Abstract
Diversion programs connect individuals to treatment instead of criminal sanctions and are used more frequently to address substance use. This systematic scoping review included 31 published empirical studies to examine evaluations of diversion programs that address substance use and misuse. We assessed the program characteristics and implementation settings, as well as the outcomes measured and study methods used. We identified five prevalent study outcomes categories: (a) Service utilization (program retention and completion), (b) Recidivism/criminality, (c) Substance use, (d) Psychosocial functioning, and (e) Other outcomes (e.g., trauma, quality of life, and housing). Findings demonstrated limited but promising evidence for improvements in recidivism, substance use, and psychosocial outcomes. The limited evidence may be partly due to heterogeneity in diversion program implementation, study designs used, and differences among subpopulations. Most studies examined postbooking diversion programs; therefore, further examination of alternative approaches such as prebooking diversion programs is needed to fully understand diversion program impacts.
Drug and alcohol use contribute substantially to global disease burden (Degenhardt et al., 2018). In the United States, estimates from the 2018 National Survey on Drug Use and Health indicated that 19.5% of people aged 12 years and older used illicit drugs in the past year and 7.4% of people aged 12 years and older met the criteria for a substance use disorder (SUD; Substance Abuse and Mental Health Services Administration, 2019). In 2018, the most used illicit substance was marijuana, with 15.9% of people aged 12 years and older using it; however, opioid misuse was the second most reported (3.7% of people aged 12 years and older). The vast majority of opioid users were misusing prescription pain relievers (97.1%), compared with 8% using heroin. The surging opioid epidemic in the past decade has resulted in an increase in opioid-related deaths (Rudd et al., 2016). It has also negatively affected children and families (Brundage & Levine, 2019) and has strained community health and social systems (Crowley et al., 2019; Florence et al., 2016; Guy et al., 2018; Hagemeier, 2018). As families and communities battle the devastation produced by this public health crisis, the National Institute on Drug Abuse (NIDA) has called for improving access to treatment and recovery services (2020).
Criminal justice is among the important social systems impacted by the opioid epidemic (Zajac et al., 2019). Alternative strategies to divert low-level drug offenders—including opioid misuse—from long-term incarceration to reduce economic, health, and social burdens have expanded in recent years (Caulkins et al., 2020). Among these alternative strategies are diversion programs that redirect people who use illicit drugs toward treatment rather than jail or other criminal sanctions (Albert et al., 2011). Substance use diversion programs stem from those that were first used to redirect individuals with mental illness toward an alternative intervention in place of criminal sanctions (Compton et al., 2017; Hartford et al., 2006). Studies have shown that mental health diversion programs can help to lessen pressure on the criminal justice system by reducing the overall number of arrests and have the potential to decrease jail costs and officer engagement expenditures (Shapiro et al., 2015). In some of those programs, police officers were integrated to help prevent and manage drug overdose challenges among persons with mental health disorders (Compton et al., 2008; Fisher et al., 2006; Pinals, 2017).
Diversion programs generally fall into two categories: (a) postbooking programs that take place after an arrest has occurred, and (b) prebooking programs such as those that occur at pre-arrest or post-overdose (Case et al., 2009; Lattimore et al., 2003). Postbooking programs have been used longer and vary based on the stage in which they are administered in the criminal justice process and by whom they are operated (e.g., jail-based vs. court-based; Broner et al., 2002). Drug courts are not the only type of postbooking programs that have been developed; however, they are prominently featured and cited as a distinct postbooking effort aimed at diverting low-level drug-involved offenders from incarceration, so that the disease of addiction can be treated in a community-based setting rather than placement in the criminal justice system (Caulkins et al., 2020).
Prebooking programs are also on the rise as a means to help individuals who have substance use and behavioral health needs to obtain treatment and prevent future entry into the criminal justice system. A number of terms have been applied to prebooking prevention and intervention programs, such as pre-arrest diversion and deflection (Charlier, 2019). Whereas these terms have previously been used interchangeably, definitions and distinctions are being advanced in the field. For example, Charlier and Reichert (2020) have proposed a five-pathway typology for deflection for people with SUDs, mental health disorders, or co-occurring disorders based on how contact with the individual is initiated. Conversely, Kopak and Gleicher (2020) distinguished between the two terms with deflection referring to citizen-initiated contacts and pre-arrest diversion referring to police-led contacts. Regardless of the approach taken, the goal of prebooking programs is to connect individuals to treatment, reduce crime, improve public safety, and improve police–community relations (Charlier, 2017). More notably, research on deflection and pre-arrest diversion is still in the early stages (Kopak & Gleicher, 2020).
