Table 2. Description of interventions and outcomes of opioid use intervention studies during a continuum of treatment (before, during, and after incarceration), 2008–2019.
| Source | Intervention | Sample size | Outcomes/conclusions |
|---|---|---|---|
| Non-specified opioid agonist treatment | |||
| Degenhardt et al., 2014 [35] | OAT in prison vs. post-release | N = 16,453 | 76.5% of individuals received OAT while incarcerated |
| Crude mortality rates at 4 weeks post-release | |||
| Among those retained in OAT post-release: 8.8 per 1,000 PY (95% CI 5.0–14.3) | |||
| Among those not in OAT post-release: 36.7 per 1,000 PY (95% CI 28.8–45.9) | |||
| OAT exposure by 4 weeks post-release reduced hazard of death by 75% (AHR = 0.25, 95% CI 0.12–0.53) | |||
| OAT receipt in prison had a short-term protective effect that decayed quickly across time | |||
| Lowest post-release mortality observed among those continuously retained in OAT post-release | |||
| Larney et al., 2014 [36] | OAT in prison | N = 16,715 | 76.9% received OAT while incarcerated |
| Mortality of opioid-dependent incarcerated persons was significantly lower among those receiving OAT in prison | |||
| Hazard of all-cause death was 74% lower among those receiving OAT in prison vs. those opioid-dependent not in OAT (AHR = 0.26, 95% CI 0.13–0.50) | |||
| Hazard of unnatural death was 87% lower among those receiving OAT in prison vs. those opioid-dependent not in OAT (AHR = 0.13, 95% CI 0.05–0.35) | |||
| Hazard of all-cause death during first 4 weeks of incarceration was 94% lower among those receiving OAT in prison vs. those opioid-dependent not in OAT (AHR = 0.06, 95% CI 0.01–0.48) | |||
| Hazard of unnatural death during first 4 weeks of incarceration was 93% lower among those receiving OAT in prison vs. those opioid-dependent not in OAT (AHR = 0.07, 95% CI 0.01–0.53) | |||
| Larney et al., 2012 [51] | OAT in prison vs. post-release | N = 375 (OAT = 331) | 80% of participants started OAT in prison, with median treatment duration of 5.5 months |
| Median duration of post-release OAT was 63 days | |||
| Participants retained in OAT post-release had lower risk of incarceration (HR = 0.80, 95% CI 0.71–0.90, p < 0.001) | |||
| Larney et al., 2016 [53] | OAT in prison | N = 8,577 | 82% retention in OAT treatment until release |
| 90% of participants received OAT prescription prior to release | |||
| 94% of participants with a prescription presented to a community clinic within 48 hours of release | |||
| Bird et al., 2015 [54] | Before vs. after OAT in prison | N = 131,472 | Before prison-based OAT (1996–2002) |
| 305 DRDs within 12 weeks post-release, with 175 deaths (57%) within 14 days post-release | |||
| 3.8 deaths per 1,000 releases (95% CI 3.4–4.2) | |||
| After prison-based OAT (2003–2007) | |||
| 154 DRDs within 12 weeks post-release, with 56% of deaths within 14 days post-release | |||
| 2.2 deaths per 1,000 releases (95% CI 1.8–2.5) | |||
| When DRD in each period was compared, a significant decrease was identified: 1.6 DRD per 1,000 releases (95% CI 1.0–2.2; p < 0.001) | |||
| Methadone maintenance treatment | |||
| Gordon et al., 2012 [38] | RCT (3 groups): Co vs. C+T vs. C+M | N = 211 (Co, 70; C+T, 70; C+M, 71) | Impact of C+M |
| More likely to enter prison treatment for opioid use vs. Co (OR = 10.6, 95% CI 2.6–42.8; p < 0.001) | |||
| More likely to complete prison treatment for opioid use vs. Co (OR = 3.5, 95% CI 1.5–8.3; p < 0.01) | |||
| Impact of C+T | |||
| More likely to complete prison treatment for opioid use vs. Co (OR = 3.6, 95% CI 1.5–8.4; p < 0.01) | |||
| Gordon et al., 2008 [39] | RCT (3 groups): Co vs. C+T vs. C+M | N = 201 (Co, 63; C+T, 68; C+M, 70) | Impact of C+M (6 months post-release) |
| More likely to start addiction treatment in prison vs. Co (p = 0.001) and C+T (p = 0.046) | |||
| Remained more days in addiction treatment vs. Co and C+T (p < 0.001; SE: 0.31) | |||
| Reported fewer days of heroin use vs. Co (p = 0.009; SE: 0.24) | |||
| Reported fewer days of criminal activity vs. Co (p = 0.025; SE: 0.29) | |||
