Major incident management by helicopter emergency medical services in south‐east Norway from 2000 to 2016: Retrospective cohort study

Helicopter emergency medical services (HEMS) and search and rescue helicopters (SAR) aim to bring specialized personnel to major incidents and transport patients to definite care, but their operational pattern remains poorly described. We aim to describe the use of HEMS and SAR in major incidents in Norway and investigate the feasibility of retrospectively collecting uniform data from incident reports.

Helicopter emergency medical services (HEMS) and search and rescue helicopters (SAR) have the potential to contribute to major incident management with transportation of equipment, personnel and patients as well as providing overhead surveillance and scene search. 3,4 A previous cross-sectional survey of all Norwegian HEMS and SAR crew members found that they seldom attended major incidents, the doctors had attended on average one whereas the rescue paramedic and pilot had attended three incidents. 5 Norway is a subarctic country, with scattered population where transport distances may be long and challenged by fjords and mountains. There is a publicly funded health care system where HEMS and fixed-wing air ambulance are part of a national air ambulance system. SAR are integrated in the air ambulance system and operated by the Royal Norwegian Air Force, but used primarily as a civilian resource. There are 12 HEMS and seven SAR bases in Norway, all staffed with a consultant anaesthesiologist, a rescue paramedic and pilot(s) and with similar medical equipment set-up.
In addition, SAR are staffed with a flight mechanic and a navigator.
Dispatch is subject to unitary coordination causing great overlap in catchment/operating areas. When required, the services have additional equipment on-base for use in incidents with special needs, for example avalanche. HEMS/SAR can provide advanced pre-hospital treatment and often has senior competence to make medical and tactical decisions. Ambulance, police and fire services are in close inter-disciplinary cooperation in most incidents in Norway. The personnel on-scene informs the emergency medical command centre what resources are needed for coordination and allocation of additional rescue services.
In an attempt to collect uniform data on HEMS/SAR use in major incidents, a consensus-based template for the use of HEMS and SAR in major incidents was developed in 2016. 6 The aim of the present study was to conduct a retrospective cohort study of Norwegian HEMS and SAR major incident management describing how HEMS and SAR are used in major incidents, their tasks and challenges to improve future management and preparedness. Furthermore, we aimed to investigate the feasibility of retrospectively collecting uniform data from incident reports.

| Setting
In this retrospective cohort study, we searched the medical database LabasNG (Normann IT) from three HEMS bases and one SAR base, for reports covering major incidents in the period from 2000 through 2016 (inclusive). The HEMS bases Lørenskog, Ål and Arendal together cover urban, mountain and coastal terrains and were thus assumed to be representative of the Norwegian HEMS. Lørenskog has two helicopters at disposal. Arendal, Ål and Rygge have one helicopter each. The SAR base at Rygge is considered a good representative of the SAR service in Norway with a mission profile of both ambulance-and SAR missions.

| Eligibility criteria
A major incident was defined as "an incident that requires the mobilization of extraordinary EMS resources and is identified as a major incident in that system." 1 In Norway, this means that the extent will vary according to resources available in the district were the incident occurs. Urban areas have more resources available; hence, they can potentially handle more patients than rural districts before extraordinary EMS resources are mobilized. Rural was defined as "characteristic of the countryside rather than the town" and urban was defined as "relating to, or characteristic of a town or city." 7

| Incident selection
LabasNG is a proprietary relational database management system. No data fields, tick-boxes or other descriptors denote a major incident.
Identification of major incidents can only be processed via free text searches. Initial mapping by International Classification of Diseases (ICD) diagnosis (Data S1) removed the incidents that clearly did not fit the description. Aborted and rejected missions were excluded as they cannot be identified as major incidents in the current registry. One author (ASJ) manually searched the remaining reports for possible eligibility. MR and MS evaluated the free text sections of all potentially eligible reports for inclusion. In cases with divergent opinions, SJS was consulted and consensus was sought through group discussion.

| Data collection
When a major incident was identified, we collected data according to variables defined in major incid entre porti ng.org 6

Editorial Comment
This report describes recent major incidents in a region where there are physician-manned helicopter ambulances.
Major incidents appear to be rare in South-East Norway according to this retrospective study. HEMS units also appear to play a major role in their management in that region.
information available in free-text area or tick-boxes; "Medium": information available in both free-text areas or tick-boxes; but more vulnerable to rater variability and "Poor": not possible to find information without a degree of speculation from the authors or not found at all.

| Ethics
The Regional Committee for Ethics in Medical Research concluded that ethical approval was not needed and gave exemption from the duty of confidentiality with the condition that no person would be recog-

| RE SULTS
The search produced a total of 31 803 missions for the study period.

