4. What do these findings mean?
The pressures on acute hospitals in winter come from many sources and are often a symptom of wider issues in the local health and social care system.
Over the period reviewed in this report, the percentage of people waiting more than four hours to be admitted, transferred or discharged in an A&E department is greater in the winter months (November to March). This has been worsening over the last five years and the latest 2015–16 data shows no change in this trend. Nationally, the four-hour A&E target has not been met for the last 27 months, and 90 per cent of patients spent less than four hours in A&E in October 2015 – five percentage points below the national target of 95 per cent.
Not everyone in A&E is waiting longer. The fact that there is no corresponding change in the median length of time patients have to wait in A&E means that during the winter months some patients are waiting ‘extreme’ lengths of time.
The reasons for this cannot simply be put down to more patients attending A&E. First, previous work from the QualityWatch programme has showed that although greater crowding in A&E is associated with longer waiting times and increased chances of a four-hour target breach occurring, it is not the only contributory factor to these outcomes. Half of those breaches occurred when an A&E unit was running at greater than 125 per cent of its usual occupancy, but over a quarter of breaches happened at a time when units were less busy than expected (Blunt, 2014). Second, as noted in this report, there is in fact a decrease in the number of attendances to A&E during the winter periods.
However, in winter there is an increase in the proportion of older people attending and in the overall proportion of people who are admitted to hospital as an emergency. Older patients and those waiting for admission tend to wait in A&E longer than other patients (Blunt, 2014). This increases the chances of the four-hour target being breached during the winter period, although the increased number of older people attending A&E appear only to account for around 11 per cent of the observed decline in performance against the four-hour target (Blunt, 2014).
Winter weather has a direct effect on need for care by the population; for example the incidence of heart attack, stroke, respiratory disease, ‘flu, falls and injuries and hypothermia (NICE, 2015) are all greater in winter. Older people are particularly susceptible as they are more likely to have complex health needs and live in under heated homes which exacerbates illness (Department of Energy and Climate Change, 2014).
In this study we did not analyse data from general practice or community services to understand whether problems in these sectors could add pressure onto emergency care in hospitals. But studies using the GP patient survey found that only 9.2 per cent of people who could not get a convenient appointment at their GP practice went to A&E or a walk-in centre (Wallis, 2013). Insufficient out-of-hours services in general practice does not seem to be associated with pressure on A&E (Blunt, 2015). And the latest results from the GP patient survey show that only 20 per cent of respondents said that they had tried to contact an NHS service when their GP surgery was closed. Of those, a third said that they went to A&E. Extrapolating from this would mean that only approximately 6 to 7 per cent of patients who needed an out-of-hours service went to A&E instead. The survey also shows that the majority (61 per cent of patients contacting the NHS out of hours) contacted an NHS service via telephone. We are not able to tell from the survey whether there are any any particular difficulties occurring in the winter months.
One of the telephone services patients can use is NHS 111. Since August 2010 (the start of the data collection period) there have been 30 million calls to NHS 111, and of these, only 8 per cent of callers were referred to A&E. There are fluctuations at the start of the data collection, but reviewing the last three financial years shows there are more likely to be referrals to A&E in the summer months than the winter ones (data not shown).
There is little information available about demand or capacity for community care. However, looking at the numbers of nurses in community services, it appears that, while the number of nurses dropped by over 2,000 between late 2009 to a low in September 2013, this number has been increasing again back to 2009 levels (QualityWatch, 2015). It should be noted that this does not show us whether or not this is sufficient to satisfy demand, particularly in winter.
Influence of inpatient service capacity
Clearly rising demand is one factor influencing the performance of acute and emergency NHS services in winter, and supply is another. There is a growing body of evidence showing that – in addition to the influence of the specific patient groups attending A&E and certain patient pathways on the ability of A&E departments to meet the four-hour target – the capacity and flow through inpatient services also has an impact. Studies by the BMA (2003) and Monitor (2015) indicate that reducing numbers of inpatient beds, rising occupancy rates and delayed transfers of care, and staffing capacity all have an effect.
