Outcomes of a Court-Based Substance Use Disorder Diversion Pilot Program: A Cost Analysis

Abstract

This research presents the outcomes of a substance use diversion program in a Midwestern county. Specifically, this work assesses the efficacy (operationalized through re-arrest) and cost-efficacy (operationalized through cost per 1% reduction in re-arrest likelihood) of the program across varying levels of American Society of Addiction Medicine levels of care (ASAM LOC). Using a sample of 430 program participants, findings illustrate that the program overall reduced the likelihood of re-arrest over both a 6- and 12-month horizon across program completers. However, more granular analyses revealed differences across ASAM LOC; the most statistically robust form of treatment was the most intensive 3.5 ASAM LOC, which led to an 82% reduction in 6-month re-arrest and an 80% reduction in 12-month re-arrest for those who completed the program relative to those who did not. The 3.5 ASAM LOC was also the most cost-effective among those explored, costing $66.04 per 1% reduction in recidivism. The outcomes of this work emphasize (a) the relevance of the ASAM LOC in determining program efficacy and (b) the need for varying levels of treatment intensity to better accommodate the needs of substance use disorder diversion participants.

Keywords

substance use disorder diversion cost analysis ASAM LOC

Introduction

Recent estimates have determined that up to 65% of incarcerated individuals have an active substance use disorder (SUD), and an additional 20% were incarcerated for a crime involving drugs or drug use (National Institutes of Drug Abuse [NIDA, 2020]). As such, individuals actively using substances constitute a substantial portion of corrections and are disproportionately represented throughout the criminal justice system in the United States. To address this, local jurisdictions have implemented various diversion programs where justice-involved individuals with SUDs are given an opportunity to substitute prosecution and incarceration for mandatory treatment participation. Drug court programs serve as an example of post-booking programs that utilize community treatment and intensive oversight to decrease confinement and stem the frequency of contacts with the police and carceral institutions for individuals with SUD.¹

The success of diversion programs for those with SUDs depends on the efforts and coordination of multiple corrections agencies, treatment service providers, and local and/or state court systems. Given the large investment of community resources into this effort, salient concerns include whether the outcomes of diversion programs are meaningful relative to their costs. Although there is some evidence that drug courts produce some cost savings (Carey & Finigan, 2004; Crumpton et al., 2004; Marchand et al., 2006), there are few published long-term cost analyses that assess the efficacy of treatment across varying levels of American Society of Addiction Medicine levels of care (ASAM LOC). This work supplements existing examinations of the cost efficacy of drug treatment programs by exploring a court-based pilot program in a Midwestern county. This particular program is unique in that it engages individuals across multiple justice-system contexts, and prescribes treatment using an ASAM LOC methodology, in direct contrast to the “one size fits all” approach applied by many drug courts. Findings of this work will lend insight into the efficacy of a novel program that emphasizes appropriate and tailored treatment instead of increased supervision and identifies the cost-efficacy by level of ASAM LOC to illuminate the benefits of varying levels of treatment intensity.

Literature Review

Cost-Effectiveness Methodology

Measuring the cost-effectiveness of substance use treatment programs is complicated by the fact that there is considerable variability in substances of use, history, trajectory of use, and outcome measures for corresponding treatments. Nonetheless, there is considerable literature that has been dedicated to measuring both the durability and costs of positive treatment outcomes (Brown, 2010; French et al., 2008; Mark et al., 2020). Sindelar and colleagues (2004) offer an explanation as to why studies may differ in findings of cost-effectiveness analysis in substance use treatment; specifically, there is little convergence on what single criteria would represent effective treatment. For example, studies have relied upon abstinence, reduced drug use, criminal desistence, and employment among other measures to determine effective treatment (Laudet & Stanick, 2010; Roebuck et al., 2003; Walton & Hall, 2016). Moreover, there is a growing body of literature that is critical of criminal justice outcomes like recidivism as an outcome measure for any form of criminal justice rehabilitation program (Butts & Schraldi, 2022; National Academies of Sciences, Engineering, and Medicine, 2022; Demleitner, 2020).

Appropriate Treatment

There are several ways to match patients to treatments for SUD and their comorbidities. A patient can be fitted to treatment based on clinical variables that can be paired with specific types of counseling. For example, traumas or psychiatric illnesses may be paired with trauma-informed care or pharmacotherapies (Gastfriend & McLellan, 1997). In contrast, the ASAM criteria match a patient by determining a level of care (LOC) prescribed by assessing biopsychosocial measures through the application of validated tools like the Addiction Severity Index or ASI (McLellan et al., 2006; Stallvik et al., 2015; Stallvik & Gastfriend, 2014). The ASAM LOC assessment has proved effective in both outcomes and costs since its dissemination in 1991 (Gastfriend & Mee-Lee, 2003; Kampman & Jarvis, 2015). This approach compares favorably to both over-matching and under-matching. In other words, implementing the LOC suggested by ASAM criteria has proven to produce more effective treatment outcomes when compared with less intensive levels of care (Stallvik et al., 2015). In addition, implementing the LOC suggested by ASAM criteria has proven to produce more effective treatment outcomes when compared to more intensive levels of care (Stallvik et al., 2015). As it is generally accepted that more intensive treatment is more expensive than less intensive, for example, inpatient versus outpatient treatment (Harrison & Asche, 1999), an overprescription for LOC would logically be less cost-effective than a lower LOC.

