Are prehospital deaths from trauma and accidental injury preventable ...

Injury, Int. J. Care Injured 48 (2017) 978–984

Contents lists available at ScienceDirect

Injury journal homepage: www.elsevier.com/locate/injury

Full length article

Are prehospital deaths from trauma and accidental injury preventable? A direct historical comparison to assess what has changed in two decades G.J. Oliver* , D.P. Walter, A.D. Redmond Humanitarian and Conflict Response Institute, Ellen Wilkinson Building, Oxford Road, University of Manchester, Manchester, M15 6JA, UK

A R T I C L E I N F O

A B S T R A C T

Article history: Accepted 20 January 2017

Background & objectives: In 1994, Hussain and Redmond revealed that up to 39% of prehospital deaths from accidental injury might have been preventable had basic first aid care been given. Since then there have been significant advances in trauma systems and care. The exclusion of prehospital deaths from the analysis of trauma registries, giv en the high rate of those, is a major limitation in prehospital research on preventable death. We have repeated the 1994 study to identify any changes over the years and potential developments to improve patient outcomes. Methods: We examined the full Coroner’s inquest files for prehospital deaths from trauma and accidental injury over a three-year period in Cheshire. Injuries were scored using the Abbreviated-Injury-Scale (AIS1990) and Injury Severity Score (ISS), and probability of survival estimated using Bull’s probits to match the original protocol. Results: One hundred and thirty-four deaths met our inclusion criteria; 79% were male, average age at death was 53.6 years. Sixty-two were found dead (FD), fifty-eight died at scene (DAS) and fourteen were dead on arrival at hospital (DOA). The predominant mechanism of injury was fall (39%). The median ISS was 29 with 58 deaths (43%) having probability of survival of >50%. Post-mortem evidence of head injury was present in 102 (76%) deaths. A bystander was on scene or present immediately after injury in 45% of cases and prior to the Emergency Medical Services (EMS) in 96%. In 93% of cases a bystander made the call for assistance, in those DAS or DOA, bystander intervention of any kind was 43%. Conclusions: The number of potentially preventable prehospital deaths remains high and unchanged. First aid intervention of any kind is infrequent. There is a potentially missed window of opportunity for bystander intervention prior to the arrival of the ambulance service, with simple first-aid manoeuvres to open the airway, preventing hypoxic brain injury and cardiac arrest. © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Prehospital Preventable death Trauma Injury

Introduction In 1994, Hussain and Redmond revealed that up to 39% of prehospital deaths from accidental injury might have been preventable had basic first aid care been given [1]. Their study was prompted by the 1988 Royal College of Surgeons (RCS) retrospective study of 1000 trauma deaths reporting on the management of patients with major injuries [2]. The RCS report highlighted significant deficiencies in the management of seriously injured patients and concluded that up to one third of hospital trauma deaths could have been prevented had the response from the emergency health system been optimal. Their report however

* Corresponding author. E-mail address: [email protected] (G.J. Oliver).

excluded deaths that occurred outside of the hospital; the assumption being that such deaths would be inevitable. Hussain and Redmond explored whether this assumption was correct. Over the two decades since their study there have been significant changes made to improve the care of the injured patient. Development has been driven by a series of reports from the RCS and the British Orthopaedic Association [3] and journal papers that highlighted deficiencies and underperforming areas in UK trauma care [4–6]. Following the National Confidential Enquiry into Patient Outcomes and Death (NCEPOD) 2007 report “Trauma: who cares?” which identified major deficiencies in both organisational and clinical elements of trauma care [7], there has been a significant push towards developing regional trauma networks [3]. From April 2012, a network based Major Trauma System has been introduced across England and these changes have already been shown to have impacted positively upon patient outcome [8].

http://dx.doi.org/10.1016/j.injury.2017.01.039 0020-1383/© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

