Table 1 Summary of Articles Included in Scoping Review of Opioid Overdose Prevention for Justice-Involved Populations

 Bennett & Holloway, 2012 [14]

To determine the impact of naloxone training on knowledge of opiate overdose and confidence and willingness to take appropriate action and to examine the use of naloxone and other harm reduction actions at the time of overdose events

Pre/post training intervention survey

521 opiate users and 4 non-opiate users sampled from 5 community sites (362) and 3 prisons (163); comparison sample of agency staff in Wales; 83% male, with a mean age of 32.9; 69% were currently in treatment

• Among the study participants, 68% had used heroin in the last 28 days, 44% had previously overdosed, 75% had witnessed an overdose in their lifetime, 30% had witnessed a fatal overdose

• Knowledge about how to recognise and respond to overdose events increased among trainees across all measures.

• Confidence to administer naloxone increased from 67 to 92% and the proportion of clients who were confident to take appropriate actions at the scene of an overdose increased from a 77% pre-training to 93% post-training; the proportion of clients who were willing to take appropriate actions increased from 91 to 97%.

• Over the course of the study, there were 28 recorded uses of naloxone, resulting in 27 recoveries and one fatality

• Training in OD management and the use of naloxone can bring about significant improvements in knowledge and willingness to take action.

• THN trainees also demonstrated that they were able to use naloxone successfully in OD events

 Cropsey, Martin, Clark, McCullumsmith, Lane, Hardy, Hendricks, & Redmond, 2013 [15]

To describe the characteristics, history of overdose, and response to overdose among a community corrections sample

Survey

478 adults under community corrections supervision in Alabama; 67% male

• 40% had lifetime history of opioid use; 40% of these had a history of opioid overdose

• OD history was associated with being female, white, higher education, and willing to receive training on Naloxone use; they were also 2–3 times more likely to have witnessed an overdose or known someone who died from opioid overdose

• In response to an overdose, those who had a history of overdose were more likely than others to provide some intervention, most often calling 911 (59%) and transporting the individual to hospital ED (33%), although 23% provided no intervention and only 4% administered Naloxone

• Past history of overdose increased willingness to be trained on naloxone administration and to intervene in an overdose situation, although prior administration of naloxone was low

 Curtis, Dietze, Aitken, Kirwan, Kinner, Butler, & Stoové, 2018 [16]

To evaluate the acceptability of THN in a cohort of male prisoners

Baseline interviews with participants in a longitudinal cohort study

380 men from the Prison and Transition Health (PATH) Cohort Study, all of whom reported regular IDU in the 6 months prior to incarceration; Victoria, Australia

• 81% of participants reported willingness to undertake THN training prior to release.

• Most were willing to resuscitate a friend using THN if they were trained (94%) and to be revived by a trained peer (91%) using THN.

• More than 10 years since first injection, having witnessed an opioid overdose in the last 5 years, having ever received alcohol or other drug treatment in prison, and injecting drugs during the current prison sentence were significantly associated with increased odds of willingness to participate in a prison THN program.

• Not specifying whether they had injected during their prison sentence was associated with decreased odds of willingness to participate in a prison THN training.

• Identification of correlates of willingness to participate in training, such as longer histories of IDU and exposure to SUD treatment in prison, provide useful information for targeting the promotion and delivery of prison-based THN programmes

 Davidson, Wagner, Tokar, & Scholar, 2019 [17]

To identify individuals incarcerated in jail who are most likely to benefit from overdose prevention and response (OPR) programs.

Survey

3781 jail inmates (3315 men, 466 women) in Los Angeles, CA; 17% of survey sample reported using opioids within the last 12 months

• 7% reported witnessing an overdose within the last 12 months

• 5% report ever having received MAT

• 39% reported interest in being trained in overdose prevention and response.

• The single largest predictor of interest in OPR was being present at an overdose in the past year.

• Overdose Prevention and Response training should be provided to all inmates who opt-in to receive training regardless of other risk factors.

• Incarceration could represent a significant opportunity to provide evidence-based treatments, including MAT.

 Gicquelais, Mezuk, Foxman, Thomas, & Bohnert, 2019 [18]

To obtain information from justice-involved individuals in a drug treatment program that can be used to inform OEND planning

Survey

514 adults sampled from residential SUD treatment program, whose treatment was prompted by the CJS and had a history of heroin use or opioid misuse in Michigan

• 56% of participants correctly identified naloxone as an opioid overdose treatment, although 68% had experienced an overdose and 79% had witnessed another person overdose.

• Two latent justice involvement classes were identified (low and high), however, justice involvement was not associated with naloxone knowledge.

• Male participants who had personally overdosed more often identified naloxone as an overdose treatment after adjustment for covariates

• All individuals with OUD in criminal justice diversion programs could benefit from OEND given the high propensity to experience and witness overdoses and low naloxone knowledge across justice involvement backgrounds among both men and women

 Holloway, Hills, & May, 2018 [19]

To apply the concept of the ‘risk environment’ to examine how witnesses respond to opiate overdose; and to examine the micro- and macro-level factors the impede the implementation of harm reduction techniques in response to an overdose

Semi-structured interviews

55 participants recruited from statutory and third sector drug treatment providers in 5 towns/cities in South Wales and in two Welsh prisons; all had ever used heroin and 95% had ever injected; 78% had ever been in prison and 47% were currently in prison

• Witnesses were amenable to assisting overdosed peers.

• Both micro- and macro-level factors impeded the successful implementation of harm reduction techniques in response to an overdose.

• At micro level, the social setting of injecting drug use, peer group drug use norms, difficulties in identifying an overdose, and panic and confusion were barriers to responding.

• Witnesses acknowledged that their own intoxication occasionally limited their ability to identify overdose and administer response techniques (e.g. CPR/naloxone).

• Some respondents did not alert EMS because they feared prosecution for either being in possession of illegal substances, administering illegal substances to the overdose victim or for having outstanding warrants for arrests, or because it might have had negative personal consequences for the victim

• Macro-level factors included laws regarding the possession of drugs and harm reduction discourse that also limited the uptake of overdose response techniques

• Context specific micro- and macro- environmental factors mitigate effective and immediate overdose intervention

• Prevention policies need to address the contextual factors that restrict IDU’s attempts to enact effective overdose response techniques through innovative measures that enable intervention.

