Background

People who are incarcerated have a significantly higher health burden compared to the general population, including infectious diseases, mental illness, and substance use disorders. The period of transition to community is a high-risk period for negative outcomes, with studies revealing mortality rates nearly 13 times higher in the weeks following release, as compared to the general population. While a number of factors contribute to this high mortality in the transition from prison to community, studies have identified that opioids contribute to nearly 1 in 8 post release fatalities. People with histories of incarceration are known to face an elevated risk of overdose, particularly in the days and weeks following release. In British Columbia (BC) (Canada’s third most populous province) approximately 70% of people who are incarcerated have been identified as having either a mental health or substance use disorder, and the most commonly reported substances used at admission are opioids and stimulants. Opioid agonist treatment (OAT) is available as the first line treatment for opioid use disorder (OUD) in BC, the most commonly available forms of which include methadone (a full opioid agonist) and buprenorphine (a partial opioid agonist), both of which are effective at reducing opioid craving and withdrawal, reducing illicit opioid use and protecting against overdose mortality.

OAT dispensation has increased in BC’s provincial prisons in recent years, with the proportion of people with OUD who had received OAT more than doubling from approximately 30% in 2015 to 65% in 2017. OAT has also been expanded in community settings in BC, with an increasing range of available medications, and new prescribing authorities assigned to nurse practitioners. For people released from prison, the timeliness of OAT dispensation is critical. Research has found that reduced tolerance following release can increase risk of overdose on the day of and in the 1–2 weeks immediately following release. Furthermore, withdrawal symptoms set in within 24–48 h of last OAT dose; therefore, expedient continuity of medication dispensation in the community is critical to reducing overdose risk. In the absence of access to OAT, people with OUD are more likely to engage with the illicit drug supply, which is increasingly contaminated by potent illicit opioids like fentanyl which have been attributed to the ongoing rise in overdose deaths in the province.

Alongside fentanyl, methamphetamine has also increasingly been detected among people who have died of illicit drug toxicity (overdose) in BC, and stimulant use is on the rise in North America among people who use opioids. This trend has also been observed among people admitted to BC’s prisons, where the prevalence of reported methamphetamine use at intake increased nearly fivefold between 2009–2017. This is particularly concerning given polysubstance use with opioids has been associated with compounded risk of adverse health effects, such as overdose. Furthermore, people with concurrent OUD and stimulant use disorder (StUD) often have greater health service needs compared to people with OUD alone, yet are less likely to be engaged and retained in care. Given OAT is a first line treatment for OUD, it remains a critical intervention that should be offered in a timely manner during the period of transition from prison to community for all people with OUD who want access to it. Due to the limited availability of pharmacological and psychosocial treatments for StUD, people who use stimulants face significant gaps in their substance use service needs, making people with concurrent OUD and StUD a priority for intervention. OAT can serve as an opportunity for engagement in care for this population, who must not be further excluded from access to evidence-based health and substance use services.

Prior studies have found that stimulant use can interfere with OAT outcomes, such as reductions in illicit opioid use and long-term retention. While studies have investigated the impact of stimulant use on OAT retention, questions about the impact of StUD on OAT dispensation following release from prison have not been explored. Given the rising prevalence of stimulant use in BC among people who have been incarcerated, this study aims to investigate how having a history of stimulant use disorder impacts the uptake of opioid agonist treatment within two days of being released from prison in BC. Analyses will examine the impact of other covariates including age, sex, and concurrent mental illness diagnoses.

Methods

Study population

This study used a 20% random sample of the general population of British Columbians, contained within the British Columbia Provincial Overdose Cohort (BC-ODC). The BC-ODC brings together administrative health data and corrections data linked through BC’s Client Roster. The Roster is comprised of records for provincial health insurance which is compulsory for all residents of BC (including Canadian citizens, permanent residents, persons on visas > 6 months and their dependents).

This analysis included a cohort of people who had a record in the client roster between January 1st 2015 and December 29th 2018 with at least one release from one of BC’s ten provincial prisons in this period. OUD was determined at the time of release using a standard algorithm that includes ICD-9 and ICD-10 codes from hospital, primary care, (ICD9 = 304.0, 305.5; ICD10 = F11) and drug dispensation records dating back to January 1^st 2010. People were required to have at least one hospitalization or one primary care visit or an OAT dispensation (See Table S1). The analyses were not pre-registered, and all results should be considered exploratory.

Outcome measure

The outcome of interest was community OAT dispensation within two days of release from a provincial prison. This time frame reflects critical window of time (24–48 h) within which OAT dispensation is necessary post-release to avoid withdrawal and subsequent return to the illicit opioid supply, and was determined in consultation with an advisory group of people with lived and living experience of incarceration, opioid use, and OAT access. OAT dispensation was retrieved from PharmaNet (provincial drug dispensation database) records and reflects medications available in BC for the treatment of OUD (See Table S2). For each release from prison, OAT dispensation within two days of release was determined to be present when there was an OAT dispensation recorded in PharmaNet on the first or second day following the date of release. PharmaNet does not distinguish between medications dispensed in the prison on the day of release vs. in community on the day of release. As such, dispensations on the day of release were not considered in the outcome.

Exposure measure

The exposure of interest was StUD diagnosis. StUD was determined at the time of release using ICD codes from hospital (ICD9 = 304.2, 304.4, 305.6, 305.7) and primary care (ICD10 = F14,F15) records dating back to January 1^st 2010. The exposure was determined at the time of release for each release. As such, the exposure was time varying, and could change from one release to the next (e.g. a person with no StUD diagnosis at their first release may have one at a future release). In order to have the StUD diagnosis assigned, one hospitalization or two primary care visits with the relevant ICD9/10 code was required within 1 year of each other.

Covariates

Variables that were hypothesized to confound the relationship between StUD diagnosis and OAT dispensations were: age (categorized), sex (female or male), health authority of residence at time of admission, number of prior incarcerations (by time of release, dating back to January 1^st 2015), year of release, and mental illness diagnosis prior to release (dating back to January 1^st 2010). Mental illness was defined based on the presence of one hospitalization record or two outpatient records within one year for anxiety, depression, schizophrenia, bipolar, personality or stress disorder (See Table S3).

Data analysis

The characteristics of the sample are presented by the exposure (StUD) (Table 1) and outcome (OAT dispensation) (Table 2) at time of release, for each release. A generalized estimating equation (GEE) was used to estimate the odds of OAT dispensation within two days post-release among people with a history of StUD compared to people without. In the GEE, a logit link and exchangeable correlation structure was used to adjust for multiple releases for the same person. Unadjusted and adjusted odds ratios with corresponding 95% confidence intervals are presented (Table 3). Multivariable models included all exposure variables outlined above, and each exposure, with the exception of sex, could vary from one release to the next.

Table 1. Demographic, geographic, health, and corrections characteristics of people with an opioid use disorder diagnosis released from provincial prisons between Jan 1 2015- Dec 29 2018 (N = 13,380 releases), by stimulant use disorder diagnosis

Table 2. Demographic, geographic, health, and corrections characteristics of people with an opioid use disorder diagnosis released from provincial prisons between Jan 1 2015- Dec 29 2018 (N = 13,380 releases), by OAT dispensation within two days of release

Table 3. Unadjusted and adjusted odds ratio estimates of OAT dispensation within two days of release among people with an opioid use disorder diagnosis who were released from provincial prisons between Jan 1 2015- Dec 29 2018 (N = 13,380 releases)

In the adjusted GEE model, StUD was associated with OAT dispensation only when mental illness was adjusted for. As such, in post-hoc analyses, an interaction term between StUD diagnosis and mental illness was tested and was statistically significant. One categorical variable was created to reflect all four levels of the 2 × 2 interaction (i.e. presenting the risk factor of interest (StUD, yes vs. no) within both levels of the suspected effect modifying variable (mental illness yes and no) (Table 4). This approach produced stratified ORs, with estimates for each stratum with one single reference category, allowing for the interaction to be estimated on an additive scale. This is a necessary approach for the determination of the separate impact of each level of both variables on the outcome. In the final model, the Bonferroni correction was used to adjust for multiple comparisons (See Table S9). In sensitivity analyses, the models were rerun with different variations of the OAT outcome to test whether the associations identified held when using different outcome definitions (i.e. dispensation within 1,3,7 days, and within 2 days, including the day of release) (See Tables S4-S8). Sensitivity analyses were also conducted among a reduced sample who had an OUD diagnosis in the 1 year prior to their release (N = 5,959), and to determine whether the interaction term held over time (by year of release). The only variable with missing data was health authority. These records were classified as “Unknown” and were included in the analysis as their own level of the health authority variable. All analyses were conducted in SAS Enterprise Guide.

Table 4. Mental illness and stimulant use disorder stratum specific unadjusted and adjusted Odds ratio estimates of OAT dispensation within two days of release among people with an opioid use disorder diagnosis who were released from provincial prisons between Jan 1 2015- Dec 29 2018 (N = 13,380 releases)

Results

Among the 1,089,677 people in the 20% random sample, there were 17,930 cases of release from prison during the study period. Of these, 14,663 had an OUD diagnosis prior to their release. An additional 1,278 releases were excluded. Of these 759(59.4%) were excluded due to intermittent sentence (sentence of < 90 days served in community under conditions of parole, and with some time (usually weekends) spent in custody). The remaining 40.6% (N = 519) of excluded records were excluded because the incarceration event lasted < 1 day. The final sample included 13,380 cases of release from provincial prisons between January 1^st 2015 and December 29^th 2018 among people with an OUD diagnosis. Of these, 37.1% (N = 4,963) had a concurrent StUD diagnosis.

