Interventions for drug‐using offenders with co‐occurring mental health problems
Abstract
Background
This review represents one from a family of three reviews focusing on interventions for drug‐using offenders. Many people under the care of the criminal justice system have co‐occurring mental health problems and drug misuse problems; it is important to identify the most effective treatments for this vulnerable population.
Objectives
To assess the effectiveness of interventions for drug‐using offenders with co‐occurring mental health problems in reducing criminal activity or drug use, or both.
This review addresses the following questions.
• Does any treatment for drug‐using offenders with co‐occurring mental health problems reduce drug use?
• Does any treatment for drug‐using offenders with co‐occurring mental health problems reduce criminal activity?
• Does the treatment setting (court, community, prison/secure establishment) affect intervention outcome(s)?
• Does the type of treatment affect treatment outcome(s)?
Search methods
We searched 12 databases up to February 2019 and checked the reference lists of included studies. We contacted experts in the field for further information.
Selection criteria
We included randomised controlled trials designed to prevent relapse of drug use and/or criminal activity among drug‐using offenders with co‐occurring mental health problems.
Data collection and analysis
We used standard methodological procedures as expected by Cochrane .
Main results
We included 13 studies with a total of 2606 participants. Interventions were delivered in prison (eight studies; 61%), in court (two studies; 15%), in the community (two studies; 15%), or at a medium secure hospital (one study; 8%). Main sources of bias were unclear risk of selection bias and high risk of detection bias.
Four studies compared a therapeutic community intervention versus (1) treatment as usual (two studies; 266 participants), providing moderate‐certainty evidence that participants who received the intervention were less likely to be involved in subsequent criminal activity (risk ratio (RR) 0.67, 95% confidence interval (CI) 0.53 to 0.84) or returned to prison (RR 0.40, 95% CI 0.24 to 0.67); (2) a cognitive‐behavioural therapy (one study; 314 participants), reporting no significant reduction in self‐reported drug use (RR 0.78, 95% CI 0.46 to 1.32), re‐arrest for any type of crime (RR 0.69, 95% CI 0.44 to 1.09), criminal activity (RR 0.74, 95% CI 0.52 to 1.05), or drug‐related crime (RR 0.87, 95% CI 0.56 to 1.36), yielding low‐certainty evidence; and (3) a waiting list control (one study; 478 participants), showing a significant reduction in return to prison for those people engaging in the therapeutic community (RR 0.60, 95% CI 0.46 to 0.79), providing moderate‐certainty evidence.
One study (235 participants) compared a mental health treatment court with an assertive case management model versus treatment as usual, showing no significant reduction at 12 months' follow‐up on an Addictive Severity Index (ASI) self‐report of drug use (mean difference (MD) 0.00, 95% CI ‐0.03 to 0.03), conviction for a new crime (RR 1.05, 95% CI 0.90 to 1.22), or re‐incarceration to jail (RR 0.79, 95% CI 0.62 to 1.01), providing low‐certainty evidence.
Four studies compared motivational interviewing/mindfulness and cognitive skills with relaxation therapy (one study), a waiting list control (one study), or treatment as usual (two studies). In comparison to relaxation training, one study reported narrative information on marijuana use at three‐month follow‐up assessment. Researchers reported a main effect < .007 with participants in the motivational interviewing group, showing fewer problems than participants in the relaxation training group, with moderate‐certainty evidence. In comparison to a waiting list control, one study reported no significant reduction in self‐reported drug use based on the ASI (MD ‐0.04, 95% CI ‐0.37 to 0.29) and on abstinence from drug use (RR 2.89, 95% CI 0.73 to 11.43), presenting low‐certainty evidence at six months (31 participants). In comparison to treatment as usual, two studies (with 40 participants) found no significant reduction in frequency of marijuana use at three months post release (MD ‐1.05, 95% CI ‐2.39 to 0.29) nor time to first arrest (MD 0.87, 95% CI ‐0.12 to 1.86), along with a small reduction in frequency of re‐arrest (MD ‐0.66, 95% CI ‐1.31 to ‐0.01) up to 36 months, yielding low‐certainty evidence; the other study with 80 participants found no significant reduction in positive drug screens at 12 months (MD ‐0.7, 95% CI ‐3.5 to 2.1), providing very low‐certainty evidence.
Two studies reported on the use of multi‐systemic therapy involving juveniles and families versus treatment as usual and adolescent substance abuse therapy. In comparing treatment as usual, researchers found no significant reduction up to seven months in drug dependence on the Drug Use Disorders Identification Test (DUDIT) score (MD ‐0.22, 95% CI ‐2.51 to 2.07) nor in arrests (RR 0.97, 95% CI 0.70 to 1.36), providing low‐certainty evidence (156 participants). In comparison to an adolescent substance abuse therapy, one study (112 participants) found significant reduction in re‐arrests up to 24 months (MD 0.24, 95% CI 0.76 to 0.28), based on low‐certainty evidence.
One study (38 participants) reported on the use of interpersonal psychotherapy in comparison to a psychoeducational intervention. Investigators found no significant reduction in self‐reported drug use at three months (RR 0.67, 95% CI 0.30 to 1.50), providing very low‐certainty evidence. The final study (29 participants) compared legal defence service and wrap‐around social work services versus legal defence service only and found no significant reductions in the number of new offences committed at 12 months (RR 0.64, 95% CI 0.07 to 6.01), yielding very low‐certainty evidence.
Authors' conclusions
Therapeutic community interventions and mental health treatment courts may help people to reduce subsequent drug use and/or criminal activity. For other interventions such as interpersonal psychotherapy, multi‐systemic therapy, legal defence wrap‐around services, and motivational interviewing, the evidence is more uncertain. Studies showed a high degree of variation, warranting a degree of caution in interpreting the magnitude of effect and the direction of benefit for treatment outcomes.
PICOs
Plain language summary
Interventions for drug‐using offenders with co‐occurring mental health problems
What is the aim?
To identify therapies to reduce drug use and/or criminal activity among criminal justice involved people with mental health problems.
What is the key message?
Therapeutic community interventions and mental health treatment courts may help people to reduce subsequent drug use and/or criminal activity.
What was studied?
Therapies identified to support criminal justice involved people with mental health and drug misuse problems.
What are the results?
￭ When men engage with a therapeutic community intervention compared to treatment as usual, they are probably less likely to be re‐arrested or return to prison (moderate‐certainty).
￭ When women engage with a therapeutic community intervention compared to a cognitive‐behavioural course, they may not be more likely to reduce drug use, or become involved in criminal activity/drug‐related crimes (low‐certainty).
￭ When men engage with a therapeutic community compared to no intervention, they are probably less likely to return to prison (moderate‐certainty).
￭ When juveniles engage with a mental health court compared to treatment as usual, they may be less likely to commit a new crime, return to prison, or take drugs (low‐certainty).
￭ When juveniles engage with motivational interviewing/mindfulness and cognitive skills, they are probably less likely to show fewer problems than receiving relaxation training (moderate‐certainty).
￭ When people engage with motivational interviewing/mindfulness and cognitive skills, they may not be more likely to report a reduction/abstinence from drug use when compared to a waiting list control (low‐certainty).
￭ We are uncertain whether people engaged in motivational interviewing/mindfulness and cognitive skills are not more likely to report a reduction in marijuana use, a positive drug test, or to be re‐arrested when compared to treatment as usual (very low‐certainty).
￭ When families and juveniles engage in multi‐systemic therapy, they may be more likely to report a reduction in drug dependence or to be re‐arrested in comparison to treatment as usual or group substance abuse therapy (low‐certainty).
￭ We are uncertain whether people involved in interpersonal psychotherapy are not more likely to use drugs again in comparison to a psychoeducational intervention (very low‐certainty).
￭ We are uncertain whether people involved in legal defence service and wrap‐around services are not more likely to commit new offences in comparison to a legal defence service only (very low‐certainty).
Sources of funding included government institutes, research bodies, or charities.
How up‐to‐date is this review?
February 2019.
