Article Text

Download PDFPDF

Influence of air ambulance doctors on on-scene times, clinical interventions, decision-making and independent paramedic practice
  1. K Roberts1,
  2. K Blethyn2,
  3. M Foreman3,
  4. A Bleetman4
  1. 1
    County Air Ambulance, Warwickshire and Northamptonshire Air Ambulance, Coventry, UK
  2. 2
    University Hospitals NHS Trust Birmingham, Birmingham, UK
  3. 3
    University of Birmingham, Birmingham, UK
  4. 4
    West Midlands HEMS Group, County Air Ambulance, Warwickshire and Northamptonshire Air Ambulance, Coventry, UK
  1. Dr K Roberts, 77 Three Acres Lane, Dickens Heath, Shirley, West Midlands B90 1NZ, UK; j.k.roberts{at}


Background: Critics of air ambulance doctors question their contribution and believe on-scene time is prolonged. Two helicopter emergency medical service (HEMS) models operate in the West Midlands, one with doctors and the other without. A study was undertaken to compare on-scene time, management and decision-making between the two units.

Method: Cases were assessed over an 18-month period, identifying on-scene time, incidence of rapid sequence induction (RSI), management of patients with a Glasgow Coma Scale (GCS) score of 3, femoral fracture, pneumothorax or those with myocardial infarction.

Results: There were 5275 HEMS activations during the study period. The presence of a doctor had no effect on on-scene time (27 (2) min vs 26 (2) min, p = NS). Advanced management of femoral fractures (nerve block, ketamine or RSI), pneumothorax (chest drain) or RSI (when patients were matched for GCS score) by doctors took no longer than conventional paramedic management. Doctors performed RSI on 38% of trauma patients and 13% of medical patients with a GCS score <15. Patients were more likely to be treated and discharged from the scene when seen by a doctor (8.7% vs 4.6%, p<0.001) and were less likely to be transported to hospital (27% vs 44%, p<0.001). For patients with a GCS score of 3, doctors were more likely to cease resuscitation efforts and confirm death at the scene (70% vs 16%, p<0.001).

Conclusions: Appropriately trained HEMS doctors provide advanced management and decision-making. This is without a negative effect on on-scene time, even when performing complex procedures. They are more likely to declare death or discharge patients at the scene, increasing the availability of this limited resource.

Statistics from

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.

The London helicopter emergency medical service (HEMS) unit, tasked to trauma cases in an urban environment, was established as the first doctor-led air ambulance service in the UK. Throughout the 1990s there was an expansion of air ambulance units, almost exclusively staffed by paramedic crews. More recently, air ambulances have incorporated doctors into the aircrew, being despatched to the entire spectrum of emergency case mix in both urban and rural (remote) environments.

The role of doctors in air ambulance teams remains controversial; critics have cited prolonged on-scene times, expense and ultimately lack of impact on patient outcome.1 2 Potential benefits of doctor-led air ambulance units include early definitive airway control, advanced analgesics/techniques, surgical procedures, provision of medical treatment beyond the scope of UK paramedic guidelines3 and advanced decision-making. Recommendations from the National Confidential Enquiry into Patient Outcome and Death (NCEPOD) report of 2007 include the provision of advanced airway management (including rapid sequence induction (RSI)) in prehospital care.4 Decision-making outside paramedic protocols—such as confirming death or treating and discharging patients at the scene—is an important benefit of experienced doctors in prehospital care but is difficult to quantify.

The Midlands is covered by four air ambulance units. The Warwickshire and Northamptonshire Air Ambulance (WNAA) began operating in October 2003. It is crewed with an appropriately trained doctor (see Appendix 1: essential and required doctor criteria) and paramedic; on occasion it is staffed by two paramedics when a doctor is not available. County Air Ambulance (CAA) was established in 1989 and operates three aircraft that serve the West Midlands, Herefordshire, Gloucestershire, Worcestershire, Shropshire, Staffordshire and the East Midlands. Each aircraft is staffed by two paramedics. At times, paramedic-only crews on either service are supported by land-based prehospital doctors of varying skill base (BASICS doctors).

This study evaluated the effects of doctors attending medical and traumatic cases on air ambulance units in a mixed rural and urban environment. We sought to identify whether the presence of a doctor on the air ambulance team has an impact on the tasking of the aircraft by ambulance control, procedures performed, length of time on-scene and patient disposal.

