An understanding of the past helps us to prepare for the future. In this, the REAL Centre's first report, we look back at the care and treatment provided by the NHS in England over the past two decades – as measured by health care activity. Measures of activity – what the NHS does – form the basis of much NHS planning. Together with measures of quality, these are used as proxies for understanding the contribution the NHS makes to population health. Understanding the drivers of this activity is therefore crucial to service and resource planning.

In this report, we provide a framework for understanding the drivers of health care activity. We describe how supply and demand side factors interact, and how policy can influence the care the NHS provides. We look at the overall trends in activity and how specific services have changed. Where possible, we estimate the proportion of any change that can be explained by four demographic factors: population size, age, gender, and proximity to death. We then explore what drives the remaining change, unexplained by demography.

Although we focus on England, many of the lessons may be relevant to the other countries of the UK. We take an in-depth look at emergency and planned hospital care, mental health, community and primary care, but do not look in detail at areas such as high cost drugs or highly specialised services. Social care, too, is beyond the scope of this initial report but will be the focus of subsequent analysis by the REAL Centre.

NHS activity since 2000

Between 2000/01 and 2017/18, the amount of NHS-funded care in England, either delivered by NHS hospitals or other providers, more than doubled – increasing by 114%, an annual average of 4.6% a year. We estimate that hospital activity increased 4.7% per year, with planned care (6.1%) growing by almost 2.5 times the rate of emergency care (2.6%).

Overall, our analysis suggests that the amount of care provided by the NHS since 2000/01 would have needed to grow by less than 1% a year to meet the demand pressures arising from demographic changes. This means that more than three-quarters of the growth in NHS care since 2000/01 is the result of other factors relating to demand, supply, political priorities and health care policies.

While overall health care activity has grown substantially, this growth has not been shared equally between services. Consultations in general practice have grown just 0.7% per year, while planned procedures increased 9.6%. The different growth rates seen between services, combined with changes to the cost of different activities, have produced a major shift in the composition of spend towards hospital-based care and away from other areas.

The expansion in activity means the population receives significantly more health care today, which has contributed to lower mortality rates for a range of conditions and, up until 2010, increasing life expectancy. However, improvements and increases in care have not been uniform over the past two decades. Between 2000/01 and 2010/11 activity rose by 5.3% a year, quality of care improved, and life expectancy grew by nearly 3 months a year for women and almost 4 months a year for men. As funding slowed after 2010 so too did activity. Between 2010/11 and 2017/18 activity grew by 3.6% a year, quality gains slowed and life expectancy has stalled overall and for some groups declined (notably for women younger than 50 in deprived areas). But it is not just life expectancy that has changed. Over the past two decades the amount of life spent living with long-term health conditions has increased and more people are living for prolonged periods with multiple long-term health problems.

New technologies, changes in prices, and the increasing availability of information about treatments have increased clinician and public expectations about what the NHS can and should provide. Medical advances have meant more treatments can be provided without the need for patients to stay overnight in hospital and more can be done in outpatient departments or as day cases. This means that while the number of people being treated in NHS hospitals has increased, the number of bed days for unplanned admissions fell 1.1% per year.

New services have been established. For example, NHS 111 and Urgent Treatment Centres (UTCs) were introduced with the aim of moderating demand in other areas of emergency care. Both have seen substantial growth. Changes to the workforce have also determined service use. Slow growth in consultations in general practice, for example, is reflected in slow growth in the number of GPs. Meanwhile, an expansion in hospital consultants has supported the rapid increase in outpatient appointments and elective procedures.

Underpinning many of these drivers, particularly around the supply of health care, are political and policy choices. The level of funding the NHS receives is a political choice. Unsurprisingly funding growth is related to activity growth. In the first decade of the century, health spending in England grew by 6.0% a year; in the second it grew by 1.3%.

Government’s shifting priorities for the NHS over the past two decades have also shaped the pattern of growth. In the early 2000s there was a focus on reducing waiting times and improving outcomes particularly for the ‘big killers’ of cancer, stroke and cardiovascular disease. A mix of additional NHS funding and a range of policies from financial incentives, to the maximum 18-week wait from referral-to-treatment target help explain the rapid growth in elective procedures. In the 2010s, in the face of lower funding growth the focus shifted to efficiency gains, contributing to the slowdown in activity growth.

Across the past two decades, much of the policy narrative was about shifting the focus of care away from acute services towards community and primary care to prevent avoidable admissions and manage care more proactively. There is little sign of this policy goal being achieved.

Lessons for the future

Looking at the overall changes in the volume and pattern of care provided over the past two decades, there has been remarkable success in some areas in expanding treatment to benefit the health of the population. Because these changes happen gradually, they are not always recognised, and the NHS can instead be portrayed as rigid or slow at adopting innovation. In fact, changes have been substantial.

However, it is also clear that some changes have been despite – rather than because of – the stated policy goals of successive governments and system leaders. Policy intent has not always been matched by a coherent programme of policy and investment decisions needed to translate rhetoric into reality – the misalignment of workforce plans and service delivery plans is an obvious and stark example.

Our analysis leads us to suggest six lessons for the future.

1. In the long term, what the health care system does is largely determined by choices. Clear population health goals are needed to help steer these choices. The contribution of demographic factors to growth is relatively small. Other drivers of changes are not well understood, but are within the influence of policymakers if we consider the full range of policy instruments available, especially over a longer time horizon. We need clearly stated population health goals against which to judge the decisions taken.
2. A better understanding of the contribution of different services to population health goals, and what this means for the future composition of NHS services, is required. Activity growth has not been spread equally across services. For example, we have seen a rapid growth in hospital services compared to community-based services. We need to better understand the contribution of different services to improving healthy life expectancy and reducing inequalities in health across the country and between population subgroups.
3. Systematic national projections of future trends in morbidity are needed to inform national and local NHS planning. If changes to the composition of NHS services are to meet the future health needs of the population, we need good projections of those needs. As the types of services needed by different groups of patients differ – from cancer to mental health – projections need to cover different forms of morbidity, and, increasingly, multi-morbidity.
4. There is a need for more sophisticated models of future activity and health outcomes, incorporating other important drivers of growth, including policy choices, as well as demographic factors. The debates about future pressures on the NHS often pay too little attention to supply-side factors, such as workforce or new technologies, in shaping the care provided. We do not have a good enough understanding of the role that drivers beyond demography play in activity growth, such as the changing composition of the workforce. Consequently, they are rarely included in future projections of activity.
5. Technology plays a major role in shaping both demand (a new treatment makes it possible to address an unmet need) and supply (drug therapies replacing complex surgery). The UK is a world leader in research and the NHS makes a vital contribution to supporting clinical trials. Innovation is inherently unpredictable and complex, but there are also several choices in what gets developed and adopted. It is vital that we better understand the forces shaping those decisions and the impacts this has on the health system.
6. Shifts in activity take time and there is a need to be realistic about what can be achieved. For example, reducing avoidable emergency admissions has been a long-standing policy goal and, since 2010, has been an aim of the various national integrated care programmes. However, although there has been some evidence of modest success, the ambitions for these programmes exceeded what has been achieved. We need a better understanding of the wider drivers of activity in order to set realistic ambitions and timeframes.

