ALCOHOL AND DRUG ABUSE INSTITUTE University of Washington  Home  |   News  |  About Us  |  Research  |  Library  |  Staff  |  Publications  |  Funding  | Training |  Links  |  Search

EXECUTIVE SUMMARY

Since the mid-1980s, state and local criminal justice systems have been inundated with felony drug cases. The saturation of state court and prison systems with drug cases and offenders convicted of drug related crimes has been at the expense of court and prison resources to manage more serious, violent felony offenders. In response to this situation, in the early 90's state and local courts in several states started to create special drug courts, also known as treatment drug courts. These courts were designed to administer cases referred for judicially supervised drug treatment and rehabilitation. In exchange for the possibility of dismissed or reduced sentences, offenders who met clearly defined criteria, and who agreed to participate, were diverted to drug court programs whose main purpose was to use the authority of the court to reduce crime by changing drug related behavior (GAO, 1997).

Key Elements of Drug Court Programs: Early in the development of Drug Court programs, the Drug Court Program Office of the Department of Justice identified a set of key elements considered crucial to the program success. These elements include:

• A unified court system with a single judge providing leadership and system-wide focus;
  
• Intervention by the drug court judge and DC team members with the offender as soon as possible after arrest;
  
• Access to comprehensive treatment based on the needs of the offender.
  
• Verification of compliance with treatment and supervision through drug testing and supervision contacts;
  
• Frequent and personal contact with the drug court judge at status hearings to assure compliance with treatment and supervision goals;
  
• Immediate and consistent responses to violations through graduated sanctions and incentives; and
  
• Integration of alcohol and other drug treatment services with justice system case processing (Drug Court Program Office, 1997).
  

Drug Court Evaluation: Over the decade a substantial number of drug court evaluations have been done. Most have focused on two outcomes: rates of retention and completion, and criminal recidivism. An Office of Justice Programs (1998) report of over 200 programs found retention rates ranging from 31-100% with an average rate of more than 70%, compared to significantly lower retention rates among criminal defendants in traditional treatment programs. Completion rates ranged from 8-95% with an average completion rate of 48%. Substantial reductions in recidivism rates, ranging from 2 to 20%, were found among programs that reported recidivism. The average recidivism rate for drug related offenses in the US is approximately 45% over a two year post-release period.

Belenko (1998), in a review of Drug Court evaluations, found wide variations in the scope, methodology and quality of the studies, and made a number of recommendation to improve the quality of drug court evaluations. He recommended that future studies document the structural and operational characteristics of the programs and examine how program characteristics contribute to participant outcomes, and that more longitudinal studies be done.

Washington State Drug Courts: Between 1994 and 1999 six counties in Washington started adult Drug Court programs: King and Pierce in 1994; Spokane in 1996; Skagit in 1997; Thurston in 1998; and Kitsap in 1999 (in addition, there are now programs in Clallam, Clark, Snohomish, Whatcom, and Yakima Counties, as well as four Tribal Drug Courts.)

Although the programs vary in a number of organizational and operational details, all reflect the key program elements, and share a common goal of reducing drug related criminal activity by judicial oversight and supervision of mandated chemical dependency treatment. Federal funding has been essential to the development and operation of drug courts in Washington. However, Federal funding was designed to be time limited with the expectation that the effective programs would be continued with State, County, and perhaps municipal funding. As Washington faces the multiple policy decisions related to continuing Drug Court programs, objective information about the organizational structure and operation of the programs and their effectiveness is essential.

In 1999, the Northwest HIDTA (Northwest High Intensity Drug Trafficking Area), working through the Division of Alcohol and Substance Abuse (DASA), contracted with the Alcohol and Drug Abuse Institute at the University of Washington to conduct a three to four year, statewide evaluation of the six adult Drug Court programs. (A program profile was also completed for the Snohomish County DC in 2001). In the course of the evaluation Federal Bryne grant monies, a supplement grant from National Institute of Justice, and support from DASA were added to the Federal HIDTA funding. The evaluation was designed to (1) describe and compare the organizational structure and operational characteristics of the six programs, and (2) examine the impact of the programs on re-arrest, conviction and incarceration rates, earned income among participants, and utilization of public resources including medical, mental health, substance abuse treatment, and vocational services.

