Because of the recent, tragic fatality on Gobbler’s Knob, the Utah Avalanche Center has received many questions about the effectiveness of avalanche airbags. The skier caught in the avalanche deployed their airbag but died anyway. So what gives? Here’s the short answer followed by a quick primer on avalanche airbags and where the statistics come from:

Answer:

According to the latest, peer-reviewed, statistical study, a deployed avalanche airbag will reduce mortality by 50% (22% vs. 11% in the study). But because 20% of the people in the dataset were not able to deploy their airbags, if you include the non-deployments, avalanche airbags reduce mortality by 41%.

In other words, avalanche airbags work, but certainly not all the time. Why?

1) About one out of four avalanche fatalities occur from the trauma of hitting trees and rocks on the way down, often at freeway speeds. In which case, all the rescue gear in the world won’t help you.

2) Airbags are much less effective if the avalanche drains into a narrow terrain trap, like a gully, where even a small avalanche can bury someone very deeply.

3) Avalanches often come down in two or more waves as different parts of the starting zone release. If someone is caught in the first wave and end up on the surface, sometimes the secondary wave can bury them even with a deployed airbag.

4) Airbags depend on the motion of the avalanche to “float” us to the surface—like shaking a bag of potato chips to make the large pieces rise to the surface. Sometimes the avalanche is too small or does not travel far enough for the “inverse segregation” process to work.

5) In about 20 percent of cases, the person caught does not—or can not—pull the trigger.

A Quick Primer on Avalanche Airbags and Statistics Behind the Study:

If we trigger an avalanche, we pull the trigger on the pack shoulder strap, which inflates the bags on the outside of our pack, making us a much larger object, and we tend to rise to the surface.

Why? Avalanche airbags don’t “float” you to the surface like a life jacket floats us to the surface of water. Instead, they work because of the “inverse segregation” process in a granular flow where larger objects rise to the top. Assignment: go shake a bag of potato chips to watch the large chips rise to the surface and the crumbs sink to the bottom. Unlike a life jacket in water, if the avalanche debris is not in motion or doesn’t travel far enough for the process to work, there’s no floatation.

Unfortunately, marketing by some manufacturers on the effectiveness was overly optimistic, especially in the early days. But even within the past couple years, I have read taglines of “97% Effective” or something similar. There’s a couple big problems with this number: first, the earlier studies included everyone that was caught in an avalanche that was wearing an airbag pack including many people who would never have been buried anyway, such as in small avalanches or when people were able to escape out of the avalanche. Second, a tagline like “97% Effective” doesn’t tell you the other half of the story; about 90 percent of people who get caught in avalanches will survive anyway even if they are not wearing airbags. Luckily for us, avalanches are surprisingly benevolent.

In order to give us more realistic numbers, a group of many of the world’s top experts on avalanche rescue and medical aspects of avalanche accidents published a 2014 study in the medical, peer review journal, Resuscitation. Researchers included Pascal Haegeli, Markus Falk, Emily Proctor, Benjemin Zweifel, Fredrick Jarry, Spencer Logan, Kalle Kronholm, Marek Biskuspic and Herman Brugger. (http://dx.doi.org/10.1016/j.resuscitation.2014.05.025)

They looked at a total of 424 documented avalanche accident victims from throughout the world who were “seriously involved” in an avalanche accident. In other words, they filtered out smaller avalanches and ones where people escaped off to the side and only looked at avalanches where an avalanche airbag had a chance to make a difference. In addition, they only considered incidents with multiple involvements including both people with and without avalanche airbags so they could make a statistically valid “apples to apples” comparisons. Because of these filters, as they admit, the dataset is skewed towards a more pessimistic effectiveness of avalanche airbags but it was statistically the only way they could test the effectiveness of the technology.

The results…drum roll please… a deployed avalanche airbag will reduce mortality by 50% (22% vs. 11% in the study). But because 20% of the people in the dataset were not able to deploy their airbags, if you include the non-deployments, avalanche airbags reduce mortality by 41%.

In other words, even in this more pessimistic dataset, avalanches are still surprisingly benevolent with a mortality of only 22%, even without an avalanche airbag. With an inflated airbag, it cuts the mortality in half, to 11%. So if you considered 100 people who die in avalanches without an airbag, an inflated airbag would save about 50 of them.

In our recent Know Before You Go video (https://vimeo.com/144545554), we included a cool animation that helps visualize these statistics. If you add everything together—fatalities by trauma, those saved by transceivers, saved by airbags, those who die in terrain traps or secondary burials, etc—our avalanche rescue gear will only save about half of the people involved in a serious avalanche accident. The take home sound bite: Avalanche rescue gear will save only half us. But at the same time, we can't afford to leave home without it.

We also have to remember that terrain consequences are extremely important. If the avalanche we may trigger will take us over a cliff, into trees, down a very large avalanche path, or into a terrain trap, then we have a low chance of survival, despite whatever rescue gear we may carry.

Finally, we have to remember that if we use any safety device to justify increased risk (or increased utility such as more powder) then the effectiveness of an avalanche airbag can be quickly nullified. The study does not address any increase in risky behavior as a result of using an avalanche airbag pack. I plan to write a future blog on this subject.