Authors' conclusions
Summary of findings
| Therapeutic community compared to treatment as usual for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with treatment as usual | Risk difference with therapeutic community | ||||
| Re‐arrests | 266 | ⊕⊕⊕⊝ | RR 0.67 | Study population | |
| 98 per 100 | 32 fewer per 100 | ||||
| Re‐incarceration | 266 | ⊕⊕⊕⊝ | RR 0.40 | Study population | |
| 59 per 100 | 36 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by one for risk of bias (blinding and selective reporting). | |||||
| Therapeutic community and aftercare compared to cognitive‐behavioural skills for drug‐using women offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using women offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with cognitive‐behavioural skills | Risk difference with therapeutic community and aftercare | ||||
| Self‐reported drug use | 314 | ⊕⊝⊝⊝ | RR 0.78 | Study population | |
| 17 per 100 | 4 fewer per 100 | ||||
| Re‐arrest for any type of crime | 314 | ⊕⊝⊝⊝ | RR 0.69 | Study population | |
| 33 per 100 | 10 fewer per 100 | ||||
| Criminal Activity | 314 | ⊕⊝⊝⊝ | RR 0.74 | Study population | |
| 33 per 100 | 9 fewer per 100 | ||||
| Drug‐related crime | 314 | ⊕⊝⊝⊝ | RR 0.87 | Study population | |
| 21 per 100 | 3 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aUnclear reporting in the paper raises concerns about the potential high risk of bias with regards to blinding and methods used in the randomisation procedure; we downgraded by one. bOne study with 95% confidence intervals through the line of no effect. | |||||
| Therapeutic community compared to waiting list control for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with waiting list control | Risk difference with therapeutic community | ||||
| Return to prison (recidivism) post parole | 478 | ⊕⊕⊕⊝ MODERATEa | RR 0.60 | Study population | |
| 40 per 100 | 16 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by one for risk of bias (randomisation process, concealment, and selective reporting). | |||||
| Mental health treatment court with assertive case management model compared to treatment as usual for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with treatment as usual | Risk difference with mental health treatment court with assertive case management model | ||||
| Conviction for a new crime | 235 | ⊕⊝⊝⊝ | RR 1.05 | Study population | |
| 72 per 100 | 4 more per 100 | ||||
| Re‐incarceration to jail | 235 | ⊕⊝⊝⊝ | RR 0.79 | Study population | |
| 71 per 100 | 15 fewer per 100 | ||||
| Self‐reported drug use | 235 | ⊕⊝⊝⊝ | ‐ | Mean self‐reported drug use was 0.08 | MD 0.00 |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by one for risk of bias (allocation concealment and blinding of assessors) and by one for imprecision. | |||||
| Motivational interviewing/mindfulness and cognitive skills compared to relaxation training for drug‐using offenders with co‐occurring mental health problems | |||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||
| Outcomes | № of participants | Certainty of the evidence | Impact |
| Self‐reported marijuana use continuous | 181 | MODERATEa | This study compared cognitive skills to a relaxation training intervention for adolescents in prison with depressed mood. Researchers measured marijuana use at 3‐months follow‐up assessment using the Risks and Consequences Questionnaire (RCQ). They report a main effect < .007, with participants in the motivational interviewing group showing fewer problems than participants in the relaxation training group. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||
| GRADE Working Group grades of evidence. | |||
| aDowngraded by one for unclear risk of bias (random allocation and blinding). | |||
| Motivational interviewing/mindfulness and cognitive skills compared to waiting list control for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with waiting list control | Risk difference with motivational interviewing and cognitive skills | ||||
| Self‐reported drug use | 31 | ⊕⊕⊝⊝ | ‐ | Mean self‐reported drug use was 0.44 | MD ‐0.04 lower |
| Abstinence from drug use | 31 | ⊕⊕⊝⊝ | RR 2.89 | Study population | |
| 15 per 100 | 29 more per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by two for optimal information size not met. | |||||
| Motivational interviewing/mindfulness and cognitive skills compared to treatment as usual for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with treatment as usual | Risk difference with motivational interviewing and cognitive skills | ||||
| Self‐reported frequency of marijuana use | 40 | ⊕⊕⊝⊝ VERY LOWa,b | ‐ | Mean self‐reported frequency of marijuana use was 1.50 | MD ‐1.05 lower |
| Arrest frequency post release | 40 | ⊕⊕⊝⊝ | ‐ | Mean arrest frequency post release was 1.47 | MD ‐0.66 lower |
| Time to first arrest or offence | 40 | ⊕⊕⊝⊝ | ‐ | Mean time to first arrest or offence was 1.6 | MD 0.87 higher |
| Positive drug screen or refusal to provide a urine sample | 84 | ⊕⊝⊝⊝ | ‐ | Mean positive drug screen or refusal to provide a urine sample was 3.25 | MD ‐0.7 lower |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by two for optimal size not met. bDowngraded by one for risk of bias (incomplete outcome measures). | |||||
| Multi‐systemic therapy involving family compared to treatment as usual for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with treatment as usual | Risk difference with multi‐systemic therapy involving family | ||||
| Drug dependence | 156 | ⊕⊕⊝⊝ | ‐ | Mean drug dependence was 3.55 | MD ‐0.22 lower |
| Arrested | 158 | ⊕⊕⊝⊝ | RR 0.97 | Study population | |
| 47 per 100 | 1 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by one for risk of bias (blinding measures) and downgraded by one for imprecision. | |||||
| Multi‐systemic therapy involving family compared to group substance abuse therapy for drug‐using adolescents with co‐occurring mental health problems | |||||
| Patient or population: drug‐using adolescents with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with group substance abuse therapy | Risk difference with multi‐systemic therapy involving family | ||||
| Arrests | 112 | ⊕⊕⊝⊝ | ‐ | Mean arrests were 1.19 SD | MD ‐0.24 SD lower |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by one for risk of bias (selective reporting of outcomes) and by one for imprecision. | |||||
| Interpersonal psychotherapy compared to a psychoeducational intervention for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with a psychoeducational intervention | Risk difference with interpersonal psychotherapy | ||||
| Substance abuse relapse post release | 38 | ⊕⊕⊝⊝ | RR 0.67 | Study population | |
| 47 per 100 | 16 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by two for optimal size not met. bDowngraded by one for risk of bias (selective reporting outcomes). | |||||
| Legal defence service and wrap‐around social work services compared to legal defence service only for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with legal defence services only | Risk difference with legal defence services and wrap‐around social work services | ||||
| Committing new offences Follow‐up: 12 months | 29 | ⊕⊕⊝⊝ | RR 0.64 | Study population | |
| 1 per 100 | 2 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by two for optimal size not met. bDowngraded for risk of bias (incomplete outcome data). | |||||
Background
This review is part of a family of three reviews providing a close examination of what works in reducing drug use and criminal activity among drug‐using offenders. These three reviews report on trials generating several publications and numerous comparisons (Perry forthcominga; Perryforthcomingc). Two of the three reviews represent a specific interest in pharmacological interventions and interventions for female offenders. All three reviews stem from a previous Cochrane systematic review (Perry 2006). We consider the effectiveness of interventions based on two key outcomes ‐ drug use and criminal activity. We have presented here the revised method for this individual review, focusing on the impact of interventions for drug‐using offenders with co‐occurring mental health problems.
Description of the condition
People involved in the criminal justice system are more likely to experience mental health problems. Many studies report different prevalence figures dependent upon the methods used to estimate prevalence (Fazel 2016). Some studies report generic figures that represent all serious mental health problems ‐ e.g. over half (64%) of jail inmates in the United States reporting serious mental health problems (Glase 2006) ‐ and others attempt to break down different types of mental health diagnoses (e.g. psychosis vs major depression). In a systematic review of 33,000 prisoners, one in seven prisoners had major depression or psychosis, with little change in rates of diagnoses over the past three decades (Fazel 2012).
Differences in the prevalence of mental health problems differentiate between males and females and by age. One study of mental health problems in jails found that more women than men (31% and 14.5%, respectively) have a serious mental health problem (Steadman 2009), and one estimate suggests that two‐thirds of juveniles in detention custody have a mental health disorder severe enough to limit their ability to function (Shufelt 2006). Moreover, violent female offenders were found to be five times more likely than male offenders to present with anxiety disorders (Waserman 2005). Other studies have reported that a greater proportion of people who have mental health problems are more likely to be arrested compared with the general population (Lamb 1998).
We also know that rates of comorbidity between mental health problems and substance misuse are high (Butler 2011). Such comorbidity worsens the prognosis of the individual psychiatric disorder and increases the likelihood of repeat offending and premature mortality after release (Chang 2015). Despite these difficulties, it is unknown how well interventions devised to deal with this comorbidity address these problems (Fazel 2002).
Description of the intervention
Many different treatments for substance misuse (e.g. detoxification, therapeutic communities) have been adopted for use in the criminal justice system. This review includes any intervention that was designed to reduce, eliminate, or prevent relapse to drug use or criminal activity, or both. This goal has resulted in the inclusion of a wide range of treatments, including mental health treatment courts with an assertive case management model, therapeutic communities, motivational interviewing (MI) with cognitive skills, use of multi‐systemic/multi‐dimensional therapy involving families and mindfulness training, legal defence service with wrap‐around social services, and interpersonal psychotherapy .
Case management evolved traditionally to address the needs of prisoner re‐entry programmes covering employment, education, health, housing, and family support via assessment and connection of clients with appropriate services (Austin 1994). Case management in the United States has been applied in Treatment Accountability for Safer Communities programmes (Marlowe 2003b); it has shown initial effectiveness but without systematic evidence in support of the process. In the United Kingdom, similar wrap‐around service provision was developed in the 1980s in an attempt to provide services that were more comprehensive by using a 'joined up' approach (Synder 2012). Wrap‐around service provision requires a team‐based approach that includes the young person, the family, and service providers in developing, implementing, and evaluating each part of any support plan (Wilson 2008).
Mental health treatment courts help to link offenders who would ordinarily be prison‐bound to long‐term community‐based treatment. They rely on mental health assessments, individualised treatment plans, and ongoing judicial monitoring to address both the mental health needs of offenders and the public safety concerns of communities. Like other problem‐solving courts such as drug courts, domestic violence courts, and community courts, mental health courts seek to address the underlying problems that contribute to criminal behaviour. Mental health courts share characteristics with crisis intervention teams, jail diversion programmes, specialised probation and parole caseloads, and a host of other collaborative initiatives intended to address the significant overrepresentation of people with mental illness in the criminal justice system.
Since the 1960s, therapeutic community interventions have been used in the United States in combination with work release programmes to rehabilitate offenders via a supportive environment over a relatively long period. Therapeutic community interventions specifically providing aftercare have modest effects on the reduction of recidivism and drug use (Mitchell 2012a; Pearson 1999), but less is known about the impact of using such schemes with people who have mental health and drug misuse problems that co‐occur (e.g. Sacks 2008).
Cognitive‐behavioural approaches, including self‐monitoring, goal‐setting, self‐control training, interpersonal skills training, relapse prevention, group work, and lifestyle modification, have shown signs of success (Lipsey 2007). Previous research based on systematic reviews has excluded evaluations focusing specifically on the needs of drug‐using offenders and/or mentally disordered offenders, but not for people with co‐occurring mental health and drug misuse problems. Motivatonal interviewing techniques are often employed to promote retention in treatment and are aimed at enhancing motivational change and reducing subsequent re‐offending (McMurran 2009; Smedslund 2011).
Multi‐systemic/multi‐dimensional therapy (MST/MDST) consists of intensive family‐ and community‐based treatment provided to adolescents with serious clinical, social, and emotional difficulties. Research on the effectiveness of MST has failed to produce findings that MST is more effective than other services in preventing restrictive out‐of‐home living arrangements, reducing arrests or convictions, or improving life and family functioning (Littell 2005). The transferability of such schemes has been questioned with variable findings when employed in different countries and contexts (Bogt 2006). MDST has also been employed via the juvenile drug court model, which is designed to address the link between substance abuse and criminal activity; it is compared in current work to manualised group‐based substance abuse treatment (adolescent group treatment (AGT)) (Dakof 2015).
Despite growing knowledge about the effectiveness of treatment programmes for offenders, it appears that no recent systematic review evidence has focused on the effectiveness of treatment for offenders with drug misuse and co‐occurring mental health problems.
How the intervention might work
Interventions delivered to drug‐using offenders under the care of the criminal justice system have varied over time. Case management is used to describe what amounts to a range of diverse practices and supervision models spanning several different services, including probation. Examples of case management have been used to co‐ordinate and integrate all aspects of community supervision, from initial offender needs assessment through to programme delivery and intended completion of the order or sentencing requirement (Partridge 2004). Similarily, wrap‐around care has several strengths in its approach, including the family‐centred and culturally sensitive tailoring of each service plan to needs, values, and talents of the individual person (Synder 2012).
