Pre-hospital management of mass casualty civilian shootings: a systematic literature review

A systematic literature search was undertaken on PubMed, Cochrane Database of Systematic Reviews and Scopus for the period of 1 January 1980 to the 20 April 2016. Medical Subject Headings terms were combined with non-indexed relevant search words to create a set of sensitive entry criteria (Table 1). Second, a systematic search of the grey literature was conducted using the Luxembourg Definition for grey literature [7]. The same entry criteria displayed in Table 1 as for the published literature were used for Web of Science, and simple search criteria were used for OpenDOAR and Evidence Search. The search was conducted over the same time period. The references from included papers were checked for additional material not found on the original search.

For the inclusion criteria the manuscripts had to contain descriptions, discussions or experiences of the pre-hospital management of civilian mass shootings as defined in the introduction, and had to be written in English, published between the 1 January 1980 and 20 April 2016 and be relevant to pre-hospital immediate medical management. Manuscripts were excluded if they only discussed mass casualty incidents (MCIs) due to blast injury.

One author (CDAT) screened the titles or abstracts of identified literature. Literature clearly not complying with the inclusion criteria was excluded. Eligibility was assessed by full text reading of the uncertain papers and inclusion was subject to consensus with a second author (MR). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline was followed [8].

Data extraction focused on identifying common themes in the case reports and other manuscripts, recording information under ten agreed questions on a Microsoft Excel Spreadsheet Version 14.5.1 (2011; Microsoft Corporation, USA) Quality was appraised using a predefined checklist of questions depicting internal and external validity and were evaluated against the Oxford Centre for Evidenced-based Medicine (CEBM) Levels of Evidence [9].

Among the 494 identified manuscripts, 73 were selected on abstract and title and after full text reading 47 were selected for inclusion in analysis (Fig. 1). Seven papers could not be retrieved.

Study design in the included manuscripts comprised case reports (n = 14), review articles (n = 4), commentaries (n = 14), government reports (n = 5), information technology report (n = 1), consensus statements (n = 2) and original research papers on matters relevant to pre-hospital management of mass casualty civilian shootings (n = 7) [1, 2, 10‐54]. The search revealed heterogeneous primary source information on seventeen mass shooting events, the majority from the USA and others from Norway, the UK, Kenya and France (Table 2). According to the Oxford UK CEBM Levels of Evidence the literature was of levels 5 (n = 27), 4 (n = 18) and 3b (n = 2).

The reports described the shooting of 1649 people with 578 deaths at 17 separate events. Fifteen events happened between 2005 and 2015 and include ten single shooter incidents and four multiple shooter attacks. Three events involved concurrent explosions (Fig. 2 and Table 3).

Fourteen incident sources describe the importance of disaster preparedness planning. Eight stated that triage was an area for improvement and four that the use of experienced practitioners achieved better triage. Thirteen incidents had communication failures. Scene safety was highlighted as an issue for twelve of the mass shootings; five of these supported Tactical Emergency Medical Services (TEMS) (Table 3).

Though many organizations have estimated the prevalence of mass shootings, accurate figures are unavailable. In the USA there have been between 62 and 78 mass shooting events between 1983 and 2013 according to the FBI definition; this involves the shooting of between 1023 and 1056 casualties, among whom 507–547 died [2, 55]. A crowd-sourced US website, which refers to an incident with four people shot rather than people killed, reports considerably higher numbers exclusively for the USA; 330 mass shooting events in 2015, 283 in 2014 and 363 in 2013 alone. Although not comprehensively documented there have been a number of devastating attacks recorded in Europe and across other continents.

The Joint Committee to Create a National Policy to Enhance Survivability from Mass Casualty Shooting [56] reported a number of key messages on pre-hospital management of mass shootings summarized by the acronym THREAT, which stands for threat suppression, haemorrhage control, rapid extraction at the scene for assessment by medical providers referring to triage and then finally transport to definitive care.

At the Utoya Island shooting in 2011, limited interagency communication and a persisting threat by the shooter resulted in EMS staff unable to get onto the island and access casualties for two hours and seven minutes [48] Similarly in the report on the Columbine Shootings of 1999, injured victims were still stranded in the school over two hours after the shooters had committed suicide. Furthermore due to a breakdown in communication a medical team almost came under friendly fire by the Special Weapons And Tactics (SWAT) team [36]. In the UK after the Cumbria shootings in 2010, Chesterman’s report identified interoperability of the police and ambulance as a “highly significant issue leading to extensive delays of ambulances reaching patients and police being left to transport the injured” [57]. This report concluded that “it would be reasonable for the public to expect the ambulance service to attend scenes where there is a residual risk” even though the police service will not be able to guarantee the safety of the staff [16].

