10 YEARS OF AVALANCHE DEATHS IN THE UNITED STATES. 2003/04 to 2012/13. Part I. Dale Atkins RECCO AB Avon, Colorado USA

Transcription

1 LEARNING OBJECTIVES 2003/04 to 2012/13 Part I Dale Atkins RECCO AB Avon, Colorado USA Lecture At the conclusion of this lecture students will be able to: I. Describe recent trends in US avalanche accidents II. Describe the typical avalanche victim III. Describe why experienced avalanche-aware people get caught and killed in avalanches IV. Describe the influence of human factors in avalanche accidents V. Present simple strategies to reduce the affect of human factors in decision-making. INTRODUCTION After reaching a twenty-year low in the late 1980s avalanche deaths in the United States soared during the 1990s (Figure 1) and spiked twice (2007/08 and 2009/10) with 36 deaths, the greatest number killed in the modern era (post 1950). Figure 1 also presents the five-year moving average that smoothes the data considerably. Since the end of the 1980s the average number of fatalities per winter rose from 11 to 30 but decreased slightly in the mid 2000s; however, in recent years the five-year average has edged back upwards to 30 deaths per winter. As more people head into the winter backcountry avalanches continue to be deadly. In the United States since 1950 avalanches have claimed 976 lives, and 29% (281) of those victims died during the last 10 winters. killed& 40" 35" 30" 25" 20" 15" 10" 5" 0" killed" 5)yr"moving"avg" 1950)51" 1952)53" 1954)55" 1956)57" 1958)59" 1960)61" 1962)63" 1964)65" 1966)67" 1968)69" 1970)71" 1972)73" 1974)75" 1976)77" 1978)79" 1980)81" 1982)83" 1984)85" 1986)87" 1988)89" 1990)91" 1992)93" 1994)95" 1996)97" 1998)99" 2000)01" 2002)03" 2004)05" 2006)07" 2008)09" 2010)11" 2012)13" Figure 1. Avalanche fatalities by winter, 1950/ /2013 and five-year moving average. 1 Dale Atkins, 2013

2 THE LAST 10 YEARS 2003/04 to 2012/13 In the past 10 years on average an estimated people are caught during each winter; 90, partly buried, 40 buried, 40 injured and 28 known killed. The median yearly direct losses to property were estimated at $100,000, though property losses varied significantly from year to year. Losses in the last 10 years ranged from a low of about $30 thousand to a high of $6 million for destruction caused by a 2008 avalanche that destroyed a portion of the highpower transmission line supplying Juneau, Alaska. However, actual costs were nearly three times greater. The city relied on diesel-powered generators for 45 days resulting in fuel costs of $10 million. Ten months later a much smaller avalanche struck the same transmission line. Repair costs tallied $1.7 million, and fuel costs totaled $3.6 million for the three weeks to run generators. (Mean losses tallied $262,000, however when fuel costs were added the mean value soars to $803,000.) At this point, a word of caution about all the accident statistics reported in this paper is necessary. These data are provisional and compiled using avalanche accident reports collected by the the Colorado Avalanche Information Center (CAIC), and by the author. During the past 20 years or so this data set has become increasingly biased towards avalanche fatalities. This trend has worsened in the past 15 years as most reported accidents involve avalanche fatalities with fewer detailed nonfatal accidents being reported. While nearly all U.S. fatalities are reported to CAIC, fewer and fewer nonfatal accidents are being reported. This absence of nonfatal data significantly affects the accident statistics, skewing the statistics against survival. We have no reason to think that the incidence of nonfatal accidents is decreasing; they are just not being reported and documented. Thus, the reader should keep this in mind when interpreting the data. Please spread the word that all accidents should be documented and reported to the and/or the Colorado Avalanche Information Center. A short avalanche accident reporting form can be found in the Snow, Weather, and Avalanche: Observational Guidelines for Avalanche Programs in the United States. THE MODERN AVALANCHE VICTIM In the United States since 2003/04, 281 people died in 237 fatal avalanche accidents. Data from these accidents have been compiled to draw a profile of the modern avalanche victim. About 9 in 10 avalanche victims are male and this statistic has remained constant for decades. This steady statistic is surprising as in the last two decades many more women are traveling in avalanche terrain. 1 Estimates are extrapolated for the US from accident data collected by the Colorado, Utah, Washington, Wyoming, Montana, Idaho, California, and New Hampshire avalanche centers. 2 Dale Atkins, 2013

