this is from April 2010, so relatively recent. A couple of observations: August/September is back-to-school for the universities. It looks like the crash rate in November-February is out of proportion to the number of cyclists, however the winter months have a lot of darkness which might be a factor. 20% of crashes are doorings. I don’t see any data on solitary cyclist crashes (undoubtedly large numbers but comparatively minor injuries?) or bike-bike crashes (likely small numbers)
The Cambridge trends correspond with international research demonstrating that as cycling participation increases, a cyclist is less likely to collide with a motor vehicle or suffer injury and death. Policies that increase bicycling appear to be an effective route to improving the safety of people bicycling.