Paramedics on both air ambulances working in the absence of doctors are required to work within the Joint Royal College Ambulance Liaison Committee (JRCALC) guidelines.3 We therefore also sought to determine whether independent practice differs between paramedics who do and those who do not routinely work alongside air ambulance doctors.


Data were collected prospectively onto separate databases, one at WNAA and one for the three aircraft operating under the CAA. The study period was 1 September 2005 to 1 March 2007 (18 months). All three aircraft operating under the CAA are considered as one operational group. Missions attended by the crews in cars were excluded.

CAA is tasked via a central control desk based at West Midlands NHS Ambulance Service. WNAA is tasked by either Warwickshire or Two Shires NHS Ambulance Service. Neither air ambulance group has a dedicated clinical crew identifying missions suitable to task the aircraft.

Missions were categorised as road traffic collision (RTC), “other” trauma, medical or leisure (equestrian or sports injuries). In the results, trauma refers to all RTC, “other” trauma and leisure.

Missions were concluded as follows:

  • “stand down” if the team was activated but cancelled before reaching the patient;

  • “aborted due to weather” if the helicopter took off from base but had to turn back due to adverse weather;

  • patients “treated and discharged” from scene by HEMS crew, requiring no immediate further medical investigation or treatment;

  • conveyed to hospital by HEMS team;

  • conveyed to hospital by land paramedics but not requiring transport or care by HEMS team;

  • declared dead at scene;

  • unsafe to transport the patient by air due to patient agitation or aggression.

The length of time on-scene was the length of time between the aircraft landing and taking off from the scene (this includes access to patient, transfer to the helicopter and packaging for flight). There is no definition of “entrapment” used in either database. As such, all these cases are considered along with other trauma or RTC cases. Due to the fact that crews were attending incidents within the same environment and the large numbers of cases in these groups, we do not feel a significant bias was introduced between crews.

To evaluate differences in patient management, we identified several life and/or limb-threatening conditions before data analysis: these were myocardial infarction; patients with a presenting Glasgow Coma Scale (GCS) score of 3; femoral fractures; patients undergoing RSI; needle decompression; or chest drain insertion. All paramedic-only missions were searched to identify the presence and activity of land-based prehospital doctors.

For procedures confined to doctor practice (RSI, sedation and surgical procedures), cases were identified in which land-based prehospital doctors attended incidents to support paramedic HEMS crews and on-scene times and procedures were compared with HEMS doctor/paramedic crews.

Analysis of data

Mann-Whitney and χ2 tests were employed. Statistical significance was accepted as p<0.05. Two standard errors of the mean are expressed when comparing lengths of time. SPSS V.15 (SPSS, Illinois, USA) was used to analyse data.


Dispatch and aircraft utilisation

The total number and category of missions, who tasked the aircraft, presenting GCS score and patient demographic data are shown in table 1. Both WNAA crews were more likely to be tasked by control as a primary response (p<0.001); missions were more likely to be medical in nature (p<0.001).

Table 1 Activations, method of dispatch, category of missions, patient demographic data and percentage of patients with an initial decrease of Glasgow Coma Scale (GCS) score by air ambulance crew

Patient demographic data were comparable between the groups, the only significant difference being the greater proportion of patients under 5 years of age seen by doctor-led WNAA (3.7%) compared with the County crews (1.6%). There was no difference between paramedic groups.

To assess the appropriateness of HEMS activation, we analysed the proportion of patients presenting with a decreased GCS score of 3–14. Both WNAA crews were more likely than para-County crews to attend these patients. When patients in (medical) cardiac arrest were excluded, there was no difference.

Differences in tasking appeared to be between different NHS ambulance controls rather than whether a unit had a doctor on board; there was no observed difference in case mix between WNAA crews, regardless of the presence of a doctor.

Patient disposal

There was a slight decrease in the number of occasions DrLed-WNAA crews were stood down before reaching an incident (fig 1). WNAA crews transferred fewer patients to hospital than Para-County. These cases were accounted for by handing a greater proportion of patients over to land ambulance crews for further management and an increase in certifying patients dead at the scene. In addition, DrLed-WNAA crews were more likely to treat and discharge patients at the scene than either paramedic crew.

Figure 1 Disposal of patient by air crews (%). *p<0.05 vs DrLed-WNAA; **p<0.001 vs Para-WNAA and DrLed-WNAA; ***p<0.0001 vs Para-WNAA and DrLed-WNAA; ‡p<0.05 vs Para-WNAA, p<0.0001 vs Para-County; no difference between Para-WNAA and DrLed-WNAA. Para-County, double paramedic crew, County Air Ambulance; Para-WNAA, double paramedic crew Warwickshire and Northamptonshire Air Ambulance; DrLed-WNAA, doctor/paramedic crew Warwickshire and Northamptonshire Air Ambulance.

