Rescue Missions for Totally Buried Avalanche Victims: Conclusions ...

4 downloads 0 Views 85KB Size Report
totally buried avalanche victims are mostly based on personal experience and ... factors determining the survival probability of totally buried victims during ...
HIGH ALTITUDE MEDICINE & BIOLOGY Volume 9, Number 3, 2008 © Mary Ann Liebert, Inc. DOI: 10.1089/ham.2007.1061

Rescue Missions for Totally Buried Avalanche Victims: Conclusions from 12 Years of Experience Matthias Hohlrieder,1,2 Stephanie Thaler,1 Walter Wuertl,2 Wolfgang Voelckel,1 Hanno Ulmer,2 Hermann Brugger,4 and Peter Mair1,2

Abstract

Hohlrieder, Matthias, Stephanie Thaler, Walter Wuertl, Wolfgang Voelckel, Hanno Ulmer, Hermann Brugger, and Peter Mair. Rescue missions for totally buried avalanche victims: conclusions from 12 years of experience. High Alt. Med. Biol. 9:229–234, 2008.—The planning and execution of avalanche rescue missions to search for totally buried avalanche victims are mostly based on personal experience and preference, as evidence-based information from literature is almost completely missing. Hence, the aim of this study was to identify major factors determining the survival probability of totally buried victims during avalanche rescue missions carried out by organized rescue teams (Austrian Mountain Rescue Service, Tyrol). During the 12-year period studied, 109 totally buried persons (56 off-piste, 53 backcountry), were rescued or recovered; 18.3% survived to hospital discharge. Median depth of burial was 1.25 m; median duration of burial was 85 min. The majority (61.6%) of the rescue missions were conducted under considerably dangerous avalanche conditions. The probability of survival was highest when located visually and lowest for those located by avalanche transceiver; survival did not significantly differ between those found by rescue dogs and those located with avalanche probes. Multivariate analysis revealed short duration of burial and off-piste terrain to be the two independent predictors of survival. Whenever companion rescue fails, snow burial in an avalanche is associated with extraordinarily high mortality. Searching the avalanche debris with probe lines seems to be equally effective as compared to searching with rescue dogs. The potential hazard for rescuers during avalanche rescue missions comes mainly from self-triggered avalanches, hence thorough mission planning and critical risk–benefit assessment are of utmost importance for risk reduction. Key words: avalanche rescue; avalanche dog; avalanche probe; rescue mission; rescue equipment Introduction

T

in winter sport activities outside organized ski resorts contributes to an epidemiologically relevant number of avalanche accidents each winter. Avalanche rescue missions are among the most challenging scenarios for mountain rescue teams. Survival probability for totally buried avalanche victims is highly time dependent (Brugger et al., 2001); therefore technical and educational efforts strive for immediate companion rescue involving avalanche transceivers (Hohlrieder et al., 2005; Radwin and Grissom, 2002). Unfortunately, more than 40% of victims are not equipped with avalanche transceivers and, moreover, HE INCREASED INTEREST

companion rescue in inexperienced hands is often ineffective (Hohlrieder et al., 2005). Consequently, each winter a significant number of organized rescue missions to search for totally buried avalanche victims are necessary in the Alps. When companion rescue has failed, only a dense network of rescue stations, emergency medical helicopters, and welltrained rescue dogs enable a reasonable probability of survival. Besides being limited in efficacy, avalanche rescue missions often carry a high risk, as rescue team members are exposed to the same avalanche hazard. From 2001 to 2002, the International Commission for Alpine Rescue reported five fatalities during avalanche rescue missions in Europe (Etter, 2004).

1Department

of Anesthesiology and Critical Care Medicine, Innsbruck Medical University, Austria. Mountain Rescue Service. 3Department of Medical Statistics, Informatics and Health Economics, Innsbruck Medical University, Austria. 4President of the International Commission for Mountain Emergency Medicine/ICAR MEDCOM Mountain Rescue Service/South Tyrolean Alpine Association. 2Austrian

229

230

HOHLRIEDER ET AL.