Females were more likely than males to have a StUD diagnosis. Nearly half of people with mental illness had a StUD diagnosis, while only approximately 15% of people without mental illness had StUD diagnosis. There were substantial regional variations in StUD diagnosis across the province, for example with approximately half of people in Vancouver Coastal region (largest urban centre in BC) and Northern region (largest rural centre) having a StUD diagnosis, and with lower proportions in the other regions of the province. Cases of release with more prior releases were more common in people with StUD diagnoses, for example, people with StUD made up approximately 30% of cases of release with 0 or 1 prior incarceration, and 45% of cases with 2 or more prior incarcerations. StUD diagnosis was increasing over time, representing approximately one-third of releases in 2015, and more than 41% of releases in 2018 (See Table 1).

Of the 13,380 cases of release, approximately 25% (N = 3,328) ended in OAT dispensation within 2 days of release. People with StUD were no less likely than people without StUD to be dispensed OAT (24.4% vs 25.1%, p = 0.375). OAT dispensation within 2 days was more likely among females compared to males (27.6% vs 24.5%, p = 0.007), among older compared to younger people (30.8% among > = 50 vs. 22.9% among < 30, p = < 0.001), and was less likely among people with mental illness compared to those without (26.4% vs. 22.1%, p < 0.001). OAT dispensation within two days of release has increased over time, with 17.8% of people accessing in 2015 compared to 29.3% in 2018. In the GEE analysis, StUD diagnosis was not associated with the outcome of OAT dispensation in unadjusted analyses, however it was in the adjusted analyses (OR(95%CI): 0.84(0.74–0.95), p = 0.006). An interaction term of StUD and mental illness and was found to be statistically significant. The GEE model was repeated adjusting for the 2 by 2 interaction of StUD and mental illness. The stratified odds ratios (with separate indicators for StUD in the presence and absence of mental illness) revealed that in the presence of mental illness, StUD was associated with lower odds of OAT dispensation (0.73(0.64–0.84), p = < 0.001), while in the absence of mental illness, this relationship did not hold (0.89(0.70–1.13), p = 0.344).

In the adjusted analyses, the odds of OAT dispensation increased with age. The youngest age group (< 30 years) had lower odds of OAT dispensation (0.63(0.47–0.86), p = 0.003) compared to the oldest age group (> = 50 years). The odds of OAT dispensation were higher in Interior Health (1.31(1.12–1.54), p = < 0.001) and Vancouver Island Health (1.64(1.40–1.93), p < 0.001)) relative to Fraser Health (the region where the majority of releases occur). The odds of OAT dispensation increased each year relative to 2015, and reached more than twice the odds of dispensation in 2018 (vs 2015) (2.15(1.82–2.52), p = < 0.001). In sensitivity analyses, the analysis was repeated using multiple variations of the OAT outcome and in all analyses, the main findings held, where StUD was associated with lower odds of OAT dispensation, only in the presence of concurrent mental illness. Among people with an OUD diagnosis in the one year prior to release (N = 5,959) a similar proportion of people accessed OAT within two days (N = 1454, 24.4%), as was found in the overall sample (N = 3328, 24.9%) in Table 2. The interaction term was tested by year and was found to be statistically significant in all years, with the exception of 2015 (p = 0.467). This could be driven by 2015 having a smaller number of releases compared to all other years, and having the lowest proportion of stimulant use disorder diagnoses and OAT access among releases, as compared to all other years.

Discussion

In this population-based study of people with OUD who were released from provincial prisons, we found that in unadjusted analyses, StUD diagnosis was not associated with a reduced odds of OAT dispensation in the two days following release. However, people with concurrent mental illness and StUD had lower odds of OAT dispensation compared to the group with mental illness alone. This suggests that among people released from BC’s prisons, people who experience concurrent health and substance use challenges alongside OUD are less likely to receive OAT post-release.

Prior studies have drawn connections between stimulant use and mental illness. For example, systematic reviews report that psychoses is highest among people who use methamphetamine or cocaine frequently and who have diagnosed dependence on these substance. However, the relationship between stimulant use and mental illness is complex, and is affected by a number of factors which vary from one person to the next, making it is difficult to draw directional or causal conclusions about this relationship. Nevertheless, stimulant use is on the rise in North America. For example, a study in the United States found that methamphetamine use nearly doubled between 2011–2017, from 18.8% to 34.2% among people with OUD. In our study, StUD diagnosis increased among people released from prison, from 32% in 2015 to 41% in 2018. This aligns with population-level analyses of drug toxicology data in BC, where methamphetamine has been detected in approximately 40% of overdose deaths. As such, the population with OUD and StUD is growing, and the health and substance use service needs of this population require increased attention. This may necessitate ongoing education for care providers, given people who use stimulants such as methamphetamine have been known to face stigma in their health care encounters which can be compounded by concurrent mental illness.

In this study, we found the odds of OAT dispensation were lowest in people aged < 30, who were approximately 35% less likely to be dispensed OAT within two days of release, compared to people aged > = 50. People aged < 30 made up the highest proportion of releases in the sample (44.2%) and therefore represent an important group whose substance use service needs must be prioritized. Prior studies have revealed that young people might be less likely to come in contact with care due to unique fears not facing adults, including concerns about involuntary detention, or disclosure of their substance use to family members. Youth are known to encounter stigma in seeking OAT and OAT dispensations are lower in youth compared to adults. In addition to efforts to expand accessibility of OAT for youth, youth advocates have recently called for a movement away from abstinence-based approaches, and ensuring the availability of confidential, peer-led interventions. These services can engage youth in harm reduction services, and may serve as a path to treatment if or when youth are ready to engage in these services.

In this analysis, we found that the odds of OAT dispensation within 2 days post-release increased over time, from 17.8% in 2015 to 29.3% in 2018. This translated to cases of release in 2018 having more than twice the odds of OAT dispensation compared to releases in 2015. This is consistent with research on OAT inside provincial prisons which found that the proportion of people receiving OAT in provincial prison in BC doubled between 2015–17. This suggests a high level of service need when transitioned to community, however we found only about 25% of people released between 2015–18 were dispensed OAT within two days of their release. This two-day window of access is particularly important given, evidence suggests withdrawal will set in within 24–48 h of last OAT dose. In cases of withdrawal without access to OAT, people are more likely to return to illicit substance use. Our team’s prior studies have found that the day of release as a time of elevated risk for overdose, further suggesting the critical need for timely OAT dispensation post-release. While nearly one third of people with OUD released from prison have access to OAT in the two days post-release, two thirds of people do not. There is a significant need for expanded interventions to reach and meet people with OUD who want access to OAT.

Efforts have been made in BC to improve the accessibility of OAT in community for people released from prison and to promote continuity of OAT prescriptions from prison to community. For example, nurses in BC’s prisons can communicate with community pharmacies and physicians to ensure they have access to complete community medical records and to bridge connections to OAT prescriptions upon release. For people facing concurrent mental illness and StUD alongside their OUD, additional low-barrier, targeted support must be provided. Given the known stigma facing people with histories of incarceration and known distrust of the health care system in this population, peer-led outreach services can play a critical role in reducing barriers to service engagement. Peer outreach workers are people who have lived experience of the same challenges facing the population, including incarceration, substance use, and/or mental health diagnoses and play an instrumental role in developing rapport with people who may need services. Prior studies have found the peer model is an effective outreach model to engage people who use substances with care. Existing peer-led programs for people released from prison such as Unlocking the Gates Services Society must be supported to expand their reach, to engage people, including those with StUD and mental illness who may face the most barriers to service access.

While OAT is effective at reducing illicit opioid use, it has also been used to connect people to other health services. For example, prior studies have found that OAT prescribers are well positioned to offer care for HCV, where prescribed medications can be effectively incorporated into OAT care. Furthermore, in some OAT clinics, physicians prescribe psychostimulants to people who use cocaine and or methamphetamine to help reduce cravings, withdrawal and to support reduced illicit stimulant use. While this practice has remained relatively limited, a recent systematic review has highlighted the effectiveness of psychostimulants in reducing illicit stimulant use and studies have shown that psychostimulant prescribing alongside OAT can promote adherence to both medications and improve psychosocial outcomes. Given the growing proportion of people released from prisons in BC who have StUD diagnoses, these medications could be prescribed alongside OAT to support reductions in illicit stimulant use.

In community, and in prison, diversified OAT options are needed to engage a wide range of people who use drugs, including people with mental illness and StUD who we have found are less likely to receive timely OAT dispensation post-release. It is possible that they have preferences for treatment, (e.g. different medications or routes of administration such as injecting or smoking) that are not currently available. As such, future research must focus on examining the service preferences of people with OUD who have concurrent StUD and mental illness. Furthermore, not everyone is ready to engage in treatment, nor abstinence, and a safer supply of alternatives to the illicit drug supply must be available to keep people alive in the context of an unregulated and unsafe illicit drug supply.

There are a number of limitations of the present study to be considered. First, we use ICD9/10 codes to define OUD and StUD. As such people who use opioids and stimulants but who have not contacted health services for care are not captured. Furthermore, the definitions of OUD and StUD rely on historical administrative health records, and do not confirm that participants met criteria for these diagnoses at the exact time of their release from incarceration. We focus on OAT dispensation within two days following release, but do not examine subsequent treatment engagement or long-term retention which could be examined in future studies in this population. Furthermore, we have not investigated the impact of stimulant use disorder diagnosis on OAT access while incarcerated, nor its impact on the continuity of OAT access between correctional settings and community. Future studies could consider the characteristics of OAT while incarcerated (e.g. dose, duration of access, timeliness of access) on continuity of OAT post-release for people with and without concurrent SUDs such as stimulant use disorder. We report on biological sex as a binary variable (male vs. female) as data on gender identity is not available.

Conclusions

Access to health and substance use services in the days immediately following release from prison is critical to reducing negative outcomes like overdose and all-cause mortality. For people with OUD, OAT remains an effective evidence-based intervention. This study demonstrates that people with mental illness and StUD have reduced odds of obtaining OAT post-release. In the context of an ongoing overdose crisis that disproportionally impacts people released from prison, there is a critical need to scale up and adapt OAT and ancillary peer support, harm reduction, treatment, and health services in order to reach people with concurrent StUD and mental illness.