A small percentage of missions were aborted due to bad weather and, in 10 cases, para-County were not able to transfer the patient safely by air due to patient agitation/aggression. No such instances were recorded with either WNAA crew.

With the exception of patients treated and discharged at the scene, there was no significant difference between Para-WNAA and DrLed-WNAA.

The stand-down rate, patient transportation to hospital by air or land and discharge from the scene was not affected by category of incident (RTC, “other” trauma, leisure or medical; data not shown).


HEMS doctors performed RSI on 41/1140 missions (3.6%). Land-based prehospital doctors supporting paramedic-only air crews performed RSI on 1.2% (8/695) of Para-WNAA missions and 0.35% (12/3440) of Para-County missions (p<0.001 HEMS doctors vs both paramedic groups, p = 0.005 between paramedic groups). The use of RSI by initial GCS score of patients with medical or traumatic injuries is shown in tables 2 and 3.

Table 2 Rapid sequence induction of medical patients according to initial Glasgow Coma Scale score
Table 3 Rapid sequence induction of trauma patients according to initial Glasgow Coma Scale score

On-scene management and decision-making

To explore complex management and decision-making, we reviewed the patients with a presenting GCS score of 3 as a separate group. A much greater proportion of patients were declared dead at the scene by HEMS doctors than by either paramedic group. Para-WNAA were 2.2–4.3 times more likely to declare death than Para-County for both medical emergencies (fig 2) and traumatic injuries (fig 3). There was no significant difference in recording a return of spontaneous cardiac activity.

Figure 2 Management of patients with medical emergency and initial Glasgow Coma Scale (GCS) of 3 by crew (%). *p<0.05 vs DrLed-WNAA; **p<0.01 vs Para-WNAA or DrLed-WNAA; ***p<0.0001 vs DrLed-WNAA; ††p<0.01 vs Para-WNAA. Rapid sequence induction (RSI) at air ambulance paramedic cases performed by land-based prehospital doctor. Para-County, double paramedic crew, County Air Ambulance; Para-WNAA, double paramedic crew Warwickshire and Northamptonshire Air Ambulance; DrLed-WNAA, doctor/paramedic crew Warwickshire and Northamptonshire Air Ambulance; ROSC, return of spontaneous cardiac output.
Figure 3 Management of patients with traumatic injuries and initial Glasgow Coma Score (GCS) of 3 by crew (%). *p<0.05 vs Para-WNAA or DrLed-WNAA; **p<0.01 vs DrLed-WNAA; ***p<0.0001 vs DrLed-WNAA; †p<0.01 vs DrLed-WNAA; ††p<0.01 vs Para-WNAA. Rapid sequence induction (RSI) at air ambulance paramedic cases performed by land-based prehospital doctor. Para-County, double paramedic crew, County Air Ambulance; Para-WNAA, double paramedic crew Warwickshire and Northamptonshire Air Ambulance; DrLed-WNAA, doctor/paramedic crew Warwickshire and Northamptonshire Air Ambulance; ROSC, return of spontaneous cardiac output.

A small proportion of patients were conveyed without intubation by Para-County but no patients were conveyed without intubation by either WNAA crew. Due to inadequate data recording, it was not possible to identify whether patients underwent intubation in 10% and 17% of medical/trauma cases, respectively, with Para-County; these patients were conveyed to hospital.

Invasive thoracic procedures, femoral fracture management and treatment of myocardial infarction

WNAA crews performed invasive thoracic procedures more frequently than CAA crews (table 4). DrLed-WNAA crews preferred formal chest drain insertion over needle decompression in almost every case. All invasive procedures were in patients with traumatic injuries.

Table 4 Managment of pneumothorax, femoral fracture and myocardial infarction by crew

Land-based prehospital doctors assisted Para-WNAA crews more frequently than Para-County crews when managing patients with femoral fracture, and these patients were significantly more likely to receive advanced methods of analgesia or anaesthesia.

There was no significant difference in the number of cases of myocardial infarction diagnosed or the methods of treatment (prehospital thrombolysis or transfer for primary coronary intervention) between crews. On three further occasions, HEMS doctors administered thrombolysis to patients in cardiac arrest. One patient had a return of spontaneous circulation.

