Describing Trauma Care Resources with GIS

Trauma, or injury, is the leading cause of death for Americans under age 40 and the third leading cause of death overall after cardiovascular disease and cancer. The burden of injury often surpasses acute clinical care to involve a lifetime of recovery and rehabilitation, especially in head and spinal cord injuries. Generally, injury is damage to the body caused by exchanges with environmental energy that are beyond the body’s resilience, resulting in fractures of bones, tearing of skin, muscle, and organs, and the often life-threatening bleeding that follows. Mechanical energy, released in motor vehicle crashes, falls, and being struck by objects (including bullets) accounts for most fatal and severe trauma. Thermal, chemical, and electrical energy sources also cause injury. Injury epidemiology attempts to study the distribution and characteristics of injuries among a population in space and time, intending to identify causes and risk factors to effectively reduce injury incidence and severity.

Persons with severe injuries require access to a comprehensive trauma care system in which a continuum of multidisciplinary resources provide appropriate, capable, and coordinated medical care within the "Golden Hour" immediately after injury occurs. At the core of this continuum are two essential components: 1) prehospital emergency medical services (EMS) including 911 emergency telephone access and rescue squads and ambulance services whose firefighters, Emergency Medical Technicians, and paramedics respond to the injury scene to expedite treatment and ambulance transport; and, 2) acute care hospitals (particularly trauma centers) where emergency physicians stabilize and surgeons repair damaged body structures and where multidisciplinary teams in critical care units manage complex, multi-organ system injuries. Other trauma care system components, described by the American College of Surgeons (ACS) and others, include enabling legislation, lead agency oversight, injury prevention programs, rehabilitation services, and information systems with data for system evaluation and performance improvement, injury surveillance, and research.

Trauma occurs in spatial distributions shaped largely by where, when, and how people live, work, travel, and recreate. Geography, demographics, economics, politics, and a variety of other influences shape the distribution of the components of a trauma care system and may significantly effect a severely injured person's access to its services. This paper presents the use of ArcView GIS (Esri, Inc., Redlands, California) to integrate and visualize data on multidisciplinary trauma care resources and on fatal and severe motor vehicle crashes and to explore their interrelationships using common GIS spatial analysis methods.

Obtaining Spatial Data Files

To map the distribution of various trauma care resources and motor vehicle crashes, three types of spatial data were needed: named point locations of Minnesota cities; center-lines for state roadways; and named areas for Minnesota state, counties, cities, and townships. The roadway and area spatial data files were obtained from the Minnesota Department of Planning's CD titled Basemap ’99 (1999). The city points were obtained from the Department's Land Management Information Center (1999). These files were readily available at no cost and provided in .dbf format.

Obtaining Attribute Data on Trauma Care Resources

Attribute data describing important characteristics of the trauma care resources were obtained from several administrative databases in Minnesota (Table 1). Each data file included a location identifier field, for example a city or township name, to enable geocoding against the Basemap 99 spatial data files. These files were readily available at no cost and provided in .dbf or convertible format.

Table 1. Description, sources, and data elements of selected trauma system components and other project-related data.

Obtaining Data on Fatal and Serious Motor Vehicle Crashes

When a motor vehicle crash is reported to a law enforcement agency in Minnesota, the responding officer completes a standardized motor vehicle accident report form. The report is submitted to the Minnesota Department of Public Safety (DPS) where the data is abstracted and entered into a database organized into three files, each containing an accident number as a common field to enable linking of the files. The crash file contains data about the crash event, including the date, time, type of crash, and roadway location. The vehicle file holds data on each vehicle involved in the crash. The person file contains data on each person involved in the crash, including their age, vehicle entrapment or ejection, injury severity, and transport to a hospital. The files contain other trauma-related data not obtained for this project. Together, these files form a hierarchical relationship of a person in a vehicle involved in a crash event. Tables 2 and 3 describe the data obtained for this project from the DPS crash and person files respectively.

Table 2. Data in the Motor Vehicle Crash File

Table 3. Data in the Motor Vehicle Person File.

Only crashes with injuries coded as "K" or "A" for September and October 1999 were obtained for this project. The crash and person data files were received in .dbf format and imported into Microsoft Access (Microsoft Corporation, Redmond, Washington) where the two files were joined on the common field, Accident Number. The crash location indicators (Route Identification and Beginning Point) were retained in each record to enable geocoding of crashes against the roadway files in Basemap ’99.

Questions of Interest

Three ACS recommended guidelines were used to identify persons in fatal and severe crashes who may need care in a Level I or II trauma center. Ambulance personnel may use these guidelines together with assessment of anatomic injury and physiologic status to triage the person to determine transport to the most appropriate receiving hospital. Queries were developed in Microsoft Access to select persons meeting one or more of these characteristics:

1. A survivor (severity coded "A") in the same vehicle as a fatality (severity coded "K")
1. A person ejected from the vehicle during the crash (ejected coded "03")
1. Person age less than 5 years or greater than 55 years.

