Identification of the technical and medical requirements for HEMS avalanche rescue missions through a 15-year retrospective analysis in a HEMS in Switzerland: a necessary step for quality improvement

We screened all HEMS primary missions from Rega-Swiss Air Ambulance between January 2001 and May 2016. [13] We included data from all patients caught by an avalanche (regardless of the burial degree), reached within 24 h of the alarm being raised, and transported by the HEMS. Individuals who were not caught by an avalanche were excluded (e.g. bystanders).

The Swiss hospital network contains 12 level-1 trauma centres, eight of which have extracorporeal life support capabilities. Rega-Swiss Air Ambulance is the major HEMS in Switzerland and has 13 bases distributed along the Alps and near to the five university hospitals in the country. Every base is equipped with a single helicopter dedicated to HEMS missions and continuously available 24/7. The standard crew includes a pilot, a paramedic, and an emergency physician. As for mountain rescue missions, a specialist from the Swiss Alpine Rescue may be added to the standard crew. In the specific case of avalanche rescue missions, the HEMS helicopter is however most of the time dispatched directly to the avalanche site to minimize the response time, and an additional helicopter from a commercial company is dispatched simultaneously to pick up an incident commander, mountain rescue specialists and avalanche dogs if required.

Data were retrospectively analysed from the digital dataset of Rega-Swiss Air Ambulance. [14] The dataset is prospectively registered for resource management and activity documentation purposes after every mission, by the pilot for the operational part and by the physician or paramedic for the medical part. The operational data included the month and year of the avalanche accident, the response time (defined as the time interval between the alarm and the arrival of the first HEMS helicopter on site), the method used to evacuate the patient from the avalanche site, the patient’s destination, and the identification number of the crewmembers involved.

Medical data included the age and gender of the patient, the pre-hospital presumptive diagnoses made by the HEMS physician, and the diagnostic and therapeutic measures applied on site or during transport. Hypothermia is an item in the predefined list of presumptive diagnoses of the database and is either determined clinically by using the Swiss staging system, or by measuring a core temperature < 35 °C with an esophageal probe, which is the only available device in our HEMS. [15] The severity of the patient’s condition was graded using the eight-level pre-hospital NACA severity score, which is assigned by the HEMS physician at the end of the rescue mission (Additional file 1). [16, 17]

Text fields of the data set were searched for the following avalanche-specific important information: burial degree, burial time, and information on the patient’s location and extrication. Complete burial was defined as the burial of at least the head and chest. [3] We defined multiple casualty accidents as those with more than one caught person. We defined summer avalanches as those occurring from June 1st to October 31st. [4, 18]

We aimed to describe the technical and medical requirements of avalanche rescue missions in terms of search, rescue, and medical competencies of the HEMS crewmembers, and to evaluate their clinical exposure to avalanche rescue missions and victims. Based on these baseline measurements, we aimed to identify and discuss operational and medical areas of avalanche rescue missions that should be the focus of QI efforts, and suggest potential QI strategies.

The data retrieved from the database were exported to Stata version 14 (Stata Corporation, College Station, TX, USA). Descriptive statistics included frequencies, mean, and standard deviation (SD), or median, interquartile range (IQR), and range as appropriate. Groups were compared using Pearson’s Chi² or Fisher’s exact tests, as appropriate. A bilateral p-value < 0.05 was considered to indicate a significant difference.

Avalanche rescue missions frequently occur in terrain that is difficult to access and, therefore, winch operation was necessary to evacuate 29% (n = 121) of the patients. This confirms the feasibility of winching both medical competencies and patients representing medical emergencies. [8] Although this information was not specified in the database, we can presume that the physician may mostly have disembarked the helicopter while hovering, as winching was used for the patient and medical crew evacuation.

Burial time is considered as the most important prognostic factor for completely buried avalanche victims, whose survival drops from 91 to 34% as burial time increases from 18 to 35 min. [20‐22] Thus, prompt companion rescue can increase the survival rate. [2, 21, 23, 24] Companion rescue is, however, not always successful. [21] Being often the first professional rescuers on scene, our crews were involved in locating and extricating patients in almost half of the avalanche accidents. Although we could not exclude an under-reporting of the extrication of victims by companions, this proportion is higher for our study than that reported by Mair et al., who found that only 12% of victims were still completely buried when the rescue teams arrived on scene. This may be explained by the shorter median response time in our study (22 min vs. 35 min for Mair et al.). [2] The response time found in our study also means that our crews arrived on-site during the very critical phase of the avalanche survival curve, during which survival drops dramatically, and during which timely extrication is crucial. [22] Our data, therefore, reinforce the importance of dispatching a HEMS helicopter directly to the avalanche scene, to rapidly deliver professional rescuers. [21, 25] These latter should be highly skilled at locating and digging out still completely buried victims, including in complex situations, such as when multiple casualties are completely buried, which was the case in 14% of the missions in our study.