On-scene times

There was no significant difference in mean (SD) on-scene times between crew types when attending leisure, medical, trauma or RTC incidents (fig 4). Doctor crews spent no longer on-scene, regardless of the patients’ presenting GCS score (GCS score 15: DrLed-WNAA 28 (2) min, Para-County 28 (1) min, Para-WNAA 27 (2) min; GCS score 3–14: DrLed-WNAA 31 (4) min, Para-County 26 (2) min, Para-WNAA 30 (3) min, all p = NS).

Figure 4 Average length of time at scene by crew and category of incident (2 standards errors of mean shown). No significant difference was seen between any group or category. Para-County, double paramedic crew, County Air Ambulance; Para-WNAA, double paramedic crew Warwickshire and Northamptonshire Air Ambulance; DrLed-WNAA, doctor/paramedic crew Warwickshire and Northamptonshire Air Ambulance; RTC, road traffic collision.

The mean (SEM) on-scene time for patients undergoing RSI was 35 (6) min by HEMS doctors (p = 0.07) and 38 (6) min by land-based prehospital doctors supporting paramedic HEMS crews (p = 0.04).

There was no significant difference between paramedic groups in mean (SEM) on-scene time when needle decompression was performed (Para-County 21 (9) min, Para-WNAA 33 (7) min, p = 0.07). HEMS doctors performed one needle chest decompression with an on-scene time of 26 min. Land-based doctors supporting paramedic air crews took nearly twice as long to perform chest drain insertion as HEMS doctor crews (51 (9) vs 30 (5) min, p = 0.01).

Patients with myocardial infarction given thrombolysis were on-scene no longer than patients in whom it was contraindicated or who were transferred for primary angioplasty (32 (7) vs 27 (6) min, p = 0.17). There was no significant difference between crew types in this respect.

For patients with femoral fractures who received morphine for analgesia, there was no difference in on-scene time between crew types. HEMS doctors providing more advanced analgesia/anaesthesia (ketamine, femoral nerve block, RSI) did so without significantly increasing on-scene time (fig 5). Land-based prehospital doctors tended to prolong on-scene time compared with HEMS doctors and techniques of analgesia/anaesthesia (fig 6).

Figure 5 On-scene time for patients with femoral fracture(s) according to crew and analgesia/anaesthesia (2 standard errors of mean shown). There was no significant difference between any groups. RSI, rapid sequence induction. Para-County, double paramedic crew, County Air Ambulance; Para-WNAA, double paramedic crew Warwickshire and Northamptonshire Air Ambulance; DrLed-WNAA, doctor/paramedic crew Warwickshire and Northamptonshire Air Ambulance.
Figure 6 Mean on-scene time of patients with femoral fracture(s) treated by helicopter emergency medical services (HEMS) or land-based prehospital doctors (2 standard errors of mean shown). DrLed-WNAA, air ambulance doctor/paramedic crew; prehospital Dr, doctor travelled to scene by land vehicle; RSI, rapid sequence induction. *p<0.05.


Appropriately trained doctors working on HEMS units provide a level of care beyond paramedic JRCALC guidelines and bring hospital-level interventions to critically injured and unwell patients. The 2007 NCEPOD report recommends the provision of advanced airway and ventilatory management, including RSI and analgesia in prehospital care.4 Doctors may also provide a higher level of decision-making and patient management at the scene.

In this large cohort of patients attended by air ambulances, we have shown that HEMS doctors provide these advanced skills without adversely affecting on-scene times; this observation is not unique to our unit.5 There was a tendency for a slight prolongation of time at the scene when patients received RSI or advanced analgesia for femoral fractures. These differences were non-significant owing to a broad spread of times which are likely to reflect the high proportion of trapped patients with multiple injuries in these groups. The observed length of time at the scene in this study is comparable with contemporary air ambulance practice in the UK: the average length of time spent on-scene by air ambulance crews assessed in the NCEPOD report was 36.9 min (personal communication, NCEPOD).

Land-based doctors supporting paramedic HEMS crews can provide advanced skills but tend to prolong on-scene time. This might be due to the time required to mobilise them to the scene, but may in part reflect the unfamiliarity between members of the team. Furthermore, the availability and competency of these voluntary land-based doctors to support the air ambulance is inconsistent.6 The provision of RSI in this study reflects this. There is a large difference in the use of RSI in patients with traumatic injuries by land-based prehospital doctors supporting the two groups of paramedics. We suspect that an increased understanding of the role of advanced medical practice, including RSI, and close working relationships with the WNAA doctors (most of whom also respond in their own vehicles) by WNAA paramedics is responsible. It is interesting to note, however, that RSI of patients with medical emergencies is almost purely the remit of HEMS doctors and provides a rationale for tasking suitably trained HEMS crews to medical emergencies.