Involvement in the criminal justice system increases with the intensity of opioid use (Winkelman et al., 2018). With recent increases in legislation and funding support, it is clear that new innovations in providing treatment and overdose prevention services are needed and evaluation of these new practices and programs are warranted (Caulkins et al., 2020). Although there are a number of literature reviews that focus on the evidence associated with mental health diversion programs (Hartford et al., 2006; Loveland & Boyle, 2007; Shapiro et al., 2015; Sirotich, 2009), very few have focused on substance use diversion programs. In one of the most comprehensive reviews, Harvey and colleagues (2007) examined studies of diversion and aftercare programs of adult drug-involved offenders. Other reviews have focused on the effectiveness of opioid maintenance treatment in prison settings (Hedrich et al., 2012), community-based opioid overdose prevention and naloxone distribution programs (Clark et al., 2014), peer recovery supports (Bassuk et al., 2016; Reif et al., 2014), and post opioid-overdose interventions (Bagley et al., 2019). Despite the contributions of these reviews to the fields of criminal justice, substance use treatment, and public health, there is a need for an updated review of the evidence across the peer-reviewed substance use diversion program literature, including examination of pre- and postbooking programs.
The current scoping review was conducted as part of an evaluation of the Quick Response Team (QRT) program. QRT was initially described as a naloxone-plus pre-arrest diversion program. Charlier and Reichert (2020) classified QRTs as a naloxone-plus deflection program after the review was completed. Therefore, only diversion programs were included in the review to inform the evaluation of QRT. This scoping review was conducted systematically with a focus on peer-reviewed empirical studies of substance use diversion programs to examine the outcomes measured and study results, as well as to set forth potential outcome measures to be used by QRTs in their program evaluations. Scoping reviews are particularly useful for examining and mapping emerging evidence until more specific questions can be posed and meta-analyses of empirical studies can be conducted (Munn et al., 2018). Through this review, we aim to (a) describe the components of substance use diversion programs, (b) describe the design of studies of substance use diversion programs, and (c) examine the outcomes measured and results. The review focused on the following research questions:
Method
The scoping review followed the evidence-based preferred reporting items for systematic reviews and meta-analyses (PRISMA) to promote transparency in the review process. PRISMA guidelines are particularly essential in examining intervention studies to ensure a rigorous process that avoids bias and benefits the field (Moher et al., 2009).
Search Strategy and Eligibility Criteria
The review was conducted using PubMed, Scopus, Web of Science, ProQuest-Criminal Justice and PsycINFO databases, with keyword searches. Search terms were “diversion program” AND (“drug abuse” OR “substance abuse” OR “drug addiction” OR “substance use disorder”). We also used the search terms “diversion program” AND (“drug abuse” OR “substance abuse” OR “drug addiction” OR “substance use disorder”) AND (“cost-analysis” OR “cost-effectiveness”). These databases and search terms were selected because they produced the most relevant titles and abstracts related to diversion intervention programs. We conducted the search in July 2019 for all records published through June 30, 2019. We only included records that were published in English in peer-reviewed journals and had full-text available, which resulted in 345 records. A total of 224 records remained after duplicates were removed. Records were exported from the database searches into the Endnote bibliographical program.
Study Selection
First, a review of publication abstracts was conducted to remove those that were not empirical studies, such as systematic reviews, conference proceedings, gray literature, theses, book chapters, protocols, or study design proposals. The abstract review also focused on excluding studies that did not address drugs or alcohol. Three studies did not include an abstract but a quick scan of the article showed that they were not empirical nor related to drugs or alcohol; therefore, they were excluded during the abstract review phase and are included in the total nonempirical and non-substance use counts in Figure 1. Four additional articles did not include an abstract but a determination could not be easily made by scanning the article, so they were included in the full-text review phase. A total of 112 articles eligible for full article review remained after the abstract review.
Systematic review process.
This review focused on studies that claimed diversion program nomenclature, meaning the authors chose not to be the arbiters of what is considered diversion programming. This broad definition was employed to capture as many empirical studies as possible in the substance use and misuse field. During the full article review, studies that did not specifically focus on substance use or misuse–related diversion programs were excluded. Additional exclusion criteria included the following: studies that did not involve program evaluation or outcomes assessment, studies about the illegal diversion of prescription drugs, studies that took place in correction facilities rather than diverting individuals from them, studies that involved university-based sanctions, and studies using diversion program participants but not focused on studying the actual diversion program. Upon full-text review, an additional five articles were determined not to be empirical studies and were excluded.
