Context and definitions

What we mean by ‘whole system flow’ and why it matters

The language used by those who have an interest in improving the flow of people or resources within or between services is often technical: batching, bottlenecks, constraints and capacity management. Yet the effects of poor flow are all too readily apparent in the daily experiences of patients, service users and people working at the front line of health and social care. Stark examples of a lack of smooth flow through a system include ambulances queuing outside hospitals, crowded emergency departments, long waits on trolleys for a bed, pressures on community and social care services, overstretched GPs, and mental health patients being transferred hundreds of miles for an inpatient bed. All of this is stressful and frustrating for staff and can be devastating for patients, service users and their families and carers.

At a time when health care is under increasing financial pressure, poor flow is both a symptom and a cause of that resourcing crisis. Delays and waits are exacerbated by deficiencies in critical areas and the resulting disruption to flow leads to an ever more suboptimal use of the resources within the health and social care system.,,

Where providers have been able to match capacity and demand and enable better flow between departments and organisations, there have been impressive results.

Work supported by the Health Foundation at Sheffield Teaching Hospitals NHS Foundation Trust and South Warwickshire NHS Foundation Trust through the Flow Cost Quality programme has delivered sustained reductions in emergency care length of stay, bed occupancy and readmissions, while improving safety and the patient experience., In the US, leading high-performing providers such as the Mayo Clinic, Seattle Children’s Hospital and Cincinnati Children’s Hospital have achieved significant productivity gains and savings using flow improvement approaches (see Box 2).

Yet most flow-related initiatives to date have focused on a small segment of the patient or service user journey, usually within acute hospitals. There is a pressing need to look beyond the hospital and to give attention to every team, service and organisation that patients and service users encounter. As well as looking at services delivered within the NHS there is also a need for consideration of social care, health promotion and other local government services. Nor should the wider determinants of health (eg age, lifestyle, environment) be forgotten – or the need to tackle the gulf between physical and mental health services.

It is a lot easier to call for a whole system approach to flow than it is to deliver it. Over the past 15 years a series of UK national bodies have made strong cases for looking beyond the hospital – to look at flow from a system-wide perspective.

The most recent – but far from the first – example is NHS England’s 2015 guide for health and social care communities on delivering urgent and emergency care. This underlined the value of whole system partnerships in improving flow. A decade earlier, in 2005, the NHS Institute for Innovation and Improvement published a guide to improving flow for system leaders, which made the point that most existing flow-related improvement work had tended to focus on single bottlenecks in the system. Future work, it said, needed to understand the flow of patients across departments, organisations and the whole system. There was also a strong emphasis on ‘whole systems working’ in many of the programmes supported by the NHS Modernisation Agency; for example, the Acute Local Improvement Partnerships, announced in 2003, used ‘a whole systems approach to follow the patient’s pathway across departmental and clinical boundaries, to deliver better care and minimise delay’. In Scotland, meanwhile, a 2007 report on patient flow in planned care made it clear that ‘the importance of getting the flow of patients right across the whole system cannot be overstated’.

Yet it has proven difficult to translate this whole system vision into reality. In Section 1.2 we consider the issues that have made it so difficult, but also describe why there are reasons to believe that change is achievable.

The challenge of achieving system-wide change

Albert Einstein once said that ‘without changing our pattern of thought, we will not be able to solve the problems we created with our current patterns of thought’. It is a quote that often crops up in articles about how to enable change in complex systems such as the NHS27 – and it does so for good reasons. Health and social care leaders and policymakers across the UK have been trying for years to cajole or nudge the various organisations and groups within their world towards a more ‘whole system’ way of working. Yet despite a succession of national initiatives – the latest in England being the Integrated Care Pioneers and New Care Models vanguard partnerships – genuine, joined-up, whole system delivery is still the exception rather than the rule.

Various elements of the system are governed, funded, inspected and regulated in silos. This reinforces significant differences, not just geographically but also culturally, between those working in hospitals and those working in community services or in primary care. As the NHS five year forward view (Forward View) stated, in relation to England, many elements of the ‘classic divide’ between ‘family doctors and hospitals, between physical and mental health, between prevention and treatment’ that characterised the NHS in 1948 remain in evidence today.