Background

Individuals who have been incarcerated typically experience significantly poorer health outcomes compared to the general population. This includes a higher prevalence of infectious diseases, mental health conditions, and substance use disorders. The period immediately following release into the community represents a time of elevated risk for adverse events. Studies show that mortality rates are nearly 13 times higher in the weeks after release when compared to the general population. While several factors contribute to this increased mortality during the transition from prison to community, opioids are implicated in nearly one in eight post-release fatalities. People with a history of incarceration are known to face an increased risk of overdose, particularly in the days and weeks after their release. In British Columbia (BC), Canada’s third most populous province, approximately 70% of incarcerated individuals have a diagnosed mental health or substance use disorder. Opioids and stimulants are the substances most frequently reported upon admission to provincial prisons. Opioid agonist treatment (OAT) is the primary recommended treatment for opioid use disorder (OUD) in BC. Common forms of OAT include methadone and buprenorphine, both of which effectively reduce opioid cravings and withdrawal, decrease illicit opioid use, and offer protection against overdose mortality.

OAT provision has increased within BC’s provincial prisons in recent years. The proportion of individuals with OUD receiving OAT more than doubled from approximately 30% in 2015 to 65% in 2017. OAT access has also expanded in community settings across BC, with a broader range of medications available and new prescribing authority granted to nurse practitioners. For individuals released from prison, the timely provision of OAT is crucial. Research indicates that reduced tolerance after release significantly increases the risk of overdose on the day of and in the one to two weeks immediately following release. Furthermore, withdrawal symptoms typically begin within 24 to 48 hours of the last OAT dose. Therefore, prompt continuation of OAT in the community is essential for reducing overdose risk. Without access to OAT, individuals with OUD are more likely to engage with the illicit drug supply, which is increasingly contaminated by potent illicit opioids like fentanyl, contributing to the ongoing rise in overdose deaths across the province.

Beyond fentanyl, methamphetamine has also been increasingly detected among individuals who have died from illicit drug toxicity (overdose) in BC. Stimulant use is rising in North America among people who also use opioids. This trend is also evident among individuals admitted to BC’s prisons, where reported methamphetamine use at intake increased nearly fivefold between 2009 and 2017. This is particularly concerning because polysubstance use involving opioids is associated with a compounded risk of negative health effects, including overdose. Moreover, individuals with concurrent OUD and stimulant use disorder (StUD) often have greater healthcare needs than those with OUD alone, yet are less likely to remain engaged in care. As OAT is a primary treatment for OUD, it remains a critical intervention that should be offered promptly during the transition from prison to community for all individuals with OUD who desire access. Due to limited pharmacological and psychosocial treatments for StUD, individuals who use stimulants face significant gaps in substance use services, making those with concurrent OUD and StUD a priority for intervention. OAT can serve as an opportunity to engage this population in care, ensuring they are not further excluded from evidence-based health and substance use services.

Previous studies have shown that stimulant use can negatively affect OAT outcomes, such as reductions in illicit opioid use and long-term treatment retention. While studies have examined the impact of stimulant use on OAT retention, the specific impact of StUD on OAT dispensation immediately following release from prison has not been thoroughly explored. Given the increasing prevalence of stimulant use in BC among incarcerated individuals, this study aims to investigate how a history of stimulant use disorder affects the uptake of opioid agonist treatment within two days of release from prison in BC. Analyses will also consider the impact of other factors, including age, sex, and concurrent mental illness diagnoses.

Methods

This study utilized a 20% random sample of the general population of British Columbians, drawn from the British Columbia Provincial Overdose Cohort (BC-ODC). The BC-ODC integrates administrative health data with corrections data, linked via BC’s Client Roster, which includes records for the province's compulsory health insurance. The analysis included individuals recorded in the client roster between January 1, 2015, and December 29, 2018, who experienced at least one release from a BC provincial prison during this period. Opioid use disorder (OUD) was identified at the time of release using a standardized algorithm incorporating ICD-9 and ICD-10 codes from hospital, primary care, and drug dispensation records dating back to January 1, 2010. Participants were required to have at least one hospitalization, one primary care visit, or an OAT dispensation to be classified as having OUD. The analyses were not preregistered and are considered exploratory.

The primary outcome of interest was community OAT dispensation within two days of release from a provincial prison. This timeframe was selected to reflect the critical 24-48 hour window post-release during which OAT dispensation is necessary to prevent withdrawal and subsequent return to illicit opioid use. This timeframe was determined in consultation with an advisory group of individuals with lived experience of incarceration, opioid use, and OAT access. OAT dispensation data were retrieved from PharmaNet, the provincial drug dispensation database, and included medications available in BC for OUD treatment. For each prison release, OAT dispensation within two days was confirmed if an OAT dispensation was recorded in PharmaNet on the first or second day following the release date. PharmaNet does not distinguish between medications dispensed in prison on the day of release versus in the community. Therefore, dispensations on the day of release were not included in the outcome measure.

The exposure of interest was a stimulant use disorder (StUD) diagnosis. StUD was identified at the time of each release using ICD codes from hospital (ICD9 = 304.2, 304.4, 305.6, 305.7) and primary care (ICD10 = F14, F15) records dating back to January 1, 2010. The exposure was time-varying, meaning an individual's StUD status could change between releases. A StUD diagnosis required one hospitalization or two primary care visits with the relevant ICD9/10 code within one year of each other. Variables hypothesized to influence the relationship between StUD diagnosis and OAT dispensations were considered as covariates. These included age (categorized), sex (female or male), health authority of residence at admission, number of prior incarcerations (up to the time of release, dating back to January 1, 2015), year of release, and mental illness diagnosis prior to release (dating back to January 1, 2010). Mental illness was defined by the presence of one hospitalization record or two outpatient records within one year for anxiety, depression, schizophrenia, bipolar, personality, or stress disorder.

The characteristics of the sample were presented based on the presence of StUD and OAT dispensation at the time of release for each release. A generalized estimating equation (GEE) was employed to estimate the odds of OAT dispensation within two days post-release for individuals with a history of StUD compared to those without. The GEE used a logit link and an exchangeable correlation structure to account for multiple releases by the same individual. Both unadjusted and adjusted odds ratios with 95% confidence intervals were calculated. Multivariable models incorporated all outlined exposure variables. An interaction term between StUD diagnosis and mental illness was tested and found to be statistically significant in post-hoc analyses, as StUD was associated with OAT dispensation only when mental illness was adjusted for in the initial GEE model. A categorical variable was created to represent all four combinations of the 2x2 interaction (StUD presence/absence combined with mental illness presence/absence), providing stratified odds ratios with a single reference category. This approach allowed for estimating the interaction on an additive scale, determining the distinct impact of each variable's level on the outcome. The Bonferroni correction was applied for multiple comparisons in the final model. Sensitivity analyses were conducted by re-running models with different OAT outcome definitions (e.g., dispensation within 1, 3, or 7 days, and within 2 days including the day of release). Additional sensitivity analyses were performed on a reduced sample with an OUD diagnosis in the year prior to release and to assess if the interaction term remained significant over time. Health authority was the only variable with missing data, and these records were included as an "Unknown" category. All analyses were performed using SAS Enterprise Guide.

Results

From the 1,089,677 individuals in the 20% random sample, 17,930 prison releases occurred during the study period. Of these, 14,663 individuals had an OUD diagnosis before release. An additional 1,278 releases were excluded, primarily due to intermittent sentences (59.4%) or incarcerations lasting less than one day (40.6%). The final sample included 13,380 prison releases between January 1, 2015, and December 29, 2018, among individuals with an OUD diagnosis. Of these releases, 37.1% (4,963) also had a concurrent StUD diagnosis.

Females exhibited a higher likelihood of having a StUD diagnosis compared to males. Nearly half of individuals with a mental illness also had a StUD diagnosis, while only approximately 15% of those without mental illness had a StUD diagnosis. Significant regional variations in StUD diagnosis were observed across the province; for instance, about half of individuals in the Vancouver Coastal region (BC’s largest urban center) and Northern region (BC’s largest rural center) had a StUD diagnosis, with lower proportions in other areas. Releases involving more prior incarcerations were more common among individuals with StUD diagnoses; for example, individuals with StUD constituted about 30% of releases with zero or one prior incarceration, but 45% of releases with two or more prior incarcerations. The prevalence of StUD diagnosis increased over time, accounting for approximately one-third of releases in 2015 and over 41% of releases in 2018.

Of the 13,380 releases, approximately 25% (3,328) resulted in OAT dispensation within two days of release. Individuals with StUD were not less likely than those without StUD to receive OAT (24.4% vs 25.1%, p = 0.375). OAT dispensation within two days was more probable among females compared to males (27.6% vs 24.5%, p = 0.007), and among older individuals compared to younger individuals (30.8% for ages ≥ 50 vs. 22.9% for ages < 30, p < 0.001). Conversely, OAT dispensation was less likely among individuals with mental illness compared to those without (26.4% vs. 22.1%, p < 0.001). OAT dispensation within two days of release increased over time, rising from 17.8% in 2015 to 29.3% in 2018. In the GEE analysis, StUD diagnosis was not associated with OAT dispensation in unadjusted analyses but became significant in adjusted analyses (OR [95% CI]: 0.84 [0.74–0.95], p = 0.006). A statistically significant interaction term was found between StUD and mental illness. When the GEE model was re-run to account for this interaction, stratified odds ratios revealed that in the presence of mental illness, StUD was associated with lower odds of OAT dispensation (0.73 [0.64–0.84], p < 0.001). However, in the absence of mental illness, this relationship did not hold (0.89 [0.70–1.13], p = 0.344).