G.J. Oliver et al. / Injury, Int. J. Care Injured 48 (2017) 978–984

Within these changes have been major advances in prehospital trauma care. The prehospital ambulance Emergency Medical Service (EMS) systems have evolved with the aim of improving performance and outcomes in the prehospital phase in key areas including new techniques, tools and procedures. These include airway management, circulatory access, control of bleeding, physiological monitoring, scoring systems and training [9]. Where death is clear, according to agreed guidelines, paramedics are also now able to declare death at the scene. Profound changes have also occurred in the nature of, and population affected by, trauma. Analysis of major injury over the last two decades within the UK by the Trauma Audit and Research Network (TARN) demonstrates an increase in the mean age of patients affected and a change in the predominant mechanism of injury from road traffic injury to low falls [10]. Rigorous analysis of hospital trauma care by TARN continues to drive progress and, following a major European collaborative effort, the uniform use of standardised inclusion and exclusion criteria, data variables and definitions have enabled accurate comparison of performance between trauma registries [11]. The exclusion of prehospital deaths from the analysis and scrutiny of trauma registries is a major limitation to the development of prehospital research [11]. Despite developments and the scrutiny of the general trauma population, remarkably little is known about changes in the demographic of the victims, nature or preventability of prehospital traumatic death. Since Hussain and Redmond’s 1994 study highlighting the preventability of prehospital deaths, only one study has made this topic its focus in the UK [12]. Death from trauma remains the leading cause of death for patients in the first four decades of life [13] and a significant percentage of deaths following trauma occur in the prehospital phase. A recent large retrospective cohort study of trauma deaths in seven Western US regions showed 37% of deaths to occur within the prehospital phase [14] but figures up to 86% have been reported [15]. In a recent study by Bakke et al., of the 86% of traumatic death occurring in the prehospital phase, the major causes identified were unchanged and remained suicide and road traffic injury. Preventable deaths are an important performance indicator [16] and measure of health service quality [17]. Analysing preventable deaths can also aid the development of mitigation strategies, improve delivery of care [18] and identify new strategies for treatment, equipment, training, and technology [19]. The prehospital care community must ensure equal rigour is applied to its research as is applied to hospital trauma to identify areas for improvement, drive change and improve outcomes. We, and others [20,21], are of the opinion that there is still a critical period between the point of injury and the arrival of the EMS, where there remains an opportunity for early intervention to save life. Within this period, which we refer to as the therapeutic vacuum, there is potential for simple first-aid manoeuvres to open the airway and possibly prevent hypoxic brain injury and cardiac arrest [1,22], apply external pressure on areas with visible profuse blood loss [20,21] and prevent hypothermia [20,21].

Aim The principal aim of the study was to assess the nature of prehospital deaths from trauma and accidental injury and their cause, two decades after the original published analysis, and identify whether there have been changes in the nature, cause and preventability of death. The secondary aim was to identify what type of interventions might prevent prehospital deaths from injury, improve patient outcome and inform first aid and prehospital education.

979

Methods In order to facilitate a direct historical comparison, this study replicated the methodology described by Hussain and Redmond [1]. All sudden, unexplained and violent or unnatural deaths that occur in the UK are referred to Her Majesty’s (HM) Coroner for the jurisdiction within which the death occurs for investigation. We wrote to several of HM Coroners in and around the area covered in the original study detailing the study and requesting their support. HM Senior Coroner for Cheshire agreed to provide access to support the study by granting access to all relevant records in his jurisdiction. Ethical approval for this study was obtained from the University of Manchester Research Ethics Committee. We retrospectively reviewed the inquest record for deaths due to injury within the Cheshire jurisdiction, occurring in a three-year period, and referred to the Coroner for investigation. Deaths occurring between 1st January 2011 and 31st December 2013 were chosen to ensure analysis of recent deaths whilst also allowing time for the completion of each inquest. We examined in detail the full inquest, including the Coroner’s report, police and ambulance statements, witness reports, the pathologist’s report and the post mortem record. The inclusion criteria were deaths within the specified three-year period occurring in the prehospital phase of care from traumatic or accidental injury. Deaths were categorised as found dead (FD), where injury was not witnessed and death was declared immediately on discovery of the body. Deaths were categorised as dead at scene (DAS) where bystanders were present at the time of injury or prior to death but the patient was not transported. Deaths were categorised as dead on arrival (DOA) where death occurred following transportation from scene but prior to hospital or when declared within the Emergency Department (ED) without successful return of spontaneous circulation. Exclusion criteria were: death in hospital, deaths due to hanging and drowning (the mechanism of death is asphyxia and not injury) and deaths from injury in the community which occurred following discharge from hospital, following treatment of the injury, on an end of life pathway or advanced care plan e.g. not for resuscitation. Deaths investigated by the Coroner occurring overseas but resident in Cheshire were also excluded since they did not reflect the UK system. For those deaths that met the inclusion criteria, cases were anonymised and given a unique study number, and data recorded using a standardised data sheet. Data on age, gender, dominating type of injury, mechanism of injury, intention of injury and prehospital EMS times were collected using the definitions agreed by the European Trauma Registries [11]. Using the autopsy report, a list of injuries sustained, airway patency, information on co-morbidities, presence of alcohol or other drugs at the time of death and cause of death was compiled. Information on how the injured was discovered and if a bystander was present was also extracted from available reports and statements, including details on the call for assistance and whether any first aid intervention of any kind was attempted. The call for assistance was not considered as physical intervention and was recorded separately. Any bystander intervention to help the injured was recorded as first aid. Best estimates were made from the available information on time from injury to discovery and the time from discovery to the call for assistance. Injuries were coded using the Abbreviated Injury Scale (AIS). A single author (GO) received training in AIS coding by the American Association of Automotive Medicine (AAAM) and coded all injury data. Hussain and Redmond used AIS-1990 to code their injury severity data; the Injury Severity Scores (ISS) and probability of survival (Ps) estimations were calculated from these. To enable direct accurate historical comparison, the now historical AIS-1990

