Evidence submitted by Alan Shackman (EPR
38)
EXECUTIVE SUMMARY
E1. This submission focuses solely on progress
on the development of the core Local NHS Care Record Service,
CRS, the area from which the vast majority of the benefits of
NPfIT will be delivered.
E2. Not only is delivery of core CRS at
least 2 years behind schedule but, as of early 2007, there remains
no definitive timescale for introducing the clinically focused
software that would take functionality in any significant way
beyond the basic patient administration functionality that was
available to the NHS when NPfIT began in 2002.
E3. Deployment of CRS cannot be said truly
to have begun at any NHS organisation in the North West &
West Midlands, North East, and East clusters because patient administration
system (PAS) solutions that have been implemented under NPfIT
are interim and will have to be replaced. Deployment of CRS proper
is not expected to take off until 2009-10. Even this date cannot
be assured given the as yet unresolved financial problems of iSoft,
the supplier of core software, and the risk that development of
the Lorenzo system will not be completed. these clusters
E4. The situation in London and South clusters
is more positive now that there is a real indication that Cerner,
the core software supplier in those clusters, has successfully
anglicised the PAS element of its Millenium product for acute
trusts. This at least gives hope that deployment of the clinical
elements of CRS will not be long further delayed.
E5. The key objective of CRS to provide
a seamless service across all care settings—GPs, community,
mental health and acute hospitals—is in danger of being
lost now that the original intention to have a single integrated
system in each cluster has been abandoned. It has been agreed
that GPs can have their system of choice; and, at any rate in
London, an additional core software supplier has been brought
in to provide for the specific needs of mental health trusts and
PCTs. Whilst there is little doubt that a solution can be found,
there are as yet no plans for how a cross-organisational, joined
up CRS is to be obtained.
E6. There is no evidence that NHS organisations
are committed to changing working practices in order to reap the
benefits of CRS as it begins to be deployed.
E7. The first major theme in suggested ways
forward is to bring additional suppliers of core CRS software
into LSP consortia. This would have a dual purpose: to move core
CRS forward more quickly particularly in the North West &
West Midlands, North East and East clusters where progress to
date has been negligible; to provide choices locally with a view
to creating greater local ownership and thereby helping to obtain
commitment to change working practices.
E8. It is suggested that current LSP contracts
be reassigned from the Secretary of State to SHAs or PCTs in order
to give those now being given responsibility for implementing
NPfIT systems the full authority and power to enable them to succeed.
E9. The final suggestion is to make use
of work undertaken under "Information for Health', the national
programme that preceded NPfIT, to provide guidance on achieving
cross-organisational information sharing.
THE SUBMITTER
Alan Shackman has over 20 years experience as
an independent consultant to the public sector, principally to
the NHS, specialising in the business/user focused elements of
preparing for and implementing IT systems. He is an expert in
the usage of IT at grass roots level in the NHS, the majority
of his assignments having been directly with NHS trusts covering
all sectors—acute, mental health, community and primary.
Over the past 15 years his work has focused on patient administration
(PAS) electronic patient record systems. In that time he has prepared
business cases and detailed definition of requirements, led procurements
of PAS and electronic patient record systems at major acute trusts
and across local health communities, developed benefits strategies
and realisation plans, and facilitated the management of change.
Alan is familiar with the products being offered
by the two suppliers of core CRS software to NPfIT having worked
on PAS replacement projects in both the South and North West and
West Midland clusters. During 2004 and 2005 he acted as NPfIT
programme facilitator for the north east sector of Greater Manchester
comprising an acute trust, a mental health trust and five PCTs.
Prior to that he helped two local health communities in Greater
Manchester and Lancashire develop and implement their strategies
under the "Information for Health" initiative.
At national level, Alan led the development
of a national output based specification for electronic patient
record for acute trusts as part of the NHS IM&T Procurement
Review of 1999. Earlier in the 1990s he was a member of the NHS
Information Management Group's panel of consultants advising trust
boards on electronic patient records. He was also a member of
the team that reviewed the national resource management programme.
He has recently been acting as an advisor on
NHS IT matters to Richard Bacon MP of the Public Accounts Committee.