Tactical Emergency Medical Support (TEMS) refers to “a non-military medical emergency service modified for the realities of the tactical environment” [58]. TEMS range from personnel trained to work in “hot” or “warm” zones, to those that operate in the safe areas but have specialist training within law enforcement in order to facilitate a greater level of co-operation between the two agencies in a time of extreme pressure. The concept of EMS operating in hazardous environments is changing [42, 59].

Supporting evidence found in this review for the effectiveness of TEMS is mostly anecdotal and from case reports [11, 15, 33]. However, a comprehensive outcome-based report from Finland shows that TEMS integrated into daily EMS could provide effective treatment in law enforcement operations only in safe working zones [60].

A recent concept by Autrey et al. in Minnesota called 3 Echo, “Enter, Evaluate and Evacuate”, looks at a streamlined method for extraction and administration of life-saving haemorrhage control to victims of active shooters. The events of the 27 September 2012 in Minneapolis put “3 Echo” into action [11]. The implementation of this framework on the day led to the safe evacuation of all critically ill patients in a time, to the knowledge of the authors, unmatched for events of this kind. The main principles of this framework are improved communication between services, a shared goal of early identification of casualties by the first wave of law enforcement and establishing safety corridors for evacuation. Instead of clearing large geographical areas such as an entire school, corridors of safety are established as a means of early access to and evacuation of casualties even before the attack has ended.

Haemorrhage is the leading preventable cause of death in trauma and causes 30–40 % of fatalities [61]. The primary principle of the Hartford Consensus was that nobody should die from uncontrolled bleeding [26]. Pre-hospital medical management at mass shootings is addressed in military settings by Tactical Casualty Combat Care (TCCC) guidelines prioritising the control of catastrophic haemorrhage [23, 24, 27, 51]. These principles are transportable to civilian response where TEMS or trained police officers need to be trained to administer care in the hyper-acute setting [42]. In Paris a damage control strategy was employed, which focused on maintaining a mean arterial pressure >60 mmhg using tourniquets, administration of tranexamic acid and prevention of hypothermia [23]. Damage control resuscitation aims to correct the triad of coagulopathy, acidosis and hypothermia, which can exacerbate bleeding [62]. Remote or “Ground zero” damage control in a pre-hospital setting includes reducing haemorrhage with tourniquets, haemostatic dressings and pelvic binders, minimising hypothermia by reducing exposure, and using thermal blankets and fluid resuscitation with red blood cell transfusion [63]; however, the current literature provides limited evidence for the use of pre-hospital blood products in a civilian setting [64].

Triage at mass shooting incidents can be complicated by the ongoing threat and the mechanism of injury [36]. Superficial wounds with minimal visible bleeding may be overlying extensive internal damage and internal haemorrhage and the same is true for the reverse; severe surface injuries do not always indicate extensive damage of internal organs [10]. The inherent challenges with gunshot wound triage are exacerbated by the volume of patients that can be seen in a mass shooting and particular problems are associated with paediatric incidents [17]. Physiological triage utilises vital signs, can be done quickly and requires little experience. In contrast, an anatomical approach relies on accurate diagnosis of injuries and may not be possible without imaging or surgery [65]. Currently the methods of triage for mass shootings are not validated by reliable evidence [66]. In many of the papers included in this review, triage was identified as an area of potential improvement or concern [13, 30, 48, 49, 53, 54].

Triage is a healthcare resource allocation system used when available resources are potentially insufficient to address the needs of patients at a mass casualty event. It attempts to provide the maximum benefit for the most casualties with the available resources [67, 68]. The mass shootings reported in the results section used a mixture of the START triage system [30], local physiological-based triage systems [54], unspecified methods of triage by experienced medical staff [48], an anatomic site of bullet entry triage methods [53] and, in some cases, moving patients rapidly to nearby hospitals where triage occurred after arrival [46].

At the Virgina Tech shooting in 2007 the rate of over-triage was 69 %, measured as patients assigned priority 1 at triage, who actually had an Injury Severity Score (ISS) of less than 15. High levels of over-triage may deprive severely ill patients of the resources needed to treat them, and also put pressure on the surge capacity of the trauma hospital. The Fair Child Massacre also noted a high over-triage rate and expressed the need for a triage tool that could address this problem [13]. Both the Fair Child and Columbine School massacres recorded the issue of uninjured ambulatory patients, processed as “priority 3 delayed”. This occupied the triage resources and prevented the identification of the severely injured. The Fort Hood report concludes that a lack of scene safety, chaotic triage organisation and communication failures led to fear amongst the emergency staff, which compromised triage and led to the inappropriate or under-triage of several patients [54]. A second attack on this base in 2014 reported similar triage issues, noting that ongoing clinical reassessment was critical [49].

