Select Committee on Public Accounts Minutes of Evidence

Memorandum submitted by David Kwo, Alan Shackman, Bernard Hunter and various NHS staff

CONTENTS

BACKGROUND: THE CENTRAL POINT OF NPfIT........................................................................................................

EXECUTIVE SUMMARY ..................................................................................................................................................

1.INTRODUCTION .............................................................................................................................................................

2.CURRENT CONTEXT .....................................................................................................................................................

3.WHAT THE NAO REPORT FAILED TO DO ................................................................................................................

4.QUESTIONS THAT THE NAO REPORT FAILED TO ASK.........................................................................................

5.CONCLUSIONS ...............................................................................................................................................................

6.RECOMMENDATIONS ..................................................................................................................................................

Annex 1—Evidence Submitted to PAC in June 2006 ..........................................................................................................

Annex 2—Information Sources Website Addresses .............................................................................................................

Annex 3—Modules to be Delivered by the LSPs .................................................................................................................

Annex 4—Economic Case of Eastern Cluster ATP ..............................................................................................................

Annex 5—Extract from a Recent PCT Report .....................................................................................................................

Annex 6—Contribution from Bernard Hunter......................................................................................................................

BACKGROUND: THE CENTRAL POINT OF NPFIT

  By way of background to NPfIT, it is useful to begin with the question: "What is the central point of NPfIT—its chief raison d'etre? Is it a shared medical record (otherwise known as the `Central Spine' or `Central Summary Care Record Service') across England?"

  The answer to this important question is simply: no. In fact, the central point of NPfIT is to provide the local Care Record Service (CRS) systems, namely: Cerner Millennium in the Southern and London Clusters and iSoft Lorenzo in the North East, North West & West Midlands, and Eastern Clusters. Compared with the local CRS, the Central Spine is a much lower priority because it is totally speculative and even if delivered is likely to result in very little clinical benefit. The Central Spine is actually a distraction and anyway cannot happen without the Local Care Record Service (or Local CRS) systems being implemented first.

  This is a subtle but critical point. The Local CRS systems, historically always known as "electronic patient record" (or "EPR") systems, are a proven technology and typically aimed at the local NHS organisation (acute hospitals, community and mental health hospitals and GP practices). They are sophisticated software systems and are quite complex because they need to cater for a range of clinicians and accomplish a wide range of tasks such as allowing the clinicians to maintain the clinical history of the patient, to request diagnostic tests, to prescribe drugs, to schedule theatres, clinics and physiotherapy, etc. and do so in intelligent ways with a great deal of embedded clinical knowledge to make clinical practice safer and evidence-based.

  These types of software products (Cerner "Millennium", GE/IDX "Carecast", etc.) are well established and proven and earlier versions of them are successfully operating in several hospitals in the UK. They are what is most needed by hospitals to support their clinical service delivery. They are what have been contracted for from the Local Service Providers (LSPs) as specified in Schedule 1.1 of the LSP contracts.

  These local CRS systems have always been costly investments (several million pounds per hospital over several years) but have been proven in the NHS and elsewhere to deliver real clinical benefits (such as reducing adverse drug events and reducing mortality) which are evidenced by sound and extensive published medical research.

  This picture is entirely different for the so-called Central Spine record, or Central Shared Summary record, which NPfIT (and the government ministers) would like the public to believe is the central point of NPfIT. It is not. The Central Spine record is just a concept, the simple one of having a summary record about a patient (with his/her key clinical data such as allergies and latest drugs) which can be shared by all clinicians needing to have it.

  The problem is that clinicians have told us medicine does not work like this. Clinicians do not just use a summary record to deliver care. They build and depend upon detailed and specific medical data that are relevant for each patient. They do not rely on some other clinicians' definition of what will be most relevant to put in a summary record. What is relevant clinically will inevitably vary from patient to patient.

  The concept of a summary Central Spine record has no scientific basis and no significant clinical support to back it up—just an overly simplistic and naive storyline about a Birmingham patient falling ill in Blackpool. In fact, no one has ever provided any figures on how often this situation is likely to arise to show whether or the investment in the Central Spine record worthwhile.

  The point here is that the Local Care Record Service is the essential building block for clinically useful health IT to support clinical care in progressive, modern and proven ways. Yes, it is difficult to implement and can take 2-3 years to roll-out across the whole hospital (or organisation), and yet it is always worthwhile—ask any of the Chief Executives of the few hospital trusts that have implemented these systems in the UK e.g. Guys & Thomas, Chelsea & Westminster, Burton Hospitals Trust, Wirral Hospitals, etc.

  These Local Care Record Service systems are the building blocks and are the point of NPfIT, and what NHS Trust Chief Executives want, need and expect. They are not waiting for a Central Spine record to run their hospitals.

  However, the Local Care Record Service systems (or the Local Service Providers' newest versions of them) are not likely to be fully deployed now (only the rudimentary patient administration elements of them will be) because NPfIT is putting in old "legacy" products in place of new modern Local Care Record Service products in its panic to show deployment and because the systems have been so late in being delivered by the LSPs.

  What is not reported widely is that the LSPs are late in delivering the new versions largely (though not wholly) because of NPfIT's own delay in providing so-called "Spine message definitions" to the LSPs during 2004 and 2005. And when they did finally arrive, they were fluid and kept changing.

  Thus, the LSPs have been delayed because of the delays and failures of NPfIT itself, but of course they are afraid to say this for fear of offending their client.

  In summary, the Local CRS agenda is the real agenda for NPfIT, not the Central Summary Care Record Service.

  The key point of the National Programme for IT is to provide both depth of clinical systems functionality and breadth of integration in terms of delivering the contracted Local CRS functions across organisations and care-settings (acute, primary, mental health, social services). This is the true vision of health IT promised by the National Programme which is embodied in the Local Service Provider contracts and it is what their price reflects.

  The trouble is, with all the delays, the LSP schedules are being down-scoped behind the NHS's back and without any accountability to the local NHS Trust chief executives to whom the original vision was promised.

EXECUTIVE SUMMARY

  This submission to the Public Accounts Committee (PAC) was produced in response to the NAO Report on "The National Programme for IT (NPfIT) in the NHS" (HC 1173 Session 2005-06 16 June 2006). It has been produced to assist the PAC in evaluating the findings of the NAO Report. It is based upon professional opinion, additional facts and references to publicly available documents.

  The NAO Report attempted to evaluate the performance of the NPfIT since its inception. The NAO Report failed to ask key questions and to explore crucial evidence regarding NPfIT, in particular:

—  NPfIT is critically behind schedule. By now, all acute hospital trusts in England should have new NPfIT patient administration systems (PAS) implemented as the essential first step in the introduction of the Local Care Record Service. As of April 2006, according to the NAO Report, the actual number was 9 hospitals. Since then, so far as we are aware, the number has increased only by 4 acute hospitals[23]. All have experienced significant operational, clinical and financial disruptions because the systems are not fit for purpose. In most cases, these serious disruptions are on-going. The prognosis for future functionality beyond the basic Patient Administration Systems phase (particularly the implementation of clinical systems) is "poor to terminal".

—  The severity of these delays—and the consequential negative impacts on the NHS, its patients, organisations and clinicians, and on the realisation of the benefits upon which the original investment business case was built—was not explored in any depth by the NAO. While the NAO reports on the relatively sparse adoption of systems through NPfIT, extraordinarily the adoption levels were not contrasted with original targets. While the current level of delays was still only emerging when the NAO Report was being prepared, the slow trajectory was already evident at that time.

—  The original business case for NPfIT was predicated on a certain level of clinical quality, as well as operational and financial benefits. These benefits are now unlikely to be realized at anywhere near the originally projected levels, calling into question the value for money of current and future NPfIT expenditures.

—  The "Spine", or the Central Shared Summary Record—or that part of it which is intended to gather patient clinical data from across the NHS—is an untested, experimental concept that is best characterised as unnecessary, unlikely to succeed and likely to cause continued delays to the Local CRS systems that are actually required to deliver the anticipated benefits.

—  Delays in Local CRS systems implementation have been caused in large measure by delays in the "Spine", but not exclusively so. The other key delay factors were: the lack of clinical engagement during the system selection process; the climate of aggression and hostility engendered by Connecting for Health (CfH); and the failure adequately to consult experienced advisors on the implementation of NHS clinical information systems.

—  The over-emphasis on very quick procurement was paradoxically a key element in this delay because rushed contracts only exacerbated the lack of clinician buy-in, the lack of informed product development and so the lack of successes in implementation.

—  Misconstrued views of the importance of the LSP role have led CfH to placate the remaining LSPs, who are under financial pressure due to lack of revenues, by giving them faster routes to revenue. This has meant short-cuts being taken by the LSPs in local implementations at the expense of the operational and financial well-being of local Trusts. For example, one LSP is not including historical patient data in its new system to save time and effort. This Trust will therefore have to keep its old system running as well as the new one so that, for example, A&E department staff can identify that a patient has been admitted previously. This is particularly important for children at risk. In this case, the LSP's approach is dangerous and analogous to building an aeroplane without life jackets under the seats—one day it will result in a safety disaster.

—  In the early LSP contract days, the transfer of risk to LSPs led to a "moral hazard" where the appearance of LSPs bearing the risk of implementation removed from hospital Trusts the local sense of urgency and responsibility. Now, the risk is being transferred the other way to the local Trusts, ironically enough, to the very people that risk transfer in the contract was meant to protect, ie the NHS.

—  In this context, it seems that CfH is colluding with the LSPs to make sure that the LSPs get paid, without caring about the damage that is caused to the Trusts either operationally, clinically, financially or in the form of lowered staff morale. CfH is now trying to legitimise this harm to the local Trusts, and washing its hands of any responsibility at the same time, by respraying a failure into a policy through making local trusts responsible for local implementation.

—  CfH continues to engender a general climate of fear resulting from the bullying that is occurring on the ground. Chief Executives—particularly the vulnerable ones with financial deficits—are being leaned on by their bosses to implement the LSP products and to keep quiet, no matter how painful it is on the ground. This is the case even where the pain is specifically caused by the LSP's inadequate products and where this inadequacy has led directly to: the loss of star-ratings (as in the case of the Nuffield Orthopaedic Centre); the inability to report activity and thus the loss of income; as well as having to pay higher prices than would be available on the open market (for example, with the Picture Archiving and Communications Systems).