In addition, queries were developed to identify several characteristics of fatal and severe crashes and the persons involved that were of interest from trauma system perspective:

1. Persons who died of their injuries after ambulance transport to a hospital. This identified severely injured trauma patients who died after unknown hospital treatment. Review of hospital medical records would be needed to explore this indicator further. To identify these persons, queries were developed to identify persons with severity coded "K" (fatality) and hospital coded as "Y" (transported to a hospital).
2. Persons trapped in the crashed vehicle that required extrication from the wreckage. This identified crashes that required rescue personnel with special crash rescue training and equipment. To identify these persons, queries were developed to identify persons with ejected coded "01" (trapped and requiring extrication).
3. Fatal and severe crashes that occurred outside the ALS ambulance service areas in cities and townships served by BLS ambulance services. This may be an indicator of the need for "intercept" agreements in which a BLS ambulance meets an ALS ambulance during transport of a severely injured person to the hospital to transfer patient care to the higher trained ALS personnel.
4. Fatal and severe crashes that occur within a specified distance (15 and 30 miles in this project) from a Level I or Level II trauma center. Since survival and optimal recovery from severe trauma is a time-sensitive, distance from trauma center care is measured in time from injury to definitive care (surgery). Greater distances may require use of helicopter air ambulances rather than traditional ground ambulance transport. This may also indicate need for interhospital transfer agreements between non-trauma centers and trauma centers.

Four steps were necessary to visualize the spatial distribution and characteristics of trauma care resources and motor vehicle crashes using ArcView GIS. First, the trauma care resources data files were prepared in .dbf format and queried in a relational database to identify crashes with characteristics of interest. Next, the trauma care resources data, the selected crash data, and the appropriate spatial data were added as themes to a view. Third, the data tables of both resources and crash attribute themes were geocoded on the common location identifier field in the spatial themes as either a city or roadway crash location point, or as a city or township area. Finally, the resulting maps were modified with the Legend Editor and Color Palette for best visualization of the characteristics of interest and prepared as a layout for printing.

Visualizing and Exploring the Relationship of Trauma Care Resources and Motor Vehicle Crashes

Of 594 "K" and "A" crashes obtained from the Department of Public Safety, 161 (29%) could be readily geocoded against the Basemap '99 roadway files and were used as the sample of crashes in this project. These crashes occurred exclusively on interstate and Minnesota trunk highways. The remainder of the crashes occurred on other roadways including state-aid highways, county roads, township roads, city streets, and other locations. These crashes would have required manual geocoding and were excluded from this demonstration project.

Themes from the trauma resources and motor vehicle crashes were combined using the spatial analysis capabilities regarding proximity and containment to explore the relationships among the sample of fatal and severe motor crashes and trauma care resources. A series of maps were created of the trauma care resources and motor vehicle crashes to address each of the questions of interest regarding the relationship of crashes to trauma care resources. Two questions and their spatial analysis methods are presented here as examples. Both are questions of spatial relationship called containment, in which features (e.g. points) are inside other features, typically areas. The first question asks which crashes occurred inside or outside the paramedic ambulance service areas. To answer this question, the Select by Theme option was used to select crashes (from the Crash theme) that are within the paramedic service area (from the Ambulance Area theme). A second operation selected those crashes outside the ALS ambulance service area. A map showing the distribution of Minnesota's licensed ambulance services and motor vehicle crashes appears in Figure 1. The second question asks which crashes occurred within the 15 and 30 miles of a Level I or Level II trauma center. To answer this question, two concentric buffers 15 miles apart were created around the hospitals with ACS Level I and Level II trauma center verification using the Create Buffers option. These buffered areas were then used to select crashes within the specified distance of a Trauma Center. A map showing the buffered Trauma Centers and the fatal and severe motor vehicle crashes appears in Figure 2.

Using GIS to Enhance Public Health Injury Surveillance

GIS may be used to monitor the incidence and characteristics of motor vehicle crashes (and other types of injury events) over time to enhance public health injury surveillance systems. Public health injury surveillance is the ongoing and systematic collection and analysis of data describing the incidence and characteristics of injuries in a described geographic area and time period. GIS is well suited to managing and integrating the data needed to describe the who, what, when, where, and how of injuries needed for injury surveillance. These injury data may then be integrated with data describing demographic, housing, socioeconomic, and other community and environmental characteristics to help identify persons, times, and places that may be at risk of certain types of injury and useful in planning, delivering, and evaluating injury prevention programs. For example, GIS could easily locate crashes causing injury to unrestrained children to help target child passenger protection programs. GIS may then be used to display this information in maps to better inform professional and public audiences with shared interest in injury prevention.

ArcView GIS is a powerful tool to integrate and manage data from multidisciplinary trauma care resources and from motor vehicle crashes. In addition, GIS’s spatial analysis capability enables the exploration of the relationships between crashes and trauma care resources to better inform discussions of the current and desired characteristics of a comprehensive trauma care system. Finally, GIS is a useful enhancement of public health injury surveillance systems.

Lynn Boergerhoff
Trauma Services
Hennepin County Medical Center
701 Park Ave
Minneapolis, MN 55455