Behavioral monitoring of DUI offenders with the Alcohol Ignition Interlock Recorder.

AIMS: To evaluate patterns of blood alcohol concentration (BAC) and driving logged on the ignition interlock recorder and to assess whether this event record is a useful outcome measure for a behavioral intervention.
DESIGN: Descriptive analyses of recorder data and multivariate analysis of the predictors of high BACs associated with a motivational intervention for driving-under-the-influence (DUI) offenders using an interlock.
SETTING: Two interlock service centers in Alberta, Canada. PARTICIPANTS: 1309 first-time and multiple DUI offenders who agreed to participate during interlock installation. INTERVENTION: A human-services (supportive guidance) intervention based on motivational interviewing and pragmatic counseling was delivered to interlock users in Calgary, but not to interlock users in Edmonton, Canada. MEASUREMENTS: This report summarizes the patterns and predictors of BAC warnings (0.02-0.039%) (20-39 mg/dl) and failures (> or = 0.04) (> or = 40 mg/dl) from more than 3 million in-vehicle breath tests. Data come from three sources: driver records, questionnaires and the interlock.
FINDINGS: From the beginning to the end of the interlock use period, there was a significant linear decline in the proportion of positive BAC driving to total driving. After controlling for prior offenses, demographics and reported drinking levels, offenders in the intervention site (Calgary) were less likely to have recorded fail BACs than were offenders in the control site (Edmonton). The temporal patterns of BAC fails with the interlock mimic the high-risk periods for DUI arrests and alcohol-involved fatal crashes.
CONCLUSIONS: The interlock successfully blocks drinking and driving during high-risk periods. Preliminary recorder data suggest the services intervention may be affecting DUI behavior.