Alan is a Chartered Engineer and Member of the
Institution of Engineering & Technology. In an earlier phase
of his career he was a member of the UK Industrial Space Committee
and represented the UK on international telecommunications standards
and regulatory forums.
COMMENTS ON
PROGRESS ON
THE DEVELOPMENT
OF THE
NHS CARE RECORD
SERVICE, CRS
Scope and context
1. This submission focuses on progress on
the development of the core Local NHS Care Record Service, CRS.
This where the overwhelming majority of NPfIT funding is allocated
and it is from the core Local CRS that the vast majority of the
benefits set out in the NPfIT Business Case will be delivered.
The submission excludes consideration of Picture Archive and Communication
Systems (PACS) because these were not originally included as core
deliverables; for the same reason it also excludes the plethora
of relatively minor acute and primary/community departmental systems
that LSPs have implemented.
2. The submission shows that:
(a) not only is delivery of core CRS up to
two years behind schedule but that, as of early 2007, there remains
no definitive timescale for introducing the clinically focused
software that would take functionality in any significant way
beyond the basic patient administration functionality that was
available to the NHS when NPfIT began in 2002;
(b) the way in which CRS is now being
introduced has been significantly altered with the result that
the original intention that CRS should "provide a... service
accessible... by health professionals whether they work in hospital,
primary care or community services"[107]
is in danger of being lost;
(c) even when systems are introduced the
danger has increased since the inception of the Programme that
changes in working processes necessary to obtain the benefits
from the investment in CRS will not be driven through.
Review of CRS deployment to date
3. By now, all 155 acute hospital trusts
in England should have implemented new NPfIT patient administration
systems (PAS) as the essential first step in the introduction
of Local CRS. As of April 2006, according to the NAO Report, the
actual number was nine hospitals. Since then, so far as I am aware,
the number has increased by only six. [108]This
gives the flavour of CRS deployment to date. The position is different
in those clusters in which iSoft is the software supplier for
core systems (North West & West Midlands, North East, and
East) and those clusters (South and London) in which Cerner is
core software supplier now that IDX the original supplier has
been replaced. iSoft dependent and Cerner dependent clusters are
therefore considered separately.
...in clusters where iSoft is core software supplier
4. iSoft was contracted as the Local CRS
software provider on the basis of its proposal to develop a new
product, Lorenzo. Phase 1 of Lorenzo was stated in iSoft's 2005
Annual Report to have been available in 2004. It was not. iSoft
now states, as reported by e-Health Insider on 11 December 2006,
that Lorenzo will be delivered to CSC, the Local Service Provider,
LSP, for the three clusters, in the first quarter of 2008, but
that it will then be some months before it becomes available to
trusts and that thereafter its introduction will be gradual, supposedly
to occur in the 2009/10 timeframe at the earliest. However, confidence
that this will be achieved must be tempered in the light of iSoft's
current difficulties as widely reported in the media: that the
company's accounting practices are being investigated by the FSA,
that its share price has collapsed and that it has yet to obtain
the long term funding it requires to meet its commitments. [109]
5. In lieu of Lorenzo, the CSC has been
offering to trusts existing pre-NPfIT, non-CRS iSoft products
known as iPM and iCM. iPM is a PAS with no clinical functionality.
iCM is an additional module which includes some clinical functionality
relevant primarily to acute trusts eg ordering pathology tests
and X-rays. Some 10 acute and mental health trusts with a pressing
need to replace their existing PAS have successfully deployed
iPM. Those without a pressing need—the vast majority—have
decided against because, even with the cash releasing benefit
to them of CfH picking up the recurring revenue costs, iPM is
considered inferior to their current PAS and certainly not a good
enough reason for taking the risk and disruption of replacing
a current working system—all the more so since it has become
clear that moving to a Lorenzo solution from around 2009 would
effectively be another major replacement exercise, not just an
upgrade.
6. Few PCTs had any corporate system for
helping them managing the community services they provide and
a considerable number, perhaps 50, in the North West and West
Midlands cluster according to its website have deployed iPM. Some,
to my personal knowledge, have rolled the system out widely to
the various community disciplines. At others, I am told, the deployment
has been more limited, the apocryphal "a couple of podiatrists
use it every other Thursday morning". iCM has some facility
for recording clinical notes but otherwise offers little to community
clinicians. PCTs will therefore have little opportunity to progress
unless and until Lorenzo becomes available, and even then, as
for acute trusts, they will presumably have to face another major
replacement project, not just an upgrade.
