NIHR DC Discover

Emergency department sign outside hospital

NIHR Signal Closing five emergency departments not linked with increased hospital admissions, though ambulance call-outs increased

Published on 23 October 2018

doi: 10.3310/signal-000665

Closure of five small emergency departments in England was not associated with change in the number of hospital admissions, urgent care attendances or deaths among the local populations. However, ambulance call-outs increased by 14% relative to comparison areas, with a four-minute increase in the time to reach a hospital with an emergency department.

Emergency departments continue to be under high pressure, while staff shortages increase patient safety concerns. One option is to close smaller sites and divert care to larger specialist centres that can provide the full scope of emergency and trauma care. However, local service reconfigurations can be concerning for communities.

This NIHR-funded study is the first to assess the impact of downgrading or closing emergency departments. Five sites where these changes happened were matched with comparable control sites. The lack of change in the overall number of deaths seems reassuring, though there are limitations related to the completeness of data and a small and uncertain increase in risk of dying from specific emergency conditions.   

Further cost analysis and study on the impact on ambulance services would be valuable.

Share your views on the research.

Why was this study needed?

In 2017-18, there were 23.8 million attendances at emergency departments in England, a 22% increase on 2008-09. There are 27,639 attendances per 100,000 of the population, which is the highest of all the UK nations.

As part of the Five Year Forward View, NHS England is transforming emergency and urgent care so that people can access the right care in the right place whenever they need it. This includes making greater use of primary care, urgent care centres and other community services; also ensuring that seriously ill patients have access to specialist emergency and major trauma centres which are known to save lives. Among NHS sustainability and transformation plans is the potential closure or downgrading of smaller emergency departments.  

Such plans often raise strong public health concerns, yet there has been little research to inform whether closures could be detrimental to local population health. This study aimed to provide evidence to inform decision-making.

What did this study do?

This study measured the impact of closing five emergency departments in Newark, Hemel Hempstead, Bishop Auckland, Hartlepool and Rochdale, between 2009 and 2011. The emergency departments were all type 1, meaning they provided 24-hour, consultant-led, multispecialty services with full resuscitation facilities. Effects on the population catchment area were compared with five control regions matched on socio-demographic characteristics.

Hospital Episode Statistics, accident and emergency and ambulance services data and mortality records from the Office for National Statistics were analysed for the two years before and after each closure.

The aim was to look at indicators including:

  • Ambulance service activity, such as the number of emergency ambulance incidents, time from calling 999 call to arrival at hospital, emergency and urgent care attendances and minor attendances.
  • Emergency hospital admissions, length of stay and deaths from serious emergency conditions within seven days. 
  • The risk of death (as a case fatality ratio) was collected for sites as the ratio of these deaths from serious emergency conditions amongst all those who died or were admitted to hospital for them.

The analyses took account of seasonal effects and other local service changes. The main limitations centred upon incompleteness of routine data collected. There were also differences in outcomes across the five different sites, reducing confidence in the precision of the results and generalisability to other areas.  

What did it find?

  • There were no changes in the total number of emergency or urgent care attendances, the number of ‘minor attendances’ without hospital treatment or investigations, total hospital admissions, length of stay, or potentially avoidable admissions. This was both in comparison to before and after the closure, and compared with control sites.
  • Compared with control sites there was a 13.9% increase (95% confidence interval [CI] +3.5% to +24.4%) in the number of ambulance call-outs. Similarly, there was a 12.3% increase (95% CI +3.5% to +21.1%) in the number of call-outs for potentially life-threatening emergencies (‘red’ incidents). Results were similar in the before-after analysis at the sites, though the difference in ambulance call-outs just fell short of significance.  
  • Following the closure, there was a 3.9-minute increase (95% CI +2.2 to +5.6) in the average time between a 999 call and arrival at hospital for ‘red’ incidents compared with before the closure. There was insufficient data on journey time at control sites to allow comparison.
  • Emergency department closure was not associated with a change in the number of out-of-hospital deaths from 16 serious emergency conditions. There was also no significant change in deaths by seven days when comparing closed with control sites (+2.2%, 95% CI -6.9% to +11.3%), or when comparing before and after closure at the closed emergency department (-4.2%, 95% CI -15.9% to +7.7%).
  • The risk of dying from specified conditions was slightly increased in this study. There was no difference when comparing before and after across the five closed sites (-0.003, 95% CI -0.017 to +0.012). However, there was finding of a small 0.023 increase (95% CI +0.009 to +0.036) when comparing closed with control sites.