The report is in two parts: first, a summary of the organizational and operational characteristics of the six programs, second, a summary of the impact of the programs on selected outcomes.

ORGANIZATIONAL AND OPERATIONAL CHARACTERISTICS OF THE COURTS

Data Generation: Between March 1999 and June 2000 on-site interviews were conducted with key staff members in the six programs, including judges, prosecuting attorneys, public defenders, drug court coordinators, and treatment program staff members. Key informants in Snohomish County were interviewed between March and April 2001. Interview topics included program goals, developmental and implementation processes, target population, inclusion-exclusion criteria, offender movement through the program, monitoring substance use, frequency of treatment and court contacts, treatment programs, and other support services. Program documents were also reviewed. These data were analyzed to develop descriptive profiles of the structure and operational characteristics of each program.

The program profiles and a table comparing the programs on a number of characteristics, are included in the Appendices. It is important to recognize that the profiles describe each program for a specific period in time. As the programs continue to evolve and change in response to local and national needs and demands, changes may have been introduced that are not reflected in the profiles. Some important areas of commonality and difference among the programs are discussed, and a number of key issues facing the programs are identified.

Commonalties Among Drug Court Programs: While there are differences among the programs in the specifics of implementation, all have strongly incorporated most of the Drug Court Program Office key elements into the ongoing operation of their programs.

Judicial Leadership and the Drug Court Team: All have one judge assigned to the DC who is responsible for the overall leadership of the program and the DC team. The degree to which the DC judges carry a full-time judicial caseload in addition to DC varies among programs, as does the consistent assignment of additional court personnel such as a Court Clerk and docket manager. All of the programs have developed a team approach to managing the legal process. Under the leadership of the DC Judge, prosecutors, defense attorneys, and a representative(s) from the treatment system jointly review each case and develop a recommended approach. Team members in all of the programs cite the team approach as a major strength of the program.

Access to Comprehensive Treatment Based on the Needs of the Offender: All the courts have attempted to put into place a comprehensive treatment system, and ensure access for offenders in the program. By their very nature DC programs have expanded the demand for treatment in a State system that is not characterized by excess capacity. Thus, timely access to comprehensive treatment services has been a common challenge among the programs.

All of the programs use one or more treatment providers to provide a carefully outlined phased approach to treatment. Minimum frequency of treatment modalities required in each phase are specified in the DC contract, as are the criteria for successful completion and movement to the next phase. The number and length of phases vary among programs but all have very specific treatment participation requirements that are clearly defined for treatment providers and participating offenders. Intensive outpatient treatment is the primary treatment modality in all programs. The use of residential treatment, a scarcer and more expensive modality, varies among the programs but all use it when deemed appropriate. All treatment programs are expected to keep the court informed about offender compliance on a regular basis to enable the court to enforce treatment requirements and to integrate treatment services with justice system case processing , another key element of DC programs.

Monitoring: A carefully specified system of random urine and breath monitoring is a common feature of all the programs. While the frequency of monitoring and specific reporting procedures vary among the programs, all require strict adherence to the monitoring program. In addition, all have developed effective communication procedures between the Court and the monitoring agency to ensure the timely exchange of information necessary to allow the Court to act swiftly in cases of continued drug use.

Frequent and Personal Contact with the Drug Court Judge: All of the programs are characterized by frequent status hearings in which the DC Judge, in direct interaction with the offender, reviews his/her participation in the program. If the offender successfully moves through the program, status hearings become less frequent, but in all of the programs the offender is seen by the Judge at least once a month to review his/her progress in the program.

Differences Among Drug Court Programs: Although the programs are similar in many important ways, there are several areas of significant differences.

