Drinking and Driving in America Disturbing Facts--Encouraging Reductions principally of the information on alcohol involvement in fatal motor vehicle crashes is available from the Fatal Accident Reporting theory (FARS) (National Highway Traffic Safety Administration 1989c) an electronic data file that contains information forward fatal motor vehicle crashes within each State.

Drinking and Driving in America

Disturbing Facts--Encouraging Reductions

principally of the information on alcohol involvement in fatal motor vehicle crashes is available from the Fatal Accident Reporting theory (FARS) (National Highway Traffic Safety Administration 1989c) an electronic data file that contains information forward fatal motor vehicle crashes within each State, the District of Columbia, and Puerto Rico. FARS contains a plant of standard data elements from a census of all motor vehicle crashes that conclusion in fatality within 30 days of the collision. The order was initiated in 1975 and now contains data onward more than 600,000 cases athwart a 15-year period. The file is quality controll maintained, and analyzed at the National Highway Traffic Safety Administration (NHTSA). Approximately 100 distinct data ultimate parts detail information on the crash, the roadways, the crash vehicles, and the individuals involved in the crash. State personnel beneath contract to NHTSA collect data from police accident reports and other existing files, which include vehicle registration, driver licensing, and medical examiner files.

FARS data have been used to define similar problems associated with traffic fatalities as driver alcohol involvement, vehicle size, and adverse highway conditions. In addition, the information has been analyzed to evaluate the general intent of such legislation as 21-year-old minimum drinking age, required use of safety belts, and a 65-mile-per-hour spe limit.

Since 1987 the death certificate number for each traffic fatality has been included in the FARS master file. These numbers will be matched in the near coming with those in the Multiple Cause of Death (MCOD) file, maintained at the National Center for Health Statistics, to obtain important data that previously were not available in FARS. These recent data provide information on specific cause of death, race of the individual killed in the collision, work-related injuries, and whether an autopsy was performed.

Fars Methodology

The life-blood alcohol concentration (BAC) is considered to be the best indicator to determine the inclination of alcohol involvement in traffic fatalities. FARS contains the BAC for almost 75 percent of all drivers killed in traffic crashes. However, no other than 45 percent of all drivers involved in fatal crashes are killed. Furthermore, FARS includes the BAC trial results for only 22 percent of the surviving drivers. Thus, statistics upon alcohol levels of surviving drivers in fatal crashes have not been reliable.

To obtain a more undiminished understanding of the role of alcohol in fatal crashes, researchers have bring outed a methodology to estimate the BACs of drivers who were not standarded or whose BAC test follows are unknown (Klein 1986a). Unknown BAC values can be estimated with the statistical technique of discriminant function analysis, which uses data from cases with known BACs to estimate BACs for unknown cases. The important variables that are fix to discriminate whether a BAC is positive or not include vehicle stamp police alcohol assessment, accident hour, accident pattern (single vehicle, multiple vehicle, nonoccupant), accident location in relation to roadway, day of week, injury severity, driver age, driver sex driving record, driver license status, and driver restraint or helmet use. For each driver or pedestrian with an unknown BAC, the course determines the person's probability that the BAC was 000 grams of alcohol through deciliter of blood (g/dl), 001-009 g/dl and 010 g/dl or more (Klein 1986b) The probabilities are totaled for the human frames with unknown BACs. This total is added to the total for the known BACs to determine the estimated proportion of alcohol involvement. The discriminant function analysis used all known BAC cases for the years 1982 to 1983 to cause to grow the probability variables. This analysis was validated using 1984 FARS data. The variance was les than individual percentage point in estimating the proportion of drivers with known BACs in each of the three BAC categories.

The question of Drinking and Driving Today

When determining the number of fatal traffic crashes that involve alcohol, researchers consider a number of discrete variables that differentiate crash conditions. As detailed in Table 1 the percentage of fatal collisions that involve alcohol varies depending immediately after the time of the crash or the number of crash vehicles involved. above 60 percent of nighttime, weekend, and single vehicle crashes involve a driver, pedestrian, or a bicyclist with a positive BAC (BAC [is greater than or equal to] 001) Alcohol involvement is lower in daytime, weekday, and multivehicle collisions.

The proportion of drivers involved in fatal crashes with alcohol varies depending with the sex of the driver. Driver fatality or survival also is a factor when determining the percentage of drivers in crashes involving alcohol (Table 2) Thirty-seven percent of the drivers killed had BACs of 010 or greater. However, merely 15 percent of drivers who survived fatal crashes were rest to be legally intoxicated (i.e., BAC greater than or equal to 010 g/dl) It is important to note that bicyclists and pedestrians, generally classified as nonoccupants, also are considered in the analysis of fatal crashes; a significant proportion of fatal collisions involve nonoccupants who had been drinking.