 Koester, Mueller, Raville, Langegger, & Binswanger, 2017 [20]

To apply the concept of the “risk environment” (i.e., social, political, economic) to understanding responses to opioid overdose within the context of a recent Good Samaritan law by describing PWIDs’ experiences of reversing overdoses and their decision whether to call for EMS support.

Semi-structured interviews and fieldwork observations

Data combined from 2 studies:

1) semi-structured interviews with 13 persons who inject drugs (5 women, 8 men);

2) fieldwork observations and qualitative interviews with 24 individuals sampled from a syringe exchange program (19 men, 5 women) in Denver, CO

• Despite being trained in OEND, most participants stated they had not called 911 (EMS) after reversing an overdose

• Most frequent reason was fear that despite the Good Samaritan law, a police response would result in arrest of the victim and/or witness for outstanding warrants or sentence violations.

• Fears were based on individual and collective experience, and reinforced by the city’s aggressive approach to managing homelessness through increased enforcement of misdemeanors and ordinances, including a camping ban, to control space.

• Participants expressed concerns that an EMS intervention would jeopardize their public housing.

• The immunity provided by the Good Samaritan law does not address individuals’ fears that their current legal status as well as the victim’s will result in arrest and incarceration. As currently conceived, the Good Samaritan law does not provide immunity for individuals who inject drugs and are already enmeshed in the CJS, or are fearful of losing their housing.

 Petterson & Madah-Amiri, 2017 [21]

To assess knowledge of opioid OD among inmates at risk of witnessing or experiencing an OD before and after a brief training session about naloxone prior to re-entry

Pre/post training intervention survey

31 current or former opioid-using offenders within 6 months of release from incarceration in Oslo, Norway; half of the participants were receiving methadone treatment prior to prison; 100% male

• Nearly every participant reported that they previously had witnessed an overdose and almost half had experienced between 1 and 10 personally.

• Participating inmates were found to have a high baseline knowledge of risk factors, symptoms and care regarding opioid overdoses on an Opioid Overdose Knowledge Scale

• A brief naloxone training session on how to recognize and respond to an opioid overdose with naloxone and how to assemble and use the device, significantly improved knowledge regarding naloxone use, effect, administration, and aftercare procedures.

• Naloxone training in the prison setting may be an effective means of improving knowledge about opioid overdose within a vulnerable group.

 Barocas, Baker, Hull, Stokes, & Westergaard, 2015 [22]

To improve understanding of the acceptability and current uptake of naloxone-based overdose prevention training among people who inject drugs who interact with the CJS.

Survey

543 individuals who inject drugs using a free multi-site syringe exchange program between June – August, 2012 in Wisconsin; 43% had a history of incarceration

• Respondents who observed an overdose were more likely to have a history of incarceration;

• Respondents who were trained to administer naloxone were more likely to have a history of incarceration.

• No participants reported receiving this training in prison or jail but received training from syringe exchange program staff.

• When naloxone training is made available through community-based syringe exchange programs, CJS-involved clients will use the service and administer naloxone in practice.

 Bird, McAuley, Munro, Hutchinson, & Taylor, 2017 [23]

To examine changes over time in receipt of THN from prisons among participants in Scotland’s ongoing Needle Exchange Surveillance Initiative (NESI) program

Secondary analysis of interview data

A demographically representative sample of between 2000 and 3000 PWID (80% having injected within the past 6 months) from across Scotland in which interviews are conducted every 2 years

• Controlling for past-year incarceration rate and average duration of incarceration, among individuals in the NESI sample who received take-home naloxone at release from prison:

° 67% were female vs. 39% were male

° 48% were younger than 35 years vs. 37% older

• The proportion whose naloxone was most recently received from prison was about 13% irrespective of recency of injecting

• The study identified heterogeneity in provision of THN by sex, age-group, homelessness, and recency of injecting, with greater provision for people who were younger than 35 years, homeless, and had injected drugs in the past 6 months

• Study examined the interface between THN and community-based-provision of naloxone and how changes in THN may reflect greater access to community-based provision.

 McAuley, Munro, Bird, Hutchinson, Goldberg, & Taylor, 2016 [24]

To address three specific evidence gaps: (1) the extent of naloxone supply to PWID; (2) supply-source (community or prisons); and (3) the carriage of naloxone among PWID.

Survey

Participants in Scotland’s Needle Exchange Surveillance Initiative (NESI) in 2011–2012 and 2013–2014; over 90% report heroin as the drug injected most often within the past 6 months.

• The proportion of NESI participants who reported that they had been prescribed naloxone within the last year increased significantly from 8% in 2011–2012 to 32% in 2013–2014.

• In contrast, the proportion of NESI participants who carried naloxone with them on the day they were interviewed decreased significantly from 16% in 2011–2012 to 5% in 2013–2014.

• The proportion of participants reporting that their last naloxone supply was made via the prison system was stable across the two surveys: 16% in 2011–2012 to 19% in 2013–2014.

• Controlling for duration of prison sentence, both community services and prisons were equally efficient at targeting their naloxone supplies to PWID.

• Carriage was lowest among those who had not injected in the previous 6-months therefore it is possible that self-reported naloxone carriage is associated with current injecting behavior and perceived risk of experiencing an overdose.

• Individuals at risk of overdose may calculate their level of risk and decision whether to carry naloxone based on their perceptions of availability of naloxone in the community (i.e., diffusion of responsibility).

• PWIDs may also be reluctant to carry naloxone on their person because of fear of coming into contact with the police. The naloxone kit provided by the NNP is in a bulky, clinically labelled yellow box, making it less discreet and less portable. It is plausible that the physical properties of naloxone kits may influence carriage rates among PWIDs

 O’Hallaran, Cullen, Njoroge, Jessop, Smith, Hope, & Ncube, 2017 [25]

To monitor the impact of the 2015 policy change to improve naloxone availability in the U.K. using national-level data on the extent of self-reported overdose and self-reported receipt of naloxone among PWID in the United Kingdom

Secondary analyses of cross-sectional surveys

3850 PWID at sentinel sites located throughout the UK that voluntarily participated in annual surveys in 2013 and 2014

• 91% of the sample injected heroin; 15% reported overdosing during the preceding year

• There were no differences in the proportion reporting OD by age or gender, but OD was more common among those who: injected multiple drugs; recently ceased addiction treatment; injected with used needles/ syringes; ever had transactional sex; had used a sexual health clinic or emergency department, and lived in Wales or No. Ireland.