Trends over the past 20 years show how widely investment and growth in services has varied, with hospital care winning out over primary and community care. Numerous factors have driven these changes, but the changing size and age structure of the population plays a relatively minor role. New technologies and drugs, as well as political, policy and planning choices, are among the most potent influences shaping investment. There has been no systematic analysis of how these factors have shaped care over the past two decades. Nor any serious overarching attempt to match the substantial investment we as a society make in the NHS with the health needs of the population.

We cannot know the future and no amount of research and modelling can eliminate this inherent uncertainty – coronavirus (COVID-19) has vividly shown this. But managing uncertainty and anticipating change is the job of policymakers and planners. Health systems are built over many years: hospitals take time to build, staff need years of training, and knowledge and relationships are developed over time. To make the best decisions and to sustain the NHS, what is needed is a strategy for investment based on evidence, analysis and modelling. This needs to be adaptable regionally to allow for differences in population health and health care systems. Above all, this needs to be guided by a ruthless focus on the founding principle of the NHS – to improve health and wellbeing and ensure that care is based on need, not the ability to pay.

---

Key points

• The amount of NHS funded care in England increased by 114% between 2000/01 and 20017/18 – a more than doubling in less than 2 decades.
• Growth in activity partly reflects demographics. Demand grew steadily as the population grew, aged, and the prevalence of co-morbidities increased.
• The growth also reflects policy decisions. Government increased funding for the NHS sharply during the 2000s (6.0% per year).
• Funding was primarily used to increase capacity, with more staff recruited and investment in new equipment and buildings – although bed capacity decreased as new technologies meant less activity required a hospital stay.
• Funding was accompanied by other reforms to increase activity – targets to reduce waiting times, activity-based payments and increasing use of the private sector – which mostly focused on hospital-based services.
• After the financial crisis of 2008, funding growth fell back. Slower growth in the supply of health care led to efforts to moderate activity growth and increase efficiency.

NHS activity has more than doubled

The Office for National Statistics (ONS) calculate that between 2000/01 and 2017/18, the overall amount of health care provided to patients by the NHS in England more than doubled – a 114% increase, or 4.6% per year (Figure 2). This is based on ONS’s aggregate measure of health care output, which combines two things: the volume of activity weighted by the cost. Cost is used to weight different activities, recognising these differ in scale and complexity – for example, an emergency admission to hospital is normally more expensive than a GP consultation.

Recognising that the quality of care also matters, the ONS further adjusts for changes in survival rates, outcomes and responsiveness to patients. Quality adjusted (QA) care provided by the NHS increased by 131% between 2000/01 and 2017/18. Growth in output was higher in the first decade, at 5.3% per year between 2000/01 and 2010/11 (5.8% QA); slowing to 3.6% per year between 2010/11 and 2017/18 (3.9% QA).

Figure 2: Health care output and quality adjusted output (index, 2000/01 = 100); annual output growth (%)

Source: ONS

Growth in output varies by type of activity

The ONS reports output for four broad service areas (more detailed reporting is complicated by changes in data recording over time). Between 2000/01 and 2017/18 there was growth in the quantity of output across each service area. The highest growth was in non-NHS provision – care funded by the NHS but provided by the independent sector, voluntary sector or local authorities – and GP prescribing, with less rapid growth in hospital and community health services and family health services (FHS), which includes primary care.

However, hospital and community health services accounts for the majority of NHS expenditure and over this period increased its share by 3.1 percentage points (pps). The contribution to overall output is calculated by weighting the quantity output growth for each sector by the respective expenditure share. Table 1 shows the annual average growth in quantity output alongside expenditure shares in 2000/01 and in 2017/18 and the contribution to overall output growth.

Table 1: Quantity output growth, expenditure share and contribution to overall output growth by sector (2000/01 to 2017/18)

	Quantity output growth	Expenditure share			Contribution to overall output growth
	Annual average growth	2000/01	2017/18	Change (pps)	Annual average (pps)
Hospitals and Community Health Services	3.8%	60.1%	63.2%	3.1	2.3
Family Health Services	2.5%	16.3%	15.6%	-0.7	0.4
GP Prescribing	6.5%	19.0%	8.5%	-10.5	1.0
Non-NHS care	11.3%	4.6%	12.7%	8.1	0.9
Whole NHS					4.6%

Source: Health Foundation analysis of ONS data, pps=percentage points.

Output is determined by both activity and cost

The NHS provides a wide variety of services with different costs. Table 2 sets out activity and costs for a selection of services provided by NHS Trusts and Foundation Trusts in 2017/18 (note, this does not include care funded by the NHS but delivered by non-NHS providers). For example, there were 85 million community health service contacts in 2017/18 at an average cost of £62 per contact. By contrast, there were only 10 million non-elective (unplanned) inpatient stays, but at an average cost of £1,726.

Table 2: Activity, unit cost and total spend by health service (delivered by NHS) (2017/18)

Service	Volume (millions)	Unit cost (£)	Total spend (£m)
All A&E	20m	£160	£3,198m
Outpatient appointments	75m	£125	£9,316m
Diagnostic imaging	10m	£90	£941m
Planned procedures	21m	£556	£11,544m
Non-elective inpatient	10m	£1,726	£17,890m
Consultations in general practice	307m	£27	£8,182m
Community prescribing	1,106m	£8	£9,095m
Community health services	85m	£62	£5,252m

Source: NHS reference costs, 2017/18, Consultations in general practice estimated from CPRD, NHS England and PSSRU; Community prescribing from Prescription cost analysis system.