The cultural divisions between the NHS and local government are often even sharper. Local authorities are less bound by central government direction, face resource pressures even more extreme than those in health care, and are driven by the need to deliver on a local democratic mandate expressed through elected members. Across the UK there is now a greater emphasis on collaborative working between local government and the NHS. A number of new initiatives such as health and wellbeing boards and the Better Care Fund in England, and health and social care partnerships in Scotland, have been set up with this aim in mind. But while these arrangements are helping to build trust and understanding between organisations, the picture is still very patchy. In some areas, the challenge of managing potentially thorny and politically contentious processes, such as the handover of patients from acute to social care settings, exacerbates tensions between providers.

During times of financial uncertainty and risk, it is not easy to encourage people and organisations to do things differently. While some see a ‘burning platform’ and are galvanised into collaborative action, others respond to the pressure of the situation by clinging ever more tightly to their established ways of working. As Peter Senge has noted, our brains tend to ‘downshift’ under pressure and we revert to our most habitual modes of behaviour. Evidence of this can be found in the tendency of some NHS providers, when faced with many competing demands, to adopt a highly bureaucratised form of management that leads to defensive and reactive behaviour and superficial displays of compliance rather than genuine efforts at improvement. Others, meanwhile, are so focused on the task of securing their immediate survival and on short-term business priorities that they do not have the headspace to think about the long-term gains that can come from working collaboratively. Quite simply, they are too busy firefighting the latest crisis to worry about anything else.

In saying this, there is room for some optimism. If you look at the factors that David Gleicher and others have suggested are necessary in order to deliver meaningful change – dissatisfaction with how things are now, a vision of what is possible, an appreciation of how change is to be implemented, and the capacity for change – there are grounds to think that we are close to tipping point on many of them.

With health and social care budgets severely stretched in every UK nation, the financial logic for tackling expensive and resource-intensive bottlenecks in the flow of patients, service users, information and equipment across the system is hard to resist. The moral and emotional case – exemplified by the human costs of delayed hospital discharge of frail, older people – is equally powerful. Moreover, there are enough inspiring examples of effective cross-organisational working – some of which are highlighted in Sections 3 to 5 – to show that real change is achievable even in the most pressured of times.

The renewed interest at national level in prevention and public health, described in England in the Forward View as being in need of ‘a radical upgrade’, is also helping to create the conditions in which local health and social care system leaders are ready to work together to improve flow – or, better still, ensure that people do not need to flow into the system at all. One indication of this is the emerging interest among public service leaders, particularly in Scotland,, in the concept of ‘failure demand’, or ‘demand caused by failure to do something or do something right for the customer’. By focusing on avoiding failure demand, a requirement is placed on health and social care leaders to work alongside their peers across the whole of the public sector, including, but not limited to, housing, education and employment.

There is also a better understanding of the capabilities and methods that can help to deliver whole system change. Senge, for instance, takes heart from the ‘extraordinary expansion in the tools to support system leaders’ to ‘see the larger system’, ‘foster reflection’ and ‘co-create the future’. He argues that the strategic use of these tools ‘at the right time, with the right spirit of openness’ can help to address ‘previously intractable situations’ and inspire confidence that change is possible. What is not in evidence, as yet, is a critical mass of people across the health and social care landscape with the capabilities to use these tools effectively.

This report aims to help local health and social care economy leaders as they begin to think about how to build the necessary capabilities to improve flow on a system-wide level. Section 1.3 describes the core elements of an archetypal local health and social care system, to which this report speaks.

Defining a health and social care system

Enabling change across an entire health and social care system is not easy. Even tasks that would appear to be fairly straightforward, such as defining what the system is, can be challenging. If you were to ask a dozen health and social care professionals to define their local system, you would likely receive a dozen different answers, even from people who worked in the same department. Ultimately, it will depend on what each person sees as being the core purpose of the system: those who see the avoidance of failure demand, for example, as being its key organising principle may define the system in broader terms than their peers focused on the operational realities of meeting the needs of the patients and service users in front of them. When change is being planned and delivered, it is important to surface such views early on to avoid potential misunderstandings and conflict at a later stage. Moreover, a shared definition of what is in and out of scope is an essential first step in understanding a system and identifying the weaknesses and constraints within it.