• Of those reporting an OD during the past year, two fifths reported 2 or more ODs and one half reported receiving naloxone.

• Those reporting naloxone receipt in the preceding year were more likely to: live in Wales or Northern Ireland; ever received used needles/syringes; ever been imprisoned (AOR = 1.59); and less likely to have injected two drug types.

• Interventions to prevent OD should promote naloxone awareness and access, and target those who: are poly-drug injectors, have ceased treatment, share needles/ syringes and whose drug use links to sexual activity.

• History of incarceration was associated with having received naloxone at last OD, controlling for other individual characteristics

 Bird & McAuley, 2019 [26]

To assess drug-related deaths before and after implementation of Scotland’s National Naloxone Program and scale-up of naloxone distribution over time.

Time series analysis pre/post implementa-tion of NNP; evaluation of program model

Opioid-related deaths among individuals released from prisons and hospitals from 2006 to 2016 in Scotland

• The primary outcome for Scotland’s NNP was a reduction from 10 to 7% in ORDs within 4 weeks of prison release, which is a reduction of 50%; secondarily, there was a reduction of 4% from 10 to 9% within 4 weeks of hospital release.

• In 6 years (2011–16), Scotland’s NNP supplied almost 36,000 naloxone kits to people at risk of opioid-related overdose.

• The distribution target of 8000 naloxone kits (20 times Scotland’s mean number of ORDs per annum in 2006–10) was met in 2014–16 when the primary outcome was halved.

• ORDs have increased since the NNP was introduced, with 709 ORDs in 2017, of whom 545 individuals (77%) were 35 years or older.

• The national program model has been adapted and implemented in England; Wales; Norway; British Columbia, Canada; progress has been slower in Australia and the U.S.

 Bird, McAuley, Perry, & Hunter, 2016 [27]

To assess the effectiveness of Scotland’s National Naloxone Programme (NNP) by comparison between two time periods, 2006–10 and 2011–13, corresponding to before and after NNP started in January 2011; and to assess cost-effectiveness of the program.

Time series analysis pre/post implementation of NNP

Individuals released from prison in Scotland in:

1) 2006–10: n = 1970;

Opioid-related deaths (ORD) n = 193;

2) 2011–2013:

n = 1212; ORDs n = 76

• In 2006–10, 9.8% of ORDs (193 of 1970) were in people released from prison within 4 weeks of death, whereas only 6.3% of ORDs in 2011–13 followed prison release (76 of 1212, P < 0.001), which is a difference of 3.5% (95% CI = 1.6–5.4%).

• This reduction in the proportion of prison release ORDs translates into 42 fewer prison release ORDs (95% CI = 19–65) during 2011–13, when 12,000 naloxone kits were issued at current prescription cost of £225,000.

• This is the first study to evaluate a national naloxone programme at a population level with before/after analyses by design at 3 years and 5 years.

• The study found that there was a 20–36% reduction in the proportion of ORDs that occurred in the 4 weeks following release from prison (from 9.8 to 6.3%) following establishment of Scotland’s NNP.

 Green, Ray, Bowman, McKenzie, & Rich, 2014 [28]

To describe two case studies of successful self-administration of intranasal naloxone during an opioid overdose.

Case study

Two people (one male and one female) with opiate use histories who self-administered intranasal naloxone following their release from prison in Rhode Island

• Describes two cases of individuals who had been trained in the high risk of overdose after release from incarceration and on how to use naloxone, which was instrumental in their successful self-administration of naloxone to reverse opioid overdose following their release from prison.

• Training of people at risk of overdose, including inmates about to be released and people who actively use drugs, as well as the members of their social and drug use networks, on the signs of overdose and how to respond with naloxone is possible, effective, and cost-effective.

 Huxley-Reicher, Maldjian, Winkelstein, Siegler, Panone, Tuazon, Nolan, Jordan, MacDonald, & Kunins, 2017 [29]

To determine rates of overdose witnessing and naloxone use among overdose rescue-trained visitors to the New York City jails.

Pre/post intervention survey

283 individuals visiting incarcerated persons, Rikers Island, New York

• 382 visitors were trained over 5 days in overdose rescue at the Rikers Island Visitors Center; of these, 283 returned to request a naloxone kit and were enrolled in the study; 226 completed the 6-month follow up.

• 40 participants (14% of the total enrolled n = 283) had witnessed at least one overdose during the study period; there was a total of 70 overdose events witnessed and 87% were known to have survived.

• Overall, 28 (10%) study participants reported administering naloxone at least once during the study period; in 17% of cases the recipient had been recently released from jail or prison

• Training visitors to incarcerated individuals in overdose rescue is an effective strategy to reach a population of potential overdose responders.

 Kobayashi, Green, Bowman, Ray, McKenzie, & Rich, 2017 [30]

To evaluate an experimental program that educated, trained and assessed at-risk, prisoners on opioid overdose prevention, recognition and layperson management with intranasal naloxone using simulation techniques.

Pre/post training intervention survey

Inmates who were within 4 weeks of release from the Rhode Island Department of Corrections in Cranston, RI; n = 85 completed baseline assessment, intervention, and follow-up assessment

• 38 (35.5%) and 75 (70.1%) subjects had personally experienced or witnessed an opioid OD, respectively; none had previously been trained to respond to ODs or obtained a naloxone rescue kit

• 44 participants (51.8%) correctly administered naloxone; 16 additional subjects (18.8%) sub- optimally administered naloxone.

• Non-indicated actions, e.g., chest compressions, were observed in 49.4% of simulations.

• Simulated resuscitative actions by 80 subjects (94.1%) were determined post-hoc to be beneficial overall for patients overdosing on opioids

• More than half of the study participants correctly delivered resuscitative doses of IN naloxone with timeliness comparable to paramedic students.

• Simulation can be applied to outreach efforts directed towards inmate target populations housed in intrinsically limiting environments and enable them to learn and practice the intervention for responding to opioid overdose, which is a high probability event following release

 Parmar, Strang, Choo, Meade, & Bird, 2016 [31]

To examine the feasibility of a large-scale naloxone distribution program for prisoners preparing for re-entry and to determine its impact on overdose rates

RCT

1685 heroin injecting offenders scheduled to be released from prison within 3 months and who had been in prison at least 7 days at study baseline; from 16 prisons in England; 98% male

• There was a high rate of consent among prisons and offenders to participate in the program; however, the study stopped early due to the finding that only one-third of naloxone administrations were to the former offender.

