Table 3. Characteristics of included studies.

Review	Quality (AMSTAR II)	Objective	Included Studies	Population	Outcomes
					Mortality	Morbidity	Substance use	Mental Health	Access to care	Retention in treatment
Supervised consumption facilities
Kennedy 2017 [65]	CRITICALLY LOW	To systematically review the literature on public health and public order outcomes of supervised consumption facilities	n = 47	People who use or inject drugs and the broader community in which SCFs were located	The majority of studies reporting on this outcome suggested a protective effect of SCF on overdose deaths	There was a decline in opioid poisoning emergency department presentations as well as ambulance attendances. One study found that SCF clients were more likely to experience an overdose within the SCF	The majority of studies reporting on this outcome showed that SCF participants were less likely to report syringe sharing behaviours as well as unsafe injection behaviours.	N/A	One study found that being advised to seek treatment by SCF staff was associated with higher rate of receiving treatment. Referral by SCF staff was associated with higher access to EDs and shorter durations of hospitalization	Four studies reported better use and uptake of addiction treatment among SCF users.
McNeil & Small 2014 [67]	NOT APPLICABLE	To develop a comprehensive understanding of safer environment interventions (supervised injection facilities, syringe exchange programs and peer interventions) informed by the experiences of people who inject drugs.	N = 21	People who inject drugs	N/A	SIFs produced social and physical settings that enable safer injection practices and reductions in risk behaviour.	N/A	N/A	Mediates access to ancillary services (e.g. food and shelter) and fostered access to broader health supports. These interventions were participants’ primary source of medical care.	N/A
Potier 2014 [66]	CRITICALLY LOW	To systematically collect and synthesize available evidence on supervised injection services benefits and harms	n = 75	Supervised injection services users	The majority of studies reported lower overdose-induced mortalities	Two studies reported lower number of calls for ambulances related to overdose among SIS users	The majority of studies showed that SIS was associated with lower rate of syringe sharing and reuse as well as public-space injection	N/A	All studies on this outcome reported a global increase in referral to addiction treatment centres and detoxification programs	N/A
Managed alcohol programs
Ezzard 2015 [69]	CRITICALLY LOW	To address the gap of evidence on the feasibility and acceptability of MAP programs in Sydney	n = 14	Alcohol dependent homeless individuals	N/A	N/A	Findings show that MAPs are significantly associated with reduced alcohol consumption	MAP programs were associated with better mental health stabilization	Findings show that MAPs are significantly associated with a reduction in emergency service contacts and hospital admissions	N/A
Muckle 2012 [68]	HIGH	To assess the effectiveness of managed alcohol regimens on their own or as compared to moderate drinking, screening and brief intervention using a harm reduction approach and traditional abstinence‐based interventions and no intervention	n = 0 (empty)	Vulnerable people aged 18 years or older who were at high risk for alcohol abuse, including those who are homeless, impoverished, mentally ill and past exposure to trauma	N/A	N/A	N/A	N/A	N/A	N/A
Nielsen 2018 [70]	CRITICALLY LOW	To review the current research evidence on harm reduction approaches and interventions for severe and chronic alcohol use in housing support initiatives.	n = 44	Individuals who are experiencing homelessness	In one trial, 8 men passed away while living in the MAP. Authors noted that clients were seriously ill prior to entering and their health deteriorated regardless of care. In another trial, 28 consecutive patients admitted to the program passed away in a two-year period. The most common diagnosis at admission were cirrhosis, malignancy, and HIV with an average time of 4 months from admission to death.	One trial found that medication compliance improved significantly among MAP clients with 88% taking their medication as prescribed 80% of the time. Two trials reported that many of the MAP clients saw improvements in their liver function tests since starting the program. One trial found that there was an increase in the number of MAP clients who met the criteria for alcohol-related liver damage.	Managed alcohol programs showed potential to increase alcohol consumption among clients. Two trials reported that clients who participated in MAP programs drank NBA alcohol on significantly less days and reported a significant reduction in withdrawal seizures than the control group. In one trial, clients reported a significant decrease in alcohol consumption, alcohol-related harms, as well as a significant improvement in mental health-related quality of life.	N/A	N/A	N/A
Pharmacological interventions for opioid use disorder