...in clusters where Cerner is core software supplier
7. Release 0, essentially the PAS element,
of the Cerner Millenium product, has now gone live at some five
acute trusts in South cluster. This was after a false start, causing
considerable operational difficulties at the trust involved (Nuffield
Orthopaedic), resulting from the LSP's eagerness to begin deployment
before, as it transpired, this product from the US had been properly
anglicised. At the time of writing post implementation reviews
are awaited from the two latest trusts to go-live (Milton Keynes
and Mid Hampshire) but early reports are encouraging. Release
0 has not been implemented at any mental health trusts or PCTs
in South cluster.
8. London, having completed just one PAS
replacement using the IDX system, has not yet undertaken any PAS
replacements at acute trusts with the Cerner Millenium system.
However, the LSP (BT) has moved away from the concept of having
a single supplier of core software and, in lieu of a suitable
product from Cerner, has adopted CSE Servelec's RiO system as
its strategic solution for mental health trusts and PCTs. RiO
offers considerable core clinical functionality as well as PAS
functions; it has now started to be deployed at a number of trusts.
9. GPs continue to use their existing systems
and have been little affected by NPfIT.
10. It is noted that responsibility for
managing further implementation of NPfIT is now being passed down
through SHA level to PCT Chief Executives.
Absence of firm plans for introducing clinical
aspects of CRS
The position regarding the all-important clinical
functionality of core CRS as at March 2007 is summarized as follows.
11. For acute and mental health trusts in
the three "iSoft" clusters (North West and West Midlands,
North East, and East). The 10 trusts who have taken iSoft's interim
solution, iPM, have the opportunity to deploy its clinical counterpart,
iCM, to obtain at least some core CRS clinical functionality in
the medium term. Those waiting for Lorenzo—the vast majority—must
wait probably until 2010-11, always assuming, of course, that
Lorenzo development is completed.
12. For acute trusts in the two "Cerner"
clusters (London and South). Little core clinical functionality
is available until Release 1 of the anglicised Cerner Millenium
product is available, timeframe as yet unclear. Fundamentally,
however, Millenium is a clinically rich, proven product working
successfully in the USA and at other overseas locations
13. For mental health trusts and PCTs in
London Cluster. Considerable core clinical functionality is potentially
available from the RiO product (see paragraph 8)
14. For mental health trusts and PCTs in
Southern Cluster. Fujitsu, the LSP, is only offering Cerner solutions:
therefore little core clinical functionality until Cerner development
available, timeframe unknown.
15. GPs will continue in the medium term
to enjoy the strong clinical functionality in their current standalone
GP system. NPfIT's initial intent for the longer term was that
the core systems being provided by the LSPs to the other care
settings would also be deployed for GPs but this strategy has
now been superseded by the GP System of Choice concept. My understanding
is that there will shortly be a formal EU procurement which will
result in CfH contracting with a number of GP system suppliers
from whom GPs will then be free to choose.
The original CRS vision is being lost
16. The delay and continuing uncertainty
of the timescale for introducing core clinical elements situation
described in paragraphs 11 to 14 means that even if things go
well from now on—by no means assured—it will only
have been possible to deliver a fraction of the intended clinical
functionality by the time NPfIT contracts expire.
17. But that is not the only concern. The
central purpose of NPfIT to provide a Local Care Record with detailed
clinical information and functions such as ordering tests and
information easily and immediately available across all local
care settings—GP, out-of-hours service, community, mental
health, acute hospitals—is being lost. With GPs having,
in effect, opted out no LSP will be able to deliver the original
intent of NPfIT of a single, truly integrated system. Nor in reality
does there appear much prospect of single systems serving the
subset of acute, mental health and community applications. Fujitsu
in the South cluster and CSC in the North West and West Midlands,
North East, and East continue to cling to this ideal but in fact
their respective software suppliers are nowhere near being able
to deliver it. BT in London which is currently achieving relatively
the most success in terms of introducing functionality of use
to clinicians (referring to the RiO implementations) has done
so by ignoring the original contracted plan to provide a single
suite of integrated software and introducing a standalone system
for PCTs and mental health trusts entirely separate from the Cerner
system for acute trusts.
