The Crash Outcome Data Evaluation System (CODES)

00723749https://doi.org/10.21949/14031221996PDFResearch PaperDOT HS 808 338Traffic crashesCrash dataCrash analysisCrash injury researchData collectionStatisticsCrash Outcome Data Evaluation System (CODES)Data linkageHawaii;Maine;Missouri;New York;Pennsylvania;United States;Utah;WisconsinJohnson, Sandra W.Walker, JonathanUnited States. Department of Transportation. National Highway Traffic Safety AdministrationNTL-SAFETY AND SECURITY-Highway SafetyNTL-SAFETY AND SECURITY-AccidentsNTL-REFERENCES AND DIRECTORIES-StatisticsBTS Data DirectoryThe CODES Technical Report presents state-specific results from the CrashOutcome Data Evaluation System project. These results confirm previous NHTSAstudies and show that safety belts and motorcycle helmets are effective inreducing fatalities and injuries. The Report also shows that safety belt andmotorcycle use in the seven CODES states (Hawaii, Maine, Missouri, New York,Pennsylvania, Utah, and Wisconsin) could save millions of dollars in directmedical costs. The CODES project represents the first time thatoccupant-specific medical outcome and cost data for all occupants involved inmotor vehicle crashes were available for highway safety evaluation. Thetechnical report provides detailed descriptions of the crash, EMS, emergencydepartment, hospital discharge and other state data files used to generate thepopulation-based information for the Report to Congress. It describes thebackground of the CODES project, the selection of the seven states, theformation of the CODES advisory committees within each state (crucial to aproject which depended on the cooperation of various data owners and data users)and the concepts of probabilistic linkage. Variations and similarities amongthe states are discussed regarding the availability of state data, filepreparation, linkage variables, the linkage process and resulting linkagerates, and validation of the linkage results. It elaborates on the uniformresearch model used and discusses the outcome variables, additional riskfactors used as covariates, models used in the logistic regressions, andmethods of computing weighted averages of odds ratios and effectiveness. Itcompares odds ratios to risk ratios and 'effectiveness' and presentsstate-specific results for the safety-belt analyses of injury and cost ofinjury. Finally, the document gives digests of other state-specific analyses,covering topics such as: data quality, additional linkages to improve theresults, data outliers (extreme values, as in inpatient charges), alcohol anddrug use, age factors, types of safety belts, and geographic patterns in crashcharacteristics. 95p. This document is an accompaniment to the Report toCongress on Benefits of Safety Belts and Motorcycle Helmets, based on datafrom The Crash Outcome Data Evaluation System (CODES), December, 1995.84