—  CfH generates what is best described as propaganda, for example in its claim to have deployed "A total of 9,600 initial deployments of software of various types..." (NAO Report, Page 3). The truth is that the vast majority of these systems are small departmental systems (such as standalone Theatre Department systems) which were not included in the original scope of NPfIT procurements and which in any event Trusts had been routinely implementing prior to NPfIT as part of their normal replacement programme. This report of deployments, as a meaningful measure of achievement, would be analogous to a spokesperson for the DeLorean corporation, prior to its going bust, declaring that they do not know how many units of motor vehicles they produced and sold during their last financial year, but instead reporting that the DeLorean canteen did produce and sell 9,600 sandwiches in the period. This hypothetical DeLorean statistic is as irrelevant as it is misleading as a measure of achievement and the NAO Report of "9,600 initial deployments" is no more meaningful and no less irrelevant and misleading.

—  One detects the distinct smell of the Emperor's New Clothes at every encounter with the National Programme for IT.

CONCLUSIONS

—  The NAO Report has failed to spot poor CfH performance in the areas of: NPfIT business case delivery; benefits delivery; implementation progress; Spine delivery; the fundamental reasons for delays; accountability to the NHS; management of LSPs; delivery by LSPs; addressing the fundamental problems and possible fraud at iSoft, and other NPfIT elements such as N3 Broadband, ETP (Electronic Transmission of Prescriptions), Choose & Book and Picture Archiving and Communications System (PACS).

—  The NAO Report has missed the fact that CfH has lost its way and has abandoned its vision to reduce NHS fragmentation of systems and care. CfH is driven by no vision at all except the desire to retain central authority using bullying tactics on both suppliers and NHS Trusts.

—  CfH has failed. Having been given the responsibility for the largest sum of money (now stated at £12.4 billion) ever allocated to a health IT programme anywhere in the world—a number which dwarfs the total NHS deficit of around £500 million per year—it has conspicuously failed to deliver.

—  The CfH mantra is "deploy anything that moves just so we can say we are deploying something" with no care for the wider picture or indeed value for money.

—  Any clinical benefit which may eventually derive from the Central Summary Care Record is likely to be miniscule compared to the evidence-based benefits that the NHS needs and which have been contracted for in the form of Local Care Record Service systems (otherwise known as Electronic Patient Record Systems) for each NHS acute, community and mental health Trust in England.

—  Not one of the LSPs has delivered to contract. The NAO should not have praised the so-called "speedy" completion of a contract procurement process which was so clearly flawed. The extraordinarily hasty way in which the original LSP contracts were let, ignoring all other considerations such as the need to achieve buy-in from clinicians and from chief executives of local Trusts, is one of the main causes of the enormous problems which have followed in attempting local deployments. The NPfIT procurements can only regarded as failures.

—  GPs have in effect opted out, putting paid to the concept of "ruthless standardisation".

—  Our conclusions cause us therefore to take serious issue with the key conclusion of the NAO Report that "The Department and NHS Connecting for Health have made substantial progress with the Programme"[24]. All the evidence, particularly that relating to delivery of the originally envisaged Local Care Records Service functionality, points the other way.

RECOMMENDATIONS

  I.  The Department of Health should take steps to:

(i)  Retain the sensible aspects of NPfIT, namely: (a) ring-fenced money for IT; and (b) discounted prices.

(ii)  Stop the tactics of bullying NHS Trust Chief Executives into taking take unfit systems and stop the climate of secrecy surrounding CfH where people are afraid to speak out, even to share lessons learnt or to suggest how problems could be solved.

(iii)  Empower the local Trust Chief Executives with final authority to decide when payments are made to LSPs.

(iv)  Recast the contractual relationship between the LSPs and each NHS Trust so that each Trust is legally the customer rather than the Secretary of State for Health.

(v)  Restore the NPfIT vision of sharing information and functionality (ie wide area, cross organisation and cross care-setting Local CRS systems) particularly across local health economies or care communities (as is already stated in the LSP contracts).

(vi)  Put the "Spine" investment on hold and ask clinicians and IT specialists to pilot and test it objectively and with scientific rigour.

(vii)  Disband the central NPfIT team within CfH in favour of allowing systems procurements and implementations to be handled by local care communities in accordance with agreed standards. This level of management would support the integration of the different care settings such as primary and community care, acute hospital care, mental health care and social services, and therefore achieve the main benefits of using IT to deliver patient-centred care.

(viii)  Make the money which is saved from reducing central overheads available locally for change management for each care community.

(ix)  Establish an open framework of CfH accountability where future decisions on NPfIT budget and central IT contract changes are made in full consultation with, and with accountability to, patients, local NHS trust management and clinicians.

(x)  All NHS trusts in the West Midlands & North West, North Eastand Eastern Clusters should be asked whether they wish iSoft to remain the sole subcontractor for the local Care Record System, with appropriate action to be taken to reflect their answer.

  II.  The NAO should:

(i)  Review its methods of investigating large IT programmes in the light of the points made in this submission eg measuring progress should be undertaken against baselines and appropriate metrics rather than merely accepting what the programme managers claim.

(ii)  Address the shortcomings of its report on NPfIT according the section 3. below.

(iii)  Ask the questions set out in section 4. below.

1.  INTRODUCTION

  This document presents evidence to the Public Accounts Committee (PAC) regarding its investigation into the NHS National Programme for IT (NPfIT) and the NAO Report on the same subject. This paper is provided at the request of Richard Bacon MP as a member of the PAC. It consolidates an earlier submitted to Richard Bacon (see Annex 1-Ev) in June 2006 and is updated here to include more recent evidence.

  The evidence contained in this document is based solely upon publicly available information ie information from websites that are accessible by anyone with access to the internet. The specific websites containing the source documents used are listed in Annex 2.

  This document has been prepared by David Kwo with assistance from Alan Shackman and contributions from various colleagues working in and around the NHS. David Kwo was formerly the Regional Implementation Director for NPfIT (London) and Chief Information Officer (CIO) for the NHS in London and Alan Shackman is a management consultant who works for various NHS organisations.

  Specifically, this paper:

—  identifies what the NAO Report on NPfIT failed to do;

—  raises questions that the NAO Report failed to ask;

—  draws conclusions regarding NPfIT; and

—  makes recommendations for improving NPfIT.

2.  CURRENT CONTEXT

  This paper focuses on the deployment of the core Local Care Record Service (CRS) originally contracted for from the Local Service Providers (LSPs) by Connecting for Health (CfH), this being the central plank of the NPfIT vision and to which the overwhelming majority of the funding is allocated. It largely excludes consideration of Picture Archive and Communication Systems (PACS) which were not originally included as a core deliverable; it also excludes the plethora of relatively minor acute and primary/community departmental systems that LSPs have implemented.

  The position with the Local Care Record Service at January 2007 is summarised as follows:

I.  For acute trusts in the three "iSoft" clusters (ie North West and West Midlands; North Eastern and Eastern): no core clinical functionality for virtually all Trusts until at least 2009-10 unless trusts are prepared in the interim to move to iSoft's pre-Lorenzo legacy PAS (Patient Administration System) product which many Trusts consider inferior to their own present PAS.

II.  For acute trusts in the two "Cerner" clusters (ie London; and Southern): little core clinical functionality until Release 1 of the Cerner product is available, timeframe as yet unclear because of delays with anglicising the current Release 0 (PAS) product.

III.  For mental health trusts and PCTs in the three "iSoft" clusters North West and West Midlands; North Eastern and Eastern): possibility of limited clinical functionality from 2007 dependent upon development of iSoft's pre-Lorenzo legacy iCM system, but no core clinical functionality from the strategic Lorenzo solution until at least 2009-10 if at all given the many doubts about the existence of Lorenzo.

IV.  For mental health trusts and PCTs in Southern Cluster: no core clinical functionality until Cerner development available, timeframe is unknown.

v.  For mental health trusts and PCTs in London Cluster: considerable core clinical functionality is available now from standalone (ie non-integrated) software, CSE Servelec's RiO system, which has been adopted by BT as its strategic solution because Cerner's integrated mental health/community product has not yet been developed.

VI.  GPs remain free to use the GP system of their choice and hence are little affected by NPfIT. There are, however, recent reports of a small number GPs in the North East and Eastern clusters implementing a system originally offered by Accenture and now by CSC.

  The situation as summarised above invites the following comments:

VII.  NPfIT is at least 2 years behind schedule and it is likely to be a further 3 years before many trusts start to have any significant clinical software available to them, always assuming software development currently being undertaken is successful. In the context of a 10 year programme this is nothing short of disastrous

VIII.  The LSP which is achieving relatively the most success in terms of introducing functionality of use to clinicians (referring to the RiO implementations in London) has done so by introducing a standalone system and ignoring the original contracted plan to provide a single suite of integrated software and adhere to "ruthless standardisation".

IX.  The central purpose of NPfIT to provide a Local Care Record is in danger of being lost. With GPs having, in effect, opted out no LSP will be able to provide a system that is integrated across all local care settings (GP, community, mental health, acute hospitals). Unless, that is, they develop what is termed an "integration engine" the function of which is to sit over standalone systems and expedite information sharing between them. The danger is compounded in the case of London with the adoption of a standalone strategic solution for mental health and PCTs.

3.  WHAT THE NAO REPORT FAILED TO DO

I.   The NAO Report failed to understand the Business Case benefits upon which the LSP contracts were based

  The Local Service Providers, LSPs, were contracted to deliver Local Care Record Systems, CRS, to NHS organisations in three Phases. The LSPs are already three years late in delivering Phase 1 and it is expected to be another two to four years before even that can be implemented, if then, due to mismanagement of the suppliers and the fact that the software products, despite over three years of preparation, are still not fit for purpose. That is, NPfIT will be at least five years late in delivering just the first element of its main programme. This element (Phase 1) is the least important element from a clinical care point of view because it contains mainly administrative functionality.

  The next two phases, Phases 2 and 3[25], are meant to provide the NHS with functionality that would enable organisations to support integrated clinical care processes (scheduling, investigating, prescribing, treating, assessing, etc.) by healthcare staff no matter in what organisation (hospital site or GP practice) or in what care-setting (primary, mental health, community, tertiary). The patient could move from one care provider or setting to another and the detailed patient record, and, importantly, the functionalities needed to care for the patient, would be available to the care provider in a consistent and standard fashion. The care provider would not have to log into and be familiar with different screens, and search for the same patient each time, and the patient would not have to be asked the same questions by different care providers at different visits.

  These benefits were the core of the business case for the high cost LSP contracts, as shown in the Eastern Cluster Business Case (see Annex 3 and 4). The high LSP costs (around £1bn per contract) were felt to be justified when originally presented to local NHS chief executives (at Trust and SHA levels) because of the new integrated care benefits that were being promised in the Cluster LSP Business Cases.

  However, these crucial benefits were not acknowledged by the NAO Report which means that the NAO assessment of NPfIT performance was not measured against proper baseline metrics (ie the expected benefits).

II.   The NAO Report failed to measure NPfIT progress using basic contracted milestones; it missed the expected shortfall at contract out-turn; and it failed to measure Value for Money.