As Table 1 showed, there was strong quantity output growth in some sectors, but they nonetheless fell as a share of expenditure. This is because there have been shifts within sectors both in terms of activity and the cost of delivering care (unit costs). Figure 3 sets out the annual average change in unit cost (in real terms) between 2007/08 and 2017/18. This shows that there is considerable variation in unit cost growth. Diagnostic imaging and community prescribing saw unit costs fall by 3.3% and 4.6% per year, compared to an increase of 5.3% for A&E attendances. A fall in unit cost does not necessarily mean doing the same thing has become less expensive, it also reflects shifts in the composition of activities.

Figure 3: Annual average change in unit costs of NHS delivered activity for selected services, 2007/08 to 2017/18

Source: NHS Reference Costs, 2007/08 and 2017/18; unit costs adjusted using GDP deflator; Consultations in general practice estimated from CPRD, NHS England and PSSRU; Prescription cost analysis system.

Overall health care output per person has increased significantly 

Taken together, the average person in England receives twice as much health care than two decades ago. Health care output per person grew 3.9% per year even before allowing for the increased quality of care received (Figure 4). Output growth per person was higher between 2000/01 and 2010/11 (4.6%) than 2010/11 and 2017/18 (2.8%).

Figure 4: Annual average growth per person: funding, input and output

Source: PESA, Table 9.11; ONS; funding in 2018/19 terms, using the GDP deflator provided by the OBR.

We now look at what has happened over the past two decades in terms funding, supply, demand and policy to explore the broad trends described.

Activity growth has been supported by growth in funding, workforce and other inputs

Funding

At the turn of the century there were concerns that the UK was falling behind other nations on health outcomes and spending. In response, the government pledged to increase spending on the health service and commissioned the Wanless review (2002), which set out a plan for increased investment in the NHS under a range of scenarios.,

Real terms (adjusting for inflation) health spending in England grew from £62.5bn in 2000/01 to £122.8bn in 2017/18 (a 96% increase, or an annual average of 4.0% a year). However, there were two very different periods of growth (Figure 5). Over the first decade health spending increased by an annual average of 6.0% – this was in keeping with what the Wanless review termed a catch-up period of growth. Although some reduction in funding growth was expected, the financial crisis led to a much sharper fall than anticipated. After 2009/10, spending grew at an annual average of just 1.3% in real terms.

Figure 5: Annual average growth in real health spending in England: total, current and capital

Source: PESA, Table 9.11; funding in 2018/19 terms, using the GDP deflator provided by the OBR.

Expansion, specialisation and then moderation

NHS funding is used to purchase inputs, including labour, goods and services, and capital. The overall volume of inputs (which is the number of each item weighted by the item’s cost) grew by 88% between 2000/01 and 2017/18. Unsurprisingly, given the direct link between funding and inputs, input growth was significantly higher in the first decade (averaging 5.0% a year between 2000/01 and 2010/11) than the second (2.0% per year).

Additional funding was used to invest in new capacity and expand supply. The number of full-time equivalent (FTE) staff working in the NHS increased by 47% (we adjust for a discontinuity in the data from 2010 when estimating growth rates), or 2.0% per year, from 874,000 in 2000 to 1,247,000 in 2019 (Figure 6). Early on, new employment contracts were introduced which increased pay, the most notable of which was a new grading and pay system for NHS staff implemented in 2004. This increased input prices but was considered necessary to attract and retain staff. Migration also played a role as the NHS became increasingly reliant on international staff. Changes to immigration rules during the period led to an increasing proportion of staff from the European economic area, although the decision to leave the European Union has seen this trend reverse.

NHS capital stock was modernised, with new hospitals and equipment such as diagnostic scanners. Funding also helped improve the IT infrastructure, with the introduction of a new patient records system.

Figure 6: Number of FTEs in NHS hospital, community health services and general practice staff

Source: NHS Digital; break in series reflects a discontinuity in the data

There were also major shifts in the mix of inputs in this period. The workforce became increasingly specialised, with growth in hospital doctors (4.6% per year) outpacing nurses (2.3%) and GPs (2.1%) (Figure 7). In order to expand capacity still further, policymakers began to make greater use of the independent sector (private providers, but also charities and local authorities). Independent sector provision – recorded within goods and services – is the main reason goods and services increased relative to labour in the late 2000s (Figure 8).

Another major shift was a reduction in the number of overnight beds. The number of acute beds grew briefly in the early 2000s, but overall beds fell sharply between 2000/01 and 2017/18 from 186,000 to 129,230. For some conditions, such as mental health disorders and learning disabilities, this reflected a policy push to treat more people in the community. Technological change, including precise anaesthetics and fibre-optics, however, reduced the need for beds more broadly, as more patients could be treated as day cases and those that did need beds could be discharged more quickly. Reducing acute beds became a key local planning objective in many health systems and a focus for annual planning.

Figure 7: Annual average growth in inputs, NHS staff by role and beds

Source: NHS Digital

Figure 8: Share of expenditure on inputs, by type

Source: ONS

However, from 2010/11 slower funding growth brought pressure on NHS inputs. Workforce growth moderated to 1.1% a year. Meanwhile, the capital budget came under pressure from 2010 as day-to-day needs were prioritised. The capital budget grew 10.1% per year between 2000/01 and 2010/11, but growth fell to 1.5% per year thereafter as day-to-day spending was prioritised.

By 2018/19, the NHS again faced significant staff shortages. Between 2010/11 and 2018/19, the NHS had become more reliant on costly agency and bank staff. Reduced growth in capital spending resulted in a rising maintenance backlog, with staff using faulty or out-of-date equipment. England still has fewer items such as Magnetic Resonance Imaging (MRI) scanners per person than other countries, despite growth in numbers between 2000/01 to 2010/11. The continued reductions in bed numbers resulted in dangerously high occupancy rates that left the services unable to deal with spikes in demand, particularly during winter months.

Demand has risen continually

Although funding and input growth slowed during the second decade of this century, factors driving demand did not abate. The population continued to grow and age, with health needs growing more complex.

In total, the population of England grew from 49.2 million in 2000 to 56.3 million in 2019, a 13.7% increase. The male population grew faster (15.1%) than the female population (12.4%). The population aged 65 and older increased by 33% (Figure 9), rising from 15.8% of the population in 2000 to 18.4% in 2019. While overall population growth was steady, the growth in the population aged 65 and older accelerated after 2011.

Between 2000 and 2019, net migration accounted for 60% of population growth, with the remaining 40% being the natural increase from births and deaths. The death rate in England fell significantly in the 2000s and, despite rising since 2011/12, remains lower than in 2000/01; the birth rate increased up to 2010/11 but has fallen since (Figure 10).

