Study 1
Saltz and colleagues (2021) conducted a randomized matched-case trial assessing the effects of California’s community-level alcohol prevention interventions on reducing alcohol-related motor vehicle crashes among drivers ages 15−30. All medium-sized cities (50,000−450,000 population) were assessed for inclusion in the study; there were 138 medium-sized cities identified from the 2000 census. The study authors then created a subset of 50 cities through a process of random selection, which was modified to avoid selecting two neighboring cities. Cities were matched via cluster analysis using a combination of demographic characteristics (e.g., population size, ethnic composition), based on 2011 census data, and the prevalence of alcohol-related problems (high consumption, alcohol-related crashes, underage drinking).
A total of 24 California cities were matched into pairs before randomly assigning 12 cities each to the treatment group and control group. Treatment group cities received high visibility, driving under the influence (DUI) enforcement operations (e.g., sobriety checkpoints and saturation patrols to provide visible enforcement of drink driving laws, public awareness media such as posters and mobile signs), from April 2013 to March 2016. The control group cities did not receive any high-visibility DUI enforcement operations during the same period. There were no statistically significant differences in demographic characteristics between the treatment group cities and the control group cities. The treatment group’s average city population size was 104,512, and 26.9 percent of the population identified as an ethnic minority. The control group’s average city population size was 101,751, and 27.8 percent of the population identified as an ethnic minority. Drivers ages 15−30 were involved in 57.6 percent of all motor vehicle crashes and 44.0 percent of single-vehicle nighttime crashes in treatment group cities during the study period, compared with 61.3 percent and 46.9 percent, respectively, in control group cities.
The outcome of interest was a monthly percentage of all motor vehicle crashes that were single-vehicle, alcohol-related nighttime crashes for drivers ages 15−30. Multilevel analyses were conducted to examine treatment effects on alcohol-related crashes among drivers. Data from the California Statewide Integrated Traffic Records System was used to create a monthly measure of alcohol-related motor vehicle crashes for each city from January 2010 through April 2016. The study authors computed the percentage of all motor vehicle crashes that were single-vehicle, alcohol-related nighttime crashes for drivers ages 15−30 (and for all drivers). A continuous time variable representing 76 months was created for the analyses, and the analyses used a random effects model. Robust standard errors were used to account for the non-normal distribution of dependent variables. The study authors also conducted subgroup analyses on treatment dosage.