Cost-Effectiveness of ASAM Programs

There are multiple studies that demonstrate both outcome and effectiveness of treatment that begins with assessing the prescribed LOC through the ASAM criteria (Baker & Gastfriend, 2004; Levine et al., 2003; Stallvik & Gastfriend, 2014). In a reported developed by the New Hampshire Department of Health and Human Services (2020), high-intensity treatment (3.5 ASAM LOC) required an average expenditure of US$2,638 to achieve a one unit change in a composite score of the National Outcomes Measures (NOMS) in SUD treatment. The costs ranged according to substance used with alcohol requiring a lower expenditure for the one unit change in outcome than heroin and other opiates. In the findings of this report, the most commonly cited positive influence on program outcome was length of time spent in a high-quality, supportive care environment. This finding is also supported by a California study that examined implementation of ASAM criteria in a Medicaid demonstration project (Mark et al., 2020). The authors noted improved retention in programs that implemented ASAM criteria to determine treatment plans and begin with residential treatment. In line with this conclusion, this study evaluates the outcomes and cost-effectiveness of a substance use diversion program that incorporates ASAM methodologies into its development.

Program Components and Current Study

The subject of this work is a pilot diversion program based in a Midwestern county. There are several meaningful characteristics that differentiate this program from other established drug courts. First, the program engaged individuals coming from many parts of the CJ system, including the county’s circuit court, adult probation, and community corrections, and capitalized on strong connections across a number of community partners that specialized in the provision of recovery services. Second, the program focused on recommending services that were in line with the severity of need for each individual; specifically, based on ASAM LOC methodologies. Resultingly, this specific diversion program would be more accurately described as a court-based SUD referral program rather than a court supervision program.

Data were collected using an intent-to-treat approach that allowed for the evaluation of program outcomes and estimated cost-effectiveness. Using a set of binary logistic analyses, this work aims to determine (a) if there are significant differences in future (6- and 12-month post-treatment) recidivism outcomes (operationalized through re-arrests)² between completers and non-completers and (b) the cost efficacy (operationalized through cost per 1% reduction in re-arrest likelihood) across the three treatment settings, including the most intensive setting (3.5 ASAM LOC), a moderate intensity setting (3.1 ASAM LOC), and the least intensive (therapeutic care, or TC).

Method

Sample

From March 2018 to June 2021, a total of 483 participants were referred to the diversion program. Across the 483 participants, missing data related to prior arrests and recidivism reduced the final sample size to 430. To be considered an eligible participant for referral to this program, individuals must be at least eighteen (18) years of age; charged with a felony offense in the county of interest; under the court-ordered supervision of community corrections, diversion services, or probation; have a diagnosed SUD as outlined in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; American Psychiatric Association, 2013); and must have had prior difficulty engaging in recovery. Court program officials determine this using a combination of actuarial risk assessments, criminal background checks, prior enrollments in a problem solving courts, and face-to-face interviews with newly referred clients. Referrals to the pilot program came from the agencies listed earlier, in addition to any of the county’s other treatment courts (e.g., mental health court, veterans’ court, and recovery court). To participate in the substance use disorder diversion program, an individual must agree to all conditions of the court program, which includes a condition of clinical evaluation by treatment provider and agree to adhere to any/all treatment recommendations borne from said evaluation. If this individual quits or withdraws from their treatment, they would be considered as being in violation of their court program and are subject to resentencing.

Measures

Recidivism

Recidivism was measured through a binary variable, and recidivism rates were reported using an intent-to-treat approach. As such, any new arrest during or after participation in the diversion pilot program counts as recidivism, regardless of program completion status. Thus, if an individual was arrested after beginning treatment through the diversion program, the recidivism variable took a value of 1. Otherwise, the value of the recidivism variable was 0. The analysis employed the measure of any new arrest because this triggered an increase/change in treatment or services in the program, regardless of the level of offense. Whether new arrests are the best measure of recidivism is debated in the literature. New arrests reflect an individual’s continued involvement in criminal activity and as a proxy for other negative outcomes such as reconviction and re-incarceration.

