Evaluation of Bicycle Injury and it's Associated Hospital Costs in San Francisco

Principal Investigator: Rochelle Dicker, MD

Co-Investigators: Wendy Max, PhD, Dahianna Lopez, RN, MSN, MPH

bike crash sffdBrief Summary of Project: Police reports are the industry standard for assessing transportation-related collisions and informing policies and interventions that address the issue. Previous studies have suggested that police reports miss a substantial portion of bicycle crashes not involving motor vehicles. Trauma centers may be the only source for surveillance and cost data on "cyclist only" injuries. No study to date has explored the health and economic impact of cyclist-only injuries. Our objective was to report medical cost of injuries from our trauma center databases to make an economic case for prevention. We aimed to compare the incidence, injury severity, admission rate, and cost for cyclist-only and bicycle-versus-auto (BVA) injuries. We hypothesized that cyclist-only injuries carry an equal or greater burden of injury and cost when compared to BVA injuries. We conducted a retrospective cohort analysis of hospital records for 3,049 patients treated for bicycle-related injuries at our city's only level I trauma center. We compared injury type by injury severity, admission status, and cost. We calculated cost using cost-to-charge ratios, ambulance, and professional fees, inflated to 2009 dollars. We analyzed data using parametric and non-parametric statistical tests.

Milestones: By September of 2010, all data related to the cost calculations-namely, hospital charges, professional fees, and ambulance charges-were collected for the years 2000 to 2009 (n=3,049). We also collected police data for the same years in order to later do probabilistic linking between datasets. We collected 10 years of data in order to increase statistical power and assess cost and injury trends over time.

By November of 2010, we conducted an exhaustive bicycle injury literature review and refined our hypotheses. We also ran preliminary calculations on the data and submitted an abstract to the Society for the Advancement of Violence and Injury Research / CDC National Conference. The abstract was accepted for presentation in February of 2011.

Results: Our major findings include the following: 1) Police records underestimate trauma-reported bicycle injury by 27% and they are significantly more sensitive to bicycling crashes that involve a vehicle (vs. bicycle falls); 2) Bicycling injuries in SF affect primarily adults; 3) Proportion of bicycling injuries have doubled over one decade; 4) Bicycle falls have a higher proportion of head, face, extremity, and skin injuries; 5) 70% of injured bicyclists did not wear a helmet; 6) The total cost of bicycle injury in SF over a decade was $35 million.

Significance: Cyclist-only injuries were admitted at 2.5 times the rate of BVA injuries and accounted for half of all bicycle-related admissions, revealing an under recognized bicycle injury pattern. Costs for care were similar. Without police data to elucidate the circumstances of the cyclist-only crashes, these injuries would continue to be underreported and underappreciated. Trauma centers can play a key role in future collaborations to define issues and develop prevention strategies for cyclist-only crashes. In line with the public health model, one recommendation that could stem from this project is to design a surveillance system for transportation-related injuries that takes into account self-report, police and hospital data, and geographic information.