January 2024Original Research: The Association Between Prior Mental Health Service Utilization and Risk

The Association Between Prior Mental Health Service Utilization and Risk of Recidivism among Incarcerated Ontario ResidentsL’association entre l’utilisation antérieure des services de santé mentale et le risque de récidivisme chez les résidents de l’Ontario incarcérésMichael Lebenbaum, MSc, PhD^1,2

, Fiona Kouyoumdjian, MD, MPH, PhD^1,3

, Anjie Huang, MSc¹

, and Paul Kurdyak, MD, PhD^1,2,4,5,6

img1

The Canadian Journal of Psychiatry / La Revue Canadienne de Psychiatrie
2024, Vol. 69(1) 21-32
© The Author(s) 2022
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/07067437221140385
TheCJP.ca | LaRCP.caimg2

Abstract
Background: There is mixed evidence on the link between mental health and addiction (MHA) history and recidivism. Few studies have examined post-release MHA care. Our objective was to examine the association between prior (pre-incarceration) MHA service use and post-release recidivism and service use.Methods: We conducted a population-based cohort study linking individuals held in provincial correctional institutions in 2010 to health administrative databases. Prior MHA service use was assigned hierarchically in order of hospitalization, emergency department visit and outpatient visit. We followed up individuals post-release for up to 5 years for the first occurrence of recidivism and MHA hospitalization, emergency department visit and outpatient visit. We use Cox-proportional hazards models to examine the association between prior MHA service use and each outcome adjusting for prior correctional involvement and demographic characteristics.Results: Among a sample consisting of 45,890 individuals, we found that prior MHA service use was moderately associated with recidivism (hazard ratio (HR): 1.20–1.50, all P < 0.001), with secondary analyses finding larger associations for addiction service use (HR range: 1.34–1.54, all P < 0.001) than for mental health service use (HR range: 1.09–1.18, all P < 0.001). We found high levels of post-release MHA hospitalization and low levels of outpatient MHA care relative to need even among individuals with prior MHA hospitalization.Discussion: Despite a high risk of recidivism and acute MHA utilization post-release, we found low access to MHA outpatient care, highlighting the necessity for greater efforts to facilitate access to care and care integration for individuals with mental health needs in correctional facilities.AbrégéContexte: Les données probantes sont mixtes en ce qui concerne le lien entre les antécédents et le récidivisme en santé mentale et dépendances (SMD). Peu d’études ont examiné les soins de SMD après la libération. Notre objectif était d’examiner l’association entre l’utilisation des services de SMD antérieure (pré-incarcération) et le récidivisme et l’utilisation des services après la libération.Méthodes: Nous avons mené une étude de cohorte dans la population liant les personnes détenues dans des institutions correctionnelles provinciales en 2010 à des bases de données administratives de la santé. L’utilisation antérieure des services de SMD était attribuée hiérarchiquement par ordre d’hospitalisation, de visite au service d’urgence et de visite ambulatoire. Nous avons suivi des personnes jusqu’à 5 ans après leur libération pour la première occurrence de récidivisme et d’hospitalisation de SMD, de visite au service d’urgence et de visite ambulatoire. Nous utilisons les modèles à risques proportionnels de Cox pour examiner l’association entre l’utilisation antérieure des services de SMD et chaque résultat s’ajustant à une implication correctionnelle antérieure et aux caractéristiques démographiques.Résultats: Dans un échantillon consistant en 4 589 personnes nous avons constaté que l’utilisation antérieure des services de SMD était modérément associée au récidivisme (Rapport de Risque (RR): 1,20–1,50, tout P < 0,001), et des analyses secondaires ont révélé des associations plus importantes d’utilisation des services de dépendance (gamme du RR: 1,34–1,54, tout P < 0,001) que pour l’utilisation des services de santé mentale (gamme du RR: 1,09–1,18, tout P < 0,001). Nous avons constaté des niveaux élevés d’hospitalisation en SMD après la libération et de faibles niveaux de soins de SMD ambulatoires relativement au besoin même chez les personnes ayant une hospitalisation antérieure en SMD.Discussion: Malgré un risque élevé de récidivisme et une utilisation aiguë des services de SMD après la libération, nous avons observé un faible accès aux soins ambulatoires de SMD, ce qui souligne la nécessité d’initiatives plus importantes pour faciliter l’accès aux soins et leur intégration pour les personnes ayant des besoins de santé mentale dans les établissements correctionnels.Keywords