Program Completion

To assess completion rates, we considered the reported outcomes of treatment. Individuals who completed their stay in treatment were considered “program completers,” and the completion variable took a value of 1. The completion variable took a value of 0 if an individual was prematurely removed from treatment as a result of violation, incarceration, or some other infraction.

Control Variables

In addition to the variables listed earlier, the analysis used several control variables shown to correlate with recidivism. These variables include the participant’s race/ethnicity,² gender, age, and number of prior arrests in the previous 12 months. The race/ethnicity variable was binary and took a value of 1 when the individual identified as a person of color. The gender variable was also a binary value and took a value of 1 if the individual identified as female. Age and number of prior arrests in the previous 12 months were both variables that took an integer value.

Analytic Strategy

To address the first of two research questions, we examined differences in recidivism across individuals who do and do not complete treatment across treatment levels. Six negative binomial regressions were used to estimate the effects of program completion across ASAM LOC on 12-month recidivism.

To determine program cost-effectiveness, we examined the average number of days in treatment by facility type and implementation costs. We then used a dose-response methodology to establish relationships between varying levels of treatment (i.e., the dose) and outcomes of interest (i.e., the response). To measure the respective dose, we consider the effects of the intensity of treatment and completion of treatment. To measure response, we consider the likelihood of 12-month post-program recidivism. The results presented outline the average number of days in treatment for completion, the reduction in recidivism, the implementation costs per person (based on cost per bed per day and the average number of days spent in treatment), and the 12-month cost-effectiveness ratio.

Results

Differences Across Program Completers and Non-Completers

Table 1 delineates differences in key variables across program completion status. For individuals enrolled in 3.5 ASAM LOC through the diversion program, more than 85% (n = 197; 85.28%) completed programming. Approximately 79% of individuals enrolled in 3.1 ASAM LOC completed treatment (n = 88; 79.28%), and 73% of individuals in TC completed treatment (n = 64; 72.73%). There were differences in the observed and expected frequencies across treatment levels, with individuals significantly more likely to complete 3.5 ASAM LOC than other treatment forms (χ² = 18.95, p = .044). With respect to completion, women were much less likely to complete treatment than men as approximately half of the women who engaged in treatment completed (n = 67; 55.83%), compared with 80% of men (n = 190; 81.97%). There were no differences by race or age; however, individuals who completed treatment had more arrests prior to treatment than individuals who did not complete treatment (t = 18.95, p = .048).

Table 1.

Descriptive Statistics Across Program Completers/Non-Completers.

Variables	(1)Complete	(2)Incomplete	t or χ²
Variables	n (%) or x¯ (SE)	n (%) or x¯ (SE)	t or χ²
Treatment level
3.5 ASAM LOC	197 (85.28%)	34 (14.72%)
3.1 ASAM LOC	88 (79.28%)	23 (20.72%)
TC	64 (72.73%)	24 (27.27%)
			χ² = 18.95**
Gender
Male	231 (81.34%)	53 (18.66%)
Female	81 (55.86%)	64 (44.14%)
			χ² = 25.65**
Race/Ethnicity
White	231 (70.64%)	96 (29.36%)
Racial/ethnic minority	81 (78.64%)	22 (21.36%)
			χ² = 2.17
Age	34.97 (0.64)	34.65 (1.03)
			t = -0.25
Prior arrests	0.45 (0.03)	0.40 (0.05)
			t = -0.86
6-Month recidivism
No re-arrest	288 (75.39%)	94 (24.61%)
Re-arrest	28 (58.33%)	20 (41.67%)
			χ² = 4.19*
12-month recidivism
No re-arrest	274 (75.27%)	90 (24.73%)
Re-arrest	42 (68.64%)	24 (36.36%)
			χ² = 2.69^†

Note. SE = standard error; ASAM LOC = American Society of Addiction Medicine levels of care; TC = therapeutic care.

†

p < .10. * p < .05. ** p <.01.

We also assessed differences in recidivism across completers and non-completers. A chi-square test illustrated differences at the 10% level in recidivism 6- and 12 months post-treatment across completers and non-completers (χ²= 4.19; p = .04 and χ² = 2.69; p = .09, respectively). Granular analyses across multiple treatment levels reveal a more thorough picture of recidivism outcomes across treatment intensity.

Tables 2 and 3 present the effects of program completion on recidivism likelihood through logistic regressions. Specifically, Table 2 outlines the odds ratios of each ASAM treatment level on 6-month post-program recidivism, while Table 3 outlines the odds ratios of each ASAM treatment level on 12-month post-program recidivism. Column 1 of these tables presents findings across all levels of treatment. Overall, Table 2 shows program completion across all levels of treatment resulted in a 66% reduction in the likelihood of recidivism 6-months after program completion relative to non-completers (odds ratio [OR] = 0.34; p = .01). Across all control variables, gender and prior arrests served as the greatest correlates of recidivism; women were 69% less likely to recidivate in the first six months post-program (OR = 0.31; p = .03), and individuals with a previous record of arrests were 845% more likely to be re-arrested (OR = 8.45; p < .01).

