Avalanche survival depends on the time of day of the accident: A retrospective observational study.

INTRODUCTION: We aimed to investigate the relationship between the time of the day and the probability of survival of completely buried avalanche victims. We explored the frequency of avalanche burials occurring after sunset, and described victims' characteristics, duration of burial and rescue circumstances compared to daytime avalanches. METHODS: In this retrospective, observational study, we analysed avalanche data from the registry of the Swiss Institute for Snow and Avalanche Research, from 1998 to 2020. RESULTS: A total of 3892 avalanche victims were included in the analysis, with 72 of the accidents (1.85%) occurring in the nighttime. Nearly 50% of the victims involved in nighttime avalanche accidents were completely buried, compared to about 25% of victims in daytime avalanches. Completely buried victims were rescued by a companion less often at night than in the daytime (15% vs. 51%, p <.001). The search and rescue of completely buried avalanche victims took longer during the nighttime compared to the daytime (median 89 min vs 20 min, p =.002). The probability of survival decreased as the day progressed; it was highest at around midday (63.0%), but decreased at sunset (40.4%) and was the lowest at midnight (28.7%). CONCLUSIONS: Avalanche accidents at night are a rare event, and probability of survival after complete burial is lower during the nighttime compared to the daytime. The most relevant reason for this is the longer duration of burial, which is explained in part by the lower rate of companion rescue and the lower rate of victim localisation with an avalanche transceiver.

Burial duration, depth and air pocket explain avalanche survival patterns in Austria and Switzerland.

AIM: To calculate the first Austrian avalanche survival curve and update a Swiss survival curve to explore survival patterns in the Alps. METHODS: Avalanche accidents occurring between 2005/06 and 2012/13 in Austria and Switzerland were collected. Completely buried victims (i.e. burial of the head and chest) in open terrain with known outcome (survived or not survived) were included in the analysis. Extrication and survival curves were calculated using the Turnbull algorithm, as in previous studies. RESULTS: 633 of the 796 completely buried victims were included (Austria n=333, Switzerland n=300). Overall survival was 56% (Austria 59%; Switzerland 52%; p=0.065). Time to extrication was shorter in Austria for victims buried ≤60min (p<0.001). The survival curves were similar and showed a rapid initial drop in survival probability and a second drop to 25-28% survival probability after burial duration of ca. 35min, where an inflection point exists and the curve levels off. In a logistic regression analysis, both duration of burial and burial depth had an independent effect on survival. Victims with an air pocket were more likely to survive, especially if buried >15min. CONCLUSION: The survival curves resembled those previously published and support the idea that underlying survival patterns are reproducible. The results are in accordance with current recommendations for management of avalanche victims and serve as a reminder that expedient companion rescue within a few minutes is critical for survival. An air pocket was shown to be a positive prognostic factor for survival.

Risk of Avalanche Involvement in Winter Backcountry Recreation: The Advantage of Small Groups.

OBJECTIVE: Avalanches are the primary hazard for winter backcountry recreationists and cause numerous deaths and injuries annually. Although recreationists usually travel in groups, there is little empirical knowledge on group-related risk factors. This study aims to explore the relative risk of avalanche accidents with respect to group size and to discuss underlying reasons for different risk levels. METHODS: We compared backcountry usage data in regions in Switzerland and Italy with avalanche accident data in these regions. RESULTS: We found higher avalanche risk for groups of 4 or more people and lower risk for people traveling alone and in groups of 2. The relative risk of group size 4, 5, and 6 was higher compared with the reference group size of 2 in the Swiss and Italian dataset. The relative risk for people traveling alone was not significantly different compared with the reference group size of 2 in the Italian dataset but was lower in the Swiss dataset. CONCLUSIONS: These findings are in accordance with avalanche safety recommendations regarding the higher risk of large groups but not regarding lower risks of people traveling alone in avalanche terrain, which is not recommended and requires great caution. Further studies on backcountry usage are necessary to improve our understanding of human behavior and risk factors. New techniques (eg, video monitoring) may be useful for acquiring reliable data on backcountry usage.

The effectiveness of avalanche airbags.

AIM: Asphyxia is the primary cause of death among avalanche victims. Avalanche airbags can lower mortality by directly reducing grade of burial, the single most important factor for survival. This study aims to provide an updated perspective on the effectiveness of this safety device. METHODS: A retrospective analysis of avalanche accidents involving at least one airbag user between 1994 and 2012 in Austria, Canada, France, Norway, Slovakia, Switzerland and the United States. A multivariate analysis was used to calculate adjusted absolute risk reduction and estimate the effectiveness of airbags on grade of burial and mortality. A univariate analysis was used to examine causes of non-deployment. RESULTS: Binomial linear regression models showed main effects for airbag use, avalanche size and injuries on critical burial, and for grade of burial, injuries and avalanche size on mortality. The adjusted risk of critical burial is 47% with non-inflated airbags and 20% with inflated airbags. The adjusted mortality is 44% for critically buried victims and 3% for non-critically buried victims. The adjusted absolute mortality reduction for inflated airbags is -11 percentage points (22% to 11%; 95% confidence interval: -4 to -18 percentage points) and adjusted risk ratio is 0.51 (95% confidence interval: 0.29 to 0.72). Overall non-inflation rate is 20%, 60% of which is attributed to deployment failure by the user. CONCLUSION: Although the impact on survival is smaller than previously reported, these results confirm the effectiveness of airbags. Non-deployment remains the most considerable limitation to effectiveness. Development of standardized data collection protocols is encouraged to facilitate further research.

The impact of avalanche rescue devices on survival.

BACKGROUND: Within Europe and North America, the median annual mortality from snow avalanches between 1994 and 2003 was 141. There are two commonly used rescue devices: the avalanche transceiver, which is intended to speed up locating a completely buried person, and the avalanche airbag, which aims to prevent the person from being completely buried. OBJECTIVE: This retrospective study aimed to evaluate whether these avalanche rescue devices had an effect on mortality. METHODS: The study population was 1504 persons who were involved in 752 avalanches either in Switzerland from 1990 to 2000 and from 2002 to 2003 (1296 persons, 86.2%) or in Austria from 1998 to 2004 (208 persons, 13.8%). RESULTS: Persons equipped with an avalanche airbag had a lower chance of dying (2.9% versus 18.9%; P=0.026, OR 0.09, n=1504). In persons who were completely buried, without visible or audible signs at the surface and who did not rescue themselves (n=317), we found a lower median duration of burial (25min versus 125min; P<0.001) and mortality (55.2% versus 70.6%; P<0.001, OR 0.26) in those using an avalanche transceiver than in those not using the device. CONCLUSIONS: Our data showed that both the avalanche airbag and the avalanche transceiver reduce mortality. However, to improve the evaluation of rescue devices in the future, the data collection procedures should be reviewed and prospective trials should be considered, as the reliability of retrospective studies is limited.