3 Men Women Gender 258 (92%) 23 (8%) Age (yr mean) Std ± 12.4 ± 16.0 Min 8 13 Max 77* 85* Table 1. Demographics of avalanche victims killed, 2003/ /13. (Note the oldest victims were killed by roof avalanches. The oldest recreationalist victims were 65, male; and 56, female.) Avalanche victims are getting older! While figure 2a shows that those suffer the most deaths, more older people are getting killed. The mean age for all victims is 35.8 (median, 33.5); which is up significantly from a decade ago. In the the 1990s the average age hovered around 32; only about 5% of victims were older than 50. Today 17.5% of avalanche victims are 50 and older. When gender is considered the average age varies significantly. The average age for men is 36, while the average age for women is 40 (M-W, P<0.0001). Figure 3 shows most avalanche victims are killed in the wintertime. January, February, and March are the most deadly months. In the past couple of years there has been an increase in spring and early summer accidents as both riders and climbers venture into snow-filled couloirs < fatalities Figure 2a. Avalanche fatalities by age groups, 2003/ /13. (N=280) Figure 4 shows avalanche fatalities by state. While avalanche deaths occurred in 13 states in the past 10 years, in previous years Arizona, Main, New Mexico, and New York also experienced deaths. Colorado has held the dubious title as leader for decades, a likely result of population, a generally weak snow cover and easy access to mountainous terrain. Interestingly, Colorado s average number killed has remained remarkably steady for two decades while deaths in other states have increased. In the early 1990s one-third of all U.S. avalanche deaths occurred in Colorado. Today that number is down to 19%, which reflects the growing recreational opportunities in other western states. While Colorado leads in total 3 Dale Atkins, 2013

4 Male Female < fatalities (percentage) Figure 2b. Avalanche fatalities by age groups and gender, 2003/ /13. (N=280) SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG fatalities Figure 3. Avalanche fatalities by month, 2003/ /13. (N=281) avalanche deaths table 2 shows Colorado is only ranked sixth when avalanche deaths are weighted by population. Alaska, Wyoming, Montana, Idaho and Utah experience significantly higher mortality rates when considering their populations. 4 Dale Atkins, 2013

5 State Deaths per year (Rank) Deaths per million per year (Rank) AK 3.6 (3) (1) CA 1.8 (8) (12) CO 5.9 (1) (6) ID 2.5 (7) (4) MT 3.3 (5) (3) ND 0.1 (13) (10) NH 0.2 (9) (9) NV 0.1 (11) (13) OR 0.2 (10) (11) UT 4.1 (2) (5) VT 0.1 (12) (8) WA 3.4 (4) (7) WY 2.8 (6) (2) Table 2. Average deaths per year and deaths per million population per year, 2003/ /13. Figures 5a and 5b show avalanche fatalities by activity group. In the last 10 years 96% of the victims were pursuing some form of recreation at the time of the accident. Most of these victims were those seeking the steep and deep. Skiers & snowboarders (combined) lead snowmobilers by a large number (Figure 5a). Nearly 1 in 2 avalanche deaths involves a skier or snowboarder. Only 1 in 3 deaths involve a snowmobiler. However because skier and snowboard deaths occur in resorts, adjacent to resorts, and in the backcountry, the skier and snowboard category can be further segregated (Figure 5b). The category miscellaneous recreation includes snowshoers, snowplayers, hikers, and a heli-skier. First the time any many years no hunters were killed. Among the non-recreation groups, the workers category includes a two workers killed by roof avalanches, one participant in an avalanche school, three ski patrollers and one avalanche forecaster. For the first time any many years no motorist or highway worker made the list fatalities CO UT AK WA MT WY ID CA NH OR NV VT ND Figure 4. Avalanche fatalities state, 2003/ /13. (N=281) 5 Dale Atkins, 2013

6 skiers & snowboarders 124 snowmobilers 94 climbers 31 misc. recreation workers residents 6 motorists fatalities Figure 5a. Avalanche fatalities by activity categories, 2003/ /13 (N=281) snowmobilers 94 backcountry skiers 54 backcountry snowboaders 30 out-fof-area skiers 18 out-fof-area snowboarders in-area skiers/snowboarders climbers 31 misc. recreation 15 motorists/highway workers patrollers residents work fatalities Figure 5b. Avalanche fatalities by expanded activity categories, 2003/ /13. (N=281) 6 Dale Atkins, 2013

