A Multi-State Integrated Data Approach to Analyzing Older Occupant Motor Vehicle Crash and Injury Risk Factors

项目总结/文摘

PROJECT SUMMARY/ABSTRACT Older persons represent the fastest-growing age group in the United States and motor vehicle crashes represent a serious public health problem for this age group. In 2014, there were over 6,800 deaths and over 191,000 non- fatal injuries treated in emergency departments for older persons in the United States. In order to fully understand the older occupant motor vehicle crash injury problem data are needed that cover the full span of the motor vehicle crash, from pre-event factors and actions through post-event injury outcomes. Police crash reports provide great detail about the roadway conditions and driver actions proceeding a crash. Police crash reports also give information regarding the event itself, such as the speed of the vehicles involved, the crash configuration, occupant seating placement, and restraint use. Hospital billing databases also contain valuable information regarding a motor vehicle crash outcomes. ICD-9-CM codes provide very descriptive data regarding injury specifics. Additionally, measures of injury severity such as the Abbreviated Injury Scale (AIS) and Injury Severity Score (ISS) can be derived from the ICD-9-CM codes helping to measure the threat to life that an occupant's injuries pose. Unfortunately, a common, unique key between these two databases, such as Social Security Number, is rarely collected. Therefore, innovative methods, such as probabilistic linkage, are needed to integrate the pre-event and event data from the crash report with the outcome data from hospital billing data to create a database which contains information from across all time points of the injury event. The overall goal of this proposal is to gain a better understanding of crash and other risk factors that contribute to increased risk of crashes and injury severity for older occupants. The following three aims will be used to support this goal: 1) to create a multi-state probabilistically linked data set with variables that have been mapped to standardized and uniform definitions; 2) to determine characteristics, risk factors, and behavioral trends among older passengers and drivers and to quantify each risk factor's contribution to the likelihood of sustaining an injury and economic impact of crashes; and 3) to determine characteristics of crashes and older occupants which may contribute to an increased risk or cost of injury beyond what is contained on a typical crash report including citation history and the presence of comorbidities and medical conditions. This project will create a multi-year, multi-state probabilistically linked database to address these aims. In addition to the traditional linkage between police crash and hospital billing databases, this project will also include integration projects with other public health databases such as driver license files, toxicology data, and citation and conviction databases.