  The NAO Report should have used publicly available data[26] to measure the progress of NPfIT in terms of the contracted modules and the contracted delivery dates. The LSPs were contracted to deliver these modules to all Trusts in the NHS in order to achieve their milestones. The NAO Report failed to use this measure and this is a major shortcoming of the document.

  By either 2010 (when LSPs are contracted completely to implement all 3 phases to all hospitals and trusts in England) or 2013 (the end of the contract, when other items were meant to be delivered, too), only Phase 1 of the 3 Phases will have any chance of being delivered (many NHS staff doubt that even Phase 1 will be delivered fully).

  However, the later the module, the more clinical it is in nature; and thus, the larger the number of clinicians who need to be engaged with it; and therefore the more time overall it will take to implement. This militates against LSPs meeting their contracted delivery milestones, given the deep gap in clinician engagement that NPfIT has allowed to develop.

  In terms of software product, CSC and Accenture[27] (who are the LSPs in the North West & West Midlands; North Eastern; and Eastern clusters) themselves reported in February 2006 that there is "no believable plan" for the development of Lorenzo, which is the software product upon which Phases 1, 2 and 3 are based for these clusters.

  Therefore, hospitals in England cannot expect Patient Administration Systems (PAS) products that are integrated across different NHS organisations but instead can only expect yet more organisation-specific PAS, which each hospital had anyway prior to NPfIT. Similarly, they can only expect yet more organisation-specific Order Communications Systems (OCS) which was also already available before NPfIT. NPfIT is thus perpetuating the fragmentation of records, functions, processes and care within the NHS which the LSP contracts were meant to overcome. This perpetuates what the NHS has always had (PAS), and what the NHS was already on course to achieve (OCS) prior to NPfIT.

  If we will only get what we've always had, or were on course to get, is the price we are paying the LSPs good value for money compared to what we were paying previously? Did we need NPfIT, with its additional £1.5bn+ central overhead costs, to bulk-buy what NHS Trusts were already buying for themselves?

III.   The NAO Report failed to recognise that the Central Spine Record is unproven and only likely to deliver relatively small benefits.

  CfH is attempting to claim that the Central Spine Summary Record will provide the cross organisation and cross care-setting integration promised in the LSP business cases. However, this is not a plausible claim because the Central Spine Summary Record offers no proven clinical benefit. It will only have partial patient record data and no (or little) clinical system functionality.

  Furthermore, due to the delay in functional deployment of the local CRS, it is unlikely to achieve any administrative let alone clinical effectiveness at all for many years.

  The concept of a summary Spine record has no scientific basis and no significant clinical support to back it up—it is just an overly simplistic and naive storyline about a Birmingham patient falling ill in Blackpool. Yet no one has ever provided any figures on how often this situation might arise or whether the incidence would be high enough to make the investment in the Spine record worthwhile. There are also severe doubts about whether the problems of patient confidentiality can be overcome.

  By contrast, the local CRS functionality is evidence-based because there is published scientific evidence showing how it improves the quality of care (eg by reducing medical errors in the form of adverse drug events).

  Table 1 below shows that when the two systems (Local CRS and Central Spine Summary Record) are compared, it is impossible to maintain that the Spine Summary Record will provide anything more than the most marginal of benefits and even then only potentially.

Table 1:  Comparison of Local CRS and Central Spine Summary Record

Feature	In Central Spine Summary Care Record	In Local Care Record Service
Provides functionality for clinicians (prescribe a drug, order a test, book a treatment, create a discharge summary, send a communication to another clinician, etc).	No	Yes
Provides detailed clinical information for that patient's episode under the patient's lead clinician.	No, Spine clinical data will always by definition be incomplete for each of its patient records.	Yes
Provides real time alerts if a patient for instance has liver dysfunction (eg high creatinine levels) thereby contra-indicating an antibiotic being prescribed (eg gentamycin) or levels of potassium when prescribing Digoxin.	No, Spine alerts will not be real-time and will always be out-of-date because the uploads are always delayed.	Yes
Is evidence-based: can offer scientific publications to justify its benefits	No, the Spine as a passive summary record has no scientific evidence behind it.	Yes, as a real-time alert at the point of care, the Local CRS (or electronic patient record) has incontrovertible evidence behind it.
Is likely to provide real clinical benefit to large numbers of patients.	No, the statistical probability of a clinician finding useful clinical data on the spine that it has not already been received has not been proven.	Yes, the probability of reducing medical errors and therefore improving patient safety is very high, at least compared to the Spine.
Has been proven to work in the NHS with NHS clinicians of all types using it.	No, the Spine summary record has never been developed or tested, even in pilot form, amongst NHS clinicians.	Yes, several dozen NHS organisations, comprising thousands of NHS clinicians, are actively using PAS, OCS and to a lesser degree, electronic prescribing with real success.

  The Local CRS has two main features: (1) real-time alerting of e.g. drug allergies at the point of decision-making (because the local Care Record Service is, when implemented, the only means for a doctor to prescribe a drug) and (2) storing data about the patient's allergies to drugs. Although the Central Spine stores data about patient allergies, at no point would the Spine actually be used to prescribe a drug, for example, because it would not provide the functionality to do this; indeed, it does not provide any functionality at all. The evidence[28]8 showing the benefits of clinical systems in terms of avoiding adverse drug events, such as preventing doctors from prescribing a drug to which the patient is allergic, is based on clinical systems that have both characteristics (1) prescribing functionality; and (2) storing patient data. CfH is trying to say that the Central Spine will reduce adverse drug events because it has characteristic (2), without mentioning that it does not have characteristic (1).

  This is not a scientific basis upon which to claim that the Central Spine will reduce adverse drug events. If CfH wants to make such (and similar) claims, then it will need to produce supporting evidence which matches the features of the intervention it is proposing with the outcome it is claiming. It is not enough to say that the Central Spine shares some characteristics (eg holding patient allergy data) with Local CRS systems and then claim that it will therefore generate the same benefits. It won't.

  The NAO should have recommended that the Spine be rigorously investigated and tested before further millions of pounds in funding and resources are put into its development. One system (Local Care Record Service) is evidence-based, the other (the Spine) is not. Both should be pursued, but one should be pursued with more confidence and commitment because the evidence justifies it. The other should be properly tested before any commitment is made to develop it or roll it out.

  A common axiom within healthcare, missed by the NAO Report, is: mere provision of information does not lead to change in clinician behaviour.

  The Spine has caused delay to LSP delivery of Local CRS systems, due to the inordinate amount of time it took for CfH to issue its Spine message definitions to the LSPs.

  The Spine has served as a decoy for the NAO, media, ministers and the NHS. People have been lulled into thinking that the central objective of the NPfIT is the Spine and that CfH is on the way to achieving something clinically beneficial, when actually the Spine is a speculative concept without empirical evidence. The empirical evidence demonstrating e.g. reductions in medical errors, cannot be attributed to the Spine because the research is based on benefits from real-time active Electronic Patient Record (EPR) alerts at the point of care, using EPR systems where electronic prescribing or ordering is the only mode of practice for the clinician. This will not happen with a passive summary record where the clinician does not have to access the record in order to prescribe and might only do so, time permitting, in the hope that there is something useful there. Statistically, the chances of finding clinically beneficial data not already held has yet to be established.

  It is important to note that the National CRS Central Spine record will not deliver the claimed benefits as it requires functionality which is only available with the Local Care Record Service solutions which contain the full and detailed patient record and advanced and intelligent functionality. By contrast, the Central Spine Record is just a passive "bucket" or repository of what will only ever be a subset of the patient medical record.

  The focus on the Spine is deluding many people into thinking that the clinical challenge in NPfIT is simply about getting the right clinical data defined for the Spine Summary by some "representative" group of clinicians, and that once this is done then all that is needed would be to link into GP systems (and eventually hospital systems) and then copy this data into the central summary record "bucket". If the clinicians don't use the Spine, that will be their fault, but CfH will have done its job, according to what CfH would have us believe.

  The Spine is clinically invalid. Clinicians have consistently told us that they consider it sheer arrogance to think that a group of clinicians can decide on behalf of all clinicians what the key data items are for all their patients and casemix and circumstances. Clinicians don't mind using data from other clinicians, their work depends upon it. What they do mind is a "committee" telling them the important data fields that they are to use for all their patient care practices and range of patient conditions.

  The Spine distracts the NHS, perhaps deliberately, into thinking that what is important in computerising the NHS is data (eg the things that the Spine may hold), rather than functionality (eg the operational systems support that the Local CRS provide). An age-old lesson in the health IT sector is that "Reliable data is best derived from systems that are relied upon". That is to say, where clinicians are using systems routinely as an integral part of treating patients, and where clinicians are in effect `forced' to use the system to take the next clinical step, then the data will generally be complete and accurate. Where the system is merely a passive tool which is available for reference but which the clinician is not required to use in order to progress treatment, then the data is likely to be of a much lower quality. This lesson has not been learned by CfH. It is trying to shortcut improvements in clinical practice by building a data bucket, rather than by working with clinicians to redesign clinical processes and embed clinical knowledge into systems in ways that clinicians deem appropriate and can use routinely as part of day to day clinical practice.

IV.  The NAO Report failed to detect the real causes of NPfIT delay

  Clinician involvement has been virtually ignored by NPfIT, in the crucial sense of engaging clinicians on the front-lines (in hospitals, GP practices, mental health facilities, etc) and getting them to take "ownership" of the NPfIT and the new systems that were being procured centrally on their behalf. NPfIT made the mistake of rushing to break a speed record to sign the contracts with the LSPs.

  While record-breaking size (£1 billion) contracts were signed in record-breaking time (in 1 year), it is clear in retrospect that time was lost, not saved. This is because the point is not how quickly one can sign a contract but rather how quickly one can implement the system and change NHS staff behaviour for the benefit of clinicians and patients.

  On this measure, NPfIT has already lost the race. Instead of taking around 2 years from the start of procurement to implementation (go-live), which is what it used to take the NHS. NPfIT has already taken over three years from the start of procurement (early 2003) and has delivered almost no core local Care Record Service functionality for acute hospitals, certainly none that is fully operational.

  This extensive delay is due to a blinkered view (on the part of both CfH and LSPs) of what implementation means in healthcare and an ignorance of past experience. There is a refusal to believe that (a) it takes time and effort to get clinicians to accept the need to redesign their practices prior to putting in new computers and (b) it takes time to make the new computer systems ready for use to support the new practice designs. These two points are widely known as clinical process redesign, or simply change management, for which there are no short cuts if one wants to implement clinical systems, not just patient administration systems, that are used by every clinician in the organisation.

