Article Text

Download PDFPDF

Reducing inappropriate emergency department attendances—a review of ambulance service attendances at a regional teaching hospital in Scotland
  1. Gareth Gordon Patton,
  2. Shobhan Thakore
  1. Correspondence to Dr Gareth Patton, Emergency Department, Ninewells Hospital, Dundee DD1 9SY, UK; garethpatton{at}


Introduction Emergency Departments (ED) in the UK have seen increasing attendance rates in recent years. Departments are now seeking strategies to reduce their attendances. A review of all ambulance attendances to the ED at Ninewells Hospital was conducted to identify if patients presenting by ambulance could be seen and treated more appropriately in other parts of the health service.

Method A retrospective review of ambulance attendances to the ED at Ninewells Hospital over 7 non-consecutive days. The ambulance patient report form and the ED notes were reviewed by the duty consultant to deem whether it was appropriate for the patient to be presented to the ED. If inappropriate, an alternative destination was suggested. Additional data was collected on the source of the ambulance call.

Results There were 910 attendances in the 7 days. 295 (32%) presented by ambulance. 32 had incomplete data and were excluded. 185 (70%) and 179 (68%) of the 263 were deemed appropriate from review of the patient report form and notes respectively. Of the inappropriate, 74.4% and 79.7% had primary care suggested as an alternative. Patients who call for their own ambulance and NHS24 had higher rates of inappropriate attendances.

Discussion The ambulance services present one-third of the patients to the ED at Ninewells Hospital. 30%–32% were found to be attending inappropriately and 74%–80% of these could have been managed in primary care. Reducing inappropriate ambulance attendances could reduce the departmental patient load by 11%.

  • Paediatric emergency med
  • resuscitation
  • emergency department
  • chest
  • abdomen
  • analgesia pain control
  • education

Statistics from


Emergency Departments (ED) throughout the UK have seen significant rises in attendances in recent years. The Department of Health figures show that new attendances have increased by 90.4% between 1987–88 and 2010–11 in England1 and 9.4% between 1999–2000 and 2008–09 in Scotland.2 Many suggestions have been made as to why this rise has come about. Health promotion and patient awareness undoubtedly play a role by increasing demand on the health service in general. Other contributory factors include the limited access to Out of Hours Primary Care and limited patient knowledge in how to access such resources, with only two-thirds of patients aware of how to access out-of-hours primary care.3 ,4 In addition the convenience of attending the ED where multiple investigations are available at any time of the day without having to pre-book for an appointment must be appealing. In parallel, the ambulance service has seen a steady annual increase in number of calls at 6.5% per year.3 By default a large percentage of these ambulance calls present to the ED. Nationally, of these presentations, 43% were discharged and over two-thirds of these did not need follow-up treatment.3 The Department of Health itself recognises ‘that some of these patients could have been cared for more appropriately and more cost-effectively by other urgent care services’.3

This pressure on the service which is at the ‘hub’ of emergency care5 has led to departments trying to reduce the number of ‘inappropriate’ attendances as a step in reaching the agreed level of reduction in attendances by 2013–14 as agreed in the national HEAT Targets. These performance targets were agreed and set by the Scottish Government and NHS Boards. By reducing these attendances you can reduce any clogging of the service and focus on those who genuinely require emergency care and allows for more efficient patient flow.

In 2010 there were 47 989 attendances to the ED in Ninewells Hospital, Dundee. Of these 15 980 were presented by the Scottish Ambulance Service (SAS), exactly one third of the total attendances. The (SAS) try independently to reduce attendances with ‘See and Treat’ Protocols being applied to 57 560 patients in Scotland in 2009/2010.6 In a bid to try and further reduce inappropriate attendances by ambulance, a review of attendances by ambulance to our ED was conducted to see if patients could be seen and treated more appropriately in other parts of the health service rather than attending the ED.


This review was conducted in Ninewells Hospital, a regional university teaching hospital in Dundee. The region's population is spread across urban and remote rural environments. Surgical and medical admissions units are separate from the ED with direct admissions. Data was collected over a 3-week period in September 2010. A retrospective review of the ambulance arrivals was conducted for 7 non-consecutive days. Each day of the week was represented.

The patients were identified as ambulance arrivals on our departmental patient record system, Symphony. The ED notes of the identified patients were collected along with the Patient Report Form (PRF) as completed by the ambulance crew. An audit form with patient demographics was completed (table 1). These notes were then reviewed by the Duty Emergency Medicine Consultant. The Duty Consultant changed throughout the course of the study. They deemed whether it was appropriate for the patient to be brought to the ED for assessment or whether they could have been seen more appropriately at another point of care. If deemed inappropriate another location was suggested. The process was applied first to the information on the PRF and then to the ED clinical notes.

Table 1

Demographics of patients included in review


In the 7 day review period there were 910 attendances to the ED. 295 (32%) patients attended by ambulance, including one by air ambulance. On average there were 42 ambulance arrivals per day. 32 of the 295 patients had incomplete data to allow full review of their notes. 27 of the 32 had no PRF (including the air ambulance attendance) and ED notes could not be located for remaining five of the patients. These 32 patients were excluded from review and hence 263 patients were reviewed.

