National Health Service (NHS) policy recognises the importance of providing appropriate care at the right time and in the right place to people who seek emergency medical aid, so that those who can benefit from an immediate response and treatment at the highest level of care are identified, and those who can wait can receive advice or be directed to an appropriate service for further care with less urgency. Department of Health policy documents^1–4 have portrayed a vision in which the spectrum of providers, such as emergency departments, general practitioners, the ambulance service, NHS Direct (the UK national nurse-led telephone helpline) and local authorities, should come together in networks to provide appropriate care to the patient. The recent White paper has reiterated this vision of an urgent care system, with the promise that patients needing urgent care “are assessed and directed, first time, to the right services for treatment or help”.⁵

This vision sounds attractive and entirely reasonable. There are clear potential benefits to patients and the NHS: faster response to those with life-threatening conditions; reassurance and self-care advice for those who do not need to attend medical services; and more efficient use of scarce resources. However, the whole system hinges on one critical factor—efficient and effective triage. Triage (French: to sort) needs to be carried out—often remotely—quickly and accurately. This needs to happen no matter what point of contact is selected, be it the local surgery, NHS Direct, the emergency department or emergency ambulance service. The traditional response to this imperative has been to err on the side of caution and tolerate a high rate of “over triage”. However, the increasing call volumes have meant that systems are under strain.

Triage systems currently in use vary across providers. However, there has been an increasing impetus to provide systematic triage, usually with the use of computer-assisted decision support software. Ambulance services, NHS Direct and many out-of-hours providers use one of a handful of such systems available in the UK today. Research evidence is equivocal about the effectiveness of such systems.^6,7 Although it has been reported that a telephone-based triage system can work safely and effectively in the general practitioner’s out-of-hours setting, other studies have found that triage in the emergency ambulance service misses patients with life-threatening conditions, while at the same time callers whose problem can wait or benefit from an alternative response are notoriously difficult to identify.⁶ Trials of telephone advice have suffered from difficulties in identifying potentially suitable calls,^8,9 and other studies have relied on the additional rostering of paramedics in ambulance control and scanning incoming calls for those appropriate to the targeted new service.^10,11 Despite this, ambulance services are increasingly implementing a range of control room and field-based initiatives, reflecting a gap between service development and research in this area.

The further development of a range of responses that are delivered by services working together can have only partial success without attention paid to the crucial point in the process: a rational, consistent, systematic and efficient triage system for all those who contact the NHS for immediate medical care.