Data Extraction and Synthesis
A data extraction spreadsheet was created in Excel and corresponded with a codebook created by the research team based on the goals of the review. The codebook listed each of the categories, a definition for each category, and a list of precodes for some variables (Table 1—Data Extraction Criteria). Three members of the research team were responsible for reviewing the inclusion criteria and coding every data category except for cost analysis. A fourth researcher having expertise in cost analysis studies coded that section, with a review of the cost analysis codes conducted by the lead author.
Coding Categories and Definitions.
The three members of the research team selected five shared articles to review and code independently. After each researcher coded the same five articles, the team met to review the codes, clarify coding procedures and definitions, and reach a consensus. The codebook was updated based on the first review and another 10 shared articles were selected for independent coding. The team met again to review codes, further refine the codebook, and ensure coding consistency among team members. The remaining articles were then evenly distributed among the research team to code separately. The team coded approximately 10 articles per week and met weekly to resolve questions that arose during coding until all of the study articles were fully coded. A second round of reviews was then conducted to further code the outcomes measured and study results. Uncertainties about how to classify the article information were discussed by team members until consensus was reached. After completing this process for the 112 articles that remained after the abstract review, 31 studies met the full criteria. Figure 1 contains a detailed diagram of the literature search process.
Results
A total of 31 articles were included in the review and data synthesis (see Appendix A). Publication dates were not limited to provide a full review of empirical studies to date. Studies ranged from 1997 to 2019, with 65% of studies (n = 20) being published in 2009 or later. The largest percentage of studies were published in journals focused on substance use and behavioral health issues (n = 15, 48%); three of the studies were in journals specifically focused on adolescent substance use and behavioral health. Criminal justice–related journals were the next most represented in the studies (n = 10, 32%). These were followed by journals related to psychology, psychiatry, counseling, and social work disciplines (n = 3, 10%); law or general legal issues (n = 2, 6%); and evaluation or research methods (n = 1, 3%).
Program Characteristics
Most of the included studies were conducted on U.S. diversion programs (77%) and they were predominantly conducted on programs in urban areas or in statewide and multi-jurisdictional programs (68%). Only 13% of studies were conducted on programs in rural and suburban areas. Geographic setting beyond the country of origin was not stated or identifiable in 19% of studies.
The review of substance use diversion program components showed that most studies examined programs focused on postbooking diversion (n = 27, 87%; Table 2). Only two studies (6%) exclusively focused on prebooking diversion programs and two (6%) more examined a combination of pre- and postbooking approaches. Most of the diversion programs described in the included studies addressed all types of substances (n = 26, 84%). Only five programs (16%) targeted a specific type of substance, with two of those five programs focused on the use of crack, cocaine, and opiates. The majority of studies focused on programs serving adult populations (n = 19, 61%), including the general population of substance-using adults as well as specific populations such as first-time offenders and prostitution offenders. Adults with co-occurring mental illness and substance use were reported in four studies (13%). The remaining eight studies (26%) focused on youth and/or emerging adults through 21 years of age. None of the included studies discussed the use of peer supports as part of diversion programming. Furthermore, only two studies (6%) explicitly reported the use of medication-assisted treatment.
Program Characteristics.
Some studies used more than one of the identified categories; therefore, totals may not equal 100%.
Study Characteristics
Study design characteristics were examined to provide an overview of the current state of substance use diversion programs and to inform future outcome studies in the field (Table 3). Eleven studies (35%) did not indicate a specific study design. Approximately, half of the studies included a comparison group (n = 16, 51%) and only three studies (10%) included a control group. The remaining studies had no comparison or control group (n = 12, 39%) and only examined data collected about participants in the diversion program being studied. Additional information about design characteristics across the studies included in the review are displayed in Appendix. A
Study Characteristics.
Some studies used more than one of the identified categories; therefore, totals may not equal 100%. bIncludes studies that examined service utilization as an outcome, predictor, or a factor in other outcomes measured.
