Comparing court-ordered DWI treatment approaches and recidivism reduction

¶ … Court-Ordered Treatments

Discussion of DWI Types of Offenders

Selection Criteria

Statistical significance

Timing, Implementation

Questionnaire Key Points

This paper will outline a research proposal designed to study various approaches to reducing drunken driving amongst those who are arrested and convicted on charges of DWI. Long a vexing problem for highway safety and the courts that deal with them, drunken driving has proven difficult for courts; drunken drivers are amongst the worst in recidivism rates amongst all convicted for crimes. Although auto deaths related to drunk driving have declined in the past 20 years from 24,094 (1987) to 17,602 (2006), there has been relatively little progress in the past 10 years, where drink has been associated with deaths in about 40% of all fatal traffic accidents (MADD, 2002).

Complicating DWI treatment is the fact that each state deals with DWI in different ways. There has been relatively little cross-comparison between methods tried in different states. The author proposes to complete a literature search of various court-ordered programs and their announced results, in order to determine the best approaches to be tested. This proposal will address three methods which appear to have had the best success in reducing DWI in other municipalities and states, and to apply those three methods in a major city with rigorous follow-up with the DWI offenders. The expected result is to be able to produce comparable data from all three methods (plus a control arm) which bring less than 10% statistical variation at the 95% confidence level.

The author believes that this study is both needed and original. There has been a good deal of literature about alcohol treatment and DWI, but most of the work has been based on a single treatment methodology. In those metastudies where there have been more than one method compared, the studies have generally drawn on different demographics; in many cases, the participants in the study came from different regions or even different countries. Comparing even the same methodologies, such as DWI courts, is difficult across regions and courts due to different methods of application.

In the researcher's opinion, it would be clearer to compare methods of treatment using demographics which are as comparable as possible. In addition, there needs to be uniformity of treatment and sentencing methods. To this end, it would be more valuable to have a statistically significant number of DWI offenders in each "arm" of a study, and to have uniformity of treatment as well.

Selection Criteria

This study will randomize those accused of DWI by a Clerk of the Court. He or she will be blinded to the name of the offender until after the decision is made.

Those who are found innocent will be dropped from the study.

Before being tried, those accused of DWI will be asked for the following additional info:

BAC at the time of arrest

Previous convictions or arrests for DWI (to be supplemented by an analysis of court records).

Age, sex and family situation

Whether the person has had exposure to any of the above-mentioned methods for DWI drivers. If "yes" to any of the three, they will not be included in the study

If the presumed DWI offender self-proclaims use of drugs other than alcohol, they will be dropped from the study. If they fail further drug screening tests for common opiates or other drugs, they will subsequently be dropped from the study.

Statistical significance

We will choose the City of Minneapolis for this study for three reasons: (1) they have several thousand DWI convictions per year, (2) they possess the ability to implement all four of the chosen methods of treatment, including DWI court, mandatory jail time and alcohol treatment and (3) there is a willingness on the part of local authorities to find the best methods of treatment, due to a high perceived problem with driving while intoxicated in the City and State.

In order to achieve statistical significance, it will be important to have a statistically meaningful number of DWI convicts in each cohort. A participation of 500 people in each "arm" of the study will lead to a 5% variation at two standard deviations, or in other words the 95% confidence level.

One of the major concerns of this study is compliance with the study protocols. Despite monitoring through the court process, the researchers expect a considerable fall-out due to lack of compliance. A recent study on compliance with alcohol treatment programs demonstrates that only half completed the court-ordered program; while this evidence was gathered in Finland, the author is confident that this compliance rate will be repeated in this study (Aalto, 2000).

