Alan M. Smith, MPH, Leslie U. Ray, MA, MPPA, Janace Pierce, MS, Clint W. Garrison, BA, Margaret Lutz, BA, Patricia M. Akers, MA, Barbara M. Stepanski, MPH, Brenda J. Dunn, Patricia A. Murrin, RN, MPH
The Safe Communities program was implemented in San Diego County in 1997 to foster local action to prevent motor vehicle deaths and injuries. This paper presents baseline regional data for each of the five regions on location of crashes by type, victim information and crash specifics. This data has been used to help decide where to target specific interventions within each coalition and has been compared with subsequent years to evaluate the effects of these interventions on the incidence of motor vehicle-related injuries.
From July 1, 1996 through June 30, 1997, there were 14,321 motor vehicle-related injury crashes in San Diego County. These crashes were responsible for injury or death to 21,549 people. Injuries are the leading cause of death and disability for people between the ages of one and 44 years, and motor vehicles are implicated in more injuries than any other cause.
In 1997, a program known as "Safe Communities" was implemented in the County of San Diego to address the problem of injuries resulting from motor vehicle crashes. A project of the National Highway Traffic Safety Administration (NHTSA), Safe Communities program funds and resources are funneled toward the development of local coalitions comprised of concerned citizens' groups, businesses, and civic leaders. There are two Safe Communities programs in the county, one focusing on the city of San Diego, while the other encompasses the rest of the county, divided into four separate regions.
Safe Communities focuses on data in determining priorities and strategies, and emphasizes evaluation and sharing of results with other coalitions. The County of San Diego Health and Human Services Agency, Emergency Medical Services Division (EMS) provides data to the Safe Communities programs for this purpose. In its role as the primary injury surveillance entity in the county, EMS collects and maintains countywide prehospital, trauma, and medical examiner's databases as well as a number of other specialized databases. The Statewide Integrated Traffic Records System (SWITRS), comprised of data on motor vehicle injury crashes to which law enforcement responded, is one of the primary sources of information on motor vehicle crashes. This database contains information on the level of the crash, the party (e.g., the automobile), and the individual victim. Crash level information includes exact location of the crash, type of collision, number of victims, which party was at fault, and other factors that contributed to the collision. Party level information includes information that relates to each vehicle involved in the crash, such as vehicle make and model, number of passengers, and number of deaths or injuries in each party. Victim level information addresses demographics, location within vehicle, restraint use, extent of injuries, and other data that applies to each person who was injured in the crash. EMS staff geocode each crash according to exact street location, and query these records as to the distribution of crashes according to any other information present in the database. In the following examples, we discuss how these mapped data were used to help the Safe Communities program focus its interventions most effectively.
Maps were created using ArcView 3.1. For choropleth maps showing rates of injury for distinct areas of the county, a dBase file was created in which the rate for each area was assigned to the corresponding code. This database file was then incorporated into the ArcView program and linked by area type (e.g., zipcode or subregional area) to the shapefile (layer) for that area type using the "join" command. Rates were grouped into categories representing equal value intervals, and the category for each area was plotted with a representative color scheme.
SWITRS data, which contained location data for each crash, was received from the California Highway Patrol as a fixed-format ASCII file. This file was then converted, using SPSS for Windows, into dBase format so that it could be read into ArcView. Dot maps were produced using the SWITRS data. Approximately 80% of incidents were able to be automatically geocoded using a data matching program developed specifically for this purpose. The remaining 20% were geocoded by hand using narrative information.
Pedestrian crashes were the most lethal type of motor vehicle-related collisions, with 5% resulting in death and 14% suffering severe injury, much higher than the 1% death rate and 4% severe injury for all other motor vehicle crashes during FY 1996/97. The coalition for the City of San Diego was particularly interested in addressing this problem because half of all pedestrian crashes in the county occurred within the San Diego city limits.
A choropleth map of pedestrian injury rates by county subregional area showed that two of the most densely populated areas also had very high crash rates. A dot map of these areas showed two specific regions of concern. Cluster A stretched along about four miles of one particularly busy thoroughfare in an older section of town, and cluster B was centered in the downtown area known as the Gaslamp district. Detailed incident maps were developed to show victim age, pedestrian action, and type of violation.
|
|
|
|
|
|
|
|
Cluster A (University Ave.) occurred where a high level of business and residential traffic came together. The victims at one end of this section were predominantly older than 55 years of age, while those at the opposite end were mostly younger than 15. The maps showed that the crashes were due primarily to pedestrian violations such as crossing not in a crosswalk, or because of failure by the driver to yield to pedestrians in a crosswalk. After examining detailed incident maps that included victim age, type of violation, and pedestrian action, the workgroup focusing on this area developed some specific intervention plans. These plans involved a combination of pedestrian education targeted at the senior population, and environmental engineering measures known as "traffic calming."
Cluster B (Gaslamp District) occurred in an area of downtown San Diego known for its bars, restaurants, and dance clubs. Visitors to downtown generally park their cars in one location and walk from place to place. Because of the number of bars in this area, one strong assumption was that the crashes were due to pedestrians walking erratically while under the influence of alcohol. An inspection of incidents plotted by violation category revealed that these crashes were caused more often by pedestrians crossing the street illegally, or by motorists failing to yield to pedestrians when turning right. This changed the group's focus in this area from alcohol to pedestrian awareness activities.
San Diego boasts what has been referred to as "the most scenic stretch of highway in the United States," in State Route 163 where it runs through Balboa Park. Along this stretch, a grassy area divides the north- and southbound lanes with large eucalyptus and oak trees along the center. At the origin of this freeway, traffic merges from three different sources, and the number of lanes quickly condenses from four to two lanes. There were high numbers of vehicle crashes in this area resulting in severe injury or death. A local trauma surgeon and his staff were very concerned and asked for more information. This came to the attention of the City Councilwoman for this district, who held hearings in order to determine if this stretch of roadway was in fact more dangerous.
A traditional dot map of crashes occurring on the freeway was not very informative because it was very difficult to discern the number of crashes along this stretch relative to other areas. By dividing the map of the highway into standard lengths and counting the crashes in each segment, the number of crashes in each segment could be represented by varying road widths. This variable road width was found to convey a better sense of relative risk of crashing in this area.
GIS has proven an essential tool for the targeted planning of interventions to prevent motor vehicle injuries. As with other analytical tools, our experience has shown that the way in which the data is presented can dramatically affect the interpretation. This in turn determines whether the most effective intervention will be used.
Alan M. Smith, MPH
Epidemiologist
County of San Diego
Health and Human Services Agency
Emergency Medical Services
6255 Mission Gorge Road
San Diego, CA 92120-3599
(619) 285-6429
asmit1he@co.san-diego.ca.us