A National Clinical Strategy for Scotland

4. Secondary and Tertiary Care

Secondary and tertiary care services have seen a very marked change over the last 10 years. There have been increases in both elective and un-scheduled admissions to secondary care, a significant increase in the care provided as a day-case, steady rises in out-patient referrals, increased accident and emergency attendances and the development of an enhanced range of services, both secondary and tertiary. At the same time, considerable work has been done through the Scottish Patient Safety and Healthcare Acquired Infection Programmes to ensure that our hospitals are amongst the world leaders in improving safety for patients and protecting them from harm.

Scotland has had success in managing increased demand, improving quality and safety, reducing wasteful variation and producing good clinical outcomes. However in order to respond to, and effectively manage, increasing demand for secondary and tertiary care and ensure sustainability of provision, NHSScotland will need to make radical changes. There is evidence to suggest that a radical approach is required to plan services differently^[11] in order to be able to continue to improve the quality and outcomes from hospital services.

There are always pressures to cling to the status quo believing, against the evidence, that current service configuration offers the best possible service provision. However, history has clearly shown that in order to achieve improvements in the quality and sustainability of care, changes in the provision of medical care are required. If change had not been progressed previously we would still have a health service that provided general practitioner emergency surgery in small community hospitals - a service that no-one would advocate now.

This section outlines significant proposals for change, and describes the evidence that we must make rapid progress on these changes if we are to maintain high quality services for the population of Scotland.

It is important to remember the challenges that are prompting change in acute hospital services:

• the potential to significantly improve outcomes
• new technology dictating the need to have centres of excellence for more complex interventions (eg. robotic assisted surgery)
• Increasing volumes of elective procedures such as cataracts and joint replacements
• pressures in recruiting highly skilled staff
• increasing demand, and
• financial considerations, and the need to invest substantially more in a locally delivered primary and community health service in response to the ageing population and the prevalence of long-term conditions. The evidence shows that a strong primary/community service achieves improved outcomes and helps to address health inequalities

In this context there are two very significant change programmes that need to occur within secondary care in order to maintain the quality of service that the people of Scotland expect. These programmes relate firstly to changes in process within acute hospital care, and secondly to the structure of acute care.

Changes in process

We have to ensure that hospitals deliver the best possible value to patients. This is not value in narrow financial terms, but is the benefit that is delivered in terms of outcomes, and how these are delivered efficiently for patients: producing outcomes that matter to them, in a way that is as safe as is possible, and minimises the disruption to their lives.

Delivery of care through reliable, safe services has been shown to promote both quality, and cost-effectiveness. It can be a way of driving out waste and variation in services, producing better services at lower cost.

There is a need for continued work on process within acute care, aiming to improve quality for patients, reduce bed usage where possible by finding alternatives to admission, and by aiding early discharge - to make sure care is more effective for patients and is delivered efficiently.

This has been shown for example in the recent work on "ERAS" - enhanced recovery after surgery. By adopting a comprehensive approach to hip replacement for example, it has been possible to produce better outcomes by pre-operative management, modern anaesthetics ensuring effective post-op pain control and early mobilisation. The impact has been that the average length of stay has reduced from ten days to less than three days: patients mobilise more quickly, with less side-effects and with better longer-term outcomes. This work is now being extended to a number of other operations, and it is likely that effective management of such standardised procedures will continue to provide benefits for both patients and costs. In order to improve efficiency and effectiveness, we will need to have a concerted programme of change across all hospitals.

A considerable amount of this is already happening, supported by improvement teams from Healthcare Improvement Scotland, and Scottish Government. We must continue to examine processes in hospitals, and work effectively to implement change where evidence of benefit exists. There are already some excellent examples of hospitals taking forward initiatives including:

Outpatients: many reviews of outpatients can be dealt with by letter, email or telephone instead of clinic appointments. Where there is a need for patient-clinician interaction we should consider, especially for rural patients, the use of tele-consultations using effective video-linking. Patient surveys show that changes of this kind are acceptable to the majority, and may significantly reduce both the burden on the patient and the work needed in hospital.