The majority of studies (n = 28, 90%) used criminal justice, law enforcement, and clinical administrative records to measure participant outcomes. The second most used quantitative data collection method was participant self-report instruments (n = 13, 42%), such as pre/postquestionnaires. Eleven (35%) of the studies incorporated more than one quantitative data collection method. Furthermore, seven studies (23%) also incorporated other methods, such as urine, hair, and breath tests (n = 2, 6%); mental health assessments (n = 2, 6%); and qualitative methods to triangulate with quantitative results, such as interviews, focus groups, observations, and document review (n = 4, 13%). The majority of studies included follow-up data collection (n = 27, 87%), with the most reported follow-up time frame being 12 months postbaseline (n = 13, 42% of follow-up studies), which was followed by the 3- to 6-month time frame (n = 8, 27% of follow-up studies). Eleven (41%) of the 27 studies that included follow-up data collection incorporated more than one follow-up point to examine program effects longitudinally.
Reported Study Results
Service utilization
A majority of the included studies (n = 19, 61%) examined service utilization in some fashion. The largest number of studies (n = 11, 35%) looked at service utilization as a predictor of participant success or completion of the program and seven studies (23%) looked at the effect of service utilization on other program outcomes, such as recidivism or substance use. Participant service utilization and retention in the diversion program was examined as an outcome in four of the studies (13%), which is described in detail in the following section. Two additional studies (6%) examined utilization of related community services, such as emergency departments and mental health therapy, as an outcome. Among the 19 studies that examined service utilization, four (13%) studied it in more than one of the ways reported above.
Program Outcomes
Program outcomes were coded for each of the included articles based on what was reported by the authors of each study to provide a structured understanding of what evaluative studies measured and what outcomes evidence was produced. Preliminary coding by the research team identified five prevalent outcomes categories: (a) Service utilization (includes program retention and completion), (b) Recidivism/criminality, (c) Substance use, (d) Psychosocial functioning, and (e) Other outcomes (e.g., trauma, quality of life, housing, and employment). Although the purpose of this scoping review was not to conduct meta-analyses of included studies, during a second round of coding the research team indicated categorically whether the results, as reported by the study authors, indicated that the desired outcome was achieved (yes/no/partial). An outcome result was coded as “partial” if the desired outcome was met for only certain subpopulations or if the findings were statistically significant but the effect size was small to moderate. Among the studies in which at least one outcome was partially met (n = 13, 42%), socioeconomic and demographic factors were consistently significant factors in the measured outcomes. Factors often associated with negative outcomes included individuals who were male, were frequent drug users, were members of racial and ethnic minority groups, had lower educational attainment, and experienced a lack of social stability. The outcomes results are discussed below and displayed in Table 4.
Program Outcomes Results.
Percentages based on the number of studies that examined the outcome. bOne study unable to report findings due to a lack of available data.
The most common outcome measured across the studies was recidivism/criminality (n = 20, 65%). Seven (35%) of the 20 studies indicated that recidivism/criminality outcomes were achieved; however, an additional 45% showed partial achievement, which may indicate promising diversion program effects on this outcome. Participant psychosocial functioning (n = 10) also showed some promise, with four of the studies (40%) finding positive outcomes and an additional three studies (30%) showing partial improvements. Substance use outcomes showed the least amount of improvement, with six of the 11 studies (55%) not finding positive results. This was followed by service utilization outcomes where only half of the studies (n = 2) found positive results; however, caution should be used in interpreting this finding as there were a limited number of studies in this outcome category. Of the eight studies that measured other related outcomes, four (50%) of the studies found positive outcome results. These included outcomes related to housing, employment, income, satisfaction with quality of life, hopefulness, and perceptions of general health and well-being.
One impetus for diversion programming is to alleviate costs within criminal justice and related systems. In addition to program outcomes based on individual participant progress, we coded articles that analyzed the cost-effectiveness of the diversion program. Costs, economic utilization, or cost-effectiveness outcomes were assessed in only three (10%) studies. Two of the studies noted higher overall measured costs attributable to the increased mental health and substance use treatment resources expended in the diversion program (Collins et al., 2019; Cowell et al., 2004). However, one study (Vander Kooi, 2015) failed to report economic implications due to the lack of available data. Short study time periods (≤12 months) were a limiting factor to understand whether the higher upfront costs would result in lower overall costs.
Discussion
This scoping review synthesized the existing peer-reviewed evaluative literature on diversion programs that serve individuals who use or misuse substances. It was based on programs and studies that claim the “diversion program” nomenclature despite debates about what is and is not considered diversion programming. The review was conducted to inform an evaluation of a local QRT program, but may also provide important information for evaluation and implementation of other similar diversion programs and deflection programs.
