Milton Keynes' CRS caused 'near melt down'

Tags: Cerner CRS Fujitsu Millennium Nicholson Security

29 Jul 2008

The deployment of a national programme care records system at Milton Keynes Hospital NHS Foundation Trust “developed into an untenable situation which resulted in near melt down of the organisation.”

According to papers from the CRS project board, obtained by E-Health Insider under the Freedom of Information Act, the trust experienced a far from smooth go-live of the Cerner Millennium system. The CRS project team described it as “eight weeks of extreme pressure and operational issues to the acute trust.”

Board papers from the CRS project board, dating from July 2005 to June 2007, show a turbulent journey before the system finally went live on 24 February 2007.

Milton Keynes was an early implementer site for the local service provider, Fujitsu, and the acute trust urgently required a new patient administration system to replace its existing legacy system, which was considered obsolete.

An initial go-live was scheduled for 16 June 2006. The system was meant to include basic PAS, clinical noting and order communications, maternity, A&E, theatres and information for analysis functionality.

However, despite a project initiation document being agreed in November 2005, the project board was unable to familiarise itself with the actual system until March 2006, when Fujitsu first delivered a system build for testing. This was just two months before the expected go-live date.

“Lack of access to a working version of Millennium is impacting on the process gap analysis,” the CRS project board noted. “The change team has to make a huge number of issues as to how CRS will work. Knowledge has been based on seeing a few demonstrations, playing the demonstration CD and requesting clarification from Cerner via Fujitsu.”

Maternity was dropped from the system requirements, after Cerner was only able to provide a webcast demonstration of the module from Kansas City. The board papers say staff were “not confident with what they witnessed.”

When EHI reported on Milton Keynes’s implementation, which also covered the community hospitals run by Milton Keynes Primary Care Trust, Fujitsu said it had benefited from the experience gained from earlier implementations.

However, the minutes from the CRS project board show problems being continuously identified and go-live dates continuing to slip. In two years of planning, the go-live date changed seven times. The cost of the slippages was almost £800,000.

Gateway reports before each planned go-live identified risks as “red light”, preventing implementation from going ahead. Issues identified included an unstable build, major concerns about the reporting functionality and a lack of security plans for the new system.

Timescales for data migration were also considered tight, with delays from other sites in the Fujitsu Domain 1 area - Nuffield Orthopaedic Centre NHS Trust; Buckinghamshire Hospitals NHS Trust and Buckinghamshire Primary Care Trust – impacting on its implementation.

The green-light to go ahead with the deployment was only given three weeks ahead of the actual go-live. However, post-go-live, doctors began to identify a series of problems, resulting in the “near melt down” described by the project board.

The board papers suggest that problems at this stage included bed availability not displaying correctly, notes for clinics going missing, patient appointments not displaying on lists, reports not printing correctly, GP labels being unavailable, clinic rebuilds being necessary, back office help being unavailable, printers not working and passwords being forgotten.

Training was also an issue, with staff needing help but trainers unable to cope with demand. The project board reported: “Despite following the correct escalation processes, there was a very slow response from service management, which has resulted in eight weeks of extreme pressure and operational issues to the acute trust.

“Staff are reporting general ‘clunkiness’ of the system, slowness when trying to do things and that it is not intuitive. This has been reported to the health authority and Fujitsu and there is an agreement that we should urgently review the processes in order to try and make the system meet operational needs better.”

The first fortnight after go-live saw 1,166 calls to the helpdesk. A letter published by disgruntled clinicians prompted a visit by NHS chief executive David Nicholson and the then-director general of NHS IT, Richard Granger.

Following the visit, NHS Connecting for Health sent a representative to work with the trust in resolving issues. The 11 key issues identified by the trust were resolved in June 2007, and the trust’s CRS project board disbanded.

The board reported: “An agreed list of issues was identified and a Fujitsu team worked with the trust and Connecting for Health to resolve these issues. Whilst the majority of these issues are resolved there are some remaining reports still to be installed into cert for testing.

“Performance is greatly improved but there are still incidences of reduction in speed and log on ability. These reports are taken seriously but we have a reliance on staff to continue reporting to the help desk.”

Joe Fernandez

Readers Comments

Thank God for the FOI!

29 Jul 08 11:45

This is horrific and, but for the FOI, would have been kept secret. Well done to EHI for getting on the case.

meltdown...?

29 Jul 08 15:49

"doctors began to identify a series of problems" - Not sure how they could have done that when no Doctors were actually using the system post Go Live.

Most of the issues were identified by the hard working CRS project team, who worked tirelessly to fix the problems very soon after they were discovered to prevent the 'meltdown'.

NewSpeak.

30 Jul 08 09:58

"...no Doctors were actually using the system post Go Live."

So, in what parallel language can it reasonably have been described as a "Go Live" if the people meant to be using it weren't actually using it?

using the system?

30 Jul 08 11:01

"So, in what parallel language can it reasonably have been described as a "Go Live" if the people meant to be using it weren't actually using it?"

In the same way that my son doesn't "use" my car but it still provides a live service to him for getting around efficiently, the doctors may not be in the driving seat but they are indirectly using the system.

re Newspeak

ted.yeoman@nhs.net

30 Jul 08 11:06

You are confusing "Go Live" with "Functional Implementation" which also gets confused with "Service Implementation" Once these have been achieved you then begin the process of "Clinical Implementatio"

Implementatio

30 Jul 08 11:55

Surely Implementatio would only be for Lorenzo not Cerner

Root Cause?