Figure 9: England population and population aged 65 and older (index, 2000/01 = 100)

Source: ONS

Figure 10: Birth and death rates

Source: ONS

The death rate fell mainly because of reductions in age-specific mortality. These can be seen in estimates of life expectancy which, in England, increased between 2000/02 and 2010/12 by 2.3 years for women (0.3% per year) and 3.1 years for men (0.4%). Unlike gains in life expectancy earlier in the 20th century, which can broadly be attributed to decreases in mortality among younger people, recent increases reflect improvements in the health and care of older people, such as better treatment for heart disease. However, increases in life expectancy have stalled since 2010/12. Between 2010/12 and 2016/18, just 0.3 years for women and 0.5 years for men were added; growth of 0.1% each year. There is no single reason for this marked slowdown, although research reveals a rising number of avoidable deaths among those younger than 50, and a widening gap in life expectancy between the richest and poorest and between the north and south of England.

Rising life expectancy is a success of health policy. However, healthy life expectancy (the number of years spent in good health) has not kept pace. Increased survivorship means more people can expect to live longer in poor health, creating additional demand for health care. The burden of disease (morbidity), as measured by the total number of years lived in disability (YLDs), increased 19% between 2000 and 2017, from around 6.5 million to 7.8 million. The age-standardised rate of morbidity increased by 2.4% over the same period, meaning the burden of morbidity is mainly driven by a larger group of people reaching old age.

Not only has the prevalence of morbidity increased, so too has the complexity. In particular, there has been a stark increase in the proportion of the population living with multiple long-term conditions. In 2000, most (58%) people aged 70 or older were living without a diagnosed long-term condition. By 2018, just 32% of those aged 70 or older had no diagnosed long-term condition, 39% had more than one, and almost one in ten (9%) had four or more (Figure 11). Multi-morbidity poses challenges to traditional ways of delivering care, which are centred around single diseases.

Figure 11: Share (%) of population aged 70+ by number of long-term conditions

Source: CPRD

A number of risk factors are well known to influence the prevalence of different diseases. Over the past two decades, there has been progress in reducing smoking rates and in reducing the prevalence of high blood pressure, but there has been an increase in the obesity rate and little change in levels of alcohol consumption.,

Policy focused on improving access, then consolidation through efficiency gains

Funding growth and market reforms to improve access

In 2000, the NHS was less well resourced than health systems in comparable European countries. The supply of health care did not match demand, with long waits for routine surgeries and low public satisfaction with the health service.46 Influenced by ideas such as public sector failure – the equivalent of market failure for public sector markets – policymakers sought to take an interventionist approach to encourage services to improve health outcomes, alongside increased funding. The main reforms to achieve this were:

• Funding – with pressure mounting to boost NHS activity, government pledged to increase health spending to the European Union average.
• Targets – an extensive set of performance targets was introduced. This included targets to reduce mortality rates from heart disease, suicide and cancer, but increasingly focused on process measures, including waiting times targets for A&E, planned hospital care and cancer treatment.
• National guidance and support – for example, improvements in health outcomes were supported by National Service Frameworks, and 10-year plans to define standards of care in major areas of medicine.
• Regulation – was introduced to promote quality. This included the introduction of NICE, tasked with evaluating treatments for use in the NHS based on cost-effectiveness, and the Commission for Health Improvement (now the Care Quality Commission).
• Reorganisation – in order to create local incentives and accountability, the government continued to organise the health system along the lines of an internal market, separating the provision of hospital services (providers) from health authority planning functions (commissioners).
• Payment – payment systems were developed for secondary care that expressly incentivised activity. Previously payment took the form of a lump sum for providing care to a population, but payment-by-results (PbR) was paid on a case basis. PbR began with elective care but expanded in the 2000s to cover most hospital activity.
• Choice – Patients were progressively given more opportunities to choose their provider for elective treatment, with the expectation that increased competition would improve efficiency. Increasingly this included the independent sector (which includes private sector providers), with independent sector treatment centres (ISTCs) introduced in 2003 and involvement expanded thereafter.

These reforms were associated with both lower waiting times and improved health outcomes; public satisfaction with health services increased.

However, towards the end of the first decade of the 2000s the emphasis on increasing activity began to abate. Indeed, there was growing concern that a forceful implementation of targets – dubbed ‘targets and terror’ – had distorted clinical priorities, created perverse incentives and scope for manipulating the system. Meanwhile, there was recognition that the pressing issue facing the NHS was now the care of a growing number of people with long-term conditions. Reforms focused on hospital care had come at the expense of other parts of the system, including primary care, community services and mental health – all of which are important for supporting those with long-term conditions. Moreover, concerns emerged that efforts to improve prevention had fallen short of the more ambitious targets set in the earlier Wanless review. The white paper, High quality care for all, signalled the government’s desire to focus the next stage of health care reform in England as much on quality as responsiveness and efficiency, and on integrated care for people with long-term conditions.

The Nicholson challenge and drive to make savings

Following the 2008 recession and resultant increase in the public sector deficit, the UK government made efforts to reduce the growth in public spending. For the NHS, the period would be dominated by the need to achieve large efficiency savings to keep pace with demand. The so-called ‘Nicholson challenge’ was for the NHS to achieve 4% increases in productivity each year with no real-terms increase in budgets.

Government imposed several measures to contain costs. Perhaps the most visible came in relation to workforce. Government introduced a public sector pay freeze (2011–2013) and subsequently capped wage growth. These policies eroded real wages (adjusted for inflation), which took a toll on morale and retention. Meanwhile, an efficiency factor was used to reduce payments under PbR. Providers were still performance managed – the coalition government had not followed through on a commitment to replace most targets introduced in the 2000s with a broader set of outcomes measures, highlighting the political challenge of removing targets once installed. Faced with lower payments for delivering care to patients, NHS providers increasingly fell into deficit.

Initially, the role of competition was re-emphasised alongside the reorganisation of the NHS initiated as part of the Health and Social Care Act 2012. However, the focus on competition would prove short lived as it became clear this approach was out of step with the needs of the increasing numbers of patients with long-term conditions. The emphasis of policy instead turned to activity moderation, efficiency and collaboration in the face of rising demand and pressure on performance against targets.

