WASHINGTON D.C. – The United States Helicopter Safety Team has released new statistics on accident rates for U.S. civil helicopters along with new data on fatal accident rates and fatalities.
For the first seven months of 2014, there have been 3.94 helicopter accidents for every 100,000 flight hours. This is a 51% decrease compared to the 2001-05 baseline determined by the USHST parent organization, the International Helicopter Safety Team. This also is a 22% reduction compared to the 5.06 rate posted in 2009 and a 12% reduction compared to the 4.46 rate posted in 2012. (See charts below for complete data.)
These results are especially encouraging because they include data from the month of July which historically tallies the most helicopter accidents in the United States during the year, according to officials..
In addition, the rate of fatal accidents is down so far in 2014 compared to 2013, which showed a troubling upward spike. During the first seven months of 2014, there were 0.53 fatal helicopter accidents per 100,000 flight hours compared to a rate of 0.86 for all of 2013.
The decrease also is reflected in the fatality rate. There were 1.79 fatalities per 100,000 flight hours in 2013 compared to 0.91 fatalities per 100,000 flight hours so far in 2014.
Safety experts at the USHST said they know that accident data from the rest of the summer has not been gathered yet and there are four months still left in the year, but the results at this time are promising, noting a stronger safety culture seems to be growing in the civil helicopter community.
The USHST and its parent organization, the IHST, promote safety and work to reduce accidents. IHST was formed in 2005 to address factors that were affecting the helicopter accident rate. Prior to 2006, the number of worldwide civil helicopter accidents was rising at a rate of 2.5% per year. Since 2006, the number of accidents worldwide has been decreasing by an annual rate of 2%, officials noted.
This announcement is statistically suspect. The claims of double digit improvement on a percentage basis are essentially unproven based on what’s presented here. Given the small number of helicopter accidents per year (or any other handy unit of time, such as flight hours), a small change in number could yield a large percentage change yet still be within the range of random variation, meaning that the apparent “change” is no change at all as far as safety is concerned. The significance of these numbers can be verified by subjecting the observed numbers to tests of significance that most high school math students in the US learn about today. A simple chi-squared or t-test of the data should do it. Fleet size would be a parameter, I think, but IHST ought to be able to get that number easily. Until an appropriate test of significance is run and a confidence level reported, I’m not sure there’s enough here to go on.