Few areas of the avalanche emergency equipment market are experiencing a boom comparable to that of the avalanche airbag segment. More and more manufacturers and suppliers are trying to capture a slice of this growing pie with innovative ideas and solutions. Although the image of avalanche airbags has changed massively in a positive direction, there is still debate about how much an avalanche airbag can increase the probability of survival in an avalanche... This question is answered by a recent international study on the effectiveness of avalanche airbags.In the following we give you a summary of the study: Here you can read a popular scientific article in the specialist magazine BergUndSteigen... (The summary reproduced here is essentially based on this article). The original publication appeared in May 2014 in ResuscitationThe authors of the study are all renowned avalanche experts: Pascal Haegeli, Markus Falk, Benjamin Zweifel, Emily Procter, Fréderic Jarry, Spencer Logan, Kalle Kronholm, Marek Biskupic, Hermann Brugger The unique selling point of airbag technology: The major safety gain made possible by avalanche airbag systems is based on the fact that they are the only avalanche emergency system that can prevent the burial of a winter sports enthusiast caught in an avalanche or at least significantly reduce the burial depth. All previous field tests with avalanche dummies equipped with airbags have confirmed the effectiveness of avalanche airbag technology. Nevertheless, there is still much and occasionally heated debate about the extent to which an avalanche airbag can increase the chances of survival of winter sports enthusiasts caught in a dangerous avalanche or reduce their mortality rate. In particular, the marketing promises of the airbag pioneer ABS "97 percent survived!" provoke opposition from prominent avalanche experts such as Dale Atkins and from the critical freeride and touring scene. Although there are several studies that compare the chances of survival between winter sports enthusiasts with and without airbags on the basis of avalanche accident statistics, there are no recent studies on this. Another shortcoming of the studies published to date was their sparse database, which in the vast majority of cases took unsatisfactorily few accidents involving airbag users into account.
The study by Pascal Haegeli and colleagues was designed to find out how effectively avalanche airbags reduce the risk of death for airbag users, based on the most up-to-date, reliable figures and avalanche parameters possible. The article by Haegerli, Falk, Zweifel and Co. published in BergUndSteigen makes it clear how crucial the questions used to evaluate the accident statistics are for the results obtained. The guiding principles of the current airbag study are the following questions:(1) How much does the use of an avalanche airbag influence my probability of death if I am caught in a serious avalanche? (Small avalanches and so-called slides, as they do not allow reliable conclusions to be drawn about the risk of death, were consequently excluded from the evaluation). (2) How high is the proportion of avalanche deaths that could be avoided if avalanche airbags were more widespread?
The authors of the study used the accident data from the 2007 study by Brugger et al. The data set of this study contains 1504 avalanche victims in open ski terrain in Switzerland and Austria. The data originates from the period from 1990 to 2005. 35 of the avalanche victims in this data set were equipped with an avalanche airbag. The evaluation of the control group showed that 81 out of 100 winter sports enthusiasts without an avalanche airbag survived the avalanche. They owe their survival to the fact that they were either (1) not buried, (2) only slightly buried or (3) rescued in time. Of the 100 percent (= 35 avalanche victims) who used an airbag, 97 percent survived. This corresponds to a mortality rate of 3 percent and appears to support the manufacturer ABS's advertising claim. The use of an airbag reduces the probability of death from 19 to 3 percent, which would correspond to a reduction of 16 percent. The so-called mortality ratio compares the relatively low remaining mortality of airbag users with the original mortality without airbag use in the control group. This results in a mortality ratio of 15 percent, which means that out of 100 airbag users, 15 would still have died if they were caught in a serious avalanche.
