Extrication Times During Avalanche Companion Rescue: A Randomized Single-Blinded Manikin Study

Abstract

Aims:

This study aimed to determine the time needed for one or two companion rescuers to access, extricate, and deliver cardiopulmonary resuscitation (CPR) to a fully buried manikin during a simulated avalanche burial scenario.

Materials and Methods:

In this randomized, single-blinded study, 18 medical students were required to extricate a manikin manually from a simulated avalanche burial of 1 m in depth, either alone or in teams of two. Each participant performed three consecutive tests with the manikin in three different positions in random order.

Results:

Median time to first manikin contact was 2.5 minutes, median time to airway access 7.2 minutes, and median time to standard position for CPR 10.1 minutes. Overall, the number of rescuers (one compared to two rescuers, 10.5 minutes vs. 9.3 minutes; p = 0.686) and the burial position of the manikin (10.8 minutes vs. 10.6 minutes vs. 8.8 minutes; p = 0.428) had no influence on extrication times. Preexisting training (6.1 minutes vs. 11.0 minutes p = 0.006) and a learning effect obtained during the experiments (12.4 minutes the first test vs. 9.3 in the third test; p = 0.017) improved all extrication times.

Conclusion:

It takes an average of 7 minutes after location of a simulated avalanche victim, buried at a depth of 1 m, to free the airway, plus a further 3 minutes to initiate CPR in standard supine position. This is more than two-thirds of the 15 minutes considered necessary for successful companion avalanche rescue. Even minimal training significantly reduced extrication times. These findings emphasize the importance of regular practice in specific extrication techniques that should be part of any training in avalanche companion rescue.

Introduction

Death from asphyxia is the leading cause for mortality in completely buried avalanche victims (Falk et al., 1994; Brugger et al., 2009; Haegeli et al., 2014; Procter et al., 2016). The chance of survival is ∼90% for those totally buried victims extricated within the first 15–18 minutes of burial (“survival phase”) but drops rapidly to about 30%–35% when burial time exceeds 35 minutes (“asphyxia phase”) (Brugger et al., 2001). The potential success of an avalanche rescue is highly time-dependent (Strapazzon et al., 2017). Companion rescue increases the chance of survival when sufficient respiration can be restored within 15–18 minutes after burial (Brugger et al., 2001, 2003). In the case of cardiac arrest, the chance of survival is markedly higher for those victims immediately resuscitated by companions, in comparison to those resuscitated subsequently by emergency medical services (Moroder et al., 2015).

Major improvements in transceiver technology and user training have helped to reduce the time taken to locate totally buried avalanche victims (Brugger et al., 2007; Ng et al., 2015). Far less attention has been given to investigating the factors that potentially affect the time needed to extricate a victim during companion rescue. We studied the times taken from location of the fully buried manikin, until complete extrication and completion of one round of standard cardiopulmonary resuscitation (CPR) during a companion rescue simulation. Specific attention was given to the times taken to access the manikin, free the airway, and start CPR in a standardized supine position. Factors that have been previously considered in the literature to affect the time to victim extrication, namely the burial position of the victim and number of companion rescuers, were included in the study design and later analyzed for influence on time to extrication (Genswein and Ragnhild, 2008a; Kornhall et al., 2016).

Materials and Methods

In this randomized, single-blinded study, medical students with varied CPR training were timed as they attempted to extricate and resuscitate fully buried manikins, either alone or in pairs, in a simulated avalanche companion rescue scenario. The Ethics Committee of the Medical University Innsbruck did not require ethical approval.

Participants and recruitment

Eighteen medical students, who regularly perform winter sports activities and had completed either basic life support (BLS) and/or advanced life support certification from Medical University Innsbruck were recruited. Before inclusion, all participants completed a questionnaire to evaluate their actual knowledge of standard BLS procedures, avalanche companion rescue, and previous formal training in mountain rescue.