A potential source of bias in this study is that CAA and WNAA are tasked by different NHS ambulance controls. There is not a clear protocol for HEMS activation used by either control; this is at the discretion of senior control staff or request by land ambulance crews for air ambulance support. Although there is a difference in the case mix (medical, trauma and leisure) between the CAA and WNAA, there is no difference in the percentage of patients with a decreased GCS score when these groups are compared directly. The GCS score correlates with prognosis, need for intervention and also has low interobserver variation.7 8 Furthermore, patterns of blunt and penetrating trauma were directly comparable with over 99% of trauma cases being caused by blunt injuries (data not shown). The lack of difference in case mix or presenting GCS score of patients attended by paramedic-only or doctor/paramedic crews on the WNAA unit shows that the presence of a doctor on the aircraft does not appear to influence dispatch decisions by the local NHS ambulance control.

Tasking of HEMS units to medical cases differs from a dedicated trauma service (such as London HEMS). This reflects the large rural areas served by the service and increases the utility of the aircraft. The tasking of doctor-led crews to these cases permits a shortened time to scene with early provision of advanced medical care. Transfer time for patients in rural environments travelling by air is greatly reduced when compared with a land ambulance. This presumably has a beneficial effect upon outcome. A recent study by Nicholl et al9 showed an association between mortality and increasing distance (and hence time) to hospital for patients transported by land ambulances. Patients not requiring hospital-level care can often be managed with a package of care at home; this is facilitated by the presence of a doctor.

Providing advanced medical skills is often given as the purpose for putting doctors onto air ambulances,10 but another important benefit is decision-making. This is difficult to quantify, although doctors are more confident to declare patients dead or to discharge patients from the scene after providing treatment. The benefits of this are multiple: the aircraft is freed to perform other missions; hospitals are relieved of some workload (often futile resuscitation attempts or patients with trivial injuries); confirmation of death at the scene allows the police to proceed with their business. These decision-making skills appear to be transferred in part to paramedics who regularly work with HEMS doctors. These paramedics are more likely to discharge patients at the scene or declare death than paramedic crews who do not regularly work with doctors, while still working within JRCALC guidelines.

Paramedic-only crews were unable to transport a small group of patients due to patient agitation or aggression. Some of these patients are the most time-critical, hypoxia and head injury being frequently encountered causes. The ability to deliver RSI to these patients has multiple benefits: securing the airway early, optimising oxygenation/ventilation, safe and rapid transport to hospital. Without RSI, these often critical patients are very difficult to manage in an aircraft and are thus sent by land ambulance, further delaying their delivery to hospital.

The observed difference in stand-down rates due to adverse weather conditions between the two groups is likely to reflect differences in the operating procedures of the two aircraft operators.

Traditionally, influence on survival in trauma cases has been the most scrutinised quality marker for air ambulances. Major trauma is still a relatively rare condition, and we have shown that doctor-led air ambulance crews improve the level of delivery of care to these patients without prolonging on-scene times. There are clearly other benefits of carrying an appropriately trained doctor on air ambulances such as advanced analgesia techniques, sedation and the ability to discharge a greater number of patients at the scene. Further work is needed to identify ways of improving the efficiency of dispatch to maximise the impact of this limited resource across the entire spectrum of medical and trauma emergencies.


Essential criteria

  • Full registration

  • Registrar level or equivalent

  • 18 months experience with acutely ill patients (which must include a substantive post in emergency medicine)

  • Regular (weekly) ongoing exposure to acutely/critically ill patients (A&E, acute medicine, acute surgery, anaesthetics/ICU, prehospital medicine)

  • Prehospital medicine experience on a regular basis

  • RSI competency

  • ALS or equivalent

  • DipIMC (within 5 years) or ATLS/BATLS/PHTLS or equivalent

  • APLS/PHPLS or equivalent


  • Experience with critically ill patients within the last 5 years

  • Emergency medicine within the last 5 years

  • Weekly prehospital medicine experience


  • DipIMC within 5 years

  • PHEC

  • Competent in-scene management

  • Competent in extrication techniques

  • Completion of RTC course

  • Completion of advanced driving course

  • Completion of Moreton fire college training or equivalent

  • Competent in prehospital acute trauma management

  • Competent in prehospital acute medical management


View Abstract


  • Competing interests: None.