18. Local CRS will remain a collection of
independent systems, not in any sense joined up, unless steps
are taken to introduce what is usually termed in IT-jargon as
an "integration engine", the function of which is to
sit over standalone systems and expedite information sharing between
them. No plans such plans have been announced. (The National Spine,
it should be noted, is designed neither to contain detailed clinical
information nor to provide the facility, for example, for GPs
to order hospital services such as pathology tests, and is therefore
not suitable for this integration engine function.)
Benefits delivery
19. Connecting for Health is focused on
delivering the systems. As and when software that supports clinical
functions is introduced it becomes essential if expected benefits
are to be obtained that all NHS organisations get to grips with
the management of change. To begin with it requires commitment
at trust Board level, commitment which it has to be said is increasingly
lacking as confidence in Connecting for Health's ability to deliver
ebbs. Assuming this can be turned round, there is then the need
for resource not only in the form of funding but also in freeing
local staff time. First and foremost it requires local "clinical
champions" to be made available at grass roots level. These
people have to be allowed significant time away from their normal
day-to-day duties to take charge of winning local support and
actually making things happen. It requires a number of senior
administrators and clinicians to think carefully through changes
to processes and procedures. It requires all staff to be released
for training. It requires organisations to be given the "space'
in their perform targets over the period of change. It requires
some measure of external, expert facilitation. Particularly in
the current financial climate it remains unclear where this resource
is to come from.
20. In 2002 the Department stated[110]
that "we will work closely with the Modernisation Agency
to change working practices so that IT is used effectively".
The Modernisation Agency ceased to exist in March 2005.
CONCLUSIONS
21. The Local Care Record Service in the
North West & West Midlands, North East, and East clusters
is dependent upon iSoft successfully completing the development
of Lorenzo. According to present plans, mass deployment of the
patient administration (PAS) element of Lorenzo will begin in
the 2009-10 timeframe, presumably with the all important clinical
elements to follow, say, from 2010-11. Given iSoft's track record
in developing Lorenzo and the company's unresolved financial problems,
the Department cannot be confident that this timescale will be
met.
22. Implementation of the Local CRS cannot
be said truly to have begun at any NHS organization in the North
West and West Midlands, North East, and East clusters. The majority
of acute and mental health trusts have implemented nothing. Some
10 acute and mental health trusts and a considerable number, around
50, PCTs have deployed as an interim measure an old, pre-NPfIT
iSoft PAS known as iPM. iPM has successfully filled an immediate
administrative need and with the addition of iCM may provide some
clinical functionality albeit primarily for acute trusts only.
Organisations which have implemented iPM have not, however, taken
any meaningful step towards CRS because moving from iPM to Lorenzo
(assuming Lorenzo development is completed) would not be a simple
upgrade but would in fact take on many of the characteristics
of a full PAS replacement.
23. There is now a real indication that
Cerner is coming good for acute trusts in the South and London
clusters and that deployment of clinical functionality in the
Millenium product will quickly follow that of the PAS deployments
achieved so far. But it must be emphasised that as yet the delivery
of clinical elements of CRS for acute trusts has not begun, nor
has any timetable for doing so been published.
24. It is not known when Cerner plans to
develop functionality within Millenium for mental health and community
(PCT) applications. Until such time mental health trusts and PCTs
in the South cluster will not be able to begin implementing CRS.
They have, however, been able genuinely to set out on the CRS
path in London because BT, the LSP, has adopted the CSE Servelec
product, RiO, as its strategic solution rather than wait for Cerner
development. Incidentally, the approach adopted in London reveals
what an opportunity is being lost, particularly to PCTs, in the
North West & West Midlands, North East, and East clusters
with the decision to use iSoft's iPM/iCM systems as interim solutions
when much more clinically rich alternatives are available.