In the adjusted analyses, the odds of OAT dispensation increased with age. The youngest age group (under 30 years) had lower odds of OAT dispensation (0.63 [0.47–0.86], p = 0.003) compared to the oldest age group (50 years and older). The odds of OAT dispensation were higher in Interior Health (1.31 [1.12–1.54], p < 0.001) and Vancouver Island Health (1.64 [1.40–1.93], p < 0.001) relative to Fraser Health, the region with the majority of releases. The odds of OAT dispensation increased each year relative to 2015, reaching more than twice the odds of dispensation in 2018 compared to 2015 (2.15 [1.82–2.52], p < 0.001). In sensitivity analyses using various definitions of the OAT outcome, the main findings consistently demonstrated that StUD was associated with lower odds of OAT dispensation only in the presence of concurrent mental illness. Among individuals with an OUD diagnosis in the year prior to release (N = 5,959), a similar proportion accessed OAT within two days (N = 1,454, 24.4%) as in the overall sample (N = 3,328, 24.9%). The interaction term was statistically significant for all years except 2015 (p = 0.467), possibly due to 2015 having fewer releases and the lowest proportions of stimulant use disorder diagnoses and OAT access compared to other years.

Discussion

This population-based study of individuals with OUD released from provincial prisons found that in unadjusted analyses, a StUD diagnosis was not associated with reduced odds of OAT dispensation in the two days following release. However, individuals with concurrent mental illness and StUD showed lower odds of OAT dispensation compared to those with mental illness alone. This suggests that among individuals released from BC's prisons, those experiencing concurrent health and substance use challenges in addition to OUD are less likely to receive OAT after release.

Previous studies have linked stimulant use with mental illness. For example, systematic reviews indicate higher rates of psychosis among frequent methamphetamine or cocaine users with diagnosed dependence. However, the relationship between stimulant use and mental illness is complex, influenced by various individual factors, making it challenging to draw definitive causal conclusions. Nevertheless, stimulant use is increasing in North America. A U.S. study found that methamphetamine use nearly doubled among individuals with OUD between 2011 and 2017. In the present study, StUD diagnosis among individuals released from prison increased from 32% in 2015 to 41% in 2018. This aligns with population-level drug toxicology data in BC, where methamphetamine has been detected in approximately 40% of overdose deaths. Consequently, the population with OUD and StUD is growing, requiring increased attention to their health and substance use service needs. This may necessitate ongoing education for healthcare providers, as individuals who use stimulants, such as methamphetamine, often face stigma in healthcare encounters, which can be compounded by concurrent mental illness.

This study found that the odds of OAT dispensation were lowest among individuals aged less than 30, who were approximately 35% less likely to receive OAT within two days of release compared to those aged 50 and older. Individuals under 30 constituted the largest proportion of releases in the sample (44.2%), highlighting them as an important group whose substance use service needs require prioritization. Prior research indicates that young people may be less likely to engage with care due to unique concerns not typically faced by adults, such as fears of involuntary detention or disclosure of substance use to family members. Youth are known to experience stigma when seeking OAT, and OAT dispensations are lower among youth compared to adults. In addition to efforts to expand OAT accessibility for youth, youth advocates have recently called for moving away from abstinence-based approaches and ensuring the availability of confidential, peer-led interventions. These services can engage youth in harm reduction, potentially serving as a pathway to treatment when they are prepared to engage.

The analysis revealed that the odds of OAT dispensation within two days post-release increased over time, from 17.8% in 2015 to 29.3% in 2018. This translates to individuals released in 2018 having more than twice the odds of OAT dispensation compared to those released in 2015. This finding aligns with research indicating that the proportion of individuals receiving OAT in BC provincial prisons doubled between 2015 and 2017. Despite this increase, only about 25% of individuals with OUD released from prison between 2015 and 2018 received OAT within two days of their release. This two-day window of access is particularly important, as evidence suggests withdrawal symptoms begin within 24 to 48 hours of the last OAT dose. In cases of withdrawal without OAT access, individuals are more likely to revert to illicit substance use. Previous studies by this research team have shown that the day of release is a time of elevated overdose risk, further emphasizing the critical need for timely OAT dispensation post-release. While nearly one-third of individuals with OUD released from prison access OAT within two days post-release, two-thirds do not. There is a significant need for expanded interventions to reach and provide OAT to individuals with OUD who desire it.

Efforts in BC have aimed to improve OAT accessibility in the community for individuals released from prison and to promote continuity of OAT prescriptions from prison to community. For instance, nurses in BC’s prisons can communicate with community pharmacies and physicians to ensure access to complete community medical records and to facilitate OAT prescription connections upon release. For individuals facing concurrent mental illness and StUD alongside their OUD, additional low-barrier, targeted support is necessary. Given the known stigma experienced by individuals with histories of incarceration and their often-distrustful relationship with the healthcare system, peer-led outreach services can play a crucial role in reducing barriers to service engagement. Peer outreach workers are individuals with lived experience of similar challenges, including incarceration, substance use, and/or mental health diagnoses, and are instrumental in building rapport with individuals who may need services. Prior studies have demonstrated the effectiveness of the peer model in engaging individuals who use substances with care. Existing peer-led programs for individuals released from prison, such as Unlocking the Gates Services Society, should be supported to expand their reach and engage individuals, including those with StUD and mental illness, who may face the most significant barriers to service access.

While OAT effectively reduces illicit opioid use, it also serves as a gateway to other health services. For example, previous studies have shown that OAT prescribers are well-positioned to offer care for Hepatitis C, as prescribed medications can be effectively integrated into OAT care. Furthermore, in some OAT clinics, physicians prescribe psychostimulants to individuals who use cocaine or methamphetamine to help reduce cravings, withdrawal symptoms, and illicit stimulant use. While this practice remains relatively limited, a recent systematic review highlighted the effectiveness of psychostimulants in reducing illicit stimulant use. Studies have also shown that psychostimulant prescribing alongside OAT can enhance adherence to both medications and improve psychosocial outcomes. Given the increasing proportion of individuals released from BC prisons with StUD diagnoses, these medications could be prescribed alongside OAT to support reductions in illicit stimulant use. In both community and prison settings, diversified OAT options are needed to engage a wide range of individuals who use drugs, including those with mental illness and StUD, whom this study found are less likely to receive timely OAT dispensation post-release. It is possible these individuals have treatment preferences (e.g., different medications or routes of administration like injecting or smoking) that are not currently available. Therefore, future research must focus on examining the service preferences of individuals with OUD who also have concurrent StUD and mental illness. Moreover, not everyone is ready to engage in treatment or abstinence, and a safer supply of alternatives to the illicit drug supply must be available to preserve lives in the context of an unregulated and unsafe illicit drug supply.

Several limitations of the present study warrant consideration. First, OUD and StUD definitions rely on ICD9/10 codes from administrative health records, meaning individuals who use opioids and stimulants but have not accessed healthcare services for these conditions are not captured. Furthermore, these definitions are based on historical records and do not confirm that participants met diagnostic criteria at the exact moment of their release from incarceration. This study focused on OAT dispensation within two days following release but did not examine subsequent treatment engagement or long-term retention, which could be explored in future studies within this population. The study also did not investigate the impact of StUD diagnosis on OAT access during incarceration or its effect on the continuity of OAT access between correctional settings and the community. Future research could examine the characteristics of OAT during incarceration (e.g., dose, duration of access, timeliness of access) on post-release OAT continuity for individuals with and without concurrent substance use disorders like stimulant use disorder. Finally, biological sex was reported as a binary variable (male vs. female) due to the unavailability of data on gender identity.

Conclusions

Access to health and substance use services in the days immediately following release from prison is vital for reducing negative outcomes such as overdose and overall mortality. For individuals with opioid use disorder, opioid agonist treatment remains an effective, evidence-based intervention. This study demonstrates that individuals with concurrent mental illness and stimulant use disorder have reduced odds of obtaining OAT post-release. In the context of an ongoing overdose crisis that disproportionately affects individuals released from prison, there is a critical need to expand and adapt OAT and associated peer support, harm reduction, treatment, and health services to effectively reach individuals facing concurrent stimulant use disorder and mental illness.

Background

Individuals who have been incarcerated typically experience significantly more health problems compared to the general population. These issues often include infectious diseases, mental health conditions, and substance use disorders. The period immediately following release from prison is particularly risky, with studies indicating that mortality rates are almost 13 times higher in the weeks after release compared to the general population. Opioid use contributes to a substantial portion of these post-release deaths, with studies showing that nearly one in eight fatalities are linked to opioids. People with a history of incarceration face an elevated risk of overdose, especially in the days and weeks following their release. In British Columbia (BC), Canada, approximately 70% of incarcerated individuals have a diagnosed mental health or substance use disorder, with opioids and stimulants being the most frequently reported substances used upon admission. Opioid agonist treatment (OAT), which includes medications like methadone and buprenorphine, is a primary and effective treatment for opioid use disorder (OUD) in BC, helping to reduce cravings, withdrawal, illicit opioid use, and the risk of fatal overdose.

OAT availability in BC's provincial prisons has notably increased in recent years, with the proportion of people with OUD receiving OAT more than doubling from about 30% in 2015 to 65% in 2017. OAT access has also expanded in community settings across BC, with a wider range of medications available and nurse practitioners now having prescribing authority. For individuals leaving prison, receiving OAT promptly is crucial. Research indicates that a reduced tolerance to opioids after release heightens overdose risk, particularly on the day of and within one to two weeks following release. Furthermore, withdrawal symptoms can begin within 24 to 48 hours of the last OAT dose. Therefore, seamless continuation of medication in the community is essential for lowering overdose risk. Without timely access to OAT, individuals with OUD are more likely to turn to the illicit drug supply, which is increasingly contaminated with powerful illicit opioids like fentanyl, a major contributor to the rising overdose deaths in the province.