Mental health treatment courts aim to identify clients early on in the criminal process, either at the jail or by court staff such as pretrial service officers or social workers in the public defender's office. Most courts have criteria related to what types of charges, criminal histories, and diagnoses will be accepted. For example, a court may accept only defendants charged with misdemeanours who have no history of violent crimes, and who have an Axis I diagnosis based on recognised diagnostic criteria. Defendants who fit the criteria based on the initial screening are usually given a more comprehensive assessment to determine their interest in participating and their community treatment needs. Defendants who agree to participate receive a treatment plan and other community supervision conditions. Cases are dismissed or the sentence is greatly reduced for those who adhere to their treatment plan for the agreed upon time, usually between six months and two years.
Since the 1960s, therapeutic community interventions have been used in the United States in combination with work release programmes to rehabilitate offenders via a supportive environment over a relatively long period. This usually encompasses the transition between being in prison and working within the community (Prendergast 2011). The ethos of a therapeutic community intervention is to focus on treatment for the whole self (not on the drug abuse per se) and underlying symptomatic problems, with residents instrumental in running the therapeutic community (Mitchell 2012a). These interventions are usually based on group activities provided to address long‐term mental illness, personality disorders, and drug addiction. The approach is usually residential, with clients and therapists living together.
Cognitive‐behavioural therapy (CBT) approaches using programmes based on psychological theory have been employed to try to help people address their offending behaviour and generally have received good support from the literature in their reduction of recidivism. This therapy is often described as a psychosocial intervention that aims to improve mental health. CBT focuses on challenging and changing unhelpful cognitive distortions (e.g. thoughts, beliefs, attitudes) and behaviours, improving emotional regulation, and developing personal coping strategies that target solving current problems. Originally, it was designed to treat depression, but its uses have been expanded to include treatment of various mental health conditions, including anxiety.
Interpersonal psychotherapy (IPT) is a brief, attachment‐focused psychotherapy that centres on resolving interpersonal problems and achieving symptomatic recovery. It is an empirically supported treatment (EST) that follows a highly structured and time‐limited approach and is intended to be completed within 12 to 16 weeks. IPT is based on the principle that relationships and life events impact mood, and that the reverse is also true.
Miller and Rollnick developed motivational interviewing as a process to motivate change in substance abusers (Miller 1991). This technique uses different strategies such as expressing empathy, avoiding arguing for change, and working on ambivalence to strengthen commitment to change. Meta‐analyses support the use of motivational interviewing as a stand‐alone treatment and in combination with more intensive programmes (Vasilaki 2006). Linked to this idea of commitment to change is the idea of self‐control, which has established links between substance use and antisocial behaviour (Malouf 2014). The theory suggests that use of mindfulness involves greater self‐awareness, which may promote thoughtful rather than reactive responding and might help to improve mood and problem behaviour (Shonin 2013).
Why it is important to do this review
Many people who are under the care of the criminal justice system have co‐occurring mental health problems and drug misuse problems. Although previous research has broadly evaluated treatment programmes for offenders, we know little about the challenges, treatments, and rehabilitation opportunities for offenders with co‐occurring mental health and drug misuse problems. We therefore believe that an evaluation of existing evidence on the impact of interventions for drug‐using offenders with co‐occurring mental health problems might be helpful in identifying treatments for reducing drug use and criminal activity in this vulnerable population.
Objectives
To assess the effectiveness of interventions for drug‐using offenders with co‐occurring mental health problems in reducing criminal activity or drug use, or both.
This review addresses the following questions.
-
Does any treatment for drug‐using offenders with co‐occurring mental health problems reduce drug use?
-
Does any treatment for drug‐using offenders with co‐occurring mental health problems reduce criminal activity?
-
Does the treatment setting (court, community, prison/secure establishment) affect intervention outcome(s)?
-
Does the type of treatment affect treatment outcome(s)?
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled trials (RCTs).
Types of participants
We included people involved in the criminal justice system with co‐occurring mental health problems and drug misuse problems regardless of gender, age, or ethnicity. Drug misuse included any study that referred to participants who used occasionally, were dependent, or were known to abuse drugs. We defined offenders as people who were involved in the criminal justice system. Individuals could reside in special hospitals, prisons, or the community or were diverted from court or placed on arrest referral schemes for treatment. The study setting could change throughout the process of the study. For example, people involved in the criminal justice system could begin in prison but progress through a work release project into a community setting. We judged offenders to have co‐occurring mental health problems when the paper explicitly stated this. We used several different mechanisms to identify study samples with mental health problems, including:
-
diagnostic gold standard tests such as criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM‐IV), or the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD‐10);
-
the nature of the intervention (e.g. mental health court); and/or
-
study authors' descriptions of participants as having a "history of psychiatric health problems" or a "serious mental disorder" with co‐occurring substance misuse.
Types of interventions
Included interventions were designed, wholly or in part, to eliminate or prevent relapse to drug use or criminal activity, or both, among participants. We included a range of interventions in the review.
Experimental interventions included in the review
-
Any pharmacological intervention (e.g. buprenorphine, methadone)
-
Any psychosocial intervention (e.g. therapeutic community, case management, cognitive‐behavioural therapy, interpersonal psychotherapy, motivational interviewing)
Control interventions included in the review
-
No treatment or waiting list control
-
Minimal and/or alternative treatment (e.g. reporting use of a similar but less intense intervention, using a different theoretical approach with the same components and/or a different alternative intervention)
-
Treatment as usual (included any study that reported a combination and/or component of (1) a psychologically based intervention (e.g. anger management, motivational interviewing, counselling, aggression replacement, family therapy), (2) an educational programme (e.g. health, substance abuse education on risky behaviour), and/or (3) life skills (e.g. financial planning, employment skills, computer skills, interpersonal skills in interviews)
Types of outcome measures
Primary outcomes
When papers reported several different follow‐up periods, we reported the longest period, as we believe this measure provides the most conservative estimate of effectiveness. We provided:
-
drug use measures reported as:
-
self‐reported drug use (unspecified drug, specific drug use not including alcohol, Addiction Severity Index composite scores); or
-
biological drug use (measured by drugs testing urine or analysing hair); and
-
-
criminal activity as measured by:
-
self‐reported or officially reported criminal activity (including arrest for any offence, drug offences, and/or re‐incarceration).
-
Search methods for identification of studies
Electronic searches
Updated searches identified records from 2014 to 6 February 2019.
-
Cochrane Central Register of Controlled Trials (CENTRAL; issues to February 2019).
-
MEDLINE (1966 to February 2019).
-
Embase (1980 to February 2019).
-
PsycINFO (1978 to February 2019).
-
SciSearch (Science Citation Index) (1974 to February 2019).
-
Social SciSearch (Social Science Citation Index) (1972 to February 2019).
-
Applied Social Sciences Index and Abstracts (ASSIA; 1987 to February 2019).
-
National Technical Information Service (NTIS; 1964 to March 2014).a
-
Sociological Abstracts (1963 to March 2014).b
-
Healthcare Management Information Consortium (HMIC; to February 2019).
-
Public Affairs Information Service (PAIS; 1972 to February 2019).
-
Criminal Justice Abstracts (1968 to February 2019).
-
Latin American Caribbean Health Sciences Literature (LILACS; 2004 to February 2019).
-
Current Controlled Trials (December 2009).c
-
SPECTR (March 2004).d
-
Cumulative Index to Nursing and Allied Health Literature (CINHAL)plus (up until February 2019).
aPaid access only ‐ insufficient resources to search.
bNot available to search through York University.
cNo longer available to search.
dNo public access through Campbell Collaboration website, which previously hosted the database.
To update the review, we restricted the search to studies that were published since the end date of the previous search (May 2014). We did not search several original databases indicated by the key at the end of the database list. One database (NTIS) was fee charging, and the other three databases (Sociological Abstracts, Current Controlled Trials, and SPECTR) were not available for searching due to changes in the provision of databases through the University of York.
We developed search strategies for each database to exploit the search engine most effectively and to make use of any controlled vocabulary. We included methodological search filters designed to identify RCTs. Whenever possible, we used filters retrieved from the InterTASC Information Specialists' Sub‐Group (ISSG) Search Filter Resource site (www.york.ac.uk/inst/crd/intertasc/). If filters were unavailable from this site, we substituted search terms based on existing versions. We did not place any language restrictions on identification and inclusion of studies in the review.
We have listed details of the update search strategies and results and the websites searched in Appendix 1,Appendix 2,Appendix 3,Appendix 4,Appendix 5,Appendix 6,Appendix 7,Appendix 8,Appendix 9,Appendix 10, and Appendix 11.
Searching other resources
Reference checking
We scrutinised the reference lists of all retrieved articles for additional references and searched the catalogues of relevant organisations.
Personal communication
We sought out experts for their knowledge of other published or unpublished studies relevant to the review.
Data collection and analysis
Selection of studies
A team of review authors independently inspected the search hits by reading the titles and abstracts. Each potentially relevant study was obtained as a full‐text article. Each article was independently assessed for inclusion. In the case of discordance, a third independent review author arbitrated. One review author undertook translation of articles not written in the English language.
We divided the screening process into two key phases. Phase one used eight key questions as reported in the original review.
Prescreening criteria: phase one
-
Is the document an empirical study? If not, exclude the document
-
Does the study evaluate an intervention, a component of which is designed to reduce, eliminate, or prevent relapse with drug‐using offenders?
-
Are participants referred by the criminal justice system at baseline?
-
Does the study report pre‐ and post‐programme measures of drug use?
-
Does the study report pre‐ and post‐programme measures of criminal behaviour?
-
Is the study an RCT?
-
Do the outcome measures refer to the same length of follow‐up for the two groups?
Papers included after phase one screening were then scrutinised for further assessment.
Prescreening criteria: phase two
-
Does the study population comprise wholly participants with diagnosed mental health problems using DSM‐IV or ICD‐10 diagnostic criteria? if yes, include the document
-
Does the study population comprise wholly participants identified on screening to have a mental health problem(s) based on intervention eligibility (e.g. mental health court)? if yes, include the document
-
When the full study population does not comprise offenders with diagnosed or presumed mental health problems, are separate results given for those participants with mental health problems? if no, exclude the document
Data extraction and management
We used data extraction forms to standardise the reporting of data from all studies obtained as potentially relevant. Two review authors independently extracted data and subsequently checked them for agreement. The narrative tables presented study details (e.g. author, year of publication, country of study origin), study methods (e.g. random assignment), participants (e.g. number in sample, age, gender, ethnicity, age, mental health status), interventions (e.g. description, duration, intensity, setting), outcomes (e.g. description, follow‐up period, reporting mechanism), and notes (e.g. country, funding).
