The Role of Controlled Substance Use in Diversion Outcomes Among Mentally Ill Offenders: A Pilot Study

Abstract

Diversion programs offer opportunities to offenders with substance abuse or mental illness to attend treatment as an alternative to incarceration. The present study identified variables associated with drug relapse and recidivism and the moderating role of substance use on recidivism in a diversion sample. Data were collected from 80 clients with psychotic disorders from a diversion program in New York City. Outcomes were examined after 6 and 12 months of program participation. Individuals who used controlled substances other than alcohol or cannabis were more likely to have a positive toxicology result than those who used alcohol or cannabis only or those with no alcohol/drug history. Individuals with schizoaffective disorder were more likely to be rearrested than individuals with other diagnoses, as were those with a violent offense (e.g., assault, robbery). Positive toxicology results were unrelated to rearrest and did not moderate recidivism, suggesting substance abuse may be only indirectly related to rearrest among diverted offenders.

Keywords

diversion diversion outcomes substance use substance abuse recidivism psychotic offenders

In recent years, the jail and prison populations in the United States have grown at an astounding rate. Between the years 1980 and 1996, there was a 200% increase in the number of federal incarcerations throughout the country, and this growth trend continued well into the 2000s (Blumstein & Beck, 1999; Bradley-Engen, Cuddeback, Gayman, Morrissey, & Mancuso, 2010). Only recently have these rates begun to decline, and even then, it has been only at the average of 0.5% per year from 2007 to 2014 (Kaeble, Glaze, Tsoutis, & Minton, 2015). This growth is particularly concerning due to the overrepresentation of individuals with mental illness in prison populations and the special needs of these individuals, which cannot always be met with the limited resources available (Bradley-Engen et al., 2010; Gunter et al., 2008; Lamb & Weinberger, 1998). Studies suggest that between 6% and 25% of prison inmates may suffer from a serious mental illness, with an even greater percentage suffering from any mental illness, and many of these individuals do not receive treatment during incarceration (Bradley-Engen et al., 2010; Fries et al., 2013; Lamb & Weinberger, 1998; Prins, 2014).

In addition to concerns about individuals with mental illness, treatment concerns also exist for the many incarcerated individuals who misuse controlled substances, which are defined for the purposes of this study as drugs or chemicals whose manufacture, possession, or use are regulated by a government, specifically, alcohol, illicit drugs, and prescription medications. Unfortunately, there is a marked overlap between these two populations. Studies suggest that up to three quarters of mentally ill incarcerated adults also have a co-occurring substance use disorder (Abram, Teplin, & McClelland, 2003; James & Glaze, 2006; Mumola & Karberg, 2006). Dually diagnosed individuals warrant special consideration, as research studies have consistently shown these individuals to be at particularly high risk for recidivism and violence after release from incarceration (Balyakina et al., 2014; Van Dorn, Volavka, & Jonson, 2012).

One strategy that has emerged in an attempt to address the needs of individuals with mental illness and/or substance use disorders is diversion programs. These programs offer an alternative to incarceration through participation in court-mandated treatment. In most cases, diversion programs are only offered to those individuals who meet specific thresholds for mental illness or substance use established by these programs and have committed relatively minor crimes such as misdemeanors or low-level felonies (Center for Health and Justice at TASC, 2013). The goal of diversion programs is to allow the participants to satisfy the court’s objectives while having their individualized special needs met in the least restrictive setting possible, thereby reducing the burden on correctional facilities.

The development of diversion programs is still relatively new, and it is only in the last two decades that these programs have begun to grow in popularity (Steadman & Naples, 2005). Steadman and colleagues (Steadman, Barbera, & Dennis, 1994; Steadman & Naples, 2005) estimated that only about 52 of these programs existed in the United States in 1992, but in 2005 that number had grown to about 300 programs operating nationally. According to a recent national survey of diversion programs, approximately 300 pretrial diversion programs existed in the United States, along with over 2,700 drug courts and 1,100 other “problem-solving” court parts such as mental health and veteran’s courts that also offer diversion opportunities (Center for Health and Justice at TASC, 2013). In the early years of diversion programs, there were frequent concerns that diverted individuals would reoffend more often than individuals who were incarcerated (Hoff, Rosenheck, Baranosky, Buchanan, & Zonana, 1999). However, research has not supported these concerns, as outcomes for diverted individuals were comparable or even better than for incarcerated individuals (e.g., Collins, Lonczak, & Clifasefi, 2017; Sirotich, 2009).

Diversion Program Effectiveness

Although no research was identified that focused specifically on diversion programs for individuals with a psychotic mental disorder, a number of studies have analyzed samples of (a) mentally ill offenders and (b) defendants with a substance use disorder. Sirotich (2009) reviewed 21 publications that investigated the effectiveness of diversion programs for mentally ill offenders, including studies of individuals with a single mental disorder as well as those with a dual diagnosis of mental disorder and substance abuse. He found that court-based diversion programs, which are typically run by mental health clinicians working within a courthouse, appeared to reduce the amount of time spent in custody for these individuals with serious mental illness. Although there was no evidence to suggest that these programs reduce recidivism in mentally ill offenders, there was also no evidence of increased recidivism. This review also found no evidence of differential effectiveness across the different types of diversion programs studied (Sirotich, 2009). Although the recidivism results of Sirotich’s study appear to imply that diversion does not accomplish its goal of reducing recidivism, they can also be interpreted to show diversion as a less prohibitive, more treatment-focused alternative to incarceration that provides the same level of effectiveness on recidivism rates as incarceration. Given the rapid expansion and diversification of diversion programs in recent years (e.g., Center for Health and Justice at TASC, 2013; Steadman & Naples, 2005), the absence of research examining particular subgroups of diversion clients (e.g., those with a psychotic mental disorder) is perhaps surprising. Several studies, however, have examined the effectiveness of diversion for patients with a primary substance use disorder with or without a co-occurring mental disorder.