30 Jul 08 13:16

Has anyone thought what might be the root cause of the issues faces at MK. Perhaps a system that should not have been implemented? Taking a serious look at the state of development and time taken to release changes and fixes for Millennium it's hardly suprising the go-live slipped so much.

Root Cause Analysis

31 Jul 08 12:09

The root cause is not too hard to see - central government cannot run efficient systems. they couldnt make cars in the UK in the 1970s, or Steel in China in the 60s or Tractors in Russia in the 50s or anywhere else at any time. The reason is not hard to see and is blindingly obvious in this case - beaurocracies always develop that look to their political masters and not to the users and consumers of the product.

Until we get the ownership and governance of the service embedded in the user base - and i dont mean the NLOP lip service - this will continue indefinately

Root Cause?

31 Jul 08 13:46

Since when have Cerner, CSC, Fujitsu, BT, Accenture, iSoft and all the others who have failed to deliver usable systems on time, been National Governments.

Root cause!%*?

31 Jul 08 17:36

IMHO if all of the suppliers and LSPs have failed, and therefore by extension CfH also, then maybe the problem is the concept rather than the execution. A silk purse cannot be manufactured from a sow's ear.

In my view the whole thing should have been built from the ground up and not imposed from the top down. Clinicians and Trust managers know what IT resources they need to care for patients and manage healthcare. They should have been allowed to chose their own suppliers, providing they could support the Spine functionality.

Competition between suppliers would have been maintained, best of breed systems could have flourished and local needs satisfied.

(post edited by ehi)

Hardly news...

01 Aug 08 09:47

Sorry, but this was reported in The Times ages ago. I was a member of staff at Milton Keynes during Implementation and Go Live and I can only say that Fujitsu are being made scapegoats. They were incredibly hard working, very professional and in some respects just weren't supported properly by CfH or by the Trust. Equally, Cerner, the supplier of the system weren't particularly well-versed in CDS or the NHS Data Dictionary and caused confusion of terms and processes. There were a number of people (from the Trust) on the project who just didn't have the right expertise and should have put their egos aside to allow work to actually be carried out. EHI doesn't appreciate that the trust hasn't told them the full story - FOI or no FOI.

Root cause!%*? don't make me laugh!

01 Aug 08 11:35

So the whole thing should have been designed from the ground up, should it? Please don't make me laugh, we tried that - remember?

If that had been the case we'd still be arguing over the definition of a care episode and defining to the nth degree the content of a handful of ADT messages.

Meanwhile the hypothecated money provided to PCT Commissioners for IM&T would have been siphoned off elsewhere, probably to pay for the PCTs' funding gap between official inflation rates and the actual consultants' and GPs' salary rises over recent years caused by centrally funded contracts. This would inevitably have led to the most powerful Acute Teaching Hospitals receiving the bulk of what IM&T funds remained leaving the PCT provider arms and MH Trusts arguing over the leftovers.

As for the root cause - wasn't this the Trust where the Chief Exec told users that they could choose for themselves whether or not to use the new system? Hardly a ringing endorsement or a sign of good stakeholder engagement from the Senior User or Executive/Customer. The real question is why the project was allowed to proceed if the Project Board had understandable reservations about the product being deployed!

Root cause

01 Aug 08 13:16

The root cause is quite simple.

The Southern Cluster/Fujitsu asked Senior Clinical representatives to evaluate the suitablity of Cerner to replace IDX - and got a thumbs up.

They did not ask any experienced PAS Managers to look at the Admin functionality within CERNER. Nor did Cerner/Fujitsu understand what a UK PAS system needs to do.

So the rout cause is suppliers not understadning what is required for PAS functionality in the UK and the NHS not doing a full evaluation of the suitability of CERNER to meet admin, as well as clinical requirments.

Joint mess up on both sides - hence lack of willingness on either side to own up and resolve problem (maybe with a UK PAS solution integrated with CERNER clinicals).

This is nothing to do with governement running IT projects - simply poor management in the process of being clear what is required from a supplier product and assessing the capability of the product to meet those requirements - lessons which I thought had been learnt from previous procurements (Winchester/AT&T Istel et al).

Will we ever learn.

Blame culture?

01 Aug 08 14:32

It seems to me that there is an attempt by some people in MK to spread the blame as far and wide as possible rather than accept responsibility.

“the acute trust urgently required a new patient administration system to replace its existing legacy system, which was considered obsolete.”

This would be the same legacy system that is still currently in use by the Oxford Radcliffe Hospitals and Heatherwood and Wexham Park Hospitals some 3 years after its urgent replacement was needed in MK.

“However, despite a project initiation document being agreed in November 2005, the project board was unable to familiarise itself with the actual system until March 2006” “This was just two months before the expected go-live date”

4 months from a PID to a “working” system – sounds pretty good to me, perhaps the expectation of a go-live date that quickly was unrealistic.

“The board papers suggest that problems at this stage included ….. passwords being forgotten.”

And who exactly should be blamed for that?; A) International socialism B) Central Government C) Department of health D) Fujitsu E) MK Programme Board F) Individual staff members G) Geoff Hurst in the 1966 World Cup Final

INCOMPETENT LOCAL NHS HOSPITAL TRUST MANAGEMENT ?

01 Aug 08 22:23

I have intimate / first-hand knowledge of what's been going on within this specific trust.

Whilst the national deployment of CRS across the U.K. could have been executed with much more care and prior consultation / consideration, individual trusts, and more to the point, their management teams and directors must also accept their share of direct responsibility and accountability.

IMHO, based on first-hand experience, the management at Milton Keynes must take some of the responsibility for the pronblems that arose with the implentation.

These are well informed comments, others with local knowledge are invited to contribute further.

(Post edited by EHI)