NHS England assumed a greater role in boosting integrated care, partnership working (across the NHS and with social care) and effectiveness. Initiatives such as multidisciplinary teams and the Better Care Fund sought to encourage the integration of care and improve collaboration to make use of resources more efficient. In addition, policymakers recognised that an onus on competition had fractured the provider landscape and made collaboration and service improvement more challenging, just when the rise in multi-morbidity meant services had to become more, not less, integrated. The policy document, the NHS Five year forward view (2014), set out a new focus for the NHS, including greater integration of care; investment in primary care; and an upgrade in prevention and public health. Many of these themes were repeated in the NHS Long term plan (2019).

---

Key points

• Growth has varied hugely across NHS services.
• Demographic factors (population size, composition by age and gender and proximity to death) explain only around a quarter of the overall growth seen – around 1% a year.
• Generally, care in hospital has seen greater growth than other settings. For example, elective procedures have grown by 9.6% each year; consultations in general practice by only 0.7%.
• In part, this is because funding, capacity increases and high-profile targets were focussed on hospital services.
• However, the amount of time patients spend in a bed in a care facility declined for all services. This is linked to technological change and treatment norms.

Introduction

In this chapter we look at trends and patterns in levels of health care activity in England by service. We analyse how these have been shaped by changes in supply, demand and policy over the past two decades.

Our approach to analysing activity

NHS care covers a broad mix of services. We look at services in five major areas of care: unplanned care, planned care, maternity admissions, primary and community care, and mental health care.

For each service, we estimate the change in activity that is explained by basic demographic factors which are relatively simple to model, and are often used to predict future service use. This recognises there are other important drivers of services use.

For all services we modelled the size of the population. Where possible, we have modelled the proportion of any change that can be explained by a combination of population size, age, gender, and proximity to death (full model). These estimates are based on ONS population and mortality data and the assumption that the amount of care needed by those at different ages, genders and proximities to death is unchanged. For example, that the average number of unplanned hospital admissions by women aged 75, who are not within 2 years of death, is constant over time.

The difference between observed activity growth and that explained by demographics, is unexplained growth. This means it is due to other factors. These are a combination of demand and supply side factors including changes in the age-specific prevalence of morbidity (demand) and changes in the availability of treatment (supply). Although it is difficult to model and quantify these factors, we attempt to identify which are most important for each service and the influence of policy on them.

Overall

Growth over the past two decades has differed widely by type of activity (Figure 12). The number of mental health admissions in acute hospitals has fallen 4.4% per year; there has been small growth in maternity admissions, consultations in general practice and community services; while elective procedures increased 9.6% per year.

Our analysis shows that demographic changes explain only a small part of the growth for most services. It explains a relatively high proportion of activity growth in a few services (major A&E attendances, consultations in general practice and community health services). This is not because these factors have a larger absolute impact in these areas, but because growth as a result of the other factors in our framework is much lower.

Figure 12: Annual growth (%) by service: projected by demographics, explained by other factors and observed

Source: Various, see Annex, *full model

Unplanned care

What is unplanned care?

Unplanned care is a necessary, highly visible and sometimes expensive point of health care provision. It includes:

• Emergency care for life threatening illnesses or accidents that require immediate, intensive and time sensitive treatment. Services include the 999 service and ambulances, and emergency departments at major hospitals.
• Urgent care for illnesses or injuries that require urgent attention but are not life threatening. Services include a phone consultation through NHS111, pharmacy advice, out-of-hours GP appointments, and urgent treatment centres (UTCs).

Table 3 shows the scale of these services in 2017/18 delivered by NHS Trusts and Foundation Trusts. This does not include care funded by the NHS but delivered by non-NHS providers.

Table 3: Unplanned care services in 2017/18

Service	Activity	Estimated unit cost	Estimated total cost
Triage services and ambulance
NHS 111 calls handled	14.6m	£12**	£179m
Calls to 999	10.2m	£7	£73m
Hear and treat or refer (help by phone)	806k	£37	£30m
Ambulance dispatches (treat on scene)	2.4m	£181	£443m
Ambulance conveyances to A&E	5.3m	£247	£1,306m
A&E attendances
Attendances at major A&Es	15.4m	£188	£2,892m
Attendance at minor A&Es	4.5m	£67	£306m
Emergency admissions
Emergency admissions to hospitals (non-electives)	9.0m	£1,698	£15,282m

Source: National schedule of NHS costs, NHS Trusts and Foundation Trusts

In 2017/18 the three largest areas of unplanned care spend (by NHS Trusts and Foundation Trusts) were emergency admissions to hospital (£15.3bn), attendances at major A&Es (£2.9bn) and ambulance transfers to hospital (£1.3bn). Together these accounted for 95% of total unplanned care spend.

Growth in unplanned care services

The chart below shows the annual growth by service, and the contribution of demographic factors from our model (Figure 13). There has been a rapid increase in activity for lower cost activities, mainly NHS 111 and calls to 999, with little explained by changes in population size. There has been lower growth in some of the services that have a higher cost, with a fall in unplanned bed days.

Figure 13: Annual growth for unplanned care services: projected by demographics, explained by other factors and observed

Source: Various, see Annex, *full model

What explains the growth?

All types of activity (with the exception of bed days) have grown more quickly than would be expected from changes in demographic factors alone. The reduction in bed days has been driven by a reduction in the number of unplanned admissions that lead to long stays in hospital.

NHS 111 (and its predecessor NHS Direct) and minor A&E departments, including UTCs, experienced high rates of growth. This is unsurprising as both NHS 111 and UTCs were set up during this period in an attempt to divert unnecessary attendances away from major A&E departments. Evidence suggests their success in doing so is mixed. An evaluation of a pilot NHS 111 identified the potential for supplier-induced demand, but there is stronger evidence that UTCs can reduce demand on A&E attendances.

Ambulance calls also increased rapidly, but the proportion of calls resulting in an ambulance transporting a patient to A&E fell from 89% in 2000/01 to 42% in 2018/19, leading to much slower growth in ambulance transports. Treatment at the scene was promoted by policy and planning as an alternative to transports. In part, this reflects the changing profile of patients – traditionally, ambulance services were designed to respond to life-threatening conditions, but by the early 2000s this accounted for only around 10% of cases, with increases in those seeking help with social care needs or long-term conditions. Changes in the skill mix and composition of staff were important in supporting more patients to be treated at the scene.

Attendances at major A&Es increased by 1.4% a year, 0.4pps above the growth expected from demographic factors alone. This is a smaller rate than many other acute services, with the lower growth driven in part by a reduction in attendances per head by younger males (Figure 14). This reflects sizeable declines in traffic and workplace accidents. By contrast, admissions increased significantly for those older than 75 (Figure 15). In 2014/15, 8.6% of A&E attendances were by people in their last 2 years of life, up from 7.2% in 2010/11.