The planning and execution of avalanche rescue missions are mostly based on the personal experience and individual preferences of the rescuers involved. This is because evidence-based information derived from analysis of avalanche rescue missions is almost completely missing in peer-reviewed literature. Hence, the aim of this retrospective study was to analyze the avalanche rescue missions conducted by the Austrian Mountain Rescue Service in Tyrol, Austria, over a 12-year period to identify major factors determining the survival probability of totally buried victims during organized avalanche rescue missions. Methods The data bases of two institutions, the Austrian Mountain Rescue Service and the Austrian Board for Alpine Safety, were searched for totally buried avalanche victims rescued on organized rescue missions in Tyrol over the 12-year period 1993 to 2005. Only victims involved in accidents in unprotected mountainous terrain during recreational or occupational activities were included. Avalanche victims involved in disastrous avalanches affecting inhabited areas were excluded from further analysis. Tyrol is one of the Austria’s nine federal states and includes 1333 square miles of mountainous terrain, with many of Europe’s busiest backcountry and off-piste skiing resorts. The region has a high incidence of avalanche accidents. A victim was considered totally buried when head and chest were covered by the snow masses, even if other parts of the body or pieces of the victim’s equipment were visible on the debris surface. Backcountry activities take place in remote wilderness terrain, whereas off-piste skiing and snowboarding are practiced in unprotected terrain located next to the outer borders of secured ski areas. During the winter months the region is served by members of the Austrian Mountain Rescue Service, organized into 93 local units dictated by geographic conditions. All are experienced mountaineers and work on standby as volunteers. In addition, all larger ski areas have professional mountain rescue personnel on duty during the opening hours, guaranteeing immediate help within the protected ski area and the neighboring off-piste terrain. The Tyrolean Mountain Rescue Service also has more than 50 dog handlers with their rescue dogs. Flying conditions permitting, these ground-operation rescue teams are supported by 12 to 15 emergency medical helicopters (HEMS), the highest density of rescue helicopters in Europe. They can airlift a rescue team consisting of emergency physician and alpine paramedic equipped with avalanche transceivers, shovels, and probes

TABLE 1.

COMPARISON Victim total (n ! 109)

Visual Transceiver Dog Probe

14 23 34 38

(12.8) (21.1) (31.2) (34.9)

OF

to the site of an avalanche within a few minutes. During periods of increased risk, rescue dogs and their handlers are on standby at some heliports, thus being available without delay. For each victim included in the study, we obtained age, sex, nationality, presence of avalanche transceiver (irrespective of whether he or she was located by the transceiver), type of activity (skiing or snowboarding, ascent or descent), the search method that eventually located the victim (visual surface search, search with avalanche transceiver, avalanche dog, or probe line), the depth and duration of snow burial, and the outcome (alive or dead at hospital discharge). Also determined were date, time, and place of the accident; length, width, and tear-off height of the avalanche, the terrain where the accident happened (backcountry or off-piste); and the number of persons totally buried. For all accidents for which the duration of burial exceeded 90 min, information available from the rescue mission record was searched for the main cause of the delay in rescue. From the daily bulletin issued by the Tyrolean Avalanche Information Service, the actual degree of avalanche hazard was obtained for each rescue mission. This avalanche bulletin provides daily information on regional avalanche hazards based on the international 5-point avalanche hazard scale. Low hazard (grade 1) means that triggering of avalanches is possible only when exerting high additional loads on extremely steep slopes. Under moderately dangerous conditions (grade 2), human-triggered avalanches are possible on steep slopes, normally, however, only with large additional impact on the snow pack. Under considerably dangerous conditions (grade 3), spontaneous avalanche descent is rather unlikely, but avalanches can be easily triggered by exerting little additional impact on the snow pack. The descent of spontaneous medium or large avalanches is possible or even likely under highly dangerous conditions (grade 4). Extremely high danger (grade 5) means that numerous large natural avalanches are likely, even in moderately steep terrain. Statistical analysis Statistical analysis was performed with the SPSS statistical package (SPSS Inc., Chicago, IL). For victim age, time of accident, number of totally buried victims, and avalanche hazard scale, we calculated the arithmetic mean. The median was used for length, width, and tear-off height of the avalanche, as well as for the duration and depth of burial. Means, medians, range, and standard deviations were calculated to describe continuous variables. The Kol-