980


was used. Using the AIS scores, an Injury Severity Score (ISS) was calculated for each case. The same method to calculate the probability of survival used by Hussain and Redmond [1] was applied. Using Bull’s probits, we identified the cases in which the age and ISS indicated a greater than 50% chance of survival. The method described by Yates [23] to identify airway obstruction and applied in Hussain and Redmond’s [1] study was also applied in our study. When the pathologist indicated that a substantial amount of blood, vomit or other material was present in the upper respiratory tract airway obstruction was taken to be present [23]. The main intended outcome measures were the same as for Hussain and Redmond to enable a direct comparison. These were: Injury Severity Score (ISS) calculated using injury data coded with AIS-1990, probability of survival estimation using Bull’s probits and the presence of neurological injury and airway obstruction on post mortem analysis. Results In the Cheshire jurisdiction, in the period from the 1st January 2011–31st December 2013, five hundred and sixty-four violent deaths were referred to the Coroner for investigation. Four hundred and thirty deaths met our exclusion criteria, of which, four hundred and one were initially excluded with an additional twenty-nine excluded on further analysis. One hundred and thirty four pre-hospital deaths were included (Fig. 1). Of the one hundred and thirty-four deaths studied, one hundred and six were male (79%) with an average age at death of 50.8 years. The remaining twenty-eight were female (21%) with an average age at death of 64 years. The average age at death overall was 53.6 years. Analysis of injury intention revealed ninety-six to be accidental, thirty-seven to be self-inflicted and one to be due to assault. Of the one hundred and thirty-four deaths, sixty-two were found dead (FD), fifty-eight died at scene (DAS) whilst fourteen were dead on arrival (DOA). Breakdown of deaths by mechanism of injury is displayed in Table 1. The most common mechanism of injury was fall (39%); 48% of these were “low falls” from the same level (body height), and the remaining from above body height. The spread of ISS is shown in Table 2; the median ISS sustained was 29. An ISS of 75 indicates maximal injury. It can result from an AIS-5 injury in three body regions or from an AIS-6 injury in one body region. AIS-6 injuries are catastrophic and are considered by

