Project Overview Key Study Findings and Recommendations

Project Overview

The current study was carried out for the Federal Highway Administration (FHWA) to develop information on the range of events causing injury to pedestrians and bicyclists. Traditional information sources such as FARS and State motor vehicle crash files primarily capture information on pedestrian and bicyclist injury events that: (1) involve a motor vehicle and (2) occur on a public roadway. Many more pedestrians and bicyclists, however, are injured as the result of falls that do not involve a motor vehicle and in locations (sidewalks, trails, parking lots, etc.) other than the roadway, but over which FHWA and local transportation departments may have some jurisdiction. While seldom reported by law enforcement officers, these cases frequently result in hospital emergency department visits. Thus, the current study used emergency department data to provide more complete information on these other types of events.

Eight hospitals in three States (California, New York, and North Carolina) participated in the study. Each collected information on injured pedestrians and bicyclists treated in their emergency department over approximately a 1-year time period, using a special two-page survey form. The form placed particular emphasis on the location of the injury event and whether or not a motor vehicle was involved. A total of 2,802 cases were reported, 2,558 of which met the study criteria for a bicyclist or pedestrian. In addition to the emergency department data, hospital discharge and motor vehicle crash data were obtained from each of the three participating States and examined in conjunction with the emergency department data.

The results presented in this report were primarily descriptive tabulations that addressed the following research questions:

• What are the frequency and characteristics of bicycle injury events that occur in non-roadway locations and/or those that do not involve a motor vehicle, and how do they differ from bicycle-motor vehicle crashes that occur on the roadway? (chapter 3)
• What are the frequency and characteristics of pedestrian injury events that occur in non-roadway locations and/or those that do not involve a motor vehicle, and how do they differ from pedestrian-motor vehicle crashes that occur on the roadway? (chapter 4)


• What role does alcohol play in each of these events? (chapter 5)
• What are the estimated frequencies of motor vehicle and non-motor vehicle, and roadway and non-roadway events causing injury to pedestrians and bicyclists? (chapter 6)

KEY STUDY FINDINGS AND RECOMMENDATIONS

Chapters 3, 4, and 5 each contain summaries of key findings that will not be repeated here. Each of the chapters expands on the injury matrix introduced in chapter 1. This matrix (see figure 1) categorized events according to their place of occurrence (roadway or non-roadway) and whether or not a motor vehicle was involved. For bicyclists, the distribution of cases based on the obtained hospital emergency department data is shown in table 64. These results

Table 64. Distribution of bicyclist injury cases by place of occurrence and motor vehicle involvement status.

suggest that statewide crash files, which are limited primarily to events that involve a motor vehicle and those that occur on the pubic roadway, are likely to capture less than a third of bicyclist injury cases serious enough to require emergency department treatment. In reality, they capture far fewer, since the results reported in chapter 6 showed that anywhere from 40 to 60 percent of the bicycle-motor vehicle cases were not reported in official State files.

Table 65 shows the distribution of pedestrian injury cases identified by the participating hospital emergency departments. These results are similar to those for bicyclists, but with a smaller percentage of the pedestrian-only or "fall" events (i.e., non-motor vehicle) occurring in the roadway, and a correspondingly larger percentage occurring on sidewalks, in parking lots, and at other non-roadway locations. Again, less than a third of the injury events serious enough to require emergency department treatment involved a motor vehicle traveling on the roadway. Furthermore, the results of chapter 6 showed that 35 to 55 percent of these events may go unreported.

Table 65. Distribution of pedestrian injury cases by place of occurrence and motor vehicle involvement status.

Tables 64 and 65 are based on the total sample of reported emergency department cases. However, even among the subset of those injured seriously enough to require hospitalization, non-motor vehicle and non-roadway events continued to play a prominent role. Sixteen percent of hospitalized bicyclists were injured in non-roadway locations and 42 percent were injured in events that did not involve a motor vehicle. For hospitalized pedestrians, the corresponding percentages were 26 percent non-roadway, 24 percent non-motor vehicle.

These findings lend strong support to previous research (summarized in chapter 2) carried out in this country, as well as in Australia, New Zealand, and a number of European countries, showing that reliance on official road accident statistics greatly underestimates the number of injured bicyclists and pedestrians. In light of the U.S. goal of increasing levels of bicycling and walking, they also reinforce the need for continued and strengthened efforts toward creating a safer environment for these non-motorized transportation modes. These efforts need to move beyond the roadway and beyond thinking about bicyclists and pedestrians only as they interact with motor vehicles. Sidewalks and trails need to be viewed as important transportation facilities in their own right, parking lots need to be built with pedestrians and bicyclists in mind, and all facilities accommodating non-motorized transportation need to be well designed and well maintained.

