Table 3. Description of intervention and related outcomes of opioid use interventions during incarceration, 2008–2019.
| Source | Intervention | Sample size | Outcomes/conclusions |
|---|---|---|---|
| Non-specified opioid agonist treatment | |||
| Green et al., 2018 [26] | Cohort: OAT in prison | N = 35 | Overdose deaths in 2016: 179, with 26 recently incarcerated (14.5%) |
| Overdose deaths in 2017: 157, with 9 recently incarcerated (5.7%) | |||
| Reduction in overdose-related mortality after the release of new model of screening and OAT treatment in 2016: 60.5% (risk ratio = 0.4, 95% CI 18.4–80.9, p = 0.01) | |||
| Marsden et al., 2017 [68] | Cohort: OAT vs. no OAT in prison | N = 12,260 | All-cause mortality lower among OAT-exposed vs. unexposed group 4 weeks post-release: 0.93 per 100 PY vs. 3.67 per 100 PY (AHR = 0.25, 95% CI 0.09–0.64) |
| Drug-related poisoning deaths lower among OAT-exposed vs. unexposed group 4 weeks post-release: 0.47 per 100 PY vs. 3.06 per 100 PY (AHR = 0.15, 95% CI 0.04–0.52) | |||
| No group difference in mortality risk following the first month | |||
| OAT-exposed group more likely to enter addiction treatment during first month post-release (OR = 2.47, 95% CI 2.31–2.65) | |||
| Prison-based OAT associated with 75% reduction in all-cause mortality and 85% reduction in fatal drug-related poisoning during first month post-release | |||
| Methadone maintenance treatment | |||
| Rich et al., 2015 [60] | RCT (2 groups): MMT vs. forced tapered withdrawal during incarceration | N = 283 | MMT access (within 1 month post-release) higher among in-prison MMT vs. forced withdrawal group (HR = 2.04, 95% CI 1.48–2.80, p < 0.001) |
| MMT initiation (within 1 month post-release) higher among in-prison MMT vs. forced withdrawal group (HR = 6.61, 95% CI 4.00–10.91, p < 0.001) | |||
| Brinkley-Rubinstein et al., 2018 [61] | RCT (2 groups): MMT vs. forced tapered withdrawal during incarceration | N = 179 | Impact at 12 months post-release of tapered withdrawal from methadone vs. MMT in prison |
| Heroin use during prior 30 days: 39.2% vs. 24.2% (OR = 2.02, 95% CI 1.01–4.04, p < 0.05) | |||
| Injection drug use during prior 30 days: 39.2% vs. 18.0% (OR = 2.95, 95% CI 1.43–6.06, p < 0.05) | |||
| Nonfatal overdose during prior 30 days: 17.7% vs. 7.0% (OR = 2.83, 95% CI 1.05–7.61, p < 0.05) | |||
| Continuous engagement in MMT: 26.0% vs. 45.2% (OR = 0.43, 95% CI 0.21–0.88, p < 0.05) | |||
| Moore et al., 2018 [62] | Clinical trial: MMT vs. forced tapered withdrawal during incarceration | N = 382 | MMT group (during incarceration) more likely to start community-based MMT within 1 day post-release (OR = 32.04, 95% CI 7.55–136.01, p < 0.001) |
| MMT group more likely to start community-based MMT within 30 days post-release (OR = 6.08, 95% CI 3.43–10.79, p < 0.001) | |||
| McMillan et al., 2008 [63] | Cohort with jail-based MMT | N = 589 | No statistically significant effect of jail-based MMT on re-incarceration (HR = 1.16, 95% CI 0.81–1.68) |
| No statistically significant effect of MMT dosage received upon release on re-incarceration rate (HR = 1.05 per additional 10 mg, 95% CI 0.99–1.12). | |||
| Data do not support the hypothesis that jail-based MMT increases or reduces re-incarceration | |||
| Marzo et al., 2009 [64] | Cohort: OAT (MMT or BPN) in prison | N = 507 (MMT, 104; BPN, 290; no OST, 113) | OAT delivered to 77.7% of opioid-dependent patients during imprisonment |
| After adjustment for confounders, MMT not associated with a reduced rate of re-incarceration (aRR = 1.28, 95% CI 0.89–1.85, p = 0.19) | |||