• There were 9 overdose deaths among offenders within 12 weeks of reentry.

• Naloxone access may introduce some risk compensation, but there is insufficient evidence to draw a conclusion.

 Wenger, Showalter, Lambdin, Leiva, Wheeler, Davidson, Coffin, Binswanger, & Kral, 2019 [32]

To evaluate a take-home naloxone program for individuals being released from jail

Surveys and program documenta-tion

637 participants who received naloxone kit upon release from jail in San Francisco, CA

• During 4 years of operation, 637 people participated; 67% received naloxone upon release, of whom only 3.5% had been previously trained in community-based OEND programs.

• Of those who received naloxone, 32% reported reversing an overdose and 44% received refills after reentry

• 190 (96%) of these individuals received their refill at a syringe access program or other community-based program and 8 (4%) received their refill at the jail during a subsequent incarceration.

• The most frequent reasons for getting a naloxone refill were that it was lost (33%), had been used to reverse an overdose (32%), had been stolen (13%), and had been given away to someone who needed it (12%).

• The study provides evidence that implementation of OEND in CJS is feasible and reaches people who have not previously been trained as well as those willing to act as overdose responders.

• Demonstrates successful collaboration among the jail, several county agencies, and community partners

• Participation in OEND programs helps individuals minimize drug-related harm and encourages them to take on new prosocial roles in their community as peer educators and “overdose responders,” as participants often teach others in their communities about overdose risk and response.

 Green, Bowman, Ray, McKenzie, Lord, & Rich, 2015 [33]

To create and test the acceptability of a new DVD on overdose prevention for former prisoners based on input and feedback from formerly incarcerated injection drug users, national experts, and overdose prevention staff.

1) Formative evaluation: Systematic review of educational videos and

2) Video development using community-based participatory process that included 3 focus groups, consultations with national expert groups and OD prevention

program staff, and ongoing presentations to correctional staff and leadership

Former prisoners and current or former injection drug users (n = 4), recruited at syringe exchange program in Providence, Rhode Island

• Review of nine videos; 3 contained theory-based learning components, and only one also contained peer-based content; none directly addressed post-incarceration overdose prevention.

• Created 19-min film, Staying Alive on the Outside, using Bandura’s Social Learning Theory and incorporates interviews, conversation and model training sessions by peers, who discuss the challenges of re-entry from prison, OUD and relapse, and misconceptions about opioid tolerance and OD.

• Viewers learn strategies to avoid overdose while using opioids and what to do in an overdose.

• Peer ‘learners’ and peer ‘trainers’ model the dissemination of education and naloxone administration.

• The theory-based DVD containing prison-specific OD information and informed by input from end-users has been disseminated to several prisons and jails as part of re-entry planning for soon-to-be-released inmates.

 Horsburgh & McAuley, 2018 [34]

To describe the development of the National Naloxone Programme in Scotland within the Scottish Prison Service with a focus on its delivery model, challenges, and developments.

Qualitative: Program description and observations

Prisons in Scotland

• Group training sessions were problematic:

° From an operational perspective, organising key personnel (i.e. trainers and participants) to be all in one place/time was problematic due to the prison regime;

° Competing priorities for prisoners led to high rate of refusal to participate;

° Limited time availability of staff, need to escort prisoners to groups

° From an individual perspective, group sessions were not always suitable for discussing emotionally charged issues related to overdose and loss in the prison setting; having to practice basic life support may have been intimidating for some in a group setting.

• Limited awareness among prison officers regarding naloxone led to lack of follow-through in placing it with prisoner possessions for collection at discharge

• 63 male and 6 female prisoners completed the Scottish Drugs Forum naloxone peer education training in 4 facilities

• Initiated a program for providing prison officers with naloxone training in order to intervene in an overdose emergency, rather than waiting for health professionals

• Because of operational challenges, prisons adopted the community NNP training model of brief interventions, delivered in a one-to-one format over 10–15 min and requiring only one member of staff to facilitate.

• The implementation of the NNP with the Scottish Prison Service has faced several challenges, which have been addressed through innovation and partnership across Scottish Prisons and the community-based programs, and has resulted in a “largely consistent model” across facilities.

 Pearce, Mathany, Rothon, Kuo, & Buxton, 2019 [35]

To understand how the THN program is implemented in two pilot correctional facilities in order to identify areas for program improvement and inform the expansion of the program to other correctional facilities in Canada.

Focus groups and interview

Two focus groups (n = 8) and one individual interview with healthcare staff who were involved in implementing THN programs in 2 correctional facilities in British Columbia, Canada

• Barriers to “train the trainer’ program included: lack of capacity, including time and staff resources, to conduct a thorough train-the-trainer program; competing healthcare priorities and high workloads since the immediate healthcare needs of persons in custody took priority over the THN program; rotating shifts that made scheduling sessions difficult

• Need to pay off-duty healthcare staff to participate in additional group train-the-trainer sessions was a financial and logistical burden

• Challenges of connecting participants to community harm reduction resources following release

• Need for “whole systems approach” that includes support from management and other correctional staff

• The implementation of the pilot program faced several logistical challenges but has the potential for improved prison population engagement and awareness of the program;

• Findings suggest that successful program implementation requires adapting resources to the needs of incarcerated populations and facility operations.

 Sondhi, Ryan, & Day, 2016 [36]

To assess potential barriers and challenges to the implementation of THN in prison.

Qualitative interviews;

focus groups; document review

Four focus groups with male prisoners who participated in a THN program (n = 26); interviews with strategic and operational prison staff (n = 17) sampled from 10 prisons within one region in England

• The distribution and implementation of THN in a prison setting was characterized by significant barriers and challenges; four main themes were identified:

° A wide range of negative and confused perceptions of THN among prison staff and prisoners;

° Inherent difficulties with the identification and engagement of eligible prisoners;

° The need to focus on individual prison processes to enhance the effective distribution of THN;

° The need for senior prison staff engagement to support “culture of change.”

• Prisoners highlighted side effects and the possible unintended consequences of being in the possession of a THN kit once released, including concerns about the possible criminal-justice consequence, fear of police, or criminal justice services.

• Another challenge to providing a harm reduction initiative stems from the mixed messages that emphasize both the desirability for complete abstinence at release and an acceptance of potential relapse to drug use at some point post-release from prison.

• Successful implementation of THN within prison requires a ‘whole system’ approach that addresses negative staff perceptions as well as clear processes to ensure eligible prisoners are trained and given access to the THN kits.