Delayed Discharge: It is recognised that tackling delayed discharge is a priority for NHSScotland. Although the situation is improving, there were over 48,000 occupied bed days used in September 2015 by patients who were clinically ready for discharge (a reduction of 4,482 on September 2014 ). Achieving significant reductions will take a determined whole-systems approach and will significantly improve the ability of hospitals to cope with winter surges of activity and year round optimum patient flow and care. This is a priority for the NHS, local authorities and Integration Joint Boards to ensure the best outcomes for people - remembering that 75% of the patients delayed in their discharge are over 75, and in older patients, prolonged stays in hospital increase the potential for loss of mobility and independence. The Scottish Government is providing a range of practical support, guidance and toolkits to help local partnerships reduce the level of patients delayed in their hospitals.

Many areas are making great progress. NHS Greater Glasgow & Clyde, for example, are implementing a policy of discharge to assess, and have invested heavily in step-down intermediate care beds. This has seen a reduction in the number of over 75 patients delayed over three days, and associated bed days, of over 70% since November last year.

Developing IT further: There are huge advantages that can be obtained by having electronic clinical notes, electronic clinical decision support, and electronic prescribing and administration systems. Such systems can improve safety, reduce wasted patient and clinician time, reduce the costs of medical records departments, and, with appropriate safeguards in place, allow for treatment across hospitals, professions and regions. For patients, online access to electronic data and services such as appointment booking would allow joint decision making and improve their satisfaction.

Reducing bed usage, where clinically appropriate: Multiple studies have shown that a proportion of patients remain as inpatients on days when no treatment or investigation is being provided, representing considerable waste. Processes that cause delay, such as waits for scans or OT assessment must be investigated, and demand and capacity balanced so that delay is significantly reduced. This should of course be preceded by assessment of need so that excessive variation in requests is reduced.

Key messages on reducing bed use in the frail elderly

Interventions which improve frail elderly patient outcomes and reduce bed use:

• Reorganisation within hospitals to provide care in special units (eg. geriatric assessment units, acute care of elderly units, orthogeriatric units)
• Multidisciplinary early discharge planning
• Clinical pathways for the most common presentations
• Comprehensive geriatric assessment
• Senior review early on in admission
• Ambulatory care (Investigating, treating and following-up patients, but avoiding their admission).

Interventions that improve frail elderly patient outcomes and might have an impact on bed use:

• Interventions to prevent delirium
• Medication review
• Treatment for malnutrition
• Exercise interventions.

Changes which introduce interventions need to consider local context and needs. Staff should be involved in planning and implementing the changes. Data needs to be collected at baseline and measured going forward to see that changes introduced are having the desired effect.

(Based on a Literature Review commissioned by Scottish Association of Medical Directors 2015)

Acute and tertiary care services need to continue the extensive work that has been undertaken in improving processes in healthcare. We should standardise process where it is appropriate to standardise^[12], and redesign flexible processes where required. While there is a financial benefit from improving process, it should be recognised that in all of the examples given above, there is direct patient benefit, and indirect patient benefit (in that more patients can potentially benefit from the fixed resource available).

Successful change in processes often requires understanding and use of improvement methodology. The work on training staff and building improvement capacity must continue. The other requirement for progressing effective change is identified clinical leadership, and the NHS requires to invest time in building clinical leadership across all disciplines to drive forward significant change in process to make care ultimately more patient centred, as well as making better use of existing resources.

Structural change in hospital services

Evidence shows that specialised procedures, concentrated on a small number of high volume sites, will improve outcomes. This conclusion is prompted by evidence that - especially for complex procedures - there is a relationship between the volume of procedures carried out by a surgeon and the likelihood of improved outcomes.

The same appears to also hold true for hospitals - indicating that the contribution to the best possible outcomes comes partially from the wider team as well as the surgeon.

Over the past 10 years there has been some concentration of services on a small number of hospital sites, allowing specialised services to produce the predicted higher quality outcomes. This has allowed the evidence base for improved clinical outcomes of specialist units to develop. For example in England, there is evidence that the centralisation of vascular services is resulting in better clinical outcomes^[13]. The UK wide mortality rates following abdominal aortic aneurysm surgery fell from 7% in 2008 to 2.4% in 2012^[14]. A report from the Vascular Society published in November 2015 shows that this has continued to fall, and is around 1.5%. This is significantly, but not wholly, related to concentrating vascular services on fewer sites.

The early evidence on the relationship between increased volume and improved outcome tended to consider only post-operative mortality. More recently however there has been accumulating evidence that outcomes that are less dramatic - but extremely important to patient wellbeing - are positively impacted upon by care provision in specialist units. For example a specialised urology unit in Germany had a five year prostate cancer survival rate which was only slightly higher than less specialised units treating lower volumes of patients.

