This disaster is why I wouldn’t ever work at a hospital. I get sufficiently stressed worrying about my company losing money. Losing lives — no. I respect hospital sysadmins immensely.
It’s not a very technical article, but still interesting. The point about the importance of the network in disaster recovery planning is essential. slashdot claims the problem was due to a spanning tree fault, although this more technical article seems to indicate a congestion problem. Hard to say.