Individual human behavior is modulated by the environment. Behaviors such as trip-making frequency, distance and time traveled, and walking patterns have been related to use of land, improved street connectivity, and higher employment and population density (Moudon, Hess et al. 1997). Evidence suggests that for traveling to work, location, neighborhood connectivity with other activity centers is important in explaining travel patterns (Cervero and Kockelman 1997). The importance of perceptions of safety, short blocks, and topography has been documented as critical in affecting non-motorized modes of transportation (Cervero and Duncan 2003). In settings where the majority of the population travels by foot, by public transport, or both, the influence of the built environment is crucial. An elemental aspect of this process involves the risk of or occurrence of injuries. While some studies have established associations between the built environment and physical activity (Handy, Boarnet et al. 2002; Hoehner, Brennan et al. 2003; Hoehner, Brennan Ramirez et al. 2005; Li, Fisher et al. 2005), others have shown that increased physical activity in neighborhoods is associated with increased neighborhood safety (Molnar, Gortmaker et al. 2004). The identification of associations between density (of outlets or employment) and injuries is important but there are virtually no studies looking comprehensively at the influence of the built environment on unintentional pedestrian-related injuries. Characteristics of the built environment such as density, diversity (mixed use of land), and design are essential for studying injuries but there are virtually no studies that address them all (Cervero and Kockelman 1997).
In order to study the association of pedestrian injuries and the built environment data sources of good quality are essential. The cities of
Cali, and
Bogotá,
Colombiahave developed injury mortality surveillance systems that are model in
Latin America. These citywide systems have been shown to be reliable I terms of the injury mortality information that they capture and have been used previously to evaluate city policies related to injury prevention and control (Villaveces, Cummings et al. 2000). The purpose of this project is to conduct a study of the built environment in two metropolitan areas of
Colombia, and to study its influence on the incidence of pedestrian-related deaths. This study will serve as a basis for a larger proposal that will be submitted to the NIH. Specifically, we will use aggregate measures of the built environment and evaluate their spatiotemporal association with mortality data on pedestrian injuries.
Methods
Data on mortality has been collected weekly since December 1992 in
Caliand since 1995 in Bogotá. These systems are administered by the Cisalva Institute of the
Universityof
El Vallein
Cali,
Colombiaand the National Institute of Legal Medicine and Forensic Sciences in Bogotá. Data items collected include individual numeric identifier, individual's age, sex, place of residence, place of injury event, occupation, mechanism of death, perpetrator information (when available or applicable), date of event, time of event (in hours) and weapon used (when applicable). Information from this data set is publicly available. In addition to this information, all events are geographically coded using ARC View software (ESRI GIS and Mapping Software 2005).
Aggregate data about the built environment will be collected from the city planning office. This data is produced on a yearly basis. Employment data is available for every trimester. Use of land information (diversity) is available yearly and includes characteristics such as residential, commercial, recreational, or mixed uses, density of selected outlets including restaurants, bars, and businesses, economic strata information (by city district), employment and unemployment information, basic sanitation and service coverage and basic social service density information. These variables collected here will serve as a proxy measure of the built environment which will later be directly evaluated using the Irvine-Minnesota measurement tool (Boarnet, Day et al. 2006) for the built environment. We plan to collect built environment data from 10 randomly selected areas of each city identifying street characteristics including pedestrian crosswalks, roadway number of lanes, and pedestrian crossing signs or signals. For these sites we will also measure land use (residential, school, public space, recreational, public building, commercial, industrial, and institutional - hospital, church, etc).
Analysis of data
The built environment variables will be merged with the already existing datasets on injury mortality. Previously geo-coded information will provide the geographic link between events and the built environment. Data on time and date of events will provide temporal information. We will use this information to correlate the rates of pedestrian injury events and homicide events with the characteristics of different neighborhoods within both cities and look at changes in mortality following specifically defined times and places were cities have been modified using major transportation engineering efforts. We will use Temporal Geographic Information Systems (TGIS) (Christakos, Bogaert et al. 2001) generated maps of the temporal and spatial distribution of injuries to establish specific patterns by of injuries and urban setting based on the information collected from each city. TGIS allows observers to estimate parameters (i.e injuries) when incomplete data is available (data on the built environment).
Because it is impossible to measure the characteristics of all areas of the built environment in a city, this methodology allows to select a sample of places and based on their characteristics predict the likelihood of events. TGIS has been previously used to test hypotheses about the spread of diseases in epidemiology (Haggett, 2000) but has not been used to study injuries. TGIS has been developed in great part at the
Universityof
North Carolinaand the resources at UNC are widely available. In addition to the value of the information concerning the built environment and injuries the use of TGIS will serve as a tool to predict future injury events and thus aid as a planning and primary prevention tool for injuries.
Click on the PowerPoint slide below for more information on Dr. Villaveces' research.
|