  Such change management is a multi-year journey which begins with the first day of procuring a new system (actually even before, with the need for business cases to justify the procurement). Clinicians who are involved in the procurement from the beginning will then "own" the decision with respect to the chosen supplier and product. It is a psychological transformation which turns a group of clinicians in a health organisation (alongside their executive and IT colleagues) from the "sceptical buyer" state of mind to the "proud owner" state of mind. Successful implementation of clinical systems in the NHS have demonstrated that the "proud owner" state of mind is essential to keep the project on track when "turbulence" is encountered in the early days of clinical systems implementations (turbulence which is perfectly natural and expected in relation to early "growth pains").

  For instance, when the local clinicians complain that the pathology results are not coming through quickly enough, or in the form in which they were used to, or without reference values presented exactly as before, or in an unreliable fashion, which are all normal concerns during early days of going live with order communications, the Medical Director can step in to calm nerves. This is because the Medical Director will have been fully committed to the project, typically after two or so years of being involved with the project from procurement to implementation.

  Such involved, informed and committed Medical Directors, Nursing Directors and Chief Executives hardly exist in the NHS today because they have been kept at arms length by CfH. CfH and LSPs have generally adopted the technocratic and autocratic attitude of "They will get product when we say it is ready and they'll use it whether they like it or not, or else their CEO will get leaned on from above". This experience of getting "leaned on from above" has recently been reported to us confidentially by a number of NHS Chief Executives (who obviously do not want to be named).

  Disengagement and disillusionment of NHS managers in this way, due to the top down approach (typified by LSPs simply coming along and presenting implementation deadlines which invariably slip), is enormously damaging and corrosive for the successful implementation of clinical systems.

  With the NPfIT implementations (even the rudimentary PAS modules) that are due to begin in early 2007, when such problems occur, which are more likely now because of over-stretched software supplier staff, the Medical Directors/Chief Executives are not so likely to step in and give the assurance to fellow clinicians that the clinical safety aspects of the systems are in the hands of reliable people, simply because they will not have invested the time in understanding the product, the supplier and the other people involved with the procurement, selection, configuration and go-live management. Unfortunately, neither CfH nor the LSPs have taken this lesson on board and nor has the NAO Report.

  To exacerbate the NPfIT delays even further, the Spine message definitions were also delayed as reported in e-Health Insider[29] (Issue 30 January 2006, our underlining):

Delays—what delays

"At my hospital we were originally supposed to go live last February. That was postponed to July and then November. We are currently looking at this July 2006 at the earliest for go-live. This has been blamed on a combination of problems with a large deployment, messages to the spine not being available and issues with the multi-campus deployment. It is very depressing for hospital staff who were keen (but sceptical) at first, but who now don't believe there will ever be a new system."

  The LSPs need these spine messages in order to modify their Local CRS systems accordingly. This added further delay to the development and implementation of Local CRS systems.

  All of these problems were avoidable and have the consequence of (a) causing delay of systems to the NHS (b) increasing the CfH internal management costs because of their expenditures in legal resources and other consultancy fees during supplier swap-outs and (c) causing the supplier incomes to be delayed.

  The third consequence, which is delayed incomes for suppliers, will have a knock-on effect in that the suppliers will focus even more on recovering their delayed incomes and therefore focus upon trigger events for payment, rather than working closely with local clinicians to gain ownership and proper process redesign and product configuration to make their systems fit for purpose.

  See Diagram 1 below for a summary of these points.

Diagram 1. The real causes of NPfIT delay. In the healthcare sector, local clinician involvement & change management has always been intrinsically linked to IT procurement - ignorance of this fact has caused massive delays and will result in massive shortfall in NPfIT delivery in terms of value for Money.

  Professor John P Kotter, a specialist in organisational change at Harvard University, wrote in the Harvard Business Review in March-April 1995:

  "The change process goes through a series of phases that, in total, usually require a considerable length of time. Skipping steps creates only the illusion of speed and never produces a satisfying result. A second very general lesson is that critical mistakes in any of the phases can have a devastating impact, slowing momentum and negating hard-won gains."

V.   The NAO Report failed to address the question of accountability, particularly in respect of the high levels of public funding which underpin NPfIT.

  Taxpayer monies, originally appropriated in 2003 for IT systems on behalf of local NHS organisations, are now being spent by central CfH senior managers without any consultation with local NHS staff. The local Trusts do not know therefore exactly what they are getting and when and most importantly, what they are not getting but were expecting as part of the initial scoping of the LSP specifications (see the Eastern Cluster's schedule of 59 modules in Annex 3). As the NAO Report has indicated, no one is currently in a position to assess the performance of the National Programme except the National Programme themselves—this means they are only accountable to themselves.

  Chief Executives of local NHS Trusts are under close scrutiny for their performance, including public Board meetings, internal and external auditors and various levels of management line reporting. CfH management only appears to be accountable to itself. CfH controls some £12 billion, including £1.5 billion or so to run themselves and to pay for management consultants and lawyers over long periods of time, apparently with little or no scrutiny. NHS Trusts control far smaller budgets but with far greater levels of scrutiny.

VI.   The NAO Report failed to spot NPfIT's mistakes with respect to Local Service Providers (LSPs).

  NPfIT misjudged the role of the LSP. LSPs were brought in to serve as large companies with deep pockets to absorb risk. However, it is clear that the LSPs do not add any value to the delivery of clinical software because they do not have the prerequisite experience or skills. They do not understand clinical process redesign, the need to involve local clinicians in clinical systems configuration, the need to have staff with experience of NHS culture, processes, systems, numbering, reporting, etc. The LSPs tried to bring in some of these skills but failed to manage them effectively. In particular, they treated the electronic patient record software suppliers adversely, deploying a master-slave culture, rather than a collaborative one, where aggressive and dominant relationships were their means of hiding healthcare ignorance.

  CfH believes that contract management is more important than change management. The senior management at CfH spend more time with lawyers and talking to suppliers about contractual matters than in engaging with Chief Executives and clinicians within the NHS. CfH is not interested in addressing the real challenge of redesigning clinical processes, configuring clinical software and change management. The philosophy is: if something is wrong, bash a supplier and if it continues to go wrong, threaten to replace them. As CfH hold the purse-strings, the suppliers do not fight back.

  The LSPs will inevitably pass their ill-fortunes with NPfIT back on to the NHS either in the form of additional charges, poor delivery, and/or poor support. At least 3 NHS Trusts have reported recently that PACS costs for their Trusts are now significantly higher under NPfIT, through their LSP, than if they were able to buy PACS themselves. CfH has created precisely the type of relationships with IT suppliers that one always strives to avoid i.e. one where any request for software change or support service from local NHS staff is turned by the LSP into a formal change request requiring additional cost for the customer.

  Misconstrued views of the importance of the LSP role have led CfH to placate the remaining LSPs, who are under financial pressure due to lack of revenues, by giving them faster routes to revenue. This has meant short-cuts being taken by the LSPs in local implementations at the expense of the operational and financial well-being of local Trusts. Such short-cuts have resulted in PAS solutions that are not fit for purpose for local Trusts and in not taking account of variations in local processes for reporting and information management. This has threatened the ability of local trusts to manage tight budgets/financial deficits with proper operational and performance reporting.

  An example of a CfH engendered short-cut that favours the LSP but is detrimental to the Trust is the fact that Fujitsu, the Cerner LSP in the Southern Cluster, refuses to load historical patient data into the new PAS system, contrary to long-established industry practice, where patient historical data is invariably loaded by the clinical information system supplier because it is essential for treating patients properly. Not having historical administrative and clinical data in new electronic patient record systems is dangerous and analogous to building an aeroplane without life jackets under the seats—one day it will result in a safety disaster.

  For the acute Trust, this means that A&E and outpatient clinic staff have to do double work because they must keep their existing A&E and outpatient clinic administration systems going in "read-only" mode to look-up historical patient data, at the same time as using their new NPfIT A&E and outpatient clinic administration modules. As an example, it is important for records of past child attendances at A&E to be flagged if the child is deemed potentially at risk of child abuse. Multiple attendances at A&E with symptoms of bruising are one of the indicators of a child at risk. If A&E reception or triage staff do not have time to use two systems because their volume of patients is very high, they may find themselves forced not to look up historical data in the old system, and thereby risk important patient data, such as that for At-Risk children, being missed.

  The hospitals are left without any legal recourse because they do not hold the contract with the LSP, but instead the Secretary of State does—a legal mechanism that further disadvantages the NHS trusts because if the LSP does not deliver workable software, the NHS trust is powerless to do anything about it.

  In the early LSP contract days, the transfer of risk to LSPs led to a "moral hazard" where the appearance of LSPs bearing the risk of implementation removed from hospital Trusts the local sense of urgency and responsibility. Now, the risk is being transferred the other way to the local Trust, ironically enough, to the very people that risk transfer in the contract was meant to protect, the NHS.

  In this context, it seems that CfH is colluding with the LSPs to make sure that they get paid, but does not care about the damage that is caused to the Trusts either operationally, clinically, financially or in the form of lowered staff morale. CfH is now trying to legitimise this harm to the local Trusts, and washing its hands of any responsibility at the same time, by respraying a failure into a policy through making local trusts responsible for local implementation. But how can the local Trust Chief Executive be expected to implement the LSP product if it is unsuitable and if the contract does not give him/her any leverage over the LSP?

  CfH continues to engender a general climate of fear resulting from the bullying that is occurring on the ground. Chief Executives—particularly the vulnerable ones with financial deficits—are being leaned on by their bosses to implement the LSP products and to keep quiet, no matter how painful it is on the ground. This is the case even where the pain is specifically caused by the LSP's inadequate products and where this inadequacy has led directly to the loss of star-ratings (as in the case of the Nuffield Orthopaedic Centre), the inability to report activity and thus the loss of income, as well as having to pay higher prices than would be available on the open market (for example, with the Picture Archiving and Communications Systems).

  CfH is operating in a climate of aggression and hostility aimed at both the suppliers and the NHS where CfH has contracted for products which do not work, with LSPs that refuse to spend the money or expend the resources to adapt their products for local reporting, or to migrate important patient data. At the same time, Chief Executives are powerless to complain because of bullying from above, because they are not the holders of the contract and thus have no legal power over the LSP, and because the official policy is that local Chief Executives are responsible for the successful implementation of a product they did not choose in the first place and which anyway does not work.

VII.   The NAO Report failed to spot NPfIT's mistakes with respect to iSoft.

  iSoft was contracted as the Local CRS software provider to three clusters (ie North West & West Midlands; North Eastern and Eastern) on the basis of its proposal to develop a new product, Lorenzo. Our understanding is that a new product was deemed necessary because at the time of the procurement (circa 2003) NPfIT did not consider iSoft's existing range of products to be a sound starting point. The following is a summary of subsequent events.