The benefit of using experienced pre-hospital experts who have seen large numbers of critically ill patients was reported by those at Utoya [48]. Since this incident a national manual for mass casualty triage has been created in Norway [69]. Two other incidents reported the benefits of having senior surgeons play a pivotal role in pre-hospital triage; the terrorist attacks in Jerusalem and the Dark Knight cinema shooting in Aurora Colorado [32, 45]. Furthermore both also supported the concept of keeping initial triage simple [31, 54].

Allocation of patients to appropriate services and maintaining good flow in these hospitals is important to avoid saturation of the receiving hospitals. Poor communication at the incident can lead to errors in patient transport, delivering patients to inferiorly equipped hospitals [52]. It has been shown that admission to a level-1 trauma centre reduces mortality for the most severe trauma patients [70]. At the Westgate shooting patient flow was maintained by the increased availability of porters and by creating of a one-direction system, keeping the emergency department clear for arriving casualties [53].

Major incident planning and preparedness focuses on the cross-service organisation of infrastructure to facilitate good communication, command and control, safety at the scene, accurate triage and the transport of patients in unpredictable environments [65].

The reviewed literature links disaster preparedness to a successful response. A lack of familiarity and unpreparedness predisposes to confusion and delay [11, 13, 23, 24, 28, 30, 32, 36, 49]. The low mortality rate in the Virginia Tech shooting of 2007 was credited to the training and collaborative planning that had been a focus since the 11 September 2001 attacks [30]. After the Aurora Dark Knight cinema shootings of 2012, the authors agreed that these incidents cannot be rehearsed too frequently and supported the idea of tabletop exercises [32]. The responders to the Minneapolis workplace shooting in 2012 had undertaken a 12-hour training curriculum, which involved tabletop exercises and hands-on, walk-through scenarios, which were highly realistic involving replica weapons and ammunition. The 3 Echo protocol training resulted in triage and evacuation of patients with reported unprecedented efficiency in a subsequent mass shooting emergency response. The interagency exercises that occurred on the morning of the attacks in Paris were considered a key factor in the success of treatment; the large hospital network coordinated by Assistance Publique-Hôpitaux de Paris was also highlighted as a strength [23].

Simple measures to ensure preparedness such as having dedicated medical equipment stores for major incidents and the value of equipment such as lightweight stretchers for the transport and tracking of injured patients have been recommended [11, 48, 71]. Targeted training can have a significant impact on perceptions of EMS preparedness. One study demonstrated that 41 % of responders felt prepared to respond to an active shooting incident before training, increasing to 89 % after tactical awareness training [29].

Communication failures are a consistent feature of post-major incident reports and at mass shootings these failures can have detrimental impacts [13, 16, 19, 30, 48, 49, 54]. At Fort Hood two patients were transported to hospitals with inadequate trauma facilities due to loss of communication [54]. Columbine in 1999 experienced problems with radio “dead zones” and a miscommunication meant that at one point it was believed that eight, not two, shooters were active in the school [36]. False alarms at the Virginia Tech shooting in 2007 caused ambulances to be evacuated unnecessarily [50]. Failure of the Norwegian Acute Medical Information System at logging communications caused issues in tracking patient pathways and destination hospitals. Tracking problems were also reported by Kaplowitz at Virgina Tech 2007 [30, 48]. Having backup communications equipment, pre-arranged disaster channels for communication and establishing communication protocols with public area authorities, such as train stations, schools and large buildings, has been suggested as a means to mitigate these issues [11, 36, 54].

Drawing direct comparison between similar incidents is sometimes difficult and anecdotal evidence in individual reports means the authority of the recommendations were uncertain in some cases. There are no suitable validated quality appraisal tools so a non-validated checklist was used. Non-English literature was excluded and there was no meta-analysis of data. Government papers and inquiries focus on either the operation of law enforcement on the motives of the shooter, and less emphasis is given to the emergency medical response [16, 19].

Uniform and comprehensive reporting of mass casualty shootings is required, along with a dedicated recording database for the occurrence of these incidents outside the USA [5]. Epidemiology in this global phenomenon is incomplete; clinical experience from shootings in many countries has not been published, including high profile events like the Sousse attacks in Tunisia 2015. This review is based on low-quality retrospective case report analysis, and anecdotal expert opinion. There is a need to design high-quality reporting and research in this area. There is likely to be reporting bias in this topic due to its sensitivity, and some documentation may be classified due to involvement of counter-terrorism agencies. Strengths of this review include having two authors with extensive pre-hospital experience, the use of a systematic approach and the incorporation of additional information found in the grey literature.