Table 2

Logit Regression, 6-Month Recidivism.

Variables	(1)	(2)	(3)	(4)
	All treatment forms	3.5 ASAM LOC	3.1 ASAM LOC	TC
	(n = 430)	(n = 231)	(n = 111)	(n = 88)
	OR (SE)	OR (SE)	OR (SE)	OR (SE)
Program completion	0.34**(0.14)	0.18** (0.11)	0.72 (0.49)	0.18^† (0.10)
Female	0.31* (0.17)	0.19* (0.12)	—	0.14 (0.19)
Racial/ ethnic minority	0.73 (0.34)	0.86 (0.44)	1.19 (0.76)	0.61 (0.57)
Age	0.98 (0.02)	0.98 (0.02)	1.00* (0.03)	0.96 (0.04)
Prior arrests	8.46**(4.07)	8.02** (4.16)	15.67* (16.72)	24.87** (30.46)
Constant	0.22 (0.19)	0.42 (0.41)	0.02* (0.03)	0.26 (0.49)

Note. Because only one woman in the sample was recommended to 3.1 ASAM LOC treatment, the gender variable was not included in this regression. ASAM LOC = American Society of Addiction Medicine levels of care; TC = therapeutic care; OR = odds ratio; SE = standard error.

†

p < .10. * p < .05. ** p < .01.

Table 3.

Logit Regression, 12-Month Recidivism.

Variables	(1)	(2)	(3)	(4)
	All treatment forms	3.5 ASAM LOC	3.1 ASAM LOC	TC
	(n = 429)	(n = 231)	(n = 110)	(n = 88)
	OR (SE)	OR (SE)	OR (SE)	OR (SE)
Program completion	0.40* (0.16)	0.20** (0.12)	0.86 (0.68)	0.15^† (0.16)
Female	0.36* (0.16)	0.25* (0.15)	—	0.19 (0.21)
Racial/ethnic minority	0.70 (0.29)	0.96 (0.45)	0.92 (0.52)	0.23 (0.22)
Age	0.96 (0.02)	0.97 (0.02)	0.98 (0.03)	0.92 (0.04)
Prior arrests	12.16**(5.44)	11.32** (5.76)	16.82* (13.23)	71.00** (90.18)
Constant	0.38 (0.30)	0.71 (0.69)	0.07* (0.08)	0.85 (1.60)

Note. Because only one woman in the sample was recommended to 3.1 ASAM treatment, the variable was not included in this regression. ASAM LOC = American Society of Addiction Medicine levels of care; TC = therapeutic care; OR = odds ratio; SE = standard error.

†

p < .10. * p < .05. ** p < .01.

These trends remained in Table 3 using 12-month recidivism as the independent variable of interest. Program completion of any level decreased the likelihood of recidivism 12 months after program participation by 60% (OR = 0.40; p = .02). Women were 64% less likely to recidivate in the first 12 months post-program (OR = 0.36; p = .02), and individuals with a previous record of arrests were twelve times more likely to be re-arrested (OR = 12.17; p < .01).

After investigating the effects of all treatment types, we next isolated the effects of each level of treatment. Column 2 of Tables 2 and 3 presents the results of 3.5 ASAM LOC treatment completion. Completion of 3.5 ASAM LOC treatment resulted in an 82% reduction in the likelihood of 6-month recidivism relative to non-completers (OR = 0.18; p = .01). Similar to the overall results, gender and prior arrests were associated with recidivism; women were 81% less likely to recidivate in the first 6 months post-program (OR = 0.19; p = .01), and individuals with a previous record of arrests were 802% more likely to be re-arrested (OR = 8.02; p < .01). These trends remained in the analysis using 12-month recidivism as the outcome variable. Program completion of any level decreased the likelihood of recidivism by 80% (OR = 0.20; p = .01). Women were 75% less likely to recidivate in the first 6 months post-program (OR = 0.25; p = .01), and individuals with a previous record of arrests were 11 times more likely to be re-arrested (OR = 11.32; p < .01).

The models that investigated the effects of 3.1 ASAM LOC treatment and TC treatment revealed differing results relative to 3.5 ASAM LOC treatment. Columns 3 and 4 of Tables 2 and 3 presents these results. Completion of 3.1 ASAM LOC treatment did not have a statistically significant effect on the likelihood of 6-month (OR = 0.72; p = .63) or 12-month recidivism (OR = 1.12; p = .86). However, completion of TC programming decreased likelihood of re-arrest 6-months post program by 82% (OR = 0.18; p = .07) and 85% 12-months post program (OR = 0.15; p = .10).

