When you call for an ambulance, you generally want it now. To you, it’s an emergency and an emergency requires an immediate response.

The reality of a modern NHS generally and Ambulance Trusts specifically mean that such an expectation is rarely met. Thankfully, in the vast majority of cases the timing of the arrival of the paramedic is of no real consequence. More important to outcome can be what happens once the paramedic alights from the ambulance and attends to the patient, how long it takes from that point until admission to A&E or the unit to which the patient is taken for necessary specialist care. These latter issues can all have a bearing on the timing, the quality and the nature of care received by the patient.

There are occasions though where, to borrow a phrase from the contract lawyers, “time is of the essence”.  By this I mean that how long it takes the ambulance to arrive with the patient may be, in and of itself, determinative of the outcome. An obvious example might be a patient who has gone into cardiac arrest, particularly an arrest which will only respond to the shock from a defibrillator in circumstances where that will not be available until the paramedic arrives at the patient’s side. The timing of the arrival may literally be the difference between that patient’s life or death. There will be numerous examples where treatment is time sensitive, having a bearing on the outcome in terms of the nature and extent of the recovery that will be achieved.

Such cases may give rise to difficult questions of causation, addressing the issue of what would have happened on the balance of probability had the ambulance and paramedic arrived at some earlier given point, and what difference that would that made to the outcome that in fact came to pass.

This piece however is limited to the more fundamental and gateway question of whether and how one may be able to establish that the arrival time of the ambulance (or the paramedic if different – perhaps in a first response vehicle) was in breach of the Ambulance Trust’s duty of care.

The existence of a duty

Perhaps surprisingly, there has been little authority that has considered this question in any particular detail, or which has sought to set out any parameters that may impinge on the issue. The only authority in the area dates back to a Court of Appeal decision from 2000 which is itself essentially confined to a determination that a duty of care was owed by what would now be an Ambulance Trust to the public.

Until this decision it was believed that the earlier decision of the then House of Lords in Hill v Chief Constable of West Yorkshire [1989] AC 53 precluded the imposition of a duty of care on the emergency services, save in certain limited circumstance. In Kent v Griffiths [2000] 2 All ER 474 Lord Woolf MR, giving the unanimous judgment of the Court of Appeal, held that the acceptance of the call and the dispatch of an ambulance (which was then automatic) established a duty of care. Any suggestion by the Defendant that public policy reasons should prevent the imposition of a duty of care were roundly rejected. On the facts of the particular case, breach of duty was made out where there was no evidence of any good reason for an established delay in the ambulance’ arrival and evidence of falsification of ambulance records relating to arrival times by staff. Indeed, breach of duty, if duty existed was admitted. In Kent at first instance the trial judge found that the ambulance arrived 40 minutes after the first call, having taken at least 14 minutes longer than was reasonable and held that the respiratory attack Mrs Kent suffered, which resulted in a miscarriage and brain damage, would be likely to have been averted had there been no unreasonable delay. So duty, breach and the causation of injury, consequent on that breach was made out.

The formulation of the duty

Whilst the existence of a duty is therefore clear, the formulation of that duty is unspecified. There is no authoritative exposition of that duty, beyond an acceptance in Kent that delay without reason or good reason will be enough to establish a breach. The duty is however in its broadest terms a duty on the relevant ambulance service to exercise reasonable skill and care in the performance of their statutory function. In the final analysis it is suggested that what the court will be seeking to identify in any given case is not whether there was delay in the dispatch or the arrival, but delay that was avoidable. Unavoidable delay, properly so understood, will not be enough to establish breach of duty.

Evidence to support breach

I would suggest that key to ascertaining whether one is likely to be able to establish a breach of duty, is a full and detailed understanding of the process that is adopted on a 999 call being connected. It is by this means that one can begin to understand whether and if so, why delay may have crept into the response time.

Before looking at this in any detail it is important to know that on 13 July 2017 NHS England introduced a new set of performance targets to apply across all 999 calls to all Ambulance Trusts in England to achieve a universally applicable response to emergency calls. These were fully implemented by November 2017. This therefore replaced a previously somewhat divergent and inconsistent approach that had hitherto been operating between different Trusts.

Importantly, in gathering evidence on this issue, all calls are recorded and the EMD’s process is captured within the computer programme with times that each critical decision is made.

The decision as to dispatch

Upon a call being connected the dispatching process commences with a Pre-Triage Sieve (PTS) conducted by reference to the Dispatch on Disposition flowchart. The purpose of this is for the emergency medical dispatcher (EMD), who receives the initial call, to identify as quickly as possible whether the call can immediately be identified as meriting a Category 1 response. This PTS function is thus limited to the prompt identification of obviously life-threatening illness or injury.