Assessment of risk of bias in included studies
The team of review authors independently assessed risk of bias of all included studies using the 'Risk of bias' assessment criteria recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).
The recommended approach for assessing risk of bias in studies included in Cochrane Reviews involves a two‐part tool that addresses four specific domains, namely, sequence generation and allocation concealment (selection bias), blinding of outcome assessors (detection bias), incomplete outcome data (attrition bias), and selective outcome reporting (reporting bias). The first portion of the tool involves describing what was reported to have happened in the study. The second portion of the tool involves assigning a judgement related to the risk of bias for that entry, in terms of low, high, or unclear risk. To make these judgements, we used the criteria indicated by the Cochrane Handbook for Systematic Reviews of Interventions, as adapted to the addiction field. See Appendix 12 for details.
The domains of sequence generation and allocation concealment (avoidance of selection bias) were addressed in the tool by a single entry for each study.
Participants and personnel cannot be blinded to the type of intervention; moreover, we think that being aware of receiving a psychosocial treatment is in itself part of the therapeutic effect; for these reasons, we did not assess risk of performance bias.
Detection bias was considered separately for objective outcomes (e.g. dropout, use of substance of abuse measured by urine analysis, participants relapsed at end of follow‐up, participants engaged in further treatments) and subjective outcomes (e.g. duration and severity of signs and symptoms of withdrawal, participant self‐reported use of substance, side effects, social functioning as integration at school or at work, family relationship).
Incomplete outcome data (avoidance of attrition bias) was considered for all outcomes except for dropout from treatment, which is very often the primary outcome measure in trials on addiction.
For studies identified in the search, the review authors attempted to contact study authors to establish whether a study protocol was available.
Measures of treatment effect
The mean differences (MD) with 95% confidence intervals (CIs) was used for continuous outcomes measured on the same scale, and the standardised mean difference (SMD) was used for continuous outcomes measured on different scales. Higher scores for continuous measures are representative of greater harm. We presented dichotomous outcomes as risk ratios (RRs), with 95% CIs.
Unit of analysis issues
To avoid double‐counting of outcome measures (e.g. arrest, parole violation) and follow‐up periods (e.g. 12 months, 18 months), we checked all trials to ensure that multiple studies reporting the same evaluation did not contribute towards multiple estimates of programme effectiveness. We followed Cochrane guidance, and where appropriate, we combined intervention and control groups to create a single pair‐wise comparison. When this was not appropriate, we selected one treatment arm and excluded the others.
Dealing with missing data
We attempted to contact study authors via email when we noted missing data in the original publication.
Assessment of heterogeneity
We assessed heterogeneity using the I² statistic and the Chi² statistic (Higgins 2011). We regarded heterogeneity as substantial if I² was greater than 50% or if the P value was lower than 0.10 for the Chi² test for heterogeneity (Deeks 2017). In keeping with the guidance provided in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2017), we distinguished the following values to denote no important heterogeneity and moderate, substantial, and considerable heterogeneity, respectively: 0% to 40%, 30% to 60%, 50% to 90%, and 75% to 100%.
Data synthesis
We used the RevMan software package to perform a series of meta‐analyses for continuous and dichotomous outcome measures (RevMan 2012). We used a random‐effects model to account for the fact that participants did not come from a single underlying population. We combined two studies of the therapeutic community and aftercare in comparison to treatment as usual.
Subgroup analysis and investigation of heterogeneity
We had planned to conduct sensitivity analyses to assess the impact of studies at high risk of bias compared with those at low or unclear risk of bias. Because of the overall high risk of bias of the included studies, this analysis was not possible.
Grading of evidence and 'Summary of findings' tables
We assessed the overall quality of the evidence for the following primary outcomes using the GRADE system: relapse, frequency of use, extent of use, any adverse events, and dropout from treatment. The GRADE Working Group developed a system for grading the quality of evidence (Schunemann 2013); this system takes into account issues related not only to internal validity but also to external validity, such as directness of results.
We have presented the main findings of the review in a 'Summary of findings' table. This transparent and simple tabular format provides key information concerning quality of evidence, magnitude of effect of the interventions examined, and sums of available data for the main outcomes.
The GRADE system uses the following criteria for assigning grades of evidence.
-
High: we are very confident that the true effect lies close to that of the estimate of the effect.
-
Moderate: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
-
Low: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
-
Very low: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
Grading is decreased for the following reasons.
-
Serious (‐1) or very serious (‐2) study limitations for risk of bias.
-
Serious (‐1) or very serious (‐2) inconsistency between study results.
-
Some (‐1) or major (‐2) uncertainty about directness (correspondence between the population, the intervention, or the outcomes measured in the studies actually found and those under consideration in our systematic review).
-
Serious (‐1) or very serious (‐2) imprecision of the pooled estimate.
-
Publication bias strongly suspected (‐1).
Results
Description of studies
Results of the search
As shown in Figure 1, our update searches identified 9653 records. We screened out 9424 references based on titles and abstracts. We examined the remaining 229 records in full text and excluded 224 of them (see Characteristics of excluded studies). We included five new trials,(Cullen 2012; Dakof 2015; Malouf 2017; McCarter 2016; Sundell 2008), along with one follow‐up study to an existing trial within the review (Lanza 2014), and we included three ongoing trials (Baldus 2011; Tinland 2013; VanDorn 2017), along with eight studies from the previous review. The total number of included studies is 13 (see Characteristics of included studies).
Study flow diagram.
Included studies
Population
The 13 included trials randomised a total of 2606 participants and were published between 1999 and 2017. Seven of the 13 trials included adult drug‐using offenders. Three studies investigated the impact on interventions with adolescents and/or youth (Dakof 2015; McCarter 2016; Stein 2011). Two studies included females only (Johnson 2012; Sacks 2008). Three studies reported on juveniles or youth involved in the criminal justice system (Dakof 2015; McCarter 2016; Stein 2011). Adult male offenders were the focus of study populations in the remaining studies, with a mean age of 30 years. In all but two studies (Cullen 2012; McCarter 2016), most participants were of white ethnic origin.
Mental health diagnoses varied across studies (see Table 1.
| Study, year | Criteria used for diagnoses | Description of mental health problem |
| Determined by a psychiatrist/psychologist on the basis of a clinical interview and observations | Mood disorder Schizophrenia Bipolar disorder Other Dual diagnosis | |
| Primary clinical diagnosis of a psychotic disorder. Diagnosis mechanism not reported | Schizophrenia Schizoaffective disorder Bipolar disorder Other psychotic disorder | |
| Diagnostic Interview Schedule for Children (DISC‐2) ‐ identifying presence of mental disorders according to the DSM‐III | Presence of mental health disorders Externalising subscales | |
| Hamilton Rating Scale for Depression Median duration of index episode in months Number of depressive episodes Number of previous suicide attempts DSM‐IV Axis I disorders using the SCID‐I/II | Criteria for a major depressive disorder at least 4 weeks after substance abuse treatment Minimum score of 18 on the Hamilton Rating Scale for Depression | |
| DSM‐IV Mini International Neuropsychiatric Interview Anxiety Sensitivity Index | Anxiety Mental health disorders Antisocial personality disorder Major depressive disorder Generalised anxiety disorder | |
| Borderline Personality Disorder Features assessed with the Personality Assessment Inventory | Affective instability Identity problems Negative relationships Impulsivity | |
| Youth Self‐Report that contain scales orientated to the DSM‐IV | Somatic complaints Anxiety and depression Social problems Internalising and externalising (thought and attention problems) | |
| DIS | Diagnosis of lifetime Axis I or Axis II mental disorder Antisocial personality disorder | |
| Global Severity Index Beck Depression Inventory Lifetime of mental health PTSD Symptom Scale ‐ Interview Posttraumatic Stress Diagnostic Scale | Depression PTSD Lifetime of mental health | |
| DSM‐IV diagnostic criteria Beck Depression Inventory Post Traumatic Stress Disorder Symptom Scale Brief Symptom Inventory Global Severity Index | Depression PTSD Psychological distress | |
| CES‐D Scale | Scores > 16 indicate presence of significant depression; 69.8% had significant depressive symptoms | |
| DSM‐IV diagnostic criteria Youth Self‐Report | Conduct disorder Internalising and externalising Total behaviour problems | |
| Not specified | Antisocial personality disorder Phobias PTSD Depression Dysthymia Attention deficit hyperactivity disorder |
CES‐D: Center for Epidemiological Studies ‐ Depression; DIS: Diagnostic Interview Schedule; DSM‐IV: Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; PTSD: post‐traumatic stress disorder; SCID: Structured Clinical Interview for DSM Disorders.
Settings
Eight studies were conducted in a secure setting (Johnson 2012; Lanza 2014; Malouf 2017; Sacks 2004; Sacks 2008; Sacks 2011; Stein 2011; Wexler 1999), two studies were conducted in community settings (Cosden 2003; Sundell 2008), and two studies were conducted in court settings (Dakof 2015; McCarter 2016). One study was conducted with a medium forensic secure hospital population in the United Kingdom (Cullen 2012). Studies were published in the United States (n = 10/13; 76%), Spain (n = 1/13; 7.6%), the United Kingdom (n = 1/13; 7.6%), and Sweden (n = 1/13; 7.6%).
Duration of trials
Trial duration varied between three‐month follow‐up in Johnson 2012, Lanza 2014, Stein 2011, and Sundell 2008, and five‐year follow‐up in Wexler 1999. Six‐month follow‐up was reported in Cosden 2003,Dakof 2015, and Sacks 2008. The remaining studies reported on outcomes at 12, 24, and 36 months (Cosden 2003; Cullen 2012; Dakof 2015; Malouf 2017; McCarter 2016; Sacks 2011; Sacks 2004). Treatment duration was most intensive (e.g. lasting between three and seven days per week) when a therapeutic community model was employed for periods of up to 12, 18, and 24 months (e.g. Sacks 2004; Sacks 2008; Sacks 2011); typically all other treatment interventions lasted between four and six months (e.g. Cullen 2012; Lanza 2014). The shortest treatment intervention was delivered in a 90‐minute session followed by a 60‐minute booster session upon release (Stein 2011).