recidivism, mental health and addictions, post-release care, population-based studyIntroductionThere were approximately 11 million individuals incarcerated across the world in 2018¹ and these individuals experience a much greater burden of disease² and higher rates of mortality³ compared to the general population. Mental health conditions are substantially over-represented, with a prior study from Ontario demonstrating an 11-fold greater rate of schizophrenia, 9-fold greater rate of mood disorders, and 5-fold greater rate of substance-related disorders,² with findings similar in other jurisdictions.^4–6Recidivism (i.e., re-arrest, re-conviction, or re-imprisonment) is common post-incarceration with re-conviction within 2 years globally ranging from ∼20% to 60% (∼35% in Canada) and is often used to judge the effectiveness of the criminal justice system.⁷ A greater risk of recidivism among individuals with mental health and addiction (MHA) conditions may contribute to their over-representation in prisons.⁴ Reviews have found small or null associations for non-substance disorder-related mental illnesses or clinical psychiatric factors on risk of recidivism; however, the included studies were largely conducted in small samples^8,9 in the USA and UK^8,9 and often focused on specific psychiatric diagnoses.^8,10–12 More recent research using large population-based samples has found that mental illnesses significantly predict recidivism.^10,11,13 Therefore, further research using large representative populations is needed to investigate the association between MHA history and recidivism.Given worsening mental health post-release is associated with greater risks of recidivism¹⁴ and some types of postrelease substance disorder treatment have been shown to reduce risk of recividism,¹⁵ a lack of post-release MHA care may be a contributing factor to the increased risks of recidivism. The post-release period is also a high-risk period for several MHA-related outcomes including over-doses,¹⁶ psychiatric hospitalizations,² and suicides.¹⁷ Despite these findings pointing to a need for greater access to MHA services post-release, little research has examined post-release MHA care beyond psychiatric hospitalizations.¹⁸ Studies to date suggest that a substantial proportion of individuals with mental illnesses or substance use disorders do not receive MHA care post-release.^19–24 To our awareness, all of the studies have been conducted in the US^19–22,25 or Western Australia,^23,24 which limits the generalizability of the results to other countries with different health care and criminal justice systems.Therefore, our primary objective was to investigate the association between severity of MHA history as measured by pre-incarceration MHA service use and risk of recidivism using population-based linked administrative health and justice system data for all residents of Ontario (population 14.6 million). Our secondary objectives were to examine psychiatric history in relation to post-release MHA care including outpatient visits, emergency department visits and hospitalizations.MethodsDataThe Ministry of the Solicitor General (MSG) provided data on individuals incarcerated in Ontario provincial prisons, which included the age, sex, and dates of admission and release from provincial custody. The MSG data were used to identify the correctional population, correctional history and recidivism. Individual records were linked using unique encoded identifiers across Ontario correctional, demographic, and healthcare databases and analyzed at ICES. This included the Registered Persons Database to confirm demographic data and obtain death dates. We also used health systems data including the National Ambulatory Care Reporting System (NACRS) to study emergency department (ED) visits, Ontario Mental Health Reporting System (OMHRS) to determine admissions to designated mental health beds, Discharge Abstract Database (DAD) to obtain admissions to non-mental health beds, and Ontario Health Insurance Plan (OHIP) Claims Database to identify visits to physicians. The linkage rate for MSG to other ICES datasets was 97.4%.²Study PopulationOur study population consisted of all individuals ages 18 years and older who were released from a provincial correctional facility from 1 January 2010 to 31 December 2010. We selected the incarceration leading to the first release in 2010 as the index incarceration. Provincial correctional facilities hold people awaiting trial or sentencing and people who are sentenced to less than 2 years and consist of jails, detention centres, and correctional centres. Exclusion criteria included transfer from a provincial correctional facility to federal prison, release due to deportation, or death, as these people would not be at risk of our post-release outcomes of interest. We further excluded individuals with data errors including inconsistencies across databases in the year of birth or sex and probable linkage errors such as release or utilization after death. In addition, we excluded non-Ontario residents given they may not have access to provincially funded care in the post-release period.Study OutcomesOur primary outcome was the occurrence of the first readmission to a provincial correctional facility (recidivism) within 30 days, 365 days, and 5 years of the date of release (index date) from the index incarceration.Secondary outcomes were the first occurrence within 30 days, 365 days, and 5 years of the index date of a MHA outpatient visit, MHA ED visit, and MHA hospitalization. MHA outpatient visits were defined as any visit in the OHIP database to a psychiatrist or any visit to a family physician accompanied by a MHA diagnostic code defined by a validated algorithm.