• It is insufficient for prisons to merely offer training and distribute kits to opiate-using prisoners without conducting a more enhanced planning and preparation process.

• Two main components must be addressed:

° Need to develop a detailed mapping of prison processes and procedures, where prison staff establish local processes to identify eligible prisoners and intervene at the most effective point in their incarceration.

° Need to incorporate a more nuanced consideration of the beliefs and perceptions of prisoners to ensure the effective distribution of THN within a prison setting, including fears about use in the community and its side effects

 Zucker, Annucci, Stancliff, & Catania, 2015 [37]

To describe a pilot program to provide training in OD prevention and naloxone to all prisoners as they re-enter the community.

Qualitative:

Program description

Minimum-security correctional facility in New York

• A pilot at a minimum-security correctional facility in New York City was initiated in February 2015. Harm Reduction Coalition staff trained inmates in the use of naloxone, as well as prison staff to provide the trainings.

• By September 2015, more than 700 inmates had been trained at one facility; about 200 had received kits. The numbers of inmates taking kits at release increased each month, suggesting growing acceptance of the program.

• Training was initiated in two other correctional facilities and were scheduled at other facilities.

• In addition, a community-based organization in the region is training family members and friends of incarcerated individuals and equipping them with naloxone free of charge.

• The state Dept. of Corrections established a statewide standing order, in conjunction with the Department of Health, which enables DOC nursing staff to administer naloxone by injection to any inmate, staff or visitor suspected of an OD without first obtaining a physician order.

• The OD prevention program was implemented through collaboration between state corrections and community providers, and has been expanded to train parole officers.

• Acceptance of the program has been augmented by the fact that many corrections staff and parole officers recognize the need for naloxone in their communities.

 Alex, Weiss, Kaba, Rosner, Lee, Lim, Venters, & MacDonald, 2017 [38]

To understand post-release death by matching electronic health records from incarcerated individuals

with vital statistics records

Secondary analysis of records data

59 individuals who were deceased within 6 weeks of release from jail between 2011 and 2012 in New York City

• Mean no. of days to death was 20; 73% died within 28 days of release; post-release all-cause mortality rate was 5.89 per 1000 PY

• Causes of death: 37% opioid overdose, 8.5% other drug overdose, 25% chronic disease, 20% assaultive trauma, 8.5% trauma related to unintentional injury, suicide, or unspecified.

• 77% of those who died from opioid overdose had a history of prior overdose or opioid detoxification

• 14% were released with methadone dose; 50% had been referred to opioid treatment within the community at release

• Patient-centered history taking is necessary as many individuals do not disclose prior drug use history

• More aggressive linkage to opioid treatment programs is needed

• Expansion of access to buprenorphine and distribution of naloxone at release from jail are needed for overdose prevention.

 Andersson, Håkansson, Krantz, & Johnson, 2020 [39]

To investigate fatal opioid intoxications in southern Sweden among people with a history of illicit drug use. The purpose of the study is: (1) to survey the deceased individuals’ contact with care-providing authorities during the year prior to death; and (2) to analyze differences in their clinical picture, relating to which opioids caused their deaths.

Secondary analysis of records data

180 opioid-related deaths where the cause of death was intoxication due to the intake of heroin, methadone, buprenorphine, or fentanyl in Sweden

• 89% of the deceased individuals had been in contact with one or more care-providing authorities in the year prior to death:

° 75% had been in contact with health care services

° 69% with the social services

° 28% with the Prison and Probation Service

° 23% had been enrolled in OST

• Sedatives were present in more than 80% of the cases.

• Individuals whose deaths were buprenorphine-related had been in contact with the social services to a significantly lesser extent during the year prior to death.

• Individuals who died from opioid-related intoxication had extensive contact with care-providing authorities, thus providing numerous opportunities to intervene with preventive and other interventions.

• Developing a broader understanding of the lives and deaths of opioid users is essential for the development and provision of effective treatment and harm reduction interventions.

 Binswanger, Nowels, Corsi, Glanz, Long, Booth, & Steiner, 2012 [40]

To understand the drug use experiences, perceptions of overdose risk, and experiences with overdose among former prisoners

Survey

29 former prison inmates recruited within 2 months of release from a community health center, an urgent care center, and addiction treatment centers, as well as by snowball sampling, in Denver, Colorado

The following themes emerged:

• Relapse to drugs and alcohol occurred in a context of poor social support, medical co-morbidity and inadequate economic resources;

• Former inmates experienced ubiquitous exposure to drugs in their living environments posing a risk of relapse;

• Intentional overdose was considered “a way out” given situational stressors, and accidental overdose was perceived as related to decreased tolerance; and

• Protective factors included structured drug treatment programs, spirituality/religion, community-based resources (including self-help groups), and family.

• Interventions to prevent overdose after release from prison may benefit from including structured treatment with gradual transition to the community, enhanced protective factors, and reductions of environmental triggers to use drugs.

 Binswanger, Stern, Yamashita, Mueller, Baggett, & Blatchford, 2016 [41]

To identify risk and protective factors for all-cause and accidental poisoning (overdose) death among individuals following release from prison

Nested case–control study of people released from prison

Cases (699 all-cause deaths, 88 were among women; and 196 additional overdose deaths, 76 were among women) between 1999 and 2009 matched 1: 1 to controls on sex, age and year of release from Washington State Department of Corrections

• Key independent risk factors for all-cause mortality included homelessness (OR = 1.53, 95% CI = 1.06, 2.23), IDU (OR = 1.54, 95% CI = 1.16, 2.06), tobacco use (OR = 1.51, 95% CI = 1.07, 2.13), cirrhosis (OR = 4.42, 95% CI = 1.63, 11.98) and psychiatric medications before release (OR = 2.38, 95% CI = 1.71, 3.30).

• Independent risk factors for OD mortality included SUD (OR = 2.33, 95% CI = 1.32, 4.11), IDU (OR = 2.43, 95% CI = 1.53, 3.86), panic disorder (OR = 3.87, 95% CI = 1.62, 9.21), psychiatric prescriptions before release (OR = 2.44, 95% CI = 1.55, 3.85), and problems with opiates/sedatives (OR = 2.81, 95% CI = 1.40, 5.63).

• SUD treatment during the index incarceration was protective for all-cause (OR = 0.67, 95% CI = 0.49, 0.91) and OD (OR = 0.57, 95% CI = 0.36, 0.90) mortality.

