Using a novel data linkage approach to investigate potential reductions in motor vehicle crash severity - An evaluation of strategic highway safety plan emphasis areas

J Safety Res. 2020 Sep:74:9-15. doi: 10.1016/j.jsr.2020.04.012. Epub 2020 May 4.

Abstract

Introduction: With the significant number of motor-vehicle fatalities occurring on the nation's roadways in recent years, there exists a need to integrate a more complete range of data sources, available at a regional or statewide level, to effectively evaluate existing safety concerns and quantify their impacts. Crash data alone does not provide ample crash-associated citation, injury, and roadway characteristics; therefore, a more cohesive dataset is required to accurately and completely analyze the true impacts of motor-vehicle crashes. Previously developed strategies linked crash data with citation and roadway inventory data to enhance the identification and optimization of highway safety strategies.

Method: The main objective of this research focused on developing a new deterministic linkage between crash and Emergency Medical Services (EMS) data, by utilizing the Massachusetts Crash Data System (CDS) and the Massachusetts Ambulance Trip Record Information System (MATRIS).

Results: After several iterations of match criterion, the validated linkage successfully matched 58.3% of MATRIS records (containing an Injury Cause of Motor Vehicle Crash) to a CDS person record (55011 linked pairs, between 2014 and 2016). The data linkage provided significant insight into injury trends in several highway safety emphasis areas such as roadway departure, speeding-related, and distraction-affected crashes. The findings from this research are twofold: (1) an established process for linking previously separate data sets, and (2) a mechanism for analysis that provides decision-makers and safety professionals with a better measure of crash outcomes.

Keywords: Crash data system (CDS); Data linkage; Emergency medical services (EMS); Traffic safety.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Accidents, Traffic / prevention & control*
  • Accidents, Traffic / statistics & numerical data
  • Emergency Medical Services / statistics & numerical data
  • Humans
  • Information Storage and Retrieval / methods*
  • Information Storage and Retrieval / statistics & numerical data
  • Massachusetts
  • Motor Vehicles*