²⁶ A MHA ED visit was defined as any NACRS record with a non-dementia MHA diagnosis (ICD-10-CA: F06-F99) in a primary diagnostic position (i.e., the most responsible diagnosis) or intentional self-injury (ICD-10-CA: X60-X84, Y10-Y19, Y28) diagnosis in any of the other 9 secondary diagnostic codes affiliated with the visit. Mental health and addiction hospitalizations included any admission in the OMHRS database excluding discharges with a primary diagnosis of dementia (DSM-IV: 290, 294) or any admission recorded in the DAD with a non-dementia MHA diagnosis (ICD-10-CA: F06-F99) in a primary position or self-harm (ICD-10-CA: X60-X84, Y10-Y19, Y28) diagnosis in a secondary position.Primary ExposureWe ascertained whether individuals had an MHA outpatient visit, MHA ED visit, or MHA hospitalization within the 5 years prior to the admission date of the index incarceration using the same definitions as for the outcomes. Individuals were assigned hierarchically to the greatest level of severity of care they received pre-incarceration in order of MHA hospitalization, MHA ED visit, MHA outpatient visit, and no prior MHA care.As a secondary analysis, we examined outpatient visits, ED visits and hospitalizations for mental health diagnoses separately from addiction diagnoses. Outpatient addiction visits were defined as any visit in the OHIP database to a psychiatrist or family physician accompanied by a diagnosis of alcoholism (ICD-8-CM: 303) or drug dependence (ICD-8-CM: 304). An addiction ED visit was defined as any NACRS record with a primary diagnosis of substance use disorders (ICD-10-CA: F10-19, F55, F63.0). Addiction hospitalizations were an OMHRS record (DSM-IV: 291–292, 303–305, 312.31) or DAD record (ICD-10-CA: F10–19, F55, F63.0) with a primary diagnosis of substance use disorder. For each type of care, all other MHA diagnosis codes were considered mental health.CovariatesWe characterized each individual’s age, sex, income quintile at the census dissemination area level (i.e., 400–700 people), rurality (i.e., community size of less than 10,000 residents), medical comorbidity, length of the index incarceration, and cumulative length of time spent incarcerated in the 5 years prior to the index incarceration. Medical comorbidity was characterized using the Johns Hopkins collapsed aggregated diagnostic groups (CADGs),²⁷ which aggregated the original 33 ADGs into 12 diagnostic groups. We did not use the psychosocial ADG given high overlap with the primary exposure and the pregnancy CADG since it is collinear with sex. Cumulative time spent incarcerated consisted of the number of days in custody in the 5 years prior to the index event date, which included days for the index incarceration.Statistical AnalysisBy MHA service use levels, we examined the distribution of covariates, the cumulative incidence of each outcome within 30 days, 365 days, and 5 years after the index date, and the occurrence of events over continuous time. By all study variables, we examined the cumulative incidence of each outcome.Our primary analysis examined the association between prior MHA service use and time to recidivism within 5 years using Cox-proportional hazards models. Secondary analyses used the same model to examine post-release MHA service outcomes. Models controlled for age, sex, neighbourhood income, rural residence, total number of days in custody, and CADGs. There were missing data only for neighbourhood income (1.7%) and rural residence (0.1%). We included a missing indicator variable to the model for neighbourhood income and rurality. The observation period for recidivism was censored at the end of follow-up or date of death, whichever was earlier. We censored health service outcomes at the end of follow-up, date of death, loss of health care coverage with a 30-day grace period, or if a more serious event occurred. That is, we further censored MHA hospitalizations if recidivism occurred, MHA ED visits if recidivism or MHA hospitalization occurred, and MHA outpatient visits if any of the other types of events occurred.Given some prior studies have found addictions to be primarily responsible for the association between MHA conditions and recidivism,^10,28,29 we conducted secondary analyses where we examined outpatient visits, ED visits and hospitalizations for mental health diagnoses separately from addiction diagnoses.Lastly, since the MHA severity variable may reflect access especially for outpatient care, we conducted a sensitivity analysis combining the MHA outpatient group with the reference group.All analyses were performed in SAS version 9.4.ResultsWe identified 71,295 incarceration episodes with a release date in 2010. After exclusion criteria were applied (Supplemental Figure 1), a total of 45,890 individuals were included in the analysis.The characteristics of the sample overall and across levels of primary exposure are presented in Table 1. A total of 66.4% of the sample had past 5 years (pre-incarceration) MHA service use with the highest level of severity being an outpatient visit for 33.9% of the cohort, followed by an ED visit for 19.1% of individuals, and a hospitalization for 13.3% of individuals.RecidivismOver the 5-year follow-up period, recidivism increased with greater severity of prior MHA service use with similar rates among individuals with a prior MHA ED visit (34.0 per 100 person years (PYs)) or hospitalization (33.0 per 100 PY) (Table 2). There were similar findings for the incidence within 30 days, 365 days and 5-years post-release (Table 3), and over continuous time as presented in the Kaplan–Meier curve (Figure 1). A