The level of formal training in avalanche rescue was classified as: (1) no avalanche rescue knowledge (participants had never received any formal avalanche safety or rescue training; (2) basic avalanche rescue knowledge (participants had completed one course in avalanche safety or rescue); and (3) advanced avalanche rescue knowledge (participants regularly trained in avalanche rescue, e.g., as a member of the Mountain Rescue Service).

Study site

The study was conducted over 2 consecutive days on a designated slope next to the organized ski area Schlick 2000, Fulpmes, Austria, at an altitude of 2136 m. The compressed snow required to simulate the dense snow of avalanche debris was created with the help of a snow groomer. Before each test cycle, three holes of 1.5 m in length, 1 m in depth, and 1 m in width were excavated in a 30° steep slope. A resuscitation manikin (Resusci Anne QCPR; Laerdal Medical, Stavanger, Norway) was placed into each hole in the preassigned body position according to the study design detailed below. Based on previous data, the three most common burial positions observed in recreational avalanche accidents (Kornhall et al., 2016) were chosen:

Position 1: head downhill, prone position (DP)

Position 2: head uphill, prone position (UP)

Position 3: head uphill, supine position (US)

The torso of each manikin was dressed in a ski suit, which was stuffed with snow to create a body-shape comparable to an adult skier in both size and weight. Each manikin was buried with the end of an avalanche probe touching the right thigh. Snow density in the simulated avalanche debris was measured using a standardized aluminum cylinder that is designed to contain 500 mL of snow. A cylinder filled with snow from the artificial avalanche was weighed using a spring scale and snow density calculated as kg/m³ (Proksch et al., 2016).

Study design

A balanced crossover design was used and each of the three different body positions (DP, UP, and US) was equally tested in the first, second, and third test cycle. A total of 36 experiments were conducted. Each test scenario was completed, either by a single participant (18 rescuer tests with 6 participants, each performing with all 3 different body positions) or by 2 participants (18 tests with 12 participants in 6 pairs, each pair performing with all 3 different body positions). Each participant was randomly assigned to a specific rescue scenario, which consisted of either a single rescuer working alone or two rescuers working as a team. Each test scenario included all different sequences of manikin body position, which were randomized in order of performance and blinded to all participants. To minimize the impact of physical exhaustion on the timed extrication efforts, the study was designed to allow each participant a break of at least 30 minutes between test scenarios.

Immediately before the experiment, all participants were given short, theoretical instructions concerning the recommended, standardized shoveling technique to be utilized in a companion rescue attempt (shoveling in a V-shaped figure starting 1.5 m downhill from the avalanche probe) (Genswein and Ragnhild, 2008b). They were briefed about the phases of extrication that would be evaluated in each experiment. All participants were informed that the avalanche probe (Ortovox Probe Alu 240; ORTOVOX Sportartikel GmbH, Taufkirchen, Germany) marked the manikin location exactly and the mark on the probe indicated a burial depth of 1 m. For extrication, the participants were equipped with commercially available aluminum shovels designed for companion rescue (Ortovox Shovel Pro Alu III; ORTOVOX Sportartikel GmbH).

Two staff members supervised each test scenario. The first staff member used an electronic stopwatch to measure three predetermined time points (Table 1) and the second documented the time that was taken to reach each specific phase of extrication. As soon as participants accessed the victim (time point of first contact = T1), they were urged to continue digging with their hands, as advised by avalanche rescue guidelines to prevent any further injury to the “victim” (manikin). Criteria for abandoning the trial included either exhaustion or inability to complete the extrication within the allotted time limit of 45 minutes.

Table 1.

Time Points

T1	First contact with the manikin
T2	Uncovering the airway of the manikin
T3	Manikin placed in standard supine position for CPR

Time points of extrication evaluated during each test scenario.

CPR, cardiopulmonary resuscitation.

To determine the first possible time point at which assessment and management of the airway was feasible (T2), the participants had to remove a red cloth from the mouth of the buried manikin. Standard position for CPR (T3) was defined as a supine horizontal position after being completely removed from the hole.