| Co more likely to test positive for opioid vs. C+M (OR = 4.68, 95% CI 1.77–12.43, p = 0.002) | |||
| Kinlock et al., 2008 [43] | RCT (3 groups): Co vs. C+T vs. C+M | N = 197 (Co, 63; C+T, 66; C+M, 68) | Impact of C+M (90 days post-release) |
| More likely to enter community treatment for opioid use vs. Co (OR = 61.7, 95% CI 16.0–237.7, p < 0.001) | |||
| Less likely to be re-incarcerated vs. Co (OR = 0.36, 95% CI 0.14–0.91, p < 0.05) | |||
| Less likely to report heroin use vs. Co (OR = 0.36, 95% CI 0.16–0.81, p < 0.05) | |||
| Less likely to engage in criminal activity vs. Co (OR = 0.34, 95% CI 016–0.73, p < 0.01) | |||
| Kinlock et al., 2009 [44] | RCT (3 groups): Co vs. C+T vs. C+M | N = 204 (Co, 64; C+T, 69; C+M, 71) | Impact of C+M (12 months post-release) |
| Remained more days in addiction treatment vs. Co and C+T (p < 0.001) | |||
| Co more likely to have a positive opioid test vs. C+M (OR = 7.1, 95% CI 1.4–11.3, p < 0.001) | |||
| Less likely to have a positive cocaine test vs. Co (p < 0.001) and C+T (p < 0.05) | |||
| Kinlock et al., 2013 [45] | RCT (sub-sample) | N = 67 | Impact of C+M |
| 74.6% of participants completed in-prison MMT treatment | |||
| Employment 3 years prior to incarceration predicted completing 1 year of community treatment post-release (p = 0.001) | |||
| Participants who completed 1 year of community-based MMT reported working over twice as many days as other participants (p = 0.003) | |||
| McKenzie et al., 2012 [46] | RCT: MMT prior to release vs. MMT referral post-release | N = 62 (arm 1, 21; arm 2, 32; arm 3, 9) | Arm 1: MMT pre- and post-release + payment of treatment costs (12 weeks) |
| Arm 2: referral to MMT upon release + payment of treatment costs (12 weeks) | |||
| Arm 3: referral to MMT upon release (no financial assistance) | |||
| Arm 1 more likely to enter MMT within 30 days post-release (86%, 41%, and 22% for arm 1, 2, and 3, respectively, p < 0.001) | |||
| Arm 1 entered MMT post-release in fewer days (2, 9, and 5 days for arm 1, 2, and 3, respectively, p = 0.03) | |||
| In the last 30 days, arm 1 reported less heroin use (3, 18, and 4 days for arm 1, 2, and 3, respectively, p = 0.008), less use of other opiates (0, 6, and 1 day, p = 0.09), less crack/cocaine use (4, 13, and 6 days, p = 0.05), and less injection drug use (2, 12, 3 days, p = 0.06) | |||
| Farrell-MacDonald et al., 2014 [52] | MMT post-release: continued MMT (MMT-C) vs. discontinued MMT (MMT-T) vs. no MMT (MMT-N) | N = 137 | MMT-C group had lower risk of return to custody vs. MMT-N (HR = 0.35, 95% CI 0.13–0.90, p < 0.05) |
| Buprenorphine maintenance treatment | |||
| Gordon et al., 2014 [37] | RCT (4 groups): B+OTP, B+CHC, C+OTP, C+CHC | N = 211 (B+OTP, 52; B+CHC, 52; C+OTP, 54; C+CHC, 53) | BPN group (B+OTP and B+CHC) more likely to enter prison treatment vs. counseling only (C+OTP and C+CHC) (AOR = 2.8, 95% CI 0.3–5.7, p = 0.006) |
| BPN group more likely to enter community treatment vs. counseling only (AOR = 1.5, 95% CI 1.1–2.1, p = 0.01) | |||
| Gordon et al., 2017 [40] | RCT (4 groups): B+OTP, B+CHC, C+OTP, C+CHC | N = 211 | Higher post-release addiction treatment retention rates among participants who initiated BPN in prison (65.9 days, SE = 12.2) vs. initiation post-release (21.8 days, SE = 7.6, p = 0.005) |
| Gordon et al., 2018 [42] | RCT (4 groups): B+OTP, B+CHC, C+OTP, C+CHC | N = 199 | No statistically significant differences in BPN treatment initiation pre- vs. post-incarceration on the following variables: proportion of individuals arrested, mean number of arrests, time to first arrest |
| Zaller et al., 2013 [47] | Clinical trial: BPN/NLX (post-release vs. in jail) | N = 44 (post-release, 32; in jail, 12) | Time until post-release appointment for addiction treatment was lower among in-jail (3.9 days) vs. post-release group (8.8 days; p = 0.1) |
| Post-release treatment duration was higher for in-jail vs. post-release group: 24 vs. 9 weeks (p = 0.007) | |||
| After 6 months, retention was higher for in-jail (83%) vs. post-release group (34%; p = 0.005) | |||