DIFFERENT MEANS

OF

LOCALIZATION

Victim survival (n ! 20) 5 0 5 10

(35.7) (0) (14.7) (26.3)

Median burial time (min) 57.5 150.0 60.0 85.0

Percentage in brackets; difference in survival (p ! 0.019); median burial time in minutes (not significantly different).

AVALANCHE RESCUE MISSIONS

231

TABLE 2. COMPARISON

OF

SURVIVORS

Survivors (n ! 20)

AND

NONSURVIVORS

Nonsurvivors (n ! 89)

Total (n ! 109)

p value

Age off-piste skiers Snowboarders Transceiver equipped Accident during descent

26.7 (14–66) 16 (80.0) 6 (30.0) 1 (5.0) 18/20 (90)

37.4 (4–75) 40 (44.9) 17 (19.1) 32/86a (37.2) 54/82a (65.9)

35.6 (4–75) 56 (51.4) 23 (21.1) 33/106a (31.1) 71/102a (70.6)

0.013 0.004 nsd 0.003 0.026

Avalanche Hazard scale Totally buried victims Avalanche length Avalanche width Tear-off height Duration of burial Depth of burial

3.12 (2–4) 1.85 (1–4) 175 (100–650) 70 (10–300) 0.8 (0.5–2.0) 37.5 (10–185) 0.9 (0.5–1.7)

2.78 (2–4) 1.84 (1–6) 300 (60–1500) 100 (15–850) 1.0 (0.2–2.0) 90.0 (20–32000) 1.5 (0.25–10.0)

2.84 (2–4) 1.8 (1–6) 300 (60–1500) 100 (10–850) 1.0 (0.2–2.0) 85 (10–32000) 1.25 (0.25–10.0)

0.035 nsd nsd nsd nsd 0.000 0.027

Localization Visual Transceiver Dog Probe

5 0 5 10

9 23 29 28

14 23 34 38

nsd 0.005 nsd nsd

(25.0) (0) (25.0) (50.0)

(10.1) (25.8) (32.6) (31.5)

(12.8) (21.1) (31.2) (34.9)

aMissing

data. Percentage and range in parentheses; nsd, not significantly different; length, width, and height of avalanche and depth of burial are expressed in meters; duration of burial is expressed in minutes.

mogorov–Smirnov test was applied to test for normal distribution. To evaluate differences between various groups of victims we used chi-square tests, t tests, Kruskal–Wallis H tests, and Mann–Whitney U tests. p-Values below 0.05 were considered significant. Logistic regression analysis was used to estimate adjusted odd rates and 95% confidence intervals for outcome (survival vs. nonsurvival). The final regression model included the type of terrain (backcountry vs. offpiste), depth (m), and duration (min) of burial. Due to the skewed distribution of depth and duration of burial, these variables were transformed logarithmically before inclusion in the regression model. Results

0.25 to 10 m), and median duration of burial was 85 min (range: 10 min to "6 h). More than three-quarters (76.7%) of the accidents happened between 11 A.M. and 3 P.M.; most victims (70.6%) were caught by the avalanche while descending. Regarding the actual avalanche hazard, only 27.3% of the rescue missions were carried out under moderately dangerous conditions (grade 2). The majority (61.6%) of the missions were conducted under considerably dangerous conditions (grade 3); 11.1% of all rescue missions had to be performed under highly dangerous conditions (grade 4). Offpiste rescue missions were performed under more dangerous conditions than were those in backcountry terrain (3.0 vs. 2.67, p ! 0.006)

Accidents During the 12-year period studied, 109 totally buried persons involved in 85 avalanche accidents were rescued or recovered by organized rescue teams in the Tyrolean Alps. Victims were predominately male (91.5%); the majority of them (78.9%) were skiers and the rest were snowboarders (21.1%). While 56 individuals (51.4%) were caught by the avalanche during off-piste activities, 53 (48.6%) were involved in backcountry activities. Median depth of burial was 1.25 m (range:

TABLE 3.

