks are for the majority.

Come on Messers Bacon and Pugh.

Nothing New

20 Oct 06 09:48

Those who have worked at the coal face in NHS IT know and can tell you the problems, but such is the climate (as the above comment makes) few are willing to challenge this appauling fiasco. The whole process is being driven by politics and money. Look on the bright side folks, with the current policy of plurality of provision and courting of the private sector, as the new providers come in perhaps they will bring their own IT offerings!

Clinicians vs bean counters and chargers

20 Oct 06 20:35

There has always and forever been a problem between clinicians driven by a Hippocratic and "socialist" or altruistic motives and the business of collecting the money to pay for it. I believe that it remains unproven that complex processes for costing, charging, and invoicing are "value fo rmoney" over an approach of "pay me how much you are grateful and if you are rich, just a bit more on behalf of the poor". So the two "information" systems for clinical care and charging are at odds philosophically and so already unlikely to fail in a "socialist" and caring society - different in a capitalist society, where illness is seen as an opportunity to make profit. Even the most money minded clinicians feel nauseated at making pure profit out of illness. Anyway clinicians want a few meaningful free text words about a patient to get their brains thinking to help make hard judgements and decisions in complex problems. IT people universally hate free text fields and love mandatory lists. Clinicians survive by sharing fuzzy information and free wide ranging discussion. IT people love to lock things up in passwords and limited access. IT people live in a word of control and numbers, clinicians live in a world of uncertainty and the wonderfully bizarre world of human behaviour. A CRS is like trying to computerise a symphony orchestra or an artist's brush strokes - it will take away the life (=death) and beauty in the pain of existence. Having said all that I love computers and what they do for my clinical practice - the computer remembers all that I forget, I can access information and expert opinion from all around the globe in seconds - and I do - only today consulting from the UK, experts in Australia and the USA. Computers can do huge amounts to make patient care safer and better - however we need clinicians who are lateral and braod thinkers matched by IT people with "Emotional Intelligence" - then we can do it. Sadly the combined skills of Cerner, Fujitsu and Clinicians appear to have produced the worst of both worlds, not the best. As ever in human existence it is getting the bright powerful minds to have the right conversations with each other that determines the right actions and outcomes. We still have all of these onboard. although I worry about the finances on all three sides, so lets get the conversations and understanding right. We can (hopefully) only get better - if it gets worse the country and its health service will be bankrupt of money and the facility to care.