Statistical analysis

Statistical analyses were conducted with SPSS (IBM, version 25). Data are expressed as medians with ranges ( = minimum − maximum) rather than as means ± standard deviations, because the data were not expected to be normally distributed. Frequencies were compared by means of Chi-square or Fischer's exact tests. Rescue times were log transformed and analyzed by means of a linear mixed model with test session as a repeated effect. We also analyzed rescue times by autoregressive covariance type and thereby were able to detect fixed effects of influencing factors. Those were body position, number of rescuers, previous training, as well as carryover effect and period effect. All p-values were two sided and a value of p ≤ 0.05 was considered statistically significant.

Results

The mean age of the 18 participants (13 male and 5 female) was 23 ± 2 years. Five males and one female participant were randomly allocated to the one rescuer scenario, two male pairs and four male-female pairs to the two rescuers group. Three of the 18 (17%) participants had advanced avalanche rescue training, 4 (22%) participants had basic avalanche rescue training, and 11 (61%) participants had no formal prior avalanche rescue training. The snow density measured during the experiment varied between 260 and 320 kg/m³ (median of 290 kg/m³) and was not different between one and two rescuers experiments or any of the different manikin positions investigated.

Extrication times

The median time intervals until first manikin contact (T1), ability to assess and free the airway (T2), and initiation of CPR in standard supine position (T3) for all experiments are shown in Table 2. The time to first manikin contact was a median 2.5 minutes (range 0.6–8.6 minutes). Extricating the patient sufficiently, the airway assessment, and management possibly took a median of 7.2 minutes (range 2.3–20.4 minutes). It took a median of 10.1 minutes (range 3.0–24.9 minutes) to bring the victim into standard position for CPR. In our study, we observed an unexpected large variation in extrication times among individual participants. This accounts for ∼61% of the totally observed random variation (variance component individual 0.177, error 0.113). The slowest participants required up to 10 times longer than the fastest to free the airway. The most time-consuming phase was from the time of first contact to the time at which access to the airway was possible (median 4.7 minutes).

Table 2.

Absolute Extrication Times

	One rescuer	Two rescuers	All
Time point T1
Head downhill prone	2.1 (1.0–4.9)	3.7 (1.1–4.7)	2.5 (1.0–4.9)
Head uphill prone	3.1 (1.5–5.5)	2.2 (1.3–5.7)	2.7 (1.3–5.7)
Head uphill supine	2.2 (1.0–8.6)	2.5 (0.6–6.8)	2.2 (0.6–8.6)
All positions	2.5 (1.0–8.6)	2.8 (0.6–6.8)	2.5 (0.6–8.6)
Time point T2
Head downhill prone	9.7 (2.3–16.7)	6.5 (4.8–19.2)	7.4 (2.3–19.2)
Head uphill prone	7.3 (5.4–12.7)	7.4 (5.7–10.9)	7.3 (5.4–12.7)
Head uphill supine	6.5 (4.2–20.4)	6.5 (2.8–9.5)	6.5 (2.8–20.4))
All positions	7.3 (2.3–20.4)	6.6 (2.8–19.2)	7.2 (2.3–20.4)
Time point T3
Head downhill prone	10.8 (3.0–17.9)	11.7 (6.6–19.2)	10.8 (3.0–19.2)
Head uphill prone	11.8 (6.5–24.9)	9.6 (7.1–22.6)	10.6 (6.5–24.9)
Head uphill supine	8.8 (6.1–22.2)	8.8 (3.6–13.5)	8.8 (3.6–22.2)
All positions	10.5 (3.0–24.9)	9.3 (3.6–22.6)	10.1 (3.0–24.9)

Comparison of extrication time points in minutes, recorded by one or two rescuers for each manikin burial position encountered.

Factors influencing extrication times

Neither the position of the manikin (10.8 minutes vs. 10.6 minutes vs. 8.8 minutes; p = 0.428) nor the number of rescuers (one compared to two, 10.5 minutes vs. 9.3 minutes; p = 0.686) significantly influenced the time taken to make first contact with manikin, free the airway, or place the manikin in standard supine position for CPR (Table 2).