Review findings indicated a great deal of heterogeneity in the implementation of diversion programs, even within multisite studies (Broner et al., 2004; Cowell et al., 2004), and the majority of studies were postbooking programs. This variation may have contributed to the limited evidence of the overall effectiveness of diversion programs on improving recidivism, substance use, and psychosocial outcomes. Whereas some studies showed improvements in individual outcomes (Clifasefi et al., 2017; Evans et al., 2009; Frisman et al., 2006; Kretschmar et al., 2016; Pollard et al., 2007), others suggested no differences between treatment and nontreatment groups (Dembo et al., 2006; Martire & Larney, 2011; Shafer et al., 2004). Nearly half of the studies also showed only partially positive effects on at least one outcome that was measured, which means that although the program may show promise in achieving the desired outcome, it did not consistently do so among all participant groups. In fact, the studies that examined the effects of key demographic, social, and substance use constructs provided the most conclusive evidence. These studies consistently found that diversion programming did not improve outcomes among individuals who were male, frequent drug users, members of racial and ethnic minority groups, had lower educational attainment, or experienced a lack of social stability (Broussard, 2012; Butzin et al., 2002; Clark et al., 2013; Kretschmar et al., 2016; Savage et al., 2015). A few studies showed that participant treatment completion was associated with positive outcomes (Evans et al., 2009; Mendoza et al., 2018; Patra et al., 2010), but many programs struggled with maintaining participant involvement in treatment over time.
Individuals needing support services associated with diversion programs were likely to have a number of social and physical challenges that take months or years to present and may go beyond the study periods represented in the included articles. Proof of efficacy may also be difficult in studies limited to comparisons of subpopulations within the program of study. The limited number of studies that examined individuals randomly assigned to treatment and control groups may be part of the reason why evidence for diversion programs serving individuals who use or misuse substances is inconclusive. Furthermore, as programs emerge within communities that are searching for an immediate response to growing substance use problems, a lack of a priori evaluation of designs may mean that these programs develop practices that are not supported by research and implementation science. Quick adoption and adaptation of untested programs could further result in program saturation with limited benefits. This review underscores that there are multiple opportunities to address previous limitations to continue moving the field of substance use diversion programming forward.
Implications for the Field and Research
To provide more evidence on the effectiveness of substance use diversion programs, additional rigorous experimental and quasi-experimental studies that collect participant data over longer periods of time with adequate sample sizes are warranted. However, to conduct these studies, it is imperative to have standardized outcome indicators that can be used across studies of substance use diversion programs. In this review, we identified outcomes that have been used in diversion evaluation studies to date. Although the studies differed in specific outcomes that were assessed, they overlapped enough to develop and review five key outcome categories (recidivism, service utilization, substance use, psychosocial functioning, and other). This succinct overview was useful as we considered aligning our QRT evaluation outcomes to what has been used and valued within the field, but we think it also provides an opportunity to consider additional outcomes that reflect the full range of potential benefits that diversion programs may have on individuals who use or misuse substances. For example, we were surprised by the limited number of studies that examined changes in substance use over time. Although reductions in arrest and recidivism may be the goal of diversion programs, decreased substance use is a key medium-term outcome that is hypothesized to lead to long-term effects. If substance use or misuse is considered a factor in why individuals come into contact with the criminal justice system, it would be useful for more program evaluations to examine connections between substance use and recidivism. In addition, researchers studying diversion programs should consider incorporating clinical records from the behavioral health providers who are responsible for delivering treatment. This may provide important evidence about linkages between arrest or recidivism and common behavioral health outcomes, such as psychosocial functioning, education and employment, housing stability, quality of life, and self-efficacy.
In making these recommendations, we acknowledge that others have considered the adoption of common outcome measures to build consistency in program evaluations. The premise of using common definitions of measurement parameters and outcomes is to allow for the pooling of data from multiple studies to better understand the overall effectiveness of an intervention approach, which may further funding requests. One example is a free resource developed by the Police, Treatment, and Community Collaborative (PTACC), an organization formed to leverage expertise across disciplines in the implementation and evaluation of diversion programs (PTACC, 2020). PTACC has created a list of outcome definitions and measurement methods for outcomes specific to treatment, services, law enforcement, and community. Even if the PTACC outcomes and methods are used in future studies, study authors should explore whether additional outcomes or measurement methods are needed to continue informing the field.