—  Phase 1 of Lorenzo was due to be available to trusts in 2005. This was not achieved. iSoft now states, as reported by e-Health Insider on 11 December 2006, that Lorenzo will be delivered to CSC 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.

—  iSoft's accounting practices are currently being investigated, its share price has collapsed and it has yet to obtain the long term funding it requires to meet its commitments.

—  In lieu of Lorenzo, the LSPs have been offering to Trusts existing pre-NPfIT iSoft products known as iPM and iCM. The former is a PAS with no clinical functionality. The latter is an additional module which does include some clinical functionality eg order entry and results reporting, relevant primarily to acute trusts. A substantial number of mental health trusts and PCTs who for various reasons did not have a corporate PAS facility have implemented iPM. The general view of acute trusts has been that iPM is no better than their existing PAS products and hence there has been virtually no interest in iPM/iCM.

—  CSC/iSoft has promised to develop iCM to include for 2007 some elements of clinical functionality relevant to mental health trusts and PCTs, eg support for patient assessments and the development and monitoring of care plans (see Annex 1).

  These facts raise a number of questions which were not addressed by the NAO Report and which remain to be answered by CfH.

—  CfH has rightly insisted that non-delivery from suppliers will not be tolerated. And indeed the principle that the position of key suppliers is not sacrosanct has been well established with the withdrawal of Accenture as the LSP for the North Eastern and Eastern Clusters and with the withdrawal of one of the Local CRS software suppliers, GE/IDX, from the London Cluster—this latter despite successful delivery of products, including compliance to the Spine, in the two London Trust contracts that pre-dated NPfIT (at UCLH and at Chelsea & Westminster). How then does CfH explain the stance it has taken with iSoft given iSoft's failure to meet contracted timescales for software development; given that neither the future of the Company nor the successful development of Lorenzo is by any means assured; and that even if Lorenzo were to become a reality it would not be available to be deployed in many Trusts until 2010?

—  What steps are being taken to ensure that iSoft and the development of Lorenzo does not fail, and what are the implications of those steps in financial and other terms? Is there not a risk that iSoft's apparent commitment to develop iCM further as an interim solution will reduce the resources available for Lorenzo development? And what happens if Lorenzo development does indeed fail?

—  Given iSoft's statement that "our aim over the next few years is to exploit fully the existing portfolio of strategic products—especially iPM and iCM—prior to the gradual introduction of Lorenzo from 2008 onwards", why has there been no protest from CfH particularly when the contract was for Lorenzo and not the inferior legacy products? How is it that products rejected back in 2002-03 as an unsound starting point are now being accepted as strategic and suitable for further development? Why did the NAO not pick up on this given that iPM/iCM were already being implemented when the NAO's Report was being prepared?

—  In declaring iPM and iCM to be "strategic products", are CSC and iSoft in fact laying the foundation for abandoning the development of Lorenzo, the totally new product iSoft is contracted to deliver? We note with concern (see Annexes 5 and 7) that Trusts are already being encouraged to use the nomenclature "Lorenzo" to describe the legacy products iPM and iCM.

—  Even more serious is the question why CfH is sanctioning an approach whereby some three to four years into the programme (and rather longer for many NHS Trusts since their pre-NPfIT initiatives and procurements were halted) the NHS still has to wait a further three to four years before even the possibility of obtaining any substantial clinical software which works? Consider first the position of acute Trusts. Few want to move to iPM and so will not have iCM available to them. They will, therefore, not be able to obtain Local Care Record System functionality before 2009-10, with even this dependent on the successful development of Lorenzo. And those who are prepared to countenance iPM appear to have little confidence that timescales will be met (see Annex 7).

—  Now consider the position of mental health trusts and PCTs. Feedback from those who have implemented iPM appears to be good (see Annex 5) with staff involved recognising the benefits to patient administration and service management. How galling, therefore, and how wasteful of people's present positive attitude that there is now to be a standstill (barring the possibility of some limited iCM development) until substantial clinical functionality becomes available, if it ever does, with Lorenzo. Given the precedent of BT's approach in London for mental health trusts and PCTs (i.e. to adopt a completely different software product, CSE Servelec's RiO system, as its strategic solution) why is CfH not insisting that CSC widen its Local Care Record System solution portfolio beyond that of iSoft to include products that have already been developed and can offer clinical benefits now? Why has CfH sanctioned an approach which clearly is good for iSoft, enabling it to obtain revenue from off-the-shelf products in advance of Lorenzo, but which holds back mental health Trusts and PCTs and risks creating disaffection amongst clinicians at grass roots level?

—  What value is the public getting from supporting the central CfH administration team and local CfH outposts until 2010 when all that is being delivered for the Local Care Record Service are old iSoft systems?

—  There is an important governance principle at stake which CfH appears to ignore and upon which the NAO did not comment. CSC and Accenture were awarded their CfH contracts on the basis of developing a new product. Yet they will be obtaining revenue from CfH for many years from old iSoft products which were actually rejected during the initial procurement. Other suppliers with products at least as good as iPM/iCM might well feel aggrieved.

—  Finally, given that Lorenzo, if ever developed, would be deployed only towards the end of the contract period (2013), should the NAO not have questioned CSC's apparent expectation for its contract to be extended from 2014?

VIII.   The NAO Report failed to spot problems in other areas of NPfIT.

  The other main sections of NPfIT are also cause for concern.

(i)  Given that millions of ordinary homes now enjoy broadband access to the internet it is hard to be impressed by the often lower speed connections of N3 Broadband and the fact that CfH themselves describe N3 Broadband as an "insecure, hostile network".

(ii)  The claims for the Electronic Transmission of Prescriptions (or ETP) appear odd given that only 1.5% of the transactions stated (less than 30,000 out of 16 million) were actually paperless and that the remainder involve printing the prescription in much the same way as GPs' standalone systems currently do (as reported in E Health Insider, 23 June 2006).

(iii)  Choose & Book continues to miss its targets.

(iv)  Whatever the success of the PACS programme, reports from the field suggest that obtaining PACS via NPfIT is more expensive than directly from suppliers. One large hospital Trust has reported that its costs of procuring PACS from NPfIT are some £600,000—£1 million higher per year than if they were to buy it directly from the supplier. Given that these extra costs will be incurred for several years and given the number Trusts who are buying PACS through NPfIT, the potential additional costs for Trusts over the life of the contracts runs into hundreds of millions of pounds.

http://www.e-health-insider.com/news/item.cfm?ID=1670

4.  QUESTIONS THAT THE NAO REPORT FAILED TO ASK

  In view of the points raised above, several basic questions remain unanswered.

I.  Given that the LSPs are contracted to fully roll-out Phase 1 Release 1 by 2004, Phase 1 Release 2 by June 2005, and Phase 2 Release 1 by June 2006, according to the Eastern Cluster Business Case document, what has actually been rolled-out (ie implemented, not just made available for implementation) to date out of the 375 or so acute (and other community) hospitals in England? This is the real measure of programme performance, not just accepting what CfH claim has been deployed with respect to miscellaneous IT systems (most of which do not pertain to the LSP contracts) without any target figures.

II.  Given that Phases 2 and 3 are due for full roll-out in 2008 and 2010 respectively, but are dependent upon full roll-out of Phase 1, what modules are expected to be implemented by contract end, and therefore what is the expected actual total cost for each LSP contract?

III.  What is the Value for Money in terms of the expected cost of each Trust system at contract end compared to pre-NPfIT market prices?

IV.  To what extent are the LSP delivery delays due to delays in CfH managing its own obligations to the LSPs upon which the LSPs were dependent for completing their software modifications?

V.  Where is the scientific evidence for the Central Spine Summary Record's purported clinical benefits to justify Spine investment?

VI.  How much has been invested in the Spine already, given that it has never been piloted or tested (and that the Scottish experience with their Spine system reports very low clinician uptake)?

VII.  Which NHS CEOs and clinicians were consulted on the decision to down-scope the LSP contracts from effectively 3 phases to 1 phase (or at best, "Phase 1 Plus") of software delivery?

VIII.  Will the LSPs be paid 66% (or whatever the correct contract value is for the phases) less as a result of the down-scoping from 3 delivery phases to 1 or will they still be paid 100% of the original contract value in order that the LSPs can recoup their losses due to delay and also reduce their costs from delivering much less product to the NHS?

IX.  Which NHS CEOs and clinicians were consulted on the decision to re-define the NPfIT main objective as the delivery of the Central Spine Summary Record, and not the delivery of the Local Care Records Service to NHS Trusts?

X.  Is this redefining of the NPfIT main objective not just a PR ploy to try and manipulate the perceptions of the ministers, the NHS, the media and the public in order to lower expectations (i.e. a Central Spine Summary Record, even if partially populated and therefore partly functional, by 2010 rather than a deeply functional and truly integrated Local CRS System for each acute trust, PCT and care community)?

XI.  Has not the time saved on LSP procurement been more than used up by the subsequent NPfIT delivery delays?

XII.  Has not the lack of clinician involvement in the software (and LSP) selection process further delayed the programme?

5.  CONCLUSIONS

I.  The NAO Report has failed to spot poor CfH/NPfIT performance in the areas of business case delivery, benefits delivery, implementation progress, Spine delivery, the fundamental reasons for delays, accountability to the NHS, management of and by LSPs, iSoft delivery, and other NPfIT elements such as N3 Broadband, ETP (Electronic Transmission of Prescriptions), Choose & Book and Picture Archiving and Communications System (PACS).

II.  The NAO Report has missed the fact that CfH/NPfIT has lost its way. CfH/NPfIT has abandoned its vision to reduce NHS fragmentation of systems and care. It is driven by no vision at all except the desire to retain central authority using bullying tactics on both suppliers and Trusts. Now it seems it is also bullying patients by not allowing them to opt out of sharing their medical records (despite its earlier promise that they could opt-out). These drivers and tactics could almost be tolerable if CfH/NPfIT were delivering the goods, but it is clearly failing on all sensible and common sense measures: workable products (there is no workable new LSP product), on time (over 3 years late now and likely to be well over 5), within budget (the billions keep adding up).

III.  The CfH/NPfIT mantra is "deploy anything that moves just so we can say we are deploying something" with little care about the wider picture or indeed value for money. This pressure, along with lack of deep understanding of the link between clinical systems and clinical culture, has led to the misguided secondary mantra which is "oh don't worry about clinical systems and integration of the NHS, the Central Spine Summary Record system will take care of that".

IV.  Any clinical benefit which may eventually derive the Spine is likely to be miniscule compared to the evidence-based benefits that the NHS needs and which have been contracted for in the form of Local Care Record Service systems (otherwise known as Electronic Patient Record Systems) for each NHS acute, community and mental health Trust in England.