Going beyond the last airbag study by Brugger, the authors wanted to obtain as reliable a picture as possible with the help of as broad and meaningful a data set as possible. To this end, they examined the data of all available avalanche accidents in which at least one person used an avalanche airbag. The data came from Canada, Switzerland, Norway, Slovakia, France and the USA. In order to rule out the possibility of the results being biased towards more-safety-than-the-airbag-can-offer, only those accidents in which the avalanche was large enough to cause a complete burial were included in the analysis. In the current study, only avalanche accidents in which (1) a person was recorded in the dangerous area of the avalanche (i.e. not just at the edge of the avalanche) and (2) those recorded were caught in a larger avalanche (= at least avalanches of size category 2 and larger.) After this "adjustment", the data set determined by Haegerli and colleagues contained 245 avalanche accidents with 424 "seriously" recorded winter sports enthusiasts. Of these, 264 people (= 58 percent) of the winter sports enthusiasts recorded had an inflated airbag. The airbag was not inflated on 61 (= 14 percent) winter sports enthusiasts and 117 winter sports enthusiasts were not equipped with an airbag (= 28 percent). It should not be forgotten that the official avalanche accident statistics contain a considerable uncertainty factor. The reason for this is simple: a not inconsiderable proportion of avalanches with a minor outcome are not reported to the avalanche centers. However, because the manufacturers of avalanche airbags - for understandable reasons - specifically look for information about avalanche accidents involving the use of airbags in order to publicize these cases, a certain statistical and media bias arises. (See also the avalanche accidents of Xavier de la Rue or Julien Lopez). It can therefore be assumed that the reporting frequency of avalanche accidents with avalanche airbags is higher than without airbags. In order to eliminate this problem as far as possible, the researchers only investigated avalanche accidents involving both people with and without avalanche airbags. In accordance with this requirement, only 106 out of 207 of the data records could be included in the analysis. In their statistical analysis of the accident data, the researchers attempted to take other survival factors - beyond the burial reduction effect of the avalanche airbag - into account in their statistical analysis. Factors such as the location of the victims, the type of avalanche, the terrain and, above all, the avalanche runout, use of avalanche transceivers, etc. were included in the statistical data analysis (for more information on the method and the complete table, see the original article in Resuscitation).
Results
Avalanche airbags only indirectly influence the probability of survival of avalanche victims, namely by (significantly) reducing the burial depth. Other significant factors that critically influence mortality are: (1) Avalanche size (the larger an avalanche, the higher the probability of being killed by the avalanche.
(2) Mechanical injuries (trauma injuries) suffered by the avalanche victim. If these factors are taken into account, the risk of critical burial for avalanche victims without an inflated airbag is 47 percent, whereas the adjusted risk for airbag users is 20.1 percent. MortalityThe mortality of an avalanche victim is largely determined by the degree of burial, the size of the avalanche and the occurrence of mechanical injuries. The adjusted mortality rate for critically buried victims is 43.8 percent, for non-critically buried victims only 2.9 percent. The adjusted mortality of a critical burial for airbag users is calculated as follows: The adjusted risk of a critical burial after deployment of the avalanche airbag is multiplied by the adjusted mortality depending on the degree of burial. [Editor's note: Sorry, the method of calculation is difficult to present in an easily understandable way; you may have to read the paragraph several times] The result of this calculation is an adjusted mortality rate for winter sports enthusiasts without an airbag of 22.2 percent, whereas the mortality rate with an inflated airbag is 11.1 percent.*** The results mean in practice:
(1) Of 100 seriously buried avalanche victims without an airbag, 22 would have died, meaning that 78 of them would have survived.
(2) Of 100 people with an inflated airbag, 11 would have died or 89 of them would have survived the avalanche accident. Interpretation of the results:Avalanche airbags could have prevented around half of the avalanche fatalities. That is a lot. But not as much as was previously assumed.
The mortality rate of airbag users is significantly higher than is often reported: 11 percent instead of the much-mentioned 3 percent!
The authors point out, however, that the results are influenced by the selection of the data set, which forms the basis of the study, because this data set included an above-average number of accidents with large avalanches and several buried victims.
One clear result of the study is that airbags can in no way guarantee survival in an avalanche.
Why do so many airbag users not deploy the airbag during an avalanche?
The strikingly large number of non-deployed avalanche airbags significantly reduces the mortality rate of airbag users: from 11 to 9 percentage points. This raises the question of the causes for the frequent non-deployment of avalanche airbags: Of the cases included in the data set, there was information on the cause of failure in 52 cases:
- 60 percent of the airbags were not deployed by the user.
- In 12 percent, maintenance errors led to system failure (e.g. incorrectly inserted cartridge).
- In 17 percent, the system failed due to a device error, e.g. due to design faults or technical failure.
- 12 percent of the airbags were destroyed during the avalanche. In relation to all avalanche accidents of avalanche airbag users evaluated, a total of two percent of the airbags were destroyed during the avalanche.