• Injection drug use and SUD are risk factors for death after release from prison.

• In-prison SUD treatment services may reduce the risk.

 Binswanger, Nguyen, Morenoff, Xu, & Harding, 2020 [42]

To examine the associations between characteristics of justice-involved individuals regarding use patterns, drug convictions and supervision setting, and overdose mortality.

Secondary analysis of records data

140,266 individuals with a history of criminal justice involvement and OUD from 2003 to 2006 in Michigan

• Among 140,266 individuals followed over a mean of 7.84 years (SD = 1.52), 14.9% of the 1131 deaths were due to overdose (102.8 per 100,000 person-years).

• Over the follow-up, 57.7% of overdose deaths occurred in the community, 28.8% on probation, and 12.8% on parole.

• Adjusted risk of overdose death was lower on probation (HR = 0.71, 95% CI = 0.60, 0.85) than in the community without probation or parole (HR = 1.00), but not significantly different on parole (HR = 1.13, 95% CI = 0.87, 1.47).

• Pre-sentence daily opioid use (HR = 3.54, 95% CI = 3.24, 3.87) was associated with increased risk of opioid-related overdose.

• Drug possession (HR = 1.11, 95% CI = 0.93, 1.31) and delivery convictions (HR = 0.92, 95% CI = 0.77, 1.09) were not significantly associated with overdose mortality.

• Given the absolute or relative risk of opioid-related overdose among justice-involved individuals, parole, probation and community settings are appropriate settings for enhanced overdose prevention interventions.

• Ensuring that individuals with pre-sentence OUD have access to harm reduction and drug treatment services may help to prevent overdose among people involved with the CJS.

 Bird, Fischbacher, Graham, & Fraser, 2015 [43]

To assess whether the introduction of a prison-based OST policy was associated with a reduction in drug-related deaths (DRD) within 14 days after prison release.

Time series analysis of pre/post intervention

Linkage of Scotland’s prisoner database with death registrations to compare periods before (1996–2002) and after (2003–07) prison-based OST was introduced.

• Before prison-based OST (1996–2002), 305 DRDs occurred in the 12 weeks after 80,200 qualifying releases, 3.8 per 1000 releases; of these, 175 (57%) occurred in the first 14 days.

• After the introduction of prison-based OST (2003–07), 154 DRDs occurred in the 12 weeks after 70,317 qualifying releases, a significantly reduced rate of 2.2 per 1000 releases.

• There was no change in the proportion that occurred in the first 14 days, either for all DRDs or for opioid-related DRDs.

• Following the introduction of a prison-based OST policy in Scotland, the rate of drug-related deaths in the 12 weeks following release fell by two-fifths.

• However, the proportion of deaths that occurred in the first 14 days did not change appreciably, suggesting that in-prison OST does not reduce early deaths after release.

 Bukten, Stavseth, Skurtveit, Tverdal, Strang, & Clausen, 2017 [44]

To estimate and compare overdose death rates at time intervals after prison release and to estimate the effect on overdose death rates over calendar time over a 15-year observation period.

Secondary analysis of records data from Norwegian Prison Registry and Norwegian Cause of Death Registry

All individuals released from prison in Norway between 1/1/2000 and 12/31/ 2014; the final sample comprised 91,090 former prisoners, released 150,090 times

• Overdose deaths accounted for 85% (n = 123) of all deaths during the first week following release (n = 145), with a peak in the 2 days immediately following release.

• Compared with week 1, the risk of overdose death was reduced by more than half during week 2 and to one-fifth in weeks 3–4.

• The risk of overdose mortality during the first 6 months post-release was almost doubled in 2000–04 compared with 2005–09.

• The risk of overdose death was highest for those incarcerated for 3–12 months compared with those who were incarcerated for shorter or longer periods, and recidivism was associated with risk of overdose death.

• There is an elevated risk of death from drug overdose among individuals released from Norwegian prisons, peaking in the first week, with the greatest risk for those serving 3–12 months compared with shorter or longer periods.

• Reductions in overdose mortality over time may be related to increases in participation in OAT and changes in patterns of drug consumption

• Collaboration among correctional services, drug treatment services, and social services is necessary to facilitate a safe release from prison.

• Provision of effective treatment, such as opioid maintenance treatment, as well as of naloxone along with harm reduction and social reintegration support in correctional settings is essential to reducing overdose deaths post-release among former inmates.

 Cepeda, Vetrova, Lyubimova, Levina, Heimer, & Niccolai, 2015 [45]

To understand the context of the post-release risk environment among formerly incarcerated PWID in Russia regarding how these risks relate to reentry, relapse to injection opioid use, and overdose.

Semi-structured in-depth interviews

25 PWID who had been incarcerated within the past 2 years who were recruited from street outreach (n = 20) and a drug treatment center (n = 5) in St. Petersburg, Russia

• Emergent themes related to the post-release environment included financial instability, negative interactions with police, return to a drug-using community, and reuniting with drug using peers.

• Almost half the sample had an opioid overdose after release, with the median time to overdose of 30 days after release.

• Many respondents relapsed to opioid use immediately after release; others who relapsed weeks or months after their release expressed more motivation to resist.

• Alcohol or stimulant use often preceded opioid relapse; alcohol use often preceded opioid overdose.

• Future post-release interventions in Russia should effectively link PWID to social, medical, and harm reduction services.

• Particular attention should be focused on helping former inmates find employment

• Overdose prevention training prior to leaving prison should also cover the heightened risk of concomitant alcohol use.

 Forsyth, Carroll, Lennox, Kinner, & 2017 [46]

To estimate the incidence and identify risk factors for mortality in adults released from prisons in Queensland, Australia

Prospective cohort study, linking baseline survey data with a national death register over up to 4.7 years in the community

1320 adults recruited in prisons within 6 weeks of expected release, between August 2008 and July 2010 in Queensland, Australia

• The rate of mortality in the cohort was higher than in the age- and sex-matched general population of Queensland for all causes (SMR = 4.0, 95% CI = 2.9–5.4] and drug-related causes (SMR = 32, 95% CI = 19–55).

• In a multivariable model, adjusting for age, sex and Indigenous status, factors associated with increased mortality risk included expecting to have average or better funds available on release (AHR = 2.9, 99% CI = 1.2–7.1), poor mental health (AHR = 2.6,99% CI = 1.1–6.1) and self-reported life-time history of overdose (AHR = 2.5, 99% CI = 1.04–6.2).