However the specialised centre had a rate of permanent severe incontinence of 8% that was a quarter of the rate of the other units, and a rate of impotence that was around one third of the other units. These are serious quality of life issues that will have an obvious impact on patients. Another review showed that centralisation of care resulted in an improvement in five-year survival from 58.6% to 68.6% for all gynaecological cancers that could be staged and graded. These changes have been most marked with endometrial and ovarian cancers^[15]. A further report considered the impact of a reconfigured regional upper gastro-intestinal cancer service: "The curative to palliative treatment ratio increased by 71%, operative morbidity fell 50%, lengths of hospital stay reduced on average by 3 days, median survival improved by 20% and overall 1 year survival improved by nearly 20%"^[16].

There are known examples of where we accept a structure that is unlikely to produce the best possible outcomes. For example, evidence from the US suggests that a surgeon doing hip replacement operations should do at least 35 operations per year. At that level of activity the occurrence of complications falls to around the minimum level - although a small further improvement is seen with increased activity beyond that level. In Scotland we provided about 7,600 hip replacements and 7,170 knee replacements in 2013/14^[17]. There were also 950 hip arthroplasty revisions, and 460 knee arthroplasty revisions. Hip and knee arthroplasty revisions are recognised to be more complex and challenging procedures, and there is a greater risk of adverse outcome for the patient. The arthroplasty project report results show that 40% of hip revision operations were carried out by surgeons who do less than ten such operations per year, and just under one third of the knee revision operations were carried out by surgeons who do less than five procedures per year. Some of the revisions will have been non-elective, but a significant proportion were not. Whilst the surgeons may have produced acceptable results in the patients, it seems to be the case that such arrangements increase the risk of adverse outcomes - a point acknowledged by the Arthroplasty Project Report.

This section presents only a fraction of the available evidence that better outcomes are obtained by organising and delivering some procedures in larger, more specialised and better resourced centres. In order to ensure that a surgeon, or a surgical team, deliver enough clinical procedures to obtain the best outcomes, that surgeon or team must provide services to a large enough population to ensure that they will retain their skills and maintain sustainable services through an adequate volume of activity. It is important to understand that a specialty delivers a range of services - some of which can be safely delivered locally by skilled staff, while others need to be delivered in a larger unit. The conclusion from this is that we need to plan individual specialties on the basis of populations - ensuring that we are planning for a large enough population to preserve high standards of specialist skills, as well as ensuring that there is an adequate range of local services which can be accessed. This would include out-patient services, diagnostics and day-case surgery.

If we are to ensure that the population are able to benefit from the better outcomes that are provided by larger volume centres, then we need to move from planning at a Health Board level to planning at a population level. This is not new. Work on the Scottish Vascular Services Framework^[18] indicated that it was not appropriate to plan to deliver a vascular service for a population of less than 800,000. This population would provide a level of need that would allow a vascular surgery service to deliver enough interventions to maintain skill levels amongst the clinicians involved. It also made effective use of resources - in order to provide 24/7 availability of the service, it needs to be of a sufficient size to have rotas that will allow a realistic prospect of recruitment of highly skilled surgeons.

The changes suggested have also been seen in cancer services. The planning of cancer services on either a national basis, or through the three regional planning groups of Health Boards has allowed specialisation of services, with improvements in outcomes for patients. Planning across larger populations has made high quality specialist services available to more patients, even though some patients have to travel further.

The management of acute heart attacks changed some years ago. Instead of treatment being provided at all local hospitals, most patients are transferred as an emergency to a small number of centres across Scotland where there are teams of specialist staff who can urgently provide specialist services designed to unblock the occluded artery that has caused the heart attack. This has led to a higher quality clinical intervention being reliably available to patients, and has contributed to improved outcomes.

For some specialties therefore, it is only by planning services across a larger population that it is possible to provide the range of services that might be required, including an effective arrangement to deliver urgent services over 24/7 time periods, effective maintenance of skills by the surgeons and other skilled clinicians, cost-effective provision of the complex technology that may be required (such as hybrid theatres) and a service that is able to deliver world-class outcomes.

