Back in March and April of this year we asked Boston Emergency Medical Services (EMS) and the Boston Police Department (BPD) to separate bike and pedestrian crash data to allow us to see where bike crashes are happening. And to their credit, each department quickly recognized the need and worked with us to meet it, partly by adding bicycle checkboxes to their incident report forms.

Now, not only do we have over six months of data to look at (May thru October 2010), but it’s all plugged into an interactive Google map created by Union member and volunteer Ben Sheldon. Thanks Ben! It will still be a year or two before we can have a large enough sample size to discern reoccurring patterns at particular intersections, but this first six-month glimpse of the data does give a sense of where crashes are happening in the city and a few other interesting things.

At a glance, where crashes happen seems to correlate more with topography and population density than the perceived “bikeyness” of the neighborhood. So Roxbury, Dorchester and Jamaica Plain have similar numbers and clusterings of crashes, but there are few crashes in the hilly parts of Highland Park (Fort Hill), Mission Hill, or the hills above Grove Hall in Roxbury. The biggest clusters of crashes by far happen in the flat and dense areas such as the South End, Back Bay, Downtown, and along Commonwealth Avenue, and the lowest numbers are in more suburban areas like Hyde Park and West Roxbury.

Helmet use

EMS has also been pretty good about recording helmet use, doing so in 76 percent of the crash reports between May and October. Around 34 percent of cyclists picked up by ambulances were reportedly wearing helmets, and that figure could be as high as 44 percent if we make some guesses about the “unknown” crashes where helmet use was not recorded.

It’s difficult to compare this 34 to 44 percent helmet use to elsewhere in the U.S., as most of the studies readily available are observational, not based on crash data, and the difference between these two collecting methods is likely vast. But that said, one count of riders in Portland, Oregon in 2009 indicated a helmet use of 76 percent.

Given that not wearing a helmet can exponentially increase a rider’s chance of dying in a crash, doing what we can to increase helmet use is important no matter what the rate is.

A few caveats

Before you start exploring the Union’s new map, you’ll also need to understand a few caveats about the data it contains.

Firstly, the data that pops up when you click on any marker is not to be taken as gospel. It’s simply not always accurate. The “Call Type,” for instance, is dependent on what the person calling 911 reports. So a person may call in to report a cyclist hit by a car, but when EMTs arrive on the scene they discover the person really just fell off their bike. That “Call Type” would still be reported as MVPED, even though a Motor Vehicle was not involved in the actual crash.

Also, all vehicle speeds and impact types are generally estimated by the victim, and are thus simply educated guesses from people that may actually be experiencing shock.

On the map itself you might also see one or two accidents in a strange place. This is often due to an occasional communication problem between the EMT’s reported location of the incident, and our method of geocoding. “WB Storrow Drive,” without a reported cross street, for instance, is difficult for us to place accurately on a map.

The Boston Cyclists Union has more work planned on this effort yet, including more analysis of the EMS data as well as matching it with data from the Boston Police and the state’s Department of Public Health. If you are interested in volunteering to help these crash reporting efforts, please email pete@bostoncyclistsunion.org.