Risk Assessment in Winter Backcountry Travel

Wilderness and Environmental Medicine, 20, 269 274 (2009) ORIGINAL RESEARCH Risk Assessment in Winter Backcountry Travel Natalie A. Silverton, MD; Scott E. McIntosh, MD; Han S. Kim, PhD, MSPH From the Division of Emergency Medicine, University of Utah, Salt Lake City, UT (Drs Silverton, McIntosh, and Kim). Objective. Risk assessment is an important part of safe backcountry travel in avalanche terrain. The purpose of this study was to determine and compare the ability of backcountry travelers to accurately estimate the avalanche danger for their destination and time of travel. Methods. We surveyed 353 winter backcountry users, asking them to rate the avalanche danger for their destination that day. We then compared this estimation to the Utah Avalanche Center daily advisory for that specific location, aspect, and elevation. Tendency to underestimate the avalanche danger was then compared across 6 different sports (backcountry skiing, backcountry snowboarding, snowshoeing, snowmobiling, out-of-bounds skiing, and out-of-bounds snowboarding) as well as across age, gender, and subject participation in an avalanche safety course. Results. A comparison across different sports, adjusted for age and gender, showed that snowshoers were 7.11 times more likely than skiers to underestimate the avalanche danger (, 2.95, 17.11). This difference was maintained after adjusting for past education in an avalanche safety course (odds ratio, 5.74;, 2.28, 14.46). Snowmobilers were also significantly more likely to underestimate the avalanche danger when compared to skiers (odds ratio, 3.11;, 1.12, 8.24), but these differences ceased to be significant when the data were adjusted for avalanche safety course (odds ratio, 2.39; 95% CI, 0.84, 6.74). While there was a trend for women and older age groups to underestimate the avalanche danger when compared to men, these trends were not significant. Conclusions. Snowshoers and snowmobilers are groups that tend to underestimate avalanche danger when traveling in the backcountry. These groups may be unknowingly assuming a higher risk and should be targeted for avalanche education and awareness. Key words: avalanche, risk assessment, safety, education, skiing, snowboarding, snowmobiling, snowshoeing Introduction Unlike a ski resort, the winter backcountry is not patrolled for hazards. Venturing into the backcountry can therefore involve a great deal of risk. In recent years, the number of avalanche fatalities in the United States has increased. 1 Efforts have been made to educate the public about the potential hazard that avalanches pose. This has been accomplished through avalanche education courses or through daily avalanche advisories. Although organizations can educate the public and provide advisories, it is ultimately up to the individual to make sound decisions. Safety in the backcountry depends in part on one s ability to correctly assess the avalanche danger. Travelers must decide whether it is safe to enter Corresponding author: Natalie A. Silverton, MD, University of Utah, Division of Emergency Medicine, 30 North Medical Dr, Salt Lake City, UT 84132 (e-mail: natalie.silverton@hsc.utah.ed). the backcountry on a particular day, where to travel, and what safety precautions and equipment to utilize. These risk assessment decisions are entirely within the control of the backcountry user. We evaluated the risk assessment skills of backcountry travelers engaging in a variety of winter sports in the Wasatch and Uinta mountains of Utah between November 2005 and March 2006. By doing so, we hoped to identify groups of backcountry travelers that could be targeted for avalanche education. Methods A convenience sample of backcountry users traveling in the Wasatch and Uinta mountains was surveyed during the 2005 06 winter season. Inclusion criteria were defined as those entering the backcountry whose destination involved travel into an area with potential

270 Silverton, McIntosh, and Kim avalanche hazard. Potential avalanche hazard was defined as traveling on any slope greater than 25u or crossing a major slide path. Users included backcountry skiers, snowboarders, snowshoers, snowmobilers, out-ofbounds skiers, and out-of-bounds snowboarders. Out-of-bounds users were defined as those leaving the ski resorts in order to travel in areas that were not controlled for avalanches. We assumed that accessing the backcountry from the ski resort requires less planning, forethought, and preparation than a traditional backcountry tour because of the ease of lift access. It is for this reason that we chose to distinguish between out-ofbounds resort skiers/snowboarders and traditional backcountry travelers. Participants were recruited at 12 different trailheads along the Wasatch Front and on a number of ski, snowshoe, and snowmobile trails in the Wasatch and Uinta mountains. Out-of-bounds skiers and snowboarders were recruited for the study at the out-ofbounds exit gates at 2 ski resorts, Brighton