It is not appropriate for all services to be planned on the basis of large populations. The volume of unscheduled medical admissions means that we will continue to require inpatient beds in local hospitals as at present. Therefore, for example, there would not be a need to plan care of the elderly services on a large-population basis. It is likely that most hospitals would require a minimum of an accident and emergency service, an acute medical admission unit and supporting inpatient wards, including care of the elderly, AHP services (such as physiotherapy/speech and language therapists, dieticians etc), outpatients, laboratories and diagnostics, critical care and day-case theatres.

This strategy proposes that for many specialties, services should be planned for a population, and delivered across a network of hospitals. Within that network, one or more hospitals would provide inpatient care, and access to specialised or complex treatment. Other hospitals in that network would not provide inpatient services, but in order to provide local access to services, would provide out-patient clinics, diagnostics and day-case surgery. The network of hospitals would be helped to work effectively by established clinical pathways, by electronic availability of clinical records, (including radiology and other test results) and by promoting strong connections between all clinicians involved in the network. In some cases - and particularly to support smaller and more rural hospitals - clinicians would work across more than one hospital to assist service delivery. Over time this will be replaced by greater adoption of telemedicine so virtual consultations can take place electronically. This allows specialist input to be delivered to remote and rural locations, and has been shown across the world to provide a very satisfactory and clinically safe service.

The outcome from these changes would be that increasingly certain services are planned across large populations, regionally or nationally as appropriate, resulting in:

• optimal clinical outcomes from fewer, specialised hospitals
• effective use of highly skilled staff
• more standardised care, through agreed clinical pathways, and optimal use of high technology equipment
• services that are much less dependent on a small number of individuals, and
• excellent centres for teaching, research and development.

Within Scotland, it is possible to see how networks of specialty services could be arranged within a grouping of say five to six hospitals. Consider the possibilities for urology for example (currently supplied on 21 sites across Scotland). In a regional model, a concentration of all inpatient beds and major surgery in one hospital would meet the requirements of a more specialised unit (with some patients being referred for more complex procedures such as robotic radical prostatectomy to a small number of nationally designated sites). If a network of sites were developed it would be possible to arrange that out-patient, diagnostic work, and minor procedures like day-case cystoscopy to be provided across all hospitals in the network. This would require the appropriate IT and communication systems to be in place to ensure information is available in a timely and efficient manner across locations. This achieves a more specialist service with improved outcomes, local access to the bulk of other clinical services, and more efficient use of the skilled staff and other specialist resources in the service. All inpatient care would be centred on one site. The reduction in the number of out of hours rotas would considerably reduce workforce pressures.

In higher volume specialties, such as orthopaedics there would be a need for a larger number of beds to deal with the volume of inpatient work. It would be appropriate therefore to have orthopaedic services on several of the sites within the cohort of hospitals. This could be in the form of the traditional orthopaedic departments, or there could be separation of elective and unscheduled care. In any event, the formation of a network could be used to address the need for a degree of specialisation, as well as mitigate the constraints provided by the limited workforce. Services like vascular surgery and interventional radiology have already, to a degree, been concentrated on fewer sites to make best use of limited skilled staff, and specialist equipment. The principle of fewer sites for some surgical specialties would apply to some medical specialties as well. The evidence of improvement in stroke outcomes from having fewer stroke units in London is hard to ignore, and should prompt consideration of how many hospitals should deliver hyper-acute stroke services, particularly across the central belt. The Royal College of Radiologists has produced a report on the future of radiology proposing that there should be regional planning of services. They are clear that image capture should take place in all hospitals, and all hospitals should employ radiologists, but by using the varied clinical expertise across a network and making use of the PACS system which allows remote access to digital scan and x-ray images, it becomes possible to deliver increased specialisation. This proposal would also help make the services more resilient by ensuring that should there be capacity problems in one hospital with regard to interpretation of results, there would be protocols across the network to provide remote assistance. It would also enable much better peer-to-peer consultation, a key component of clinical decision support. A key component to developing this model is further work on the radiology information systems to ensure excellent connectivity across Health Board boundaries.

Rural general hospitals

It is important in the context of this strategy to recognise the significant contribution of rural general hospitals to the provision of healthcare in our more remote areas. Despite small volumes of activity, they have to be capable of providing primary emergency care for the complete spectrum of emergencies, and appropriate onward referral when required. It is essential that these hospitals are supported to maintain emergency and elective services: this requires Boards to collaborate to ensure that these hospitals are supported - the success of this has been described earlier, describing the networking of clinicians in the north of Scotland to ensure that, as far as possible, specialist services can safely and effectively be provided in the rural hospitals, often by visiting specialists. Further developments will be enhanced by increasing use of IT.