Outcome measures
A total of 5 of 13 (38%) trials reported drug outcomes (Cullen 2012; Johnson 2012; Lanza 2014; Stein 2011; Sundell 2008), 5 of 13 (38%) reported crime outcomes (Dakof 2015; McCarter 2016; Sacks 2004; Sacks 2011; Wexler 1999), and 3 of 13 (23%) reported both drug and crime outcomes (Cosden 2003; Malouf 2017; Sacks 2008).
Interventions
Therapuetic interventions and aftercare
Four studies compared a therapeutic community (TC) intervention with aftercare versus treatment as usual (Sacks 2004; Sacks 2011), another intervention (Sacks 2008), or no intervention (Wexler 1999). Sacks 2004 compared a modified TC residential treatment programme using a cognitive‐behavioural curriculum to change attitudes and lifestyles versus a programme of intensive psychiatric services with medication, weekly individual therapy and counselling, and specialised groups of cognitive‐behavioural work, anger management, therapy and education, domestic violence, parenting, and weekly drug/alcohol therapy sessions.
Sacks 2008 evaluated a modified TC group with programme activities supplemented by peer‐led activities on weekends in comparison to an intensive outpatient programme that consisted of an educational programme on substance abuse treatment.
Sacks 2011 consisted of a re‐entry residential TC programme where participants worked in the community and saved money for independent living. This was compared to participants who were released to a community corrections facility during the day; they left the facility to go to work, receive treatment, and report to parole officers. This group engaged with brokering community‐based services and directly received support and counselling services. A weekly relapse prevention group and daily medication monitoring were provided. Psychiatric and substance abuse services were provided by outside agencies (community parole officers helped clients choose). The Wexler study compared a TC treatment programme with aftercare in the community versus a waiting list control.
Mental health court
One study compared use of a mental health court and case management to treatment as usual (Cosden 2003). The mental health treatment court (MHTC) consisted of case management and assertive community treatment (ACT) provided via a case management model. This model included weekly or bi‐weekly court supervision and frequent contact with case managers, followed by treatment as usual (if required), and compared this to treatment as usual, which included traditional court proceedings and county mental health services (Cosden 2003).
Motivational interviewing, mindfulness, and cognitive skills
Four studies compared motivational interviewing, mindfulness, and cognitive skills to no intervention (Lanza 2014), another intervention (Stein 2011), or treatment as usual (Cullen 2012; Malouf 2017). Stein 2011 was a manualised motivational intervention focused on empathy ‐ not arguing and developing discrepancy; self‐efficacy; and personal choice, and compared this approach to a relaxation intervention that included progressive muscle relaxation, use of guided imagery, and feedback on use of techniques.
Malouf 2017 used a manualised group intervention for jail inmates nearing release into the community. The intervention incorporated and adapted elements from several mindfulness‐based interventions (MBIs), including acceptance and commitment therapy, mindfulness‐based relapse prevention (MBRP), and dialectical behavioural therapy (DBT), and was compared to programmes that were normally available within the prison (e.g. anger management financial planning, health education).
Lanza 2014 used cognitive‐behavioural therapy (CBT) to change behaviour through cognitive restructuring and compared to ACT, which aimed to construct an alternative context in which behaviour aligned with one’s values is more likely to occur.
Multi‐systemic therapy including families
Two studies compared multi‐systemic therapy including families versus treatment as usual (in Sundell 2008) and another intervention (in Dakof 2015). Sundell 2008 compared an intensive family‐ and community‐based treatment to support prosocial development versus individual counselling, family therapy, addiction treatment, and special education services.
Dakof 2015 compared an intervention that involved therapists who worked individually with each family in four areas of treatment (adolescent, parent, family, and community) versus adolescent group therapy based on cognitive‐behavioural therapy and motivational interviewing.
Legal defence and social work
One study compared legal defence and wrap‐around social work to legal defence service only (McCarter 2016). The wrap‐around approach provides a collaborative and co‐ordinated response of service providers that organises and streamlines service delivery. This includes attending any team meeting with or on behalf of youth, providing service referrals, and connecting families and guardians to local providers for appropriate mental health, substance abuse, and educational services and support. This was compared to provision of only legal defence service.
Interpersonal psychotherapy
One study compared interpersonal psychotherapy versus another intervention (Johnson 2012). Study participants in the intervention group received manualised group and individual sessions in prison for treatment of substance misuse and mental health problems. These approaches were compared to an attention‐matched manualised in‐prison and post‐release psychoeducation course on mental health and drug problems.
Excluded studies
We excluded 224 full‐text studies. (See Characteristics of excluded studies for further details.) Reasons for exclusion were lack of criminal justice involvement in referral to the intervention; lack of reporting of relevant drug or crime outcome measures, or both, at pre‐ and post‐intervention periods; and allocation of participants to study groups that were not strictly randomised or did not contain original trial data.
Risk of bias in included studies
Allocation
Randomisation
All 13 studies were described as randomised. Nine of the included studies reported on how the randomisation sequence was generated and were judged as having low risk of bias (Cosden 2003; Dakof 2015; Johnson 2012; Lanza 2014; Malouf 2017; McCarter 2016; Sacks 2011; Stein 2011; Sundell 2008). The remaining four studies did not report how the randomisation sequence of participants was generated (Cullen 2012; Sacks 2004; Sacks 2008; Wexler 1999).
Characteristics at baseline
Eight of the 13 studies were similar in terms of drug use at baseline (Cullen 2012; Dakof 2015; Johnson 2012; McCarter 2016; Sacks 2008; Sacks 2011; Stein 2011; Wexler 1999); four studies were rated unclear (Cosden 2003; Lanza 2014; Malouf 2017; Sundell 2008); and one study showed comparable baseline differences (Sacks 2004). For similarity on criminal justice measures, nine studies were rated as similar (Cosden 2003; Cullen 2012; Dakof 2015; Johnson 2012; McCarter 2016; Sacks 2008; Sacks 2011; Sacks 2004; Wexler 1999), and four were rated as unclear (Lanza 2014; Malouf 2017; Stein 2011; Sundell 2008).
Allocation concealment
Of the 13 studies, only two adequately reported that the allocation process was concealed (Johnson 2012; Sundell 2008 ). The remaining 11 (85%) studies were rated as unclear.
Blinding
We assessed risk of detection bias across subjective and objective measures (see Appendix 12). We rated eight studies as having unclear risk (Cosden 2003; McCarter 2016; Sacks 2004; Sacks 2008; Sacks 2011; Stein 2011; Sundell 2008; Wexler 1999); two studies as having low risk (Cullen 2012; Lanza 2014); and the remaining three studies as having high risk of bias.
Incomplete outcome data
Loss to follow‐up was reported to a differing extent in the included studies. We rated six studies as having low risk with limited attrition noted (Johnson 2012; Lanza 2014; Sacks 2004; Stein 2011; Sundell 2008; Wexler 1999); three studies as having unclear risk (Dakof 2015; Sacks 2008; Sacks 2011); and four studies as having high risk of bias (Cosden 2003; Cullen 2012; Malouf 2017; McCarter 2016).
Selective reporting
We rated five of the thirteen trials as having unclear risk of bias (McCarter 2016; Sacks 2004; Stein 2011; Sundell 2008; Wexler 1999); six studies as having low risk (Cosden 2003; Cullen 2012; Lanza 2014; Malouf 2017; Sacks 2008; Sacks 2011), and two studies as having high risk of bias (Dakof 2015; Johnson 2012).
See Figure 2 and Figure 3 for details.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Effects of interventions
See: Summary of findings for the main comparison Therapeutic community compared to treatment as usual for drug‐using offenders with co‐occurring mental illness; Summary of findings 2 Therapeutic community and aftercare compared to cognitive behavioural skills for drug using women offenders with co‐occurring mental illness; Summary of findings 3 Therapeutic community compared to waiting list control for drug‐using offenders with co‐occurring mental illness; Summary of findings 4 Mental health treatment court with assertive case management model compared to treatment as usual for drug‐using offenders with co‐occurring mental illness; Summary of findings 5 Motivational interviewing/mindfulness and cognitive skills compared to relaxation training for drug‐using offenders with co‐occurring mental illness; Summary of findings 6 Motivational interviewing/mindfulness and cognitive skills compared to waiting list control for drug‐using offenders with co‐occurring mental illness; Summary of findings 7 Motivational interviewing/mindfulness and cognitive skills compared to treatment as usual for drug‐using offenders with co‐occurring mental illness; Summary of findings 8 Multi‐systemic therapy involving family and juveniles compared to treatment as usual for drug‐using offenders with co‐occurring mental illness; Summary of findings 9 Multi‐systemic therapy involving family compared to group substance abuse therapy for drug‐using adolescents with co‐occurring mental illness; Summary of findings 10 Interpersonal psychotherapy compared to a psychoeducational intervention for drug‐using offenders with co‐occurring mental illness; Summary of findings 11 Legal defence service and wrap‐around social work services compared to legal defence service only for drug‐using offenders with co‐occurring mental illness
1. Therapeutic community and aftercare versus treatment as usual
See summary of findings Table for the main comparison.
Impact on self‐reported drug use
This was not reported.
Impact on self‐reported criminal activity
Sacks 2011 and Sacks 2004 were combined to show a significant reduction in the number of re‐arrests (risk ratio (RR) 0.67, 95% confidence interval (CI) 0.53 to 0.84) and re‐incarcerations (RR 0.40, 95% CI 0.24 to 0.67), with moderate‐certainty evidence at 12‐month follow‐up (266 participants; see Analysis 1.1).
2. Therapeutic community and aftercare versus cognitive‐behavioural skills for drug‐using women
See summary of findings Table 2.
Impact on self‐reported drug use
Sacks 2008 showed no significant reduction in self‐reported drug use (RR 0.78, 95% CI 0.46 to 1.32), with low‐certainty evidence at six‐month follow‐up (314 participants; see Analysis 2.1).
Impact on self‐reported criminal activity
Sacks 2008 showed no significant reduction in re‐arrest for any type of crime (RR 0.69, 95% CI 0.44 to 1.09), criminal activity (RR 0.74, 95% CI 0.52 to 1.05), or drug‐related crime (RR 0.87, 95% CI 0.56 to 1.36), with low‐certainty evidence at six‐month follow‐up (314 participants; see Analysis 2.2,Analysis 2.3, and Analysis 2.4).
3. Therapeutic community versus waiting list control
See summary of findings Table 3.
Impact on self‐reported drug use
This was not reported.
