In the fall of 2005, staff from DOHMH's Division of Epidemiology, DOT's Bicycle and Pedestrian Safety Units, and NYPD's Traffic Control Division established a multiagency team to develop the report. The team chose to focus on bicyclist fatalities and serious injuries (defined as an injury that requires the bicyclist to be taken to the hospital), based on completeness of existing data sources. Analyses were conducted on bicyclist deaths that occurred in NYC from 1996 through 2005. Because of a longer lag time in receiving data on serious injuries, these analyses were conducted through 2003 only.
Reconciliation of multiple data sources
The primary data source for bicyclist fatalities was the DOT Fatality Database, which is reconciled on a monthly basis with NYPD investigation reports. Fatalities identified from this source were cross-referenced with death certificates maintained by the Office of Vital Statistics at DOHMH to confirm the cause of death and identify any additional bicyclist fatalities. This cross match identified an additional 13 fatalities not found in the DOT Fatality Database. For these deaths, cause of death was verified using medical examiner records.
Development of the analytic plan
In addition to person, place, time, and circumstance factors for each crash, the multiagency team examined motor vehicle type (small vs. large), contributing crash factors as defined by crash scene investigators, helmet use, presence of bicycle lanes, and “dooring” (accidents caused by a bicyclist hitting an open motor vehicle door or trying to avoid one). Spatial analyses were also incorporated into the investigation, and medical examiner files were reviewed to ascertain bicyclist characteristics such as helmet use.
Analysis and interpretation
Throughout the fall of 2005 and spring of 2006, the multiagency team reviewed findings and refined its analysis. Additional data sources were examined to understand bicycling trends in NYC and provide national comparison data. Because of the interdisciplinary composition of the team, analysis and interpretation required thorough discussion of data sources and variables.
The study identified 225 bicyclist deaths in NYC from 1996 through 2005 and 3,462 serious injuries between 1996 and 2003. The annual death rate remained steady at 2.8 deaths per million population, comparable to the national rate of 2.7 deaths per million, despite much higher rates of bicycling. The majority (91%) of bicyclists who died were male, and among men, the highest death rate (8.3 per million) was in the 45–54 age group. Among children aged 5–14, the death rate was five times higher for boys than for girls.
Most fatalities (92%) involved contact with a moving motor vehicle. Almost all fatal crashes (89%) occurred at intersections, and more than half (53%) occurred on large, arterial roads, even though such roads make up only 10% of NYC's road system. Large vehicles, including trucks and buses, were involved in almost one-third (32%) of fatal crashes, although such vehicles comprise only 5% to 17% of vehicles on city roadways. Nearly all fatalities (94%) involved poor driving or bicycle riding practices, particularly driver inattention and disregard of traffic signals and signs. Among those for whom helmet use was known (59%), almost all bicyclists who died (97%) were not wearing a helmet. Nearly three-quarters (74%) of fatal crashes involved a head injury. Only one fatal crash with a moving motor vehicle occurred when a bicyclist was in a marked bike lane.
Spatial analyses identified three locations in NYC where three or more fatalities occurred in very close proximity (within 1,000 feet of one another) and an additional six locations where three or more fatalities occurred in close proximity (within a quarter mile of one another). Several locations were also identified where multiple serious injuries occurred. DOT investigators immediately carried out safety evaluations of all these locations.
Serious injuries declined by 46% over the study period, against a backdrop of increased bicycling, as indicated by annual counts and population-based surveys. Thirty percent of crashes resulting in serious injuries occurred mid-block, compared with 11% for fatal crashes. While large vehicles were involved in nearly a third of fatal crashes (32%), they accounted for only 7% of serious injury crashes. Documentation of helmet use for serious injuries was even more incomplete than for fatal crashes (32%); nonetheless, among those with documented information, helmet-wearing rates were more than four times higher than rates for fatal crashes (13% vs. 3%).
Development of recommendations and action steps
As part of the report-writing process, team members worked with the leadership of their respective agencies to develop action steps to simultaneously promote bicycling and increase bicyclist safety. The team identified 38 action steps and organized them into five key areas: (1) bicycle infrastructure, (2) motorist and bicyclist awareness, (3) investigation and enforcement, (4) legislation, and (5) improvement of data collection, analysis, and reporting. Among these action steps were commitments to complete 240 miles of bicycle lanes over four years, conduct a citywide public awareness campaign on roadway safety, and increase data collection efforts to better understand bicycle ridership. While a small number of these initiatives were already under consideration, report findings created a sense of momentum among agency leaders and galvanized their commitment to the initiatives. The team also developed general recommendations for motorists, bicyclists, health-care providers, and community-based organizations toward improving bicyclist safety in NYC.