and The Canyons. These resorts were selected because access to the backcountry from these areas is easy, unlimited, and very popular. Both ski resorts do make efforts to inform the public about avalanche danger with graphic warning signs at the access gates. Data were collected using a written questionnaire. Information collected included demographics, mode of travel, and the location of the participant s destination. Participants were then asked to assess the avalanche danger that they expected for that destination. The aspect and elevation of the participant s destination were researched by the authors. Avalanche forecasts for each day of data collection were obtained from the Utah Avalanche Center (UAC) daily advisory for that morning and compared to the participant s estimation of the avalanche danger for the specific aspect and elevation of their destination. The UAC provides daily avalanche condition forecasts for the Wasatch and Uinta mountain ranges during the winter season. These forecasts are compiled by professional avalanche forecasters who monitor the snow conditions on a daily basis. The UAC uses a standard ranking system for avalanche danger, as follows: low, moderate, considerable, high, or extreme. Definitions for these terms can be found on the UAC website (www.avalanche.org/,uac/). For example, on a given day, south-facing slopes may be reported as presenting low danger, while north-facing slopes could be reported as presenting considerable danger. In that case, if a skier s destination was a north-facing slope, his/ her own assessment of the danger level was compared to the UAC forecast of considerable danger. The difference between perceived danger and that reported in the UAC advisory was then used to assess the participant s ability to correctly assess the avalanche danger. Participants were also asked if they had taken an educational avalanche safety course and if they had accessed the UAC daily advisory that morning. They were also asked to identify the type of avalanche safety equipment they were carrying. The latter information was presented in a separate paper entitled Avalanche Safety Practices in Utah, in which we compared the use of avalanche safety equipment among the different groups of backcountry travelers. 2 The survey took 3 to 5 minutes to complete. The study team recorded the date, time, location, weather conditions, whether the participant was traveling to or from his destination, and the type of equipment used (skis, snowshoes, etc). (The entire survey and the additional participant information recorded by the study team can be viewed online at http://www.wemjournal.org.) Participants were screened so that they only completed the survey once. We made efforts to collect data in all weather conditions and at various levels of avalanche hazard. Participants who overestimated the avalanche danger (reported that the danger was high, for example, when the UAC report said it was low) were grouped with those who correctly estimated the danger. Overestimating danger would encourage the backcountry traveler to avoid terrain and therefore decrease the risk of an avalanche accident. In contrast, underestimating the danger (stating that the danger was lower than what was reported) could actually increase the risk of traveling in avalanche terrain and the risk of being caught in an avalanche. Participants evaluation of the avalanche danger was then compared across age, sex, and sport using a logistical regression statistical technique. We chose backcountry skiers as a comparison group, because traditionally many of the pioneers of avalanche safety and many of the proponents of avalanche education have been backcountry skiers. Also, in our prior study we found that backcountry skiers were more prepared than other groups of backcountry travelers, as they tended more often to travel with an avalanche transceiver, a shovel, and a partner. 2 Logistic regression was also used to adjust for a prior history of an avalanche safety course, because this could influence the participant s ability to assess the avalanche danger. Odds ratios and s for underestimating the avalanche danger were then compared. All statistical tests were 2- sided with a significance level of a 5.05. The study was reviewed and approved by the University of Utah Institutional Review Board. SAS 9.1.2 (SAS Institute, Cary, NC) was used to analyze data. Results Three hundred and eighty-eight backcountry travelers were approached to participate in the survey. Of these, 35

