Motor vehicle crashes are a leading cause of death and injury in the United States, accounting for 33,561 fatalities and 2.4 million injuries in 2012. Although risk factors for fatal crashes are welldocumented in transportation safety research, fewer studies exist regarding risk factors for nonfatal injury crashes. Several States have developed systems that link crash reports to medical data to create a source of nonfatal injury data; however, these systems vary from State to State. One example of a State linkage system is the Crash Outcome Data Evaluation System (CODES), which was originated by the National Highway Traffic Safety Administration in 1992. In CODES, NHTSA provided partial funding and technical assistance to State linkage programs through cooperative agreements in participating States. In 2009, NHTSA advised funded CODES programs to plan for transition to independent State systems–without NHTSA funding or technical assistance. NHTSA finalized that transition in 2013. Outside of CODES, some States have conducted linkage projects, and other motor vehicle safety organizations also have expressed interest in linkage programs. From 2010 to 2012 the Centers for Disease Control and Prevention and NHTSA formed a partnership to explore the feasibility and benefits of using linked data in CODES to determine risk factors for motor vehicle injury and to design and evaluate interventions to address these risk factors. This report describes a program appraisal of data linkage programs that was conducted during the final year of the partnership. The purpose was to examine State CODES and other State linkage programs to better understand the facilitators and barriers to linking and analyzing data to produce useful products. This information can be used to guide future improvements to a State’s data linkage system and to implement a new State data linkage system. The program appraisal was supported by the University of Maryland National Study Center for Trauma and EMS (NSC) under its CODES cooperative agreement as a resource center. The study team was composed of CDC, NHTSA, and NSC specialists who jointly designed and conducted the study to document information that could improve current and future programs. The study team conducted a literature review regarding organizational theory to determine what theory would provide a framework for studying State data linkage systems. Because such programs involve a coalition of organizations and an information technology system, the CCAT and the ITIM were selected and used to develop a framework for the study of the following nine theory constructs. * Community context * Lead agency/convener group * Coalition membership * Leadership and staffing * Pooled member and external resources * Member engagement * Implementation of strategies/information technology infrastructure * Outcomes * Community capacity. Data collection consisted of collecting primary data (a survey of States conducting linkage programs, followed by focus groups with the surveyed States to further explore the survey results) and obtaining secondary data on program operations and products (available State Traffic Records Assessments and bibliographies of publications using linked data). CDC obtained approval for primary data collection from the Office of Management and Budget (OMB) under OMB clearance number 0920-0879. The survey was administered to 16 CODES States and 14 States that were known to have State data linkage systems outside of CODES during December 2012. Five focus groups were also conducted with these same States during May 2013 to explore the survey results in more detail. Twenty-five States (response rate 86%) completed the survey, and 19 of the responding States participated in follow-up focus groups (response rate 76%). Results of the survey indicated that the majority of responding States identified motor vehicle injury as a high priority health problem for their State. States consistently reported having crash, hospital inpatient, and emergency department data available to link, and most States were successful in linking those datasets. Through data results and the construct applications, the team noted key facilitators and barriers to collecting, linking, analyzing, and reporting data . Key facilitators included: * Characteristics of the community, such as the importance placed on motor vehicle safety, that either facilitate or inhibit the linkage of data and the use of the resulting database; * Housing overall responsibility for the data linkage program in one organization; * Establishing a coalition that includes all data owners; * Establishing memorandums of understanding and/or undergoing institutional review board review to obtain data; * Establishing matching variables in the data; * Interacting with other State and local organizations interested in motor vehicle safety; * Staffing with personnel that have the appropriate skills, including a high level of expertise with the linkage software packages, epidemiology, statistics, knowledge of traffic safety, data sources being linked, and presentation and marketing skills to ensure data are used; * Providing technical assistance and training in linking and analyzing linked data; * Providing stable funding; and * Participating in a community of practice of data linkage practitioners. Barriers noted included: * Insufficient funding, * Staffing turnover, * Lack of process documentation, * Long lag times in obtaining source data for linkage, * Statutory requirements for obtaining and reporting data, * Complex linkage techniques such as probabilistic linkage, and * Marketing linked data so that others understand how they can be used to increase traffic safety. State data linkage systems that link police crash reports to medical records are important sources of data on non-fatal motor vehicle crashes and injury severity from those crashes. The information obtained in this study on facilitators and barriers to effective State linkage systems can be used by States to modify existing data-linkage systems to improve the efficiency and success rate of data linkage, conduct more analysis of linked data, and increase dissemination of the analysis. This information can also be used by States, when starting a new data linkage program, to design a program based on the experiences of other States. (Author/publisher)
Abstract