Intended for healthcare professionals

Education And Debate

NHS emergency response to 999 calls: alternatives for cases that are neither life threatening nor serious

BMJ 2002; 325 doi: (Published 10 August 2002) Cite this as: BMJ 2002;325:330
  1. Helen Snooks, senior lecturer (h.a.snooks{at},
  2. Susan Williams, research fellowb,
  3. Robert Crouch, consultant nurseb,
  4. Theresa Foster, research associatec,
  5. Chris Hartley-Sharpe, senior operations officerc,
  6. Jeremy Dale, directord
  1. aCentre for Postgraduate Studies, Swansea Clinical School, University of Wales, Swansea SA2 8PP
  2. bSchool of Nursing and Midwifery, University of Southampton, Southampton SO17 1BJ
  3. cLondon Ambulance Service NHS Trust, London SE1 8SD
  4. dCentre for Primary Health Care Studies, University of Warwick, Coventry CV4 7AL
  1. Correspondence to: H Snooks
  • Accepted 14 December 2001

Ambulance services and emergency departments are under increasing pressure as the number of emergency calls continues to rise—but in many cases, patients do not need immediate clinical care. Helen Snooks and colleagues consider the alternatives to the standard NHS response and review the current literature

The number of emergency (999) calls received by ambulance services in the United Kingdom has risen consistently over recent years. Ambulance services must respond to calls immediately by sending vehicles staffed by paramedics, with flashing lights and sirens. All patients have to be taken to an accident and emergency department. This response is not always appropriate, and it can resultin inefficient use of resources and unnecessary risks to the general public, patients, and paramedics.

The NHS Plan and the recent consultation document Reforming Emergency Care have emphasised the importance of trying new approaches to deliver appropriate care. 1 2 They highlight the need to consider new ways to integrate the ambulance response to 999 calls into the overall system that deals with emergencies.

Summary points

Demands on emergency services and inappropriate requests for emergency ambulances are increasing

Ambulance services must respond to calls immediately by sending vehicles with flashing lights and sirens, staffed by paramedics

Many ambulance services want to develop alternatives to the standard response to all 999 calls

Evidence about the safety and effectiveness of alternatives is weak and few rigorous trials have been reported

Studies show that alternative responses are needed but that the work involved in their development is complex

Evidence of the need to change

In England, demand through the 999 telephone system for services has risen by 40% since 1990.3 Problems of overcrowding and high attendance have also been noted in emergency departments and in primary care. Concerns have been expressed over the number of home visits requested at night and whether all such visits are necessary.4 Several studies have reported a high proportion of ambulance call outs that do not warrant an emergency ambulance—a problem that seems to be common internationally. In the United Kingdom, around 40% of 999 calls are estimated not to need an emergency response.5

Lights and sirens are used by ambulance crews to shorten response times to 999 calls, even though a substantial proportion of cases may not subsequently be found to have needed immediate clinical care. Use of lights and sirens places the general public, patients, and ambulance crews at riskof injury or death from collisions and increases financial costs.6

Response to need for change

Some patients do benefit from a rapid response. The benefits of early treatment for patients with myocardial infarction or those in cardiac arrest are reflected in targets set out in the national service framework for the treatment of coronary heart disease.7 The Department of Health has set a response target of eight minutes for calls about patients with immediately life threatening conditions in England and Wales.

Recommendations to enable ambulance services to respond quickly to patients who can benefit from early intervention, as well as to provide a more appropriate service to the full range of 999 callers, have been made. These state that “local health authorities and ambulance services should be free to make whatever response is right for the patient's clinical need, by ambulance or otherwise” for calls about patients with conditions that are neither life threatening nor serious (category C).8 Some ambulance services in the United Kingdom are trying alternative responses for these patients.9

Sources and selection criteria

We reviewed the literature on alternatives to current emergency ambulance service provision—from the initial point of contact with the ambulance service throughout the prehospital phase of care. We searched Medline, BIDS, Healthplan, and Helmis for papers published in English in 1975-2001 (box 1). We also manually searched relevant journals and cross checked with the bibliographies of published reviews and original articles. We appraised papers for relevance, rigour of methods, and validity of findings, on the basis of their results (see table on 8 9 1214 1622 In general, we did not include preliminary research findings published in the form of abstracts, as theylacked detail and were difficult to appraise fully.

