Alcohol Impaired Driving in the Natural Environment

Overview

This is a pilot study to set up a larger investigation examining predictors of the decision to drive after consuming alcohol. All participants will carry a study provided smartphone and breathalyzer device for the 2 week period of the study. The intervention is that participants are randomly assigned to one of 2 breathalyzer feedback conditions - one where they receive a warning that their results indicate they should not drive and one where they receive no feedback. The study is designed to provide information needed for a larger version with a similar protocol, but also to provide an initial test of project hypotheses as well.

Despite previous success in reducing alcohol impaired driving (AID) and fatal crashes, rates have not decreased significantly since the 1990s. Developing novel approaches to prevention/intervention are likely required to produce further progress. This project is designed to evaluate a potential intervention for AID using mobile technology. In addition, the project is designed to provide pilot data for a large scale investigation. Recent laboratory work by the PI has demonstrated the utility of assessing AID risk factors when participants are intoxicated. We identified four risk factors that, measured under intoxication, are cross-sectional predictors of AID: AID attitudes (Morris et al., 2014), impulsivity (McCarthy et al., 2012), behavioral economic demand (Amlung et al., 2016) and subjective intoxication (Amlung et al., 2014). The proposed project will evaluate these factors as prospective predictors of AID and extend this work outside the laboratory and into participants' natural drinking environment using Ambulatory Assessment (AA). AA is a set of techniques used to collect data from individuals in the course of their daily life. AA methods provide increased ecological validity across multiple modalities (e.g., self-report, physiological measures). No study to date has used AA to examine AID. We will employ a combination of three AA methods: ecological momentary assessment, geospatial technology, and portable breath analysis. Ecological Momentary Assessment (EMA) is a form of AA that allows participants to self-report on their current thoughts, feelings, and behaviors in their natural environment (Shiffman et al., 2008). EMA avoids most pitfalls of retrospective reports, and is increasingly used to study temporally proximal influences on substance use behaviors (Shiffman, 2009). Despite the importance of such influences to AID decisions (Quinn \& Fromme, 2012), no study to date has applied EMA methods to the study of AID. Advancements in mobile technology, such as global position systems (GPS) and portable breathalizers, allow for the collection of objective data on participant location and alcohol use in real-time. We will combine location data (e.g., drinking venue, distance from home) and in-the-moment breathalyzer data with self-report EMA data from each participant drinking episode. This combination will allow for a more complete assessment of event-level factors that contribute to AID decisions. This pilot project will test the hypothesis that temporally proximal measures of AID risk factors (subjective impairment, perceived risk) collected during a drinking episode will improve prediction of AID, over and above trait measures of AID risk factors. The pilot project will also test the potential for AA to function as an intervention for AID. In this pilot, participants will be randomly assigned to either receive some feedback from their breathalyzer (a warning that they have consumed too much to drive safely) or not to receive any feedback. Note that participants will not be told their BAC, due to potential risks of specific feedback (increased consumption, titrating drinking to drive at slightly below .08). Instead, feedback will be general ("Your results indicate that you have drank too much to drive safely"), and participants will not be aware of what BAC triggers these warnings. Comparisons of all participants AID behavior at baseline (from a Timeline Follow-back assessment) with their AID behavior during study participation will be used to test whether the increased self-monitoring of AA can alter AID decisions. In addition, comparisons between groups (feedback vs. none) will test for potential effects of BAC feedback on AID decisions. Support for these hypotheses could lead to novel, cost-effective interventions to reduce AID using mobile technology. This pilot will enroll 40 moderate-to-heavy drinking young adults. Participants will complete a brief laboratory session of AID risk measures, followed by two weeks of AA. They will be randomly assigned to one of two conditions (BAC feedback vs. none). Interview assessments of AID will be conducted at baseline and at the completion of AA assessments, with interviewers blind to condition.

Conditions

Interventions

Eligibility

Locations (1)

Outcomes