Beyond fentanyl, methamphetamine is increasingly detected among individuals who have died from illicit drug toxicity (overdose) in BC, and stimulant use is generally rising in North America among people who use opioids. This trend is also evident among individuals admitted to BC’s prisons, where reported methamphetamine use at intake increased almost fivefold between 2009 and 2017. Polysubstance use, particularly involving opioids and stimulants, is concerning because it can heighten the risk of adverse health effects, including overdose. People with co-occurring OUD and stimulant use disorder (StUD) often have greater healthcare needs than those with OUD alone, yet they are less likely to engage with and remain in care. OAT remains a vital intervention for OUD and should be offered promptly to all individuals with OUD transitioning from prison to community, provided they desire access. Given the limited pharmacological and psychosocial treatments available for StUD, individuals who use stimulants face significant gaps in their substance use service needs. This makes individuals with concurrent OUD and StUD a priority for intervention, and OAT can serve as an opportunity to engage this population in care, ensuring they are not excluded from evidence-based health and substance use services.

Previous research suggests that stimulant use can negatively affect OAT outcomes, such as reducing illicit opioid use and long-term treatment retention. While studies have examined how stimulant use impacts OAT retention, the specific effect of StUD on OAT dispensation immediately following release from prison has not been thoroughly explored. Due to the increasing prevalence of stimulant use among incarcerated individuals in BC, this study aimed to investigate how a history of StUD influences the initiation of OAT within two days of release from a BC provincial prison. The analysis also considered the influence of other factors, including age, sex, and co-occurring mental illness diagnoses.

Methods

Study Population

This study analyzed a 20% random sample of British Columbia’s general population, drawn from the British Columbia Provincial Overdose Cohort (BC-ODC). The BC-ODC integrates administrative health data with corrections data, linked via BC’s Client Roster. This Roster contains records for provincial health insurance, which is mandatory for all BC residents, including Canadian citizens, permanent residents, and individuals on visas exceeding six months, along with their dependents.

The analysis focused on a group of individuals listed in the client roster who experienced at least one release from one of BC's ten provincial prisons between January 1, 2015, and December 29, 2018. Opioid Use Disorder (OUD) status was determined at the time of release using a standard algorithm that incorporates ICD-9 and ICD-10 codes from hospital, primary care, and drug dispensation records dating back to January 1, 2010. For an OUD diagnosis, individuals needed at least one hospitalization, one primary care visit, or an OAT dispensation. The analyses were not preregistered, and all results should be regarded as exploratory.

Outcome Measure

The primary outcome of interest was the dispensation of community OAT within two days following release from a provincial prison. This two-day timeframe is considered crucial (24–48 hours) because OAT dispensation is necessary within this period post-release to prevent withdrawal symptoms and the subsequent return to illicit opioid use. This timeframe was established in collaboration with an advisory group comprising individuals with lived experience of incarceration, opioid use, and OAT access. OAT dispensation data was obtained from PharmaNet, the provincial drug dispensation database, and reflects medications available in BC for OUD treatment. For each prison release, OAT dispensation within two days was confirmed if a record appeared in PharmaNet on the first or second day after the release date. PharmaNet does not distinguish between medications dispensed within the prison on the release day versus those dispensed in the community on the release day. Therefore, dispensations recorded on the day of release were not included in the outcome measure.

Exposure Measure

The exposure of interest was a diagnosis of Stimulant Use Disorder (StUD). StUD was identified at the time of each release using ICD codes from hospital and primary care records dating back to January 1, 2010. As the exposure was assessed at the time of each release, it was considered time-varying, meaning an individual's StUD status could change between different release events. To be assigned an StUD diagnosis, an individual required at least one hospitalization or two primary care visits with the relevant ICD-9/10 codes within a one-year period.

Covariates

Variables hypothesized to influence the relationship between StUD diagnosis and OAT dispensation included age (categorized), sex (female or male), health authority of residence at the time of admission, number of prior incarcerations (up to the time of release, dating back to January 1, 2015), year of release, and mental illness diagnosis prior to release (dating back to January 1, 2010). Mental illness was defined by the presence of at least one hospitalization record or two outpatient records within one year for conditions such as anxiety, depression, schizophrenia, bipolar disorder, personality disorder, or stress disorder.

Data Analysis

The characteristics of the study sample were described based on the presence of StUD and OAT dispensation at the time of each release. A generalized estimating equation (GEE) was employed to estimate the likelihood of OAT dispensation within two days post-release for individuals with a history of StUD compared to those without. The GEE model utilized a logit link and an exchangeable correlation structure to account for multiple releases by the same individual. Both unadjusted and adjusted odds ratios, along with their 95% confidence intervals, were calculated. Multivariable models incorporated all previously mentioned exposure variables, with each exposure, except sex, being able to vary from one release event to the next.

In the adjusted GEE model, StUD was found to be associated with OAT dispensation only when mental illness was also considered. Consequently, in subsequent analyses, an interaction term between StUD diagnosis and mental illness was tested and found to be statistically significant. A single categorical variable was created to represent the four possible combinations of StUD and mental illness (e.g., StUD present with mental illness, StUD absent with mental illness). This approach generated stratified odds ratios, providing estimates for each stratum relative to a single reference category. This method allowed for the assessment of the independent impact of each variable's level on the outcome. In the final model, the Bonferroni correction was applied to adjust for multiple comparisons. Sensitivity analyses were conducted by re-running the models with different definitions of the OAT outcome (e.g., dispensation within 1, 3, or 7 days, or within 2 days including the day of release) to confirm the consistency of the findings. Additional sensitivity analyses were performed on a reduced sample of individuals with an OUD diagnosis in the year preceding their release and to determine if the interaction term remained significant over time (by year of release). Health authority was the only variable with missing data, and these records were included in the analysis as an "Unknown" category. All statistical analyses were performed using SAS Enterprise Guide.

Results

From a 20% random sample of 1,089,677 individuals, 17,930 prison release events occurred during the study period. Of these, 14,663 involved individuals diagnosed with OUD before their release. An additional 1,278 release events were excluded; 759 (59.4%) were due to intermittent sentences, and 519 (40.6%) involved incarceration events lasting less than one day. The final study sample included 13,380 release events from provincial prisons between January 1, 2015, and December 29, 2018, among individuals with an OUD diagnosis. Of these, 37.1% (4,963 events) involved a concurrent StUD diagnosis.

Females were more likely than males to have a StUD diagnosis. Nearly half of individuals with mental illness also had an StUD diagnosis, while only about 15% of those without mental illness had an StUD diagnosis. Significant regional differences in StUD diagnoses were observed across the province; for instance, approximately half of individuals in the Vancouver Coastal and Northern regions had an StUD diagnosis, with lower proportions in other areas. Release events involving a higher number of prior incarcerations were more common among individuals with StUD diagnoses. For example, individuals with StUD accounted for about 30% of releases with 0 or 1 prior incarceration, but 45% of releases with 2 or more prior incarcerations. The prevalence of StUD diagnoses increased over time, accounting for roughly one-third of releases in 2015 and over 41% of releases in 2018.

Of the 13,380 release events, approximately 25% (3,328 events) resulted in OAT dispensation within two days of release. Individuals with StUD were not less likely than those without StUD to receive OAT (24.4% vs. 25.1%, p=0.375). OAT dispensation within two days was more probable among females compared to males (27.6% vs. 24.5%, p=0.007) and among older individuals compared to younger ones (30.8% for those ≥50 vs. 22.9% for those <30, p<0.001). Conversely, it was less likely among individuals with mental illness compared to those without (22.1% vs. 26.4%, p<0.001). OAT dispensation within two days of release increased over time, from 17.8% in 2015 to 29.3% in 2018. The GEE analysis initially showed no association between StUD diagnosis and OAT dispensation in unadjusted analyses. However, in adjusted analyses, StUD diagnosis was associated with OAT dispensation (OR(95%CI): 0.84(0.74–0.95), p=0.006). A statistically significant interaction was found between StUD and mental illness. When the GEE model was re-run to account for this interaction, stratified odds ratios revealed that among individuals with mental illness, StUD was associated with lower odds of OAT dispensation (0.73(0.64–0.84), p<0.001). This relationship did not hold true in the absence of mental illness (0.89(0.70–1.13), p=0.344).

In the adjusted analyses, the likelihood of OAT dispensation increased with age. The youngest age group (<30 years) had lower odds of OAT dispensation (0.63(0.47–0.86), p=0.003) compared to the oldest age group (≥50 years). OAT dispensation odds were higher in the Interior Health region (1.31(1.12–1.54), p<0.001) and Vancouver Island Health region (1.64(1.40–1.93), p<0.001) relative to Fraser Health, which accounts for the majority of releases. The odds of OAT dispensation increased annually compared to 2015, more than doubling by 2018 (2.15(1.82–2.52), p<0.001). Sensitivity analyses, which used various definitions for the OAT outcome, consistently supported the main findings: StUD was associated with lower odds of OAT dispensation primarily in the presence of concurrent mental illness. A similar proportion of individuals accessed OAT within two days among a reduced sample with an OUD diagnosis in the year prior to release (24.4%). The interaction term also remained statistically significant across all years except 2015 (p=0.467), possibly due to fewer releases and lower rates of StUD diagnoses and OAT access in that year.

Discussion

This population-based study of individuals with OUD released from provincial prisons found that while StUD diagnosis alone was not associated with a reduced likelihood of OAT dispensation in the two days following release in unadjusted analyses, individuals with concurrent mental illness and StUD had significantly lower odds of receiving OAT compared to those with mental illness alone. This finding suggests that individuals leaving BC’s prisons who face complex health and substance use challenges, in addition to OUD, are less likely to obtain OAT shortly after their release.

The rising prevalence of stimulant use, particularly methamphetamine, is a growing concern. Studies indicate that methamphetamine use has significantly increased among people with OUD in North America. This trend is consistent with the study’s finding that StUD diagnoses among individuals released from BC prisons rose from 32% in 2015 to 41% in 2018, aligning with drug toxicology data showing methamphetamine in a high percentage of overdose deaths in BC. The population with co-occurring OUD and StUD is expanding, necessitating greater attention to their unique health and substance use service needs. This may require ongoing education for healthcare providers, as individuals who use stimulants often encounter stigma in healthcare settings, a challenge that can be compounded by co-occurring mental illness.