Very few studies were identified that focused on prebooking programs that occurred pre-arrest, pre-charge, or post-overdose, which may reflect findings from a recent study of local criminal justice practices that these practices are just starting to be implemented in communities (Lattimore et al., 2020). However, as local policing and criminal justice innovations begin to permeate the field, such as the post opioid-overdose interventions described by Bagley and colleagues (2019), a new set of research is needed to understand what is and is not effective. Moreover, the current mixed evidence in diversion programming may not necessarily reflect the full range of outcomes associated with prebooking programs. For example, two prebooking studies included in our review showed promise in reducing arrest rates, jail bookings, and time in jail, as well as lowered odds of prison incarceration (Clifasefi et al., 2017; Collins et al., 2019). The participants also showed significant reductions in legal costs. Additional development and study of prebooking programs, program participant utilization, and the effect on arrest rates in addition to client-level outcomes would greatly contribute to this approach that is still in its infancy. More importantly, it may help to move from a set of promising practices to evidence-based programs.
Another potential area of development is the use of diversion programs to address specific substances or combinations thereof. Only five studies (16%) included in this review focused on the use, arrest, or treatment of a specific substance. With rising rates of opioid and methamphetamine usage, diversion programs tailored to the unique needs of these populations may require different outcomes and measurement parameters to assess effective means of intervention rather than a broad-based approach to substance use. The QRT evaluations currently being conducted by the authors of this article and by other researchers in nearby states greatly affected by the opioid epidemic may provide critical initial evidence regarding pre-arrest approaches to combat this public health crisis. Published results from studies of QRT, as well as other pre-arrest diversion and deflection programs, may also provide evidence about essential program implementation components and best practices that result in positive outcomes when addressing specific substances.
The NIDA (NIDA, National Institutes of Health, U.S. Department of Health and Human Services, 2019) reports that an estimated 65% of the U.S. prison population has an SUD and that an additional 20% were under the influence of drugs or alcohol at the time of their crime. In addition, the Federal Bureau of Prisons (2020) reported that nearly half (46%) of the current inmates are incarcerated on drug offenses, which includes a wide range of drug use, possession, and distribution charges. Diversion programs provide an alternative to treating individuals with SUDs, with the potential to reduce substance-related criminal justice costs (Logan et al., 2004; McCollister et al., 2004). However, findings from our review demonstrated an overall lack of cost analysis related to substance use diversion programs although the interest of cost reduction may be a motivation for local communities to implement diversion programming. Additional studies that examine the costs and potential cost-effectiveness in substance use diversion programming are needed in the criminal justice, health, and social sciences fields. A comparison of pre- and postbooking programs would be especially useful for determining the best point at which to intervene with diversion.
Limitations
The studies included in this review may not necessarily include all empirical studies on diversion programs despite our efforts to use broad search parameters and inclusion criteria. This review also may not include studies of substance use interventions that do not specifically claim the diversion program nomenclature but involve law enforcement, criminal justice, health care, and treatment providers. This especially includes recent increases in the use of deflection or other pre-arrest approaches that are written about in the gray literature but for which there were no peer-reviewed studies at the time of our search. In addition, we were only able to analyze the studies based on what the authors reported in the articles. Therefore, there may be additional information related to the diversion programs and cross-system partnerships that were not explicitly addressed in the review articles. We only included the primary outcome indicators and results highlighted by the study authors, but it does not mean that additional outcome indicators were not used in these diversion program studies. The pooling of individual studies and meta-analysis provides the highest form of evidence to determine the effectiveness and generalizability of public health interventions (Muellerleile & Mullen, 2006). Although this review was not designed to pool the studies for quantitative analysis, it is important to note for future research that the notable differences in outcomes, study designs, and populations may make it difficult to do so until more outcome studies are available using consistent measurement frameworks. This issue may also be mitigated by conducting separate reviews for specific populations (e.g., youth vs. adult) or types of diversion programs; however, given the purpose of our review and the limited number of articles that met the review inclusion criteria, we chose not to differentiate between these.
Conclusion
Diversion programs and studies aimed at addressing substance use and misuse have increased in the past decade and have primarily targeted the full range of substances. This review of the empirical literature synthesized the current state of evidence behind these programs. We also conducted this review so that it can be used by communities and researchers who continue to pursue diversion programming as one of many responses to the growing substance use problem. This review did not produce definitive conclusions about the effectiveness of substance use diversion programs, but this may be due in part to variations in how programs are implemented and the methodological limitations of studies conducted to date. To fully integrate diversion programming into community responses to this public health crisis, further rigorous study is needed. It is especially necessary to continue examining how programs are being implemented, the impact of these programs on client and community outcomes, and the potential cost-effectiveness of this approach over time.