Previous formal training improved extrication times. Participants with previous training were able to free airways in a median of 6.1 minutes (range 2.3–7.2 minutes), whereas those without formal training took a median of 11.0 minutes (range 4.2–20.4 minutes; p = 0.006). Participants significantly improved their extrication times during the three consecutive experiments as depicted in Table 3.

Table 3.

Time Improvement of Consecutive Test Cycles

Time point	Absolute time	Difference	Difference	p
T1	First	First − Second	First − Third
Median	2.6	−0.8	−0.6	0.841
T2	First	First − Second	First − Third
Median	8.0	−1.7	−0.9	0.114
T3	First	First − Second	First − Third
Median	12.4	−3.0	−3.1	0.018

Absolute times (in minutes) to reach each extrication time point (T1, T2, and T3) in the first test cycle of each participant are given in the column “First.” The columns “First − Second” and “First − Third” show the time differences in minutes between the medians of the first to second test cycle and the first to the third test cycle, respectively. The extrication times in the second and third tests were significantly shorter than in the first test (p = 0.017).

Discussion

This study highlights the need to determine crucial time points during companion rescue of a completely buried avalanche victim. It shows the effectiveness of training to reduce extrication times significantly.

In our field experiments, times to first contact with the manikin, uncovering the airway and placing the manikin in standard supine position for CPR were ∼2, 7, and 10 minutes, respectively. We chose a burial depth of 1 m for our experiments, because depths of 0.8 m (Switzerland) and 1.0 m (Canada) represent the average depths of burial in victims involved in recreational avalanche accidents (Harvey and Zweifel, 2008; Haegeli et al., 2014).

There are limited data available pertaining to the exact times taken to reach different points in the process of extrication of completely buried avalanche victims. The majority of published reports from field experiments and avalanche accidents provide data only on the “duration of burial,” that is time from the avalanche until uncovering the victim's face. These data do not differentiate between the time needed for location and extrication of each victim. A study analyzed data from organized rescue missions after reported avalanche accidents in Tyrol and Switzerland from 1995 to 2001. The authors reported that on average, the full extrication of a victim buried at a depth of 1 m will require ∼10 minutes, when dug out by a team of rescue professionals (Slotta-Bachmayr, 2005).

Similarly, in previous field experiments, another study found a median time to first contact with the victim of 6 minutes, a median time until “head free” of 8 minutes and a median time to completely extricate a victim of more than 20 minutes (Genswein and Ragnhild, 2008a). In these experiments the victims were not simulated by a manikin, but by a sack of firewood (Genswein and Ragnhild, 2008b). The marked differences in time intervals compared to our study can be explained by potential differences in the snow density of the simulated avalanche debris and marked differences in the chosen range of burial depth and the number of rescuers involved (Genswein and Ragnhild, 2008a).

There is not only a lack of robust data concerning extrication times in avalanche accidents but also due to a lack of standardization among studies, it is also impossible to compare the few data available. As an important step to enable evaluation and comparison of techniques and equipment for extrication, we suggest the establishment of a standardized protocol for data acquisition and documentation in field experiments. This protocol should include the number and training status of participants, the equipment used, snow density measurement of the avalanche site, and the burial depth and body position of the sham victim (manikin). This should include clearly defined time points of interest during the process of extrication. From a medical point of view, first access to the airway and ability to start mouth-to-mouth ventilation and external chest compressions are critical time points. Therefore, they should be part of any standardized protocol for data acquisition and documentation.

At each extrication time point or phase of interest, a high level of interindividual variation was found. For the slowest participant, it took ten times longer to free the manikin's airway compared to the fastest participant. Wide interindividual variation in extrication times was also found in almost all previous studies evaluating extrication times (Edgerly and Atkins, 2006; Genswein, 2009). It is reasonable to assume, that such a wide variation in extrication times will also be found in the general population that may be involved in attempted companion rescue in real-life avalanche accidents. This wide variation is at least partly due to factors such as physical strength and physical fitness, which cannot be influenced by rescue training or knowledge. In addition to previous rescue training, the level of individual physical fitness is known to be a secondary factor with proven significant impact on the duration of burial in field experiments (Quine, 2010).