The study also observed that individuals under 30 years old had the lowest odds of receiving OAT, being approximately 35% less likely to receive it within two days of release compared to individuals aged 50 or older. This younger age group constitutes the largest proportion of releases in the sample (44.2%), underscoring the importance of prioritizing their substance use service needs. Previous research indicates that younger individuals may be less likely to seek care due to unique fears, such as concerns about involuntary detention or disclosure to family members. Youth are known to experience stigma when seeking OAT, and OAT dispensation rates are generally lower in youth than in adults. Beyond expanding OAT accessibility for youth, advocates have called for a shift away from strictly abstinence-based approaches and for the availability of confidential, peer-led interventions. These services can engage youth in harm reduction and potentially guide them towards treatment when they are ready.

Over time, OAT dispensation within two days post-release increased, rising from 17.8% in 2015 to 29.3% in 2018. This meant that individuals released in 2018 were more than twice as likely to receive OAT compared to those released in 2015. This aligns with findings that OAT access within provincial prisons in BC also doubled between 2015 and 2017. Despite these improvements, only about 25% of individuals released between 2015 and 2018 received OAT within two days of release. This two-day window is critical because withdrawal symptoms can manifest within 24 to 48 hours of the last OAT dose, increasing the likelihood of returning to illicit substance use without timely access. Prior studies have also highlighted the day of release as a period of heightened overdose risk, further emphasizing the urgent need for prompt OAT dispensation. While nearly one-third of individuals with OUD released from prison access OAT within two days, two-thirds do not, pointing to a significant need for expanded interventions to reach individuals with OUD who desire OAT.

Efforts in BC have aimed to enhance OAT accessibility for individuals released from prison and to ensure continuity of OAT prescriptions from correctional facilities to the community. For example, prison nurses in BC can communicate with community pharmacies and physicians to facilitate access to comprehensive medical records and arrange OAT prescriptions upon release. For individuals facing concurrent mental illness and StUD in addition to OUD, additional low-barrier, targeted support is essential. Given the known stigma associated with a history of incarceration and a general distrust of the healthcare system within this population, peer-led outreach services can play a vital role in reducing barriers to engaging with services. Peer outreach workers, who have personal experience with incarceration, substance use, or mental health diagnoses, are instrumental in building trust with individuals who may need services. Research indicates that the peer model is an effective outreach strategy for engaging individuals who use substances in care. Existing peer-led programs for individuals released from prison, such as Unlocking the Gates Services Society, should receive support to expand their reach and engage those who face the most significant barriers to service access, including individuals with StUD and mental illness. OAT is effective not only in reducing illicit opioid use but also in connecting individuals to other health services. For instance, OAT prescribers are well-positioned to offer care for Hepatitis C, as relevant medications can be integrated into OAT care. Additionally, some OAT clinics prescribe psychostimulants to individuals who use cocaine or methamphetamine to help reduce cravings and withdrawal and to support a decrease in illicit stimulant use. While this practice is still somewhat limited, recent reviews highlight the effectiveness of psychostimulants in reducing illicit stimulant use, and studies suggest that prescribing psychostimulants alongside OAT can improve adherence to both medications and enhance psychosocial outcomes. Given the increasing proportion of individuals released from BC prisons with StUD diagnoses, these medications could be prescribed alongside OAT to support reductions in illicit stimulant use.

A variety of OAT options are needed in both community and prison settings to engage a diverse range of individuals who use drugs, especially those with mental illness and StUD, who, as this study found, are less likely to receive timely OAT dispensation post-release. It is possible that these individuals have treatment preferences (e.g., different medications or routes of administration like injecting or smoking) that are not currently available. Therefore, future research should focus on understanding the service preferences of individuals with OUD who also have concurrent StUD and mental illness. Moreover, not everyone is ready or desires to engage in treatment or abstinence. In the context of an unregulated and unsafe illicit drug supply, a safer supply of alternatives must be available to preserve lives. This study has several limitations. First, OUD and StUD definitions rely on ICD-9/10 codes from administrative health records, meaning individuals who use opioids and stimulants but have not accessed health services for care are not included. Furthermore, these definitions are based on historical records and do not confirm that participants met diagnostic criteria at the exact time of their release. This study focused solely on OAT dispensation within two days of release and did not examine subsequent treatment engagement or long-term retention, which could be areas for future research. The study also did not investigate the impact of StUD diagnosis on OAT access during incarceration or its effect on the continuity of OAT access between correctional settings and the community. Future studies could explore how characteristics of OAT during incarceration (e.g., dose, duration, timeliness of access) influence post-release OAT continuity for individuals with and without concurrent substance use disorders. Finally, biological sex was reported as a binary variable (male vs. female) due to the unavailability of data on gender identity.

Conclusions

Access to health and substance use services immediately following release from prison is essential for reducing negative outcomes, including overdose and overall mortality. For individuals with OUD, OAT remains an effective, evidence-based intervention. This study highlights that individuals with co-occurring mental illness and StUD have lower odds of obtaining OAT after release. In the context of an ongoing overdose crisis disproportionately affecting individuals released from prison, there is a critical need to expand and adapt OAT and supportive services such as peer support, harm reduction, treatment, and general health services to effectively reach individuals facing concurrent StUD and mental illness.

Background

People in prison often face more health problems than the general population, including infectious diseases, mental health issues, and substance use disorders. The time shortly after being released from prison is particularly risky, with studies showing that mortality rates are nearly 13 times higher in the weeks following release compared to the general population. Opioids contribute to a significant number of these deaths. Individuals with a history of incarceration face a high risk of overdose, especially soon after release, because their tolerance to drugs may decrease while in prison. In British Columbia (BC), Canada, about 70% of people in provincial prisons have a mental health or substance use disorder, with opioids and stimulants being the most commonly reported substances used upon entry. Opioid agonist treatment (OAT) is a main treatment for opioid use disorder (OUD) in BC, available as medications like methadone and buprenorphine. These treatments help reduce cravings and withdrawal symptoms, decrease illegal opioid use, and protect against overdose.

Access to OAT has increased in BC’s provincial prisons and in community settings. For those released from prison, receiving OAT quickly is vital. Research indicates that a lower drug tolerance after release can increase overdose risk immediately following release and in the first week or two. Also, withdrawal symptoms can start within 24 to 48 hours after the last OAT dose. Therefore, continuing OAT without delay in the community is essential to prevent overdose. Without OAT, people with OUD are more likely to use illegal drugs, which are often contaminated with powerful substances like fentanyl, contributing to the rise in overdose deaths.

Alongside fentanyl, methamphetamine use has also risen among people who have died from illicit drug toxicity (overdose) in BC. Stimulant use is increasing in North America among people who also use opioids. This trend is also seen among those admitted to BC’s prisons, where reported methamphetamine use at intake nearly quadrupled between 2009 and 2017. This is concerning because using both opioids and stimulants together increases the risk of serious health problems, including overdose. People with both OUD and stimulant use disorder (StUD) often have greater healthcare needs than those with OUD alone, yet they are less likely to stay connected to care. OAT is a primary treatment for OUD and should be offered promptly to all who want it during the transition from prison to community. Since there are few medical or counseling treatments for StUD, people who use stimulants face significant gaps in meeting their substance use service needs. This makes individuals with both OUD and StUD a priority for specialized support. OAT can provide an opportunity to connect this group with necessary health services.

Previous studies have shown that stimulant use can negatively affect OAT outcomes, such as reducing illicit opioid use and long-term participation in treatment. While studies have looked at how stimulant use impacts staying in OAT, questions about how StUD affects getting OAT after prison release have not been fully explored. Given the rising number of people with stimulant use in BC who have been incarcerated, this study aimed to investigate how a history of StUD affects receiving OAT within two days of being released from prison in BC. The analysis also looked at other factors such as age, sex, and other mental health diagnoses.

Methods

Study population

This study used a 20% random sample of people in British Columbia, taken from a larger cohort called the British Columbia Provincial Overdose Cohort (BC-ODC). This cohort combines health and corrections data through BC’s Client Roster, which includes records for provincial health insurance that all residents must have.

The study focused on people recorded in the client roster between January 1, 2015, and December 29, 2018, who were released from one of BC’s ten provincial prisons at least once during this time. Opioid use disorder (OUD) was identified at the time of release using standard medical codes from hospital and primary care records, and drug prescription records dating back to January 1, 2010. Individuals needed at least one hospitalization, one primary care visit, or an OAT prescription to be considered as having OUD. The analyses were exploratory, meaning they were not planned in advance.

Outcome measure

The main outcome of interest was receiving OAT in the community within two days of being released from a provincial prison. This two-day period is crucial (24-48 hours) because OAT is needed quickly after release to prevent withdrawal and a return to using illegal opioids. This timeframe was chosen after consulting with an advisory group of people who had personal experience with incarceration, opioid use, and OAT access. Information about OAT dispensation came from PharmaNet, the provincial drug prescription database, and included medications available in BC for OUD treatment. For each prison release, OAT dispensation within two days was recorded if a prescription appeared in PharmaNet on the first or second day after release. Prescriptions on the day of release were not included, as PharmaNet does not differentiate between those given in prison and those given in the community on the same day.

Exposure measure

The main factor being examined was a diagnosis of stimulant use disorder (StUD). StUD was identified at the time of release using medical codes from hospital and primary care records dating back to January 1, 2010. This factor could change for a person with each new release (e.g., someone might not have an StUD diagnosis at their first release but have one at a later release). To be assigned an StUD diagnosis, an individual needed one hospitalization or two primary care visits with the relevant medical code within one year of each other.