The fact that in three percent of cases a device or system error prevented the airbag from being deployed is strikingly high. [Editor's note: This raises the question of whether the current airbags available on the market have similarly high failure rates. There is some evidence to suggest that the failure rate of products from established manufacturers is lower. Of course, nothing can yet be said about the suppliers entering the market with their new products]. Do airbags increase the willingness of their users to take risks? In contrast to the freeriding and ski touring sector, the phenomenon of so-called risk compensation has been well researched in other areas, such as road traffic. Risk compensation means that the safety gain from safety programs or techniques is offset by less cautious behaviour on the part of users, who rightly feel "safer". It seems difficult to determine reliable empirical figures for the user group of avalanche airbag users. However, it can be cautiously concluded that such risk compensation, if it occurs, would reduce the expected benefit of the avalanche airbag. Limitations of the studyThe authors make the limited informative value of their study transparent: The biggest shortcoming of the data set of accidents involving avalanche airbag users was the significantly limited sample. In addition, the data set contained an above-average number of large avalanches.
Conclusion
The team of authors of the study led by Pascal Haegerli and colleagues draws the following conclusion:
- Avalanche airbags are a useful and valuable avalanche emergency device. Nevertheless, the reduction in mortality is lower than previously assumed.
- Avalanche victims who were seriously caught in a major avalanche show a reduction in mortality from 22 to 11 percentage points in the case of an inflated airbag.
- Non-inflated airbags are the most important limiting factor of avalanche airbags. If non-inflated airbags are included in the results of the study, the risk of death is only reduced from 22 to 13 percent.
- 60 percent of non-inflated airbags are due to a non-deployed system. This proves how important it is that users are familiar with their airbag and that it must also be regularly maintained!
- The safety gains that result for the airbag user can be lost again if the airbag users drive on more extreme terrain because of this safety gain, because they feel well protected with the airbag. Thank you The PG team would like to thank the researchers for this meaningful and long overdue research study, despite its limitations. Thank you 2***
Thank you to my colleague Patrick Wehowsky for the additional/correcting explanations regarding the method of calculating the adjusted mortality, because this method is difficult to understand in the text The following section is from Patrick: "To come to a conclusion regarding the mortality of avalanche victims with airbags vs. In order to arrive at a statement regarding the mortality of avalanche victims with airbags or without airbags (control group), two burial categories are distinguished for each group (airbag/non-airbag), which are later summarized again.
One category comprises the number of cases of critical burial (in percent), the other the number of cases (in percent) of non-critical burial. A critical burial is when breathing was severely impaired (head under the snow and airway impaired). The whole thing is explained here using the example of non-airbag users: As was to be expected, the risk of being "critically buried" without an airbag is quite high: specifically 47%. If there is a critical burial, the risk of dying as a result of the burial is quite high, namely 43.8 percent in relation to the study. These two values are now multiplied together to give the so-called "adjusted mortality". In the case of non-airbag users, this is 20.5% for "critical burial". However, this is only one side of the coin.
In addition to the 47% of "critical burials", the 53% of "non-critical burials" must also be taken into account. In this group, the risk of dying due to "non-critical burial" is quite low. Only 2.9 percent of people who are "not critically buried" die as a result of mechanical injuries. These two percentages are now multiplied together again - the result is an adjusted mortality rate of 1.5 percent, i.e. if I am "not critically buried", I am highly unlikely to die (98 out of 100 people do not die). In order to arrive at an adjusted mortality for a group, the adjusted mortality for "critically buried" and "non-critically buried" people must be added together.
For the group of non-airbag users, the adjusted mortality is therefore 20.5%+1.5% =22.1% Incidentally, this is an error both in the graph and in the text at BergUndSteigen). If we now compare the group of non-airbag users with the group of airbag users, one point stands out. At 20.1 percent, the probability of being critically buried (as theoretically assumed) is also empirically significantly lower than in the control group without airbags (47 percent).* And therefore the risk of dying is also significantly lower in the airbag group; the adjusted mortality rate here is 11.1 percent (compared to 22.1 percent in the control group). [*The deviations in the percentage figures for the adjusted mortality rate for "non-critically buried" 1.5% in the control group compared to 2.3% in the airbag group are not significant and are due to the general problem of data collection.