In most avalanche rescue operations, the duration of burial is a key parameter, which not only determines whether a victim dies but also the type of triage to be administered for best neurological outcome (Slotta-Bachmayr, 2005). A large time difference between “first contact,” “airway access/head free” and “CPR standard position/completely extricated” was found in our data and had been found in the data of previous studies. A universally accepted definition and documentation of the time point when “burial” ends is therefore necessary to allow clear communication within the scientific community and enable a reproducible comparison of burial duration among rescuers. The “duration of burial” has been defined clearly as “time between burial and uncovering the face” only since 2017 (Kottmann et al., 2017). Many of the analyses based on duration of burial published in the recent years must be interpreted cautiously whenever duration of burial is not defined in more detail or defined time points are not used for analysis.

A difference of 5–10 minutes between the first contact with on avalanche victim and airway access must be expected based on current data. On the contrary, a victim completely extricated in <20 minutes, may have benefited from early airway management, preventing the development of asphyxia after only a few minutes of burial and ensuring far better outcome. Survival curves calculated on the basis of data using “duration of burial,” without exact definitions of these specific time points of interest (specifically T2) must be interpreted cautiously. Asphyxia is the leading cause of early death after snow burial. Freeing the airway of a buried victim is the most important intervention to promote survival and should be targeted in practice. The current definition of “duration of burial” should perhaps be considered as the time taken to access and effectively free the airway of an avalanche victim.

In experimental field studies, after airway access had been established, it took an additional 3–12 minutes to extricate an avalanche victim completely to enable the initiation of CPR in the standard supine position (Genswein and Ragnhild, 2008a; Genswein, 2009). These data instill uncertainty regarding whether or not it is always completely necessary to fully extricate the victim, before providing initial rescue breaths and starting CPR if indicated. The argument is that, starting ventilation as soon as possible should be the priority, instead of completely digging out the victim. Alternative techniques for external chest compression, such as over-head CPR and reverse CPR in prone position, may also be considered as they sometimes can be provided with only limited access to the thorax (Mazer et al., 2003). The significant differences between extrication time points demonstrated in previous field experiments (Genswein and Ragnhild, 2008a; Genswein, 2009) and in our own data justify further evaluation concerning the advantages of using alternative CPR techniques and rescue breaths at the earliest possible opportunity for victims in atypical burial positions (Blancher et al., 2018).

In our study, previous training and a possible training effect evident during the three experimental cycles significantly improved all extrication time points of interest. Due to the wide interindividual variation in extrication times, statistical analysis using medians and means can miss a significant effect of training on extrication times. We chose a statistical analysis comparing data by each participant at different time points to demonstrate a significant intraindividual training effect in the course of the experiment. As in our study, a previous study found a positive effect of 15-minute training on the effectiveness of avalanche rescue in otherwise untrained volunteers (Genswein, 2009). Because even limited training had a significant effect, we recommend practical training in shoveling techniques during avalanche rescue courses.

Duration of burial in avalanche companion rescue includes time taken to locate the victim or victims (search with avalanche transceivers and probing) and time for extrication. Using modern transceiver technology, the time to locate an avalanche victim is approximately 3–5 minutes in experimental field studies (Genswein and Ragnhild, 2008a), but in real avalanche rescue scenarios search time ranged from 5 to 10 minutes (Genswein and Ragnhild, 2008b; Genswein, 2009). It has been repeatedly suggested that extrication requires “the major part of the time of companion rescue efforts” (Genswein and Ragnhild, 2008a; Schindelwig et al., 2017).

Nevertheless, current training courses in companion rescue focus on searching with transceivers and normally include only a theoretical training module in digging and extrication techniques (Genswein and Ragnhild, 2008b; Genswein, 2009). Our data indicate that extrication takes at least as much time as transceiver search. Even minimal training could significantly shorten this time interval. Courses in companion avalanche rescue should teach digging techniques that allow rapid access to the victim's airway, in the three most common burial positions.