PREDICTION

Backcountry Depth (per log unit) Duration (per log unit)

OF

SURVIVAL

Localization Of all victims studied, 31.1% were equipped with avalanche transceivers; none of them could be successfully rescued by companions before arrival of the rescue teams. Table 1 shows the localization methods, the survival rates, and the median burial times. The probabilities of survival were significantly different, being highest when located visually and lowest for those located by avalanche transceiver.

WITH

LOGISTIC REGRESSION ANALYSIS

Odds ratio

95% CI

p value

3.7 3.8 8.1

1.1–12.8 0.4–36.3 1.8–35.8

0.042 0.251 0.006

232 When compared directly, survival probability did not significantly differ between those found by rescue dogs and those located with avalanche probes. Survival Overall survival was low; only 20 (18.3%) of 109 victims survived to hospital discharge. The factors studied for their influence on survival are summarized in Table 2. In univariate analysis, younger age, lesser depth of burial, shorter time of burial, being caught during descent, search under more dangerous conditions, and burial in off-piste terrain were significantly associated with survival. The localization method had no significant influence on survival, apart from the fact that victims located with their transceiver after failed companion rescue had an extraordinarily poor prognosis, shown by the fact that all of them died. Avalanche transceivers were almost exclusively used during backcountry activities. Transceiver carriers were rescued out of longer (300 vs. 250 m, p ! 0.044) and wider (250 vs. 90 m, p ! 0.003) avalanches and from greater depths of burial (1.7 vs. 1.0, p # 0.001).ˇIn general, the amount of debris that had to be searched had no significant influence on survival. In multivariate analysis, only the duration of burial and off-piste terrain proved to be independent predictors of survival (Table 3). Adjusted for depth and duration of burial, backcountry accidents were associated with a 3.7-fold risk of nonsurvival as compared to off-piste accidents. Rescue delay Duration of burial exceeded 90 min in 52 victims (47.7%). In 37 of these, an obvious reason for the prolonged burial time could be identified. The respective causes of delay were delayed alarming of rescue forces because the accident was not witnessed or the whole group was affected (n ! 19), accidents that occurred in a remote area and poor weather conditions not allowing helicopter use (n ! 8), extremely great depth of burial exceeding 4 m (n ! 8), and the impossibility of conducting a primary rescue mission because of an unacceptable avalanche hazard for the rescuers (n ! 2). Discussion In this retrospective study the overall survival of totally buried avalanche victims rescued by organized rescue parties was 18.3%. Comparable survival rates for organized avalanche rescue missions have been reported previously (Brugger et al., 1997; Tschirky et al., 2007). Available data clearly indicate that, whenever companion rescue fails, snow burial in an avalanche is associated with extraordinarily high mortality. The two most important determinants of survival elicited in our analysis were the duration of snow burial and the terrain where the accident happened. Time to extrication has already been described as a major determinant of survival (Brugger et al., 2001; Brugger and Durrer, 2002); the correlation between burial duration and survival is typically nonlinear (Brugger et al., 2001). The survival probability plotted against the burial duration shows two distinct decreases: a first drop between 15 and 35 min after avalanche descent when about 60% of all victims die and a second between 90