• The study found that people released from prison in Queensland, Australia are at increased risk of death, particularly due to drug-related causes.

• Those at greatest risk of death are characterized by poor physical and mental health and a history of risky substance use, including lifetime history of overdose.

 Hacker, Jones, Brink, Wilson, Cherna, Dalton, & Hulsey, 2018 [47]

(1) To describe the demographic characteristics and the opioid epidemic in Allegheny County;

(2) To identify possible points for intervention, recognizing that overdose decedents may have used various public human services before their death;

(3) To determine the temporal relationship between overdose mortality and incarceration or the use of mental health or SUD services;

(4) To recommend potentially beneficial interventions.

Secondary analysis of records data

Records of 1399 individuals who died of opioid overdoses from 2008 to 2014 who were matched to records of their premortem incarcerations and use of mental health and SUD services in Alleghany, PA

• Of the 1399 decedents, 957 (68.4%) had a public human service encounter before overdose death.

• Of these 957 decedents, 531 (55.5%) had ever been incarcerated in the county jail, 616 (64.4%) had ever used a mental health service, and 702 (73.4%) had ever used a substance use disorder service.

• Of 211 (22%) decedents incarcerated in the year before their overdose death, 54 (25.6%) overdosed within 30 days of their last release from jail.

• Of 510 decedents using mental health services in the year before death, 231 (45.3%) overdosed within 30 days of their last use of the services.

• Of 350 decedents using SUD services in the year before their overdose death, 134 (38.3%) overdosed within 30 days of their last use of the services.

• The large number of decedents who had encounters with either mental health or SUD services close to the time of their overdose deaths suggests that these encounters may be an important opportunity for intervention.

• Effective screening and brief intervention procedures, especially as part of mental health treatment, can identify active drug use and potential overdose risk.

• Merging data on overdose mortality with data on use of public human services can be a useful strategy to identify trends in, and factors contributing to, the opioid epidemic; to target interventions; and to stimulate collaboration among public health and community providers to address the epidemic

 Keen, Young, Borschmann, & Kinner, 2020 [48]

To determine the incidence, predictors and clinical characteristics of NFOD following release from prison.

Secondary analysis of records data

1307 adults who had participated in RCT of a case-management intervention to increase engagement with primary care and mental healthcare after release from prison in Queensland, Australia

• Approximately 8% of participants had at least one NFOD during a median of 2.9 years of follow-up

• The crude incidence rate (IR) of NFOD was 47.6 (95% CI 41.1–55.0) per 1000 person-years and was highest in the first 14 days after release from prison (IR = 296 per 1000 person-years, 95% CI 206–426).

• In multivariate analyses, NFOD after release from prison was positively associated with a recent history of SUD, dual diagnosis of mental illness and SUD, lifetime history of injecting drug use, lifetime history of NFOD, being dispensed benzodiazepines after release, a shorter index incarceration, and low perceived social support.

• 33% of those who experienced an NFOD after index release had not previously overdosed

• The risk of NFOD was lower for people with high-risk alcohol use and while incarcerated

• Individuals released from prison are at high risk of non-fatal overdose, particularly in the first 14 days after release.

• Providing coordinated transitional care between prison and the community is needed to reduce the risk of overdose.

 Kinner, Milloy, Wood, Qi, Zhang, & Kerr, 2012 [49]

To identify risk and protective factors for NFOD among a cohort of illicit drug users in Vancouver, Canada, according to recent incarceration.

Prospective cohort study

2515 community-recruited illicit drug users followed from 1996 to 2010 in Vancouver, Canada

• One third of participants (n = 829, 33.0%) reported at least one recent NFOD; those recently incarcerated were significantly more likely to report recent NFOD (OR = 2.13, 95% CI 1.89–2.40, p < 0.001).

• Among those recently incarcerated, risk factors independently and positively associated with NFOD included daily use of heroin, benzodiazepines, cocaine or methamphetamine, binge drug use, public injecting and previous NFOD.

• Older age, methadone maintenance treatment, and HIV+ status were protective against NFOD.

• There is an urgent need to develop and implement evidence-based preventive interventions for ex-prisons that target those with modifiable risk factors.

 Larochelle, Bernstein, Bernson, Land, Stopka, Rose, Bharel, Liebschutz, & Walley, 2019 [50]

To identify potential touchpoints for intervention with individuals at risk of overdose, including those within the CJS.

Secondary analysis of records data

General population of Massachusetts aged 11 and older in 2014 with non-missing data on sex and age; N = 6,717,390; analysis of individuals who died from opioid overdose from 2011 to 2015

• Past 12-month exposure to any touchpoint was identified in 2.7% of person-months and for 51.8% of opioid overdose deaths.

• Opioid overdose SMRs were 12.6 (95% CI: 11.1, 14.1) for opioid prescription and 68.4 (95% CI: 62.4, 74.5) for critical encounter touchpoints.

• SMR = 30.0 (95% CI: 24.8,35.3) for individuals released from prison or jail

• PAFs were 0.19 (95% CI: 0.17, 0.21) for opioid prescription and 0.37 (95% CI: 0.34, 0.39) for critical encounter touchpoints.

• Eight candidate touchpoints were associated with increased risk of fatal opioid overdose, and collectively identified more than half of all opioid overdose deaths.

• Medical care, public health, and CJS encounters could serve as touchpoints to identify and intervene with individuals at high-risk of opioid overdose death, although the relative risk of opioid overdose death and proportion of deaths that could be averted at such touchpoints are unknown

 Moore, Winter, Indig, Greenberg, & Kinner, 2013 [51]

To estimate the prevalence and correlates of lifetime NFOD among prisoners in from two states in Australia

Secondary analyses of cross-sectional surveys

2288 adults in prison that were included in the 2009 New South Wales (NSW) Inmate Health Survey and the Passports Study from New South Wales and Queensland, Australia

• In both NSW and Queensland, 23% of participants reported a lifetime history of NFOD and prisoners with a history of IDU use were significantly more likely to report lifetime NFOD.

• The lifetime prevalence of NFOD among prisoners with a history of IDU was significantly higher in NSW than in Queensland (44% vs. 35%; p < 0.01).

• Independent correlates of lifetime NFOD were similar across the two states and included having attempted suicide, injected heroin or other opioids.

• The risk of NFOD among prisoners with a history of injecting drug use is high.