In light of these results, there are two important notes: first, because only one woman was assigned to 3.1 ASAM LOC treatment, this variable was not included in the 3.1 ASAM LOC model. Second, prior arrests were highly related to recidivism for individuals who participated in TC treatment; individuals with a prior record of arrests were 25 times (OR = 24.87; p < .01) and 71 times (OR = 71.00; p < .01) more likely to be re-arrested 6- and 12 months after treatment in the sample of individuals who participated in TC treatment. This could indicate that a more relaxed TC treatment leads to more opportunities for recidivism for those with prior arrests, as a less intensive treatment program may lead to greater opportunities to slip into old habits.

Cost-Efficacy Across Treatment Levels

In light of the meaningful differences in efficacy across treatment outcomes, we also considered the cost efficacy of programming across treatment levels. Table 4 displays the results of the cost analysis across the three treatment modalities. The following section identifies differences in cost-efficacy across treatment levels based on the average number of days in treatment, annual implementation costs, and the 12-month reduction in recidivism. The resulting metric of cost efficacy is the cost per % reduction in recidivism for a client to complete at each treatment level.

Table 4.

Cost Analysis by Treatment Level.

Treatment level	(1)	(2)	(3)	(4)
Treatment level	Average # of days in treatment	Implementation costs (annual)	12-month recidivism reduction (%)	Cost per % reduction in recidivism
3.5 ASAM LOC	23.67	$5,283.85	80	$66.04
3.1 ASAM LOC	26.13	$4,272.25	14	$305.16
TC	113.22	$7,381.94	85	$78.52

Note. ASAM LOC = American Society of Addiction Medicine levels of care; TC = therapeutic care.

For the highest intensity treatment, the average number of days in treatment was approximately 24 days, and the implementation cost was US$5,283.85. Those that completed the residential treatment program were 80% less likely to recidivate relative to those who did not complete it. The 12-month cost-effectiveness ratio for high-intensity, clinically managed residential care was $66.04, which translates to US$66.04 per 1% reduction in recidivism.

At the 3.1 ASAM LOC, the average number of days in treatment was 26, and the implementation cost was $4,272.25 for a treatment regimen completion. Those that completed the residential treatment program were 14% less likely to recidivate relative to those who did not complete treatment. The 12-month cost-effectiveness ratio for high-intensity, clinically managed residential care was US$305.16, which translates to US$305.16 spent for a 1% reduction in recidivism.

For TC treatment, the average number of days in treatment was 113, and the implementation cost was approximately US$7,381.94 for a completed stay in TC. Those that completed the residential treatment program were 85% less likely to recidivate relative to those who did not complete it. The 12-month cost-effectiveness ratio for high-intensity, clinically managed residential care was $78.52, which translates to US$78.52 spent for a 1% reduction in recidivism.

Discussion

Diversion programs are an important part of the growing movement in the criminal justice system aimed at providing alternatives to incarceration for substance and justice-involved individuals. Previous research has shown court-based programs, and drug courts in particular, have the potential to lower costs relative to incarceration (Downey & Roman, 2010; Hiller et al., 2021; Trood et al., 2022), although they often present negative externalities including increased supervision and, consequently, an increased likelihood of technical violations (Hamilton, 2010).

This analysis builds on the previous literature by exploring the effectiveness and cost-effectiveness of a pilot drug diversion program that focuses on appropriate treatment provision rather than heightened supervision. The advantage of our study lies in both the long-term nature of the study and the multiple treatment regimens employed in this pilot program. Individuals were enrolled in one of three programs, ranging in program length and treatment intensity. The shortest, most intensive treatment program (3.5 ASAM LOC) takes just 24 days to complete on average, while the longest, least intensive program (TC) spans 113 days on average. We not only explore the reduction in recidivism associated with each of the three diversion programs at the 12-month level but also the cost-effectiveness of each program in terms of dollars spent per 1% reduction in recidivism. As programs of this nature are difficult to implement and require buy-in from a number of community stakeholders, the cost-effectiveness of treatment is of primary interest when evaluating the success of a given program.

We find that this particular pilot program has a meaningful ability to reduce recidivism over both a 6- and 12-month horizon for individuals who complete treatment relative to those who do not. The most statistically robust form of treatment is the most intensive 3.5 ASAM LOC, leading to an 82% reduction in 6-month recidivism and an 80% reduction in 12-month recidivism relative to those who fail to complete the program at the .01% significance level. Furthermore, due to the short time span of this treatment type (24 days vs. 26 days vs. 113 days) and its high efficacy, this program is the most cost-effective among those explored in the 4-year substance use diversion program, costing just US$66.04 per 1% reduction in recidivism. The 3.1 ASAM LOC has both a smaller effect size (28% reduction in 6-month and 14% reduction in 12-month recidivism) and fails to meet the standard for statistical significance, even at the 10% significance level.