Hospital bed capacity
One study found that there was an increased risk of having no beds available for patients requiring immediate admission when average bed occupancy rates exceed about 85 per cent (Bagust and others, 1999). The study also found that acute hospitals can expect regular bed shortages and periodic bed crises if average bed occupancy rises to 90 per cent or more. Using quarterly data, we can see that hospital bed capacity (i.e. the number of beds available) increased during the winter months in the period reviewed, but there are also peaks in bed occupancy at the same time (quarter 1 for any given financial year). In the last three winters the occupancy rate reached 90 per cent. This suggests that winter months bring particular issues in making beds available for admitted patients.
Norovirus, which causes diarrhoea and vomiting, is a particularly contagious disease, and often occurs in winter. If it occurs in a hospital, beds are often closed (Guidelines for the Management of Norovirus Outbreaks, 2012). However, more beds are also closed due to delayed transfer of care than due to norovirus during the winter months, but the picture of how delayed transfer of care relates to bed occupancy is not clear when looking at the official discharge data collected throughout the year.
It has been suggested that the overall increasing number of delayed transfers of care has contributed, along with a decrease in the number of beds, to the overall increased bed occupancy. However, the link during the winter months is not obvious. The number of patients delayed actually drops in December and returns to pre-Christmas levels in January. There is no discernible increase in the number of days delayed during this time. Yet occupancy at this period is at its highest.
While delayed transfers of care are most often due to the NHS, some are a result of reductions in the availability of social care. Between 2009–10 and 2013–14, there has been a 16 per cent reduction in real-terms net expenditure on social care for adults, which means almost 300,000 fewer older adults have received publicly funded social care services (Holder, 2014). However, over a similar period (Figure 3.6), delayed transfers of care attributable to social care have remained fairly static.
Impact on services outside of A&E
In this report, we have also explored winter pressures in ambulance services and inpatient services as well as in A&E units, but because of data limitations did not analyse the impact in primary or community or social care sectors.
Impact on ambulance services
Pressures in A&E departments can reduce their ability to take new patients who arrive by ambulance. Peaks in the number of ambulances queuing do appear to coincide with breaches of the four-hour A&E waiting target. They also correspond to periods when there is a reduction in ambulance services’ ability to respond to emergency calls within eight minutes.
However, more work is needed to look into how much impact each of these factors have on ambulance response times, especially as there is no systematic information on ambulance queuing outside of the winter period. This work could also bring in staff sickness absence rates. In addition, the ambulance queuing measure used is not nuanced: it only looks at ‘extreme’ waits, and it might be the case that a large increase in people waiting in ambulances just a little bit longer without actually breaching the 30-minute threshold has a more significant impact on ambulance response times. It should also be noted that this study only considered the national picture, and more in-depth analysis locally – including looking at events such as A&E diverts and closures – will provide a richer picture on what is happening within the system.
Impact on inpatient care
During the winter months, increasing bed occupancy, as well as other factors, can result in planned and urgent inpatient care being cancelled.
The number of urgent operations cancelled is highest in winter - peaking at over 400 a month - compared to the rest of the year. Numbers of urgent operations cancelled twice or more are relatively low: less than 16 in any given month throughout the year. As a result, these data are more susceptible to random fluctuations and so it is hard to discern whether there is a winter pattern.
The number and proportion of elective operations being cancelled peak in either quarter 3 (September to December) or quarter 4 (January to March) of any given financial year, and increase from a quarterly average low of 0.9 per cent in quarter 2 to a quarterly average of 1.2 per cent in quarter 4. This may not sound like a large increase, but does mean an average quarterly difference of over 4,000 operations across England.
Capacity within urgent care can be managed by transferring patients out of critical care where possible, and this does seem to be linked to winter pressures as the number of patients transferred out of critical care (for non-clinical reasons) peak during the winter months.
Was 2014–15 a particularly bad winter?