For example, in a longitudinal study, Hoff and colleagues (1999) followed 352 criminal defendants with a substance use disorder, some of whom also had a comorbid psychiatric diagnosis. They contrasted defendants placed in a diversion program to those who were not, following the sample for 1 year to track the number of days spent incarcerated for rearrests during that time. On average, individuals in diversion spent 39 days incarcerated during the follow-up period, whereas nondiverted individuals spent an average of 141 days incarcerated. However, dually diagnosed offenders spent more time incarcerated than those with only a substance use disorder, regardless of diversion status. Several other studies have specifically focused on individuals dually diagnosed with both mental health and substance use disorders and the relationship between substance abuse and diversion outcomes.

In a quasi-experimental longitudinal study, Shafer, Arthur, and Franczak (2004) examined which individuals are most successful in diversion programs by investigating the outcomes of 248 dually diagnosed clients in two Arizona programs. The researchers conducted a series of three interviews over a 12-month period following the individuals’ entrance into the diversion program. They reported a decline in number of arrests and the rate of self-reported violence for all participants, with no significant difference in recidivism rates based on diversionary status (Shafer et al., 2004).

Frisman and colleagues (2006) also used a quasi-experimental design to examine the effectiveness of a diversion program in Connecticut for 211 offenders diagnosed with both substance abuse and mental illness. Interviews were conducted with clients at the time of diversion, 3 months after the baseline assessment, and 12 months after the baseline assessment. The authors examined a range of outcomes, including mental health, substance use, quality of life, and recidivism. They found a significant difference in the time to reincarceration between the groups, with only 17% of individuals in the diversion program being reincarcerated within the first 6 months compared with 30% of individuals who were not in the diversion program. However, scores from a self-report inventory indicated that mental health symptoms improved more for those individuals who were not in the diversion program compared with individuals who were diverted. There were no differences between groups on measures of drug use, alcohol abuse, any substance abuse, frequency of rearrest, or frequency of violent rearrest during the follow-up period.

Broner, Lattimore, Cowell, and Schlenger (2004) examined the effectiveness of jail diversion programs for 1,966 individuals with co-occurring disorders in eight sites across the United States. All participants had a diagnosis of a psychotic or mood disorder and a comorbid substance use disorder. Results of the study showed that on average, individuals in one of the diversion programs spent more time in the community between the baseline and follow-up interviews than those who were not in a diversion program. However, the researchers did not find any association between diversion program participation and recidivism rates for the overall sample or for any of the individual sites involved in the study.

The Present Study

The present study sought to build on previous research about diversion program effectiveness for individuals with serious mental illness by examining the role of substance use in differentiating successful from unsuccessful program participants. While the primary criterion for inclusion in the mental health diversion program from which we selected participants was the presence of a serious mental illness, diversion programs routinely monitor substance abuse through drug testing. Positive test results have substantial adverse effects on program outcomes, which highlights the need for accuracy in drug test results and interventions that decrease the likelihood of renewed substance abuse.

Our study aimed to address the negative findings in previous research by identifying which participants may experience difficulty complying with court monitoring and be more likely to recidivate, with a particular focus on individuals with a psychotic disorder. This goal was addressed by monitoring the progress and outcomes of a group of diversion program participants with a psychotic mental disorder to identify the characteristics of individuals with successful outcomes. Of particular interest was the potential moderating role of controlled substance use, as assessed by random urine toxicology testing.

Method

Participants

The participants in this study entered a diversion program, Queens Treatment Alternatives for Safer Communities (TASC) Mental Health Diversion Program in New York City, between January 2011 and June 2014. Individuals are referred to TASC for a mental health assessment by the Queens County District Attorney’s Office. Once a referral is made, a TASC staff member evaluates the individual for enrollment in the program based on his or her mental health, substance abuse, and criminal history. This evaluation is performed by a TASC mental health case manager, who is always under the supervision of the clinical director, a licensed clinical psychologist. The clinical director also meets with most clients during the assessment process to confirm the diagnostic findings of the case manager. TASC staff members assess individuals by means of a clinical interview, along with several structured objective measures including a specialized diagnostic evaluation scale, the Colorado symptom index (CSI; Shern et al., 1994), and the posttraumatic stress disorder (PTSD) Checklist–Civilian Version (PCL-C; Weathers et al., 1994). During the interview, individuals are asked questions regarding demographic information, substance use history (e.g., any use, first age of use, frequency of use), mental health history (e.g., current and past symptoms), and treatment history. In addition to self-report, TASC staff members also gather information from collateral sources such as family or friends of the individual, current or past treatment providers, and records of incarceration or previous psychiatric hospitalizations. The length of assessment varies depending on the complexity of an individual’s diagnosis and the ease with which collateral sources can be reached, but the assessment process generally occurs over a period of 2 to 3 weeks. Individuals are diagnosed with mental health and/or substance use disorders at the conclusion of the assessment process and a psychosocial report is prepared for the court’s reference by the TASC case managers and clinical director. If the client is suitable for diversion and chooses to participate in the program, he or she is required to enter a guilty plea and fulfill a court mandate that contains the program’s expectations and requirements with regard to drug use and mental health treatment. A treatment mandate typically lasts for a minimum of 9 months for a misdemeanor and 12 months for a felony, although the length of the mandate is ultimately at the discretion of the judge.

