Date:   January 20, 1999

By:     Reid K. Hester, Ph.D., Scott Walters, M.A.,

           Michael Sergent, M.A., & Bo Miller, M.S.W., Research Division

This project has had two goals: to provide cost-effective treatment for DWI offenders in Torrance County and to evaluate its effectiveness. Past research has shown brief interventions to hold great promise in reducing alcohol-related problems in DWI populations. Building on this base, this project has examined the impact of providing a brief intervention and empirically supported self-help materials for offenders.

Participants received a brief motivational intervention known as the Drinker’s Check-up (DCU). They were also given a choice of two empirically supported self-help options: The Addictive Behaviors Change workbook (Horvath, 1996) or Behavioral Self-Control Training either via the Miller & Munoz (1984) book or the Behavioral Self-Control Training Program for Windows (BSCPWIN).

This project began with the recruitment and training of therapists. First we advertised for and interviewed a number of potential therapists. We chose four counselors who met our criteria of licensure, professional training, and clinical experience. We invited these four to participate in a 2 day training program in Motivational Interviewing and the clinical protocol. The two counselors we chose for the project, Mike Sergent and Mary Beth Theisen, both had excellent credentials, clinical experience, and demonstrated good skills in the training workshop. We were also pleased with the performance of the other counselors and encouraged them to continue using the skills they had learned in the workshop in their clinical practices in the Moriarty area.

Once the counselors were selected we provided them with additional training in assessment and implementation of the protocol. We also developed forms for logging participation, noncompliance, feedback to the participants, etc.. We provided the counselors training in the use of the BSCPWIN software and guidelines in recommending the different self-help materials. Copyrighted assessment materials were ordered, received, and distributed to the counselors. Copies of the self-help materials were also acquired and provided to the counselors.

Prior to receiving the first referral, we made a change in the treatment program assessment protocol. On the basis of a factor analysis published in September, 1996 on the SOCRATES instrument we adopted the newer, shorter version. We also dropped the neuropsychological screening component in light of other recent data indicating that the component does not substantively contribute to outcome. In its place we added the World Health Organization’s Alcohol Use Disorders Identification Test (AUDIT) and the Drinker’s Inventory of Consequences (DrInC). Both are measures of negative consequences and have good psychometric properties.

Making these changes in the protocol resulted in an assessment that is more similar to the Motivational Enhancement Therapy intervention that was used with great success in the Project MATCH collaborative study.

Prior to receiving the first referral, we also developed a liaison with the courts. Through telephone contacts and face-to-face meetings, we learned what the court personnel needed to facilitate referrals. They also learned what we needed from them to facilitate a smooth referral process. That liaison began in October, 1996 and has continued on a regular basis to the present.

In Year 1 (11/96-11/97) the two counselors, Mike Sergent and Mary Beth Theisen, received close clinical supervision on their cases on a regularly scheduled basis. They also had access to additional supervision whenever needed. Sessions were audio taped and those tapes were reviewed in supervision sessions. The closeness of this supervision helped to ensure close adherence to the clinical protocol. Although supervision was more intense during the initial phases of the project, it continued on a regular basis throughout year 1. In Year 2 counselors were supervised on an as needed basis. We decreased the intensity of supervision because of the high level of proficiency demonstrated by the counselors.

Referrals for the overall project began in mid-November, 1996. The 12 month follow-ups for those clients referred in the first year (11/96-11/97) were completed by early December, 1998. This report documents the outcomes of this first cohort of DWI offenders referred to the program.

There were no new referrals to the program year 3 (11/98-11/99). Although there were a number of meetings with the DWI Coordinator and court personnel, referrals stopped entirely by May, 1998. Since then we have been involved in completing the 12 month follow-ups for those referred in year 1.

Forty-one participants completed pretreatment measures. Thirty-three (81%) were male. The mean age was 37.8 (SD = 13.7, range = 18-72 years) and the mean years of education completed was 11.3 (SD = 1.8). Fifty-one percent reported living with a spouse or partner, 27% lived alone, 17% lived with parents, and two people (5%) lived with children only. Thirty-seven percent were single, having never been married, and 39% reported being currently married and living with spouse. One person (2%) was widowed and 20% were divorced.