In rural and island settings it is more difficult to maintain high quality clinical services across a wide range of specialties. In many cases there is not the activity to justify the employment of specialists. Specialists may not wish to work in more remote areas where their valuable acquired skills may decrease through lack of use. However there are examples of where a regional approach to the planning of services for more remote hospitals has improved services significantly. For example, stroke services in the Western Isles are shared between local clinicians and a stroke specialist in another health board. The stroke specialist does regular "virtual" ward rounds with the local clinicians in order to provide expert input. The specialist visits the islands regularly to maintain good relationships with clinicians he is working with, and to support teaching and protocol development. This approach can be much more widely used if there is a firm commitment to regional planning of services from all concerned, ensuring that we significantly reduce inequity of access to expert care.

A review of available literature on reconfiguring clinical services was undertaken as part of this strategy development. Whilst the detail may require further expert advice, analysis suggested that, for the services considered, there are advantages as described in adopting a national, regional or local approach to planning of services for relevant populations. It must be emphasised that regional planning of a service does not mean that there would only be a regional delivery of that service.

Specialty	Proposal
Cancer (less common cancers and related oncology)	Low volume and specialist nature makes the case for planning for less common cancers on a national basis, though actual delivery of services might be on more than one site.
Neurosurgery	Low volume, high complexity, need to support major trauma centres - all suggests a national, clinically integrated planning approach.
Burns care	Decreasing numbers of severe burns, and increasing success of specialised units suggests that national planning would be appropriate.
Cancer (higher volume)	Services planned across regions as now, with more specialist services available in cancer centres.
Orthopaedics	Planned regionally to provide for emergency trauma work and expanding volume of elective work for an ageing population.
Radiology	Planned across regions - regional planning approach may support change as described above, assuming technology issues addressed. Could help standardise use of radiology. Links to interventional radiology need to be considered.
Paediatrics	Planned regionally, ensuring good local access to day-case and diagnostics, community support, but regional planning of specialist inpatient services.
Urology	Regional planning with reduction of inpatient sites, but retained local access to out-patient, diagnostic and day-case surgery. Emergency pathways must be established (though low volume).
Stroke	Evidence from London shows regional planning of stroke services and reduction in number of sites resulted in improved survival. Needs further evaluation in Scottish context, and in view of emerging possibilities of thrombectomy for stroke. This will require a regional approach.
Cardiology	Already has strong regional component - requires to be planned across regional network of services.
Ophthalmology	Mostly local Board level planning at present - this requires review. Some high volume services such as cataract surgery may be dealt with by elective centres, though there will be a need to plan remaining services across populations. Some care (eg. stable glaucoma), could be transferred to high street optometrists.
Oral and maxillofacial surgery	Includes high volume of day-case work, and small volumes of highly complex work - so suitable for a regional planning approach. Could have highly specialised out-reach staff delivering services across multiple hospitals.
Neonatal	Being reviewed currently by maternity and neonatal review.
Maternity	Being reviewed currently by maternity and neonatal review.
General surgery	Potential for joint local/regional approach to planning. Should review pathways for emergency out of hours surgery - may benefit from more specialist centres - but workload considerations.
E.N.T	Includes high volume of day-case work, and small volumes of highly complex work - so suitable for a regional planning approach.
Gynaecology	Includes high volume of day-case work, and small volumes of highly complex work (cancers, endometriosis) - so suitable for a regional planning approach.
Intensive care	Should relate to trauma centres[19], and elective surgery requirements.
Mental Health	Services planned across regions, delivered locally: Some tertiary level services planned nationally (eg high secure, specialist in-patients, CAMHs)

This is not, of course, an exhaustive list of specialties concerned. Where specialties have not been listed, it is because of a relative lack of evidence on reconfiguration. This must not be confused with evidence of a lack of benefit - it is likely that the same basic principles apply, especially where there are low volumes of cases, or complex interventions involved.

It seems appropriate to conclude that some core services which do not involve highly specialised interventions, and have moderate to high levels of demand, should continue to be planned for at a local level. This would include a number of specialties such as care of the elderly and palliative care.