Eligibility Criteria: While eligibility criteria in all the programs reflect broad Drug Court Program Office guidelines, they are also shaped by local politics and community attitudes. The two major areas of difference are definitions of previous violent offenses, and the range of other drug related offenses that are considered eligible for DC. Federal guidelines prohibit the inclusion of individuals with a history of violent offenses, but individual programs have some discretion in identifying such offenses. A number of the programs accept individuals with previous history of misdemeanor domestic violence on a case-by-case basis, while others exclude individuals with any history of violence. Some consider only adult violent offenses while others consider juvenile offenses that included violence. Similarly, some programs consider a broad range of property crime offenses in which substance abuse played a significant role as eligible for participation in DC while other programs define eligible crimes more narrowly.

Response to Violations: One of the key elements of DC programs is immediate and consistent response to violations through the use of graduated sanction and incentives. The programs in the study vary widely in both timing and consistency of responses to violations. Several factors seem to contribute to these variations including: (1) how clearly the Drug Court team has considered their approach to sanctions and developed a clearly articulated set of graduated sanctions and violations to which the sanctions will apply; (2) the timeliness with which violations are reported to the DC and the Judge; and (3) the organization of the program that either facilitates or inhibits the enforcement of a range of sanctions. The programs that have a clearly articulated set of graduated sanctions and strong working relationships with law-enforcement are better able to use sanctions than those who do not.

Relationships with Law Enforcement: The programs vary widely in the relationships they have with law enforcement agencies. In several of the programs, law enforcement agencies, including local representatives from the Department of Corrections (DOC), were involved in program planning and are an integral part of the program. In others, there has been very little law enforcement involvement. All of the programs need the support of law enforcement in a number of ways, the most obvious of which in the bench warrant process that is invoked when an offender fails to appear for DC status hearings. Developing relationships with law enforcement is a goal of several of the programs.

Data Systems: The programs vary widely in the data systems they use to document and track the program. While a standardized data system for all of the programs has been a goal for the last three years, the system is not in place and the programs use a broad array of local systems in an attempt to meet their data needs.

COMMON ISSUES FACING THE DRUG COURT PROGRAMS
Developing a Stable and Adequate Funding Base: At the present time the Drug Court programs are funded by multiple sources. Most continue to rely heavily on Federal funding provided by planning and implementation grants from the Department of Justice/Drug Court Program Office (DOJ/DCPO) which are designed to be time limited. Other important funding sources include Federal monies from Northwest HIDTA and Bryne grants, State and County funding. County funding in most programs has been largely in the form of providing the Court personnel including the Drug Court Judge, and, in some programs, the County has provided monies for urine and breath testing. The use of Bryne grant funds is now restricted to treatment costs. As programs reach the time limits of DOJ/DCPO funding, they are increasingly dependent on State and County monies. While the State Legislature has approved some funding for specific programs in the last two years, neither the State nor the counties have yet institutionalized funding for Drug Court program. Thus, the programs exist in an uncertain financial environment. At the same time, public demands are growing for the kinds of alternatives Drug Court programs provide to deal with drug related crime.

Developing and Supporting Adequate Data Systems: Given the multi-system nature of Drug Court programs, the cross-system communication that is key to successful program functioning, and the importance of demonstrating program effectiveness, data systems that support system integration and program evaluation are essential. At the present time, most of the programs do not have the resources to support adequate data system development, and there is little to no standardization of data system across the programs. While several efforts have been made to address data system development, it remains a problem area for most of the programs.

ANALYSES OF ADMINISTRATIVE DATA SETS
The eleven data sets analyzed are:

• Arrests, from the Washington State Patrol.
  
• Superior Court filings, from the Office of the Administrator for the Courts.
  
• Time incarcerated in prisons, from the Department of Corrections.
  
• Earned income, from the Department of Employment Security.
  
• Length of time in drug courts, from the respective drug court programs.
  
• Chemical dependency treatment, from the Division of Alcohol and Substance Abuse.
  
• Vocational training services, from the Division of Vocational Rehabilitation.
  
• Mental health services, from the Division of Mental Health.
  
