Sounds about right, although it's disputed on the basis of the small sample size:

http://news.nationalgeographic.com/n...death-science/