Participants were included in the present study if they had a felony index offense, attended at least one intake appointment at a treatment facility mandated by TASC (i.e., began the treatment mandate), and were diagnosed with a psychotic mental disorder based on the assessment of the TASC intake clinician as defined by the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2000, 2013). Diagnoses that fell under the umbrella term “psychotic disorder” for the purposes of this study included schizophrenia, schizoaffective disorder, major depressive disorder with psychotic features, bipolar disorder with psychotic features, and psychotic disorder not otherwise specified. Individuals were not eligible for the present study if they were referred to TASC but did not enter a guilty plea or make intake at a treatment facility. Intake was defined as attending a first intake appointment at an outpatient clinic or entering an inpatient treatment facility. Clients who did not enter a plea have no further interaction with TASC staff after the initial assessment, and TASC does not keep follow-up records of clients who did not enter treatment; thus, data were unavailable for these individuals to be used as a comparison group.

A total of 80 participants were selected for the current study based on our inclusion criteria. The sample consisted of 80 individuals between the ages of 18 and 83 with a mean age of 38.0 (SD = 13.3) and included 64 men (80.0%) and 16 women (20.0%). Most were Caucasian (33.8%, n = 27) or African American (37.5%, n = 30), but participants also identified as Hispanic (17.5%, n = 14), Asian (7.5%, n = 6), and “Other” (3.8%, n = 3). Participants had one of five diagnoses: schizophrenia (32.5%, n = 26), schizoaffective disorder (28.8%, n = 23), major depressive disorder with psychotic features (15.0%, n = 12), bipolar disorder with psychotic features (18.8%, n = 15), or psychotic disorder not otherwise specified (5.0%, n = 4). These categories were collapsed for statistical analyses into three groups: schizophrenia, schizoaffective disorder, and mood disorders with psychotic features. The majority of participants had a history of prior mental health treatment (90.0%, n = 72). Only 5.0% (n = 4) of participants denied any history of controlled substance use, whereas 37.5% (n = 30) reported a history of alcohol and/or marijuana use but denied using any other controlled substances, and 57.5% (n = 46) reported a history of other forms of substance use. More than two thirds of the participants (68.8%, n = 55) were diagnosed with a substance use disorder, but only 46.3% (n = 37) reported a history of substance abuse treatment. This history of treatment was almost exclusively reported by those individuals with a substance use disorder; only one individual who was not diagnosed with a current substance use disorder self-reported a history of substance abuse treatment.

Index offenses varied widely, but the most serious charge for 43.8% (n = 35) of individuals was a violent offense (e.g., assault, robbery), whereas 36.3% (n = 29) had a nonviolent, nondrug charge (e.g., criminal contempt, harassment) and 20.0% (n = 16) had a drug-related charge (e.g., criminal sale or possession of controlled substances). Roughly, three-fourths of the participants (73.8%, n = 59) had at least one prior arrest.

Procedures

The current study was approved by the Fordham University Institutional Review Board for Research on Human Subjects as of December 2013 and extended for approved data collection through December 2015. Data were extracted from records kept by the TASC program. These records included progress notes and summary reports written by TASC staff (case managers, interns, externs, and the clinical director), correspondence from psychiatric treatment facilities, records of urinalysis test results, and any other records relevant to treatment progress. Data were also collected on demographic variables, psychiatric diagnosis, index offense data, prior arrests, and substance use history for each participant at the time of the TASC intake evaluation.

Substance use history was divided into three categories: those who denied ever using controlled substances, those who reported a history of alcohol and/or marijuana use only, and those who reported a history of using drugs other than alcohol and marijuana. Alcohol and marijuana were considered as a group based on three factors: (a) the frequency of their use in the general population, (b) the differences in legality between these and other controlled substances, and (c) the fact that many participants used both while not using any other type of controlled substance. Importantly, not all participants who fell into these three substance use history categories had an identified substance use disorder. This distinction was considered in our analyses by examining a history of substance abuse treatment and the presence of a diagnosed substance use disorder separately from the categories of identified substance use.

To assess treatment outcomes, data were collected over the 12 months following intake for each participant, including urine toxicology results and rearrest incidence. Positive urine toxicology results were defined as a positive result for any controlled substance, including alcohol. Testing was conducted at regular, but variable intervals (typically several times per month) for all individuals identified as suffering from a substance use disorder, although some individuals without an identified substance use disorder were also required to submit to random drug testing (e.g., when a drug or alcohol problem was suspected but not diagnosed). Testing was typically conducted at the TASC offices, as well as by substance abuse treatment programs (if such a program was required). If a positive toxicology test result was contested by the individual, testing was confirmed by a laboratory analysis. Laboratory tests that resulted in a negative outcome were considered “negative” toxicology results, whereas positive urine toxicology screens that were uncontested, or that were confirmed by an outside laboratory, were considered “positive” findings.

When an individual has a positive toxicology result, programmatic response varies depending on circumstances such as the individual’s current treatment plan and the judge in front of whom the individual’s case is heard. Generally, if a client who is involved in outpatient treatment tests positive, outpatient substance abuse treatment services will be added or increased. Repeated positive toxicology results after treatment services have been increased will lead to inpatient treatment in the form of short-term rehabilitation or long-term residential services. If an individual in a long-term residential treatment program has a positive toxicology result, the individual is either discharged or subject to a behavioral contract at the treatment program’s discretion. Treatment contracts vary by facility and the severity of a relapse or other behavioral issue, but they generally include increased chore duties in the treatment facility and/or writing assignments related to the circumstances of relapse, along with a loss of privileges such as day passes out of the facility for a period of up to 30 days.