Given the results of the above-mentioned study, it will be important to separately monitor the performance of men and women as subgroups, in order to see whether there is a significant sex-based difference in compliance and recidivism, depending on program (Lapham, 2000). There is a concern that by decreasing the subsets even further by sex, confidence levels could increase to 10%; even if these do not meet statistical relevance criteria (P

Timing, Implementation

Each of these methods will require time for monitoring. Since each method will take approximately 6 months plus an additional follow-up period of 6 months after "treatment," (except for the control group) a total of one year will be required for the adequate monitoring of each group. Although this researcher has approached the necessary authorities, it is anticipated that it will take three months from the beginning of the study in order to implement the randomization process, insure mandatory sentencing, prepare the treatment programs and alert the DWI court of potential increases in their workload.

There will be three primary measurement points: (1) intake, when the DWI offenders are entered into one of the DWI treatment cohorts and data is collected, (2) active period, where the convicted offenders are undergoing court-ordered jail sentences, DWI treatment or other treatments as prescribed by the DWI court judge, and (3) follow-up 6 months after the programs are completed, in order to assess recidivism and therefore effectiveness of the program.

After completion of the program, and after an additional 6 months, the participants in the trial will receive a mailed questionnaire with telephone follow-up. This post-test-only factorial design with both mail and telephone follow up is the most realistic way to receive feedback, as it has been found that participants in such studies tend to be more honest about their drinking and driving habits than those in face-to-face interviews.

Literature Review

The rate of recidivism differs significantly depending on the type of offender. Female offenders are generally re-arrested less often than males (Lapham, 2000). This may tie to overall death figures in driving accidents, which seem to occur more with men than women (Enterline, 1961).

DWI treatment is highly political, and highly contested. There is a considerable debate between those who feel that drunk drivers should be punished, and those that feel that they should be rehabilitated, with states like Virginia passing the death penalty for those who kill others while driving drunk (Jump, 2007).

Most literature about court and legal treatment of DWI offenders is community-based. There is a good deal of information generated in a few areas of the country where there is particular public pressure against drunk drivers. New Mexico, for example, has the highest DWI rate per capita of any state in the U.S. The Bernalillo County Court (which includes Albuquerque) has a particularly tough DWI court, patterned on work done in Minneapolis and Dade Country, Florida (NDCI, 2007). According to the National Drug Court Institute, authorities at Columbia University have found "drug courts provide the most comprehensive and effective control of drug-using offenders. (NDCI, 2007)"

Alcohol treatment centers are difficult to run, and have a relatively low "cure" rate. While these statistics are discouraging, it is clear that the combination of court orders and the program has a possibility of greater effect (Lapham S.S., 1997).

Evidence on jail time is more difficult to assess. Jail time for short periods, as little as 48 hours, appears to have an effect on those drivers who are not chronic alcoholics (NHSTA, 2007). There is a concern, based on studies in Norway and Sweden, that longer jail sentences may not have an increased effectiveness for chronic alcoholics who are convicted multiple times of DWI.

Conclusion

This study is both relevant to current needs, and compelling in that there are no studies found by the author which assess the effectiveness of the above-mentioned programs within the same region or demographic profile. The researchers' ability to work with local authorities in Minneapolis will assure that the sample selection is as close to random as possible, and the follow-up mechanisms should provide the most realistic feedback, when supplemented by additional access to court records after the study questionnaires are received.

Further statistical analysis of the data may point to tendencies between subgroups, particularly men and women, and those who are "occasional" heavy drinkers/drivers from those who are chronic alcoholics with multiple arrests and convictions for DWI.

Appendix

Questionnaire Key Points

Information for each participant

Section 1: Demographics From police report

Section 2: Previous arrest/conviction record From court records

Section 3: Self-declared alcohol habits From questionnaires before/after

Section 4: Treatment program adhesion From court records, follow-up

Section 5: Recidivism From court and police records

Section 6: Reduction in drinking/driving From questionnaires after Bibliography

Aalto, M. a. (2000). COMPLIANCE RATE and ASSOCIATED FACTORS for ENTERING an ALCOHOL BRIEF INTERVENTION TREATMENT PROGRAMME. Alcohol and Alcoholism, 372-376.