The patients encountered by our crews showed a wide range of pathologies and severity. Despite our shorter response and burial times (35 min vs. 360 min described by Mair et al.), 29% of our patients were in CA, which is higher than reported by Mair et al. [26, 27] While the latter did not describe burial times for victims extricated by companions, they found a slightly higher rate of victims extricated by companions (62% vs. 59%), which might at least in part explain the different rates of patients in CA.

The management of completely buried avalanche victims in CA is complex and rescuers should follow the dedicated algorithm. [4] Cardiopulmonary resuscitation (CPR) was applied in 51% of the CA patients, whereas this value was 100% in the study by Mair et al. [2] In our study, all patients with a NACA score of 6 and in CA were transported to a hospital under CPR or after the return of spontaneous circulation. However, the low rate of CPR in patients declared dead on site (NACA 7) is worrisome, notably as previous publications have shown that at least some, but possibly a significant proportion of CA victims might have been declared dead on site without sufficient reason. [5, 28]

The patients encountered by our crews showed a wide range of pathologies and severity and required a large panel of procedures, including advanced procedures to be performed on site (e.g. tracheal intubation, pneumothorax decompression, administration of pharmacological agents over an i.v. or i.o. access). This confirms that advanced life support (ALS) is feasible at avalanche scenes, as recommended by international guidelines. [3, 4, 29, 30]

However, the low proportion of our patients who were monitored might appear suboptimal. This might be explained, at least partially, by its feasibility in difficult conditions and the low reliability of some of the equipment under very cold environments. Monitoring of any kind was more frequently used in CA patients, possibly because this is an essential part of their assessment and because of its relatively smaller benefit in less severely injured patients, in a system where a health care professional will constantly be beside the patient. [4, 31] The stress of the situation, as well as the cold environment might explain partially the surprisingly low rate of end-tidal CO₂ (ETCO₂) measurement for intubated patients. Although its monitoring was higher in patients with a NACA score of 6 than in patients with a NACA score of 7 (49% vs 9%) and was also used more systematically since 2010, EtCO₂ should be monitored in every intubated patient, as it allows to confirm the tube placement and to evaluate the efficiency of CPR. Core temperature was available for only 19% of CA patients. This can be partially explained by the fact that its measurement is considered mandatory only for CA patients with long burial times. [3, 4]

As new technologies are being developed to improve the reliability of monitoring under cold environments, other devices have come up to facilitate the work of the rescuers. Among these, two have been introduced at our HEMS during recent years with a direct implication in the management of avalanche victims: (1) an intraosseous access device, which is a valuable and efficient alternative to intravenous access in cold environments, and therefore especially attractive in avalanche rescue and; (2) a mechanical chest compression device that has been shown not only to allow efficient CPR under difficult conditions, but also contributes to the safety of the rescuers by allowing them to quickly clear the avalanche scene with a patient in CA, or performing on flight CPR, as in the case of hypothermic avalanche victims. [7, 25, 32‐35] Those two devices (and especially the mechanical chest compression device) have been used frequently since their introduction. This confirms the feasibility and attractiveness of their use in the setting of avalanche rescue missions.

In our study, only 27% of the injured patients were immobilized with a vacuum mattress, and 11% had a cervical collar. In a previous study, a spine fracture was found in 7% of avalanche victims and was confirmed to be the cause of death in 5.6% of cases. [30] This apparently low rate of immobilization could be partially explained by the application of evolving international guidelines, suggesting that rapid extrication, transportation and stabilisation of the patient should be prioritized over immobilization. [36, 37] Crewmembers should also rely on solid clinical experience, as “best practice” management must be balanced with the environmental context.

Finally, in accidents involving multiple casualties (32%) or multiple CA patients (5%), complex medical triage skills are required to allocate the available human, material, and transport resources to those patients with the greatest chance of survival. [38, 39]

We showed that our crews have extremely low clinical exposure to avalanche rescue missions and victims. To our knowledge, it is the first time that data on this are made available. This finding is important considering the variety of technical and medical skills that need to be applied under challenging conditions.

In our HEMS, physicians had a proportionally lower exposure than paramedics. As most of the paramedics spend 100% of their time with the HEMS and are assigned there for several years, most of the shifts are covered by physicians who are assigned there for a limited period (6–12 months), with the remainder of the shifts being covered by physicians with a part-time activity over several years (15 shifts/year to 50% activity rate).