The second outcome variable, rearrest incidence, was defined as any arrest for a new offense during the individual’s participation in the TASC program. Incidents related to TASC program participation, such as being returned to court on a warrant (e.g., for absconding from a residential treatment facility) were not included, as they were more likely to reflect treatment noncompliance than they were to reflect new criminal activity. Rearrest data were collected from TASC records. TASC obtains rearrest information from participant self-report as well as communication from an individual’s treatment program, the court, or the district attorney’s office and can verify or discover this information through the online public Webcriminal database maintained by the New York State Unified Court System.

Statistical Analyses

Independent samples t-tests and Fisher’s Exact tests were used to identify variables associated with each outcome variable at both 6 months and 12 months following the intake evaluation. Variables analyzed included demographic (e.g., gender, age, education), clinical (e.g., diagnosis, controlled substance use history), and criminal history (e.g., history of prior arrests, violent index offense). The moderating relationship between substance abuse on recidivism was assessed using logistic regression analysis examining the impact of positive toxicology test results on other reoffense.

Results

Controlled substance use and criminal history variables, along with several demographic variables such as gender, race, and diagnosis, were examined for their association with the two outcome variables: positive urine toxicology test results and new arrests. Outcomes for each participant were assessed cumulatively at both 6 and 12 months, while the individual was under the supervision of the diversion program. At 6 months, 32.5% (n = 26) of participants had tested positive for drugs or alcohol at least once. By 12 months, this number had increased to 40.0% (n = 32). Arrest rates were much lower as 8.8% (n = 7) of participants had been arrested for a new offense within the first 6 months of TASC participation and 15.0% (n = 12) had been arrested within the first 12 months.

Associations With Outcomes

Table 1 shows the results of Fisher’s Exact tests of participants based upon the incidence of positive urine toxicology screening after 6 and 12 months in the program. Three variables were significantly associated with a positive toxicology result at 6 months: type of controlled substance use history (based on the three categories described above), p = .03; φ = .29, prior arrest history, p = .05; φ = .27, and a history of mental health treatment, p = .05; φ = .23. None of the other variables tested, including gender, race, mental health or substance diagnosis, or index offense type, showed significant associations with a positive toxicology at 6 months.

Table 1.

Associations With Positive Toxicology Outcomes at 6 and 12 Months.

	Six-month toxicology results			Twelve-month toxicology results
	Positive	Negative	φ/p	Positive	Negative	φ/p
	(n = 26)	(n = 54)	φ/p	(n = 32)	(n = 48)	φ/p
Gender			φ = –.15			φ = –.22
Male	23 (35.9%)	41 (64.1%)	p = .24	29 (45.3%)	35 (54.7%)	p = .09
Female	3 (18.8%)	13 (81.3%)		3 (18.8%)	13 (81.3%)
Race			φ = .17			φ = .17
Caucasian	6 (22.2%)	21 (77.8%)	p = .53	9 (33.3%)	18 (66.7%)	p = .50
African American	11 (36.7%)	19 (63.3%)		12 (40%)	18 (60%)
Hispanic/Latino	6 (42.9%)	8 (57.1%)		8 (57.1%)	6 (42.9%)
Other	3 (33.3%)	6 (32.5%)		3 (33.3%)	6 (66.7%)
Education			φ = .27			φ = .26
Less than HS	13 (48.1%)	14 (51.9%)	p = .06	15 (55.6%)	12 (44.4%)	p = .07
HS or GED	8 (32%)	17 (68%)		10 (40%)	15 (60%)
Some college	5 (17.9%)	23 (82.1%)		7 (25%)	21 (75%)
Diagnosis			φ = .09			φ = .05
Schizophrenia	9 (34.6%)	17 (65.4%)	p = .73	11 (42.3%)	15 (57.7%)	p = .93
Schizoaffective	8 (34.8%)	15 (65.2%)		9 (39.1%)	14 (60.9%)
Mood disorder	7 (25.9%)	20 (74.1%)		10 (37%)	17 (63%)
Prior MH treatment			φ = .23			φ = .19
Yes	26 (36.1%)	46 (63.9%)	p = .05	31 (43.1%)	41 (56.9%)	p = .14
No	0 (0%)	8 (100%)		1 (12.5%)	7 (87.5%)
Substance diagnosis			φ = .18			φ = .22
Yes	21 (38.2%)	34 (61.8%)	p = .13	26 (47.3%)	29 (52.7%)	p = .05
No	5 (20%)	20 (80%)		6 (24%)	19 (76%)
Substance type			φ = .29			φ = .31
None	0 (0%)	4 (100%)	p = .03	0 (0%)	4 (100%)	p = .02
Alcohol/marijuana	6 (20%)	24 (80%)		8 (26.7%)	22 (73.3%)
Other drugs	20 (43.5%)	26 (56.5%)		24 (52.2%)	22 (47.8%)
Prior SA treatment			φ = .05			φ = .11
Yes	13 (35.1%)	24 (64.9%)	p = .81	17 (45.9%)	20 (54.1%)	p = .36
No	13 (30.2%)	(69.8%)		15 (34.9%)	28 (65.1%)
Index offense type			φ = .19			φ = .30
Violent	14 (40%)	21 (60%)	p = .23	18 (51.4%)	17 (48.6%)	p = .03
Drug	6 (37.5%)	10 (62.5%)		8 (50%)	8 (50%)
Other	6 (20.7%)	23 (79.3%)		6 (20.7%)	23 (79.3%)
Prior arrest			φ = .27			φ = .26
None	4 (19%)	17 (81%)	p = .05	5 (23.8%)	16 (76.2%)	p = .07
1-2 arrests	7 (25%)	21 (75%)		10 (35.7%)	18 (64.3%)
3 or more	15 (48.4%)	16 (51.6%)		17 (54.8%)	14 (45.2%)

Note. Percentages reflect proportion within row. HS = high school; GED = General Education Development; MH = mental health; SA = substance abuse.