While each set of local leaders will define their systems according to their local context and priorities for action, it is useful to have an archetype in mind when describing, as this report seeks to do, the capabilities and resources needed to improve flow on a whole system basis. Our archetypal system, illustrated in Figure 1, is focused on the organisations that will be involved once a need for care has been identified. This care system sits within a wider system that influences the health and wellbeing of the public. The relationship between these two distinct but interlocking systems is crucial. From a health and social care provider perspective, a close and transparent relationship between them and other partners with a wellbeing focus – which provides scope for joint working, information sharing and peer challenge – will help to ensure that any care system redesign activity is consistent with the needs of the wider population. It may be, after all, that resources allocated towards optimising primary and acute care journeys could have a bigger impact if they were used to address an underlying cause of ill health in the area.

Figure 1 is based on a typical system in England and is designed to illustrate the possible extent of a system-wide approach to improving flow, in terms of the number of organisations, professionals, patients and service users involved. Figure 2 illustrates an example of the care journeys that run through a local health and social care economy. It shows the many teams potentially involved.

Figure 1: Anyborough, England – a local health and social care economy

Figure 2: Anyborough local health and social care economy care journeys

It can be difficult, even for experienced system leaders, to navigate the landscape illustrated in the Figure 1. While this makes flow improvement hard, methods for understanding and improving flow (see Box 1 and Section 3.2) can help to make sense of the landscape. AQuA’s work with organisations in the north west of England has led it to conclude that a focus on patient flow can be an effective way of helping people see and understand the complexity of the system in which they are working. Flow improvement methods provide a means of fostering greater collaboration within and between organisations and designing care models that will better meet the needs of the local population.

The benefit of examining the system through a care journey lens, as shown in Figure 2, is that it allows system leaders to see how services are connected and where constraints may exist. It also allows them to start to consider what activities might be amenable to flow improvement approaches and what resources and capabilities they will require. It may be that approaches which have been primarily used to improve flow in acute contexts may need to be adapted, or may not be applicable at all, to the type of challenges faced by organisations focused on prevention and continuing care.

Identifying flow within the system

The concept of flow is closely associated with the approach to quality and productivity improvement known as ‘lean’ or the Toyota Production System. In their definitive book on the subject, Lean thinking, James Womack and Daniel Jones use health care as a prime example of the lack of flow in a system.

‘What happens when you go to your doctor? Usually you make an appointment some days ahead, then arrive at the appointed time and sit in a waiting room. When the doctor sees you, usually behind schedule, she or he makes a judgement about what your problem is likely to be. You are then routed to the appropriate specialist, quite possibly on another day, certainly after sitting in another waiting room. Your specialist will need to order tests… requiring another wait and then another visit to review the results… If you are unlucky and require hospital treatment, you enter a whole new world of disconnected processes and waiting.’

Lean practitioners argue that the absence of flow arises out of the ‘batching’ of patients, service users and routine tasks, so that they are seen or completed at the same time by members of staff. For patients and service users, this can be incredibly frustrating, as it means that they often have to wait in a queue until the next stage in the process is ready to begin. It is also an enormous source of potential error, duplication and waste.

As well as looking at the flow of patients and service users through a set of care processes, it is important to look at the flow of the information, resources and staff that need to come together to enable effective care of these individuals. In acute settings, the flow of staff to the patient can be critical, for example, having an early senior clinical decision maker available on arrival in an emergency department. Effective flow of information across a system also matters: for example, if all professionals treating a particular patient had access to a shared care record it would significantly reduce waste and delays.

An effective flow of resources is also essential so that a lack of finance in one part of the system – for example, in social care or domiciliary services – does not mean that patients and service users experience a delay in discharge and an unnecessary stay in hospital because there is nowhere for them to go. These different types of flow need to be made visible and purposefully designed and managed to ensure they are mutually supportive.

Understanding flow across the whole system

Once the different flows have been identified, further work is needed to understand variations in demand and capacity within the system and the root causes of them. While there are examples of analysing the flow of patients, service users, data and resources within specific services or organisations, rarely has this been done across a whole system.

Analysing flow across a whole system is a major undertaking. This is especially so given the lack of easily accessible cross-organisational data, and the shortage of analytical capability. Yet, as discussed in Key area 3, such analysis is critical.

