Evidence submitted by Statis Ltd (EPR
05)
STALIS LTD
Our credentials for making this submission to
the Select Committee are that Stalis Ltd, based in Oxfordshire,
has for the past 25 years provided a range of IT systems and services
to the NHS whose long standing clients include Moorfields Eye
Hospital NHS Trust, The Oxford Radcliffe Hospitals NHS Trust and
Sheffield Children's NHS Trust as well as the independent Capio
hospital group.
Following the introduction of the NPfIT Stalis
has majored on assisting Trusts and their LSPs with Data migration,
cleansing and the provision of interim Patient Administration
Systems (PAS). Over the past three years Stalis has undertaken
contracts with BT (for 9 London PCTs to migrate and archive data
from legacy Child health systems), Accenture (for 4 Essex based
PCTs to migrate and archive data from legacy Child health systems)
and Fujitsu (directly contracting with 4 NHS Hospital Trusts to
migrate, cleanse and archive data from PAS, A&E, Theatres
and Maternity modules). We are also in the final stages with Moorfields
of replacing their mission critical PAS due to the fact that BT
and its main systems partner IDX (and now Cerner) could not provide
an adequate solution to meet the Trust's needs.
Based on this past and current experience of
working with LSPs, their system suppliers and NHS Trusts we respectfully
submit this assessment as evidence of both the failings with the
NPfIT and the opportunities for the programme to be salvaged for
the benefit of the NHS, tax payers and patients. Specifically
the response addresses the item:
Current progress on the development of the
NHS Care Records Service and the National Data Spine and why delivery
of the new systems is up to two years behind schedule.
INTRODUCTION
1. The purpose of this submission is to
provide a synopsis of the background to and lessons that might
be learned from the NPfIT and its management by Connecting for
Health (CfH). It further indicates how the current problems with
the failure to deliver Care Record Services (CRS) to NHS hospital
Trusts can be turned round within the current investment levels.
2. As one of, if not the, most ambitious
public sector ICT project ever undertaken it is not surprising
that NPfIT has so far fallen far short on its overall delivery
goals, and in particular, the Care Record Service (CRS). In 2006
the National Audit Office report, indicated that the programme
is at least two years late. This is not to say all aspects are
suffering the same delays but the key components of the National
Spine, Choose & Book and, most of all, the Care Record Service
(CRS) for acute hospitals is significantly adrift from its original
goal.
Reflections on NPfIT so far
THE CONTRACTING
PROCESS
3. Given the magnitude of the undertaking
and the history of failure with previous public sector IT projects
(Passport Office, Child Benefits etc) the programme was placed
in the hands of an externally recruited senior IT figureRichard
Granger from Deloitte & Touche. With a launch budget of £2.3
billion and a timescale of approximately two years to put in place
all the necessary suppliers and contracts, NPfIT was launched.
This was exceedingly ambitious by any standards and the "haste"
contributed to a serious lack of NPfIT, supplier and solution
due diligence. The speed of procurement also precluded user (Trust
and specifically Clinician buy in to the solution). This lack
of buy in caused much of the delay during the early months of
deployment.
4. It is to Mr Granger's (and the CfH team)
credit that he achieved the early "political" goals
in that contracts were placed more or less within the two years
and the total contract value was held within the £2.3 billion).
However, all this was accomplished at a price and against some
ill founded concepts as set down below:-
5. Mr Granger had never undertaken a programme
approaching this magnitude of scale and complexityin fairness,
few if any haveand had no prior knowledge of the healthcare
IT requirements within the NHS.
6. From the outset Mr. Granger, and those
within government directly involved, appeared to believe that
all current systems operating within the NHS were in need of replacement.
(At a public meeting with the IT industry at the outset of the
programme Mr Granger was heard to describe current systems as
"rubbish") and that no UK suppliers had the capability
of meeting the needs. Furthermore, Mr Granger told the supplier
industry that "UK companies had not served the NHS well and
so could not expect much business". This demonstrated the
lack of understanding at the time that was instrumental in guiding
the programme.