7 The overall experience level of avalanche victims varies from very little to the knowledgeable, seasoned mountaineer, backcountry skier to professional ski patroller and mountain guide The majority of victims tend to be competent and experienced winter adventurers. In the last 15-plus 0 0 5% 0 1% 9% 29% 42% 15% years the majority of avalanche victims had at least some level of avalanche awareness training, and many of these victims had advanced levels (several courses, or multi-day courses and several seasons recreating in avalanche terrain) of avalanche awareness. Though threequarters (72%) of victims had at least Avalanche Awareness Training advanced some none novice intermediate advanced Activity Skill Level Table 2. Relationship between avalanche awareness training and activity skill level, U.S. avalanche fatalities, 1997/98 to 2006/07 (n=86) Note: old data is presented but little has changed in recent years. some level of avalanche awareness training, most victim s avalanche training lags far behind their activity skill level. THE AVALANCHE SETTING There is a recent trend toward more accidents in the backcountry and occurring further into the backcountry than even just five years ago. Today backcountry accidents out number accidents in and near developed areas almost four to one. (The backcountry traditionally has been defined as any area where rescuers are more than 30 minutes away.) Nearly four in five accidents (78%) occurred in the backcountry and about one in five (22%) occurred in or near developed areas (within 2 miles). When looking at only skiing and snowboarding accidents 40% of accidents occurred within 2 miles of developed areas. In 213 accidents the avalanche type and trigger were reported, and during the last ten years nearly all fatal-avalanches (94%) were human triggered either by the victim, a member of the victim s party, or in at least two cases by another party. Nearly all fatal avalanches (93%) involved slab avalanches (Table 3). The data also show that most killer avalanches are small to medium in size, and usually result in one fatality. Of the 237 avalanche accidents, 203 (86%) claimed one life; 27, two lives; 5, three lives; 1, four lives; and 1, five lives. The worst accident in the modern era (post 1950) occurred on Mount Rainier where 11 died in an ice avalanche on June 21, Dale Atkins, 2013

8 SURVIVAL STATISTICS 100% Time A completely buried victim has a poor chance of survival. To obtain a better picture of the associated mortality and morbidity of avalanche victims, we will also look to experiences from other parts of the world. Burial is defined as any victim whose head is buried under the snow even though the rest of the body may be exposed. The Swiss have the best dataset of buried avalanche victims, and Figure 6 shows survival decreases with increasing burial time. In the first 15 minutes, about 9 in 10 avalanche victims are found alive. There is a good chance the victim can be found alive, if the victim s companions can perform a speedy rescue. After 15 minutes survival plummets. Survival probability At 30 minutes, an equal number are found dead as alive. After 30 minutes the survival rate continues to diminish. At 45 minutes only about 25% survive. At two hours fewer than one in ten buried victims survive. Time is the enemy of the buried victim. A speedy rescue is essential to finding a buried victim alive. 80% 60% 40% 20% 0% Time buried under avalanche (min) Figure 6. Data from Switzerland for survival probability for completely buried avalanche victims in open areas (n=638) in relation time buried (minutes) under the snow. Median extrication time was 37 minutes. The dotted curve represents the survival function for completely buried avalanche victims in open areas (n=422) based on the Swiss data for as calculated by Brugger and Falk. United States Canada Switzerland duration of burial, min no. extricated no. (%) survived no. extricated no. (%) survived no. extricated no. (%) survived (85.60) (89.5) (93.6) (37.3) (36.2) (71.1) (30.4) 29 7 (24.1) (43.9) (10.5) (4.0) (16.3) (0) (0) (0) (25.0%) (25.0%) (7.1%) ALL (39.2) ( (46.9) Table 3. Proportion of people that survived complete burials, by duration. US data (2001/ /11). Canada and Switzerland data from Haegeli et al Dale Atkins, 2013