| Initiating BPN/NLX prior to release from incarceration increased engagement and retention in community-based treatment | |||
| Past 30 days heroin use was higher among those receiving BPN/NLX post-release (34%) vs. in jail (0%; p = 0.08) | |||
| Past 30 days alcohol use was higher among those receiving BPN/NLX post-release (47%) vs. in jail (10%; p = 0.14) | |||
| Past 30 days injection drug use was higher among those receiving BPN/NLX post-release (39%) vs. in jail (0%; p = 0.05) | |||
| Past 30 days arrest was higher among those receiving BPN/NLX post-release (24%) vs. in jail (0%; p = 0.08) | |||
| Injectable extended-release naltrexone | |||
| Gordon et al., 2015 [41] | Clinical trial: XR-NTX | N = 27 | 37% of participants completed all 6 monthly post-release XR-NTX injections |
| Among participants who completed all 6 injections (n = 10), none reported opioid use, re-arrest, or re-incarceration during the study. | |||
| Among participants who did not complete all 6 injections (n = 16), 62.5% reported opioid use during the study (10/16), 31.3% reported re-arrest (5/16), and 18.8% reported re-incarceration (3/16) | |||
| Results were not statistically significant. | |||
| Lee et al., 2016 [48] | RCT (2 groups): XR-NTX vs. OAT (MMT/BPN) 1 week prior to release | N = 308 | During the 24-week treatment phase, participants assigned to XR-NTX (compared to those assigned to usual treatment) |
| Longer median time to opioid relapse (10.5 vs. 5.0 weeks, p < 0.001; HR = 0.49, 95% CI 0.36–0.68) | |||
| Lower rate of opioid relapse (43% vs. 64% of participants, p < 0.001; OR = 0.43, 95% CI 0.28–0.65) | |||
| Higher rate of opioid-negative urine samples (74% vs. 56%, p < 0.001; OR = 2.30, 95% CI 1.48–3.54) | |||
| Friedmann et al., 2018 [49] | RCT (2 groups): pre-release vs. post-release XR-NTX | N = 15 | Pre-release XR-NTX group had better treatment retention rate vs. post-release XR-NTX group |
| 100% of participants received the first XR-NTX injection vs. 67% of post-release group | |||
| 78% of participants received more than 1 injection vs. 17% of post-release group | |||
| 22% of participants received all 6 injections vs. no participants in the post-release group | |||
| Pre-release XR-NTX group had greater abstinence vs. post-release XR-NTX group | |||
| Confirmed abstinence 4 weeks post-release (OR = 5.6, 95% CI 0.8–37.9, p = 0.08) | |||
| Pre-release XR-NTX group had increased time to relapse vs. post-release XR-NTX group | |||
| 9 weeks vs. 5 weeks | |||
| Soares et al., 2018 [50] | RCT (2 groups): XR-NTX vs. OAT (MMT/BPN) 1 week prior to release | N = 297 | No significant difference in overall healthcare utilization (IRR = 0.88, 95% CI 0.63–1.23, p = 0.45) |
| XR-NTX group had fewer medical/surgical hospital admissions during the treatment phase (IRR = 0.37, 95% CI 0.16–0.88, p = 0.02) and throughout the course of the study (IRR = 0.55, 95% CI 0.30–1.00, p = 0.05) | |||
| Lincoln et al., 2018 [55] | Cohort: pre-release vs. post-release XR-NTX | N = 67 | Receiving XR-NTX prior to release from jail increased treatment retention vs. post-release XR-NTX |
| 4 weeks post-release: 55% vs. 25% | |||
| 8 weeks post-release: 36% vs. 25% | |||
| 24 weeks post-release: 21% vs. 15% | |||
AHR, adjusted hazard ratio; AOR, adjusted odds ratio; B+OTP, BPN in prison and continued at an opioid treatment program; B+CHC, BPN in prison and continued at a community health center; BPN, buprenorphine; C+M, counseling and MMT in prison + referral to MMT upon release; C+OTP, counseling in prison and initiation of BPN at an opioid treatment program; C+CHC, counseling in prison and initiation of BPN a community health center; C+T, counseling in prison + referral to MMT upon release; Co, counseling in prison; DRD, drug-related death; HR, hazard ratio; IRR, incidence rate ratio; MMT, methadone maintenance treatment; NLX, naloxone; OR, odds ratio; OAT, opioid agonist treatment; PY, person-years; RCT, randomized control trial; XR-NTX, injectable extended-release naltrexone.