At 12 months, the relationship between type of controlled substance use and a positive toxicology remained significant, p = .02; φ = .31, but the associations with mental health treatment, p = .14; φ = .19, and prior arrests, p = .07, φ = .26, did not. Significant associations were also observed for index offense (violent, drug-related, and other), p = .03; φ = .30, and substance use disorder diagnosis, p = .05; φ = .22.

Table 2 shows the results of Fisher’s Exact tests addressing the association between demographic, clinical, and criminal justice variables and new arrests within the first 6 and 12 months in the TASC program. Only one variable was significantly associated with a new arrest at 6 months: clinical mental health diagnosis, p = .03; φ = .31. Individuals diagnosed with schizoaffective disorder had higher rates of rearrest than individuals diagnosed with schizophrenia or a mood disorder with psychotic features (21.7% vs. 7.7% and 0%, respectively; four individuals diagnosed with psychotic disorder not otherwise specified were omitted from this analysis).

Table 2.

Associations With Rearrest at 6 and 12 Months.

	Six-month			Twelve-month
	Yes	No	φ/p	Yes	No	φ/p
	(n = 7)	(n = 73)	φ/p	(n = 12)	(n = 68)	φ/p
Gender			φ = .07			φ = –.04
Male	5 (7.8%)	59 (92.2%)	p = .62	10 (15.6%)	54 (84.4%)	p = 1.00
Female	3 (18.8%)	13 (81.3%)		2 (12.5%)	14 (87.5%)
Race			φ = .05			φ = .07
Caucasian	2 (7.4%)	25 (92.6%)	p = .97	4 (14.8%)	23 (85.2%)	p = .93
African American	3 (10%)	27 (90%)		4 (13.3%)	26 (86.7%)
Hispanic/Latino	1 (7.1%)	13 (92.9%)		2 (14.3%)	12 (85.7%)
Other	1 (11.1%)	8 (88.9%)		2 (22.2%)	7 (77.8%)
Education			φ = .16			φ = .09
Less than HS	4 (14.8%)	23 (85.2%)	p = .36	5 (18.5%)	22 (81.5%)	p = .71
HS or GED	1 (4%)	24 (96%)		4 (16%)	21 (84%)
Some college	2 (7.1%)	26 (92.9%)		3 (10.7%)	25 (89.3%)
Diagnosis			φ = .31			φ = .30
Schizophrenia	2 (7.7%)	24 (92.3%)	p = .03	2 (7.7%)	24 (92.3%)	p = .03
Schizoaffective	5 (21.7%)	18 (78.3%)		7 (30.4%)	16 (69.6%)
Mood disorder	0 (0%)	27 (100%)		2 (7.4%)	25 (92.6%)
Prior MH treatment			φ = .10			φ = .14
Yes	7 (9.7%)	65 (90.3%)	p = 1.0	12 (16.7%)	60 (83.3%)	p = .60
No	0 (0%)	8 (100%)		0 (0%)	8 (100%)
Substance diagnosis			φ = .11			φ = .21
Yes	6 (10.9%)	49 (89.1%)	p = .43	11 (20%)	44 (80%)	p = .09
No	1 (4%)	24 (96%)		1 (4%)	24 (96%)
Substance type			φ = .10			φ = .11
None	0 (0%)	4 (100%)	p = .67	0 (0%)	4 (100%)	p = .61
Alcohol/marijuana	2 (6.7%)	28 (93.3%)		4 (13.3%)	26 (86.7%)
Other drugs	5 (10.9%)	41 (89.1%)		8 (17.4%)	38 (82.6%)
Prior SA treatment			φ = –.02			φ = .17
Yes	3 (8.1%)	34 (91.9%)	p = 1.0	8 (21.6%)	29 (78.4%)	p = .20
No	4 (9.3%)	39 (90.7%)		4 (9.3%)	39 (90.7%)
Index offense type			φ = .24			φ = .28
Violent	5 (14.3%)	30 (85.7%)	p = .11	9 (25.7%)	26 (74.3%)	p = .04
Drug	2 (12.5%)	14 (87.5%)		2 (12.5%)	14 (87.5%)
Other	0 (0%)	29 (100%)		1 (3.4%)	28 (96.6%)
Prior arrest			φ = .23			φ = .10
None	3 (14.3%)	18 (85.7%)	p = .13	3 (14.3%)	18 (85.7%)	p = .65
1-2 arrests	0 (0%)	28 (100%)		3 (10.7%)	25 (89.3%)
3 or more	4 (12.9%)	27 (87.1%)		6 (19.4%)	25 (80.6%)

Note. Percentages reflect proportion rearrested within row. HS = high school; GED = General Education Development; MH = mental health, SA = substance abuse.