The TC program is the least intensive; however, its impact on recidivism reduction (83% over 6 months and 85% over 12 months) is greater than the 3.5 ASAM LOC, albeit less tightly identified as results are significant at the 10% significance level. However, the 3.5 ASAM LOC is clearly the most efficacious, as the potential increase in recidivism reduction from the TC program comes at a cost of an additional US$12.48 per 1% reduction in recidivism, a 19% increase in cost relative to the 3.5 ASAM LOC. Thus, it appears that short-intensive treatment programs have an advantage over other forms of treatment as (a) they have the highest completion rates among the three treatment options and (b) they have the highest cost-adjusted impact on reducing recidivism for individuals who complete the program.

The ability to evaluate cost outcomes over a 12-month horizon, along with the unique attributes of this court-based diversion program, separates our study from previous investigations of drug court programs. We find that the ASAM LOC provides a unique framework to examine the efficacy of diversion programming.

In light of these findings, there are several limitations of the research. First, we acknowledge that recidivism, and particularly re-arrests, presents a flawed measure of discontinuation of substance use (National Academies of Sciences, Engineering, and Medicine, 2022). Specifically, using re-arrests as a measure of recidivism is flawed, as arrests and convictions are not the same. Furthermore, we lack a level of specificity regarding the re-arrest—for example, a low-level misdemeanor or even a technical violation of supervision or a warrant for failing to appear for a court date may have different implications regarding the cost savings of the intervention relative to more serious crimes. Finally, an arrest or conviction for criminal activity involving substance use does not definitively determine that substance use treatment is a necessary or helpful intervention. Many people use substances and people recover without programming (Granfield & Cloud, 2001). As a result, we would like to emphasize that the results of this research should be considered with this caveat in mind.

In addition, there were methodological concerns with the present analysis that may have affected our results. We are unable to establish causality using current methods of this research. Ideally, groups would be randomly assigned, so a third group of individuals who were not offered any form of treatment could serve as the baseline against which outcomes associated with some treatment and successful completion of treatment could be compared. However, by measuring treatment effectiveness as we have, we are likely attenuating the true treatment effect associated with these programs as incomplete treatment should have a greater impact on recidivism reduction than no program participation at all. In addition, we did not have data to differentiate individuals who quit the program from those who were involuntarily removed. Thus, our measurements likely reflect a low-end estimate of the effectiveness of treatment on recidivism. Future research should make efforts to assess these differences in full to better understand the effects of similar programming.

The potential presence of omitted variable bias may have also impacted our results. That is, there may be factors correlated with program completion that may in turn influence recidivism. A follow-up with participants outlining the reasoning behind program non-completion could shed light on mechanisms driving our results. In particular, the non-completion rates for women were concerning, but certainly in line with treatment literature and results. The reasons for the disparity are complex and multifaceted; programs that address trauma, mental health, child care, family reunification, and stigma are treatment elements are identified as gender-responsive, and research has shown that gender-responsive care leads to greater engagement in treatment for women (Salisbury & Van Voorhis, 2009). Although this study made no attempt to identify these elements, it would be helpful to understand what structural barriers are in the way of individuals completing programs (financial need, family obligations, travel limitations) and if these same barriers coincide with drivers of re-arrest. This gap poses an avenue for future research.

Finally, these results may not be generalizable across space and time. We considered the results of a diversion effort in a Midwestern county during the emergence of the COVID-19 pandemic, which may have affected the results of the program across the dates considered. In addition, our sample size was particularly small once the analyses were delineated across ASAM level. Future work examining the cost-effectiveness outside of an urban setting in the Midwest, outside of the emergence of a global pandemic, and with a more robust sample size would help to solidify these results as a general phenomenon and not specific to this unique to this place or time period.

Conclusion

The majority of justice-involved individuals suffer from a substance use disorder, which has led to increased interest in utilizing drug courts to re-direct individuals from the carceral system. One hurdle to expanding these programs is the limited availability of resources, both social and fiscal, which are required to create successful diversion programs. As such, it is of paramount importance to implement diversion programs that are effective in improving outcomes for justice-involved individuals by reducing future recidivism which also utilize scarce resources in a cost-effective manner. This study examines a substance use diversion pilot program in a Midwestern county to understand the most efficacious means of reducing recidivism for individuals that are substance- and justice- involved.

This research provides evidence in support of several treatment regimes, with both a short, intensive program (3.5 ASAM LOC) and a longer, less intensive program (TC) leading to statistically significant reductions in recidivism for program completers relative to non-completers. However, once costs are considered, the shorter and more intensive program is more effective in terms of recidivism reduction per dollar spent. The shorter program also has the advantage of a higher completion rate over the longer program (85% vs. 73%), meaning individuals who enter this treatment regime are more likely to experience the benefits associated with this form of treatment.