In addition to documenting the role of non-motor vehicle and non-roadway events resulting in injury to pedestrians and bicyclists, the descriptive tables and figures contained in this report provided insight into the nature of these events and potential countermeasures that could improve pedestrian and bicyclist safety. Although the conclusions that can be drawn from a descriptive analysis are limited, the following areas stand out and may warrant further investigation:

• Alcohol was a significant contributing factor in both pedestrian and bicyclist injury events, especially those involving a motor vehicle. Nearly a third of pedestrian-motor vehicle crash victims ages 25-44, and 22 percent of those ages 45-64, were reported to have been drinking. For bicyclists, 15 percent of adults age 20 and above had been drinking. Alcohol use was more prevalent among males and during the late evening and early morning hours. It was also associated with a higher rate of hospitalization. Future public education efforts should target the dangers of drinking and walking and drinking and bicycling as well as drinking and driving.
• Sidewalks were a frequent site of bicyclist and pedestrian injury events not involving motor vehicles. Although the vast majority of both bicycle-motor vehicle and pedestrian-motor vehicle collisions occurred on the roadway, 43 percent of the bicycle-only events occurred off the roadway, and half of these were on sidewalks. For pedestrians, 78 percent of non-collision events occurred off the roadway and over half (58 percent) of these were on sidewalks. Young children were particularly overrepresented in sidewalk injury events, as were senior pedestrians age 65+.
• Equipment such as in-line skates and skateboards can make being a pedestrian particularly hazardous. Of the nearly 200 pedestrian-only events that occurred in the roadway, one-fourth involved the use of in-line skates and an additional 5 percent involved persons on skateboards. Off-road, the situation was not much safer: 12 percent of the injured pedestrians were using in-line skates and 3 percent were using skateboards. These percentages are combined for all ages and would be higher for teens or young adults. While education efforts might help to alert young people to the dangers of these activities, a better alternative might be to provide a safer environment for skating, such as a network of well-maintained off-road trails.
• In climates where winter months are accompanied by snow and ice, off-road locations such as sidewalks, parking lots, and driveways can be especially hazardous for pedestrians. In Buffalo, New York, which experienced considerable snow and ice during the winter of 1995-96, over a fourth of all pedestrian injuries reported by the hospitals during the entire year of data collection were icy weather related. The vast majority of these injury events occurred off the roadway and did not involve a motor vehicle. In addition to clearing roadways and making them safe for motor vehicle travel, sidewalks, driveways, and parking lots need to be made as safe as possible for pedestrian travel. Too often, roadways are cleared at the expense of sidewalks, and little, if anything, is done to help pedestrians negotiate parking lots once they arrive at their destinations. Middle-aged adults were more susceptible to icy weather-related falls than were either young persons or older adults, perhaps due to their greater exposure.
• For this sample of injured bicyclists, reported helmet use was highest among children under 15 years of age, and among those injured in bicycle-only events occurring off the roadway. However, emergency department personnel were unable to determine helmet use in about 20 percent of the cases, and may have relied on self-reports or observations of head/face injuries in many instances. Assessing helmet use is clearly a difficult task in an emergency department setting. And in the absence of any exposure or control data, it is not an appropriate data source for evaluating the effectiveness of helmets in preventing or lessening the severity of head injuries.
• An incidental but intriguing result was the discovery that 7 of the 91 pedestrian-motor vehicle cases identified in the two North Carolina emergency department files were matched to bicycle-motor vehicle crashes in the State crash file. This suggests that emergency department personnel may not always be aware that the individual they are treating was riding a bicycle at the time he or she was struck. As a result, hospital-based sources may overestimate the number of pedestrian-motor vehicle cases, while underestimating the number of bicyclist-motor vehicle cases. Further research (e.g., a follow-up telephone survey of individuals identified in hospital or emergency department records as pedestrians) could help to clarify the issue.

Two final comments are in order. The current study was not very successful in producing firm projections of the overall numbers of injured pedestrians and bicyclists. Possible reasons for this are outlined in the discussion section of chapter 6, and include the limited sampling of hospitals within the selected States and the difficulty in defining and capturing information on pedestrian-only events. However, the difficulties experienced in this study also underscore the desirability of establishing routine linkages between hospital and motor vehicle crash databases, and for requiring E-coding of hospital inpatient as well as outpatient (emergency department) cases. Linked together, police and hospital databases can provide more complete information on a much broader range of pedestrian and bicyclist injury events. Efforts such as CODES (Crash Outcome Data Evaluation System) initiated by the National Highway Traffic Safety Administration are making this possibility a reality in a number of States.

The current study was also limited in that no exposure data were collected for analysis in conjunction with the pedestrian and bicyclist injury data. Relevant exposure data has been a long-standing need in the area of non-motorized safety research. In the absence of such data, it is not possible to draw definitive conclusions regarding the level of risk associated with specific locations, behaviors, etc. The safety of riding a bicycle on the sidewalk, for example, or walking in a parking lot, is best assessed if information is available on the total numbers of individuals bicycling on sidewalks or walking in parking lots, i.e., those who are uninjured as well as injured. One of the few studies to collect such information was carried out by the Consumer Products Safety Commission (Rodgers, 1993). Information on injured bicyclists was collected from cases reported to a national sample of hospital emergency departments, and exposure data from a national telephone survey. Combining the two sources of information, the Commission was able to conclude, for example, that the risk of injury for children riding bicycles in the street was about eight times greater than riding on bicycle paths, and nearly two times greater than riding on sidewalks.
What this study has provided is information on the numbers of bicyclists and pedestrians being injured, the types and locations of events that cause these injuries, the nature of the injuries, and some details on the characteristics of these events and the persons injured. In doing so, it has broadened the usual definition of pedestrian and bicyclist, at least from the traditional highway safety perspective, and has encouraged a more comprehensive approach to creating a safer environment and promoting greater use of these non-motorized transportation modes.