| Wickersham et al., 2013 [65] | Cohort: methadone dose at release <80 mg/daily vs. ≥80 mg/daily | N = 27 | Methadone dose of ≥80 mg/daily associated with treatment retention |
| At 12 months post-release, 21.4% of participants on <80 mg/daily were retained vs. 61.5% of those on ≥80 mg/daily (p < 0.01) | |||
| Wickersham et al., 2013 [66] | Cohort: methadone dose at release <80 mg/daily vs. ≥80 mg/daily | N = 72 | Methadone dose of ≥80 mg/daily associated with treatment retention |
| At 12 months post-release, 29.0% of participants on ≤80 mg/daily were retained vs. 73.1% of those on >80 mg/daily (p < 0.001) | |||
| Westerberg et al., 2016 [67] | Cohort: MMT in jail vs. opioid detox | N = 960 | Participants who received MMT in prison less likely to be re-incarcerated after 1 year vs. opioid detox group (53.4% vs. 72.2%, p < 0.001) |
| Among those re-incarcerated within 1 year, the number of days to rebooking was longer for participants who received MMT in prison vs. opioid detox group (275.6 days [SD 124.9] vs. 236.3 days [SD 131.2], p = 0.035) | |||
| 97.8% of participants who received MMT in prison continued MMT post-release | |||
| Buprenorphine maintenance treatment | |||
| Sheard et al., 2009 [56] | RCT (2 groups): BPN vs. DHC in prison | N = 90 (BPN, 42; DHC, 48) | At 5 days post-detox: BPN group more likely to have a negative opioid test (RR = 1.61, 95% CI 1.02–2.56, p = 0.04) |
| At 1, 3, and 6 months: no statistically significant differences found between the groups | |||
| Magura et al., 2009 [57] | RCT (2 groups): BPN vs. MMT in jail | N = 116 (BPN, 60; MMT, 56) | While incarcerated, BPN group more likely to report to assigned addiction treatment post-release vs. MMT (48% vs. 14%, p < 0.001) |
| While incarcerated, BPN group more likely to report intention to continue addiction treatment post-release vs. MMT (93% vs. 44%, p < 0.001) | |||
| BPN group less likely to withdraw voluntarily from medication while in jail vs. MMT (3% vs. 16%, p < 0.05) | |||
| Awgu et al., 2010 [58] | RCT (2 groups): BPN vs. MMT in jail | N = 114 (BPN, 60; MMT, 54) | MMT patients reported more side effects than BPN patients, including depression (p < 0.01), constipation (p < 0.01), confusion (p < 0.01), and fatigue/weakness (p < 0.05) |
| MMT patients reported more opioid withdrawal symptoms (85% vs. 53%, p < 0.001) | |||
| Wright et al., 2011 [59] | RCT: MMT vs. BPN | N = 289 (MMT, 148; BPN, 141) | MMT vs. BPN had equal clinical effectiveness in achieving abstinence at follow-up |
| Predictors of abstinence: continued incarceration and abstinence during previously measured period | |||
| Impact of incarceration vs. release on abstinence | |||
| At 8 days post-detoxification, participants still in prison were more likely to be abstinent than those released to the community: OR = 15.2 (95% CI 4.2–55.3, p < 0.001) | |||
| At 1 month post-detoxification, participants still in prison were more likely to be abstinent than those released to the community: OR = 7.0 (95% CI 2.2–22.2, p = 0.001) | |||
| Impact of abstinence on previous measurement of current abstinence | |||
| At 1 month of follow-up, abstinence at 8 days of follow-up vs. not: OR = 4.5 (95% CI 1.9–10.3, p < 0.001) | |||
| At 3 of months follow-up, abstinence at 1 month of follow-up vs. not: OR = 8.6 (95% CI 3.2–23.3, p < 0.001) | |||
| At 6 months of follow-up, abstinence at 3 months of follow-up vs. not: OR = 32.8 (95% CI 6.1–176.6, p < 0.001) | |||
AHR, adjusted hazard ratio; aRR, adjusted relative risk; BPN, buprenorphine; DHC, dihydrocodeine; HR, hazard ratio; MMT, methadone maintenance treatment; OR, odds ratio; OST, opioid substitution treatment; OAT, opioid agonist treatment; PY, person-years; RR, relative risk.