majority of individuals with a prior MHA ED visit (69.8%) or hospitalization (67.8%) returned to prison within 5 years.Differences in recidivism across prior levels of MHA service use were statistically significant in Cox-proportional hazards models. There was a significantly greater rate of recidivism for individuals with a prior MHA outpatient visit (hazards ratio (HR): 1.20, 95% Confidence Interval (CI): 1.17, 1.24), ED visit (HR: 1.47, 95% CI: 1.42, 1.52), or hospitalization (HR: 1.50, CI: 1.44, 1.56) (Table 4, full model in Supplemental Table 1), compared with those with no prior MHA health care use. Secondary analyses demonstrated that associations with recidivism for all levels of addiction care (hospital/ED/outpatient admission/visit HR range: 1.34–1.54, CI range: 1.30, 1.59, reference group: individuals without prior addiction service use) were greater than with prior mental health service use (hospital/ED/outpatient admission/visit HR range 1.09–1.18, CI range: 1.06–1.24, reference group: individuals without prior mental health service) (Supplemental Table 2).Secondary Outcomes: MHA Service Use Post-releaseThe post-release period was a high-risk period for MHA ED and hospitalizations, especially in the first 30 days and 365 days and for individuals with a prior MHA ED visits or hospitalization (Table 3, Supplementary Figure 2(a) to (c)). However, a large proportion of individuals with prior MHA service did not receive a post-release MHA outpatient visit within 365 days (26.3–49.8%) (Table 3).In results from Cox-proportional hazards models, all levels of prior MHA service use had large associations with time to post-release MHA outpatient visit (HR range: 2.83–4.48, P < 0.001) and there were sizeable associations for the occurrence of a post-release MHA ED visit with a prior MHA ED visit (HR: 4.54, P < 0.001) or hospitalization (HR: 6.86, P < 0.001) (Table 4). While having a prior MHA ED visit had a large association with the incidence of a post-release hospitalization (HR = 4.71, P < 0.001), the association with a prior hospitalization was substantially higher (HR = 17.84, P < 0.001). Conclusions from secondary analyses that divided prior MHA service use into mental health service use and addiction service use, were largely the same, with the exception that a prior mental health hospitalization (HR: 12.42, P < 0.001) rather than a prior addiction hospitalization (HR: 2.96, P < 0.001) was largely responsible for the association with post-release hospitalization (Supplemental Table 2).Table 1. Number and Proportion of Individuals Released from Ontario Provincial Correctional Facilities in 2010 by Pre-incarceration Mental Health Service use.
Tab1Table 2. Demographic, Clinical and Correctional Factors and Outcomes During the 5-year Post-release Period, among Individuals Released from Ontario Provincial Correctional Facilities in 2010.
Tab2Tab2aIn an exploratory analysis, combining the MHA outpatient group with the reference group, the results were very similar to those for the primary analysis and conclusions unchanged (Supplemental Table 3).DiscussionTwo out of 3 individuals (66.4%) had evidence of MHA health care utilization in the 5 years prior to correctional involvement, with 13.3% and 19.1% having had at least one prior MHA hospitalization or prior MHA ED visit, respectively. Recidivism was high and most likely to occur amongst those with a prior MHA ED visit or hospitalization, followed by a prior outpatient visit; these associations remained statistically significant even after controlling for duration of prior incarceration and sociodemographic characteristics. However, secondary analyses suggest that prior addiction service use rather than prior mental health service use is the greater driver of association between prior MHA service use and recidivism. To our knowledge, this is one of the few studies outside of the US to examine the utilization of MHA care post-release. We found access to outpatient care was low for individuals with evidence of high-severity MHA use, suggesting a lack of care continuity between correctional institutions and the community.Implications and Relation to Previous WorkWe found elevated risks of recidivism among individuals with prior MHA service use and this risk remains elevated for several years. Although prior MHA service use appears to be an important risk factor for recidivism, even after accounting for prior incarceration history and sociodemographic characteristics, secondary analyses found this risk is largely driven by prior addiction service use (HR range: 1.34–1.53) rather than by mental health service use (HR range: 1.10–1.18). It is unclear from this study whether these associations are causal; the association may reflect multiple factors such as mental health symptoms, access to care and impacts of mental health treatment. These findings are in contrast to 3 large administrative-based studies, which found an important role for non-substance mental illnesses.^10,11,13 However, one of these studies did not examine substance use disorders¹³ and a factor complicating the results of these studies is that personality disorders are highly underestimated using administrative health care databases,^4,10,11,30 have high comorbidity with other mental illnesses, and can confound the association between nonpersonality or substance use disorder mental illnesses and risk of recidivism.