In sum, the integration of cost analysis has helped to identify the efficacy of treatment in reducing recidivism across the ASAM spectrum. The findings of this work are largely positive, reflecting the importance of appropriate treatment in the context of substance use diversion programming. Drug court programs often engage in intensive supervision, which can consist of frequent court appearances, drug screenings, mandatory participation in group therapy and recovery meetings, education requirements, employment training or coaching, and parenting classes to name a few. None of these activities are a direct measure of desistance from substance use, the presumed goal. As such, diversion programs that result in increased criminal justice involvement act in opposition to the goals of effective treatment and recovery. We advocate for further research in the vein of health and health outcomes research, in line with literature supporting the treatment of SUD as a medical condition rather than a moral failing or criminal proclivity.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Ariel L. Roddy, PhD, is an assistant professor in the Department of Criminology and Criminal Justice at Northern Arizona University. Her work centers on the experiences of substance- and justice-involved individuals in their reentry and recovery, with special attention to the intersectional concerns of marginalized groups. She is Ojibwe and a member of the Sault Ste. Marie Tribe.

Tristan Nighswander, PhD, is an assistant teaching professor in the Economics Department of the W.A. Franke College of Business at Northern Arizona University. He conducts behavioral economic research in the areas of macroeconomics, inequality, and criminal justice.

Juliette Roddy, PhD, is the endowed Wurlger Chair of Criminal Justice and Behavioral Health in the Department of Criminology and Criminal Justice at Northern Arizona University and a Transformational Fellow at the Northern Arizona Regional Behavioral Health Authority Institute (NARBHA). She is an economist who conducts community-based research at the intersection of behavioral health and criminal justice. Dr. Roddy is a member of the federally recognized tribe of Sault Ste. Marie Chippewa Indians.

Bradley Ray, PhD, is a senior researcher at RTI International. He is a sociologist who conducts research at the intersection of criminal-legal and behavioral health systems and on overdose prevention through harm reduction strategies.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). American Psychiatric Publishing.

Baker

S. L.

Gastfriend

D. R.

(2004). Reliability of multidimensional substance abuse treatment matching. Journal of Addictive Diseases, 22(suppl. 1), 45–60. https://doi.org/10.1300/J069v22S01_04

Brown

R. T.

(2010). Systematic review of the impact of adult drug treatment courts. Translational Research, 155(6), 263–274. https://doi.org/10.1016/j.trsl.2010.03.001

Butts

J. A.

Schiraldi

V. N.

(2018). Recidivism reconsidered: Preserving the community justice mission of community corrections. Papers from the Executive Session on Community Corrections.

Carey

S. M.

Finigan

M. W.

(2004). A Detailed Cost Analysis in a Mature Drug Court Setting: A Cost-Benefit Evaluation of the Multnomah County Drug Court. Journal of Contemporary Criminal Justice, 20(3), 315–338. https://doi.org/10.1177/1043986204266893

Crumpton

Carey

S. M.

Finigan

M. W.

(2004). Enhancing Cost Analysis of Drug Courts: The Transactional and Institutional Cost Analysis Approach. NPC Research.

Demleitner

(2020). The U.S. Sentencing Commission’s Recidivism Studies: Myopic, Misleading, and Doubling Down on Imprisonment. Federal Sentencing Reporter. https://doi.org/10.1525/fsr.2020.33.1-2.11

Downey

P. M.

Roman

(2010). A Bayesian meta-analysis of drug court cost-effectiveness. The Urban Institute.

French

M. T.

Popovici

Tapsell

(2008). The economic costs of substance abuse treatment: Updated estimates and cost bands for program assessment and reimbursement. Journal of Substance Abuse Treatment, 35(4), 462–469. https://doi.org/10.1016/j.jsat.2007.12.008

10.

Gastfriend

D. R.

McLellan

A. T.

(1997). Treatment matching. Theoretic basis and practical implications. The Medical Clinics of North America, 81(4), 945–966. https://doi.org/10.1016/s0025-7125(05)70557-5

11.

Gastfriend

D. R.

Mee-Lee

(2003). The ASAM patient placement criteria: Context, concepts and continuing development. Journal of Addictive Diseases, 22(Suppl), 1–8.

12.

Granfield

Cloud

(2001). Social context and “Natural Recovery”: The role of social capital in the resolution of drug-associated problems. Substance Use & Misuse, 36(11), 1543–1570. https://doi.org/10.1081/JA-100106963

13.

Hamilton

(2010). Do reentry courts reduce recidivism? Results from the Harlem Parole Reentry Court. Center for Court Innovation.

14.