At 12 months, mental health diagnosis was again associated with rearrest although the rates of rearrest were generally higher, p = .03; φ = .30, with 30.4% of clients diagnosed with schizoaffective disorder rearrested compared with 7.7% of those with schizophrenia and 7.4% of those with a mood disorder. In addition, index offense type was associated with rearrest at 12 months, p = .04; φ = .28. Individuals who had been referred to TASC following a violent offense were more likely to be rearrested within the first 12 months than those charged with drug offenses or other felony offenses (primarily theft-related; 25.7% vs. 12.5% and 3.4%, respectively). There was no association between a positive urine toxicology test result and rearrest at either 6 or 12 months, p = 1.0; φ = –.03 and p = .21; φ = .16.

Table 3 shows the results of independent-samples t tests assessing the relationship between age and toxicology and rearrest outcomes at both 6 and 12 months. Notably, mean age did not differ significantly between those individuals with or without a positive toxicology at 6 (p = .49) or 12 months (p = .26). The same null results were found for those with and without a rearrest at both 6 (p = .73) and 12 months (p = .35).

Table 3.

Age Associations With Toxicology and Rearrest Outcomes at 6 and 12 Months.

	Six-month toxicology results			Twelve-month toxicology results
	Yes (n = 26)	No (n = 54)	t/p	Yes (n = 32)	No (n = 48)	t/p
Mean age	36.5 (10.8)	38.7 (14.5)	t = 0.69p = .49	35.9 (11.2)	39.3 (14.6)	t = 1.14p = .26
	Six-month rearrest			Twelve-month rearrest
	Yes (n = 7)	No (n = 73)	t/p	Yes (n = 12)	No (n = 68)	t/p
Mean age	36.3 (14.0)	38.1 (13.4)	t = 0.34p = .73	34.6 (14.2)	38.5 (13.2)	t = 0.95p = .35

Note. Standard deviations are in parentheses.

Moderating Effect of Substance Misuse

To examine the impact of drug use on rearrest, we used logistic regression analyses to examine the interaction between a positive urine toxicology test result and the two variables with the strongest associations with rearrest: clinical diagnosis (schizoaffective vs. other) and violent index offense. These analyses indicated that, at both 6- and 12-month outcomes, diagnosis was the only significant variable associated with rearrest (see Table 4). There was no significant increase in the predictive models when positive toxicology test results were added to each of the models, nor was there a significant interaction between positive toxicology test results and either diagnosis or violent index offense (data available upon request).

Table 4.

Logistic Regression Models Predicting Rearrest at 6 and 12 Months.

Variable	B	Wald	p	OR [95% CI]
Rearrest at 6 months (n = 7)
Schizoaffective disorder	2.06	5.00	.03	7.87 [1.29, 47.98]
Violent index offense	1.46	2.39	.12	4.32 [0.67, 27.35]
Positive urine toxicology	−0.68	0.47	.49	0.51 [0.07, 3.52]
Rearrest at 12 months (n = 11)
Schizoaffective disorder	1.68	5.48	.02	5.38 [1.32, 22.01]
Violent index offense	1.36	3.05	.08	3.89 [0.85, 17.83]
Positive urine toxicology	0.46	0.39	.53	1.58 [0.38, 6.69]

Note. OR = odds ratio; CI = confidence interval.

Discussion

The results of the current study highlight the importance of three variables in the outcomes of offenders suffering from psychosis in a diversion program: type of controlled substance use, clinical diagnosis, and violent index offense. When analyzing positive urine toxicology outcomes after both 6 and 12 months in the diversion program, there was a significant relationship with substance use history. Individuals who had a history of using drugs other than alcohol or marijuana were more likely to test positive while under supervision than those with a history of only alcohol and marijuana use; and both groups of individuals (i.e., any history of controlled substance use) were more likely to test positive than those who had no reported history of controlled substance use. Notably, this relationship was not the same for individuals formally diagnosed with a substance use disorder diagnosis as those with a substance use disorder diagnosis were only more likely to show positive toxicology results than nondiagnosed individuals at 12 months. Index offense type also showed a significant association with positive urine toxicology after 12 months in the program; individuals referred to TASC following a violent offense were more likely to test positive for controlled substances after 12 months than individuals with other charges.

Identifying associations with rearrest was more successful when analyzing 12-month outcomes than when analyzing 6-month outcomes, in large part because the rate of rearrest within the first 6 months of program participation was quite low (8.8%). At 12 months, clinical mental health diagnosis (individuals with a diagnosis of schizoaffective disorder were more likely to be rearrested than individuals with other diagnoses) and a violent index offense were both significantly associated with rearrest. Surprisingly, positive urine toxicology screening (i.e., use of controlled substances) was not, nor did positive urine toxicology test results moderate the relationship between diagnosis or violent index offense and rearrest. The lack of any discernible impact of renewed drug or alcohol use on rearrest rates was unexpected, and raises questions about the nature of any relationship between controlled substance use and criminal behavior. Because much of the literature that drove this hypothesis focused on violent behavior (e.g., Van Dorn et al., 2012), which was very uncommon in this sample (at least during the study follow-up period), the unique impact of renewed drug or alcohol use on nonviolent criminal behavior may be less powerful. Alternatively, drug/alcohol use may moderate other variables that were not available in this study (e.g., impulsivity, cognitive functioning).

The lack of a moderating effect of renewed drug/alcohol use does not necessarily diminish the importance of substance abuse in diversion program outcomes. Although measured indirectly, an individual’s drug of choice appears to significantly affect the likelihood of abstaining from controlled substance use while in a diversion setting. One goal of diversion is for participants to satisfy the court’s objectives while having their individualized needs met in the least restrictive setting possible. This goal is adversely affected by renewed drug or alcohol use and may result in termination from treatment programs or even “failure” in the diversion program. These findings suggest that greater concern (or tolerance, in some cases) may be warranted for those individuals with a more varied drug use history, as ability to abstain was significantly poorer among those individuals who had used drugs other than marijuana and alcohol. Of course, a more in-depth analysis of the specific drugs used and the frequency and extent of their use would likely provide a more nuanced understanding of the role played by substance abuse in criminal behavior and may have helped identify the hypothesized moderating relationship on rearrest. Future research with much larger samples would be needed to adequately examine drug use in more detail.