Nine (22%) of the total sample of 41 did not return for either follow-up point and were not included in the statistical analyses. Thirty-two (78%) clients completed at least the 3-month follow-up and 24 (59%) completed 12-month follow-up point. Additionally, 3 (7%) of the clients had their third follow-up at approximately 7 months because they were leaving the area (e.g., joining the military) and would not be available for a 12-month follow-up. This relatively low follow-up rate at 12 months prevents us from having a high level of confidence in this follow-up point, because individuals who cannot be located for follow-ups typically have worse outcomes than those who can be located. We examined clients’ drinking characteristics at pretreatment to see if any predicted whether an individual would or would not be found at follow-up. Our examination suggested that those clients who were heavy episodic drinkers at intake were less likely to be available for follow-up while the heavy steady drinkers were more likely to be available. However, none of these comparisons were significant when we took into account the number of multiple contrasts tested.

Participants reported having had their first drink at an average age of 17.4 (SD = 6.0). As measured by the Brief Drinker Profile, participants reported drinking, on average, 22.0 drinks per week (SD = 41.7, range 0-236.4), 10.9 drinks per drinking day (SD = 16.1) and, achieved an average weekly peak BAC of 152mg% (SD = 250). Episodic drinking was less frequent, and in the three months prior to treatment averaged 12.4 (SD = 7.6) drinks per episode, with an episodic peak BAC averaging 209mg% (SD = 181). Our measure of a standard drink in this project was the equivalent of 10oz. of 5% beer or .5oz pure ethanol. In other words, a typical beer would be 1.2 standard drinks and a 6 pack of beer would equal 7.2 standard drinks. Estimates of peak BACs were based on the individuals gender, weight, and time taken to consume the amounts reported. We used the BAC calculation tables from the BSCPWIN software program for these calculations.

Scores on the Michigan Alcoholism Screening Test (MAST) averaged 9.7 (SD = 8.7), indicating a moderately problematic sample.

Four repeated measures analyses of variance (AVOVAs) were used to test for differential changes over time in average drinks per week, weekly peak BAC, drinks per heaviest episode and peak episodic BAC. These analyses contain all clients who returned for at least one follow-up point. Results showed a main effect for time in three of the four areas, with significant differences in mean drinks per week, F(2,42) = 13.06, p = .000, weekly peak BAC, F(2,42) = 13.42, p = .000, and peak episodic BAC, F(2,38) = 9.48, p = .000. A change in drinks per heaviest episode was not significant. Outcomes on drinking variables of interest are presented in the following graphics. The data presented in graphics represent the 41 clients at pretreatment, the 32 who completed the 3-month follow-up and the 24 who completed the 12-month follow-up.

This study was done in a naturalistic setting and, as such, has a number of limitations. First, because of the lack of a control or comparison group, it is difficult to determine whether the changes in drinking behaviors are due solely to this intervention. Second, the 59% follow-up rate at the 12 month follow-up is about 11% below the level necessary to have significant confidence in the validity of the 12 month outcomes. This is because other research has suggested that those who are not found at follow-up tend to have worse outcomes than those who can be located. Third, some participants in this intervention were also referred to other treatments in addition to this one. This makes it difficult to determine whether the outcomes are a result of this intervention alone. Fourth, some multiple offenders were referred to this program even though such prior offenses had not been identified during the legal process.

With these provisos in mind, the data suggest a significant reduction in drinking by the DWI offenders in this program. This is true both for total consumption and peak BAC levels, the latter of which put offenders at greatest risk for DWI recidivism. Regular drinking is defined as that which the individual drinks in a typical week. Episodic drinking reflects a non-steady pattern of occasional drinking that is 5 or more drinks more than their regular pattern. Individuals can have either a steady pattern or episodic pattern alone or a combined pattern. The steady drinking pattern results suggest that the individuals in this trial become primarily abstinent. However, some continued to drink at least once in the three months prior to follow-up in a way that elevated their BACs to a range (e.g., 90mg% and 73mg% at 3 and 12 month follow-ups, respectively) that puts them at risk for DWI. Nevertheless, their steady patterns of heavy drinking virtually cease to exist and their episodic peak BACs decline to a level less than one third that of their pretreatment levels. These are clinically significant outcomes and are similar to those seen in controlled clinical trials in the treatment literature.

Given the impact of this treatment protocol on the drinking behaviors, we recommend that it be continued and made a routine part of the DWI screening and treatment process. Unfortunately, just the opposite has happened. This project has not received a new referral since May,1998 in spite of several meetings with the DWI Coordinator and Court personnel.

We also recommend that probation periods be extended to encompass the entire follow-up period of 12 months. This will improve the follow-up rates and provide a better picture of the longer-term outcomes of this treatment. There is therapeutic value in follow-up.