Covariates

Variables that were thought to affect the link between an StUD diagnosis and OAT prescriptions included: age (grouped into categories), sex (female or male), the health region where the person lived at the time of admission, the number of previous incarcerations (up to the time of release, dating back to January 1, 2015), the year of release, and a mental illness diagnosis before release (dating back to January 1, 2010). Mental illness was defined as having one hospitalization record or two outpatient records within one year for conditions such as anxiety, depression, schizophrenia, bipolar disorder, personality disorder, or stress disorder.

Data analysis

The characteristics of the study group were presented based on whether they had StUD and whether they received OAT at the time of each release. A statistical method called a generalized estimating equation (GEE) was used to estimate the likelihood of receiving OAT within two days after release for people with a history of StUD compared to those without. This method accounted for the fact that the same person might have multiple releases. Unadjusted and adjusted odds ratios, along with their 95% confidence intervals, were calculated. The adjusted models included all the listed exposure variables, and most of these variables, except sex, could vary for a person across different releases.

In the adjusted GEE model, StUD was linked to OAT dispensation only when mental illness was also considered. Therefore, a further analysis looked at how StUD and mental illness interacted, and this interaction was found to be statistically significant. A single categorical variable was created to show all four combinations of StUD (yes/no) and mental illness (yes/no). This approach provided separate odds ratios for each combination, with a single reference group, which helped understand the individual impact of each variable. For the final model, a statistical correction (Bonferroni) was used to adjust for multiple comparisons. Additional analyses were performed to see if the findings held true with different definitions of the OAT outcome (e.g., dispensation within 1, 3, or 7 days, or within 2 days including the day of release). These analyses were also conducted on a smaller group of people who had an OUD diagnosis in the year before their release, and to check if the interaction between StUD and mental illness remained significant over different years. Only health authority data had missing information, which was categorized as "Unknown" and included in the analysis. All analyses were conducted using SAS Enterprise Guide software.

Results

From a sample of nearly 1.1 million people, there were 17,930 prison releases during the study period. Of these, 14,663 individuals had an OUD diagnosis before their release. After excluding some cases (e.g., short sentences or very short incarcerations), the final study group included 13,380 releases from provincial prisons between 2015 and 2018 for people with an OUD diagnosis. Among these releases, 37.1% (4,963) also had a concurrent StUD diagnosis.

Females were more likely than males to have an StUD diagnosis. Nearly half of individuals with a mental illness also had an StUD diagnosis, compared to only about 15% of those without a mental illness. StUD diagnoses varied significantly by region across the province; for instance, about half of people in Vancouver Coastal and Northern regions had an StUD diagnosis, while other regions had lower proportions. Releases involving more prior incarcerations were more common among people with StUD diagnoses. StUD diagnoses also increased over time, accounting for about one-third of releases in 2015 and over 41% in 2018.

Of the 13,380 prison releases, approximately 25% (3,328) resulted in OAT dispensation within two days of release. People with StUD were about as likely as those without StUD to receive OAT (24.4% vs 25.1%). Receiving OAT within two days was more likely among females (27.6% vs 24.5%), and among older individuals (30.8% for those 50 and older vs. 22.9% for those under 30). It was less likely among people with a mental illness (22.1% vs. 26.4%). OAT dispensation within two days of release increased over time, from 17.8% in 2015 to 29.3% in 2018. In the statistical analysis, StUD diagnosis was not initially linked to OAT dispensation. However, when other factors were adjusted, StUD was associated with lower odds of OAT dispensation. A statistically significant interaction was found between StUD and mental illness. When this interaction was considered, the analysis showed that for individuals with a mental illness, StUD was associated with a lower likelihood of receiving OAT. However, this relationship did not hold true for those without a mental illness.

In the adjusted analyses, the likelihood of receiving OAT increased with age. Individuals under 30 were less likely to receive OAT compared to those 50 and older. The odds of receiving OAT were higher in the Interior Health and Vancouver Island Health regions compared to Fraser Health. The likelihood of OAT dispensation increased each year relative to 2015, more than doubling by 2018. In additional analyses with different definitions of the OAT outcome, the main findings remained consistent: StUD was linked to lower odds of OAT dispensation only when a concurrent mental illness was present. The interaction between StUD and mental illness was significant in all years except 2015, which might be due to fewer releases and lower rates of StUD diagnoses and OAT access that year.

Discussion

This study of people with OUD released from provincial prisons found that while an StUD diagnosis alone was not initially linked to a lower chance of receiving OAT within two days of release, people with both a mental illness and StUD had reduced odds of getting OAT compared to those with a mental illness alone. This suggests that among individuals released from BC’s prisons, those facing multiple health and substance use challenges alongside OUD are less likely to receive OAT after release.

Previous research has noted connections between stimulant use and mental illness. For example, studies report higher rates of psychosis among frequent methamphetamine or cocaine users who are dependent on these substances. However, the relationship is complex and influenced by many individual factors, making it hard to draw direct cause-and-effect conclusions. Nevertheless, stimulant use is increasing in North America. In this study, StUD diagnoses rose among people released from prison, from 32% in 2015 to 41% in 2018. This matches province-wide data showing methamphetamine in about 40% of overdose deaths in BC. The population with both OUD and StUD is growing, requiring more attention to their health and substance use service needs. This may mean more training for healthcare providers, as people who use stimulants like methamphetamine often face stigma in healthcare, which can be worse if they also have a mental illness.

This study also found that the odds of receiving OAT were lowest for people under 30, who were about 35% less likely to receive OAT within two days of release compared to those aged 50 and older. Individuals under 30 made up the largest group in the study sample (44.2%), highlighting them as an important group whose substance use service needs must be prioritized. Previous studies suggest that young people may be less likely to seek care due to unique fears, such as concerns about involuntary detention or their substance use being revealed to family. Youth are known to encounter stigma when seeking OAT, and OAT dispensation rates are lower in youth compared to adults. Besides expanding OAT accessibility for youth, advocates have recently called for moving away from only abstinence-focused approaches and ensuring confidential, peer-led interventions. These services can engage youth in harm reduction and potentially lead them to treatment when they are ready.

The analysis showed that the odds of receiving OAT within two days of release increased over time, from 17.8% in 2015 to 29.3% in 2018. This means that individuals released in 2018 were more than twice as likely to receive OAT compared to those released in 2015. This aligns with research showing that the proportion of people receiving OAT in BC provincial prisons doubled between 2015 and 2017. While this suggests a high need for services during the transition to the community, only about 25% of people released between 2015 and 2018 received OAT within two days of release. This two-day window is critical because withdrawal symptoms can begin within 24 to 48 hours after the last OAT dose. Without OAT to manage withdrawal, individuals are more likely to return to using illicit substances. Previous studies have also shown that the day of release carries a high risk for overdose, further emphasizing the urgent need for timely OAT dispensation after release. Although nearly one-third of people with OUD released from prison access OAT in the two days after release, two-thirds do not. There is a significant need for expanded interventions to reach and support people with OUD who want access to OAT.

Efforts have been made in BC to improve OAT accessibility in the community for people released from prison and to ensure OAT prescriptions continue from prison to community. For example, nurses in BC’s prisons can communicate with community pharmacies and doctors to ensure complete medical records are available and to arrange OAT prescriptions upon release. For people facing concurrent mental illness and StUD alongside their OUD, additional easy-to-access, targeted support must be provided. Given the known stigma faced by people with a history of incarceration and their distrust of the healthcare system, peer-led outreach services can greatly help reduce barriers to engaging with services. Peer outreach workers are individuals who have personal experience with similar challenges, such as incarceration, substance use, or mental health diagnoses, and they play a vital role in building trust with people who need services. Studies have shown that the peer model is an effective outreach approach to connect people who use substances with care. Existing peer-led programs for people released from prison, such as Unlocking the Gates Services Society, should receive support to expand their reach and engage individuals, especially those with StUD and mental illness, who may face the most barriers to service access.

Both in the community and in prison, a wider variety of OAT options are needed to engage diverse groups of people who use drugs, including those with mental illness and StUD who were found less likely to receive timely OAT after release. It is possible these individuals have preferences for treatment (e.g., different medications or ways of taking them, like injecting or smoking) that are not currently available. Therefore, future research should focus on understanding the service preferences of people with OUD who also have StUD and mental illness. Furthermore, not everyone is ready for treatment or abstinence, and a safer supply of alternatives to the illicit drug supply must be available to keep people alive in the context of an unregulated and dangerous illicit drug supply.

There are several limitations to this study. First, OUD and StUD diagnoses were based on medical codes, meaning people who use opioids and stimulants but have not sought healthcare services for these conditions were not included. Also, the definitions of OUD and StUD rely on past health records and do not confirm that individuals met diagnostic criteria at the exact moment of their release. The study focused on OAT dispensation within two days after release but did not examine subsequent treatment engagement or long-term participation, which could be explored in future studies. Furthermore, the study did not investigate how StUD diagnosis affects OAT access while incarcerated, nor its impact on the continuity of OAT access between correctional facilities and the community. Future studies could consider characteristics of OAT during incarceration (e.g., dose, duration of access, timeliness) and their effect on post-release OAT continuity for people with and without concurrent substance use disorders like StUD. Biological sex was reported as a binary variable (male vs. female) because data on gender identity were not available.

Conclusions

Access to health and substance use services in the days immediately following release from prison is vital for reducing negative outcomes such as overdose and death from all causes. For people with OUD, OAT remains an effective, evidence-based intervention. This study shows that people with mental illness and StUD have a lower chance of getting OAT after release. In the context of an ongoing overdose crisis that disproportionately affects people released from prison, there is a critical need to expand and adapt OAT and related services like peer support, harm reduction, treatment, and health services to reach individuals with concurrent StUD and mental illness.

Background

People who have been in prison often have more health problems than others. These can include infections, mental health issues, and problems with drug or alcohol use. The time right after someone leaves prison and returns to the community is very risky. Studies show that death rates are almost 13 times higher in the weeks after release compared to the general population. Opioid use plays a big role in these deaths, causing nearly 1 in 8 deaths after release. People who have been in prison are known to have a higher chance of overdose, especially soon after they get out.