Three questions are initially asked. The first of these questions, “Is the patient breathing” will, if answered in the negative, result in the EMD presenting the call to Dispatch for allocation as a Category 1 response. The second question is “Is the patient conscious”, but regardless of how that is answered, it is only if the third question “Does their breathing sound noisy” is answered in the affirmative, that this results in the call being presented by the EMD to Dispatch at that point for allocation as a Category 1 response.

On occasions how a question is answered by the caller may impact on the response. Sadly, a caller’s response to what are carefully worded specific questions is not always as clear as a “yes” or “no”. Any answer interpreted as a “don’t know” will not result in a dispatch request being put through at that point.

Assuming the call is not identified as Category 1 on the initial PTS, there is a further opportunity for an emergency (as opposed to urgent) dispatch being requested if the next open question “What is the problem” produces a response which the EMD interprets as one of a number of specified problems which include choking and drowning listed in a Nature of Call list or if the caller uses one of a range of key words or phrases which are listed in alphabetical order for the EMD to scan.

If the initial questions do not illicit a Category 1 response, the EMD will move to a triage process as contained in the Advance Medical Priority Dispatch System (AMPDS) which is a computer-based set of progressive screens, where the EMD is led through a succession of questions which will lead to a categorisation of the call. An initially lower categorisation can and sometimes is re-categorised to a higher level of priority if additional information is gleaned which merits that, with the EMD having the ability to refer to a clinician for advice and guidance based on the information elicited. This gives rise to the possibility of exploring whether a call, not categorised as Category 1 from the Dispatch on Disposition process, should nonetheless have been recategorized in the light of additional information.  To argue this will likely require expert evidence as to why the clinician considering the information ought to have realised that the call in fact related to a life-threatening situation as opposed to one that was only urgent or of even lower priority.

Delays in dispatch once requested

Assuming that a Category 1 dispatch is requested, as a caller one would hope that the ambulance would be dispatched and en-route to the scene immediately. Sadly, commonly this is not the case. In recent times a particular problem has arisen relating to the availability of resources. We are all familiar with hearing in the news about ambulances, occupied with patients transported to hospital from the community waiting, sometimes for many hours, to handover their patient to staff in A&E. Until such time as that crew is clear from their previous job that resource is unavailable for dispatch. On occasions this can result in no resource being available to dispatch or there being a need to choose or prioritise one Category 1 over another Category 1. It is suggested that any delay that results from this situation is very unlikely to result in a finding of breach of duty. This delay, consequent upon inadequacy of resource, is likely to be regarded as unavoidable. It will only be if resourcing can, in and of itself, be objectively demonstrated to be insufficient to enable the Trust to meet its obligations to exercise reasonable skill and care that this could potentially result in a finding of breach of duty. An example may be where the number of ambulances in service is patently insufficient to properly serve a population, perhaps lower than had previously been operated. It is suggested that such situations will be vanishingly small in number.

Delays after dispatch

Assuming that an ambulance is then available and is dispatched, operation performance standards require a specific response to be achieved. Performance standards (formerly performance targets) have been in operation since the 1970s. For Category 1 and 2 calls, which are the type most likely to be relevant for the purposes of this article, there is a stated mean arrival time and a 90th centile time.  However, these are performance standards that are generally applicable to all calls and not a specific requirement for any given response. So, if any particular response takes longer than the mean or perhaps even is more than the 90th centile response time, that will not be any real evidence of breach of duty.

There will be a huge number of variables, including but not limited to the geographical area, the split of rural over urban locations, the location of the specific emergency, its location relative to the location of available resources, the weather, the day of week, the time of the day, the impact of traffic, roadworks, obstructions etc. Evidence may be required to address what might reasonably be expected having regard to any particular variables that apply to the given situation and against which the actual arrival time can be assessed. It is suggested that it would have to be a significant and unexplained delay beyond this to support a breach of duty, as it was in Kent.

As long ago as March 2018 a review of the 2017 ARP concluded that “there remains scope to improve the accuracy, speed and efficiency of early prediction still further and that a group should be established to continually review and make recommendations about the future design of pre-triage processes.” This may provide opportunity to explore what has been done, if anything, in the years since 2018 to improve the system’s design and efficiency. This may be a more useful enquiry within an inquest setting, where the Coroner may be more interested in missed opportunities and preventing future deaths as opposed to matters which are strictly causative.


In the context of civil litigation, it is clear that whilst an apparent delay in the arrival of an ambulance may be easy to identify in theory, proving that that was due to a breach of duty will continue to prove a real hurdle to the success of such claims.

It is suggested that there may be a role for expert evidence to address, from either side of the litigation fence, the breach issue. Ideally this would come from someone with particular knowledge and experience of the call handler’s function and the computer programme that they operate. I have seen opinions from Dr Tim Kilner a Senior Lecturer in Paramedic Science, albeit more in the context of the conduct of paramedics once they have arrived at the scene. I see no reason why expertise could not be sought on the process prior to the point of arrival where that is a potential issue.