|Date added||June 6, 2010|
|Category||2010 CMRSC XX Niagara|
|Author/Auteur||Garreth Rempel, Jonathan D. Regehr, Jeannette Montufar|
This paper describes the approach and results of a feasibility study to apply data linking for injury prevention in the Alberta commercial motor vehicle (CMV) workplace. The study, which is the first of its kind in Canada, is an initial step to help understand the causes and characteristics of CMV collisions in Alberta in order to determine appropriate future countermeasures. Governments today use many databases for research, policy-making, planning, and programming. There are myriad databases available for CMV safety analysis; however, while individual databases may support specific needs and analyses, it has proven to be difficult to use these databases together to create an “enhanced knowledge-base” that can help improve road safety, particularly involving CMV workplace-related issues. Data linking facilitates database integration and allows highway officials to select and evaluate tailored countermeasures and road safety strategies specific to the causal factors of traffic collisions at individual locations and assess and calculate economic cost forecasts for different collision types. However, due to the complexity of this issue and the wide variety of existing data owners and databases, it is necessary to first understand the types of data available for analysis to assess the feasibility of applying data linking for injury prevention. The data linking feasibility assessment described in this paper comprises four parts: (1) a literature review of data linking to determine benefits and challenges, (2) an inventory and characterization of available databases, (3) data linking architecture development, and (4) the identification of challenges that may constrain data linking feasibility. Common challenges associated with data linking are policy and legislative issues, technical issues, and institutional and administrative issues. The data inventory and characterization involves identifying core and peripheral databases. Core databases are considered critical for data linking efforts and peripheral databases are those which, if linked to the core databases, would be relevant for only very specific queries or applications. Data linking agents for each database are identified with the collision database serving as the central component. The architecture developed uses the Haddon matrix to identify efficient data linkages between pre-collision, collision, and post-collision databases that describe the driver, vehicle/carrier, and road/environment. Applying data linking for injury prevention in the Alberta CMV workplace is feasible and doing so provides a more holistic view about the causes and characteristics of CMV collisions. This view facilitates the evaluation of collision countermeasures, improved collision reduction strategies, better investment prioritization, and more informed policy and regulatory direction. Further, the feasibility assessment approach described in this paper can be used as guidance for other jurisdictions interested in similar data linking studies.
Garreth Rempel, Jonathan D. Regehr and Jeannette Montufar