25. The key objective of CRS to provide
a seamless service across all care settings—GPs, community,
mental health and acute hospitals—is in danger of being
lost now that the original intention to have a single integrated
system in each cluster has been abandoned. It has been agreed
that GPs can have their system of choice; and, at any rate in
London, an additional core software supplier has been brought
in to provide for the specific needs of mental health trusts and
PCTs. Whilst there is little doubt that a solution can be found,
there are as yet no plans for how a cross-organisational, joined
up CRS is to be obtained.
26. There is no evidence that NHS organisations
are committed to changing working practices in order to reap the
benefits of CRS as it begins to be deployed.
SUGGESTIONS ON
THE WAY
FORWARD
27. Before formulating any suggestions for
the way forward it is worth observing that the following principles
have been established by Connecting for Health:
(a) that a non-performing software supplier
will not be tolerated: witness the dismissal of IDX from the Fujitsu-led
LSP consortium in South cluster;
(b) that an LSP can have more than one core
software supplier: witness BT now using CSE Servelec in addition
to Cerner;
(c) that trusts may be offered choice: witness
plans for establishing a catalogue of additional systems suppliers.
[111](My
understanding is that CfH may be intending that this catalogue
be used only for non-core systems but there is no reason in principle
why choice should not also be available for core CRS.)
Bearing in mind the principles set out in paragraph
27 and the conclusions of paragraphs 21 to 26 the following questions
need to be asked and the results of more detailed investigation
acted upon.
28. What is the benefit to the NHS in keeping
iSoft as core CRS software supplier to CSC in the North West &
West Midlands, North East and East clusters given the company's
inability to date to deliver any CRS systems and that the future
availability of Lorenzo is by no means assured? Even if iSoft
is retained, why are NHS organisations not being offered the option
of an alternative supplier with a more proven product?
29. Should both other LSPs (Fujitsu and
CSC) be encouraged to follow BT's example in bringing a specialist
supplier of mental health and community (PCT) applications into
its consortium?
30. Can funding from NPfIT can be justified
for any further implementations of iSoft's interim iPM solution
given that iPM cannot truly be said to be a step towards CRS?
In the event that Lorenzo development is not successfully completed,
what would be the consequence to those trusts and PCTs which having
already implemented iPM are now tied in to iSoft?
31. Should the putative catalogue of additional
systems be extended to core systems thereby enabling local communities
to choose the core CRS supplier or combination of suppliers of
their choice? Would so doing create a greater feeling of local
ownership and thereby help obtain commitment to changing working
practices in order to reap the full benefits of CRS?
32. Can the current LSP contracts be reassigned
from the Secretary of State to SHAs or PCTs in order to give those
now being given responsibility for implementing NPfIT systems
the full authority and power to enable them to succeed?
33. Might the Local Information Strategy
(LIS) plans developed under "Information for Health', the
national programme that preceded NPfIT, provide guidance on achieving
cross-organisational information sharing? (It should be noted
that LIS work was very much focused on sharing information across
entire local health communities and that a number of local communities
were well on the way to getting systems in place when projects
were put on hold in 2002.)
34. Some of the questions in paragraphs
28 to 33 overlap; all are difficult to answer. A working party
should be set up as a matter of urgency. It should be chaired
by a senior member of the Department. It should be drawn from
expert witnesses who have made submissions to the Health Select
Committee and to the Committee of Public Account's pending report
on NPfIT, representatives of the NHS at SHA and trust level, representatives
of Connecting for Health.
Alan Shackman
March 2007
107 Connecting for Health Business Plan 2005-06. Back
108
The way in which information on the progress of implementation
is presented in the Connecting for Health website is poor. Users
are directed to the individual Clusters' sites for detailed data.
Such data, however, either does not appear or is in too summary
a format. It is not possible, for example, to obtain detail of
which precise elements of the Local Care Record Service has been
implemented, and where and when. Back
109
"We have secured bank funding until late 2007, but it is
clear that if the business is to prosper we must soon put in place
long-term funding arrangements", quote from Chairman's report,
iSoft interim results 11 December 2006. Back
110
"Delivering 21st Century IT Support for the NHS",
Department of Health, 2002. Back
111
As reported in E-Health Insider on 1 March 2007. Back
| 