Impact on self‐reported criminal activity
Wexler 1999 showed a significant reduction (but a trend towards favouring) return to prison in favour of the therapeutic community intervention (RR 0.60, 95% CI 0.46 to 0.79), with moderate‐certainty evidence at 36‐month follow‐up (478 participants; see Analysis 3.1).
4. Mental health treatment court with assertive case management model versus treatment as usual
See summary of findings Table 4.
Impact on self‐reported drug use
Cosden 2003 showed no significant reduction in Addiction Severity Index (ASI)‐self‐reported drug use (mean difference (MD) 0.00, 95% CI ‐0.03 to 0.03), with low‐certainty evidence at 12‐month follow‐up (235 participants; see Analysis 4.3).
Impact on self‐reported criminal activity
Cosden 2003 showed no significant reduction in conviction for a new crime (RR 1.05, 95% CI 0.90 to 1.22) or re‐incarceration to jail (RR 0.79, 95% CI 0.62 to 1.01), with low‐certainty evidence at 12‐month follow‐up (235 participants; see Analysis 4.1 and Analysis 4.2).
5. Motivational interviewing/mindfulness and cognitive skills versus relaxation therapy
See summary of findings Table 5.
Impact on self‐reported drug use
Stein 2011 compared cognitive skills to a relaxation training intervention for adolescents in prison with depressed mood. This study measured marijuana use at three‐month follow‐up assessment using the Risks and Consequences Questionnaire (RCQ). Researchers reported a main effect < .007, with participants in the motivational interviewing group showing fewer problems than participants in the relaxation training group. No further numerical information is available (moderate‐certainty of evidence; 181 participants).
Impact on self‐reported criminal activity
This was not reported.
6. Motivational interviewing/mindfulness and cognitive skills versus waiting list control
See summary of findings Table 6.
Impact on self‐reported drug use
Lanza 2014 reported no significant reduction in self‐reported drug use based on the ASI (MD ‐0.04, 95% CI ‐0.37 to 0.29) and abstinence from drug use (RR 2.89, 95% CI 0.73 to 11.43), with low‐certainty evidence at six months (31 participants; see Analysis 5.1 and Analysis 5.2).
Impact on self‐reported criminal activity
Studies did not assess this outcome.
7. Motivational interviewing/mindfulness and cognitive skills versus treatment as usual
See summary of findings Table 7.
Impact on self‐reported drug use
Malouf 2017 found no significant reduction in frequency of marijuana use (MD ‐1.05, 95% CI ‐2.39 to 0.29), with very low‐certainty evidence at three months post release (40 participants; see Analysis 6.1).
Cullen 2012 found no significant reduction in positive drug screens (MD ‐0.7, 95% CI ‐3.5 to 2.1), with very low‐certainty evidence at 12 months (84 participants; see Analysis 6.4).
Impact on self‐reported criminal activity
Malouf 2017 found a significant reduction in frequency of re‐arrest (MD ‐0.66, 95% CI ‐1.31 to ‐0.01) but not in time to first arrest (MD 0.87, 95% CI ‐0.12 to 1.86), with very low‐certainty evidence up to 36 months (40 participants; see Analysis 6.2 and Analysis 6.3).
8. Multi‐systemic therapy (involving family) and juveniles versus treatment as usual
See summary of findings Table 8.
Impact on self‐reported drug use
Sundell 2008 found no significant reduction in drug dependence on the Drug Use Disorders Identification Test (DUDIT) score (MD ‐0.22, 95% CI ‐2.51 to 2.07), with low‐certainty evidence up to seven months (156 participants; see Analysis 7.2).
Impact on self‐reported criminal activity
Sundell 2008 found no significant reduction in arrests (RR 0.97, 95% CI 0.70 to 1.36), with low‐certainty evidence up to seven months (158 participants; see Analysis 7.1).
9. Multi‐systemic therapy (involving family) versus adolescent group substance abuse therapy
See summary of findings Table 9.
Impact on self‐reported drug use
This was not reported.
Impact on self‐reported criminal activity
Dakof 2015 reported no significant reduction in re‐arrests (MD ‐0.24, 95% CI ‐0.76 to 0.28), with low‐certainty evidence up to 24 months (112 participants; see Analysis 8.1).
10. Interpersonal psychotherapy versus a psychoeducational intervention
See summary of findings Table 10.
Impact on self‐reported drug use
Johnson 2012 reported no significant reduction in self‐reported drug use (RR 0.67, 95% CI 0.30 to 1.50), with very low‐certainty evidence up to three months (38 participants; see Analysis 9.1).
Impact on self‐reported criminal activity
This was not reported.
11. Legal defence service and wrap‐around social work services versus legal defence service only
See summary of findings Table 11.
Impact on self‐reported drug use
This was not reported.
Impact on self‐reported criminal activity
McCarter 2016 reported no significant reduction in the number of new offences committed (RR 0.64, 95% CI 0.07 to 6.01), with very low‐certainty evidence up to 12 months (29 participants; Analysis 10.1).
Discussion
Summary of main results
This systematic review provides evidence from 13 trials involving 2606 participants and evaluating 11 different comparisons; one pooled analysis was possible. Certainty of evidence was generally low. Most interventions were delivered in prison‐based (eight studies; 61%), court (two studies; 15%), community (two studies; 15%), or medium secure hospital (one study; 8%) settings. Most studies compared an intervention versus treatment as usual or another intervention (11/13 studies; 84%).
The 11 different treatment comparisons were divided into:
-
therapeutic community and aftercare versus treatment as usual (Sacks 2004; Sacks 2011);
-
therapeutic community and aftercare versus a cognitive‐behavioural skills course (Sacks 2008); and
-
therapeutic community and aftercare versus a waiting list control (Wexler 1999).
Two studies comparing therapeutic community interventions reported a significant reduction in subsequent re‐incarceration and criminal activity compared to treatment as usual (Sacks 2004; Sacks 2011), with moderate‐certainty evidence. Sacks 2008 adapted a therapeutic community treatment for women offenders compared to a cognitive‐behavioural skills course. This study compared women assigned to therapeutic community treatment or standard treatment versus a cognitive‐behavioural recovery and relapse prevention curriculum, referred to in the system as the 'intensive outpatient programme' (Sacks 2008), with low‐certainty evidence. At six months, researchers found that both groups improved significantly on variables of mental health, substance use, criminal behaviour, and HIV risk. Study authors noted that further exploration of each model for different offender groups is required to permit more precise utility of each model. They concluded that these preliminary findings suggest the importance of providing gender‐specific sensitive and comprehensive approaches within the correctional system to respond to the complex substance abuse needs of female offenders (Sacks 2008). Therapeutic community treatment was found to be more beneficial than cognitive‐behavioural therapy, lengthening time spent in the community before subsequent re‐incarceration (Sacks 2008). This finding partially supports previous research suggesting that the combination of therapeutic community treatment and aftercare release seems to produce the most consistent and successful results among offenders who do not have co‐occurring mental health problems (Mitchell 2012a). Although this is not addressed within this review, clients who remained in treatment for the longest period appeared to benefit the most (Sacks 2004). These differences seem to be borne out for up to 36 months when compared to people who received nothing, suggesting that over time, the impact of the intervention eventually became dissipated (Wexler 1999), with moderate‐certainty evidence. Only one of the four studies reported on outcomes of drug use (in women) and found no reductions following the intervention in comparison to attending a cognitive skills course (Sacks 2008). We do not know whether drug use is reduced in men with co‐occurring mental health problems.
Mental health treatment court (MHTC) and use of an assertive case management model versus treatment as usual
People under the care of the criminal justice system in both groups showed improvement across a range of outcomes in life satisfaction, a decrease in distress levels, and improvement in independent living. Overall, the pattern of criminal activity suggested that both groups spent time in jail but for different reasons. The individual under the care of the MHTC was more likely to be 'booked' for a crime and not convicted and to have more convictions due to probation violation in comparison to individuals who had received only treatment as usual. Those people receiving treatment as usual were more likely to be convicted of a new offence (Cosden 2003), with low‐certainty evidence.
Motivational interviewing/mindfulness and cognitive skills versus a waiting list control; motivational interviewing/mindfulness and cognitive skills versus relaxation training; motivational interviewing/mindfulness and cognitive skills versus treatment as usual
See Cullen 2012,Lanza 2014,Malouf 2017, and Stein 2011.
Four studies of motivational interviewing/mindfulness and cognitive skills compared to a waiting list control, relaxation training, and treatment as usual reported moderate‐ to very low‐certainty evidence. No significant differences were noted across these studies, suggesting that use of such skills may not reduce subsequent drug use and/or criminal activity in comparison to any of the alternatives. In addition, one of the four studies was a pilot randomised controlled trial (RCT) of motivational interviewing versus treatment as usual, which suggests that larger studies are required to support any future findings. Use of self‐reported measures often contaminated by social desirability bias means that confidence in these results may be limited (moderate‐certainty evidence; Malouf 2017).
Multi‐systemic therapy (MST) involving families versus treatment as usual; MST involving families versus group‐based substance abuse therapy
Two studies of multi‐systemic therapy for juveniles included families and compared treatment as usual or an alternative group‐based substance abuse therapy (Dakof 2015; Sundell 2008). Findings show that MST did not support short‐term effectiveness relative to services usually available for conduct disordered youths in Sweden (Sundell 2008). This outcome is contrary to other work conducted in the United States and Norway but similar to work performed in Canada (Cunningham 2002). Sundell 2008 highlighted the importance of measuring and monitoring fidelity during transportation and delivery of interventions to other settings and different countries worldwide; the importance of the impact of social context should not be underestimated (low‐certainty evidence).
Interpersonal psychotherapy versus psychoeducational intervention
One pilot study of interpersonal psychotherapy in comparison to a psychoeducational intervention showed no significant reduction in subsequent drug use. However, these results should be interpreted with caution, given the small sample and the short follow‐up period (very low‐certainty evidence; Johnson 2012).
Legal defence work and wrap‐around social services versus legal defence work only
One pilot study of legal defence work and wrap‐around social services in comparison to legal defence work with juveniles did not reveal any significant reduction in subsequent return to prison in the 12‐month follow‐up period. Study authors argue that holistic representation services can help to provide protective factors that might strengthen underlying risks and needs of young people, which might contribute to additional court involvement and/or re‐offending in the future (very low‐certainty evidence; McCarter 2016).
Succesful treatment elements and dealing with complex co‐occurring problems
In terms of addressing some of the complex issues of individuals with mental health problems and co‐occurring substance abuse, the evidence from this systematic review provides sparse information.