Supplemental Material
sj-docx-1-jod-10.1177_00220426211000330 – Supplemental material for Diversion Programs for Individuals Who Use Substances: A Review of the Literature
Supplemental material, sj-docx-1-jod-10.1177_00220426211000330 for Diversion Programs for Individuals Who Use Substances: A Review of the Literature by Robin Lindquist-Grantz, Peter Mallow, Leah Dean, Michelle Lydenberg and Jennifer Chubinski in Journal of Drug Issues
Footnotes
Appendix
List of Review Articles Included (n = 31)*.
| Study | Study Design1,2 | Target Population | Sample Size | Methods | Follow-up Period | Outcomes/Variables Measured 3 |
|---|---|---|---|---|---|---|
| Broner, Lattimore, Cowell, & Schlenger (2004) | Quasi-experimental non-equivalent comparison group study 2 | Adults with co-occurring mental health and substance use disorders | Diverted=971; non-diverted=995 | Questionnaires | 3, 12 months | Service utilization, recidivism/criminality, psychosocial functioning, substance use |
| Broussard (2012) | Retrospective cross-section design | Past program participants | 38 | Administrative data | Graduation or termination | Service utilization |
| Brown (2011) | Matched case-cohort study design 2 | Circuit court offenders | Offenders in program=137; offenders not in program=274 | Administrative/clinical data | First subsequent offense after index offense | Recidivism/criminality |
| Butzin, Saum, & Scarpitti (2002) | Longitudinal study | First-time offenders | 116 | Administrative data | 12, 24 months | Service utilization |
| Clark, McCullumsmith, Redmond, Hardy, Waesche, Osula, & Cropsey (2013) | n/a | Felony offenders (TASC participants) | 21,419 | Administrative data | n/a | Service utilization |
| Clifasefi, Lonczak, & Collins (2017) | Single-arm within subjects study | low-level drug offenders, prostitution offenders | 176 | Administrative data | Housing, employment, income/benefits=18 months post-program referral; recidivism=6 months post-program referral | Service utilization, recidivism/criminality, other |
| Collins, Lonczak & Clifasefi (2019) | Nonequivalent-groups longitudinal quasi-experimental field trial design 2 | Low-level drug and prostitution offenders | Total=316; LEAD= 202; comparison=114 | Administrative data | n/a | Recidivism/criminality, cost-effectiveness |
| Cowell, Broner & Dupont (2004) | Observational study 2 | People with co-occurring serious mental illness and substance use or dependence disorders | 4 sites | Administrative data, questionnaires, interviews | n/a | Recidivism/criminality, substance use, psychosocial functioning, other, cost-effectiveness |
| Dembo et al. (2012) | n/a 2 | Youth drug users with no significant arrest history arrested on misdemeanor charges | 100 | Questionnaires | 3 months | Substance use, other |
| Dembo et al. (2007) | Longitudinal study | Diversion program youth | 137 | Questionnaires, urine/hair samples | 11-14 months | Recidivism/criminality, substance use, psychosocial functioning |
| Dembo, Wareham, Poythress, Cook, & Schmeidler (2006) | Longitudinal study 1 | Arrested youth (≥11 years old) | 164 | Administrative data, questionnaires | 12 months | Recidivism/criminality, psychosocial functioning |
| Downs et al. (1997) | Longitudinal study 2 | Adolescents with chronic substance use | Total=188; randomly selected=100; diversion program=88 |
Questionnaires, Administrative data | 12 months | Substance use |
| Evans, Li, & Hser (2009) | Multi-site prospective treatment outcome study 2 | Treatment dropouts | 926 | Administrative data, questionnaires, open-end questions | 3, 12 months (outcomes focus on 12-month follow-up) | Service utilization, recidivism/criminality, substance use, psychosocial functioning |
| Frisman, Lin, Sturges, Levinson, Baranoski, & Pollard (2006) | Longitudinal quasi-experimental study 2 | Criminal defendants with psychiatric disorders (most with co-occurring substance use problems) | Total=211; defendants in seven criminal courts with diversion programs = 113; defendants in five courts without diversion programs = 98 | Questionnaires | 3, 12 months | Recidivism/criminality, substance use, psychosocial functioning, other |
| Hall et al. (2009) | Experimental design – four groups 1 | Adults convicted of nonviolent drug offenses | 136 | Urine and breath samples, documented treatment plan activities, administrative data, questionnaires | 2 times per week for 26 weeks | Service utilization, substance use |
| Hoff et al. (1999) | Retrospective cohort with two groups 2 | Offenders with substance abuse disorders who committed “minor” crime | 347 | Administrative data, mental health assessment | 1 year after index arrest | Recidivism/criminality |