What does current guidance say on this issue?

The NHS Five Year Forward View (2014) discussed the viability of smaller hospitals, overall, and planned to consider:

  • the costs of delivering safe, efficient services in smaller compared to larger providers
  • new organisational models for smaller acute hospitals
  • new sustainable models of medical staffing

The Royal College of Emergency Medicine (2017) suggest that decisions to downgrade emergency departments or centralise to larger sites should consider the needs of the local community and the effect that diverted patient flow would have on primary care and ambulance services. They consider that reconfiguration should not be driven by workforce issues.

NICE guidelines on emergency and acute medical care (2018) found low-quality evidence regarding restricted emergency department opening or closures.

What are the implications?

This study has found little to suggest that selected emergency department closures have adverse impacts on communities. For example, the lack of change in the number of emergency or urgent care attendances or emergency hospital admissions, despite an expected increase in ambulance call-outs, is reassuring.

This study did not show a change in the number of deaths overall and specifically none for those living furthest from an emergency department. Although there was a small increase in `risk of death’ from serious emergency conditions in sites with closures compared with control sites, this could have been linked to changes in admission thresholds. 

The nature of reconfigurations locally will to some extent determine how the emergency services can adapt to change. Evaluation of whole systems changes like this can provide important insights for emergency and urgent care services if planned closures are contemplated.

Citation and Funding

Knowles E, Shephard N, Stone T, et al. Closing five emergency departments in England between 2009 and 2011: the closed controlled interrupted time-series analysis. Health Serv Deliv Res. 2018;6(27).

This project was funded by the National Institute for Health Research Health Services and Delivery Research Programme (project number 13/10/42).

Bibliography

NHS website. When to go to A&E. London: Department of Health; 2018.

NHS website. When to visit an urgent care centre (walk-in centre or minor injury unit). London: Department of Health; 2018.

NHS England. Five Year Forward View. NHS England; 2014

NHS England. Next steps on the NHS Five Year Forward View. NHS England; 2017.

NHS England. Urgent treatment centres - principles and standards. NHS England; 2017.

NICE. Emergency and acute medical care in over 16s: service delivery and organisation.  NG94. London: National Institute for Health and Clinical Excellence; 2018.

RCEM. Reconfiguring emergency medicine services: service design and delivery. London: The Royal College of Emergency Medicine; 2017.

Why was this study needed?

In 2017-18, there were 23.8 million attendances at emergency departments in England, a 22% increase on 2008-09. There are 27,639 attendances per 100,000 of the population, which is the highest of all the UK nations.

As part of the Five Year Forward View, NHS England is transforming emergency and urgent care so that people can access the right care in the right place whenever they need it. This includes making greater use of primary care, urgent care centres and other community services; also ensuring that seriously ill patients have access to specialist emergency and major trauma centres which are known to save lives. Among NHS sustainability and transformation plans is the potential closure or downgrading of smaller emergency departments.  

Such plans often raise strong public health concerns, yet there has been little research to inform whether closures could be detrimental to local population health. This study aimed to provide evidence to inform decision-making.

What did this study do?

This study measured the impact of closing five emergency departments in Newark, Hemel Hempstead, Bishop Auckland, Hartlepool and Rochdale, between 2009 and 2011. The emergency departments were all type 1, meaning they provided 24-hour, consultant-led, multispecialty services with full resuscitation facilities. Effects on the population catchment area were compared with five control regions matched on socio-demographic characteristics.

Hospital Episode Statistics, accident and emergency and ambulance services data and mortality records from the Office for National Statistics were analysed for the two years before and after each closure.