Harrison

P. A.

Asche

S. E.

(1999). Comparison of substance abuse treatment outcomes for inpatients and outpatients. Journal of Substance Abuse Treatment, 17(3), 207–220. https://doi.org/10.1016/S0740-5472(99)00004-5

15.

Hiller

M. L.

Belenko

Dennis

Estrada

Cain

Mackin

J. R.

Kagan

Pappacena

(2021). The impact of Juvenile Drug Treatment Courts (JDTC) implementing Federal Evidence-Based Guidelines on recidivism and substance use: Multisite Randomized Controlled Trial (RCT) and Regression Discontinuity (RDD) Designs. Health & Justice, 9(1), 1–15.

16.

Kampman

Jarvis

(2015). American Society of Addiction Medicine (ASAM) National Practice Guideline for the use of medications in the treatment of addiction involving opioid use. Journal of Addiction Medicine, 9(5), 358–367. https://doi.org/10.1097/ADM.0000000000000166

17.

Laudet

A. B.

Stanick

(2010). Predictors of motivation for abstinence at the end of outpatient substance abuse treatment. Journal of Substance Abuse Treatment, 38(4), 317–327. https://doi.org/10.1016/j.jsat.2010.01.007

18.

Levine

H. J.

Turner

Reif

Janas

Gastfriend

D. R.

(2003). Determining service variations between and within ASAM levels of care. Journal of Addictive Diseases, 22(Suppl. 1), 19–25. https://doi.org/10.1300/j069v22s01_02

19.

Marchand

Waller

Carey

S. M.

(2006). Barry county adult drug court outcome and cost evaluation. NPC Research.

20.

Mark

T. L.

Dowd

W. N.

Council

C. L.

(2020). Tracking the quality of addiction treatment over time and across states: Using the federal government’s “signs” of higher quality. RTI Press. https://doi.org/10.3768/rtipress.2020.rr.0040.2007

21.

McLellan

A. T.

Cacciola

J. C.

Alterman

A. I.

Rikoon

S. H.

Carise

(2006). The Addiction Severity Index at 25: Origins, contributions and transitions. The American Journal on Addictions, 15(2), 113–124. https://doi.org/10.1080/10550490500528316

22.

National Academies of Sciences, Engineering, and Medicine. (2022). The Limits of Recidivism: Measuring Success After Prison. Washington, DC: The National Academies Press. https://doi.org/10.17226/26459.

23.

National Institutes of Drug Abuse (NIDA) (2020, June 1). Criminal Justice DrugFacts. https://nida.nih.gov/publications/drugfacts/criminal-justice

24.

New Hampshire Department of Health and Human Services. (2020). Cost effectiveness of substance use disorder treatment in New Hampshire. https://www.dhhs.nh.gov/sites/g/files/ehbemt476/files/documents/2021-12/rsa-12-j4.pdf

25.

Roebuck

M. C.

French

M. T.

McLellan

A. T.

(2003). DATStats: Results from 85 studies using the Drug Abuse Treatment Cost Analysis Program. Journal of Substance Abuse Treatment, 25(1), 51–57. https://doi.org/10.1016/s0740-5472(03)00067-9

26.

Salisbury

E. J.

Van Voorhis

(2009). Gendered pathways: A quantitative investigation of women probationers’ paths to incarceration. Criminal Justice and Behavior, 36(6), 541–566. http://dx.doi.org.proxy1.cl.msu.edu/10.1177/0093854809334076

27.

Sindelar

J. L.

Jofre-Bonet

French

M. T.

McLellan

A. T.

(2004). Cost-effectiveness analysis of addiction treatment: Paradoxes of multiple outcomes. Drug and Alcohol Dependence, 73(1), 41–50. https://doi.org/10.1016/j.drugalcdep.2003.09.002

28.

Stallvik

Gastfriend

D. R.

(2014). Predictive and convergent validity of the ASAM criteria software in Norway. Addiction Research & Theory, 22(6), 515–523. https://doi.org/10.3109/16066359.2014.910512

29.

Stallvik

Gastfriend

D. R.

Nordahl

H. M.

(2015). Matching patients with substance use disorder to optimal level of care with the ASAM Criteria software. Journal of Substance Use, 20(6), 389–398. https://doi.org/10.3109/14659891.2014.934305

30.

Trood

M. D.

Spivak

B. L.

Ogloff

J. R. P.

(2022). The effects of judicial supervision on recidivism of offenders in Australia and New Zealand: A systematic review and meta-analysis. Psychiatry, Psychology, and Law, 29(5), 651–678. https://doi.org/10.1080/13218719.2021.1956385

31.

Walton

M. T.

Hall

M. T.

(2016). The effects of employment interventions on addiction treatment outcomes: A review of the literature. Journal of Social Work Practice in the Addictions, 16(4), 358–384. https://doi.org/10.1080/1533256X.2016.1235429