V.  Not one of the LSPs has delivered to contract. The NAO should not have praised the so-called "speedy" completion of a contract procurement process which was so clearly flawed. The extraordinarily hasty way in which the original LSP contracts were let, ignoring all other considerations such as the need to achieve buy-in from clinicians and from chief executives of local Trusts, is one of the main causes of the enormous problems which have followed in attempting local deployments. The NPfIT procurements can only regarded as failures.

VI.  GPs have in effect opted out, putting paid to the concept of "ruthless standardisation".

VII.  Our conclusions cause us therefore to take serious issue with the key conclusion of the NAO Report that "The Department and NHS Connecting for Health have made substantial progress with the Programme"[30]. All the evidence, particularly that relating to delivery of the originally envisaged core Care Records Service functionality, points the other way.

6.  RECOMMENDATIONS

I.   The Department of Health should take steps to:

(a)  Retain the sensible aspects of NPfIT, namely: (a) ring-fenced money for IT; and (b) discounted prices.

(b)  Stop the tactics of bullying NHS Trust Chief Executives into taking take unfit systems and stop the climate of secrecy surrounding CfH where people are afraid to speak out, even to share lessons learnt or to suggest how problems could be solved.

(c)  Empower the local Trust Chief Executives with final authority to decide when payments are made to LSPs.

(d)  Recast the contractual relationship between the LSPs and each NHS Trust so that each Trust is legally the customer rather than the Secretary of State for Health.

(e)  Restore the NPfIT vision of sharing information and functionality (ie wide area, cross organisation and cross care-setting Local CRS systems) particularly across local health economies or care communities (as is already stated in the LSP contracts).

(f)  Put the "Spine" investment on hold and ask clinicians and IT specialists to pilot and test it objectively and with scientific rigour.

(g)  Disband the central NPfIT team within CfH in favour of allowing systems procurements and implementations to be handled by local care communities in accordance with agreed standards. This level of management would support the integration of the different care settings such as primary and community care, acute hospital care, mental health care and social services, and therefore achieve the main benefits of using IT to deliver patient-centred care.

(h)  Make the money which is saved from reducing central overheads available locally for change management for each care community.

(i)  Establish an open framework of CfH accountability where future decisions on NPfIT budget and central IT contract changes are made in full consultation with, and with accountability to, patients, local NHS trust management and clinicians.

(j)  All NHS trusts in the West Midlands & North West, North Eastand Eastern Clusters should be asked whether they wish iSoft to remain the sole subcontractor for the local Care Record System, with appropriate action to be taken to reflect their answer.

II.   The NAO should:

(i)  Review its methods of investigating large IT programmes in the light of the points made in this submission e.g. measuring progress should be undertaken against baselines and appropriate metrics rather than merely accepting what the programme managers claim.

(ii)  Address the shortcomings of its report on NPfIT according the section 3. above.

(iii)  Ask the questions set out in section 4. above.

Annex 2—Information Sources Website Addresses

—  Eastern Cluster LSP Business Case (ATP2)

  http://www.portal.nscsha.nhs.uk/imt/Document%20Library/AtP%20Eastern%20cluster%20v0.20%20(no%20finance%20case).doc

—  London LSP Business Case (ATP2) Sign-Off

  http://www.royalfree.nhs.uk/DOC/appxe2.doc

—  Report on Serious Untoward Incident at Homerton and Newham 2004 using Cerner EPR

  http://www.newhampct.nhs.uk/docs/board/060117.pdf

—  Spine Message Delays

  http://www.e-health-insider.com/news/item.cfm?ID=1670

Annex 3—Modules to be Delivered by the LSPs

SUMMARY OF MODULE FUNCTIONS

101  User Environment/Tools

  The aim of this module is to identify the core toolset which is required from each LSP to control the user environment and the operation of the component functions of Local ICRS Service and local system tailoring which will be required to support the diversity of clinical and patient needs which the LSP Services will address.The following components are required:

—  tools to enable the collection of locally-defined information

—  tools to support the reporting and analysis of information

—  parameter controls to support the behaviour of the solution to meet needs within a Cluster and the diversity of clinical practice

—  controls to support information governance, including but not limited to access and confidentiality

—  controls to support the operation and activation of workflow as an inherent function.

102—Patient Index

  Typically, each Trust and GP practice currently holds its own patient index. The new service shall provide a method to ensure that all systems and services for which the LSP is responsible use the national PDS as their unique source of patient demographic information although local Systems will need to hold temporary supplementary information.

103—Prevention, screening and surveillance

  The purpose of this module is to support national and local promotion, prevention, screening and surveillance programmes. The service requirements specified in this module should enable:

—  the comprehensive identification of persons at risk of developing particular problems, to enable implementation of systematic prevention programmes, which will lead to a reduction in the incidence of these problems or conditions

—  improvements in the coverage of screening programmes in order to detect the incidence of problems and conditions at an early stage, which will lead to improvements in the subsequent outcomes of care.

104—Assessment

  This module specifies the requirements for the service to support all clinical assessments across all care settings, incorporating the specific requirements of the Single Assessment Process (SAP)—with links to social and other care agencies. ICRS seeks to address, specifically, the generic functional requirements that arise from the distributed nature of care services and the need to deliver patient -centred, integrated, evidence-based care. NSFs are driving the latter and require the development of increasingly shared (integrated) assessments and care planning within and between services.

  Scope

—  Support for the documentation of uni- and multi-disciplinary assessments in a structured form eg templates linked to the Patient Record

—  Support for integration of structured assessments within care plans and care pathways, as required

—  Support for access to assessment documentation

—  Support for access to clinical knowledge eg guidelines and protocols to be available during assessment.

  Components

—  Structured assessment templates for specific types of clinical assessment, with the ability to create and amend templates (including an Audit Trail for amendments)

—  Common data dictionary of data items to enable sharing of common information across each assessment template

—  Incorporation of different types of clinical assessment and evaluation information, including graphical, textual, numeric, audio and video data.

105—Integrated Care Pathways and Care Planning

  An Integrated Care Pathway or ICP describes a process within health and Social Care, which maps out a pre-defined set of activities and records care delivered and the variations between planned and actual care. ICPs will be used to support "whole systems" processes spanning Primary Care and Secondary Care service boundaries. ICPs are largely based on conditions or diagnoses. The development of ICPs is a complex process and, from a clinical perspective, will take time to develop. The current reliance on paper-based care pathways has made the task of defining pathways of care more difficult in terms of design and application as a real-time tool to assist in delivery of evidence-based care. The incorporation of ICPs within the LSP Services will enable ICPs to become an active tool to assist in the delivery of care incorporating clinical decision support to identify actions, reminders and guidance at the point of care, across the continuum of care.

106—Clinical Documentation, including Clinical Noting and Clinical Correspondence

  The purpose of this module is to provide functionality which will support the recording of structured and, where necessary, unstructured clinical notes, summaries and letters. Key elements of clinical documentation covered by these requirements include:

—  Clinical notes, including, but not limited to: operation notes; medical history, where this is not part of a structured assessment; treatment notes; and documented observations; referral letters; out-patient clinic letters

—  Alerts from systems in one sector to another that a patient is undertaking an Episode of care (eg GP system alerted to patient's admission to A&E)

—  Immediate discharge summaries, including ToTakeAways (TTA)

—  Final discharge summaries

—  Copies of the above to patients, either in electronic or printed format.

107—Care managementThis module specifies the requirements for the management of specific types of care event, independent of care setting. Requirements are provided for unscheduled care management, domiciliary care management, ambulatory care management, bed management and demand or access management. With the shift from acute-based care to community and primary -based care, the traditional approach to provision of out-patient and in-patient management wholly within a hospital environment is no longer valid.

108—Scheduling

  Scheduling will often involve the scheduling of resources from more than one organisation, across a range of care delivery environments. Effective scheduling will also require the adequate capture of demand data and prediction of future capacity use. The scheduling solution proffered must be flexible enough to accommodate existing and emerging working practices, across the whole range of health care delivery environments and locations.

  Components

  The scheduling functionality must provide a process in which events that need to occur in order to deliver patient care are assigned a date, time and place in the future when the resources that are required to carry out those events are available. It is to be available at three levels:

—  Departmental scheduling (intra-departmental)

—  Enterprise-wide scheduling (inter-departmental and intra-organisational)

—  Scheduling across a healthcare community (inter-organisational).

109—eBooking compliance

  To support eBooking, the following components are required:

—  Changes to Primary Care systems to make them support eBooking

—  Changes to Secondary Care systems to make them support eBooking.

  Note that the LSPs, and not the e-Booking service provider, are responsible for connecting all GP and local systems, and have responsibility for data migration.

110—Requesting and order communications

  Orders are used to request services or goods, and may result in results being reported back. Orders may be fulfilled by electronic systems, manually, or by a combination of both. In order to deliver an order, whether it is styled as an order or request, the initiator may also need to take a sample or schedule a procedure. Requests can be placed for diagnostic and investigative services. This is not just for pathology/radiology tests, but also for other diagnostic services (eg, audiology, cardiology, endoscopy, pulmonary function and neurophysiology) and for other goods and services.The following components are required:

—  Order definition

—  Order creation

—  Order routing

—  Sample collection

—  Order receipt

—  Order enquiries

—  Order management.

111—Results reporting

  This module specifies the requirement for results reports to be made available to the requestor and/or other authorised persons, including the patient. Results are generated by diagnostic and investigative services. They are generally provided in response to orders/requests. Results may be generated by electronic systems or be provided manually, or by a combination of both. Results can be provided by a range of departments; not just by pathology/radiology departments, but also by other services (eg, audiology, cardiology, pulmonary function and neurophysiology).The following components are required:

—  Results definition

—  Results generation

—  Results reporting

—  Reports and alerts routing.

112—Decision support

  ICRS will enable Clinicians to make decisions based on the best-available patient information and currently-accepted evidence of best practice. ICRS will also provide managers with quality summarised data for service planning. The following components are required:

—  Elective—structured access to reference material

—  Passive—implementation of local protocols

—  Active—alerts

—  Clinical management—management and maintenance of protocols

—  Service development—forward planning of clinical services.

113—Prescribing and pharmacy

  Prescribing and administering drugs to patients is a key care process. Both processes, if inadequately informed, can also cause serious risks to patient safety. This module describes the core functionality required to allow and support the safe prescribing of drugs by Clinicians, as well as assisting in managing the dispensing and administration of drugs (mainly in the hospital setting), and monitoring and presenting each patient's drug history and compliance. The scope of this module includes all prescribing and drug use across the NHS: in Primary Care; in the Acute Care sector; and by community practitioners, as well as provision of drugs in the community. The following components are required:

—  Reviewing medication history prescribing

—  Prescribing

—  Repeat prescriptions

—  Decision support

—  Dispensing and administration.