9 Table 3 shows that at least a few lucky victims do survive for extended periods of time under the snow. (Note: The Canadian and Switzerland statistics for 36+ minutes combine longersurviving incidents into the one class. The US data is segregated to 360 minutes and then longer.) The US data shows that 2.4% (7 of 286) of buried victims survived burials longer than 5 hours. Burial times were 5, 5, 6, 8, 10, 23 and 24 hours. Rescuers should not prematurely shift attitudes from rescue to recovery after just a few hours or even after more than half a day. A very few lucky victims have been known to survive much longer. Burial Depth For many years researchers in the United States felt that survival was dependent on both burial time and burial depth; however, today most researchers feel burial time is most important. Deeper burials take longer to dig which leads to longer burial times. Table 4 shows that those killed in avalanches tend to be buried deeper. The mean burial depth for all buried avalanche victims is 3.8 feet. This is the depth that transceiver users should practice at. To date, no one in the burial (feet) alive dead composite N mean stdev median max min Table 4. Summary statistics for buried burial depth for victims in direct contact with snow, 2003/ /13. United States has survived a burial deeper than 10 feet. However, elsewhere in the world, lucky victims have survived deeper burials. In Switzerland two mountain guides buried 21 and 23 feet survived (F. Tshirky and B. Durrer M.D., personal communication). In France an off-piste skier survived a 26-foot burial. (C. Larcher personal communication). Rescuers dug for four hours to uncover the men. Injuries Avalanche victims both buried and not buried get injured and sometimes those injuries are very serious. The medical community agrees that on average that 75% of buried victims die of asphyxiation, about 25% from trauma, and very few from hypothermia. These average values vary with geographical location and the terrain over which the avalanche runs. Traumatic injury serious and often body-altering injuries is usually the a result of the terrain where the avalanche occurred. Data from Europe, Utah and British Columbia shows that serious traumatic injuries occur to about 25% of avalanche survivors. The British Columbia data also shows that the death rate by trauma varies by the activity from a low of 9% for snowmobilers to a high of 42% for ice climbers. Common traumatic injuries include fractures and dislocations to arms, legs and pelvis; soft tissue injuries to the belly and chest including major lacerations to internal and external organs; and head and face injuries. Victims swept into rocks and trees will suffer major traumatic injuries that can quickly lead to death. In 2011 a backcountry skier s leg was severed when he was swept through trees. He died within 15 minutes. Though hypothermia rarely kills avalanche victims, nearly all buried and seriously injured victims will incur hypothermia. 9 Dale Atkins, 2013

10 The International Commission for Alpine Rescue views an avalanche accident and especially when a burial is known or suspected as a medical emergency and recommends early notification of rescuers to be a essential action. RESCUE STATISTICS Type of Rescue A buried victim s chance for survival directly relates not only to length of time and burial depth but also to the type of rescue. Table 5 compiles the statistics on survival as a function of the type of rescue. Obviously, buried victims rescued by party members or another group nearby the accident site have a much better chance of survival than those found by organized Self Rescue Found by Companions Found by Organized Rescue Team Alive # 9 (8%) # 91 (81%) # 13 (11%) 113 Dead # # 97 (43%) # 130 (57%) 227 Table 5. Type of rescue for buried avalanche victims in direct contact with snow, 2003/ /13 (read left to right). rescue groups time being the major influencing factor. Of all those found alive, companions rescued 81%; an organized rescue party rescued 11%. However, when Table 5 is read vertically rather than horizontally the statistics become discouraging. Of those Found by Companions only 49% (91 of 188) were found alive. This means simply one s best chance of live recovery is by their companions; however, one s chances of being found alive by companions is only 50/50. Another way of saying this is that companions find a dead person for every one person found alive. Method of Rescue Table 6 describe the method of rescue for 281 buried avalanche victims; unfortunately the data is becoming more biased toward mortality because of less reporting of burials with survivors. Fifty-two percent of victims who were buried with a body part, like a hand or foot, or an attached object, like a ski tip, protruding from the snow were found alive. This statistic shows the advantages of a shallow burial less time required to search and shorter digging time. Of the fatalities in this category, some were traveling solo with no one to spot the hand or ski tip and rescue them. In several incidents the search was abandoned prematurely to seek organized rescue. When rescue teams arrived they found a skis or hand sticking out of the snow. The avalanche transceiver (aka beacons) has been used to find more victims in the past 10 years than any other method, and the use of transceivers reveals both good news and bad news. A transceiver is the best method for a companion to find a completely friend, if carried and used correctly. The bad news is the mortality rate of reported incidents over the last 10 years has been rising steadily and now stands at 66% (82 of 125). There are several likely reasons for this increase. First, more people are using transceivers; however, these users knowingly or unknowingly may take greater risks when wearing a transceiver, and eventually their luck runs out. There were also several cases in which the signal was quickly Total 10 Dale Atkins, 2013

11 Method Found Alive Found Dead Total Attached object or body part Spot probe Coarse or fine probe Rescue Transceiver Avalanche dog Voice Other (digging, RECCO) Found after a long time span Not found, not recovered 4 4 Inside vehicle 1 1 Inside structure 0 Totals Table 6. Method of rescue for buried avalanche victims, 2003/ /13. located but companions did not carry shovels. In other cases, members in the group could not locate signal and had to wait for trained rescuers to arrive. Also, are reports of companions who took 1-2+ hours to locate their friend inferring they could not use their transceiver effectively, or maybe fast enough to save a life. Despite improvements in usability, users continue to not practice, make mistakes, forget to carry one, or even purposely choose not to carry a transceiver. In the past four winters there have been at least six cases where very experienced and avalanche savvy people chose purposely to not take their transceiver. All died in avalanche burials. An avalanche rescue transceiver and shovel in trained hands of a companion are a buried victim s best hope for survival, but it is still no guarantee of a live recovery. Even in textbook rescues the signal is quickly located and the victim is dug out in a short time most victims do not survive. Organized probe lines have recovered the next greatest number of buried victims, but because of the time required, nearly all (91%) are recovered dead. Only 3 were found alive by this method, 32 were recovered dead. Despite the insulating properties of snow, eight victims buried shallowly were able to yell and be heard by their rescuers (voice). The Other category includes two victims found alive and four dead by random digging plus three victims found by a RECCO detector. In all cases the victims were deceased. In one case the search effort could not be initiated until 10 days after the accident; however, only on scene for minutes the detector quickly located the man s buried snowmobile. He was found immediately next to his machine. In a second case the detector was used to find the body of an avalanche victim whose transceiver had been turned off, but then eight feet of snow reburied the body. In a third case, the victim died of severe trauma before the rescue was started. 11 Dale Atkins, 2013