In 2014–15 £700 million was provided by the Government (2014) in order to help relieve winter pressures by using the money to provide extra staff and beds and to release capacity through minimising delayed discharges.
As widely reported at the time, in the winter of 2014–15 performance against the four-hour A&E waiting time target decreased dramatically from an already low point. However, there were no similar dips in performance when looking at other A&E waiting time measures, such as the average time spent in A&E and the average time until treatment began.7 This suggests that those waiting ‘extreme’ lengths of time were particularly adversely affected in 2014–15.
Ambulance queuing – measured by the number of ambulance handovers delayed by 30 minutes – increased considerably in 2014–15. A knock-on effect on ambulance response times was also observed, decreasing well below levels seen in previous winters.
Despite all this, not all services experienced a particularly dramatic effect in the winter of 2014–15. There do not seem to be any discernible changes in the number of cancelled operations, or the number of people transferred out of critical care (for nonclinical reasons) outside of an approved group However, a lack of increases in critical care transfers outside of an approved group could have been offset by an increase in transfers within an approved group in 2014–15.
There was also little variation in the extreme ‘winter pressure’ effects. A&E diverts and closures were maintained at the same levels as in previous winters (although local analysis may provide a different view), while trusts reporting operational problems did increase in 2014–15 compared with the previous two winters, but not to the levels seen in 2010–11 and 2011–12 (although this may be susceptible to changes in reporting of data).
So, while the £700 million funding may have maintained performance against certain performance measures, there were clear problems in the winter of 2014–15 that the money was unable to prevent.
What do we know about performance in the 2015-16 winter?
The latest information from Public Health England shows that, for 2015–16:
- all regions are in the two least severe levels of cold weather alerts
- in January there was an indication that the influenza season started
- levels of lower respiratory tract infections appear to be within seasonal expectations
- levels of norovirus and rotavirus incidence appear to be lower than in previous years and reports of outbreaks of diarrhoea and vomiting in hospitals are at lower levels than in previous years.
While it is too early to understand the full impact that winter pressures will have on the NHS in 2015–16, we have entered this period from a historically poor starting point on the four-hour A&E target, trolley waits, ambulance response times, bed occupancy and on delayed transfers of care.
It should be noted that there has not been the same injection of funds specifically for dealing with winter pressures as in previous years. Nationally, trusts are reporting budget deficits that are a first call on spare cash that might have been reserved to support winter preparations (NHS Confederation, 2015).
Reporting on winter pressures has changed this year. Trusts that have previously encountered ‘heightened operational pressure’ are required to submit data for every measure, whereas the other trusts are only required to submit a subset of this data.
Notably, the measures that have changed are:
- ambulance handovers waiting more than 30 minutes
- four-hour waits in A&E
- cancelled operations
- number of beds closed due to delayed transfers of care.
The future outside of the winter months
Problems that were usually confined to the winter months are now increasingly being experienced at other times of the year. The four-hour A&E target has not been met nationally for over two years. Patients waiting to be placed on a ward after a decision to admit (trolley waits) is also an increasing issue, with the number and proportion of patients affected in summer 2015 being at the same levels seen in the winter of 2012–13, and worse than those in the winters of 2010–11 and 2011–12.
However, the proportion of elective operations cancelled and the number of urgent operations cancelled does not seem to be deteriorating. It should be noted that there may be other effects upon a service under strain in terms of inpatient activity, and there are growing concerns about the recent changes in the amount of time patients are waiting for treatment and the number of patients waiting for treatment (QualityWatch, 2015).
With ever-increasing demand on the NHS, performance levels that were historically only observed in the winter period are becoming increasingly common at other times of the year. Recent analysis suggests that if admission rates continue to increase at the rate they have, in the absence of other changes it is estimated that 17,000 new beds will be needed by 2022 (Smith, 2014).
So problems with emergency care performance usually only seen during winter seem to have become ‘the new normal’ for the rest of the year. But uncertainty remains over what our future understanding of ‘winter pressures’ might be.