115—Digital imaging including specification for a picture archiving and communications system (PACS) solution

  This module specifies the requirements for the service provider to enable management and distribution of digital images used for clinical purposes. An integrated care record shall include a wide range of non-textual information; eg, graphs, scans, etc. An important element within the record shall be digital images. This section describes the requirements for collection, management and presentation of digital images, functionality commonly referred to as a Picture Archiving and Communications System (PACS). The following components are required:

—  Initially restricted to static radiological images and associated reports, but rapidly expanding to cover other disciplines, including dermatology, orthopaedic surgery, endoscopy and cardiology

—  Needs to link to existing (or replacement) radiology management systems demographic and administrative/scheduling functions (eg Patient Administration Systems/functions) to allow for pre-fetching and auto-routing

—  Supports transfer of individual images/reports (or the images/reports for an individual) between health communities—generally through a remote viewing process, rather than actual transfer.

116—Document Management

  Today a large proportion of care records are held on paper. ICRS will increasingly reduce the amount of paper needed. In some settings, at least, the need for paper could be eliminated if documents could be managed by capturing them electronically and making them part of the Patient Record. For this module the following components are required:

—  Document creation and capture

—  Indexing and profiling

—  File system services/storage

—  Document viewing, annotation and editing

—  Tracking of paper-based documents and X-ray films.

Other components

  There are a number of other components which are required to be delivered by the LSPs. These components do not actually provide additional functionality over and above that described already. For instance Maternity services (Module 118) will require "patient index", "scheduling", "results reporting" etc. The full list of other components is as follows:

114—Diagnostic and Investigative Services

117—Financial Payments to Service Providers

118—Maternity

119—Social Care

120—Dental Services

121—Maintain Patient Details

122—Emergency\Unscheduled Care

123—eHealth and Clinical Development

124—Information to support secondary analysis and reporting

125—Surgical Interventions

151—Primary and Community Care

153—Acute Services

154—Ambulance Services

160—National Service Frameworks

161—Mental Health

162—Diabetes

163—Cancer

164—Coronary Heart Disease

165—Older people

166—Children's Services

167—Renal services

  The functionality within these components will be delivered in accordance with the timescales described above.

  1.  The following text describes the roll out of functionality by phase.

Phase 1 Release 1—Available June 2004—Complete by December 2004

  2.  The principle purpose of this phase is to install systems, hardware and software that will form the framework on which future functionality will be built.

  3.  Nationally we will see the establishment of a Personal Demographics Service (PDS) that will store demographic data for every individual in England in one central data store accessible through local systems. In parallel the Personal Spine Information Services (Spine) will be established. This will form the basis of the lifelong health record for every patient in England. In order for these two services to work effectively and to be accessible from the start each person registered will be assigned a single unique identifying number (NHS Number) that will form the common link between local and national applications.

  4.  There will also be a secure Access Control Service (ACS) which will register and authenticate users, (including patients), provide a single sign on and record the consent of the patient to clinicians accessing their personal health record.

  5.  Whilst these three key services are put in place, additional work will be undertaken to set up an infrastructure to handle and process the data that will flow between the various local and national systems. These data "messages" will be processed through a Transaction Messaging System (TMS). Each message will contain tagged data in XML format; the tags indicating the route the message should take through the system—perhaps from a GP to a pharmacy in the case of electronic prescribing or one PCT to another in the case of GP: GP transfers. The key message processing functionality that will be built in this phase is:

—  Initial e-booking

—  Prescribing

—  Basic Patient Information

—  Birth and death notification

—  Allergies.

  6.  ICRS is thus providing the enabling technology for two key initiatives, eBooking and Electronic Transfer of Prescriptions, as well as starting to share basic patient information across the NHS clinical community; given some basic facts it will be possible to clearly identify the patient and get a snap shot of medical history, reducing the scope for errors and improving the service to patients.

  7.  There will clearly be a need for user support and a helpdesk will be available from the start of the phase. This will expand in size and scope through each subsequent phase.

Phase 1 Release 2—Available December 2004—Complete by June 2005

  8.  The next release of phase 1 is concerned with building on this infrastructure to start managing more complex business processes and handle messages from different clinical situations. This will deliver;

—  Full e-booking functionality

—  Outpatient clinic letters

—  Inpatient discharge summaries

—  Report of the Single Assessment for elderly people

—  Diagnostic imaging and pathology result (flag and locator)

—  Screening results

—  Recording of care episode events

—  Routing of orders for some blood tests and diagnostic images.

  9.  These new services will start to change the way people work within the health service. There will be a gradual migration towards electronic working with easier, faster access to patient demographic details and medical history, with routine use of electronic indexing systems.

  10.  GP's will routinely be using electronic prescribing tools with the roll out of electronic prescription transfer to the community pharmacy, reducing the potential for error and fraud as well as enhancing the patient experience.

  11.  Maternity units will start to use systems that automatically record birth details and demographics and link the baby's records to the mother's.

  12.  All healthcare staff will have access to a wide range of digital libraries and support systems to assist in diagnosis and treatment.

  13.  Clinicians will routinely capture clinical notes electronically, saving time and making key health information available to others directly involved in delivering care.

Phase 2 Release 1—Available June 2005—Complete by December 2006

  14.  This phase sees the consolidation of the process and the enhancement of functionality to deliver;

—  NSF assessment and review record

—  Secondary uses of spine data*  Planning and recording of the total care journey—integrated care pathways

—  Full linking and electronic transfer of correspondence

—  Pathology and image order and result (HL7 messaging)

—  Integration of dental services.

  15.  By this time, electronic support for business processes will be commonplace within the NHS;

  16.  Clinical teams will be able to create and share assessments for specific National Service Frameworks such as Cancer and Diabetes, leading to faster more relevant treatment

  17.  Secondary users such as medical researchers will have access to a growing pool of pseudonymised data about health events and outcomes, improving both the quality and speed of research

  18.  Clinicians and others involved in delivering care will be able to maintain patient-specific care pathways, tailoring care to reflect both the patients needs, and any specialist skills available

  19.   A significant proportion of ordering of services such as Pathology and Diagnostic Imaging, and the subsequent distribution of results will be electronic, leading to faster and more predictable reporting.

  20.  Dental practitioners will have access to the patient health records through the Spine, with the benefit of added knowledge about a patient's medical history facilitating more effective care.

Phase 2 Release 2—Available June 2006—Complete by December 2008

  21.  At this time there will be greater sophistication in the technology available with an increased level of integration and seamless care in three key areas;

—  More sophisticated Access Control Services

—  Extensions of the eBooking Service

—  Links to remote care settings.

  22.  Patients will be able to place elements of their medical history in a virtual "sealed envelope", allowing them to more closely control access to their data.

  23.  Clinicians too will be able to seal particular elements of data, protecting patients from information which might be harmful, and there will be audited "break glass" functionality enabling them to gain access even to "sealed" data in emergencies.

  24.  Health care professionals and patients will be able to extend the eBooking concept to a wider range of health encounters both inside and outside the NHS, giving greater choice, control and certainty to patients.

  25.  Health care delivered through remote settings such as ambulances, NHSDirect, walk-in centres or non-NHS facilities will be integrated with ICRS.

  26.  Tertiary care centres will routinely be linked to each other, secondary care and to primary care.

Phase 3—June 2010

  27.  Significant changes will take place over this phase, with further enhancement of systems and processes to ensure seamless care. The scope of this phase is deliberately open; clinical priorities will change, leading to new processes and in turn new functionality and messages. The trends, however, are clear:

—  The majority of clinical events and episodes will be recorded electronically; the majority of clinical processes will be supported by IT, reducing costs and providing a faster, more effective patient journey

—  Clinicians will have widespread access to information and integrated decision support at the point and time of need, with links into the knowledge base, improving decision making and hence clinical outcomes

—  This information will enable the development of individually tailored care pathways with linked tracking and exception reporting and dynamic monitoring of progress; Workflow will trigger alerts to clinicians both when planned events fail to take place, and when expected outcomes are not achieved.

  28.  Systems will be "self aware" with sophisticated error and pathway deviation monitoring enabling consistent care and a consequent reduction in mistakes due to human error.

  29.  However, it must be noted that these are the minimum levels of functionality that must be achieved by the given dates. They do not preclude procurement and rollout at a faster rate, subject to local priorities and affordability.

  30.  Implementation of ICRS is expected to take place via the procurement and implementation of "undles"of ICRS functions, covering both the core services (within the scope of this AtP2) and additional services (the case for which will be set out in AtP2 supplementaries). The core service "undles"are described in Appendix E1 (Annex 4 Core Service Bundles), in relation to the ICRS functions they comprise and the implementation phasing to which the bundles relate.

  31.  Eastern will aim to push the pace of implementation across many locations simultaneously. An urgent driver is the age of existing systems and status of some of the support contracts. The more important driver is that Eastern is seeking to gain benefits from a common ICT infrastructure to drive service redesign benefits as early as possible, where affordability permits.

Annex 4—Economic Case of Eastern Cluster ATP

(Extract showing the Importance of Cross-Organisation and Cross-Setting Benefits)