Several successful treatment elements were reported throughout these trials, and several key themes can be identified.
First, we noted that the issue of treatment engagement was important. In the mental health court trial, informal support from family and friends encouraged the engagement of clients within the community to longer‐term gain, but more research is required to assess whether interventions that empower families can enhance and sustain outcomes longer than non‐family‐based interventions (Cosden 2003; Dakof 2015).
Second, programmes that were specifically adapted to the needs of mental health clients tended to include a cognitive‐behavioural curriculum that emphasised criminal thinking and behaviour alongside psychoeducational classes. The purpose of combining these two types of mechanisms is to enhance an individual's ability to recognise and understand his/her substance misuse and mental health problems in greater detail (Sacks 2004).
Third, the longer an individual is engaged in treatment, the better is the outcome(s) (Wexler 1999).
Overall completeness and applicability of evidence
General applicability
Applicability of the evidence is hindered in general by the range of trials covering various different treatment options, making it inappropriate to pool study results. Most trials were conducted within the US judicial system; therefore, they are limited in their generalisability to criminal justice systems outside the United States. Three trials conducted in Spain (Lanza 2014), the United Kingdom (Cullen 2012), and Sweden provide a European perspective but with moderate‐certainty evidence (Sundell 2008). As a result, study findings must be interpreted with caution.
Mental health information
Although this review specifically sought to identify studies including participants with co‐occurring mental health problems, study descriptions of mental ill health varied (see Table 1). Cosden 2003 used a psychiatrist or a psychologist to conduct a clinical interview to determine a mental health diagnosis alongside substance misuse. This resulted in a mental health court sample of individuals diagnosed with various mental health problems, including mood disorder, schizophrenia, bipolar disorder, and dual diagnosis. Other papers referred to use of Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM‐IV) diagnostic criteria, akin to the Youth Self‐Report measure (Dakof 2015; McCarter 2016; Sacks 2011; Sundell 2008), but subsequently provided little information with regards to individual mental health needs. Demographic information in Sacks 2004 revealed other aspects of mental health prognosis, including lifetime mental health treatment, lifetime patient care, and prescribed medication.
The Wexler 1999 series of studies reported a range of diagnoses, including antisocial personality disorder, phobias, post‐traumatic stress disorder, depression, dysthymia, and attention deficit disorder, but did not describe how these diagnoses were confirmed or assessed within the population.
Six of the 13 trials reported on change in mental health well‐being. Three trials reported on use of the Beck Depression Inventory, the Global Severity Index, and the Posttraumatic Diagnostic Scale (Sacks 2004; Sacks 2008; Sacks 2011). Another study reported on depression but used the Hamilton Rating Scale for Depression (Johnson 2012). Four studies reported the presence of mood disorder alongside schizophrenia, general anxiety disorder, and/or antisocial personality disorder (Cosden 2003; Cullen 2012; Lanza 2014; Malouf 2017). Four studies discussed differential effects of treatment on the severity of depression (Cosden 2003; Johnson 2012; McCarter 2016; Stein 2011). Cosden 2003 noted that further understanding of how to help clients with serious mental health problems through different levels of treatment is needed. Johnson 2012 noted that participants undergoing interpersonal psychotherapy had significantly reduced levels of depression and substance misuse over attention‐matched controls. Study authors noted that the intensity of treatment delivered once the individual is released into the community is key to maintaining good outcomes. However, they go on to state that people under the care of the criminal justice system often experience delays in treatment and service provision on release, and they suggest that alternative services such as phone treatment might be helpful in providing more intensive and useful post‐release treatment in times of crisis.
Quality of the evidence
We rated eight of 13 (62%) studies as having unclear risk of bias in more than four of eight domains. The main limiting factor was lack of reporting evidence, which prevented review authors from making a clear judgement on bias. Given that the imprecision of reporting lowers the quality of evidence, further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. In addition, several specific limitations related to study design (and leading to problems of selection bias) were described, and sample sizes were small. Stein 2011 and Cullen 2012 were noted as being relatively underpowered. Replication of these studies is required to enhance the generalisability and external validity of study findings.
Similar modest sample sizes were reported, with some trials referred to as pilot studies (e.g. Malouf 2017; McCarter 2016). Sacks 2011 and Cosden 2003 suggest that larger samples should be used to provide a more precise estimate of effect. Small sample sizes limit the generalisability of the sample population to other settings, and few studies collected longitudinal data sufficient to support the ongoing use of such schemes without additional larger trials commissioned (Cullen 2012; Dakof 2015; McCarter 2016). Cosden 2003 also reported on the possibility of outcome bias, as the interviewer was not blind to the outcome condition of the participant nor to loss to follow‐up (25% of the study sample were lost to follow‐up) at 12 months.
Another possible selection bias concern in the series of Wexler studies was that participants were randomly assigned to prison therapeutic community treatment and regular prison conditions but not to aftercare (Prendergast 2003; Prendergast 2004; Wexler 1999). Study authors noted that possible differences in personal motivation may account for some of the positive outcomes associated with participants' continued support for aftercare services. Subsequently, these participants were noted as having the highest 'readiness scores', which suggests that motivation creates an important consideration for client selection (Wexler 1999). Cullen 2012 reported on the use of randomisation within sites, which may have led to contamination across treatment groups, and the likelihood that further selection bias might have arisen from the fact that declining patients were more unwell and/or antisocial, and that these factors might influence treatment outcomes (Cullen 2012).
Overall we judged the certainty of evidence as moderate to very low for all included interventions.
Potential biases in the review process
Besides the limitations already discussed, the search method was limited to databases that could be accessed via the University of York, and extensive website searches were not conducted. We did not search specific trial registers. As a result, some literature may have been omitted from this updated version.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Comparison 1 Therapeutic community and aftercare vs treatment as usual, Outcome 1 Criminal activity.
Comparison 2 Therapeutic community and aftercare vs cognitive‐behavioural therapy, Outcome 1 Self‐reported drug use at 6 months.
Comparison 2 Therapeutic community and aftercare vs cognitive‐behavioural therapy, Outcome 2 Arrested any for 6 months.
Comparison 2 Therapeutic community and aftercare vs cognitive‐behavioural therapy, Outcome 3 Criminal activity at 6 months.
Comparison 2 Therapeutic community and aftercare vs cognitive‐behavioural therapy, Outcome 4 Drug‐related crime.
Comparison 3 Therapeutic community vs waiting list control, Outcome 1 Re‐incarceration at 36 months.
Comparison 4 Mental health treatment court with assertive case management vs treatment as usual, Outcome 1 Committing a new crime.
Comparison 4 Mental health treatment court with assertive case management vs treatment as usual, Outcome 2 Re‐incarceration to jail at 12 months.
Comparison 4 Mental health treatment court with assertive case management vs treatment as usual, Outcome 3 ASI drug use at 12 months.
Comparison 5 Motivational interviewing and cognitive skills vs waiting list control, Outcome 1 Abstinence from drug use at 6 months.
Comparison 5 Motivational interviewing and cognitive skills vs waiting list control, Outcome 2 ASI drug score at 6 months.
Comparison 6 Motivational interviewing and cognitive skills vs treatment as usual, Outcome 1 Marjuana frequency at 3 months.
Comparison 6 Motivational interviewing and cognitive skills vs treatment as usual, Outcome 2 Arrest frequency 3 years post release.
Comparison 6 Motivational interviewing and cognitive skills vs treatment as usual, Outcome 3 Time to first arrest or offence 36 months post.
Comparison 6 Motivational interviewing and cognitive skills vs treatment as usual, Outcome 4 Positive drug screen at 12 months.
Comparison 7 Multi‐systemic therapy vs treatment as usual, Outcome 1 Arrest by police.
Comparison 7 Multi‐systemic therapy vs treatment as usual, Outcome 2 DUDIT scores.
Comparison 8 Multi‐systemic therapy vs adolescent substance treatment group, Outcome 1 Arrests between 6 and 24 months.
Comparison 9 Interpersonal psychotherapy vs psychoeducational controls, Outcome 1 Substance abuse relapse at 3 months.
Comparison 10 Legal defence services and wrap‐around social work services vs legal defence work only, Outcome 1 Number of new offences committed at 12 months.
