Return to NHTSA homepage

Banner with logos of contributing organizations: FHWA, FMCSA, NHTSA, GHSA, and NCSA

Back   Table of Contents   Forward

MMUCC
M
ODEL MINIMUM UNIFORM CRASH CRITERIA GUIDELINE

Improving Crash Data for Safer Highways


APPENDIX D


Example of a Highway Safety Linked Data System

An example of data linkage (see figure below) is the Crash Outcome Data Evaluation System (CODES) project that evolved from the Intermodal Surface Transportation Efficiency Act (ISTEA) of 1991. ISTEA mandated that NHTSA prepare a report to Congress about the benefits of safety belt and motorcycle helmet use. To obtain the crash outcome information necessary for this report, NHTSA sponsored the CODES project and awarded grants to Hawaii, Maine, Missouri, New York, Pennsylvania, Utah, and Wisconsin. Currently, a total of 27 states have received CODES funding. Each grantee links available state crash, EMS, emergency department, hospital discharge, insurance, and other traffic records and carries out analyses using the linked data to support their state’s highway safety problem identification, planning/policies, and program management.

Flow chart: CODES Model

Successful linkage, using probabilistic techniques, of person-specific crash and injury records relies on indirect identifiers that are sufficient to discriminate among the crashes and the persons involved in each vehicle involved in a crash.


Data Elements That Discriminate Among the Different Crashes

Date, time, geographic location and type of vehicle are commonly used to discriminate among the crashes. Time should be recorded in hours and minutes, particularly for urban areas where multiple crashes occur within a given hour. Latitude and longitude are useful for assigning the location of a crash. When coordinates are not available on the police crash report, coordinates are used that indicate the center of a jurisdiction (city, town, county). Health care facilities and EMS agencies are assigned coordinates according to their address. Type of vehicle is defined in general terms as motor vehicle, motorcycle, pedestrian, pedalcycle, etc.


Data Elements That Discriminate between the Persons Involved in a Crash

Date of birth (or age), sex, person type (driver, passenger, rider), vehicle number in crash, injury type, severity (KABCO) are used to discriminate among the persons involved in a crash. Direct identifiers such as name, initials, or social security number (sometimes truncated) are not usually available because of privacy and confidentiality policies.


Unique Identifiers For Other Types of Highway Safety-Related Linkages

Linkages between the crash and administrative files such as driver licensing, vehicle registration and roadway inventory files are more likely to have access to a unique number such as a driver license number, vehicle plate number or roadway marker. Indirect identifiers are also useful to compensate for potential data collection errors.


Banner with logos of contributing organizations: FHWA, FMCSA, NHTSA, GHSA, and NCSA

Back   Table of Contents   Forward