Probabilistic Linkage of Trauma Registry and CRASH Databases and Quality Improvement of Data Elements for the Presence of Drugs and Alcohol in Injured Drivers

The May 2012 Kentucky Traffic Records Assessment report by the National Highway Traffic Safety Administration Technical Assessment Team recommend an additional dataset that links the CRASH and injury surveillance/KY Trauma Registry (KTR) systems (at sections 1-C and 2-F). This project will produce the recommended dataset and also address the report’s section 2-F recommendation for assessment of KTR e-coding quality. As part of the project we will 1) perform probabilistic linkage with KTR and CRASH databases to integrate these components of the KY Traffic Records System and expand data sharing; 2) assess integration of the KTR and CRASH systems; 3) assess the completeness of the data elements in the linked KTR and CRASH records pertaining to presence of drugs and alcohol in injured drivers; and 4) develop quality improvement indicators for drug and alcohol use data elements within the KTR system. With current support from the Kentucky Transportation Cabinet’s Office of Highway Safety, the KTR system has shown measurable progress. The number of participating facilities increased from 3 in 2008 to 16 in 2011, and the total number of trauma cases reported to the KTR increased 30% from 6,568 to 8,509. The expansion of the KTR and the improvement of its data quality elements are directly related to the improvement of the Statewide Injury Surveillance System and can significantly improve the completeness, accuracy and consistency of Kentucky traffic records, especially the crash records for the most severely injured crash victims. The Kentucky Trauma Registry has not yet been linked to the collision data. This project proposes to: 1. Perform a probabilistic data linkage between the 2011 CRASH database and the 2011 Kentucky Trauma Registry database using data elements such as incident time, date, location, crash victim’s date of birth, gender, and others; 2. Assess the quality of integration between CRASH and Trauma Registry systems and outline recommendations for quality improvement of the two databases with regard to future probabilistic linkages for analytical purposes; 3. Evaluate the accuracy and completeness of CRASH and trauma records on the following data elements: - Person type (crash data elements: person type, position in/on vehicle; trauma registry data elements: fourth-digit subdivisions used with ICD-9 external-cause-of-injury codes to identify the injured person) - Presence of drugs in injured drivers: (crash data elements: contributing human factor=drug involvement, test offered, chemical test, tested for drugs, results; trauma registry data elements: drug use indicator) - Presence of alcohol in injured drivers: (crash data elements: contributing human factor=alcohol involvement, suspected drinking driver, method of determination, test offered, chemical test, tested for alcohol, test taken by, test sent to, results; trauma registry data elements: alcohol use indicator); 4. Develop and pilot-test quality improvement indicators for drug and alcohol use data elements in the Trauma Registry system; 5. Pilot-test new data elements for specific blood alcohol levels and specific drug classes within the state Trauma Registry database to improve the comprehensive surveillance of the presence of drugs and alcohol among injured trauma-activated drivers.