many as unsurvivable [10]. Twenty-nine of the prehospital deaths had an ISS score of 75; in all twenty-nine deaths an ISS score of 75 was due to an AIS score of 6 in at least one body region (100%). In twenty-one cases an AIS-6 injury was coded in the Head/Neck body region, in sixteen cases for the Chest body region and in one case for the Abdomen. An ISS > 15 is the definition of major trauma. In 22 deaths (16%) the ISS was below this score. Probability of survival estimation using Bull’s probits indicated that fifty-eight (43%) cases had a probability of survival that was greater than 50%. In thirteen of these there was evidence of airway obstruction at autopsy. In a further fourteen cases there was evidence of significant ingestion of alcohol at the time of death. The twenty-two cases with ISS < 15 were included in this number. The role of bystanders in these fifty-eight deaths is shown in Table 4. The remaining seventy-six were considered unsurvivable (57%); twenty-nine cases had a maximal ISS score of 75 and were considered unsurvivable, while the remaining forty-seven had ISS scores that indicated that the probability of survival was less than 50%. Neurological injury was found in one hundred and two cases (76%). Post mortem evidence of airway obstruction was noted in twenty-six cases, twenty of which were associated with neurological injury (77%). Bystander presence on scene, during or immediately after the point of injury, was recorded in sixty cases (45%). A passer-by was on scene prior to arrival of the EMS in an additional sixty-eight cases (96%). The time from injury to discovery varied from cases where a bystander was present on scene immediately, to discovery days after injury (Table 3). The call for emergency medical assistance was made in eleven cases by someone directly involved in the incident, in one hundred and fourteen cases by a bystander or passer-by, in eight cases by a member of another emergency service, and in one case no call was made. In six deaths no bystander was present. Bystander intervention was attempted in thirty-three cases (25%) and not attempted in one hundred and one (75%). When only those DAS or DOA are considered and those FD are excluded, bystander first aid of any kind was attempted in thirty-one of seventy-two (43%) cases. Prehospital time points were found and extracted from fifty cases. In these cases the time from alarm to the arrival of the emergency medical services was on average 8.4 min, median 6.5 min (IQR 5–12).

Fig. 1. Flow diagram detailing inclusion and exclusion of cases.


981

Table 1 Breakdown of mechanism of injury. Traffic: motor vehicle injury (car, van, heavy transport vehicle, bus) Traffic: motorcycle injury Traffic: bicycle injury Traffic: pedestrian Traffic: other (ship, airplane, train) Shot by handgun, shotgun, rifle or other firearm of any dimension Stabbed by knife, sword, dagger, other pointed or sharp object Struck or hit by blunt object (Tree, bar, stone, human body part, metal, other) Low energy fall (fall at the same level) High energy fall (fall from a higher level) Other: Burns Other: Industrial accident

Table 2 Variation in Injury Severity Score. ISS

Number

0 < ISS < 16 16 ISS < 21 21 ISS < 31 31 ISS < 41 41 ISS < 51 51 ISS < 61 61 ISS < 71 ISS = 75

22 11 37 14 11 7 3 29

Table 3 Time from injury to discovery. Bystander present immediately Minutes From minutes to an hour Hours Days

60 9 11 43 11

Discussion Several significant changes in the nature of prehospital deaths between the 1987–1990 North Staffordshire and 2011–2013 Cheshire cohorts have been noted. A potential limit to the historical comparison is that the same Coroner’s jurisdiction has not been compared. However, figures from the office for national statistics show the neighbouring counties of Cheshire and Staffordshire to rank very closely alongside each other in population, area and density, though the Stoke on Trent and North Staffordshire Coroner’s jurisdiction is smaller than the Cheshire Coroner’s jurisdiction that we studied. Based on the Ministry of Justice 2011 Coroners’ statistics, three thousand six hundred and twenty two deaths were reported with one hundred and nine receiving a verdict of death due to accident and misadventure in

13% (17) 45% (6) 2% (3) 8% (11) 13% (17) 5% (6) 8% (10) 2% (2) 20% (27) 19% (25) 6% (8) 2% (2)

Stoke-on-Trent and North Staffordshire. In Cheshire four thousand seven hundred and eleven deaths were reported with one hundred and sixty one from accident and misadventure. A significant change was observed in the number of deaths excluded due to hanging. Of the four hundred and ninety-seven violent deaths reported in North Staffordshire between 1987 and 1990, thirty (6%) were deaths due to suicide by hanging and excluded. This figure is significantly higher in Cheshire between 2011 and 2013. Of the five hundred and sixty-four violent deaths reported, one hundred and thirty-nine (25%) were deaths from suicide by hanging and excluded (p < 0.05). The number of prehospital deaths with probability of survival greater than 50% has not changed significantly since the previous analysis remaining high at 43% from 39% (p = 0.45). A major criticism of Hussain and Redmond’s study was the methodology used to determine the probability of survival [24]. It was correctly pointed out that Bull’s data were based only on hospital deaths and were then some 30 years old. Our results therefore do not confirm that 43% of prehospital deaths are preventable but rather, through replication of methodology and direct historical comparison, that this figure has not changed significantly. This suggests that despite dramatic improvements in the care of the injured patient, no improvement has occurred in minimising the number of deaths considered as preventable by Hussain and Redmond. Future work involving probability of survival estimation should use updated and refined methods. Reliably reporting which deaths would be judged preventable by current standards is not currently possible. A recent systematic literature review of studies on preventable prehospital deaths by the authors highlights the heterogeneity in methodology, terminology and definitions of “preventable” between studies on prehospital trauma death [25]. There is no current consensus on what can be considered preventable in the prehospital phase, and there has been a call for an agreed ontology and consensus on the definition of “preventable” and how this is determined. It must include consideration of which time interval or phase, and under what circumstances, the death occurred. Within an ontology, the criteria used to judge the appropriateness of care delivered, both for opportunities to improve care and for the detection of errors as