7. The original concept was that all legacy
systems would be replaced in a few yearsa feat that was
never realistic and demonstrated a total lack of understanding
within NPfIT / CfH. The "Legacy Systems Team" within
CfH quickly came to recognise the value of the existing systems
and the magnitude of the task facing the NHS in replacing these.
CfH then renamed the team to "The Existing Supplier Team"
reflecting that some/many legacy systems are strategic and will
need to be retained for many years.
8. CfH made clear that they considered only
companies with no previous track record in the NHS were likely
to be successful, and in particular, US companies. This view prevailed
during the early stages of the procurement process thus encouraging
companies to participate who would otherwise have seen their lack
of experience as a material weakness (rightly) in any such project.
9. Only large global service companies (Local
Service Providers(LSP)) would be considered as main contractors
as they were the only ones likely to have the human resources
to scale up for the projects and the balance sheets to take on
contracts with severe financial penalties. The issue of whether
they had NHS IT experience was not viewed as key at the final
selection/contract stage.
10. Because Mr Granger had no previous experience
in healthcare he initially failed to appreciate the complexity
and diverse nature of the requirements. Similarly, none of the
LSPs had in-depth healthcare IT expertise in the UK, or really
understood the needs of the NHS. Thus for the first three years
of the programme (two years of procurement and the first year
of implementation) NPfIT was driven in an environment of ignorance
of the true NHS environment.
11. The funding for the projects did not
include the provision of funds for the migration and cleansing
of data from existing systems to the CRS or even acknowledge the
significance of this part of the program. In many cases, it was
not even understood that data migration into the new systems would
be required as part of the ongoing support of the Trusts' operational
systems. NHS patient data is generally of a poor standard with
most hospitals having electronic records that contain many duplications,
incomplete fields of data and data stored in areas of the systems
for which the systems were not designed. Trusts have records spanning
between 10 and 20 years so the need to clean these is significant
to the success of NPfIT and the CRS.
IMPLEMENTATION EXPERIENCE
12. The LSPs commenced the programme with
little or no experience in UK healthcare and little experience
anywhere of the systems required by the NHS. Although this has
improved with some LSPs it is not consistent across the NHS and
remains an issue today.
13. The LSPs appeared to place their reliance
on what their sub-contractors committed to deliver and placed
their trust in the contract terms to hold them safe. LSPs also
fell into the trap of believing the NHS would accept the systems
without question and would abide by project timescales imposed
by the LSPs. The naivety of this contributed to the early implementation
experiences when it became apparent Trusts would not accept (and
could not for resource and operational reasons) inferior systems
compared to what they already had.
14. There was little experience of the existing
"legacy" systems. In conjunction with this, there was
little appreciation of the number and variety of systems to support
clinical applications that were "held together" by the
existing legacy Patient Administration Systems (PAS) due to be
replaced.
15. The LSPs built project teams very quickly
in the naive belief that the NHS would accept their delivery schedules
and that the software would be ready. They, the LSPs did not have
the skill or experience to evaluate the software suppliers, hence
many of the problems with iSOFT, IDX and Cerner, are systemic
and continue to this daythe US company Cerner cannot provide
the reporting capabilities necessary to support NHS Trusts (as
is evidenced from the early users in the Southern Cluster).
16. Both IDX and Cerner systems have never
been successfully implemented in the UK, being systems configured
for the US market (which is one focussed on patient billing in
discrete establishments rather than the UK model of focus on administrative/clinical
systems over multiple site establishments).
17. Large teams of project staff worked
to a rigid methodology that was ill fitted to the NHS and the
implementation of clinical systems. Many Trusts acknowledged that
there was a huge culture gap between the NHS approach and the
LSP approach.
18. The NHS was not prepared for the magnitude
of change required, and is still not after more than three years.
Trusts are unprepared for the tasks required by NPfIT to create
a secure patient information system and for the data migration
involved.
19. Because of the delays and the unsuitability
of the CRS software especially at the hospital level, Trusts have
lost three years of progress and are now being forced into implementations
where they may lose important functions in their systems.
