barnaby, I think the issue is the percentages... here are some stats I found:
-----------------
I then checked out the US Product Safety Commission for announcements & recalls. Their Safety Alert confirms the AAP findings and adds some numbers:
The CPSC estimates that in 2001 there were 91,870 hospital emergency room-treated injuries associated with trampolines
About 93 percent of the victims were under 15 years of age, and 11 percent were under 5 years of age
Since 1990, CPSC has received reports of 6 deaths of children under age 15 involving trampolines (study date not specified, likely from 1999 or 2000)
Another CPSC study states 11 deaths from 1990-1999
Compared to biking, the death rates sound pretty good. In 2006, 77 people died from bicycle accidents (Im ignoring the 90% who died in collisions with cars the total number dead in 2006 is 770). The difference is an estimated 3 million trampolines in use vs. 85 million bike riders (and about 540,000 total annual bike injuries).
Additionally, children are actively taught to be safe on a bike by learning about traffic and common sense safety, with the majority of serious head injuries avoidable by the use of a relatively cheap helmet. Some 45-88% of bicycle brain injuries can be prevented with a helmet, and similar increases in safety for trampolines *might* convince me that its worth the risk (more on risk in a bit).
source:
http://www.tinygrass.com/2008/05/trampolines-safety-liability/