³¹ Consistent with one study that examined violent recidivism,¹⁰ we found an association between prior mental health service use and recidivism while controlling for prior addiction service use in the secondary analysis that split up mental health from addiction care (Supplemental Table 2), but this association was smaller in our study and we did not distinguish between service use for personality disorders and other mental illnesses. Our findings are more in line with a systematic review⁹ and with other studies,^{12,28,29,31,32} which found a small or absent role for mental illnesses other than substance use disorders or personality disorders in predicting recidivism. One study found no association with mental health history in the absence of substance use disorder²⁸ and another found greater recidivism only among individuals with dual diagnoses or substance abuse alone relative to individuals with mental illness alone.²⁹ Therefore, our study findings and those of other studies also support the validity of substance use specifically (rather than mental health overall) as a key factor predicting recidivism in conceptual models and prediction tools, such as the Risk Needs Responsivity model, which also includes criminal history, antisocial personality pattern, procriminal attitudes, social supports for crime, family/marital relationships, school/work, and prosocial recreational activities.³³ In addition, research to date suggests that although not highly predictive or typically a criminogenic factor, mental health issues may be a separate goal for treatment providers.³³Table 3. Number of and Cumulative Incidence of Individuals Receiving Each Type of Post-release Care, by Time Post-release and Prior MHA Service Use (N, %) for Individuals Released from Ontario Provincial Correctional Facilities in 2010.
Tab3Our findings combined with research demonstrating that substance use disorders are greatly elevated among prison populations⁶ and that some types of substance use disorder treatment may reduce recidivism,¹⁵ suggest that increasing access in custody and post-release to evidence-based treatment for substance use disorders is essential to improving both health and public safety. This includes therapeutic communities (i.e., drug-free residential settings focused on mutual aid), naltrexone¹⁵ and possibly other less studied interventions such as diversion programs, which direct individuals with substance disorders who commit low-level offences towards community-based substance and social services rather than the criminal justice system, reducing recidivism and costs to the criminal justice system.³⁴We found greater post-release MHA outpatient care among individuals with prior MHA service use with the highest levels among those with prior acute service utilization. However, these levels of care were likely low relative to need, especially for individuals with a prior hospitalization. In Ontario, 60% of individuals with a psychiatric hospitalization receive an MHA outpatient visit within 30-days of discharge.³⁵ By comparison, in our sample, only 33.9% of individuals with a prior hospitalization received outpatient MHA care within 30 days of release and approximately one quarter did not receive outpatient care within 1 year. This much lower level of utilization occurred despite the incarcerated population with a prior psychiatric hospitalization having nearly the same elevated risk of MHA hospitalization within 30 days and a 25% greater risk within 1 year as the general population discharged from a psychiatric hospital.³⁵ These results highlight that access to care may be low, even in a system of universal access to physician and hospital care, with limited integration between the corrections and health care system. Prior studies examining mental health and addiction services were conducted in the US^20–22 and Western Australia^23,24 and have found similar or worse access to care. Whether prompt post-release care has the potential to reduce poor health outcomes such as MHA hospitalizations and overdoses deserves attention in future research studies.Fig1Table 4. Cox-proportional Hazard Model Examining the Association Between Prior MHA Service Use, Recidivism, and Mental Health Service use Outcomes, for Individuals Released from Ontario Provincial Correctional Facilities in 2010.
Tab4Strengths and LimitationsThe strengths of this study included the use of whole population administrative data that enabled the examination of a large sample of all individuals held in provincial correctional custody. Both prior MHA service use and post-release utilization were assessed using non-correctional administrative data; therefore, assessment of the primary exposure and outcomes were not biased by correctional involvement pre- or post-release¹³ or recall bias. Lastly, residents had access to a publicly funded health care system that covers the costs for all types of care examined. Therefore, access to insurance or affordability is a smaller issue than in prior studies on postrelease care conducted in the US³⁶ and the results are likely generalizable to other universal health care systems.This study is not without limitations. First, since only provincial corrections data were available, the sample only includes individuals awaiting trial or sentencing and those with sentences of less than 2 years (∼64% of adults in prison nationally).