Limitations

The study was a manikin study in an artificial avalanche. We attempted to simulate all aspects of companion rescue but could not include all factors such as psychological stress. The manikin cannot represent all components of a real human avalanche victim. The median snow density in our study was 290/m³. This is typical for dry avalanches but may not be representative of wet avalanches that occur at ambient temperatures above 0°C. In 1985, a study, performed in British Columbia, Canada, analyzed 95 avalanches of which 55% were dry avalanches with an average density of 330 kg/m³, 24% were moist avalanches with average density 420 kg/m³, and 21% were wet avalanches with average density 500 kg/m³ (McClung and Schaerer, 1985). All of the 95 avalanches recorded were observed at the same path, which was characterized by a vertical drop of 1000 m at an incline of 35°. This is not typical for most avalanches triggered by backcountry skiers. We assume that 290 kg/m³ was a representative density of typical avalanche debris for typical avalanches in an area with a continental climate and mainly dry avalanches (Haegeli et al., 2011).

For logistic reasons, we limited our sample size to 18 participants. Despite extensive efforts, we were only able to include five female participants. Because of this low number, we refrained from performing any statistical analysis of sex-specific aspects of extrication. To limit the number of scenarios, we only compared one to two rescuer groups even though we are aware of the fact that larger groups exist.

Conclusion

Ethics Approval and Consent to Participate

The Ethics Committee of the Medical University Innsbruck was contacted and informed about the study, but did not require an ethical approval. All participants gave written informed consent to the participation in the study and to publication of the data.

Footnotes

Acknowledgments

The authors thank the ski area Schlick 2000 in Fulpmes, Austria for helping in setting up the simulation and providing transport, the snow groomer, and catering during the experiment. They also thank the Austrian Alpine Club for lending the avalanche shovels and probes to us for the duration of the study. There was no funding supporting this article.

Authors' Contributions

All authors made substantial contributions to conception and design, and/or acquisition of data, and/or analysis and interpretation of data. All authors participated in drafting the article or revising it critically for important intellectual content. All authors gave final approval of the version to be submitted and any revised version. B.W. and P.M.: study conception and design, acquisition of data, analysis and interpretation of data, drafting of article, and critical revision. L.M., A.B., S.E., G.P., G.S., and H.B.: study conception and design, acquisition of data, and critical revision. M.B.: acquisition of data and critical revision. R.T.: acquisition of data, language correction, and critical revision. M.F.: study conception and design, statistical analysis, and interpretation of data.

Author Disclosure Statement

None of the authors have a conflict of interest to declare. The authors declare that they have no competing interests.

References

Blancher

, Albasini

, Elsensohn

, Zafren

, Holzl

, McLaughlin

, Wheeler

3rd , Roy

, Brugger

, Greene

, and Paal

. (2018). Management of multi-casualty incidents in mountain rescue: Evidence-based guidelines of the International Commission for Mountain Emergency Medicine (ICAR MEDCOM). High Alt Med Biol, 19:131–140.

Brugger

, Durrer

, Adler-Kastner

, Falk

, and Tschirky

. (2001). Field management of avalanche victims. Resuscitation, 51:7–15.

Brugger

, Etter

, Boyd

, and Falk

. (2009). Causes of death from avalanche. Wilderness Environ Med, 20:93–96.

Brugger

, Etter

, Zweifel

, Mair

, Hohlrieder

, Ellerton

, Elsensohn

, Boyd

, Sumann

, and Falk

. (2007). The impact of avalanche rescue devices on survival. Resuscitation, 75:476–483.

Brugger

, Sumann

, Meister

, Adler-Kastner

, Mair

, Gunga

, Schobersberger

, and Falk

. (2003). Hypoxia and hypercapnia during respiration into an artificial air pocket in snow: Implications for avalanche survival. Resuscitation, 58:81–88.