• Washington State death records, from the Department of Health.
  
• Medicaid claims, from the Medical Assistance Administration.
  
• Superior Court convictions, from the Office of the Administrator for the Courts.
  In addition, the two integrative analyses were performed:
  
• Analyses to determine whether there was a relationship between the amount of chemical dependence treatment received and outcomes.
  
• Analyses to examine whether graduation from the drug court program in less than a year was related to outcomes (this analysis used data only from King County).

All quantitative analyses have used participant identifying data from the courts, and archival data from state administrative data sets. No data have been collected directly from subjects. All procedures have been reviewed and approved by the DSHS Human Research Review Board. The study design has been to use the naturally occurring groups formed by program and self selection. This introduces almost certain bias in the composition of the groups. The effect of the bias is to make it difficult to determine whether group differences result from the bias or from the effects of the drug court programs. Results must be interpreted with these design limitations in mind.

METHODOLOGY
Identifying Subjects: The subjects used in most of these analyses were those identified in August-September, 2000. Where identifiers were available (which included the three oldest, largest courts), we obtained data for all offenders who had been referred to the drug court, even if they had not entered the court.
Comparing Results Across Counties. County drug court programs, and the contexts in which they operate, differ from one another in many ways, including criteria for admission, types and backgrounds of offenders admitted, retention policies, types and amounts of treatment and other interventions offered, policies and practices in dealing with violations, graduation criteria, and county arrest and conviction rates, among many others. Further, a single court can vary on the same factors across time. The effect of all the differences is to render any cross-county comparisons of sizes of results inappropriate. We report these differences as part of the descriptions of results, but we explicitly do not intend to suggest that observed differences have any implications about program merit. We do look at patterns of outcomes across counties, either relative sizes of outcomes across groups within counties, and/or patterns of changes across time for groups in each county. Given the many differences among the counties, we feel that when we find common patterns across counties, the likelihood of the pattern being a reliable result is increased.

Strengths And Weaknesses: Strengths. Relative to most other research on drug courts, this study has three major strengths:
First, it is a multi-site study. There are at least three counties with analyzable data (sometimes six counties), which allows comparisons of patterns of results across counties. If similar patterns of results occur across sites, it lends greater confidence to the results, and suggests it is reasonable to expect to see this result in additional drug courts.

Second, using data from administrative data sets, we are able to examine a substantial number of outcome measures, many of them of central interest to drug court evaluation. These include several indices of criminal justice involvement, earned income, several measures of social services received, and several measures of health and/or medical treatment system involvement. This gives us a singularly rich and multifaceted view of drug court outcome.

Finally, we have unusually long baseline and follow-up periods for most subjects (that is, data for a long period of time prior to and following referral to the courts). Further, because the data are from archival sources, they are relatively "objective," at least compared to the usual self-report data.

Weaknesses. In common with the majority of other drug court evaluations, this study also has a major weakness in comparison to generally accepted research standards: we do not have an adequate control or comparison group. When subjects are selected or self-selected into groups (rather than being randomly assigned), there is a likelihood that the groups will differ on characteristics such as motivation, intelligence, work history or skills, social support, or any of a multitude of other factors that could influence their outcomes. This makes it extremely difficult to determine whether the program or the pre-existing differences are what are leading to differences in outcomes.

While attractive for some reasons, administrative data sets also have the weakness of being limited to the data available, both in terms of the types of information available, and its accuracy and completeness. In addition, data sets may not be current with the time the analyses are being done.
Definition Of Drug Court Outcome Groups. Subjects were divided into these naturally occurring outcome groups:

• Ineligibles are persons who passed an initial legal screen and were referred to the court, but on closer examination were found to be ineligible on either legal or clinical grounds.
  
• Opt Outs are persons who met all criteria, and were offered entry to the court, but who personally declined to participate.
  
• Did Not Finish (DNF) are individuals who were admitted to a drug court program but either failed or dropped out.
  
• Graduates are individuals who graduated from a drug court.
  