| Kretschmar, Butcher, Flannery, & Singer (2016) | n/a 2 | Juvenile-justice involved youth with behavioral health issues | 2545 | Diagnostic assessment, questionnaires, administrative data, surveys | Program termination, adjudications 12 months post-enrollment | Service utilization, recidivism/criminality, psychosocial functioning, other |
| Liang & Long (2013) | n/a | DUI and drug court offenders who had completed the program (females v. males) | 683 | Administrative data | n/a | Service utilization |
| Lulham (2009) | n/a 1 | Adults with an illicit drug problem who are suitable for bail | 2396 | Administrative data | 24 months | Recidivism/criminality |
| Martire & Larney (2011) | n/a | Adult defendants with a history of problematic alcohol use | 202 | Administrative data | 12 months | Service utilization, recidivism/criminality, substance use, psychosocial functioning, other |
| McSweeney et al. (2016) | Comparative design 2 | Program completers | 3319 | Administrative data | 12 months after program exit | Service utilization, recidivism/criminality |
| Mendoza, Anthony, & Roe-Sepowitz (2018) | Secondary research design | Defendants 18-20 years old (emerging adults) | 1260 | Administrative data, questionnaires | 12 months | Service utilization, recidivism/criminality, |
| Newton-Taylor et al. (2009) | n/a 2 | Non-violent criminal offenders who abuse crack/cocaine or opiates | 365 | Administrative data, interviews | First court appearance, clinical assessment, at least one return to court, graduation | Service utilization |
| Passey et al. (2006) | n/a | Participants accepted into MERIT pilot program | 238 | Administrative data, interviews | Entry, Exit, 3-9 months after exit | Service utilization |
| Patra, Gliksman, Fischer, Newton-Taylor, Belenko, Ferrari, . . . Rehm (2010) | n/a 1 | Adults who used crack, cocaine, or heroin and had not received other drug treatment in the previous 6 months | 365 | Administrative data | n/a | Service utilization, recidivism/criminality |
| Pollard, Schuster, Hsiu-Ju, & Frisman (2007) | n/a | Women admitted to diversion program | 128 | Questionnaires, focus groups, document review, observation, administrative data | 6, 12 months | Recidivism/criminality, substance use, psychosocial functioning, other |
| Rysavy et al. (2011) | n/a | Alleged offenders | 484 | Administrative data (assessment interview) | Program completion | Service utilization |
| Savage, DuRant, Brendan Clark, Nair, & Cropsey (2015) | Secondary data analysis 2 | Adolescent mothers (ages 16-21) | Total=1080; Mothers=403 | Administrative data | Program completion | Service utilization |
| Shafer, Arthur, & Franczak (2004) | Quasi-experimental, longitudinal panel study 2 | Individuals with co-occurring SMI and SUD who were booked on misdemeanor or low-level felony charges | 248 | Questionnaires, administrative and clinical data | 3, 12 months | Service utilization, recidivism/criminality, substance use, psychosocial functioning, other |
| Shaffer et al.(2009) | Non-equivalent control group design 2 |
Female drug court participants | Total=171; High-risk drug court=91; Probationers=80 | Administrative data (courts) | 24 months | Recidivism/criminality |
| Vander Kooi (2015) | One-group ex post facto design | Juveniles arrested on drug charges and their parents | 160 | Administrative data | During program, graduation, 30 days post-graduation, 90 days post-graduation, 12 months post-graduation | Service utilization, recidivism/criminality, cost-effectiveness |
Study information listed based on descriptions provided by study authors; “n/a” indicates no information was explicitly stated by the authors.
Study used control group.
Study used comparison group.
Includes studies that examined service utilization as an outcome, predictor, or a factor in other outcomes measured.
Acknowledgements
Special thanks to the University of Cincinnati Institute for Policy Research Co-Directors (Kimberly Downing, PhD and Eric Rademacher, PhD) and graduate students (Morgan Grever, Stephen Diko, and Caroline George) who provided support for this review. The authors also acknowledge and thank Interact for Health staff members, Sonya Carrico and Lisa Myers, who support substance use–related projects and informed the development of this review.
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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Interact for Health, an independent foundation that serves 20 counties in Ohio, Kentucky, and Indiana.
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