The annual probability of the Rega physicians and paramedics to be involved in an avalanche rescue mission is proportional to their occupation rate and depends on the HEMS base where they are employed. The probability is higher in the HEMS bases located closer to the mountains (6 bases having 1–4 missions/year) than in the bases located near the big cities (5 bases having less than 1mission/year).

Although the probability is highest in the two HEMS bases that cover 60% of the avalanche accidents (and having 5–8 missions/year), the number of missions and managed patients remains low when distributed among the providers. However, our data is insufficient to analyse whether exposure influenced patient outcome.

The data presented in this study allowed us to identify three areas (HEMS operations and technical rescue, medical management of avalanche victims, data collection for avalanche rescue missions) that should be the focus of QI efforts and to propose multifaceted interventions for HEMS operating avalanche rescue missions in similar settings (Table 4). [11, 12]

As a preliminary requirement, rescuers intervening on an avalanche scene should be comfortable in a mountainous environment and trained to work in difficult outdoor conditions.

The goal of any intervention in the field of operational and technical aspects of avalanche rescue is to improve time efficiency while maintaining or increasing safety. Considering that our response times correspond to the asphyxia phase, reducing them may directly and favorably influence time to extrication, and thus also patient outcome. Thus, it may be monitored as a quality indicator of avalanche rescue in the future.

On an organizational level, avalanche rescue readiness should be high, both from a material and individual point-of-view. Individual (e.g., clothes, avalanche beacon) and avalanche-specific search and rescue materials (e.g., helicopter’s external avalanche beacon, RECCO® device, probes, shovels) should be readily available and checked (e.g., through the use of a checklist, on a daily basis at the beginning of the shift), and their use incorporated into regular training by the crewmembers.

Additionally, standard operating procedures (SOPs) for avalanche rescue should be available, incorporated in training and used. Next to the mandatory personal safety equipment, we suggest that the physician should systematically wear the harness dedicated to winch operations when on avalanche rescue missions. These measures may improve time efficiency and increase safety by relieving the minds of the rescuers from organizational concerns and allow them to concentrate more on risk assessment and on the technical and medical aspects of the rescue. [6, 29]

On an individual level, specific training should be carried out to achieve a high level of performance in locating and digging out of avalanche victims, including single and multiple casualty scenarios. The crewmembers should also be regularly trained in winch procedures, to increase safety and efficiency.

Our data, as those of other studies, suggest potential improvements in the medical management of avalanche victims. [5, 28] Because of the low clinical exposure to avalanche victims and complexity and specificity of their management, we focused our initiatives on provider education and decision-making support. For example, guidelines and clinical practice recommendations (or even medical SOPs) could be disseminated to the providers at the beginning of the winter. Also, targeted practical training can be achieved through workshops and field training. [40] Additionally, indoor simulations for the management of avalanche victims, as well as the sharing with all the providers of the HEMS of experiences collected through a continuous audit of avalanche rescue missions may partially substitute the lack of experience. [41‐44]

The use of the existing Avalanche Victim Resuscitation Checklist as a clinical decision support could improve adherence to the recommendations, as well as with information transmission on the scene and to the hospital teams. However, this has not been evaluated yet. [4, 9, 10, 25, 45]

Finally, research in the field of avalanche rescue and victim management is hindered by the low incidence of cases, as well as the heterogeneity, quality, and completeness of the collected datasets among the different HEMS. To address this concern, a template for uniform data documentation and reporting for avalanche rescue missions should be developed through consensus of experts. It would allow for better comparison of registries data and upcoming studies among different HEMS. [46]

To measure QI, as well as to provide feedback of clinical performance measurements to the providers and identify new areas that should be the focus of QI efforts, quality indicators for avalanche rescue and medical management of avalanche victims should be developed, monitored and reported prospectively as part of a continuous quality improvement program. [12]

Our study has some limitations. Firstly, the quality and completeness of some data reported may have suffered from the retrospective design of our study. Under-reporting of some variables in the database cannot be ruled out. This may especially be the case for some items that are not must report items, as are the avalanche specific information. The latter were reported on a voluntary basis by the providers into free text fields. This might account for several differences between our data and previous studies. [2, 21] Moreover, the overall reliability of the information regarding most of the items was high, and their completeness was about 100%.

Secondly, diagnoses were only presumptive, and classified in pre-defined categories. Nevertheless, completeness of presumptive diagnoses was high, as this was missing for only two patients with a NACA score ≥ 1. However, considering that diagnoses were made under time pressure, in a difficult environment and with fully dressed patients, significant injuries may have been overlooked, especially in severely injured patients (NACA score ≥ 4) as reported by Hasler & al. [27] Although greater completeness and precision of diagnoses would have been appreciated, this would have required the analysis of hospital or forensic data, which were not available. The absence of definitive diagnosis however did not impact the conclusions of our study, which was specifically quality-oriented.