Following review of the PRFs by the duty consultant, 185 (70%) of the 263 attendances were deemed appropriate to attend the ED and 179 (68%) deemed appropriate after review of the ED notes. Of those deemed as inappropriate, a large percentage had Primary Care suggested as a more appropriate destination for review (74.4% from PRF and 79.7% from ED notes). Primary Care review includes General Practice or Out Of Hours Service. Table 2 breaks down the suggested alternate destinations. Other destinations included Minor Injury Unit review, Treat at Home or individual cases that could have bypassed the ED.

Table 2

Suggested alternative destinations for inappropriate attendances

Of those patients who were deemed inappropriate and attended the ED, a certain percentage went on to be admitted to hospital for further investigation or management. Of these inappropriate patients 35 (44.9%) of the 78 in the PRF group and 40 (47.6%) of the 84 ED Note group were admitted. Most of these patients were admitted via the Acute Medical Receiving Ward, although Paediatrics, Surgical, other medical specialties and the ED Short Stay Ward also received admissions. We do not have follow-up details on the admitted patients to review their length of stay in hospital.

Data was also collected on the source of the call to the SAS. Unfortunately from the data it was not possible to identify the source of all of the calls and these were labelled as not clear. Other source, as shown in table 3, refers to community carers, schools, other hospitals or care home.

Table 3

Source of call to SAS


This study has shown that the ambulance service present with approximately one-third of all patients seen in our ED. This proportion may be higher in units where GP's admit via the ED. Of these patients 30%–32% were subsequently found to be attending inappropriately to the ED. This group of patients could have been managed in a more appropriate setting. 74%–80% of these could have been seen and managed by primary care.

In excess of 50% of the inappropriate patient group were discharged directly from the ED. It was noted that 45% to 48% of the inappropriate attendances went on to subsequently be admitted. If these patients had been seen in primary care it is difficult to say if these figures for admission would be higher or lower. Primary Care would have a greater understanding of the individual patient's general condition and social support and this may reduce the likelihood of admission. They also have the ability to initiate a management plan at home and arrange review in the community rather than admit. Even if a GP were to decide to admit a patient this would, in our setting, mean a direct admission to a bed on a receiving ward. This pathway would be safe and be more comfortable and efficient for individual patients.

To reduce the inappropriate presentations, we should consider the source of referral as one part of a strategy to prevent their presentation. On review of table 3 showing the source of the call, it is apparent that the patients themselves and NHS 24 both have greater percentage attendance rates that were deemed inappropriate rather than appropriate. Patient education about how to access the healthcare system appropriately may reduce the default use of dialling 999 for an ambulance. This level of education will require a large scale national campaign. However it is also necessary to have triage systems in ambulance control to prioritise emergency calls effectively. It may even be necessary at times to deliver education to callers about the appropriate use of the 999 system at the time of the call as a form of immediate feedback. This would clearly need to be done selectively and sensitively. While these interventions may be possible any change resulting from them is likely to take time to produce significant results.

Referrals from NHS24 also have a higher inappropriate rate than appropriate. The use of this resource has filled the gap left by the introduction of the new General Practitioner (GP) contract. Patients no longer have rapid access to their own GP out of hours and are instead advised to call NHS24 for health advice. The clinical evaluation of a patient over the phone can be extremely difficult. Algorithms are used to support decision making and these will inevitably err on the side of caution. Calls to NHS24 have increased by 2% between 2006/07 and 2008/09 but calls leading to a 999 response have increased by 16%.2

It is difficult to see a national education campaign working quickly and it is also hard to see how telephone triage could be made less risk averse. This logically leads to an examination of what could be done at scene to prevent transport to the ED. Once an ambulance has been dispatched the options open to the crew can be limited, leading to ED attendance. Telephone based assessments are difficult but the arrival of an ambulance crew gives the opportunity for a face to face consultation. At this point some form of decision support would be extremely useful in directing patients to the most appropriate destination for their needs. The close percentage of agreement in levels of appropriate and inappropriate attendances following review of the information in the PRF and the ED notes supports this suggestion. At scene the crew are extracting a sufficient and similar level of information to staff in the ED. With this level of information at scene the crew can consider the best options for their patient and discuss these with the relevant experienced practitioner. It is important that crews are encouraged to use this support. Our recommendation would be that crews are able to contact either a senior Emergency Physician or a GP for advice by using existing and underutilised lines of communication such as telephone or radio call. Formalisation and education in the use of these sources of decision support may result in fewer transfers to the ED, with an increase in the number of consultations conducted by Primary Care. It is likely that any such increase in Primary Care activity would be diluted across a number of local practices rather than affect one practice in isolation.

This study suggests that a large proportion of emergency ambulance attendances could or should be assessed in another part of the healthcare system rather than the ED. Patient journeys could be improved by introducing a system that allows the ambulance crew to seek decision support from an experienced practitioner.


Taken in context, our department sees 47 989 patients annually. 15 980 (33.3%) of these present by ambulance. One-third of these presentations could be seen in another part of the healthcare system or given self-care advice and left at scene. If a system was introduced where the ambulance service had the ability to access other parts of the healthcare system a significant reduction in unnecessary ambulance journeys and ED attendances could be achieved. This could reduce our annual patient load by 11%.



  • Contributors The initial conception and design idea was generated by ST. Notes were reviewed by the departmental (duty) consultants. Data review and interpretation was performed by GP. Literature and background searches were performed by GP. The first draft was written by GP. ST reviewed and edited the drafts. Both authors, GP and ST agreed on the final draft.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement The data collected and used is based and held in the Emergency Department, Ninewells Hospital. All data collected was used and included in the article.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.