The Crash Outcome Data Evaluation System (CODES) is a State-based program initiated in 1992 by the National Highway Traffic Safety Administration (NHTSA) under NHTSA’s National Center for Statistics and Analysis. CODES links State crash report data to State medical outcome data using an advanced probabilistic linkage methodology. Probabilistic linkage is a powerful method for combining information from different databases into a single dataset for analysis. Desired information about study subjects is often contained in two or more databases, and if a unique key to allow perfect matching does not exist between these databases, it is not possible to combine the information directly. Rather than relying on a unique key to combine records, probabilistic linkage makes use of fields that are common to each database. The CODES Data Network, a supplemental program for States that had successfully completed a CODES startup, involved cooperative agreements between NHTSA and participating States. The agreements allowed NHTSA to provide technical assistance and limited funding that would not have otherwise been made available to States. In 2013, NHTSA discontinued the CODES Data Network and transitioned the CODES program to full State-level responsibility. NHTSA encouraged transitioning States to continue the program independently and to seek and secure other funding sources, including NHTSA funds via State grants when appropriate. CODES data reside in the States where the linkage originated, and NHTSA does not own or disseminate CODES data files. Queries may be directed to State projects (Table 5). To preserve knowledge about the program and help inform current and future linkage projects, this paper summarizes the history of CODES, as well as the transition plan. This report also describes the CODES Data Network output from 2009 to 2014. Special studies completed by the CODES Data Network from 2009 to 2014 included: * a motorcycle helmet study using data from 18 States; * an alcohol reporting data request using data from 7 States; * a two-State collaboration on comparing injury measurement methods; and, * a final technical report on methodology and pooled data applications. Collaborations with other Federal agencies from 2009 to 2014 were: * a collaboration with the Centers for Disease Control and Prevention involving data analyses and a program analysis of linkage projects with a survey and focus groups; and, * data support to the National Transportation Safety Board for a study on single-unit trucks. State-specific data applications shared by grantees in 2009 to 2012 covered: * objectives of problem identification, traffic safety decision-makers support, traffic safety legislation support, and public education; and, * subject matter areas such as child safety, impaired driving, motorcycle safety, seat belts, young drivers, crash factors, injuries, older occupants, pedestrians, data dissemination, and coalition building. (Author/publisher)
Abstract