³⁷ Therefore, our study would exclude many violent offenders (e.gs., sexual assault, aggravated assault, homicide), which requires further study given mental health and addiction may be an important factor predicting recidivism and post-release care in these populations. Second, we lacked data on the use of non-physician or hospital-based MHA services including visits to psychologists or the use of community-based addiction care. We do not expect this to be a significant issue for psychologists, whose services are not covered by the public health care system in Ontario, which results in substantial barriers to access for this population. However, this was likely a significant issue for addiction services, much of which is provided in the community by non-physician providers. Therefore, the association between outpatient addiction services in secondary analyses should be interpreted with caution since it relied exclusively on physician-based care. Third, there may be heterogeneous associations across specific mental health and addiction conditions, which were not examined. Fourth, although the use of administrative databases rather than structured clinical assessments eliminated biases due to patient recall, given the diagnoses recorded are not validated and multiple diagnoses were not recorded, administrative databases may introduce measurement error in the assessment of psychiatric diagnoses, likely underestimating the prevalence of some psychiatric conditions.¹⁰ This underestimation is likely most substantial for antisocial personality disorder, which structured assessments demonstrate to be highly prevalent in this population⁴ and much greater than estimated using administrative data from Canada³⁰ and Sweden.¹⁰ Lastly, as a result of relying on mental health service utilization to proxy mental illness, any association between the underlying mental illness and recidivism may have been attenuated by receipt of mental health services.ConclusionsThere is a substantial gradient with increasing recidivism with greater severity of pre-incarceration MHA service use. However, these risks primarily occur among individuals with greater prior addiction service use. Despite a high risk of poor outcomes including recidivism and acute MHA utilization post-release, we found low access to outpatient care in prison and post-release. Therefore, access to care appears to be poorly aligned with need. Whether poor access to care contributes to the high rates of recidivism or hospitalizations post-release is unknown, but given the complex needs of these individuals, more efforts are needed to facilitate access to care.AcknowledgementsNoneAuthor’s ContributionsPK conceptualized and designed the study, interpreted the results, drafted the initial manuscript, revised the manuscript, and approved the final manuscript as submitted. ML designed the study, interpreted the results, drafted the initial manuscript, revised the manuscript, and approved the final manuscript as submitted. FK conceptualized and designed the study, interpreted the results, revised the manuscript, and approved the final manuscript as submitted. AH designed the study, analyzed the data, interpreted the results, revised the manuscript, and approved the final manuscript as submitted. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.FundingThe author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Canadian Institute for Health Research, Ontario Ministry of Health and Long-Term CareEthics Approval and Consent to ParticipateICES is an independent, non-profit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze health care and demographic data, without consent, for health system evaluation and improvement. The use of ICES data in this project was authorized under section 45 of Ontario’s Personal Health Information Protection Act. Research Ethics Board approval for the use of corrections data was obtained from the Hamilton Integrated Research Ethics Board (REB #4575).Declaration of Conflicting InterestsThe author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.FundingThis study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health (MOH). The opinions, results and conclusions reported in this paper are those of the authors and are independent from the funding sources. Dr Kurdyak received funding from the Canadian Institutes of Health Research. Michael Lebenbaum was supported by a Vanier Canada Graduate Scholarship from the Social Sciences and Humanities Research CouncilDisclaimersParts of this material are based on data and information compiled and provided by MOH, Canadian Institute for Health Information. The analyses, conclusions, opinions and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred.Availability of DataWe are not able to share data because of restrictions specified in our Research Agreement with the Ministry of the Solicitor General and in the data-sharing agreements of ICES. People who would like access to data from the Ministry should direct requests to SolGenResearch@ontario.ca. Access to data at ICES can be granted to those who meet pre-specified criteria for confidential access, available at www.ices.on.ca/. Requests to access ICES data for research purposes may be submitted to ICES’ Data and Analytic Services, with information available at http://www.ices.on.ca/DAS and contact: das@ices.on.ca.ORCID iDsMichael Lebenbaum