Bahji 2018 [41]	CRITICALLY LOW	To identify pharmacological interventions for the prevention and treatment of opioid overdose	n = 8	Patients with an established opioid use disorder	In one trial, exposure to any opioid agonist treatment for more than 7 days significantly reduced the risk of mortality	N/A	Extended-release naltrexone was significantly associated with a higher percentage of opioid-negative urine drug test as well as a significant reduction in illicit opioid use. One trial found a positive effect of extended-release naltrexone on overdose events. There was a significant difference favouring the diacetylmorphine group on overdose events compared to the methadone or hydromorphone group Methadone was found to have greater positive effect on illicit drug use compared to forced withdrawal. However, there was no significant difference between groups on overdose events There was no significant difference between the buprenorphine and methadone groups on overdose events. There was no significant difference between the LAAM and methadone groups on overdose outcomes. However, the LAAM group reported significantly lower rates of illicit opioid use compared to the methadone group	N/A	N/A	Extended-release naltrexone was significantly associated with greater retention rates in treatment. Compared to hydromorphone or methadone, diacetylmorphine showed a significant improvement in retention in treatment as well as a significant reduction in illicit opioid use. Methadone was found to have greater positive effect on retention in treatment compared to forced withdrawal
Clark 2002 [42]	LOW	To compare the efficacy and acceptability of LAAM maintenance with methadone maintenance in the treatment of heroin dependence	n = 18	Heroin dependents or patients in opioid replacement therapy for heroin dependence	No significant between group differences were found on mortality of all causes	N/A	Continuous abstinence of at least 4 weeks was significantly lower among LAAM patients compared to methadone patients. No significant difference was found between groups on the cessation of all opioid substitution therapy.	N/A	N/A	Patients were more likely to have stopped using LAAM than methadone by the end of the study. Almost twice the number of LAAM patients dropped out due to side effects compared to methadone patients.
Ferri 2006 [43]	CRITICALLY LOW	To assess heroin prescription effectiveness	n = 4	Chronic heroin-dependent individuals	Two studies reported fatalities with no difference between groups and no relation to treatment	N/A	One study showed that heroin helped people avoid illicit opiates. One study did not find a significant difference. Two studies found no difference between groups on using other substances	N/A	N/A	Two studies found that heroin alone and heroin + methadone are better than methadone alone in retaining patients in treatment. In one study, patients in the heroin arm remained in treatment longer than the methadone arm.
Gowing 2011 [44]	HIGH	To assess the effect of oral substitution treatment on risk behaviours and rates of HIV	n = 38	Opioid dependent drug users	N/A	Six studies reported decreased HIV risk and positive HIV seroconversion after entry to substitution treatment	All studies showed that substitution treatments were beneficial in significantly decreasing the proportion of participants using illicit opioids, the frequency of injecting, and sharing injecting equipment.	N/A	N/A	N/A
Jones 2012 [45]	CRITICALLY LOW	To review literature regarding outcomes following maternal treatment with buprenorphine	n = 44 (Total)	Buprenorphine maintained pregnant women and their offspring exposed in utero to buprenorphine	N/A	N/A	No significant differences were found between buprenorphine and methadone on any drug use measures at time of delivery, whereas findings were inconsistent during pregnancy	N/A	N/A	No significant differences were found between buprenorphine and methadone on treatment completion.
Karki 2016 [46]	CRITICALLY LOW	To explore relevant literature regarding the possible impact on methadone maintenance treatment on HIV risk behaviours	n = 12	HIV high risk injection drug users	N/A	All studies reported a significant association between MMT and reduced sex and drug-related HIV risk behaviours	The majority of studies reported that MMT significantly reduced the likelihood of frequent heroin injection, syringe borrowing, non-fatal overdose and pubic injection	N/A	N/A	N/A
Kirchmayer 2002 [47]	CRITICALLY LOW	To evaluate the efficacy of naltrexone maintenance treatment in preventing relapse	n = 11	Heroin dependent in- and out-patients, or former heroin addicts dependent on methadone	N/A	N/A	In most cases, no significant difference between naltrexone versus placebo or other alternative treatment was found on the use of opioid under treatment	N/A	N/A	In most cases, no significant difference between naltrexone versus placebo or other alternative treatment was found on successful completion of treatment
Klimas 2019 [63]	LOW	To assess the efficacy of slow release oral morphine (SROM) as a treatment for opioid use disorder (OUD).	n = 4	Persons with opioid use disorder as defined in the DSM-IV	N/A	No difference in incidence of adverse events (81% SROM vs 79% methadone)	No difference between SROM and methadone in reducing opioid use (RR = 0.96; 95% CI: 0.61 to 1.52, p = 0.86, I2 = 50%)	In one study SROM was associated with fewer adverse mental symptoms	N/A	Difference in dropouts was not statistically significant between participants in the SROM vs methadone (RR = 0.98; 95% CI: 0.94 to 1.02, p = 0.34)
Larney 2014 [48]	MODERATE	To assess the efficacy and adverse events of naltrexone implants when used to treat opioid dependence	n = 9	Opioid dependents	No significant differences in mortality rates between naltrexone implants and TAU or buprenorphine maintenance treatment. Another trial reported no evidence of increased risk of death due to overdose after naltrexone treatment	One trial reported no opioid-related overdose requiring ED for patients with naltrexone implant or oral naltrexone. Another trial reported no significant differences in number of self-reported overdoses	Naltrexone implants significantly decreased opioid use compared to placebo or oral naltrexone. However, patients with naltrexone implants reported significantly higher use of non-opioid drugs than those on oral naltrexone	N/A	N/A	Two studies showed that naltrexone implants significantly increased treatment retention than placebo implants and oral naltrexone.