The aim was to look at indicators including:

  • Ambulance service activity, such as the number of emergency ambulance incidents, time from calling 999 call to arrival at hospital, emergency and urgent care attendances and minor attendances.
  • Emergency hospital admissions, length of stay and deaths from serious emergency conditions within seven days. 
  • The risk of death (as a case fatality ratio) was collected for sites as the ratio of these deaths from serious emergency conditions amongst all those who died or were admitted to hospital for them.

The analyses took account of seasonal effects and other local service changes. The main limitations centred upon incompleteness of routine data collected. There were also differences in outcomes across the five different sites, reducing confidence in the precision of the results and generalisability to other areas.  

What did it find?

  • There were no changes in the total number of emergency or urgent care attendances, the number of ‘minor attendances’ without hospital treatment or investigations, total hospital admissions, length of stay, or potentially avoidable admissions. This was both in comparison to before and after the closure, and compared with control sites.
  • Compared with control sites there was a 13.9% increase (95% confidence interval [CI] +3.5% to +24.4%) in the number of ambulance call-outs. Similarly, there was a 12.3% increase (95% CI +3.5% to +21.1%) in the number of call-outs for potentially life-threatening emergencies (‘red’ incidents). Results were similar in the before-after analysis at the sites, though the difference in ambulance call-outs just fell short of significance.  
  • Following the closure, there was a 3.9-minute increase (95% CI +2.2 to +5.6) in the average time between a 999 call and arrival at hospital for ‘red’ incidents compared with before the closure. There was insufficient data on journey time at control sites to allow comparison.
  • Emergency department closure was not associated with a change in the number of out-of-hospital deaths from 16 serious emergency conditions. There was also no significant change in deaths by seven days when comparing closed with control sites (+2.2%, 95% CI -6.9% to +11.3%), or when comparing before and after closure at the closed emergency department (-4.2%, 95% CI -15.9% to +7.7%).
  • The risk of dying from specified conditions was slightly increased in this study. There was no difference when comparing before and after across the five closed sites (-0.003, 95% CI -0.017 to +0.012). However, there was finding of a small 0.023 increase (95% CI +0.009 to +0.036) when comparing closed with control sites.

What does current guidance say on this issue?

The NHS Five Year Forward View (2014) discussed the viability of smaller hospitals, overall, and planned to consider:

  • the costs of delivering safe, efficient services in smaller compared to larger providers
  • new organisational models for smaller acute hospitals
  • new sustainable models of medical staffing

The Royal College of Emergency Medicine (2017) suggest that decisions to downgrade emergency departments or centralise to larger sites should consider the needs of the local community and the effect that diverted patient flow would have on primary care and ambulance services. They consider that reconfiguration should not be driven by workforce issues.

NICE guidelines on emergency and acute medical care (2018) found low-quality evidence regarding restricted emergency department opening or closures.

What are the implications?

This study has found little to suggest that selected emergency department closures have adverse impacts on communities. For example, the lack of change in the number of emergency or urgent care attendances or emergency hospital admissions, despite an expected increase in ambulance call-outs, is reassuring.

This study did not show a change in the number of deaths overall and specifically none for those living furthest from an emergency department. Although there was a small increase in `risk of death’ from serious emergency conditions in sites with closures compared with control sites, this could have been linked to changes in admission thresholds. 

The nature of reconfigurations locally will to some extent determine how the emergency services can adapt to change. Evaluation of whole systems changes like this can provide important insights for emergency and urgent care services if planned closures are contemplated.

Citation and Funding

Knowles E, Shephard N, Stone T, et al. Closing five emergency departments in England between 2009 and 2011: the closed controlled interrupted time-series analysis. Health Serv Deliv Res. 2018;6(27).

This project was funded by the National Institute for Health Research Health Services and Delivery Research Programme (project number 13/10/42).

Bibliography

NHS website. When to go to A&E. London: Department of Health; 2018.

NHS website. When to visit an urgent care centre (walk-in centre or minor injury unit). London: Department of Health; 2018.

NHS England. Five Year Forward View. NHS England; 2014

NHS England. Next steps on the NHS Five Year Forward View. NHS England; 2017.

