The University of Massachusetts Amherst
University of Massachusetts Amherst

Search Google Appliance

Links

Case study assessment of crash data challenges: Linking databases for analysis of injury specifics and crash compatibility issues

TitleCase study assessment of crash data challenges: Linking databases for analysis of injury specifics and crash compatibility issues
Publication TypeJournal Article
Year of Publication2006
AuthorsBenavente M, Knodler J.Michael A, Rothenberg H
JournalTransportation Research Record
Volume1953
Start Page180
Pagination180-186
Abstract

Data from crash reports do not generally include sufficient injury information for analysis of particulars of injuries resulting from crashes. In the Massachusetts Crash Data System, the primary field that relates to injury status is based on a well-known police-reported scale. More specific data on injuries and their associated hospital charges can be obtained with crash data from police reports at the crash scene being complemented with injury data from the medical care system. A database that includes information about the crash and specifics about the injuries will allow for the identification of possible injury patterns associated with particular crashes and their associated hospital charges. However, the lack of unique identifiers to link records in the different databases as well as data quality issues, such as misreporting and missing values, make challenging the process of combining these two databases and analyzing the data. This paper focuses on the challenges in the process of linking, through use of the Massachusetts Crash Outcome Data Evaluation System, crash records to inpatient hospital records. The paper illustrates the findings through a particular application analyzing crash compatibility issues. A mismatch crash occurs when a large vehicle collides with a smaller one, and size and weight incompatibilities could make the passengers traveling in the small vehicle more vulnerable to severe injury. Probabilistic linkage and multiple imputation were used to complete the linkage of 12 months of crash data to hospital inpatient discharge data.

URLhttp://dx.doi.org/10.3141/1953-21
DOI10.3141/1953-21