Edgerly

, and Atkins

. (2006). Strategic shoveling: The next frontier in companion rescue. Proceedings of the International Snow Science Workshop, October 1–6, 2006, Telluride, CO. pp. 522–529.

Falk

, Brugger

, and Adler-Kastner

. (1994). Avalanche survival chances. Nature, 368:21.

Genswein

. (2009). Will a guest ever be able to save your life? Proceedings of the International Snow Science Workshop 2009, Davos, Switzerland. pp. 666–670.

Genswein

, and Ragnhild

. (2008a). The efficiency of companion rescue with minimal training. Proceedings of the International Snow Science Workshop, Whistler, BC, Canada. pp. 581–590.

10.

Genswein

, and Ragnhild

. (2008b). The V-shaped snow conveyor belt. Proceedings of the International Snow Science Workshop, Whistler, BC, Canada. pp. 571–578.

11.

Haegeli

, Falk

, Brugger

, Etter

, and Boyd

. (2011). Comparison of avalanche survival patterns in Canada and Switzerland. CMAJ, 183:789–795.

12.

Haegeli

, Falk

, Procter

, Zweifel

, Jarry

, Logan

, Kronholm

, Biskupic

, and Brugger

. (2014). The effectiveness of avalanche airbags. Resuscitation, 85:1197–1203.

13.

Harvey

, and Zweifel

. (2008). New trends of recreational avalanche accidents in Switzerland. Proceedings of the International Snow Science Workshop, Whistler, BC, Canada. pp. 9–15.

14.

Kornhall

, Logan

, and Dolven

. (2016). Body positioning of buried avalanche victims. Wilderness Environ Med, 27:321–325.

15.

Kottmann

, Blancher

, Pasquier

, and Brugger

. (2017). Avalanche Victim Resuscitation Checklist adaption to the 2015 ERC Resuscitation guidelines. Resuscitation, 113:e3–e4.

16.

Mazer

, Weisfeldt

, Bai

, Cardinale

, Arora

, Ma

, Sciacca

, Chong

, and Rabbani

. (2003). Reverse CPR: A pilot study of CPR in the prone position. Resuscitation, 57:279–285.

17.

McClung

, and Schaerer

. (1985). Characteristics of flowing snow and avalanche impact pressures. Ann Glaciol, 6:9–14.

18.

Moroder

, Mair

, Brugger

, Voelckel

, and Mair

. (2015). Outcome of avalanche victims with out-of-hospital cardiac arrest. Resuscitation, 89:114–118.

19.

, Smith

, Wheeler

, and McIntosh

. (2015). Advanced avalanche safety equipment of backcountry users: Current trends and perceptions. Wilderness Environ Med, 26:417–421.

20.

Procter

, Strapazzon

, Dal Cappello

, Zweifel

, Wurtele

, Renner

, Falk

, and Brugger

. (2016). Burial duration, depth and air pocket explain avalanche survival patterns in Austria and Switzerland. Resuscitation, 105:173–176.

21.

Proksch

, Rutter

, Fierz

, and Schneebeli

. (2016). Intercomparison of snow density measurements: Bias, precision, and vertical resolution. Cryosphere, 10:371–384.

22.

Quine

, and Latosuo

. (2010). Shoveling in the name of science. 2010 International Snow Science Workshop, Squaw Valley, CA.

23.

Schindelwig

, Hoffmann

, Mossner

, and Nachbauer

. (2017). Does avalanche shovel shape affect excavation time: A pilot study. Sports (Basel), 5:31.

24.

Slotta-Bachmayr

. (2005). How burial time of avalanche victims is influenced by rescue method: An analysis of search reports from the Alps. Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards, 34:341–352.

25.

Strapazzon

, Paal

, Schweizer

, Falk

, Reuter

, Schenk

, Gatterer

, Grasegger

, Dal Cappello

, Malacrida

, Riess

, and Brugger

. (2017). Effects of snow properties on humans breathing into an artificial air pocket—An experimental field study. Sci Rep, 7:17675.