12 Trained avalanche-search dogs are capable of locating buried victims, but like all tools of organized rescue teams there have been very few live rescues. Since 1950 there have only been eight reported live recoveries. Six of the eight involved trained dogs: four at ski areas, one along a highway and one in the backcountry. Two other burials involved personal dogs who found their owners in the backcountry. Burial depths of those found alive ranged from 1 to 5 feet. Burial depths of those killed have ranged up to 15 feet. In the search of a destroyed house a rescue dog alerted for victim buried 35 feet. However, the median burial depth for all burials is only 3 feet. It should be noted that dogs find scent and not people, though sometimes they find both. On numerous occasions dogs alerted in the vicinity, which may extend out as far as feet from a buried victim. These scent clues then led to searchers finding victims with other technologies. For 10 buried victims, the search was called off after several days. The bodies were recovered weeks to months later. In the category not found, not recovered, four victims were never found. The four shared a rope and were swept into a glacial crevasse. CONCLUSIONS Some important conclusions can be drawn from this review. 1. Avalanche victims trigger their own avalanche. Very seldom is an accident a case of wrong-place-wrong-time; victims choose to visit hazardous slopes during unstable periods. Avalanche survival is best practiced by avoidance: learning where avalanches occur and when avalanches might occur so steeper terrain can be avoided when conditions are dangerous. Conventional wisdom says the best way to learn avoidance is through a three-part process of education, deliberate practice, and experience. 2. Knowledge, deliberate practice and experience are critical to staying alive in avalanche terrain; however, these same qualities when misused can lead directly to accidents and deaths. A sobering reminder is the simple fact that nearly threefourths of avalanche victims had at least some level of avalanche awareness training, and that nearly one in three victims had a significant amount of training. Sometimes bad decisions result in bad outcomes, but more unsettling is that sometimes even good decisions result in bad outcomes. 3. People caught in avalanches get hurt, and often seriously. People working and recreating in avalanche terrain should be ready to call for help, provide first aid to an injured friend and to attract rescuers to their location. 4. Burial statistics do not favor the buried victim more will die than will live. The bottom line is that only one in four completely buried no part sticking above the surface and not in vehicles or buildings victims survives. This 25% survival rate of totally buried victims has remained consistent for decades. Even when an object 12 Dale Atkins, 2013

13 or body part is visible on the snow surface the survival rate is still poor. Table 6 shows a dismal survival rate for the buried victim at only 53%. This value is suspect because for the 30 years prior to this 10-year study period the survival rate has hovered around 60 65%. It seems the bias caused by not reporting, or incomplete (missing data) reporting, of buried survivors has caused the general survival rate to drop. 5. Surviving an avalanche burial is luck. Devices like transceivers, airbags, Recco reflectors, and Avalungs help put one in a place to be lucky, but these devices do not guarantee survival. One should never rely on luck to survive. 6. Some lucky buried victims do survive for many hours. No rescue should be abandoned prematurely on the assumption that the victim could not possibly be alive. For even up to 24 hours, some victims have been found alive, and no avalanche victim should ever be denied this small chance at life. When we consider the large numbers of people exposed to avalanche hazard each winter compared to the annual average number of avalanche fatalities (30 for the last five years), there is a low probability that someone will be caught in an avalanche at a given time and location. However, once a victim is buried, chances for survival are low about one in four. Thus, prevention of accidents is the key to saving lives. No rescue method or device is an adequate substitute for proper route finding and decision-making. But when an accident does occur, speed of recovery is still the critical factor in live rescues. More than three-fourths (81%) of all live rescues were made by the victim s companions. While one s best chance of being found alive is in the hands of companions, everyone must remember the actual chances of being found alive even by companions is only slightly better than 50/50. Pick your friends carefully. 13 Dale Atkins, 2013