Table 4 Bystander role in the 58 deaths with a probability of survival greater than 50%.

Bystander Presence

Call for assistance

Bystander Intervention

During/immediately after the point of injury

Not immediately on scene, arrival prior to EMS.

No bystander present

19 (33%)

36 (62%)

3 (5%)

Person involved

Bystander

Emergency services

No call

0 (0%)

53 (91%)

4 (7%)

1 (2%)

Attempted

Not attempted

19 (33%)

39 (67%)

982


part of preventability categorisation, would vary according to the phase and so provider of care and circumstances within the prehospital period. The reliability of the calculation of probability of survival has advanced significantly through the work of the trauma registries [26]. The logistic regression model based on age, gender, Injury Severity Score (ISS) and Glasgow Coma Score (GCS) developed by TARN [27] and repeatedly updated to remain relevant to the changing trauma population is a significant evolution away from Bull’s probits. The fundamental message from Hussain and Redmond’s paper was the potential opportunity for the application of simple airway manoeuvres such as ‘jaw thrust’ that could open the airway and possibly prevent hypoxic cardiac arrest following injury. This was based on their results showing a very high percentage of cases with neurological injury (74%), a high number of cases with post mortem evidence of airway obstruction (59%) and a strong association between the presence of airway obstruction and neurological injury (80%). This message has been echoed by others [22,28]. Our results indicate that neurological injury remains present in a high percentage of cases (77%) and, whilst the correlation between airway obstruction and neurological injury remained high (77%), the presence of evidence of airway obstruction at post mortem was significantly lower (19%, p < 0.05). The reliability of using neurological injury and airway obstruction at post mortem as the indicator for missed opportunity and need for airway support can be questioned. Increasing attention is being given within the prehospital community to a phenomenon known as impact brain apnoea [29]. This describes a physiological response to brain injury, which can’t necessarily be detected in gross pathological changes at post mortem. Within the critical phase of head injury, arbitrarily defined as the first ten minutes following injury [30], concussive pressure transmitted to the brainstem can lead to apnoea and a subsequent stress-related massive sympathetic discharge. The combined effects of hypoxia, hypercarbia, acidosis and blood pressure surge, as well as direct catecholamine effects on tissue lead to synergistic injurious effects. The first responder at the scene of injury must provide ventilatory support early, or multi-systemic hypoxic injury and cardiac arrest will occur. Physician’s eye witness accounts have documented the dramatic alteration of clinical course in apnoeic patients with GCS 3 and fixed dilated pupils following head injury [31] with the delivery of bystander ventilator support. The critical phase of head injury, and how we respond to it, determines life or death at the scene [30]. Even with good prehospital response times, which averaged 8.4 min, in cases where the call for assistance is made immediately after injury; what occurs before arrival, in the time-critical period of the therapeutic vacuum, lies in the hands of bystanders. Within the therapeutic vacuum following the point of injury, a key role performed in the majority of cases (93%) by bystanders, and fundamental to the EMS response, is the call for assistance. There is a key opportunity for the Emergency Dispatch Centre to guide bystander intervention with the delivery and effectiveness of post-dispatch instructions (PDI). Studies focussing on this phase within the pre-hospital period suggest improved outcomes are achieved with dispatcher assistance [32]. Whether directed by PDI or prior first aid training, the role of the bystander or passer-by is crucial. The importance of airway support [20,30] and controlling bleeding [21,33] through compression on areas of visible profuse blood loss, to prevent early death have been highlighted. Both the techniques of simple airway management and pressure on visible bleeding are routinely taught during first aid training [21]. Our data demonstrated that a bystander was on-scene immediately or within minutes in 52% of cases. This is consistent