Figure 14: Attendances at major A&E departments by age and gender, rate per 1,000 (2017/18)

Source: HES

Figure 15: Annual growth in rate of attendances at major A&E departments by age and gender (2007/08 to 2017/18)

Source: HES

Emergency admissions to hospital have grown by around 1.1pps a year more than expected from demographic factors alone. The growth in the rate of attendances per head was higher for women than men (Figure 16). Males aged 5–17 have seen very low rates of growth, as have men and women aged 65–74. Growth was higher among the very young (aged 0–4) and the elderly (85+) (Figure 17). The rate of emergency admissions for those within a year of death is 1,400 per 1,000 population, compared to 69.5 for those who survive beyond 2 years. 

Figure 16: Emergency hospital admissions by age and gender, rate per 1,000 (2017/18)

Source: HES

Figure 17: Annual growth in rate of emergency hospital admissions by age and gender (2000/01 to 2017/18)

Source: HES

One reason unplanned admissions may have grown faster than our modelling predicted based on demographic factors would be if admissions thresholds had reduced. However, research shows the opposite: admissions thresholds have actually increased. Instead, the increase seems to be the result of a greater severity of need, with a greater proportion of those attending therefore requiring an admission.

This may have been driven by changes in population health and the increase in the proportion of patients with multiple long-term conditions. By 2015/16, one in three emergency patients admitted for an overnight stay had five or more health conditions, up from one in ten in 2006/07. Research has shown that an unintended consequence of rising survival rates is a larger population of patients who are frail and at high risk of further hospital admissions. However, the pattern of emergency admission growth across age groups (eg the growth in admissions for 0–4 year olds) suggests other factors are at play too.

There have been various initiatives to reduce growth in emergency admissions. For example, attempts to make care more integrated – the Integrated Care Pilots, Pioneers and New Models of Care – were partly aimed at improving efficiency, often with a focus on reducing emergency admissions. Despite this, emergency admission growth in the second decade was 1.5% a year – 0.4 percentage points a year above that expected from population growth and ageing – compared to 2.2% a year in the first decade. 

While the policy aims for unplanned care have remained fairly consistent, policy aims in other parts of the system have limited their success. For example, the decision to reduce bed capacity has resulted in high occupancy rates, which have limited the ability of A&E departments to meet the 4 hour waiting time target intended to improve patient satisfaction and outcomes. Difficulties in discharging patients to appropriate care in the community have compounded this. Without a whole system approach, these trends will likely continue.

Planned care

What is planned care?

Planned care – largely hospital care that is not due to an accident or emergency – is arranged in advance and generally follows a referral from a GP. After a referral for treatment a patient has an outpatient appointment (here this excludes appointments for procedures, diagnostics and mental health). There the care team decide what course of treatment to follow and whether the patient requires a follow up appointment, a diagnostic test or an elective procedure. If it is a procedure, this can take the form of:

• an outpatient procedure (only at the hospital for the duration of the consultation) 
• a day case (given a hospital bed but not staying overnight), or
• an inpatient procedure (an overnight hospital stay).  

Table 4 shows the scale of these services in 2017/18 delivered by NHS Trusts and Foundation Trusts. This does not include care funded by the NHS but delivered by non-NHS providers.

Table 4: Costs of planned care, 2017/18 

Service	Activity	Estimated unit cost	Estimated total cost
Outpatient appointments§§	74.5m	£125	£9,305m
Diagnostics
Diagnostic imaging	10.5m	£90	£941m
Elective procedures
Outpatient procedures	13.2m	£140	£1,841m
Regular attendance	0.3m	£327	£93m
Day cases	5.9m	£742	£4,404m
Electives¶¶	1.3m	£4,023	£5,323m

Source: National schedule of NHS costs, NHS Trusts and Foundation Trusts

In 2017/18 the three largest areas of planned care spend (by NHS Trusts and Foundation Trusts) were outpatient appointments (£9.3bn), elective procedures (£5.3bn) and day cases (£4.4bn). Outpatient appointments were by far the largest area of activity.

Growth in planned care services

The chart below shows the annual growth by service, and the contribution of demographic factors from our model. This shows that there has been a rapid increase in all planned activities (with the exception of bed days). Growth has been most pronounced in elective procedures, followed by diagnostic tests, and outpatient appointments. 

Figure 18: Annual growth for planned care services: projected by demographics, explained by other factors and observed

Source: Various, see Annex; *full model

What explains the growth?

Much of the growth is not explained by the demographic factors in our model. This implies that the growth in planned care was driven by other factors, including funding, supply and policy.

Outpatient appointments have grown by 2.8pps per year above what is explained by demographic factors. Outpatient appointments increase with age, with those aged 75 and older accounting for nearly one in five appointments in 2017/18 (Figure 19). The rate of appointments for those close to death rose more quickly than for the rest of the population.

Figure 19: Outpatient appointments by age and gender, rate per 1,000 (2017/18)

Source: HES

Figure 20: Annual growth in rate of outpatient appointments by age and gender (2003/04 to 2017/18)

Source: HES

Some patients require a follow up appointment, and many require several. Growth in first appointments (4.0% per year) was marginally higher than follow ups (3.8%) over the whole period, meaning the rate of follow ups to first appointments delivered each year has fallen from 2.39 in 2000/01 to 2.33 in 2017/18.

From 2010/11 onwards, however, follow ups grew faster than first appointments. This has led to efforts to limit the number of follow up appointments, particularly those not deemed clinically necessary or desired by patients. Another proposal is to make greater use of digital appointments, something that may gain traction following COVID-19.

One of the most significant service developments of the past two decades has been an increase in diagnostics tests, driven by technological advances and investment in capital. From 2008/09 to 2018/19, we see a 75% increase in tests (5.8% per year). Imaging tests, including CT and MRI scans, account for most diagnostics. Our ability to diagnose increasing numbers of conditions has contributed to the rise in other areas of planned care.

Elective procedures (including outpatient, inpatient and day cases) have increased by 8.7pps each year above what is expected by demographic factors. This growth is larger than any other service area in our analysis. The proportion of elective procedures for those within 2 years of death has fallen steadily, from 14.0% of procedures in 2000/01, to 9.6% in 2014/15. A greater share of care, therefore, would seem to be for those with less acute needs.