https://orcid.org/0000-0002-6319-3466Fiona Kouyoumdjian

https://orcid.org/0000-0002-6869-7516Anjie Huang

https://orcid.org/0000-0001-9148-1038Paul Kurdyak

https://orcid.org/0000-0001-8115-7437Supplemental MaterialSupplemental material for this article is available online.References

Walmsley R. World Prison Population List. Twelfth Edition. 2018. Available at: www.prisonstudies.org/sites/default/files/resources/downloads/wppl12.pdf

Kouyoumdjian FG, Cheng SY, Fung K, et al. The health care utilization of people in prison and after prison release: a population-based cohort study in Ontario, Canada. PLoS One. 2018;13(8):e0201592.

Kouyoumdjian FG, Andreev EM, Borschmann R, et al. Do people who experience incarceration age more quickly? Exploratory analyses using retrospective cohort data on mortality from Ontario, Canada. PLoS One. 2017;12(4):e0175837.

Fazel S, Danesh J. Serious mental disorder in 23 000 prisoners: a systematic review of 62 surveys. Lancet. 2002;359(9306):545-550.

Fazel S, Seewald K. Severe mental illness in 33 588 prisoners worldwide: systematic review and meta-regression analysis. Br J Psychiatry. 2012;200(5):364-373.

Fazel S, Bains P, Doll H. Substance abuse and dependence in prisoners: a systematic review. Addiction. 2006;101(2):181-191.

Yukhnenko D, Sridhar S, Fazel S. A systematic review of criminal recidivism rates worldwide: 3-year update [version 2; peer review: 2 approved.

Fazel S, Yu R. Psychotic disorders and repeat offending: systematic review and meta-analysis. Schizophr Bull. 2011;37(4):800-810.

Bonta J, Blais J, Wilson HA. A theoretically informed meta-analysis of the risk for general and violent recidivism for mentally disordered offenders. Aggress Violent Behav. 2014;19(3):278-287.

Chang Z, Larsson H, Lichtenstein P, et al. Psychiatric disorders and violent reoffending: a national cohort study of convicted prisoners in Sweden. Lancet Psychiatry. 2015;2(10):891-900.

Jones, RM, Manetsch, M, & Gerritsen, C, et al. Patterns and predictors of reincarceration among prisoners with serious mental illness: a cohort study: modèles et Prédicteurs de Réincarcération Chez les Prisonniers Souffrant de Maladie Mentale grave: une Étude de Cohorte. Can J Psychiatry. 2020;66(6):0706743720970829.

Mahentharan, M, Kouyoumdjian, FG, & de Oliveira, C, et al. Prevalence and predictors of reincarceration after correctional center release: a population-based comparison of individuals with and without schizophrenia in Ontario, Canada: Prévalence et prédicteurs de la réincarcération après la libération d’un centre. Can J Psychiatry. 2020;66(4):0706743720953018.

Baillargeon J, Binswanger IA, Penn JV, et al. Psychiatric disorders and repeat incarcerations: the revolving prison door. Am J Psychiatry. 2009;166(1):103-109.

Wallace, D, & Wang, X. Does in-prison physical and mental health impact recidivism? SSM Popul Health. 2020;11:11100569.

Holloway, KR, & Bennett, TH. Drug interventions. In: What works in crime prevention and rehabilitation. Springer New York, NY: Springer; 2016:219-236.

Binswanger IA, Blatchford PJ, Mueller SR, et al. Mortality after prison release: opioid overdose and other causes of death, risk factors, and time trends from 1999 to 2009. Ann Intern Med. 2013;159(9):592-600.

Kariminia A, Law MG, Butler TG, et al. Suicide risk among recently released prisoners in New South Wales, Australia. Med J Aust. 2007;187(7):387-390.

Baillargeon J, Hoge SK, Penn JV. Addressing the challenge of community reentry among released inmates with serious mental illness. Am J Community Psychol. 2010;46(3–4):361-375.

Mallik-Kane K, Visher CA. Health and prisoner reentry: How physical, mental, and substance abuse conditions shape the process of reintegration. Washington, DC: Urban Institute Justice Policy Center; 2008.

Lovell D, Gagliardi GJ, Peterson PD. Recidivism and use of services among persons with mental illness after release from prison. Psychiatr Serv. 2002;53(10):1290-1296.

Begun AL, Early TJ, Hodge A. Mental health and substance abuse service engagement by men and women during community reentry following incarceration. Adm Policy Ment Health. 2016;43(2):207-218.

Morrissey JP, Steadman HJ, Dalton KM, et al. Medicaid enrollment and mental health service use following release of jail detainees with severe mental illness. Psychiatr Serv. 2006;57(6):809-815.