with the one study found by Tannvik et al., [20] in their recent review on first aid provided by laypeople to trauma victims, which reported a bystander presence of 59% [34,35]. However, a recent Norwegian study by Bakke et al. found bystander presence at 97% [36]. Bystander intervention in any form was only found in thirtythree cases (25%) in our study. Tannvik et al. reported that the frequency of any kind of first aid being delivered by bystanders ranged from 11% to 65% [21,37]. Bakke et al. recorded much higher rates with correct airway intervention and haemorrhage control in 76% and 81% of appropriate cases respectively [36]. When cases where the patient was found dead are excluded and only patients who died on scene or had died on arrival to the Emergency Department are considered, bystander intervention of any kind was still only attempted in thirty-one (43%) cases. Our results highlight a significant missed opportunity for bystander intervention and of attempts at any form of basic life support. This is particularly apparent in the fifty-eight deaths with a probability of survival of greater than 50% where intervention was attempted in only nineteen (33%) cases. Estimations from other studies suggest that an increased focus on layperson first aid is likely to be worthwhile [20]. Two autopsy based studies estimated that layperson bystander first aid intervention could lead to a reduction in mortality of 4.5% [21] and 1.8% [33]. Ambulance-based studies reporting on the relationship between the frequency and quality of bystander trauma care indicate that the frequency and quality of first aid intervention decreases in the more severely injured patient [35] but that the correct delivery of first aid increases with higher levels of first aid training [34]. The potential for first aid to improve patient outcomes following injury shown by Hussain and Redmond led to discussions with the Ministry of Transport regarding implementation of first aid training as part of learning to drive. The continued potential benefit for first aid and the key role fellow citizens can play following injury, raises the need to reassess whether first aid training should be integrated more into everyday life. Bakke et al. reported that 35% of bystanders overall (87.5% of those with a known training status), had received first aid training. What level of first aid training should be delivered in prehospital trauma care, or concerns by the lay public that incorrect first aid might adversely affect the patient, can be addressed by further study. Significant changes within the prehospital trauma population are seen between the two studies in the age of the population affected, and in the dominant mechanism of injury. These changes mirror changes already documented in the hospital trauma population [10]; whilst these changes might be expected they have not been demonstrated in the prehospital population. The average age of death across both sexes has increased from 41.9 years to 53.8 years. In the hospital trauma population the average age increased from 36.1 in 1990 to 53.8 in 2013 [10]. The elderly comprise the most rapidly expanding section of the population and advances in medical care mean that many enjoy better health, mobility and independence longer than ever before, thereby increasing their exposure to the possibility of injury. The dominant mechanism of injury has shifted dramatically from road traffic accidents in Hussain and Redmond’s study; reduced from 57% to 28% in our study. Falls only contributed to 15% in the original study but have risen to 39% to become the dominant mechanism of injury. This shift is also one that is well documented in the hospital trauma population, where the most common causative mechanism of injury has changed from road traffic injury, causing 59% of injury in 1990, to low falls, which made up 39% in 2013 [10]. Road and vehicle safety has increased whilst the risk of mortality has increases with age. The probability of significant injury as a consequence of falls, lower physiological reserve and the ability to recover are also diminished with age [38].


983

Conclusion

Reference

The population affected by, and major mechanism of, prehospital death from trauma and accidental injury has changed dramatically in line with observed and documented changes in the hospital trauma population. The number of potentially preventable prehospital deaths seems to have remained high and unchanged, in spite of major advances in trauma care over the previous two decades. Bystanders witness the event in the majority of cases but, while making the call for assistance, first aid intervention of any kind is infrequent. It is clear that there is a window of opportunity following injury for the provision of bystander first aid before the arrival of the emergency medical services that could potentially improve outcomes, particularly in the context of head injury. Lay person airway management and ventilatory support during the current therapeutic vacuum could significantly improve survival. The prehospital care community needs to form a consensus on how to define and determine the preventability of death in the prehospital phase to enable detailed study of each of the prehospital phases in order to improve current practice.