It is vitally important to the public that services required in an emergency are of high quality and structured so as to deliver the best possible outcomes. Responses to emergencies have been improved with better response times from the Scottish Ambulance Service, and the continuing evolution of the highly trained paramedic role. The current work on the evolution of major trauma pathways for critically injured patients will improve outcomes, and projections suggest this will save an additional 40-50 lives per year. The emergency pathway that has been in place for some years in relation to the management of myocardial infarction (heart attacks) has shown that directing the patient with a myocardial infarction beyond the most local Accident and Emergency Department to a specialised centre with 24 hour a day capability to perform coronary artery interventions immediately has resulted in a significant reduction in mortality from heart attack. Emergency stroke pathways have been developed to ensure rapid scanning, and administration of appropriate medication to relieve symptoms caused by occlusion of carotid arteries. These examples indicate that, for a proportion of patients, and particularly for those who are most unwell, transfer to the most local A&E department may not be the best possible option. Instead, we need to focus on the development of emergency care pathways that are responsive to different local and clinical contexts to achieve best possible outcomes.

It is appropriate for A&E services to be available at hospitals locally - but it is vital that Scottish Ambulance Service staff have the capability and are supported to respond promptly in making an early diagnosis and streaming the patient to the most appropriate emergency pathway to ensure rapid access to high quality definitive care.

Separation of elective and unscheduled care: Diagnostic and Elective Treatment centres

We know that speedy access is important to patients. Prompt treatment reduces anxiety, leads to better outcomes and avoids further clinical deterioration. Scotland stands among the best in the world in delivering prompt and effective heath care. For example the conversion of the Golden Jubilee hospital to a high volume, specialist centre for a relatively narrow set of surgical conditions has helped reduce the need for buying over flow capacity from the private sector. It also has driven increasingly high quality care, with lower rates of complications for procedures than those that are carried out in lower volume hospitals. Given that there will be an increasing need for a range of age-related surgical interventions as the population changes (e.g. hip replacement, knee replacements, cataract surgery) there is a need for planning for increased diagnostic and treatment capacity. This has been recognised by Scottish Government who have recently pledged a total of £200 million over the next five years to expand capacity across a number of hospitals. The expansion capacity that will be provided should be considered when regional planning processes are developing options for the consolidation of some services into fewer centres of excellence. The geographical spread of the proposed developments offers significant potential across most of Scotland and particularly in Health Boards where the forecasted population changes will be largest.

These new facilities will be designed to adopt best practice in the clinical delivery of services with the latest technology and enhanced recovery techniques. The new centres will significantly reduce the chances of cancellation and the use of the private sector. We will also wish to ensure that this investment in elective care leverages in benefits for the wider community with greater operational efficiency and with the promotion of smooth flow through the entire healthcare system.

Summary

The concluding principles from this section are:

Most care will be provided locally with the expansion of primary care avoiding many having to access secondary care at all.

Most local hospitals will be able, as now, to provide emergency services, including accident and emergency services, out-patient, diagnostic and day-case services across a range of specialties.

Using a network of hospital sites, some specialties will provide inpatient services in a smaller number of hospitals. This will allow hospitals to develop a degree of specialisation in some specialties to ensure high quality outcomes.

The evidence suggests that secondary and acute care services should be planned on a population basis - which could be either regionally or nationally. Further work needs to be undertaken to establish which services might better be planned regionally or nationally for local delivery across a relevant hospital network. It should be noted that planning a service regionally or nationally also supports and underpins local delivery of services in that specialty.

The proposal is not that large hospitals would harbour all of the specialist inpatient units. This would be a wasteful use of our current hospitals. But it will be possible for services to be planned regionally in a way that sees the advantages of specialisation, and identify hospitals which will become more specialist centres of care. This will improve patient outcomes, will ensure that there is an equitable standardisation of services, will make best possible use of skilled staff, and may result in reduced costs that will help sustain services. The reduction in the number of inpatient units will reduce the requirement for the employment of locum staff, preserving a significant resource for more effective delivery of value to patients. The changes proposed will support the delivery of 24/7 emergency care in more specialist services.

These changes will be complex and must be accompanied by the improvements in process that have been described earlier. It may mean that clinicians have to change their ways of working. It will require improvements in the electronic transfer of patient information. But experience elsewhere has shown it is possible, and it does result in improvements to patient outcomes, and it will make services more sustainable. It is imperative that progress is made on these changes as soon as possible if we are to maintain a high quality service to all patients that will be sustainable in the challenging times ahead.