HOHLRIEDER ET AL. and 120 min after avalanche descent when another 20% of victims die (Brugger et al., 2001). Interestingly, in the period between 35 and 90 min after avalanche descent, almost none of the victims die after surviving the initial asphyxia period (Brugger et al., 2001). Considering this typical time course of survival probability, the proposed operational target for organized avalanche rescue missions is to extricate all victims within 90 min, whereas companion rescue should be successful within 15 min (Brugger et al., 2001; Brugger et al., 1996; Brugger and Durrer, 2002). In our analysis, the 90-min limit was exceeded in about 50% of all cases studied, suggesting that there is room for improvement in planning and executing avalanche rescue missions in our region. However, when determining the actual reasons for prolonged burial time, data indicate that, for more than two-thirds of all victims, rescue delay was caused by factors independent of mission planning and execution (for example, delayed alarming of the rescue party because the accident was not witnessed). With the methods of localization available today (predominantly rescue dogs, avalanche transceivers, and probing), it is possible to localize the majority of totally buried victims within the suggested time limit whenever an organized rescue mission can be initiated. The terrain where the accident happens is the second major determinant of survival with an almost fourfold probability of survival for victims involved in off-piste accidents. Improved survival in off-piste accidents is at least partly due to the significantly better logistic background for rescue efforts in the immediate neighborhood of a ski resort. Off-piste accidents are normally witnessed, and hence rescue efforts by on-duty members of the Mountain Rescue Service and ski-resort staff are started immediately. By contrast, in backcountry terrain the personnel for probing an avalanche and the availability of rescue dogs are often limited, and rescue during the first 90 min after avalanche descent normally depends on helicopter transport. However, companion rescue may be another factor contributing to the difference in survival probability between off-piste and backcountry terrain. In backcountry avalanche accidents, companion rescue often occurs before organized rescue efforts start, whereas it is very unusual in off-piste terrain. It is possible that in backcountry accidents many of the easier burial situations are successfully dealt with by companion rescue before professional rescue personnel arrives at the scene. The victims remaining to be rescued by organized forces are mainly those not equipped with an avalanche transceiver and those involving difficult burial scenarios, that is, extraordinarily deep burial or particularly large avalanches. Such victims have very poor survival chances anyway. This retrospective study revealed that searching the avalanche debris with probe lines was equally effective compared with searching with rescue dogs. This is surprising considering the fact that searching with rescue dogs is widely considered to be the most rapid and efficient means of localizing totally buried avalanche victims (Slotta-Bachmayr, 2005). This assumption is based mainly on common sense, theoretical considerations, and experience from rescue training. To the best of our knowledge, the performance of rescue dogs under the difficult and stressful situation of a reallife rescue mission has not been independently evaluated so far. On the other hand, the effectiveness of rescue dogs may

AVALANCHE RESCUE MISSIONS be limited by the fact that it is seldom possible to get them to the accident site in sufficiently large numbers as fast as needed. Taken together, the potential superiority of a search with rescue dogs is not translated into improved survival or localization rates in real-life avalanche accidents. Therefore, our results suggest that it may be unwise to base avalanche rescue strategy solely or predominately on searches with rescue dogs, although this is an attractive option as probe lines expose a markedly larger number of rescue personnel to potential risk. Our results indicate that we strongly need a universal and more rapidly available search technique that has reliable and reproducible efficacy and does not depend on human or animal performance. Among the techniques currently being evaluated, the radar-based RECCO detection system #www.recco.com" might be such a technique. However, in our dataset no victim was localized using the RECCO detector. Victims for whom companion rescue failed and who were localized later during an organized rescue mission with the help of their transceiver had an extraordinarily poor prognosis. Failed companion rescue is typically associated with a particularly difficult burial situation, as indicated by a greater depth of burial in larger avalanches, the accident often being unwitnessed or involving the whole group in remote backcountry terrain. Although the presence of an avalanche transceiver cannot improve survival in such situations, avalanche transceivers are of unique value for the rescuers as they allow rapid localization of the deceased victim with minimal exposure to avalanche hazard. In this regard, it should be noted that this study analyzed only organized rescue missions. Thus, the number of victims found and rescued alive by their companions employing transceivers is unknown. In avalanche rescue missions, the rescuers are exposed to the same avalanche hazard as the victim; depending on the actual avalanche hazard and the topographic conditions, the mission may entail a high risk. Our analysis shows that the majority of rescue missions had to be carried out when the regional avalanche bulletin predicted a considerable avalanche hazard, that is, in situations where large natural avalanches are indeed rare, although only minor additional load to the snow pack (for example, the weight of a single rescuer) can trigger an avalanche descent. Obviously, thorough rescue mission planning is essential for risk reduction, because a rescue mission should be reasonably safe whenever additional load to hazardous slopes can be avoided. Three out of five fatalities during avalanche rescue missions documented by the International Commission for Alpine Rescue (Etter, 2004) occurred while approaching the site of the accident and were triggered by the rescuers themselves. In contrast, fatalities due to spontaneous natural avalanches affecting rescuers while searching the avalanche debris were less common. The Swiss Federal Institute for Snow and Avalanche Research observed that in Switzerland off-piste skiers tended to be less cautious than backcountry skiers, as they more often triggered avalanches during extremely hazardous conditions (Harvey et al., 2007). This seems to be the same in our sample, as rescue missions for off-piste victims had to be carried out under significantly more dangerous conditions, and rescue missions necessary under highly hazardous conditions were performed mainly in off-piste terrain.