While a lot of policy interventions, innovations and finances (such as the new models of care, vanguards and the Better Care Fund) are now developing and rightly aimed at cutting the number of admissions and providing more and better services outside of the hospital, some of these are only just starting to embed and it will probably be some years before their impact is felt.
Data sources used in this report
- Health and Social Care Information Centre (2015a) NHS Hospital and Community Health Service (HCHS) monthly workforce statistics. Analysed based on data accessed 9 December 2015
- Health and Social Care Information Centre (2015b) NHS Sickness Absence Rates. Analysed based on data accessed 17 December 2015
- NHS England (2015a) A&E attendances and emergency admissions statistics. Analysed based on data accessed 30 November 2015
- NHS England (2015b). Bed availability and occupancy. Analysed based on data accessed 8 December 2015
- NHS England (2015c) Winter Daily Situation Reports: winter daily SitRep 2015–16 data. Analysed based on data accessed 19 November 2015
- NHS England (2015d) Delayed transfers of care. Analysed based on data accessed 22 December 2015
- NHS England (2015e) Ambulance quality indicators. Analysed based on data accessed 5 January 2016
- NHS England (2015f) A&E attendances and emergency admissions. Analysed based on data accessed 1 December 2015
- NHS England (2015g) Critical care bed capacity and urgent operations cancelled. Analysed based on data accessed 2 December 2015
- NHS England (2015h) Critical care bed capacity and urgent operations cancelled. Analysed based on data accessed 19 December 2015.
About this publication
Suggested citationFisher E and Dorning H (2016) Winter pressures: what’s going on behind the scenes? Nuffield Trust and Health Foundation
The Nuffield Trust is an independent health charity. We aim to improve the quality of health care in the UK by providing evidence-based research and policy analysis and informing and generating debate.
The Health Foundation
The Health Foundation is an independent charity working to improve the quality of healthcare in the UK. We are here to support people working in healthcare practice and policy to make lasting improvements to health services. We carry out research and in-depth policy analysis, fund improvement programmes in the NHS, support and develop leaders and share evidence to encourage wider change.
QualityWatch, a Nuffield Trust and Health Foundation research programme, is providing independent scrutiny into how the quality of health and social care is changing over time.
- The A&E conversion rate is calculated as the number of A&E attendances that result in an emergency admission via A&E. (Back to text)
- The number of patients delayed in this measure is a proxy for all patients as it is a snapshot view determined by the number of patients who had a delayed transfer of care at midnight on the last Thursday of the reporting period. (Back to text)
- There have been some ongoing operational issues and concerns from local health and social care systems about the delayed transfer of care data return, and there has been ongoing work to refresh the guidance to remove ambiguity and improve operational clarity (NHS England Delayed transfers of care guidance, 2015). (Back to text)
- Individual acute NHS trusts are required to specify the geographically related hospitals to which they transfer patients for capacity reasons alone. This is their ‘approved group’. ‘Transfer groups’ are specific to each trust, and arrangements may not be reciprocal (Intensive Care Society, 2011). (Back to text)
- An A&E divert is an agreed temporary diversion of patients to other A&E departments to provide temporary respite (i.e. not to meet a clinical need). This includes diversions between hospitals that are part of the same trust, but geographically separate. An A&E closure is any unplanned, unilateral closure of an A&E department (type 1, 2 or 3) to admissions without consultation, which occurred without agreement neither from neighbouring trusts nor from the ambulance trust. This is irrespective of whether the A&E department is still accepting patients arriving on foot (NHS England Winter Daily Situation Report guidance). (Back to text)
- Trusts are asked to consider, based on their returns against all the measures collected in the Winter Daily Situation Reports as well as any other factors (e.g. staffing issues, adverse weather conditions), whether the trust has experienced ‘serious operational problems’ . Trusts are to use their judgement on what else constitutes a ‘serious operational problem’. If they respond yes, they are asked to provide more information (NHS England Winter Daily Situation Report Guidance). (Back to text)
- In these cases, the average waiting time referred to is the median. (Back to text)