Figure 21: Elective procedures by age and gender, rate per 1,000 (2017/18)

Source: HES

Figure 22: Annual growth in rate of elective procedures by age and gender (2000/01 to 2017/18)

Source: HES

Growth has been driven by outpatient procedures and, to a lesser extent, day cases. Day cases grew 6.5% per year between 2000/01 and 2017/18, while outpatient procedures grew 21.5% between 2003/04 (the first year of data) and 2017/18. Inpatient admissions declined 1.5% per year. As such, the proportion of elective procedures delivered in an outpatient setting increased from 15% in 2003/04 to 62% in 2017/18 (Figure 23).

Figure 23: Share (%) of elective procedures, by type

Source: NHS Digital, HES

This has been facilitated, in part, by technological changes that mean more patients can be treated without an overnight stay and discharged more quickly. An example is cataract surgery. Technological advances have allowed a shift from invasive surgery with a high complication rate and long recovery time, to a much quicker and safer surgery with shorter recovery. As the surgery was refined and became established, the threshold for procedures declined, leading to an increase in the number of surgeries performed. This and other similar changes meant that the proportion of cataract surgeries as inpatients fell from 15% in 2000 to 1% in 2017.

Overall, there has been a marked increase in the amount of planned care delivered. This has been supported and driven by increases in the supply of inputs, particularly an increase in the number of specialist doctors (8.2% growth per year across the period).

Over time, the health system has come under pressure not just to reduce mortality rates of diseases like cancer and heart disease, but also to improve quality of life through treatments such as hip or knee replacements. Increasing numbers of patients are eligible for these procedures, which are cost-effective and are being offered at lower levels of severity than in the past. This in turn changes societal expectations of the ‘right’ level of care. 

However, the amount of care provided is ultimately limited by capacity, underpinned by value judgements over what is feasible and desirable. This may be less obvious when funding and activity are growing, but recent slowdowns in funding growth have resulted in difficult decisions over care thresholds, with some local planners reducing access to procedures such as hip replacements.

Maternity admissions

What is maternity care?

One of the most common reasons for a hospital admission in England is having a baby. Table 5 sets out the elective and non-elective costs of maternity activity in 2017/18.

Table 5: Cost of maternity inpatient care, 2017/18 

Service	Activity	Estimated unit cost	Estimated total cost
Elective – Maternity	12,500	£3,157	£39.5m
Non-elective (long stay) – Maternity	411,000	£3,686	£1,586m
Non-elective (short stay) – Maternity	818,000	£944	£771.9m

Source: National schedule of NHS costs, NHS Trusts and Foundation Trusts.

Growth in maternity admissions

Maternity admissions are recorded distinctly from elective (planned) or emergency (unplanned) admissions. This category includes admissions of a pregnant or recently pregnant woman to a maternity ward, except when the intention is to terminate the pregnancy.

The chart below shows the annual growth in maternity admissions by number of admissions and bed days. Here, the contribution of demographics is accounted for differently from other service areas: we use births instead. This shows that there has been an increase in maternity admissions (0.5% per year), but less than expected based on the number of births. Bed days fell 1.1% per year.

Figure 24: Annual growth for maternity admissions: projected by demographics, explained by other factors and observed

Source: Various, see Annex.

What explains the growth?

Maternity admissions were 9.3% higher in 2017/18 than in 2000/01 (0.5% growth per year). Admissions grew 2.2% per year between 2000/01 and 2010/11, but fell 1.8% per year thereafter. This reflects the fertility rate, which increased in the in the 2000s before falling again (see chapter 3 on demand).

Our model shows a substantial reduction in the number of maternity bed days. This has been driven by: 

• a reduction in the ratio of admissions to births, from 1.76 admissions per birth in 2001/02 to 1.68 admissions per birth in 2017/18, and
• a reduction in the number of bed days per admission, from 3.40 bed days per birth in 2001/02 to 2.56 in 2017/18.

There has been a rise in zero day admissions (those not lasting a day), with the proportion increasing from 34% in 2000/01 to 38% in 2017/18. This reflects trends in other countries and may be driven by a combination of patient choice, greater flexibility and safety, or a drive to improve efficiency.

There has also been a substantial reduction in maternity admissions for women younger than 27 and growth in admissions for women aged older than 27, mirroring the changing age profile of women giving birth. The older age profile of mothers may contribute to more complicated and challenging births.

Primary care and community services

What are primary care and community services?

Primary care and community services are provided outside of hospital. Primary care is the most used part of the health care service and has a relatively low cost per consultation (Table 6). While a broader definition includes areas such as dentistry and community pharmacy, here we use primary care to mean general practice. General practice is where most health needs are identified, monitored, treated. A patient may be prescribed medication or, where necessary, referred to specialist services. General practice plays a crucial role in keeping people well, and in keeping them out of hospital.

Community services cover a range of teams, from district nurses and therapists to palliative care. They are delivered close to home, including in people’s homes. Teams also provide preventative interventions, such as smoking cessation services, often in partnership with local government or the third sector.

The table shows the estimated scale of primary care and community care services in 2017/18.

Table 6: Primary care activity and costs, 2017/18

Service	Activity	Estimated unit Cost	Estimated total cost
Consultations in general practice††† (2018)	307m	£27	£8,182m
Community prescribing	1,106m	£8.20	£9,083m
Community services	88.3m	£62	£5,480m

Source: NHS England, Appointments in General Practice; CHE analysis of PSSRU, Unit Costs of Health and Social Care; Prescription Cost Analysis - England, 2017 and 2018; NHS reference costs.

Growth in primary care and community services

Figure 25 shows the annual growth by service, adjusted for changes in the size of the population. This shows that there has been a rapid increase in community prescribing (prescribing outside of the hospital) but almost no growth per person in primary care or community health services. This is despite a strong primary care system being crucial to preventing illness and supporting people to manage their own conditions, and despite policy ambitions to move care from hospitals to the community.

Figure 25 Annual growth for primary care and community services: projected by demographics, explained by other factors and observed

Source: various, see Annex; *full model

What explains the growth?

Unlike other service areas, demographic changes do explain the majority of observed growth in consultations in general practice and community services. Contacts with community services in 2017/18, up from 75.7 million in 2004/05 (Figure 25), increased by 11% between 2004/05 and 2017/18 or 0.9% per year. Considering population growth, contacts per person grew just 1%, 0.1% per year. Therefore, nearly all the growth in community contacts can be explained by population growth.