20. The project approach by the LSPs has
been to deliver to an implementation plan which has been ratified
by the NHS at the cluster level, but with no recourse or discussion
with existing suppliers. The existing system suppliers have also
been held at arms length from the LSPs and their partners delivering
the replacement systems thus ensuring that communication is at
best convoluted or at worst, non- existent. We have had experience
of this as a supplier of data migration services, not contracted
through the LSP, being prevented from direct communication with
the system supplier, and we believe this has also been the case
for the Trusts. The situation has improved in some areas but largely
due to local initiatives rather than a deliberate policy.
21. Although the UK healthcare IT supplier
industry contained a wealth of skill, experience and useful systems
these were largely ignored by the LSPs in the first two years
of implementation. This has changed somewhat but is still far
from ideal in that the LSPs do not make communication with outsiders
easy. This may be due to the conflict of interests in areas where
suppliers are in direct competition with the LSP for additional
services.
Lessons for future success of NPfIT
22. Attempt to introduce any major IT programme
against political led time schedules is fraught with danger and
should not be the basis for any future implementation timescales.
Although projects need targets, and there is nothing wrong in
having challenging timescales, project goals must be realistic
and those for NPfIT in general and CRS in particular were not.
23. Governments, of whatever political persuasion,
will likely wish to change their strategies on public sector health
care and so continually alter the data set and reporting requirementsas
evidenced by the 18 week wait policy. Specification creep is a
nightmare for NPfIT, Trusts and IT suppliers to deal with in a
project of this magnitude. Such changes also underpin the need
for NHS specific Patient Administration Systems rather than non
UK solutions which are designed for very different healthcare
environments such as the US.
24. Most healthcare systems vary significantly
country to country especially in respect of administrative, billing
and HR rules and requirements. It is these issues that cause the
most problems for healthcare providers and their suppliers. Systems
are typically designed for their home markets and not for a multi
country/world model approach.
25. The US healthcare market is very different
from the public sector systems found in the UK, Ireland and parts
of continental Europe. As such, any systems built in the US face
the prospect of the greatest changes.
26. Replacing working PAS and the like is
counterproductive. It is better to leave the PAS functions to
local suppliers whereas clinical applications are more widely
applicable and so can be added and allow faster progress to be
made.
27. Centrally driven projects are rarely
successful. It would have been far better (as supported by the
BCS) for CfH to set the standards, control the national applications
like the Spine, and allow the Trusts to choose from an accredited
supplier portfolio who meet the national standards (as advocated
for GP systems). This would have created local impetus, ownership
of the solution, and a vibrant competitive UK industry in which
the overseas firms could have competed but not dominated.
28. The success to the implementation of
systems is not having large global corporations. It is recognising
that the software must be fit for purpose and that this can come
from small companies with appropriate healthcare expertise within
the local economy. Large resources in service companies for volume
roll-out can be trained.
Impact assessment
29. Overall the programme remains in place
despite all the setbacks and failures. It is to the Government's
credit that they continue to support NPfIT and give encouragement
to all stakeholders to stay the course. The message that NPfIT
is material to the NHS and its reforms remains an essential underpinning
of the investment
30. For the NHS as a whole and its staff,
NPfIT has shown few benefits with the exception of improved networking
(part of the original plan) and GP payments, email and PACS (all
added later). By contrast many NHS Trusts have been significantly
disadvantaged by having to wait for the delivery of systems that
could have been provided by smaller UK suppliers (iSOFT notwithstanding).
31. Although the "central" message
is that the tax payer has been protected this is not true in that
ultimately the tax payer picks up the cost of delay and the increased
cost of the programme (now estimated at £12 billion). Delays
to delivery, the waste of NHS resources due to repetitive work
following inadequate software deliveries, and the lack of progress
in improving workflows all has a cost to the NHS and thus the
tax payer
32. The failure to modernise the NHS to
the extent required is in part due to the fact that NPfIT has
not met its objectives. However, NPfIT is not wholly at fault.
The NHS has not applied itself to the programme due in part to
the series of NHS re-organisations during the past four years.
33. NHS staff have and continue to suffer
from the burden of having inappropriate systems "not fully
fit for purpose" forced on them. The NPfIT made no allowance
for the fact that NHS staff and clinicians are under pressure
and have little time to devote to the magnitude of the change
especially when LSP's are forcing delivery dates on them as they
now are due to the financial pressures caused by non payment.