While it is important to be pragmatic about the time and resource available for analysing the system, experience from other sectors underscores the importance of doing so. Analysis is especially important when the systems are too big and complex for people to easily see through their direct experience, or to be able to predict how they will respond to change. Consequently, there is a strong case for investing substantially in system analysis before making changes to care processes and services, especially given the potential cost and quality implications of any changes.

Some common approaches that have been used in the UK and other countries to understand flows across complex organisations or care journeys with many variables and interrelationships are described in Box 1.

Box 1: Methods for understanding system flow

SIMULATION AND MODELLING

Simulation and modelling of patient or service user flow can provide insight into where bottlenecks occur in a health care system. They allow service planners to evaluate the benefits and pitfalls of potential improvements before enacting them.

Simulation has been widely used in manufacturing and in the logistics sector with the aim of optimising throughput and profitability. Gatwick Airport, for example, has used simulation of passenger flow through the check-in process to increase understanding of variation and where bottlenecks were occurring. Gatwick was able to make changes to the process as a result and has seen an improved check-in process with reduced queue times and improved airline efficiency.

In health care, simulation and modelling approaches have been used to manage bed capacity, schedule staff, manage admission and scheduling procedures, and to test the value or functionality of new initiatives and services before they are implemented.

The effectiveness of these approaches is often contingent on the quality of the process mapping used to inform them, as well as the robustness of the data used to populate them. Finding sufficient funding and staff with the right skills to build and run simulations is a further challenge.

Resources

  • SAASoft has developed a whole system dynamics simulation for education purposes, with five interdependent components: home, hospital, intermediate care, care home and cemetery.
  • The Basic Building Blocks methodology published by the Scottish government offers a systematic approach to the demand and capacity analysis of existing care journeys. Its tools can be used for simulation modelling.

Examples from the peer reviewed literature

  • In England, researchers used simulation to create a ‘perfect world model’ for accident and emergency (A&E) care – not as it is, but as it could be. Importantly, the ‘efficiency gap’ between the ‘perfect world’ and the ‘real world’ was used to identify the location of bottlenecks in the current ‘whole hospital’ care journey and brainstorm ideas for improvement.
  • An English primary care commissioning organisation focused on improving the use of unscheduled care and support efficiency gains in the local hospital. A model of the system was developed to help set usage targets at the micro-level of the hospital. The model drew on a small number of readily available key data items. The model emphasised that primary care had an important role in changing the culture, communication and care provided within A&E and other unscheduled services.
  • A Swedish hospital has used a simulation model to support discussions about the resources, capacity and work methods that would be required on a maternity ward that was shortly to be built.
  • In Canada, a simplified, low-cost simulation platform, developed using spreadsheets, was found to be as effective in predicting patient flow patterns as more expensive commercial software packages.

VALUE STREAM MAPPING

Value stream mapping (VSM) is an approach that produces a visual map of a system or process.

It is often used by multidisciplinary teams to improve processes as part of lean/continuous improvement projects.

Using VSM, a team can produce a visual map of the ‘current state’, identifying all the steps in a patient or service user’s care journey.

The team then focuses on the ‘future state’, which often represents a significant change in the way the system currently operates. This means that the team needs to develop an implementation strategy to make the future state a reality.

Using VSM can result in streamlined work processes, reduced costs and increased quality.

Resources

  • The NHS Institute for Innovation and Improvement produced a guide to using VSM.
  • Another useful guide to VSM can be found on NHS Scotland’s Quality Improvement Hub.

Examples from the peer reviewed literature

  • In Ireland, researchers used lean principles and the theory of constraints to identify bottlenecks in patient journeys through A&E. For each stage of the patient journey, average times were compared and disproportionate delays were identified using a significance test.
  • A value stream map and the five focusing steps of the theory of constraints were used to analyse these bottlenecks.
  • A US multidisciplinary team analysed the steps required to treat patients with acute ischemic stroke and developed a streamlined treatment protocol.

QUEUEING THEORY

Queuing theory, or the study of waiting lines, or queues, can help to understand and address mismatches between service demand and capacity. Usually a mathematical model is constructed to help predict queue lengths and waiting times. Historical data are analysed to explore how to provide optimal service while minimising waiting, thus providing an objective method of determining staffing needs during a specific time period. Popular in other industries, queuing theory has also been used in health care, particularly by hospitals wanting to understand waiting times for unscheduled care or the time spent waiting for specific equipment, surgery or laboratory results. It is also applicable to wider systems of care or transitions.