233 Conclusion In summary, our data underline that, whenever companion rescue fails, snow burial in an avalanche is associated with extraordinarily high mortality. Multivariate analysis showed short duration of burial and off-piste terrain to be the two independent predictors of survival. Despite the theoretical superiority of searching with rescue dogs, searching the avalanche debris with probe lines was seen to be equally effective in this study. Our data suggest that in the majority of cases the potential hazard for rescuers during avalanche rescue missions comes from self-triggered avalanches. Consequently, thorough mission planning and critical risk–benefit assessment are of utmost importance for risk reduction. Acknowledgment We thank Peter Veider, managing director of the Austrian Mountain Rescue Service in Tyrol for his support and valuable advice. References Brugger H., and Durrer B. (2002). On-site treatment of avalanche victims ICAR-MEDCOM-recommendation. High Alt. Med. Biol. 3:421–425. Brugger H., Durrer B., and Adler-Kastner L. (1996). On-site triage of avalanche victims with asystole by the emergency doctor. Resuscitation. 31:11–16. Brugger H., Durrer B., Adler-Kastner L., Falk M., and Tschirky F. (2001). Field management of avalanche victims. Resuscitation. 51:7–15. Brugger H., Falk M., and Adler-Kastner L. (1997). [Avalanche emergency. New aspects of the pathophysiology and therapy of buried avalanche victims.] Wien Klin. Wochenschr. 109:145–159. Etter H.J. (2004). Report of the Avalanche Subcommission at the general meeting of the International Commission of Alpine Rescue. Zakopane, Poland. Harvey S., Rhyner J., and Ammann W.J. (2007). Vom weißen Rausch zum schwarzen Tag - Lawinenrisiko beim Variantenfahren. #www.slf.ch" Hohlrieder M., Mair P., Wuertl W., and Brugger H. (2005). The impact of avalanche transceivers on mortality from avalanche accidents. High Alt. Med. Biol. 6:72–77. Radwin M.I., and Grissom C.K. (2002). Technological advances in avalanche survival. Wilderness Environ. Med. 13:143–152. Slotta-Bachmayr L. (2005). How burial time of avalanche victims is influenced by rescue method: an analysis of search reports from the Alps. Nat. Hazards. 34:341–352. Tschirky F., Brabec B., and Kern M. (2007). Lawinenrettungsgeräte—Stand der Entwicklungen, Erfolge und Misserfolge. In: Yearbook 2001 of the Society of Alpine and High Altitude Medicine. H. Brugger, G. Sumann, W. Schobersberger, and G. Flora, eds. Innsbruck, Austria; pp. 101–125.

Address reprint requests to: Matthias Hohlrieder, MD Department of Anesthesiology and Critical Care Medicine Innsbruck Medical University Anichstrasse 35, 6020 Innsbruck, Austria E-mail: [email protected] Received September 28, 2007; accepted in final form February 12, 2008.