It is well known that the LSPs are faced with significant losses
which exacerbates the pressure on Trusts to "go-live"
with systems that are not fit for purpose nor rigorously tested.
34. Ruling out the UK supplier industry
from the programme was a fundamental mistake and all but destroyed
the sector. This has meant that much of the ICT skills and software
development potential that could have help build our health IT
industry into a world class industry was handed to mainly US companies
(Cerner, IDX, Accenture, CSC). Now, due to the delivery slippage
and underperformance the value of UK suppliers and their systems
is being recognised and has the potential to save the programme.
35. The belief that only large companies
could handle the programme was flawed. It is true that large volumes
of human resources are needed for managing and implementing multiple
concurrent projects. However, the production and supply of NHS
specific software does NOT require large companies. On the contrary
the majority of successful systems in the NHS come from relatively
small suppliers. If NPfIT had appreciated that the basic software
could be provided by UK suppliers the LSPs could have been trained
to implement the software to meet the CRS requirements.
The Way ahead
36. The NPfIT has achieved some considerable
and measurable improvements in the provision and application of
modern IT systems for the benefit of the NHS and its patients.
Most notable among these are PACS (imaging systems), the BT provided
N3 broadband network, and some GP/Community systems. Choose &
Book and the Spine have made some progress but both remain some
way from their original goals. It is noteworthy that the most
progress towards Choose & Book has been made by Trusts enhancing
their existing PAS, not replacing for one of the new NPfIT solutions.
37. The Care Record Service (CRS) is the
one programme within NPfIT that has the worst record of delivery
and yet it is, perhaps, the most needed. Although LSP CSC continues
to implement the legacy iSOFT system to the benefit of those Trusts
with archaic legacy systems it is not the modern futuristic solution
(Lorenzo) originally contracted. Likewise, in the London and Southern
clusters, both Fujitsu and BT are underperforming due to the Cerner
Millennium product needing far more change than expected, and
there are doubts that it can ever meet the underlying and changing
needs of the NHS for a robust NHS workflow related PAS which underpins
CRS.
38. However, much of the time that has been
lost could be made up if the strategy was changed. What we respectfully
suggest is that consideration be given to the following:-
39. There are many Trusts with perfectly
adequate PAS that could be supported for many years to come. Instead
of replacing these they should be retained and clinical systems
added to them. This offers significant benefits
(a) Although the data cleansing would still
need to be undertaken there would be no data migration
(b) The Trust's operational support is unaffected
and resources are not wasted on a one year cycle of replacement
(c) Clinical systems can be implemented to
the benefit of clinicians and patient care far sooner than with
the current programme
40. Those Trusts that need to replace their
PAS should use pre existing UK developed systems (of which there
are several) that already meet the NHS requirements especially
for reporting. This strategy would allow BT and Fujitsu particularly
to recover lost ground and would provide a sound IT platform on
which to implement the Cerner clinical applications. It would
also overcome the lack of reporting capability in the Cerner product.
41. CSC has demonstrated with iSOFT that,
despite the latter's inability to deliver its new system Lorenzo,
the deployment of a UK built PAS (iPM) can meet NPfIT needs. CSC
has implemented more PAS than the rest of the LSPs combined (including
Accenture prior to their departure). CSC is showing that a UK
PAS with a separate solution for clinical applications is viable
(the iSOFT iCM product originated from the US).
42. Such a strategy would also save the
programme significant new expenditure. We understand that when
CfH/DoH requires changes (usually known as Data Set Change NoticesDSCNs)
the charge made by Fujitsu/Cerner for even minor amendments is
substantially in £six figuresthis would be avoided.
43. Trusts should be free to choose which
PAS suits them best from an approved list of suppliers which are
committed to meeting the mandatory standards of the NPfIT .
44. With this approach the modernisation
of IT within the NHS can gather pacewithout this the CRS
programme will continue to under-perform to the detriment of the
NHS, its patients and tax payers.
R Roger Wallhouse
Chairman, Stalis Ltd
March 2007
|