Examples from the peer reviewed literature

  • A hospital in England used queuing theory to analyse one year’s worth of data to help understand the practical challenges associated with variation in patient demand for services and length of stay. The analysis found that daily bed shortages are mostly influenced by the timing of arrival and discharge of patients with a short length of stay, and that bed shortages around holiday periods are not due solely to increased demand, but also a reduction in staff and service capacity in and out of hospital around these times.
  • In Canada, researchers used queuing theory at an organisational level to analyse the relationship between patient flow to A&E and patient flow to the inpatient unit. They then used the model to estimate the average waiting time for patients and the resources needed in unscheduled and inpatient care. The model was used to analyse the potential impacts on waiting time and resources of an alternative way of accessing unscheduled care and this helped managers plan the resources needed to enhance patient flow.
  • The Scottish Whole System Patient Flow programme has also been informed by queuing theory (see Box 5).

Experience of improving flow at a system level

In recent years, there has been a significant growth in thinking about how to improve flow within health care processes and systems. One of the leading experts in the field, Eugene Litvak suggests that to improve flow in a health care setting it is necessary to:

  • understand variation in the performance of a process over time and its sources
  • separate patient flows into appropriate streams
  • redesign work processes for those streams to smooth out the flow
  • match capacity with estimated demand.

Around the world, a number of hospitals have worked to redesign care journeys in order to improve flow using these principles (see Box 2).

Within the UK, the Royal Bolton Hospital took a similar approach between 2004 and 2010. This led to improvements in quality and productivity. For example, a redesign of the process for patients with fractured hips reduced length of stay by 33% and reduced standardised mortality by 50%. An audit concluded that there had also been a 42% reduction in paperwork for the staff involved.

More recently, flow improvement programmes have been implemented across Wales and Scotland (see Box 3).

The Welsh Patient Flow programme, which involved all health boards with general hospitals that admit emergency patients and the Welsh Ambulance Service, has succeeded in delivering some improvements in flow in local pathways. It has also generated some valuable learning about the challenges involved in using a national breakthrough collaborative model to improve flow across multiple sites at the same time.

However, while primary, community and social care services have been involved in the Welsh programme, much of the improvement activity has focused on the acute sector. The same is true of the Scottish Whole System Patient Flow programme, which got underway in 2013, a few months after the Welsh programme.

The challenge now is to build on this work to improve flow within hospitals, and develop approaches which look at flow across the whole of the health and social care system. This would involve the smoothing of demand upstream – in particular in general practice – and the development of community resources downstream to allow a smooth and safe flow of patients out of hospital once they are fit for discharge.

Box 2: International examples of flow improvement programmes

Cincinnati Children’s Hospital implemented a series of measures to match capacity with demand and improve quality. It reported a cost avoidance of $100m in capital costs and an increase in its margin of more than $100m annually.

‘Esther’ in Jönköping in Sweden was led by a team of physicians, nurses and other providers who joined together to improve patient flow and coordination of care for older patients within a six-municipality region. It reported significant reductions in hospital admissions, days spent in hospital by heart failure patients, and waiting times for referral appointments with a range of specialists.

Intermountain Healthcare is an integrated system in Utah and Idaho in the US that consists of 23 hospitals and 160 clinics and has a workforce of 32,000. Its improvement journey, which began in the late 1980s, has been influenced by Deming’s insight that the best way to reduce costs is to improve quality. Intermountain has driven improvement by focusing on measuring, understanding and managing variation among clinicians delivering care. In the past 20 years it has delivered more than 100 clinical improvement initiatives that it reports have improved outcomes and reduced costs. The introduction of an elective labour induction protocol, for example, has helped to reduce the rate of caesarean sections and saved around $50m each year in Utah.

Lee Memorial Health System in Fort Myers, Florida, reported savings of $5.3m by adopting lean principles across the organisation. It also recorded improvements in unscheduled admission rates and overall patient flow.

The Mayo Clinic in the US used variability methodology to analyse surgeries over a three-month period and construct models of the resources used for scheduled and unscheduled cases. Guidelines were implemented to smooth the daily schedule and minimise variation. It reported a range of improvements in its productivity and resource use. Overtime staffing decreased by 27%, the number of elective scheduled same-day changes decreased by 70% and net operating income improved by 38%.