1—Benefits-Should Cost v LSP

			Individual Trust procurements ("should cost")		Cluster ICRS procurement
Overall aim + subsidiary "SMART" objectives	Notable ICRS phase 1 & 2 functionality required locally	Relative weight	Raw score	Weighted score	Raw score	Weighted score	Commentary
1.  Improve the experience of patients at all stages of care from the initial contact, through referrals, to scheduled treatment and back home as described in the NHS plan.
By 2005, patients should expect that an NHS organisation providing care already have any existing demographic information it needs for care to hand. They should only be asked to confirm personal demographics (name address etc). Any changes that are made to this record should automatically update the record for any future episodes of care.	Integration across cluster organisations, and with PDS, Spine.	9	1	9	3	27	Cross boundary issues likely with "should cost".
From 2005, patients will be able to securely access their own records through a range of channels.	Many components plus integration across cluster organisations.	5	1	5	2	10	More difficult with "should cost" to provide access to comprehensive records.
From 2006, every patient requiring an elective procedure will be given a choice of provider, whenever it is practical, depending on the patient's condition and treatment required, in accordance with National policy.	Integration with e-bookings.	7	1	7	3	21	High risk with "Should cost" of lack of integration between primary and secondary care and within each care setting.
From 2005, every patient requiring an outpatient clinic or day case appointment will be provided with a confirmed date, time and place, at the time of referral or through the booking management service, in accordance with National policy.	Integration with e-bookings.	7	1	7	3	21	High risk with "Should cost" of lack of integration between primary and secondary care and within each care setting.
From 2007 all test results (Pathology and Radiology as a minimum) should be held electronically (Including those tests that are performed at remote locations from where the patient's treatment is performed).	Order entry, results reporting.	9	2	18	2	18	Risk will remain in place for both options as it will rely on total integration.
From 2007 all order for all tests should be recorded electronically. Including those tests that are performed at remote locations to where the patient's treatment is performed.	Clinical correspondence, assessments, scheduling, results reporting, PACS.	8	2	16	3	24	Higher chance of information falling between the cracks with "should cost".
From 2007, patients who need to make a series of visits for investigations and treatments, will be able to agree and book a mutually convenient schedule of contacts, based around the clinic, theatre schedule of the relevant clinicians.	Scheduling, E-booking.	7	2	14	3	21	Risk with "Should cost" of lack of integration between primary and secondary care and within each care setting.
From 2007, clinical interventions should never be postponed to another date because the information required for treatment is unavailable in electronic format.	Scheduling, Results Reporting, Order entry.	6	2	12	3	18	Medium risk for "should cost" of systems not all being integrated.
2.  Improve the quality of care by enabling standards to be implemented across the NHS and Clinical and Social Care networks.
By 2006, patient level information will routinely be available to support clinical governance relating to NSF's which span care settings.	Decision support as part of ICPs, order entry, Clinical correspondence etc.	9	1	9	2	18	Risk with "should cost" of systems not all being integrated.
By 2007, ICRS will have had a significant impact in reducing the impact of clinical error. This will arise from the improved quality of records and access to them, legible clinician orders and use of defined order sets, together with the increasing use of clinical decision support.	Clinical correspondence, Order Entry, Decision Support, results reporting.	10	2	20	3	30	Some risk with "should cost" of information not being available regarding previous tests conducted at another organisation.
By 2010, ICRS will have reduced the number of avoidable adverse incidents arising from medications management.	e-prescribing.	9	2	18	3	27	In "should cost" some risk of not picking up interactions with drugs recorded on other organisations' systems.
By 2006, investigations will not need to be repeated because information relating to earlier tests is not available in electronic format. (Pathology and Radiology as a minimum).	Results reporting, clinical correspondence.	9	2	18	3	27	Some risk with "should cost" of information not being available regarding previous tests conducted at another organisation.
By 2005, provide all clinical staff will have the infrastructure made available to them to quickly and easily access the latest evidence and best practice.	Access to electronic reference material.	9	3	27	3	27	Both options will provide this functionality.
By 2007, patient's medical records from their registered GP practice will be available to clinicians at an acute or community hospital where routine medical treatment is provided.	Many components plus integration across cluster organisations.	8	0	0	2	16	High risk for "Should cost" due to legacy systems and lack of spine record.
3.  Enable effective access to clinical and administrative information across care providers and locations to support NHS clinical priorities.
By 2007, a summary of a patient's recent medical history from acute or community hospitals will be available to clinicians at within the community.	Integration across cluster organisations + with PDS, Spine.	8	1	8	3	24	More difficult with 'should cost' to provide access to comprehensive records.
4.  Reduce the fragmentation of care through improved consistency and coherence of systems and records.
By 2005, provide clinical staff with fast and convenient access to the summary of previously electronically recorded healthcare interventions delivered (regardless of by whom) for any one patient and to the detail about patients in their care and which are held on the National Spine.	Integration & messaging across cluster organisations + with PDS, Spine.	5	1	5	3	15	More difficult with "should cost" to provide seamless information exchange during referrals, discharges & transfers.
By 2010, General Practitioners and others actively providing care (eg within a clinical network) will be able to access and maintain a consistent and detailed patient record irrespective of care location.	ICPs within and between organisations	8	1	8	3	24	More difficult with "should cost" to provide integrated pathways across different organisations.
By 2008, ICRS should contribute to a 25% reduction in nurse's administrative workload.	Many components plus integration across cluster organisations.	6	2	12	3	18	More risk with "should cost" due to lack of full integration.
By 2008, ICRS should contribute to a 20% reduction in doctor's administrative workload.	Many components plus integration across cluster organisations.	6	2	12	3	18	More risk with "should cost" due to lack of full integration.
By 2006, significantly reduce the time spent obtaining missing/lost patient information during referrals, discharges and transfers.	Clinical correspondence, integration across communities.	5	1	5	2	10	More difficult with 'should cost' to pull together information from different organisations.
5.  Improve health policy development and health research through increased availability, improved quality and speed of retrieval of data.
By 2010, facilitate the operation of multi-disciplinary and multi-organisational integrated care pathways.	Facilities for the abstraction, management and reporting of information.	3	2	6	3	9	More difficult with "should cost" to pull together information from different organisations.
By 2008, provide up to date, accurate and fully anonymised patient data for policy development where the source data is less than 3 months old at time of making it available.	Facilities for the abstraction, management and reporting of information.	5	2	10	3	15	More difficult with "should cost" to pull together information from different organisations.
Total			246		438

 

 

Key to relative weighting	Key to raw score
1 = least important through To 10 = most important	0  Does not deliver goal at all 1  Low delivery of goal 2  Medium delivery of goal 3  Maximum delivery of goal

Annex 5—Extract from a Recent PCT Report

  Caveat. The word "Lorenzo" in this extract, which is taken verbatim from a PCT report, is no longer being used strictly to refer to the new product that iSoft is supposed to be developing but instead is being used to rebadge the old ISoft products, iPM and iCM, thereby creating the impression that Lorenzo is in fact a reality. The author of the extract is aware of this but explains that he wished to avoid confusing his readership and so employed what increasingly is becoming common terminology.

EXTRACT FROM A RECENT UPDATE REPORT ON NPFIT FROM A PCT TO THE LOCAL PEC[31] IN A CLUSTER SERVED BY CSC/ISOFT

Reference Solution—Lorenzo

  There has been a considerable amount of activity over the last 15 months to implement the NPfIT Reference Solution, usually referred to as the Lorenzo system. The roll out of Phase 1 of the system is now complete, and we have recently completed a joint exercise between the local NHS and the NPfIT Local Service Provider to capture the Lessons Learned from that project.

  There have been some excellent examples of PCT staff using this system to fully support patient care, and the majority of services are now using the system. There has been some reluctance to use the system in a few services, and the project team has been very active to provide support before and during the go live period.

Future Outlook

  Over the next 3 months a major exercise will be undertaken to plan the implementation of the next phase of Lorenzo, iClinical Manager, which will be available for deployment from May 2007. This will provide significant clinical functionality for the PCT's clinical services, such as the ability to support patient assessments, and the development and monitoring of care plans.

  Further major releases are planned as Lorenzo Enterprise 3.5 (including e-prescribing) available from February 2009, and Lorenzo Enterprise 4.0 (including support for Integrated Care Pathways) available from May 2010.

Annex 6—Contribution from Bernard Hunter

  Bernard Hunter is a health IT consultant with over 15 years experience of working for the NHS.

  "Some thoughts—largely from an acute trust perspective.

   (a)  I always thought the whole thing was TOO big—human beings and human systems can only cope with so much.

   (b)  the surprising thing is that it appears to be the software that has been letting the project down; for some reason, the private sector seems not to be very good at developing big healthcare systems—much of the legacy stuff dates back to NHS-development days (notwithstanding some recent front-end sexing up); the database design (data model) is specified, the requirements have been known and documented for decades, so—yes—that has been a surprise.

   (c)  having said that, in London, things really seemed to be coming together with BT and IDX, until GE lost interest (assuming that is what happened)

   (d)  where I thought it would fall down, and this remains untested, is in local implementations.

—  are there enough good people to manage local implementations?

—  will local data quality problems—horrible to resolve, but must be resolved before migration—kill the whole thing?

   (e) the talk now is of local involvement/control, and the massive gains are being ignored in all the statements (in favour of "why should we move from existing if what is on offer is no better")—sounds like a death-knell to me; I can see the baby following the bathwater!

   (f)  My guess re the future: they will keep the centralised services and design, but go back to 1999 or so, to the precursor of NPfIT (was it called "Information for health"?), requiring trusts to implement system according to given standards, sharing data through the spine in that way. I just hope those contracts are as good as Grainger says, and the NHS does not have to pick up the tab.

   (g)  to some extent, implementation at a particular trust will take as long as it takes, so shoe-horning into a regional timetable is not helpful—another reason for local control

   (h)  the lack of experienced people could partly to be addressed by having region-wide shared implementation services

Bernard"

Annex 7—Comments from an Acute Trust Director of IM&T

  Note that the author of these comments, inserted below verbatim, has enclosed "Lorenzo" in parentheses to indicate his awareness that the word is no longer being used strictly to refer to the new product that iSoft is supposed to be developing but instead is being used to rebadge the old ISoft products, iPM and iCM, thereby creating the impression that Lorenzo is in fact a reality.

  Our first major implementation is for the replacement PAS and we are attempting to agree our `go live' date. This has become problematic because the date depends on training being able to be delivered in a three month window where there can be no slippage. Commencement of the training is dependent upon CSC delivering LE2.2 of Lorenzo for testing by the NHS, testing and resolution of issues identified and agreement to deploy on a specific date. There is not a lot of confidence that all these milestones will be delivered on the dates due. This is probably down to previously poor CSC delivery of releases and version and the NHS capacity to respond to issues effectively.

  There is evidence that CSC performance is improving and that the "Lorenzo" product is fit for purpose. This needs to be put into context however, in that a fully integrated solution will not be available until, I believe, 2010 at the earliest.

  PACS, I think, is going to be an NPfIT success but not without hard work.

  Choose and Book application is working but performance is poor. My detailed knowledge of the Choose and Book application is not great but there have been some problems with adequacy of functionality and the Trust has put hundreds of man hours into "workarounds".

  Summary spine record.... not available, although we have had some discussion with a local PCT and CfH in respect of the piloting of a GP summary record.

  A final point may be that agreeing contractual changes is difficult in the extreme and the logistics of involving the NHS, CfH, LSP and often the 3rd party provider are somewhat horrendous. It took approximately 4 months to agree a CCN to deliver PACS in advance of the main deployment to one of our sites.

---

---

23   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

24   NAO Report, Summary, Conclusions & Recommendations 7. Back

25   As described on page 5 in Part A above and described in detail in Annex 1 below. Back

26   See the Eastern Cluster LSP Business Case at Annex 3 and 4. http://www.portal.nscsha.nhs.uk/imt/Document%20Library/AtP%20Eastern%20cluster%20v0.20%20(no%20finance%20case).doc Back

27   Accenture has since pulled out of its LSP contracts. Back

28   This evidence is provided in the following publication: "Improving Safety With Information Technology", Bates and Gawande, New England Journal of Medicine 2003; 348(25);2526-2534. Back

29   This publicly available information is accessible at the following website address: Back

30   NAO Report, Summary, Conclusions & Recommendations 7. Back

31   Professional Executive Committee, comprising, amongst others, local GPs Back

---

Prepared 17 April 2007