One unexpected finding was that individuals with schizoaffective disorder had higher rates of rearrest (at both time points) than those with other psychotic disorder diagnoses. Although not anticipated, this finding likely reflects the greater difficulty these individuals have in managing the combination of psychotic and mood symptoms compared with individuals with less complex symptom patterns. The diagnosis of schizoaffective disorder has long been a source of controversy (e.g., Pope, Lipinski, Cohen, & Axelrod, 1980; Stip et al., 2005), but some research has demonstrated higher levels of functioning and better outcomes among this subgroup of individuals with a psychotic disorder (e.g., Harrow, Grossman, Herbener, & Davies, 2000). This literature points to alternative explanations for the study findings, such as a greater ability to renew criminal behavior among individuals who are less impaired by their illness. Further research disentangling these possible explanations, including measures of symptom severity (that were not available in the current study) might help disentangle these and other alternative explanations. Regardless of the reason for these findings, individuals with schizoaffective disorder may need more aggressive management strategies to successfully complete a diversion program and remain at liberty in the community.

Another noteworthy finding was the null result of our age analyses. Age plays a central role in the considerations of mental illness (e.g., individuals may become more accustomed to their symptoms, or symptoms may worsen with age), substance use (e.g., individuals may try more diverse drugs over time), and arrest (e.g., younger individuals have more current police contact; Hirschi & Gottfredson, 1983). It is therefore surprising that age did not appear to play a role in our measured outcomes.

Limitations

The current study was limited by several factors. The first of these was the modest sample size. With only 80 participants, our ability to distinguish trends in the outcome variables was limited. This was particularly true for variables associated with rearrest as only 12 individuals had been rearrested within the first 12 months they were under program supervision. Our modest sample size also limited the power of our exploratory logistic regression analyses. In addition, a larger sample would have enhanced our ability to disentangle complex variables, such as the nature and extent of renewed drug or alcohol use.

Urine toxicology testing can identify multiple drugs of abuse, yet we collapsed this complex variable into positive versus negative test results, essentially equating “minor” substance use (e.g., a single drink of alcohol) with more serious substance use (e.g., daily crack cocaine use). Likewise, although data regarding number of positive toxicology findings was available, we collapsed this variable into a single dichotomous variable for data analysis (though analysis as a continuous variable did not alter the findings). An additional concern regarding the toxicology result variable was that some patients with a history of controlled substance use, but without a substance use disorder, were required to participate in random drug testing, while other patients with no known history of substance misuse were not administered urine drug testing. While this does raise the possibility of undetected drug use by those patients without a history of drug use and not required to participate in drug testing, this possibility appears unlikely.

Additional, potentially important variables were not available for study including the nature of charges that gave rise to rearrest, the occurrence of criminal behavior that did not lead to an arrest, and the occurrence of rearrest after diversion program completion. These and other study limitations (e.g., the absence of a measure of symptom severity, medication compliance measures, and psychopathy or comorbid personality disorder diagnoses) may have further obscured potentially interesting results. Finally, the lack of a comparison with a sample of diversion program participants without psychosis prevents any assessment of whether these findings would differ in less severely ill individuals.

These limitations notwithstanding, these results suggest that diversion programs may not be best conceived as a “one size fits all” approach. Some individuals may benefit from a strong focus on treating problematic substance use, whereas other individuals may benefit more from a more aggressive mental health treatment focus, perhaps with greater tolerance for “slips” in their abstinence. Currently available literature fails to examine the impact of drug or mental health court leniency in these areas on recidivism and substance use outcomes, suggesting an important avenue for future research. The present study, while preliminary, suggests that a stronger focus on mental health (and perhaps criminogenic) needs may be more beneficial to individuals with a psychotic mental disorder, as positive urine toxicology findings were not associated with rearrest. However, there is little doubt that some individuals require more aggressive management of their substance abuse and an attitude of tolerance is likely to be more harmful than beneficial for those individuals. Without continued attention to these complex and interwoven issues, research on diversion will likely continue to show mixed outcomes and, by extension, provide relatively little guidance to service providers who treat these challenging individuals.

Footnotes

Authors’ Note

Jennifer K. Boland is now at the Department of Psychology, Sam Houston State University, Huntsville, Texas, USA.

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.

References

Abram

K. H.

Teplin

L. A.

McClelland

G. M.

(2003). Comorbidity of severe psychiatric disorders and substance use disorders among women in jail. American Journal of Psychiatry, 150, 1007-1010. doi:10.1176/appi.ajp.160.5.1007

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author.

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

Balyakina

Mann

Ellison

Silvernell

Fulda

K. G.

Sarai

S. K.

Cardarelli

(2014). Risk of future offense among probationers with co-occurring substance use and mental health disorders. Community Mental Health Journal, 50, 288-295. doi:10.1007/s10597-013-9624-4

Blumstein

Beck

A. J.

(1999). Population growth in U.S. prisons, 1980-1996. Crime and Justice, 26, 17-61.