• Active cases are those still involved with the drug court program. Active groups are used in some analyses, but not in others.

Outcome Vs. Process Variables: Some of the data sets that follow may be thought of as reflecting drug court outcomes, in particular, arrests, court filings, incarceration, earned income, and possibly mortality. Others give a view of the process involved in the court programs, these being the time in drug court, utilization of Chemical Dependency services, time to admission to service modalities, use of vocational rehabilitation services, and use of medical and mental health services. Vocational rehabilitation, Medicaid utilization, and possibly mental health could become outcomes if programs had goals or objectives concerning their use, but at this point seem to be mostly descriptive.

SUMMARY OF FINDINGS

Arrests

• Graduates have fewer re-arrests than any of the other outcome groups (significantly less than the Opt Outs).

Court Filings

• Only Pierce shows any statistically significant differences among outcome groups on new court filings: there Graduates, Actives and Did Not Finishes are similar, and different from Opt Outs and Ineligibles. Across counties the Graduates have the lowest post court filing rate. However, all groups show declines in filings following court contact (except Spokane, due to dating practices).

Convictions

• Offenders who graduate from drug court are less likely than offenders in any other group to be re-convicted in the three years following referral to drug court.

Prison Time

• Graduates have zero rates of imprisonment in the post drug court referral period in all counties except King. The post drug court referral imprisonment rates for Graduates in King and Pierce counties are significantly lower than the Ineligible, Opt Out and Did Not Finish group rates.

Earned Income

• Graduate groups show systematic and substantial increases in incomes, with some tail-off in the third year. Graduates are the only group to show this improvement.

Mortality

• There are no differences in mortality rates between counties or compared to general population.

Time Spent in Drug Courts

• In general the counties have similar lengths of stay for the groups. However, King County begins graduating participants in as little as 6 months, and graduates over 20% by one year. Other courts begin graduating at one year.

Chemical Dependency Treatment Service Utilization

• Not all Graduates have a TARGET record of Outpatient treatment.
  
• Counties vary in mean levels of Outpatient treatment offered to Graduates during drug court: means range from 67 to 114.
  
• Counties vary in the use of services for outcome groups both during and after drug court.

Chemical Dependency Treatment and Outcomes

• The service data included in TARGET appear to be incomplete
• Overall, based on the available data, the amount of group and individual treatment being received by drug court participants appears to be adequate.
• There is mixed evidence for a relationship between higher amounts of group and individual treatment and lower risk of re-arrest. This relationship holds for Graduates, but for DNFs the effect is reversed (more therapy is associated with higher risk for re-arrest), and there is no relationship for Opt Outs or Ineligibles.
• For Graduates but not for DNFs, there is support for a relationship between higher amounts of group and individual therapy being related to lower risk of re-conviction.
• For Graduates and DNFs there is no relationship between higher amounts of treatment and increased income.

Drug Court Length of Stay and Outcomes (King County)

• There are drug court participants who can be graduated with less than a year, even substantially less that a year, in the court program, with every reason to anticipate that their long-term outcomes will be among the best expected from the program. Whatever criteria King County is using to identify such participants seem to be effective.

Vocational Rehabilitation

• Overall rates for applications for the use of DVR services are very low, about 2% in King and Pierce counties, and 4% in Spokane County, and are similar for all groups pre court referral.
• Application rates for Graduates in King, and Graduates and Did Not Finishes in Spokane, roughly double, into the 8% to 11% range (toward the high values in Spokane).

Mental Health

• Use rates for Inpatient services are generally very low, one half percent or less., and do not differ significantly across the three counties or between outcome groups.
• Outpatient use is generally between 2% and 5%, occasionally 8% to 12%. There is little change between pre and post court referral (except in King, where there is a general upward trend).
• Outpatient use varies across counties, but not much within counties across outcome groups.

