Study 1
Holder and colleagues (2000) conducted a five-year quasi-experimental evaluation of the Community Trials Intervention to Reduce High Risk Drinking (RHRD) from 1992 through 1996. A longitudinal multiple time-series analysis was used to assess the impact of RHRD across three intervention communities with populations of approximately 100,000 in northern California, southern California, and South Carolina. The three sites were selected because they were interested in testing RHRD strategies. Matched communities were used as the comparison groups and did not receive the intervention components.
The five RHRD program components (community mobilization and awareness; responsible beverage service; reduced underage drinking; reduced drinking and driving; and stricter alcohol access) were implemented in phases within each intervention community. Data were collected throughout the 12-month post-intervention period. Data sources included: traffic crash records; emergency room surveys; local news coverage of alcohol-related topics; intoxicated patron and underage decoy surveys; roadside surveys conducted on weekend evenings; and a community telephone survey (including self-reported measures of drinking and of drinking while driving).
Approximately 120 random digit dialing telephone surveys of individuals over the age of 18 from intervention and comparison sites per month were conducted over 66 months. Six self-reported measures were obtained from the surveys. Every three months the proportion of survey respondents that reported alcohol consumption was assessed, with the number of observations varying by outcome measure. Multivariate analysis was used to control for differences, with authors reporting that the differences in outcomes were minimal.
Traffic record data on motor vehicle crashes were obtained from the California Statewide Integrated Traffic Reporting System and the South Carolina Department of Public Safety. Monthly aggregated crash rates were estimated for each community for nighttime crashes (8PM to 4PM), driving under the influence (DUI) citations, and daytime crashes (4AM to 8PM). Emergency department (ED) surveys were conducted in one northern California intervention-comparison matched pair and one South Carolina intervention site. Permission could not be obtained to conduct surveys at the other sites. Over the course of the study, 7,817 injury cases were admitted to EDs and 5,941 interviews were attempted, with an overall 58 percent response rate. Archived hospital data were also collected to measure the observed number of hospitalized assault cases from matched intervention-comparison sites in northern and southern California (data were not available from South Carolina). No subgroup analyses were conducted.
Study 2
Treno and colleagues (2007) conducted a quasi-experimental time-series comparison study to assess the efficacy of implementing RHRD at the neighborhood level in two roughly comparable communities in Sacramento, California (referred to as the South and North neighborhoods). The study, the Sacramento Neighborhood Alcohol Prevention Project (SNAPP), used environmental approaches developed in the earlier Community Trials Intervention study conducted by Holder et al. (2000) that focused on whether RHRD could be applied at the neighborhood level.
Using geo-statistical techniques, SNAPP assessed whether South and North census block groups (n = 37) changed with respect to all other block groups in Sacramento (n = 243). Data from the California Department of Alcoholic Beverage Control indicated that the two study areas had higher concentrations of bars and off-site alcohol outlets per roadway mile. Data from the 2000 Census revealed that the study areas also had similar rates of assaults (per 1,000 population). Thus, although not comparable to the city of Sacramento at large, results from the North and South sites were roughly comparable to each other.
Based on these findings, SNAPP was implemented from July 2000 to July 2001 (year 1) in the South neighborhood, while the North neighborhood served as the no-treatment comparison site. A second SNAPP targeting the North was conducted from January 2002 to January 2003 (year 2) to provide for comparison and to test the interventions’ long-term impact. Data were also collected from Sacramento at large to serve as a control for historical conditions at both sites.
Researchers conducted a process evaluation to track mobilization, community awareness, and law enforcement activities through the SNAPP management information system. Changes in alcohol access were monitored using Apparent Minor and Pseudo Intoxicated Patron Surveys, which were conducted in two stages: (1) an off-site establishment scouting survey was conducted to provide information on the distributor and surrounding environments; and (2) a buyer and a driver team was sent to purchase alcohol from selected sites and to obtain descriptive information about the clerk, the attempt, and the environment. The establishments’ alcohol license data were obtained from the Department of Alcoholic Beverage Control. The service frequency of alcohol to actors pretending to be intoxicated during purchases was documented.
An outcome evaluation of SNAPP was conducted using data on alcohol-related injuries and police incidents, which were indexed by police calls for assaults and public drunkenness and by Emergency Medical Services (EMS) reports of assaults, motor vehicle accidents, and other alcohol-related incidents. Data were aggregated on a monthly basis. Sacramento City Police crime data were collected between January 1996 and December 2003. Sacramento Fire Department and EMS data were collected between January 1995 and December 2003. Police data included seven years of monthly data (84 observations), while EMS data included eight years of monthly data (96 observations).
Statistical analyses of data were conducted using seemingly unrelated regression equation (SURE) models that treat each time series separately. These analyses also controlled for possible autocorrelation due to the nesting of communities. No subgroup analyses were conducted.