Risk Assessment in Winter Backcountry Travel 271 Table 1. Demographics of No. of female No. of male Total No. of Mean age (y) Age range (y) Ski 26 79 105 38 21 83 Snowboard 2 28 30 32 23 50 Snowmobile 7 39 46 36 13 57 Snow shoe 23 33 56 45 16 74 OOB skier* 14 58 72 28 13 54 OOB snowboarder* 6 38 44 25 13 40 Total 78 275 353 *OOB indicates out of bounds. (9%) declined to participate. Table 1 shows the demographics of study. Table 2 shows the percentage of who either correctly estimated or overestimated the avalanche danger and the percentage of who underestimated the avalanche danger. The Figure shows a graphical representation of the percentage of underestimating the avalanche danger, separated by sport. A comparison across different sports and adjusted for age and sex (Table 3) showed that snowshoers were 7.11 times more likely than skiers to underestimate the avalanche danger (, 2.95, 17.11). This difference was maintained after adjusting for past education in an avalanche safety course (odds ratio, 5.74;, 2.28, 14.46) (Table 4). Snowmobilers were also significantly more likely to underestimate the avalanche danger when compared to skiers (odds ratio, 3.11; 95% CI, 1.12, 8.24) (Table 3), but these differences ceased to be significant when the data were adjusted for avalanche safety course (odds ratio, 2.39;, 0.84, 6.74) (Table 4). While there was a trend for women to underestimate the avalanche danger when compared to men, this trend was not significant (Tables 3 and 4). When odds ratios were compared across different age groups (Tables 3 and 4), there was a trend indicating that older age groups tend to underestimate the avalanche danger. This trend, however, was not significant. Table 2. Estimation of avalanche danger by sport, gender, age category, and avalanche course participation No. of % of Participants correctly estimating or overestimating the avalanche danger % of Participants underestimating the avalanche danger Ski 106 90.6 (85.0, 96.1) 9.4 (3.8, 15.0) Snowboard 30 86.7 (74.5, 98.8) 13.3 (1.2, 25.5) Snowmobile 44 77.3 (64.9, 89.7) 22.7 (10.3, 35.1) Snow shoe 55 54.5 (41.4, 67.7) 45.5 (32.3, 58.7) OOB skier* 66 92.4 (86.0, 98.8) 7.6 (1.2, 14.0) OOB snowboarder* 43 88.7 (78.8, 98.0) 11.6 (2.0, 21.1) Female 75 74.7 (64.8, 84.5) 25.3 (15.5, 35.1) Male 269 85.1 (80.9, 89.4) 14.9 (10.6, 19.2) Under 20 y of age 38 94.7 (87.6, 100) 5.3 (0.0, 12.4) 20 29 y of age 98 83.7 (76.4, 91.0) 16.3 (9.0, 23.6) 30 39 y of age 96 88.5 (82.1, 94.9) 11.5 (5.1, 17.9) 40 49 y of age 61 75.4 (64.6, 86.2) 24.6 (13.8, 35.4) 50+ y of age 51 70.6 (58.1, 83.1) 29.4 (16.9, 41.9) Prior avalanche course No 157 77.7 (71.2, 84.2) 22.3 (15.8, 28.8) Yes 187 87.2 (82.4, 92.0) 12.9 (8.1, 17.7) *OOB indicates out of bounds.

272 Silverton, McIntosh, and Kim Figure. Percent of underestimating the avalanche danger, with s. *OOB indicates out of bounds. Discussion While it may be true that ignorance is bliss, sometimes what you do not see can hurt you. In this study, we have shown that snowshoers and snowmobilers often underestimate the avalanche danger when traveling in the backcountry. In addition, these groups are less likely than other backcountry users to carry an avalanche transceiver or a shovel, to travel with a partner, or to have taken an avalanche safety course, 2 inadequate safety practices that may be a reflection of this tendency to underestimate the avalanche danger. Avalanche fatality statistics show that these groups are just as likely to die in an avalanche as other backcountry travelers, if not more so. Snowshoers and snowmobilers represent 17% and 28%, respectively, of the total number of avalanche fatalities in Utah in the last 9 years. 3 11 The proportion of fatalities of these groups is 2 to 3 times the number of skier fatalities during the same time period. Because there are no current data on the number of participant-days for each of these sports, it is unclear whether this discrepancy represents actual differences in fatality rates or simply a relative increase in the absolute number of snowshoers and snowmobilers compared to skiers. Regardless of the reason, these data indicate that close to half of the avalanche fatalities in Utah involve snowshoers and snowmobilers, and additional efforts need to be made to raise awareness of avalanche hazards among these groups. Out-of-bounds resort skiers and snowboarders tended to be fairly good at correctly estimating or at least overestimating the avalanche danger when compared to snowshoers and snowmobilers. The reason for this may be that the 2 ski resorts at which data were collected post signs that describe the avalanche danger for the day. These postings could be an effective means of alerting out-of-bounds skiers and snowboarders to the avalanche hazard for the day. In the past 9 years, 8 of 36 avalanche fatalities in Utah (22%) have involved out-of-bounds skiers or snowboarders. 3 11 This number is comparable to the numbers of snowshoer and snowmobiler deaths and indicates that other factors may be involved in the decision-making process that leads to risk-taking behavior and avalanche fatalities. Significant efforts have been made in this country to raise awareness of avalanche safety, both through media