• An understanding of the risk factors for NFOD in this population can inform targeted, evidence-based interventions to reduce this risk.

 Pizzicato, Drake, Domer-Shank, Johnson, & Viner, 2018 [52]

To determine overdose mortality rates among offenders after release from the Philadelphia jail system.

Retrospective cohort study linking incarceration data with OD fatality and death records

82,780 incarcerated individuals released from the criminal justice system between 2010 and 2016 in Philadelphia; 80.2% male

• Of the sample, 2522 (3%) died from any cause, of which 33% died from OD

• Individuals released from incarceration had higher risk of OD death compared to the non-incarcerated population (SMR: 5.29, 95% CI 4.93–5.65), and risk was greatest during the first 2 weeks following release (SMR: 36.91, 95% CI: 29.92–43.90).

• Among released individuals, black, non-Hispanic individuals (HR: 0.17, 95% CI: 0.14–0.19) and Hispanic individuals (HR: 0.41, 95% CI: 0.34–0.50) were at lower risk for OD than white, non-Hispanic individuals.

• Individuals released with a serious mental illness were at higher risk of overdose (HR: 1.54, 95% CI: 1.27–1.87) than those without.

• Previously incarcerated individuals are at high risk of OD death following release from a local criminal justice systems, especially in the earliest weeks following release.

• Prevention measures including behavioral health treatment and referral and take-home naloxone may reduce overdose mortality after release.

 Ranapurwala, Shanahan, Alexandridis, Proescholdbell, Naumann, & Edwards, 2018 [53]

To examine differences in rates of opioid overdose death (OOD) between former North Carolina (NC) inmates and NC residents and evaluate factors associated with post release OOD.

Retrospective cohort study

229,274 former prison inmates released from 2000 to 2015 in North Carolina

• Of the sample, 1329 died from opioid OD after release. At 2-weeks, 1-year, and complete follow-up after release, the respective OD risk among former inmates was 40 (95% CI = 30, 51), 11 (95% CI = 9.5, 12), and 8.3 (95% CI = 7.8, 8.7) times as high as general NC residents; the corresponding heroin overdose death risk among former inmates was 74 (95% CI = 43, 106), 18 (95% CI = 15, 21), and 14 (95% CI = 13, 16) times as high as general NC residents, respectively.

• Former inmates at greatest opioid OD risk were those within the first 2 weeks after release, aged 26 to 50 years, male, White, with more than 2 previous prison terms, and who received in-prison mental health and SUD treatment.

• Former inmates are highly vulnerable to opioid overdose fatality and need urgent prevention measures.

 Spittal, Forsyth, Borschmann, Young, & Kinner, 2019 [54]

To identify modifiable risk and protective factors for external cause and cause-specific mortality after release from prison.

Secondary analysis of data from a retrospective cohort study and records data

572 inmates released from prison between 1994 and 2007 (n = 286 cases, n = 286 matched controls) in Queensland, Australia

• Factors associated with increased risk of external cause mortality of cases vs. controls included use of heroin and other opioids in the community (OR = 2.20, 95% CI: 1.41–3.43, p < 0.001), a prescription for antidepressants during the current prison sentence (OR = 1.94, 95% CI: 1.02–3.67, p = 0.042), a history of alcohol use in the community (OR = 1.54, 95% CI: 1.05–2.26, p = 0.028), and having ever served two or more custodial sentences (OR = 1.51, 95% CI: 1.01–2.25, p = 0.045).

• Being married (OR = 0.45, 95% CI: 0.29–0.70, p < 0.001) was protective.

• Fewer predictors were associated with cause-specific mortality.

• The study identified several behavioral, psychosocial, and clinical markers associated with mortality from preventable causes (i.e., drug overdose, suicide, accidents, violence) in people released from prison.

• Interventions that could be targeted at those at increased risk of external cause mortality include SUD treatment and harm reduction programs, improving transitional support programs and continuity of care for mental health, diversion and drug reform for repeat incarceration, and nurturing stable relationships during incarceration.

• The period of imprisonment and shortly after release provides a unique opportunity to improve the long-term health of ex-prisoners.

 Wagner, Liu, Davidson, Cuevas-Mota, Armenta, & Garfein, 2015 [55]

To identify venues where high-risk PWID could be targeted by OEND interventions.

Secondary analyses of baseline data from a cohort study

573 PWIDs sampled from community sites in San Diego, CA

• 41.5% reported past heroin/ opioid overdose, and 7.9% had at least one heroin/opioid overdose in the past 6 months

• A higher proportion of participants with past 6-month overdose had been arrested for any reason (43.2% vs. 25.7%), had been arrested for drug possession (27.3% vs. 7.3%), and had their syringes confiscated by police (16.3% vs. 8.5%)

• Individuals who had been arrested for drug possession in the past 6 months had 4 times the odds of reporting a recent heroin/opioid overdose.

• Identifying venues outside of those that traditionally target services to PWIDs (i.e., syringe exchange programs) is critical to implementing OEND interventions at a scale sufficient to address the growing epidemic of heroin/opioid-related deaths

 Winter, Stoové, Degenhardt, Hellard, Spelman, Jenkinson, McCarthy, & Kinner, 2015 [56]

This study aimed: (1) to estimate the incidence of self-reported NFOD at three discrete time periods following release from prison, among all released prisoners and among PWID, and (2) to identify the pre-release predictors of non-fatal overdose among PWID

Longitudinal cohort study with structured interviews at 1, 3, 6 months post-release from prison

1051 prisoners from selected prisons from August, 2008 to July, 2010 who: (1) expected release within 6 weeks, (2) were sentenced, and (3) imprisoned for at least 4 weeks. Participants were generally representative of all persons released from prison in Queensland, Australia during the recruitment period; women were oversampled

• The incidence of reported overdose was highest between 1 and 3 months post-release: 37.8 per 100 person-years (PY) among PWID; 24.5/100 PY among all ex-prisoners.

• In adjusted analyses, the risk of post-release NFOD was higher for PWID who reported:

° being unemployed for > 6 months before prison

° having been removed from family as a child

° using benzodiazepines and/or pharmaceutical opiates at least weekly in the 3 months prior to prison

° ever receiving OST

° having pre-release psychological distress or a lifetime mental disorder

• Risky alcohol use in the year before prison was protective.

• Imprisonment is an opportunity to initiate targeted preventive interventions such as OST, overdose prevention training and peer-delivered naloxone for those with a high risk of overdose.