Keywords used in literature search

  • Ambulances

  • Ambulances-history

  • Ambulances-standards

  • Emergency-medical-technicians

  • Emergency-medical-services

  • Emergency-medical services utilization

  • Emergency service

  • Transportation of patients

  • Prehospital care


Ten papers described the profile and outcomes of emergency patients who were not transported by the ambulance crew that attended them or evaluated an alternative to the current emergency response (table). We found no papers from countries other than the United Kingdom or United States.

Alternatives in the ambulance control centre

Prioritising 999 calls

Priority dispatch systems have recently been introduced across the United Kingdom to triage calls in ambulance control. These will potentially enable

  • Faster responses to calls about patients with life threatening conditions

  • Care to start as soon as a call is received

  • Lights and sirens to be used in fewer cases10

Research evidence about the safety and accuracy of call prioritisation is limited and conflicting. Undertriage (calls assigned a priority below their actual clinical need) as assessed by a clinical panel was reported to be low (4/571, 0.7%), but a higher rate was found in a subsequent study in which undertriage was assessed by ambulance crews (28/104, 27%). 11 12 In these two British studies, the priority dispatch systems were used only in “shadow form”—a simulation. In an operational “real life” context, call takers may behave differently, and they may be more reluctant to delay the response by designating calls as low priority.

Telephone advice

Telephone consultation seems to be able to provide an alternative response to non-serious 999 callers. Alternatives could include referral to an appropriate primary care agency or self care advice. Systems led by nurses are safe and effective in out of hours settings.13 In the United Kingdom, such systems are in use in general practice, in emergency departments, and by NHS Direct (box 2). Many services provided by NHS Direct have been set up within or alongside ambulance services. Opportunities for further integration, including the smooth transfer of calls in each direction, are available. Noevidence has yet been published about the safety of passing Category C 999 calls to NHS Direct or whether this service has reduced the number of 999 calls made to ambulance services.14

NHS Direct (

  • Confidential telephone advice line staffed by nurses

  • Open 24 hours a day, 365 days of the year

  • Point of access for information and advice about health care

  • Gives guidance on what to do about specific health worries

  • Provides health advice for more general queries—“Why have I been given this me

  • dicine?” and “Will exercising help my condition?”

  • Gives health related information, such as the location of the nearest late night pharmacist

  • Nurse advises on whom to contact for the right treatment on the basis of symptoms described and information supplied


Preliminary information from a recent study showed that nurses and paramedics in the ambulance control centre can successfully provide telephone advice to 999 callers categorised by priority dispatch systems as non-urgent.15 The study assessed the advice service in shadow form, and comparisons with a control group suggested that the service is potentially safe and acceptable. It could result in more appropriate care and a more efficient service.In America, early evaluation of a pilot study of a similar advice service showed that the number of ambulance journeys could be reduced without compromising patient safety.16 The findings of these studies need to be confirmed in larger, “live,” controlled trials.

Alternative vehicles

A study from the United States reported on the issue of providing alternative responses to patients who ask for an emergency ambulance for minor medical problems.17 Of 626 patients surveyed, 11% were judged not to have needed ambulance transportation to an emergency department. The most common reason patients gave for asking for an ambulance was lack of any alternative transport, although 82% were willing to use an alternative if one was available. The authors concluded that unnecessary use of emergency ambulances would decline if alternatives were provided. No indication was given of how patients who could safely be offered alternativetransportation could be identified before an ambulance is dispatched.

No published studies from the United Kingdom have addressed this issue. The results of early evaluations of alternatives, such as bicycles and vehicles unable to carry stretchers, by the LondonAmbulance Service seem to be encouraging.9

On-scene alternatives


In 1998-9, the proportion of non-transported patients, those patients attended by an emergency ambulance after a 999 call who were left at the scene, across England and Wales was 17%.3 Despite this, few ambulance services in the United Kingdom have any policies or protocols about non-transportation, and none provide training for crews on whether to leave patients at home.18 In fact, most ambulance crews in the United Kingdom, as in the United States, may leave patients at the scene only when they refuseto travel.

Despite differences between the emergency medical service systems in the United Kingdom and United States, issues of non-transportation seem remarkably similar. One American study reported the proportion of non-transported patients as 26%.19 Considerable concerns have been expressed about the risks of litigation associated with non-transportation.20

Some useful information about non-transported patients inthe United States has been published, but no such data are available in the United Kingdom. In the United States, non-transported patients are often younger than those taken to hospital, and they often have conditions such as minor trauma or even no illness or injury. 19 21 Interestingly, some non-transported patients were reported as disorientated.