Bradley-Engen

M. S.

Cuddeback

G. S.

Gayman

M. D.

Morrissey

J. P.

Mancuso

(2010). Trends in state prison admission of offenders with serious mental illness. Psychiatric Services, 61, 1263-1265. doi:10.1176/appi.ps.61.12.1263

Broner

Lattimore

Cowell

A. J.

Schlenger

W. E.

(2004). Effects of diversion on adults with co-occurring mental illness and substance use: Outcomes from a national multi-site study. Behavioral Sciences & the Law, 22, 519-541. doi:10.1002/bsl.605

Center for Health and Justice at TASC. (2013). No entry: A National Survey of Criminal Justice Diversion Programs and Initiatives. Chicago, IL: Author.

Collins

S. E.

Lonczak

H. S.

Clifasefi

S. L.

(2017). Seattle’s Law Enforcement Assisted Diversion (LEAD): Program effects on recidivism outcomes. Evaluation and Program Planning, 64, 49-56.

10.

Fries

B. E.

Schmorrow

Lang

S. W.

Margolis

P. M.

Heany

Brown

G. P.

. . . Hirdes

J. P.

(2013). Symptoms and treatment of mental illness among prisoners: A study of Michigan state prisons. International Journal of Law and Psychiatry, 36, 316-325. doi:10.1016/j.ijlp.2013.04.008

11.

Frisman

L. K.

Lin

Sturges

G. E.

Levinson

Baranoski

M. V.

Pollard

(2006). Outcomes of court-based jail diversion programs for people with co-occurring disorders. Journal of Dual Diagnosis, 2, 5-26. doi:10.1300/J374v02n02_02

12.

Gunter

T. D.

Arndt

Wenman

Allen

Loveless

Sieleni

Black

(2008). Frequency of mental and addictive disorders among 320 men and women entering the Iowa prison system: Use of the MINI-Plus. Journal of the American Academy of Psychiatry and the Law, 36, 27-34.

13.

Harrow

Grossman

L. S.

Herbener

E. S.

Davies

E. W.

(2000). Ten-year outcome: Patients with schizoaffective disorders, schizophrenia, affective disorders and mood-incongruent psychotic symptoms. British Journal of Psychiatry, 177, 421-426. doi:10.1192/bjp.177.5.421

14.

Hirschi

Gottfredson

(1983). Age and the explanation of crime. American Journal of Sociology, 89, 552-584.

15.

Hoff

R. A.

Rosenheck

R. A.

Baranosky

M. V.

Buchanan

Zonana

(1999). Diversion from jail of detainees with substance abuse: The interaction with dual diagnosis. The American Journal on Addictions, 8, 201-210. doi:10.1080/105504999305811

16.

James

D. J.

Glaze

L. E.

(2006). Mental health problems of prison and jail inmates (NCJ 213600). Washington, DC: Bureau of Justice Statistics.

17.

Kaeble

Glaze

Tsoutis

Minton

(2015). Correctional populations in the United States, 2014. Washington, DC: Bureau of Justice Statistics.

18.

Lamb

R. H.

Weinberger

L. E.

(1998). Persons with severe mental illness in jails and prisons: A review. Psychiatric Services, 49, 483-492.

19.

Mumola

C. J.

Karberg

J. C.

(2006). Drug use and dependence, state and federal prisoners, 2004. Washington, DC: U.S. Department of Justice. doi:10.1037/e560272006-001

20.

Pope

H. G.

Lipinski

J. F.

Cohen

B. M.

Axelrod

D. T.

(1980). “Schizoaffective disorder”: An invalid diagnosis? A comparison of schizoaffective disorder, schizophrenia, and affective disorder. American Journal of Psychiatry, 137, 921-927.

21.

Prins

S. J.

(2014). Prevalence of mental illnesses in U.S. state prisons: A systematic review. Psychiatric Services, 65, 862-872.

22.

Shafer

M. S.

Arthur

Franczak

(2004). An analysis of post-booking jail diversion programming for persons with co-occurring disorders. Behavioral Sciences & the Law, 22, 771-785. doi:10.1002/bsl.603

23.

Shern

D. L.

Wilson

N. Z.

Coen

A. S.

Patrick

D. C.

Foster

Bartsch

D. A.

Demmler

(1994). Client outcomes II: Longitudinal client data from the Colorado Treatment Outcome Study. The Milbank Quarterly, 72, 123-148. doi:10.2307/3350341

24.

Sirotich

(2009). The criminal justice outcomes of jail diversion programs for persons with mental illness: A review of the evidence. Journal of the American Academy of Psychiatry and the Law, 37, 461-472.

25.

Steadman

H. J.

Barbera

S. S.

Dennis

D. L.

(1994). A national survey of jail diversion programs for mentally ill detainees. Hosp Community Psychiatry, 45, 1109-1113.

26.

Steadman

H. J.

Naples

(2005). Assessing the effectiveness of jail diversion programs for persons with serious mental illness and co-occurring substance use disorders. Behavioral Sciences & the Law, 23, 163-170. doi:10.1002/bsl.640

27.

Stip

Sepehry

A. A.

Prouteau

Briand

Nicole

Lalonde

Lesage

(2005). Cognitive discernible factors between schizophrenia and schizoaffective disorder. Brain and Cognition, 59, 292-295. doi:10.1016/j.bandc.2005.07.003

28.

Van Dorn

Volavka

Jonson

. (2012). Mental disorder and violence: Is there a relationship beyond substance use? Social Psychiatry & Psychiatric Epidemiology, 47, 487-503. doi:10.1007/s00127-011-0356-x

29.

Weathers

F. W.

Litz

B. T.

Herman

Huska

Keane

(1994). The PTSD checklist-civilian version (PCL-C). Boston, MA: National Center for PTSD.