[1] Hussain LM, Redmond AD. Are pre-hospital deaths from accidental injury preventable. BMJ 1994;308:1077–80. [2] Saleh M. Commission on the Provision of Surgical Services: the management of patients with major injuries. Ann R Coll Surg Engl 1989;71:58. [3] Sleat G, Willett K. Evolution of trauma care in the UK: current developments and future expectations. Injury 2011;42:838–40. [4] Yates DW, Svoboda P, Kantorova I. The influence of medical care on the death rate from trauma in england and south moravia. Eur J Trauma 2002;28:304–9. [5] Davenport RA, Tai N, West A, Bouamra O, Aylwin C, Woodford M, et al. A major trauma centre is a specialty hospital not a hospital of specialties. Br J Surg 2010;97:109–17. [6] Gabbe BJ, Biostat GD, Lecky FE, Bouamra O, Woodford M, Jenks T, et al. The effect of an organized trauma system on mortality in major trauma involving serious head injury: a comparison of the United kingdom and victoria, australia. Ann Surg 2011;253:138–43. [7] (NCEPOD) NCEiPOaD. Trauma: who cares? 2007. [8] Cole E, Lecky F, West A, Smith N, Brohi K, Davenport R, et al. The impact of a pan-regional inclusive trauma system on quality of care. Ann Surg 2016;264 (1):188–94. [9] Williamson K, Ramesh R, Grabinsky A. Advances in prehospital trauma care. Int J Crit Illn Inj Sci 2011;1:44–50. [10] Kehoe A, Smith JE, Edwards A, Yates D, Lecky F. The changing face of major trauma in the UK. Emerg Med J 2015;32:911–5. [11] Ringdal KG, Coats TJ, Lefering R, Di Bartolomeo S, Steen PA, Røise O, et al. The Utstein template for uniform reporting of data following major trauma: a joint revision by SCANTEM, TARN, DGU-TR and RITG. Scand J Trauma Resuscitation Emergency Med 2008;16:7. [12] Limb D, McGowan A, Fairfield JE, Piggot TJD. Prehospital deaths in the Yorkshire health region. J Accid Emergency Med 1996;13:248–50. [13] Pfeifer R, Tarkin IS, Rocos B, Pape HC. Patterns of mortality and causes of death in polytrauma patients–has anything changed. Injury 2009;40:907–11. [14] Dean D, Wetzel B, White N, Kuppermann N, Wang NE, Haukoos JS, et al. From 9-1-1 call to death: evaluating traumatic deaths in seven regions for early recognition of high-risk patients. J trauma Acute Care Surg 2014;76:846–53. [15] Bakke HK, Wisborg T. Rural high north: a high rate of fatal injury and prehospital death. World J Surg 2011;35:1615–20. [16] Teixeira PG, Inaba K, Hadjizacharia P, Brown C, Salim A, Rhee P, et al. Preventable or potentially preventable mortality at a mature trauma center. J Trauma-Inj Infect Crit Care December 2007;63:1338–47. [17] Rutstein DD, Berenberg W, Chalmers TC, Child 3rd CG, Fishman AP, Perrin EB. Measuring the quality of medical care. New Engl J Med 1976;294:582–8. [18] Vioque SM, Kim PK, McMaster J, Gallagher J, Allen SR, Holena DN, et al. Classifying errors in preventable and potentially preventable trauma deaths: a 9-year review using the Joint Commission's standardized methodology. Am J Surg 2014;208:187–94. [19] Kelly JF, Ritenour AE, McLaughlin DF, Bagg KA, Apodaca AN, Mallak CT, et al. Injury severity and causes of death from operation iraqi freedom and operation enduring freedom: 2003–2004 versus 2006. J Trauma 2008;64:S21– 26 discussion S6–7. [20] Tannvik TD, Bakke HK, Wisborg T. A systematic literature review on first aid provided by laypeople to trauma victims. Acta Anaesthesiol Scand 2012;56:1222–7. [21] Ashour A, Cameron P, Bernard S, Fitzgerald M, Smith K, Walker T. Could bystander first-aid prevent trauma deaths at the scene of injury? Emergency Med Australasia: EMA 2007;19:163–8. [22] Trunkey D. Towards optimal trauma care. Arch Emerg Med 1985;2:181–95. [23] Yates DW. Airway patency in fatal accidents. Br Med J 1977;2:1249–51. [24] Nicholl J. Pre-hospital deaths from accidental injury. BMJ 1994;309:57. [25] Oliver GJ, Walter DP. A call for consensus on methodology and terminology to improve comparability in the study of preventable prehospital trauma deaths: a systematic literature review. Acad Emerg Med 2016;23:503–10. [26] Lecky F, Woodford M, Edwards A, Bouamra O, Coats T. Trauma scoring systems and databases. Br J Anaesth 2014;113:286–94. [27] Bouamra O, Wrotchford A, Hollis S, Vail A, Woodford M, Lecky F. A new approach to outcome prediction in trauma: a comparison with the TRISS model. J Trauma 2006;61:701–10. [28] Ryan M, Stella J, Chiu H, Ragg M. Injury patterns and preventability in prehospital motor vehicle crash fatalities in Victoria. Emergency Med Australasia: EMA 2004;16:274–9. [29] Wilson MH, Hinds J, Grier G, Burns B, Carley S, Davies G. Impact brain apnoea – a forgotten cause of cardiovascular collapse in trauma. Resuscitation 2016;105:52–8. [30] Atkinson JL. The neglected prehospital phase of head injury: apnea and catecholamine surge. Mayo Clin Proc 2000;75:37–47. [31] Levine JE, Becker D, Chun T. Reversal of incipient brain death from head-injury apnea at the scene of accidents. N Engl J Med 1979;301:109. [32] Bohm K, Vaillancourt C, Charette ML, Dunford J, Castren M. In patients with out-of-hospital cardiac arrest, does the provision of dispatch cardiopulmonary resuscitation instructions as opposed to no instructions improve outcome: a systematic review of the literature. Resuscitation 2011;82:1490–5. [33] Henriksson E, Ostrom M, Eriksson A. Preventability of vehicle-related fatalities. Accident; analysis and prevention 2001;33:467–75.