Lobmaier 2008 [49]	LOW	To evaluate the effectiveness of sustained-release naltrexone and its adverse effects	n = 17	Adults or adolescents with opioid dependence	N/A	N/A	No significant difference between groups was found on “wanting heroin” but a significant reduction favouring both naltrexone groups compared to placebo was found on “needing heroin”. No significant between group difference was found on severity of opioid and cocaine use	No significant between group difference was found on depression	N/A	No statistically significant difference in retention in treatment was found between those receiving naltrexone depot (lower or higher dose) or those receiving placebo. However, a significant difference between the high dose and the placebo groups was found on time to drop out
Maglione 2018 [64]	LOW	To evaluate the effects of MAT (using buprenorphine, buprenorphine plus naloxone, methadone, or naltrexone) for OUD on functional outcomes compared to wait-list, placebo, treatment without medication, any other comparator, or each other (e.g., buprenorphine vs naltrexone)	n = 40	Adults 18 years or older using medication-assisted treatment (MAT) for OUD—methadone, buprenorphine, buprenorphine plus naloxone, or naltrexone	N/A	Buprenorphine patients had a significantly lower prevalence of fatigue compared to methadone. No difference in insomnia between buprenorphine and methadone participants	N/A	N/A	N/A	N/A
Mattick 2009 [51]	LOW	To evaluate the effects of methadone maintenance treatment (MMT) compared with treatments that did not involve opioid replacement therapy	n = 11	Individuals with opioid dependence	No significant between-group difference was detected on mortality	N/A	Methadone was shown to significantly reduce heroin use compared to control conditions	N/A	N/A	Methadone had a superior retention rate compared with control conditions
Mattick 2014 [50]	LOW	To evaluate buprenorphine maintenance compared to placebo and to methadone	n = 31	Individuals dependent on heroin or other opioids	Three studies reported no deaths. One study reported a 20% mortality rate for the control group. One study reported two death with no relation to treatment condition.	N/A	Patients receiving high dosage of buprenorphine showed significantly less heroin use than the placebo group. Patients receiving medium dosage of buprenorphine showed significantly less cocaine and benzodiazepines use than the placebo group. No other significant differences were detected for different dosages.	N/A	N/A	Patients receiving flexible or low-dose methadone reported higher rates of retention in treatment than those with buprenorphine, whereas no significant differences were found on medium or high-dose groups. There was a significant benefit favouring patients receiving buprenorphine at any dose compared to those receiving placebo on treatment retention
Minozzi 2011 [52]	LOW	To evaluate the effects of naltrexone maintenance treatment versus placebo or other treatments in preventing relapse	n = 13	In-patients and out-patients dependent on heroin, or former heroin addicts dependent on methadone and participating in naltrexone treatment for opioid dependence	N/A	N/A	No significant difference between naltrexone versus placebo or no pharmacological treatment was found on abstinence. No significant difference between naltrexone versus psychotherapy was found on abstinence	N/A	N/A	No significant difference between naltrexone or naltrexone plus psychotherapy versus any control was found on retention in treatment
Platt 2017 [53]	MODERATE	To assess the effects of needle syringe programs and opioid substitution therapy, alone or in combination, for preventing acquisition of HCV	n = 28	People who inject drugs (opioids and/or stimulants)	N/A	Opioid substitution therapy (OST) was significantly associated with a reduction in HCV infection compared to no OST. High-coverage of needle syringe programs NSP was marginally associated with a reduction in HCV infection compared with lower coverage or no coverage, whereas no significant difference was found between low-coverage NSP versus no NSP. The combination of OST and NSP was significantly associated with a reduction in HCV infection.	N/A	N/A	N/A	N/A