In British Columbia (BC), Canada, about 70 out of every 100 people in prison have a mental health problem or a substance use disorder. The most common drugs they report using when they enter prison are opioids and stimulants. Opioid Agonist Treatment (OAT) is the main way to treat opioid use disorder in BC. OAT medicines like methadone and buprenorphine help reduce cravings, stop withdrawal, cut down on illegal opioid use, and prevent overdose deaths.

More people in BC's prisons have been getting OAT in recent years. The number of people with opioid use disorder who received OAT more than doubled from 2015 to 2017. OAT has also become more available in communities in BC, with more types of medicine and more health workers like nurses able to prescribe it. For people leaving prison, it is very important to get OAT quickly. This is because tolerance to opioids goes down after being away from them, which raises the risk of overdose on the day of release and in the next one to two weeks. Also, withdrawal symptoms can start within one or two days after the last OAT dose. So, getting the medicine quickly in the community is key to preventing overdose. Without OAT, people with opioid use disorder are more likely to use street drugs, which are often mixed with strong drugs like fentanyl. Fentanyl has caused a rise in overdose deaths.

Along with fentanyl, methamphetamine has also been found more often in people who have died from drug overdose in BC. The use of stimulants is going up among people who use opioids in North America. This is also true for people entering BC's prisons, where reported methamphetamine use went up almost five times between 2009 and 2017. This is worrying because using both opioids and stimulants can lead to even higher health risks, like overdose. People who use both opioids and stimulants often need more health services than people who only use opioids, but they are less likely to stay in care. OAT is a very important treatment for opioid use disorder and should be given quickly to all who want it when leaving prison. Because there are not many medicines or other treatments for stimulant use disorder, people who use stimulants face big gaps in the help they need. This makes people who use both opioids and stimulants a top group to help. OAT can be a way to get these people into care, and they should not be left out from getting health services that are proven to work.

Studies have shown that stimulant use can make OAT less effective in reducing illegal opioid use and in helping people stay in treatment long-term. While some studies have looked at how stimulant use affects staying on OAT, questions about how stimulant use disorder affects getting OAT right after leaving prison have not been answered. Because more and more people who have been in prison in BC are using stimulants, this study looked at how having a history of stimulant use disorder impacts getting OAT within two days of leaving prison in BC. The study also looked at other factors like age, sex, and having a mental illness.

Methods

This study looked at records for 20 out of every 100 people living in British Columbia. These records came from health data and prison data, linked together. The study included people who were released from one of BC's ten provincial prisons between January 1, 2015, and December 29, 2018. Opioid use disorder was identified using health records, including hospital visits, doctor visits, and medicine records.

The main thing the study measured was whether a person received OAT in the community within two days of leaving prison. This two-day time frame is important because it is when OAT is needed to prevent withdrawal and going back to using street drugs. The time frame was chosen with advice from people who have been in prison, used opioids, and received OAT. Records from the provincial medicine database showed when OAT was given.

The study also looked at whether a person had a stimulant use disorder. This was found using health records from hospitals and doctor visits. The study also considered other factors that might affect OAT access. These included a person's age, whether they were female or male, where they lived, how many times they had been in prison before, the year they were released, and if they had a mental illness. The study looked at these factors to see how they might influence whether a person got OAT.

Results

Among the people in the study, 13,380 were released from provincial prisons between January 1, 2015, and December 29, 2018, and had an opioid use disorder. Out of these, 37.1% (about 4 out of 10) also had a stimulant use disorder.

More females than males had a stimulant use disorder. Almost half of the people with a mental illness also had a stimulant use disorder, while only about 15% of people without a mental illness did. The number of people with stimulant use disorder varied across different parts of BC. For example, about half the people in Vancouver Coastal and Northern regions had a stimulant use disorder. People with stimulant use disorder had also been in prison more often. The number of people with stimulant use disorder went up over time, from about one-third of releases in 2015 to more than 41% in 2018.

About 25% (1 in 4) of the people released received OAT within two days. People with stimulant use disorder were just as likely as those without it to get OAT (24.4% vs 25.1%). Getting OAT within two days was more likely for females than males, and for older people compared to younger people. It was less likely for people with a mental illness. Getting OAT within two days of release increased over time, from 17.8% in 2015 to 29.3% in 2018.

When other factors were considered, having a stimulant use disorder alone was not linked to getting OAT. However, when mental illness was considered, a link was found. People with both mental illness and stimulant use disorder were less likely to get OAT. For people with mental illness, having a stimulant use disorder meant they were less likely to get OAT. But for people without mental illness, stimulant use disorder did not change their chances of getting OAT.

Older people were more likely to get OAT. Those under 30 were about 35% less likely to get OAT than those 50 or older. Getting OAT was more common in some regions of BC than others. The chances of getting OAT increased each year, more than doubling by 2018 compared to 2015. The study's main findings remained true even when different ways of looking at OAT access were used. The link between stimulant use disorder and mental illness also held true for most years, but not 2015, possibly because there were fewer releases that year and less stimulant use disorder and OAT access overall.

Discussion

This study looked at people with opioid use disorder who were released from prisons in BC. It found that having a stimulant use disorder alone was not linked to getting OAT quickly after release. However, people who had both a mental illness and a stimulant use disorder were less likely to get OAT. This means that people leaving prison who have more health and substance use challenges alongside opioid use disorder are less likely to receive OAT after release.

Other studies have shown a link between stimulant use and mental illness, though it is a complex one. Stimulant use is increasing in North America. This study found that stimulant use disorder also increased among people released from prison in BC, from 32% in 2015 to 41% in 2018. This matches findings that methamphetamine is found in about 40% of overdose deaths in BC. Since the number of people with both opioid use disorder and stimulant use disorder is growing, their health and substance use needs require more attention. Health workers may need more training, as people who use stimulants, especially with mental illness, can face judgment in health care settings.

The study found that people under 30 were less likely to get OAT quickly after release, compared to those 50 or older. This younger group made up the largest number of releases (44.2%), so their needs for substance use services are very important. Younger people might be less likely to seek care due to concerns about being held against their will or family finding out about their drug use. They can also face judgment when trying to get OAT. Besides making OAT easier for young people to get, there have been calls for services that are not focused on stopping all drug use, and that offer confidential help from others with similar experiences. These services can help young people with safer drug use practices and can be a path to treatment when they are ready.

The study also showed that OAT access within two days after release increased over time, from 17.8% in 2015 to 29.3% in 2018. This means that by 2018, people released were more than twice as likely to get OAT compared to 2015. This matches other research showing more OAT use inside BC prisons. Even with this increase, only about 1 in 4 people released between 2015 and 2018 got OAT within two days. This two-day window is very important because withdrawal can start within one to two days of the last OAT dose. Without OAT during withdrawal, people are more likely to use illegal substances. Prior studies show the day of release is a time of high overdose risk, highlighting the critical need for quick OAT access. While almost one-third of people with opioid use disorder released from prison get OAT within two days, two-thirds do not. There is a great need for more help to reach and support people with opioid use disorder who want OAT.

Efforts have been made in BC to make OAT more available in the community for people leaving prison and to keep OAT prescriptions going from prison to the community. For example, nurses in BC prisons can talk with community pharmacies and doctors to make sure people have full medical records and connections for OAT when they are released. For people facing mental illness and stimulant use disorder alongside their opioid use disorder, extra, easy-to-access support is needed. Since people who have been in prison often face judgment and do not trust the health care system, services led by peers can be very helpful. Peer workers are people who have gone through similar challenges, like being in prison, using drugs, or having mental health issues. They are very important in building trust with people who need help. Studies have shown that peer-led programs are good at connecting people who use substances with care. Programs like Unlocking the Gates Services Society, which are run by peers for people leaving prison, need more support to reach more people, including those with stimulant use disorder and mental illness who may face the biggest hurdles to getting help.

OAT can also connect people to other health services. For example, doctors who prescribe OAT can also offer care for Hepatitis C. Also, some OAT clinics offer certain medicines to people who use cocaine or methamphetamine to help with cravings and withdrawal, and to support less use of illegal stimulants. While this practice is not widespread, studies have shown these medicines can reduce illegal stimulant use and help people stick to both OAT and other medicines, improving their overall well-being. Given the growing number of people released from prisons in BC who have stimulant use disorder, these medicines could be given along with OAT to help reduce illegal stimulant use.

Both in the community and in prison, more choices for OAT are needed to help many different people who use drugs, including those with mental illness and stimulant use disorder, who this study found are less likely to get OAT quickly after release. It is possible they prefer different treatments (like different medicines or ways to take them) that are not currently available. Future research needs to look at what kinds of services people with opioid use disorder, stimulant use disorder, and mental illness prefer. Also, not everyone is ready for treatment or to stop using drugs completely. A safer supply of drugs, as an alternative to the illegal street drug supply, is needed to keep people alive given the current unsafe street drug market.

This study has some limits. It used medical codes to identify opioid use disorder and stimulant use disorder. This means it did not include people who use opioids or stimulants but have not seen a health worker for care. Also, these definitions rely on old records and do not confirm that people had these disorders exactly when they were released from prison. The study focused on getting OAT within two days after release, but did not look at how long people stayed in treatment later. It also did not look at how stimulant use disorder affects getting OAT while in prison or keeping OAT going between prison and the community. Future studies could look at how OAT in prison affects getting OAT after release for people with and without stimulant use disorder. The study only reported male and female, as information on other gender identities was not available.

Conclusions

Getting health and substance use services in the days right after leaving prison is very important to prevent bad outcomes like overdose and death. For people with opioid use disorder, OAT is a proven treatment that works. This study shows that people with mental illness and stimulant use disorder are less likely to get OAT after release from prison. With the ongoing overdose crisis, which affects people released from prison more, there is a clear need to increase and adapt OAT services, along with peer support, harm reduction, treatment, and other health services. This is especially true for people who have both stimulant use disorder and mental illness.