| Therapeutic community compared to treatment as usual for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with treatment as usual | Risk difference with therapeutic community | ||||
| Re‐arrests | 266 | ⊕⊕⊕⊝ | RR 0.67 | Study population | |
| 98 per 100 | 32 fewer per 100 | ||||
| Re‐incarceration | 266 | ⊕⊕⊕⊝ | RR 0.40 | Study population | |
| 59 per 100 | 36 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by one for risk of bias (blinding and selective reporting). | |||||
| Therapeutic community and aftercare compared to cognitive‐behavioural skills for drug‐using women offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using women offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with cognitive‐behavioural skills | Risk difference with therapeutic community and aftercare | ||||
| Self‐reported drug use | 314 | ⊕⊝⊝⊝ | RR 0.78 | Study population | |
| 17 per 100 | 4 fewer per 100 | ||||
| Re‐arrest for any type of crime | 314 | ⊕⊝⊝⊝ | RR 0.69 | Study population | |
| 33 per 100 | 10 fewer per 100 | ||||
| Criminal Activity | 314 | ⊕⊝⊝⊝ | RR 0.74 | Study population | |
| 33 per 100 | 9 fewer per 100 | ||||
| Drug‐related crime | 314 | ⊕⊝⊝⊝ | RR 0.87 | Study population | |
| 21 per 100 | 3 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aUnclear reporting in the paper raises concerns about the potential high risk of bias with regards to blinding and methods used in the randomisation procedure; we downgraded by one. bOne study with 95% confidence intervals through the line of no effect. | |||||
| Therapeutic community compared to waiting list control for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with waiting list control | Risk difference with therapeutic community | ||||
| Return to prison (recidivism) post parole | 478 | ⊕⊕⊕⊝ MODERATEa | RR 0.60 | Study population | |
| 40 per 100 | 16 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by one for risk of bias (randomisation process, concealment, and selective reporting). | |||||
| Mental health treatment court with assertive case management model compared to treatment as usual for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with treatment as usual | Risk difference with mental health treatment court with assertive case management model | ||||
| Conviction for a new crime | 235 | ⊕⊝⊝⊝ | RR 1.05 | Study population | |
| 72 per 100 | 4 more per 100 | ||||
| Re‐incarceration to jail | 235 | ⊕⊝⊝⊝ | RR 0.79 | Study population | |
| 71 per 100 | 15 fewer per 100 | ||||
| Self‐reported drug use | 235 | ⊕⊝⊝⊝ | ‐ | Mean self‐reported drug use was 0.08 | MD 0.00 |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by one for risk of bias (allocation concealment and blinding of assessors) and by one for imprecision. | |||||
| Motivational interviewing/mindfulness and cognitive skills compared to relaxation training for drug‐using offenders with co‐occurring mental health problems | |||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||
| Outcomes | № of participants | Certainty of the evidence | Impact |
| Self‐reported marijuana use continuous | 181 | MODERATEa | This study compared cognitive skills to a relaxation training intervention for adolescents in prison with depressed mood. Researchers measured marijuana use at 3‐months follow‐up assessment using the Risks and Consequences Questionnaire (RCQ). They report a main effect < .007, with participants in the motivational interviewing group showing fewer problems than participants in the relaxation training group. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||
| GRADE Working Group grades of evidence. | |||
| aDowngraded by one for unclear risk of bias (random allocation and blinding). | |||
| Motivational interviewing/mindfulness and cognitive skills compared to waiting list control for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with waiting list control | Risk difference with motivational interviewing and cognitive skills | ||||
| Self‐reported drug use | 31 | ⊕⊕⊝⊝ | ‐ | Mean self‐reported drug use was 0.44 | MD ‐0.04 lower |
| Abstinence from drug use | 31 | ⊕⊕⊝⊝ | RR 2.89 | Study population | |
| 15 per 100 | 29 more per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by two for optimal information size not met. | |||||
| Motivational interviewing/mindfulness and cognitive skills compared to treatment as usual for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with treatment as usual | Risk difference with motivational interviewing and cognitive skills | ||||
| Self‐reported frequency of marijuana use | 40 | ⊕⊕⊝⊝ VERY LOWa,b | ‐ | Mean self‐reported frequency of marijuana use was 1.50 | MD ‐1.05 lower |
| Arrest frequency post release | 40 | ⊕⊕⊝⊝ | ‐ | Mean arrest frequency post release was 1.47 | MD ‐0.66 lower |
| Time to first arrest or offence | 40 | ⊕⊕⊝⊝ | ‐ | Mean time to first arrest or offence was 1.6 | MD 0.87 higher |
| Positive drug screen or refusal to provide a urine sample | 84 | ⊕⊝⊝⊝ | ‐ | Mean positive drug screen or refusal to provide a urine sample was 3.25 | MD ‐0.7 lower |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by two for optimal size not met. bDowngraded by one for risk of bias (incomplete outcome measures). | |||||
| Multi‐systemic therapy involving family compared to treatment as usual for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with treatment as usual | Risk difference with multi‐systemic therapy involving family | ||||
| Drug dependence | 156 | ⊕⊕⊝⊝ | ‐ | Mean drug dependence was 3.55 | MD ‐0.22 lower |
| Arrested | 158 | ⊕⊕⊝⊝ | RR 0.97 | Study population | |
| 47 per 100 | 1 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by one for risk of bias (blinding measures) and downgraded by one for imprecision. | |||||
| Multi‐systemic therapy involving family compared to group substance abuse therapy for drug‐using adolescents with co‐occurring mental health problems | |||||
| Patient or population: drug‐using adolescents with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with group substance abuse therapy | Risk difference with multi‐systemic therapy involving family | ||||
| Arrests | 112 | ⊕⊕⊝⊝ | ‐ | Mean arrests were 1.19 SD | MD ‐0.24 SD lower |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by one for risk of bias (selective reporting of outcomes) and by one for imprecision. | |||||
| Interpersonal psychotherapy compared to a psychoeducational intervention for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with a psychoeducational intervention | Risk difference with interpersonal psychotherapy | ||||
| Substance abuse relapse post release | 38 | ⊕⊕⊝⊝ | RR 0.67 | Study population | |
| 47 per 100 | 16 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by two for optimal size not met. bDowngraded by one for risk of bias (selective reporting outcomes). | |||||
| Legal defence service and wrap‐around social work services compared to legal defence service only for drug‐using offenders with co‐occurring mental health problems | |||||
| Patient or population: drug‐using offenders with co‐occurring mental health problems | |||||
| Outcomes | № of participants | Certainty of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
| Risk with legal defence services only | Risk difference with legal defence services and wrap‐around social work services | ||||
| Committing new offences Follow‐up: 12 months | 29 | ⊕⊕⊝⊝ | RR 0.64 | Study population | |
| 1 per 100 | 2 fewer per 100 | ||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence. | |||||
| aDowngraded by two for optimal size not met. bDowngraded for risk of bias (incomplete outcome data). | |||||
| Study, year | Criteria used for diagnoses | Description of mental health problem |
| Determined by a psychiatrist/psychologist on the basis of a clinical interview and observations | Mood disorder Schizophrenia Bipolar disorder Other Dual diagnosis | |
| Primary clinical diagnosis of a psychotic disorder. Diagnosis mechanism not reported | Schizophrenia Schizoaffective disorder Bipolar disorder Other psychotic disorder | |
| Diagnostic Interview Schedule for Children (DISC‐2) ‐ identifying presence of mental disorders according to the DSM‐III | Presence of mental health disorders Externalising subscales | |
| Hamilton Rating Scale for Depression Median duration of index episode in months Number of depressive episodes Number of previous suicide attempts DSM‐IV Axis I disorders using the SCID‐I/II | Criteria for a major depressive disorder at least 4 weeks after substance abuse treatment Minimum score of 18 on the Hamilton Rating Scale for Depression | |
| DSM‐IV Mini International Neuropsychiatric Interview Anxiety Sensitivity Index | Anxiety Mental health disorders Antisocial personality disorder Major depressive disorder Generalised anxiety disorder | |
| Borderline Personality Disorder Features assessed with the Personality Assessment Inventory | Affective instability Identity problems Negative relationships Impulsivity | |
| Youth Self‐Report that contain scales orientated to the DSM‐IV | Somatic complaints Anxiety and depression Social problems Internalising and externalising (thought and attention problems) | |
| DIS | Diagnosis of lifetime Axis I or Axis II mental disorder Antisocial personality disorder | |
| Global Severity Index Beck Depression Inventory Lifetime of mental health PTSD Symptom Scale ‐ Interview Posttraumatic Stress Diagnostic Scale | Depression PTSD Lifetime of mental health | |
| DSM‐IV diagnostic criteria Beck Depression Inventory Post Traumatic Stress Disorder Symptom Scale Brief Symptom Inventory Global Severity Index | Depression PTSD Psychological distress | |
| CES‐D Scale | Scores > 16 indicate presence of significant depression; 69.8% had significant depressive symptoms | |
| DSM‐IV diagnostic criteria Youth Self‐Report | Conduct disorder Internalising and externalising Total behaviour problems | |
| Not specified | Antisocial personality disorder Phobias PTSD Depression Dysthymia Attention deficit hyperactivity disorder | |
| CES‐D: Center for Epidemiological Studies ‐ Depression; DIS: Diagnostic Interview Schedule; DSM‐IV: Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; PTSD: post‐traumatic stress disorder; SCID: Structured Clinical Interview for DSM Disorders. | ||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Criminal activity Show forest plot | 2 | 532 | Risk Ratio (M‐H, Random, 95% CI) | 0.57 [0.42, 0.77] |
| 1.1 Any criminal activity | 2 | 266 | Risk Ratio (M‐H, Random, 95% CI) | 0.67 [0.53, 0.84] |
| 1.2 Re‐incarceration | 2 | 266 | Risk Ratio (M‐H, Random, 95% CI) | 0.40 [0.24, 0.67] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Self‐reported drug use at 6 months Show forest plot | 1 | 314 | Risk Ratio (M‐H, Random, 95% CI) | 0.78 [0.46, 1.32] |
| 2 Arrested any for 6 months Show forest plot | 1 | 314 | Risk Ratio (M‐H, Random, 95% CI) | 0.69 [0.44, 1.09] |
| 3 Criminal activity at 6 months Show forest plot | 1 | 314 | Risk Ratio (M‐H, Random, 95% CI) | 0.74 [0.52, 1.05] |
| 4 Drug‐related crime Show forest plot | 1 | 314 | Risk Ratio (M‐H, Random, 95% CI) | 0.87 [0.56, 1.36] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Re‐incarceration at 36 months Show forest plot | 1 | 478 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.60 [0.46, 0.79] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Committing a new crime Show forest plot | 1 | 235 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.79 [0.62, 1.01] |
| 2 Re‐incarceration to jail at 12 months Show forest plot | 1 | 235 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.06 [0.91, 1.24] |
| 3 ASI drug use at 12 months Show forest plot | 1 | 235 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [‐0.03, 0.03] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Abstinence from drug use at 6 months Show forest plot | 1 | 31 | Risk Ratio (M‐H, Random, 95% CI) | 2.89 [0.73, 11.43] |
| 2 ASI drug score at 6 months Show forest plot | 1 | 31 | Mean Difference (IV, Random, 95% CI) | ‐0.04 [‐0.37, 0.29] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Marjuana frequency at 3 months Show forest plot | 1 | 40 | Mean Difference (IV, Random, 95% CI) | ‐1.05 [‐2.39, 0.29] |
| 2 Arrest frequency 3 years post release Show forest plot | 1 | 31 | Mean Difference (IV, Random, 95% CI) | ‐0.66 [‐1.31, ‐0.01] |
| 3 Time to first arrest or offence 36 months post Show forest plot | 1 | 31 | Mean Difference (IV, Fixed, 95% CI) | 0.87 [‐0.12, 1.86] |
| 4 Positive drug screen at 12 months Show forest plot | 1 | 84 | Mean Difference (IV, Fixed, 95% CI) | ‐0.70 [‐3.50, 2.10] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Arrest by police Show forest plot | 1 | 156 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.97 [0.70, 1.36] |
| 2 DUDIT scores Show forest plot | 1 | 156 | Mean Difference (IV, Random, 95% CI) | ‐0.22 [‐2.51, 2.07] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Arrests between 6 and 24 months Show forest plot | 1 | 112 | Mean Difference (IV, Fixed, 95% CI) | ‐0.24 [‐0.76, 0.28] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Substance abuse relapse at 3 months Show forest plot | 1 | 38 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.67 [0.30, 1.50] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Number of new offences committed at 12 months Show forest plot | 1 | 29 | Risk Ratio (M‐H, Random, 95% CI) | 0.64 [0.07, 6.01] |