Medicaid

• In the two areas of greatest use of Medicaid services, Chemical Dependency and Medical Outpatient services, the Graduates tend to have the highest level of use, with the DNF group generally second.
• DNFs are the highest utilizers of Emergency and Medical Inpatient services.
• In the post period, the utilization percentages for Medical Outpatient are considerably higher than for Chemical Dependency services, suggesting that there is a potential additional resource for supporting drug court Chemical Dependency treatment for a sizable group, up to as many as 20%, of court participants.
• Pre-referral to drug court the quarterly percentages of utilization of Medicaid reimbursements tend to be between 0% and 2% for most categories. Use rates for Chemical Dependency and Emergency services are generally higher, up to 4%, and rates for Medical Outpatient higher yet, ranging between about 10% and 19%.
• Post drug court referral quarterly percentages show little change from pre rates, except Graduates and DNFs show increased percentages of Chemical Dependency service reimbursements (3% - 6% in King and Pierce, and 6% - 10% in Spokane) and Medical Outpatient reimbursement percentages (between 15% and 26%).
• Mean quarterly reimbursements vary widely across counties and especially across Medicaid categories.
  • Pre-referral mean quarterly Chemical Dependency reimbursements range between $0 to $20, post-referral $15 - $30 in King and Pierce, and $30 - $90 in Spokane.
    
  • Reimbursement for Emergency and Medical Inpatient services are highly variable in all three counties, tending to range between $0 and $100, with some higher values, in all three counties, pre and post-referral.
    
  • Medical Outpatient pre-referral reimbursement group means are largely between $40 and $100, while post-referral levels are between $50 - $150.

CONCLUSIONS AND RECOMMENDATIONS

• The overall pattern of results across all four outcome variables (arrests, court filings, incarceration time, and earned income, excluding mortality) and for the three drug courts with sufficient follow-up for interpretation (King, Pierce and Spokane courts) is that participants who graduate have better outcomes that those in any other group. These differences are not always statistically significant, and when they are, the Graduates are generally not significantly better than all other groups. Where statistically significant differences were found and effect sizes were calculated, these were generally small.

• In general, the gains made by the Graduates after referral to drug court were retained over a three year post-referral period, although on some measures there were weakening patterns in the third year post-referral.

• Among the other groups (DNF, Opt Out, and Ineligible) there were no consistent orderings on relative outcomes. There is some tendency for the DNFs to show the worst results, but this is not reliable across counties or outcome variables.

• The patterns of results vary in several ways across the counties. No single court shows overall superiority.

• Although Graduates as a group have significantly superior outcomes compared to the other outcome groups, membership in the graduated group is the result of a process that will automatically select a subgroup of participants with better outcomes. Thus the extent to which the drug courts are themselves producing or contributing to these improved outcomes, as opposed to them resulting from offender characteristics, is difficult to determine.

• Medicaid is underutilized as a source of funding for Chemical Dependency treatment in the drug courts.

• Appropriately chosen participants can be graduated substantially earlier than one year with no reduction in favorable outcomes. Outcomes for participants who stay longer than about 15 months, and especially longer than 26 months, become increasingly less favorable, hinting that there may be a length of stay beyond which improvement is minimal.

• Drug court participants appear to be receiving, on average, as much treatment as is useful. There does not appear to be a strong relationship between amount of treatment received and outcomes: Graduates sometimes show the expected patterns, but other groups do not. It is clear that other factors are as important for outcomes as amount of treatment, and in this context it is likely that there is a level beyond which more treatment does not improve results. That level may be lower than mean levels being provided now (although there may be other reasons for continuing treatment).

• We were not able to check on the relationships among violations, sanctions, and outcomes. The general thinking is that quick and standardized responses to violations is the most effective policy. Courts varied in practices on this matter. In those courts lacking it, a more standardized approach might be considered. A study of this issue would be desirable, but would hinge on the availability of data.

• Non-Chemical Dependency services that might be desirable for drug court participants, e.g., mental health counseling and vocational training, are not frequently used. A more aggressive approach to providing these ancillary services might improve retention and graduation rates, and subsequent outcomes. A closer examination of the relationships between these services and outcomes might be beneficial.