Limitations There are several limitations to our study. Different Coroners’ jurisdictions were studied in this direct historical comparison. The jurisdictions were felt to be comparable, HM Senior Coroner for Cheshire kindly agreed to support this study and undertaking the study in Cheshire was logistically achievable. Some data were subjectively interpreted with single author best estimates based on the available information; this was particularly the case with data on time from injury to discovery based on available information from the witness statements and emergency services reports. Due to the direct historical comparison, the now historic Bulls probits and AIS 1990 coding dictionary were used. These do not provide accurate indication of injury severity or probability of survival estimation by today’s standards and have no role in future research. Conflict of interest All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf and declare: no support from any organisation for the submitted work; GO, DW and AR have received research grants from the British Red Cross through the University of Manchester, GO received non financial support in the form of travel and course paid by the British Red Cross to attend the American Association of Automotive Medicine Abbreviated Injury Scale coding course deemed essential to conduct the research; no other relationships or activities that could appear to have influenced the submitted work. Funding The British Red Cross provided funding for this research activity. It had no involvement with the design, undertaking or writing of this review. Acknowledgements We thank Mr N Rheinberg, HM Senior Coroner for Cheshire, and his staff for their support and help with this study.

984


[34] Pelinka LE, Thierbach AR, Reuter S, Mauritz W. Bystander trauma care–effect of the level of training. Resuscitation 2004;61:289–96. [35] Thierbach AR, Pelinka LE, Reuter S, Mauritz W. Comparison of bystander trauma care for moderate versus severe injury. Resuscitation 2004;60:271–7. [36] Bakke HK, Steinvik T, Eidissen SI, Gilbert M, Wisborg T. Bystander first aid in trauma prevalence and quality: a prospective observational study. Acta Anaesthesiol Scand 2015;59:1187–93.

[37] Khorasani-Zavareh D, Haglund BJ, Mohammadi R, Naghavi M, Laflamme L. Traffic injury deaths in West Azarbaijan province of Iran: a cross-sectional interview-based study on victims' characteristics and pre-hospital care. Int J Inj Contr Saf Promot 2009;16:119–26. [38] Bruijns SR, Guly HR, Bouamra O, Lecky F, Lee WA. The value of traditional vital signs, shock index, and age-based markers in predicting trauma mortality. J Trauma Acute Care Surg 2013;74:1432–7.