Studies in the United States have looked at the appropriateness and outcomes of non-transportation. In Selden's study in 1991, when non-transported cases were considered against the standard criteria for release from care, non-transportation was appropriate in 78% of cases.18 Inadequate documentation was the most common reason for inappropriate release. Three other studies in the United States described serious and occasionally fatal outcomes.2123 Up to 65% of patients left at the scene needed further medical help within the week after the emergency medical service attended, with up to 20% needing emergency care and admission to hospital. Follow up rates were low (59-67%), and more of those lost to follow up may have had adverse outcomes than those who were traced. No studies evaluating non-transportation in the United Kingdom have been published.

Field triage and diagnosis by paramedics

Paramedics' abilities to determine patients' need for transportation to emergency departments have been assessed in the United States by retrospective review of case notes. Results from a general population found subsequent events that indicated the need for transportation in at least 3% of cases,24 although findings from a study focusing on hypoglycaemic patients were more encouraging.25

Further research on protocols that allow patients to be left at home is underway in the UnitedStates and United Kingdom, with early results mirroring those reported above. Emergency services generally seem to make the correct decision about whether to release patients who do not need emergency treatment rather than taking them to hospital, but such decisions do put a few patients at serious risk.26 A trial of “treat and release” protocols in Albuquerque was suspended recently owing to safety concerns.27

Although accuracy of triage of seriously injured patients to appropriate care has been well researched, evidence about the safety and effectiveness of prehospital triage to minor treatment centres has not been published in the United Kingdom. An evaluation of protocols that allowed crews to take patients to similar clinics in California found that protocols were used for only a small proportion of eligible patients.28 Some patients needed immediate referral to an emergency department, although triage was assessed as generally appropriate and patient satisfaction was high. The lack of a control group in this study meant that the overall benefits were not clear. A randomised controlled trial of triage and transportation to minor injury units of patients in two parts of south east England is currently under way; results are due in 2002.29


Current developments in the ambulance service provide opportunities for it to work with other healthcare providers to optimise the response to emergency calls about patients with non-serious conditions. Research about the clinical effectiveness and cost effectiveness of prioritised dispatch, telephone advice, and on-scene triage and referral is lacking.

The studies we reviewed were mainly conducted in the United States, although papers and reports from the United Kingdom indicate that concerns about emergency services' workloads and the management of patients are similar for both countries. We found no published studies from other parts of the world, although other systems probably face similar challenges. Many descriptive and exploratory studies were identified; none of the few trials we identified were randomised or controlled. Weaknesses in the methods were apparent in all of the studies, particularly with regard to measurement of appropriateness. Severalabstracts reported the ability of paramedics to triage patients for release from care and stated that doctors disagreed with the decisions made by paramedics in the field about the most appropriate care. We did not include the abstracts because of their brevity and the preliminary nature of the findings.

Research in this area does not identify a “gold standard” of appropriate care. Various methods were used to assess appropriateness, including:

  • Identification of treatments subsequently given

  • Review against protocol

  • Comparison with doctors' opinions

Each method has its weaknesses, with individual doctors' judgments shown to be unreliable in related research about appropriateness of care. Research leading to full publication rather than thepublication of abstracts is clearly needed in this difficult area.

Evidence about the need to develop alternatives that are more appropriate to the current 999 response in the United Kingdom is strong. The benefits of developing more appropriate responses to patients who call 999 with non-emergency problems could accrue to these patients, patients with life threatening conditions, and the NHS. If we view the 999 service as a component of the emergency care system, we could develop a more strategic response, in which people calling about non-emergency situations are diverted to a service that provides more appropriate care. This would allow the ambulance service to respond quickly to patients who would benefit from early intervention. An integrated single point of access to immediate care, such as NHS Direct, may enable an appropriate response to be triggered for the full range of cases. Any exploration of this option, however, must take into account the effect on response times for life threatening emergencies.


We thank Dr Frances Maggs-Rapport and Professor John Williams for providing additional editing advice in the latter stages of preparation of this paper.

Contributors: SW and HS carried out the primary literature searches for this review, although other authors also provided references. All authors contributed to the writing of drafts. HS took the lead in producing the final version submitted for publication and will act as guarantor.


  • Funding NHS Executive Primary/Secondary Care Interface Programme

  • Competing interests JD and RC own shares in and have acted as clinical consultants to the Plain Software Company, which produces clinical decision support software.

    A summary of published studies appears on

    Embedded Image


View Abstract