Seattle Children’s Hospital used Integrated Facility Design, an adaptation of the Toyota 3P process, to design a surgery centre with reduced variation and improved cost-effectiveness. Using this approach resulted in completion 3.5 months ahead of schedule, with estimated savings of $30m in project costs and improved patient throughput.

Virginia Mason Medical Center in Seattle is a leading example of the application of the Toyota Production System principles, or ‘lean’, to improve flow. After adopting lean as its management system in 2002, the center has reported appreciable and sustained improvements in clinical outcomes, safety, patient satisfaction, process indicators, staff engagement and costs. The prevalence of hospital acquired pressure ulcers, for example, fell from 5% to 1.7% between 2007 and 2012. Liability claims also dropped by over half and the centre achieved positive financial margins each year through efficiency savings, after previously losing money in consecutive years. It has also been named as ‘top hospital of the decade’ by Leapfrog and has been ranked in the top 1% of US hospitals for quality and efficiency.

Its success has been underpinned by an improvement approach – the Virginia Mason Production System – which seeks to standardise processes where possible, streamlining repetitive aspects of care to reduce waste and free up staff time with patients. All 5,500 staff at the centre are trained in the approach. The emphasis is on creating a culture of learning throughout the organisation, which can be applied successfully to drive continuous improvement.,,,

Box 3: UK examples of national flow improvement programmes

The Scottish government has developed two programmes that focus on flow, primarily within acute systems:

  • Launched in 2013, the Whole System Patient Flow programme, which has been delivered in collaboration with the Institute for Healthcare Optimization (IHO), contains a number of acute-focused workstreams. The programme draws upon IHO Variability Methodology® and ‘classic queuing theory’ to describe and achieve ‘optimal flow’. Four territorial health boards have well-established projects; a further six (of a total of 14) have completed a Scottish Patient Flow Assessment and are starting their own pilot projects.
  • The Unscheduled Care programme, launched in May 2015, is focused on achieving the four-hour emergency Access Standard across Scotland through six essential actions. The programme has adopted a collaborative approach underpinned by measurement for improvement and other quality improvement approaches. The building blocks of the programme involve six high-level themes, which are managed both individually and collectively. NHS Scotland has reported that this whole system approach has helped to improve flow for over 40,000 people in the last year: long waits of 8 hours and 12 hours have improved by 92% and 100% respectively.

The Welsh 1000 Lives Patient Flow programme was launched in June 2013 and ran until August 2015. It aimed to develop organisational capability and improve the effectiveness, efficacy and efficiency of the system for managing the care and flow of patients from the point of unscheduled entry, through diagnosis and treatment to discharge. Participants in the national roll-out of the programme included the Welsh Ambulance Services Trust (WAST) and the six Local Welsh Health Boards (LHBs) with general hospitals that admit emergency patients.

The programme was publicised as a ‘Breakthrough Collaborative’, modelled on the work of the Institute for Healthcare Improvement, and its conceptualisation and design was informed by the Health Foundation’s Flow Cost Quality programme. It had three main components: national learning events, a computer-based training course, and local workshops at each of the participating sites. The programme has achieved some improvements to patient flow within certain pathways and has led to the sustained use by some LHBs of the Big Room process and the A3 structured problem-solving process (see Case study 1 for details of these processes). However, it has proven more difficult – certainly within the relatively limited time and resources allocated to the programme – to deliver wide-scale improvements across participating sites.

Summary

This section has highlighted the importance of improving flow across whole health and social care systems. It has summarised what is known about how to develop a deeper understanding of flows and offered international and UK-based examples of what can be achieved through the application of flow improvement methods. While these methods have significant potential to help address the challenges faced by the system, realising this will require long-term commitment and investment.

Section 2 proposes an organising framework to guide such efforts. It identifies four levels of action that need to be woven into a coherent strategy to realise the potential of flow improvement.


See glossary for definition of these terms, as well as others used in this report.

§ Eugene Litvak is at the Institute for Healthcare Optimization, which is supporting the Scottish Whole System Patient Flow programme.

‘upstream’ refers to services encountered early in a care journey (eg primary care), while downstream refers to those encountered at a later stage (eg secondary or tertiary care). See glossary for more details.

Previous Next