Sodhi-Berry N, Knuiman M, Alan J, et al. Pre-and postsentence mental health service use by a population cohort of older offenders (≥ 45 years) in Western Australia. Soc Psychiatry Psychiatr Epidemiol. 2015;50(7):1097-1110.

Sodhi–Berry N, Knuiman M, Preen DB, et al. Predictors of post–sentence mental health service use in a population cohort of first–time adult offenders in Western Australia. Crim Behav Ment Health. 2015;25(5):355-374.

Hartwell S, Deng X, Fisher W, et al. Predictors of accessing substance abuse services among individuals with mental disorders released from correctional custody. J Dual Diagn. 2013;9(1):11-22.

Steele LSLS, Glazier RHRH, Lin EE, et al. Using administrative data to measure ambulatory mental health service provision in primary care. Med Care. 2004;42(10):960-965.

John Hopkins Bloomberg School of Public Health. The Johns Hopkins ACG® System Excerpt from Version 11.0 Technical Reference Guide. 2014.

Rezansoff SN, Moniruzzaman A, Gress C, et al. Psychiatric diagnoses and multiyear criminal recidivism in a Canadian provincial offender population. Psychol Public Policy Law. 2013;19(4):443.

Wilson AB, Draine J, Hadley T, et al. Examining the impact of mental illness and substance use on recidivism in a county jail. Int J Law Psychiatry. 2011;34(4):264-268.

Hensel, JM, Casiano, H, & Chartier, MJ, et al. Prevalence of mental disorders among all justice-involved: a population-level study in Canada. Int J Law Psychiatry. 2020;68:68101523.

O’Driscoll C, Larney S, Indig D, et al. The impact of personality disorders, substance use and other mental illness on re-offending. J Forensic Psychiatry Psychol. 2012;23(3): 382-391.

Grann M, Danesh J, Fazel S. The association between psychiatric diagnosis and violent re-offending in adult offenders in the community. BMC Psychiatry. 2008;8(1):1-7.

Bonta J, Andrews DA. Risk-need-responsivity model for offender assessment and rehabilitation. Rehabilitation (Bonn). 2007;6(1):1-22.

Bernard CL, Rao IJ, Robison KK, et al. Health outcomes and cost-effectiveness of diversion programs for low-level drug offenders: a model-based analysis. PLoS Med. 2020;17(10): e1003239.

Gatov, E, Koziel, N, & Kurdyak, P, et al. Discharge and postdischarge outcomes of psychiatric inpatients with a lifetime history of exposure to interpersonal trauma: a population-based study. Gen Hosp Psychiatry. 2020;65:6582-6590.

Osborn R, Squires D, Doty MM, et al. In new survey of eleven countries, US adults still struggle with access to and affordability of health care. Health Aff. 2016;35(12):2327-2336.

Malakieh A. Adult and youth correctional statistics in Canada, 2017/2018. 2019 [accessed 2021 November 6] https://www150.statcan.gc.ca/n1/pub/85-002-x/2019001/article/00010-eng.htm

¹ ICES, Toronto, ON, Canada² Institute of Health Policy, Management and Evaluation (IHPME), University of Toronto, Toronto, Ontario, Canada³ Department of Family Medicine, McMaster University, Hamilton, ON, Canada⁴ Institute for Mental Health Policy Research, Centre for Addiction and Mental Health, Toronto, ON, Canada⁵ Department of Psychiatry, Faculty of Medicine, University of Toronto, Toronto, ON, Canada⁶ Dalla Lana School of Public Health, University of Toronto, Toronto, ON, CanadaCorresponding author:
Paul Kurdyak, ICES, Toronto, ON, Canada; Institute of Health Policy, Management and Evaluation (IHPME), University of Toronto, Toronto, Ontario, Canada; Institute for Mental Health Policy Research, Centre for Addiction and Mental Health, Toronto, ON, Canada.
Email: Paul.Kurdyak@camh.caJournal CoverOtsuka LundbeckEditorial BoardJournal InfoPerspective: A Call for an Evidence-Based Strategy Against the Overdose CrisisCommentary: The Burden of Psychosis in Black Communities in Canada: More than a Feeling,Original Research: Ten-Year Trends in Lithium Prescribing in Alberta, CanadaOriginal Research: The Association Between Prior Mental Health Service Utilization and RiskOriginal Research: Experiences of Sexual and Reproductive Health Care Access for WomenOriginal Research: Cost-Effectiveness of Group Transdiagnostic Cognitive BehaviouralOriginal Research: Culturally Adapted Cognitive Behaviour Therapy (CaCBT)Sage Peer ReviewersSage Author Services