Risk Assessment in Winter Backcountry Travel 273 Table 3. Likelihood of underestimating the avalanche danger represented by the odds ratio [All sports are compared to skiers. Data are adjusted for sport, age, and sex using analysis of variance.] Table 4. Likelihood of underestimating the avalanche danger represented by the odds ratio [All sports are compared to skiers. Data are adjusted for sport, age, sex, and prior avalanche course participation.] Odds ratio Odds ratio OOB skiers{ 1.11 (.35 3.54) OOB snowboarders{ 2.03 (.60 6.83) Snowshoers 7.11 (2.96 17.11)** Snowboarders 1.76 (.49 6.27) Snowmobilers 3.11 (1.17 8.24)** 20 29 vs under 20 y of age 3.39 (.68 16.84) 30 39 vs under 20 y of age 2.26 (.44 11.53) 40 49 vs under 20 y of age 4.09 (.77 21.80) 50+ vs under 20 y of age 4.11 (.75 22.54) Gender female vs male 1.53 (.76 3.06) {OOB indicates out of bounds. **5 Statistically significant. campaigns and avalanche safety courses. The goal of avalanche education is to provide the backcountry traveler with insight into the hazards of travel in avalanche terrain. This study identifies snowshoers and snowmobilers as 2 groups of backcountry travelers who tend to underestimate the avalanche danger. These groups should be targeted for education so that individual snowshoers and snowmobilers can make more knowledgeable and informed decisions about the amount of risk they are willing to undertake while traveling in the backcountry. LIMITATIONS The primary limitation of this study was that it employed a convenience sample of backcountry users. We attempted to sample at multiple trailheads, on multiple days, in all weather conditions, and with varied avalanche conditions throughout the season. However, the sample may not have captured an absolutely representative study population. Also, many of our accessed the UAC avalanche daily advisory on the day of travel. It is possible, and indeed very likely, that a large number of who correctly estimated the avalanche danger were able to do so because they had seen the report prior to venturing into the backcountry. We chose not to distinguish between those who were able to correctly estimate the avalanche danger by making their own assessment (through weather observation or snow pit studies) and those who gained their knowledge by reading the morning report. In this way we were measuring assessment of the danger but not the means by which this assessment was made. OOB skiers{ 0.89 (.27 2.94) OOB snowboarders{ 1.61 (.46 5.64) Snowshoers 5.74 (2.28 14.46)** Snowboarders 1.64 (.46 5.88) Snowmobilers 2.39 (.84 6.74) 20 29 vs under 20 y of age 3.73 (.74 18.7) 30 39 vs under 20 y of age 2.69 (.51 13.94) 40 49 vs under 20 y of age 4.54 (.84 24.44) 50+ vs under 20 y of age 4.83 (.87 26.81) Female vs male 1.47 (.73 2.95) Avalanche course: no vs yes 1.65 (.84 3.26) {OOB indicates out of bounds. **5 Statistically significant. Conclusion Snowshoers and snowmobilers tend to underestimate avalanche danger when traveling in the backcountry. This finding may help direct avalanche education organizations in their efforts to raise awareness of the dangers implicit in backcountry travel. References 1. Page CE, Atkins D, Shockley LW, Yaron M. Avalanche deaths in the United States: a 45-year analysis. Wilderness Environ Med. 1999;10:146 151. 2. Silverton N, McIntosh S, Kim H. Avalanche safety practices in Utah. Wilderness Environ Med. 2007;18: 264 270. 3. Tremper B. Snow and avalanches in Utah: annual report 1997 1998. Available from: http://www.avalanche.org/,uafc/ season%20reports/season%20report%2098%20acrobat.pdf. Accessed July 28, 2006. 4. Tremper B. Snow and avalanches: annual report 1998 1999. Available from: http://www.avalanche.org/,uafc/ season%20reports/season%20report%2099%20acrobat.pdf. Accessed July 28, 2006. 5. Tremper B. Snow and avalanches: annual report 1999 2000. Available from: http://www.avalanche.org/,uafc/season% 20reports/Season%20Report%202000.pdf. Accessed July 28, 2006. 6. Tremper B. Snow and avalanches: annual report 2000 2001, Forest Service Utah Avalanche Center. 2001. Available from: http://www.avalanche.org/,uafc/season %20reports/Season%20Report%202001.pdf. Accessed July 28, 2006. 7. Tremper B. Snow and avalanches: annual report 2001 2002, Forest Service Utah Avalanche Center. 2002. Avail-

274 Silverton, McIntosh, and Kim able from: http://www.avalanche.org/%7euac/season_report/ SeasonReport2002.pdf. Accessed July 28, 2006. 8. Tremper B. Snow and avalanches: annual report 2002 2003, Forest Service Utah Avalanche Center. 2003. Available from: http://www.avalanche.org/%7euac/ season_report/seasonreport02-03.pdf. Accessed July 28, 2006. 9. Tremper B. Snow and avalanches: annual report 2003 2004, Forest Service Utah Avalanche Center. 2004. Available from: http://www.avalanche.org/%7euac/season_report/ SeasonReport2004_PDF/Season_Report_2003-04.pdf. Accessed July 28, 2006. 10. Tremper B. Snow and avalanches: annual report 2004 2005, Forest Service Utah Avalanche Center. 2005. Available from: http://www.avalanche.org/%7euac/season_report/ SeasonReport2004-05/AnnualReport04-05.pdf. Accessed July 28, 2006. 11. Tremper B. Avalanche incidents and accidents 2005 06. 2006. Available from: http://www.avalanche.org/%7euac/ accidents_04-05.htm. Accessed August 1, 2006.