F I G U R E 1 Mission flowchart
After exclusion by ICD diagnosis, aborted and rejected missions 21 524 All recorded missions 31 803 After initial screening (one author) 265 After 2nd screening (two authors) Total included (after removing duplicates) 50 Disagree 161 Included 70 Excluded 34 Included after consensus (all authors) 27 We identified only three incidents (6%) where HEMS/SAR was the first medical resource on scene, but in 33 incidents (66%) they brought the first (or only) doctor. In two of the incidents (4%), HEMS/ SAR was the only resource in the acute phase, a train accident in a mountainous area not accessible by road and a helicopter crash in a mountainous region.

| Challenges for HEMS/SAR
Weather was considered a hazard on-scene in 7 (14%) and on-going fires in 6 (12%) incidents. Difficult landing site was the most common challenge (n = 5, 10%), but in the majority of incidents there were no reported hazards. Communication problems were reported in 6 (12%) incidents (see Table 4 for a summary reported challenges).  Blunt injuries were the most dominating injuries (n = 37, 74%).

| Inclusion of reported, unreported and missing data
Multiple questions from the template and the survey were overlapping (6 from the template and 12 from the survey) and others were general background information (8 and 34, respectively). The majority of information was found in the free-text area where the anaesthesiologist reported a description of the incident, response and patient treatment. This is subject to rater variability. The availability ratio "Good":"Medium":"Poor" was 13:12:3 (Data S2, column D-F).
Data depicting coordinating roles and triage remain unreported, as this was not systematically recorded in LABAS. Road traffic accidents (RTAs) were the most common type of incident and summer the busiest season, echoing findings from other studies. [9][10][11] Norway is a country dominated by rural areas in a sub-arctic environment with potential for decompensated scenes given the austerity of the environment. The capacity to manage a major incident varies with local resources and is why we differentiated urban and rural incidents. A majority of incidents occurred in rural areas as these resources are more easily overwhelmed. Other countries will have different profile of distances, HEMS/EMS coverage and crew combination, but RTAs will probably be a leading cause of trauma and a warm climate may make them more prone to major incidents. 12 Arguments for a more widespread use of ground units may be wise in some countries, but considered not so relevant in Norway. The Norwegian population is scattered and transport distances are long and challenged by fjords and mountain areas, making HEMS/SAR effective in reducing transportation time for severely injured patients in rural areas. HEMS/SAR are vulnerable to weather 13-15 but in most incidents there were no recorded hazards or safety challenges. Aircraft crowding and "Hot zone" hazard were all related to the twin-terrorist attack in the governmental building and Utøya island. 16 This was the largest incident in this material both regarding resources and persons involved, injured and dead thereby being an outlier in our data. 16,17 Although HEMS/SAR are seldomly the first crew on-scene, they often bring the first doctor. 18,19 The first crew on scene will often have a role in keeping overview, triage and perform logistical and tactical communication with the other agencies.

| D ISCUSS I ON
Furthermore, the other crews will focus on the most severely injured patients identified by first crew on-scene. 19 The median number of helicopters participating in major incidents was three, Transportation from scene to casualty clearing station 2 (4%) Transportation from scene to trauma unit 12 (24%) Transportation from scene to regional trauma centre 26 (52%) Transportation from casualty clearing station to trauma unit 3 (6%) Transportation from casualty clearing station to regional trauma centre

(6%)
Transportation from trauma unit to regional trauma centre  18,20,21 Norway has no official policy on "Stay and Play" vs "Scope and Run." This depends on the condition of the patient, provider competence and transport time to hospital. All HEMS/SAR transports to hospital were from rural incidents. HEMS/SAR may contribute with transport of personnel and equipment to scene, although this study shows that HEMS/SAR rarely bring additional equipment.
When needed, this may be brought by civil protection services and non-governmental organizations. In the majority of included major incidents, other rescue agencies were present. When a major incident occurs, multiple agencies with different roles operate in parallel in chaotic environments. 22,23 Therefore, it is important to have implemented major incident management plans and ensure that inter-agency training frequently occur.
In this study, we wanted to investigate the feasibility of retrospectively collecting uniform data from the incident reports. We originally planned to include information regarding triage and coordinating roles. We interpreted from free text field annotations that informal major incident triage has been performed, but the application of formal triage standards was not described. The Norwegian standard for mass-casualty triage was developed during the study period and was published in 2013. 24 The complexity of defining a major incident remains a controversy in the field of disaster medicine research where several definitions exist and no definition is uniformly accepted. 1,25,26 We applied the definition used in the previous cross-sectional study and Delphi study in which the variables in the current study originated. 1,5 The definition focus on medical major incidents but as this study shows, all rescue services work together in the complexity of a major incident. We have not been able to quantify other rescue services participating as the current registry provides no information on this.
There is no exact space that mentions major incidents. The prehospital experience and knowledge of Norwegian geography in the author group were used to achieve consensus on which incidents to include.
There are multiple reporting templates available. 27 The EMS society should agree on a common template to enable more homogenous data reporting as major incidents are rare and prospective studies will be hard to conduct. This may inflict recall bias and the quality of the entered data varies.
HEMS/SAR will naturally record data on patients they treat and transport, but not patients handled by other rescue organizations. Median NACA of all patients involved in major incident will probably be lower as missing data most likely occur in patients with lower NACA score.
The score was set by the doctor reporting in LABAS and is a subjective score for patient severity. Although it may be subject for rater variability, it has shown to reliably predict mortality and the need for advanced interventions. 28 This was a retrospective study and we may have missed incidents, thereby underestimating our reported major incident incidence. Unfortunately, the current data system does not allow analysis of aborted or rejected mission requests and incidents where helicopters did not participate because of weather, technical issues etc remain unknown. The total number of patients involved in the major incidents included is difficult to establish, as the exact number not always was reported.

| CON CLUS ION
Major incidents are rare and operations are characterized by extensive inter-disciplinary cooperation. HEMS play a central role in medical management and should be included in major incident plans.
Future research should focus on systematic data gathering and a system for sharing lessons learned for major incident planners to make resilient plans that include HEMS/SAR involvement and help HEMS/SAR crews identify important areas of training.

ACK N OWLED G EM ENTS
We thank Prof. Jo Røislien for statistical quality assurance and comments on final manuscript.