NHS England. Urgent treatment centres - principles and standards. NHS England; 2017.

NICE. Emergency and acute medical care in over 16s: service delivery and organisation.  NG94. London: National Institute for Health and Clinical Excellence; 2018.

RCEM. Reconfiguring emergency medicine services: service design and delivery. London: The Royal College of Emergency Medicine; 2017.

Closing five Emergency Departments in England between 2009 and 2011: the closED controlled interrupted time-series analysis

Published on 1 August 2018

Knowles E, Shephard N, Stone T, Bishop-Edwards L, Hirst E, Abouzeid L, Mason S & Nicholl J

Health Services and Delivery Research Volume 6 Issue 27 , 2018

Background In recent years, a number of emergency departments (EDs) have closed or have been replaced by another facility such as an urgent care centre. With further reorganisation of EDs expected, this study aimed to provide research evidence to inform the public, the NHS and policy-makers when considering local closures. Objective To understand the impact of ED closures/downgrades on populations and emergency care providers. Design A controlled interrupted time series of monthly data to assess changes in the patterns of mortality in local populations and changes in local emergency care service activity and performance, following the closure of type 1 EDs. Setting The populations of interest were in the resident catchment areas of five EDs that closed between 2009 and 2011 (in Newark, Hemel Hempstead, Bishop Auckland, Hartlepool and Rochdale) and of five control areas. Main outcome measures The primary outcome measures were ambulance service incident volumes and times, the number of emergency and urgent care attendances at EDs, the number of emergency hospital admissions, mortality, and case fatality ratios. Data sources Data were sourced from the Office for National Statistics, Hospital Episode Statistics (HES) accident and emergency, HES admitted patient care and ambulance service computer-aided dispatch records. Results There was significant heterogeneity among sites in the results for most of the outcome measures, but the overall findings were as follows: there is evidence of an increase, on average, in the total number of incidents attended by an ambulance following 999 calls, and those categorised as potentially serious emergency incidents; there is no statistically reliable evidence of changes in the number of attendances at emergency or urgent care services or emergency hospital admissions; there is no statistically reliable evidence of any change in the number of deaths from a set of emergency conditions following the ED closure in any site, although, on average, there was a small increase in an indicator of the ‘risk of death’ in the closure areas compared with the control areas. Limitations Unavailable or unreliable data hindered some of the analysis regarding ED and ambulance service performance. Conclusions Overall, across the five areas studied, there was no statistically reliable evidence that the reorganisation of emergency care was associated with an increase in population mortality. This suggests that any negative effects caused by increased journey time to the ED can be offset by other factors; for example, if other new services are introduced and care becomes more effective than it used to be, or if the care received at the now-nearest hospital is more effective than that provided at the hospital where the ED closed. However, there may be implications of reorganisation for NHS emergency care providers, with ambulance services appearing to experience a greater burden. Future work Understanding why effects vary between sites is necessary. It is also necessary to understand the impact on patient experience. Economic evaluation to understand the cost implications of such reorganisation is also desirable. Funding The National Institute for Health Research Health Services and Delivery Research programme.

Expert commentary

Service reconfiguration is an inevitable part of modern healthcare. It is well-recognised that concentration of expertise in high-volume specialist centres leads to improved outcomes for life-threatening conditions; any disadvantage associated with an increased journey time is more than offset by the better care that is delivered on arrival.

However, reconfigurations are often unpopular with clinicians and local communities. There is concern regarding the wider impact of centralising urgent care services, and the downgrading or closure of emergency departments, with a lack of contemporary evidence to inform policy.

The findings from this study are broadly reassuring in terms of the population effects of reconfiguration. Whilst the benefits demonstrated by specialist centres were not seen, there was also no convincing evidence that patient outcomes were worse. The study lends cautious support to clinically justified and appropriately planned and supported reconfigurations. However, there are important implications for other services, particularly ambulances, and these need to be considered carefully in any future plans. 

Jonathan Benger, Professor of Emergency Care, University of the West of England; Consultant in Emergency Medicine, University Hospitals Bristol NHS Foundation Trust