The rate of consultations in general practice in 2018/19 (5.0 consultations per person) was the same as in 2000/01, though this hides small fluctuations over time. The lack of growth has in part been explained by a lack of growth in the workforce. Indeed, the number of full-time equivalent GPs per head of population has fallen since 2010. This has led to a shift in who delivers care. The proportion of consultations delivered by other staff, including nurses, increased from 28% in 2000/01 to 41% in 2018/19, with an associated fall in the proportion delivered by GPs.

The rate of consultations in general practice generally increases with age (Figure 26) and is higher for women than men. The rate of consultations among younger people (up to the age of 35) has fallen (Figure 27). By contrast, the rate of use among older people, especially those aged 75 and older, has increased significantly. This is likely explained by the rising number of patients with long-term conditions, many of whom have complex needs and require longer consultations.

Figure 26: Consultations in general practice by age and gender, rate per 1,000 (2017/18)

Source: CPRD

Figure 27: Annual growth in rate of consultations in general practice by age and gender (2000/01 to 2017/18)

Source: CPRD

In contrast to consultations in general practice and community services, little of the growth in community prescribing can be explained by population growth. The number of community prescriptions increased by 96% from 2000/01 to 2018/19, or 3.8% per year (Figure 25).

One notable factor in explaining the growth is cost. Figure 28 looks at the change in the number of prescriptions and also the change in total spend for the top ten drugs by volume as defined by the British National Formulary (BNF). This shows the increase in cardiovascular drugs and drugs for the central nervous system has come with lower spending, reflecting a fall in the cost per drug as items come off patent and are replaced by cheaper generic alternatives.

Figure 28: Change in number of prescriptions and total spend, 2008–2018

Source: NHS Digital, Prescription Cost Analysis; Costs adjusted using GDP deflator

Policies, such as Transforming Community Services, emphasised the need to shift care outside of hospital, both by bolstering primary care and community services.87 Specific initiatives aimed to improve the management of long-term conditions (eg the Quality Outcome Framework), access to care (eg situating GP surgeries in A&E departments) and workforce capacity (eg task shifting to practice nurses). However, these have not been backed sufficiently by funding increases or, in general, with enough political attention. Community services in particular lack the clear metrics (eg waiting time targets) that other services have, making it more difficult to monitor progress and understand where there may be gaps.

Mental health services

What are mental health services?

Mental health services support people with a range of mental health conditions, from anxiety and depression to severe psychoses. Mental health is both a diagnosis and a specialty. Initially, most mental health conditions are discussed with a GP, and those with a high level of need are referred onto specialist services. Alternatively, a patient may present at hospital and from there be treated and/or signposted to a mental health service. At hospital, those who require intensive care may be admitted; those with less intensive needs may be seen as an outpatient. Consequently, care can take place as a hospital inpatient, outpatient, or by specialist services in the community.

Table 7 shows the scale of mental health services in 2017/18 delivered by NHS Trusts and Foundation Trusts. Note, significantly more outpatient attendances are recorded by Hospital Episode Statistics (HES) data, which includes care funded by the NHS but delivered by non-NHS providers.

Table 7: Mental health activity and costs, 2017/18

Service	Activity	Estimated unit cost	Estimated total cost
Hospital mental health
MH elective and non elective finished consultant episodes (FCEs)****	135,000	£1,886	£254m
MH outpatient appointments††††	67,000	£167	£11.3m
Community mental health
Specialist MH care (initial assessment)	874,000	£307	£268m
Specialist MH care	236m	£18	£4,227m
Adult IAPT (initial assessment)	781,000	£121	£95m
Adult IAPT	849,000	£353	£300m
MH care in secure facilities (initial assessment)	35,000	£870	£30m
MH care in secure facilities	1.4m	£521	£726m
Other specialist MH services	6.6m	£232	£1,535m

Source: NHS reference costs

As shown in Table 7, the majority of mental health spend (more than 90%) is delivered by specialist mental health services (community mental health). Community mental health covers a range of services, including specialist services for adults and older people, the Improving Access to Psychological Therapies (IAPT) programme and services such as drug and alcohol rehabilitation and Child and Adolescent Mental Health Services (CAMHS).

Growth in mental health services

The chart below shows the annual growth by service. This shows that there has been a rapid decrease in admissions and bed days, but strong growth in outpatient appointments. However, it is worth noting that outpatient growth stalled after 2010/11.

Figure 29: Annual growth for mental health services: projected by demographics, explained by other factors and observed

Source: Various, see Annex.

What explains the growth?

In community mental health, the biggest component of care, we are limited in our ability to explore and explain the growth in activity as a reclassification means data prior to 2011/12 are not comparable. Community mental health is also defined in relation to a mix of contacts and care clusters – here we present an aggregated index.

The amount of care increased in 2012/13, with 16% growth that year, but has since fallen (Figure 30). Activity was 9% higher in 2017/18 than in 2011/12, or 1.5% per year. The increase in population size explains much of this growth.

Figure 30: Community mental health services; activity and activity per person (index, 2011/12 = 100); annual growth in activity

Source: NHS Reference costs

There have been elements of growth in provision. Considering IAPT, a programme of talking therapies for people with low-moderate level depression and anxiety, the trend is very different. The number of people entering treatment per year grew 152% over this period, or 17% per year. Just 5% of this growth can be explained by changes in the size of the population, with the remaining 95% unexplained. This rapid growth is unsurprising as this programme was set up as a result of a major policy drive to increase the availability of psychological therapies. Decreased stigma around mental health and increased willingness to seek treatment have likely further contributed to the growth.

Considering care in acute hospitals, mental health admissions and bed days have both seen a decline. The number of mental health admissions declined 53% between 2000/01 and 2017/18, a steady 4.4% per year. By contrast, our model estimated that appointments would increase 0.4% per year. This is due to a drive to treat patients in the community or outpatient settings, which led to a marked reduction in mental health beds. This is reflected by weak growth in FTE hospital doctors and a reduction in mental health nurses over the period. However, there is some evidence the shift away from inpatient care may have gone too far.

Mental health outpatient treatments have increased by 116% between 2003/04 and 2017/18 or 6% per year. Notably, growth was rapid in the first decade, increasing 12% per year between 2003/04 and 2010/11, but was flat thereafter. Unlike mental health admissions, then, the other factors in our framework beyond demographic factors contributed to an increase in appointments and amounted to 4.7% of growth per year. The rate of outpatient appointments has increased most for young females (aged 5–17). This may reflect a rise in eating disorders among this group.

---