Roozen 2006 [54]	CRITICALLY LOW	To summarize available and recent evidence on the effectiveness of naltrexone treatment	n = 7	Participants with alcohol or opiate abuse or dependence	N/A	N/A	Two studies reported no statistically significant medium- term difference on abstinence between naltrexone and the placebo groups, whereas one study reported a significant benefit favouring naltrexone. One study reported a significant benefit favouring naltrexone group compared to the placebo group on abstinence, whereas one study showed no significant difference No significant difference was found on continuous abstinence outcomes. There is a significant difference favouring the naltrexone group on drinking days compared to placebo. Abstinence was significantly higher for patients receiving naltrexone and supportive therapy.	N/A	N/A	There is a significant medium term benefit favouring patients receiving naltrexone on relapse rates compared to the placebo group There is moderate evidence favouring patients receiving naltrexone long term on relapse rates
Saulle 2017 [55]	LOW	To compare the effectiveness of opioid substitution therapy OST with supervised dosing relative to dispensing of medication for off-site consumption	n = 6	People “diagnosed as opioid dependent and receiving opioid substitution treatment with either buprenorphine or methadone	One study found that all-cause mortality was lower in the supervised methadone group. However, after adjustment insufficient evidence existed to support a protective effect.	Two trials found no significant difference between supervised and unsupervised therapy groups in serious adverse events requiring hospitalization	One study found no significant difference between supervised and unsupervised therapy in self-reported heroin use at three months or ASI’s composite score	N/A	N/A	There was no significant difference between between supervised and unsupervised therapy groups in retention rates across time
Simoens 2005 [56]	LOW	To evaluate the effectiveness of maintenance treatment with methadone or buprenorphine in treating opiate dependence	n = 48	Opiate dependent subjects aged 18 years old or over	N/A	N/A	Maintenance interventions with methadone or buprenorphine has been proven to be effective in reducing illicit opiate use and stimulating abstinence. No significant differences between methadone and buprenorphine were found in reducing illicit opiate use, cocaine use, or severity of withdrawal symptoms	N/A	N/A	Maintenance interventions with methadone or buprenorphine has been proven to be effective in promoting retention in treatment. Four studies found no significant difference between methadone and buprenorphine in retention in treatment, whereas three found additional benefit favouring methadone.
Sordo 2017 [57]	CRITICALLY LOW	To compare the risk for all cause and overdose mortality in people with opioid dependence during and after substitution treatment with methadone or buprenorphine and to characterise trends in risk of mortality after initiation and cessation of treatment.	n = 20	People with opiate dependence during and after substitution treatment with methadone or buprenorphine	Time spent in OST with methadone was associated with a significant reduction of mortalities. The mortality rate in treatment was less than a third of the rate out of treatment. Three studies showed that OST with buprenorphine could be associated with a reduction in mortality rates.	N/A	N/A	N/A	N/A	N/A
Standiford Helm 2008 [58]	CRITICALLY LOW	To evaluate and update the available evidence regarding the use of agonist/ antagonists to provide office- based opioid treatment for addiction.	n = 20	Patients with opioid dependence (whether they are in or out of treatment)	N/A	N/A	The combination of buprenorphine and naltrexone was found to significantly improve opioid dependence outcomes and had a significantly greater benefit on these outcomes compared to clonidine. Office-based treatment, levomethadyl, buprenorphine (with or without psychosocial treatments), and high dose methadone were found to be more effective than low dose methadone. No significant difference was found between methadone and buprenorphine. Tramadol compared favourably to buprenorphine in managing acute withdrawal symptoms/	N/A	N/A	As compared with low dose methadone, participants taking levomethadyl acetate had a higher rate of continuous abstinence from opioids, and those taking buprenorphine and high dose methadone had a trend towards higher rate of continuous abstinence.
Strang 2015 [59]	CRITICALLY LOW	To synthesise published findings for treatment with SIH for refractory heroin-dependence through systematic review and meta-analysis, and to examine the political and scientific response to these findings.	n = 6	Individuals with heroin dependence unresponsive to standard treatments	There was a positive but not significant effect favouring supervised injectable heroin SIH compared to oral methadone maintenance treatment on mortality	Five trials showed a significant higher risk of side effects in the SIH group compared to the oral MMT group	There was an overall positive effect favouring supervised injectable heroin SIH compared to oral MMT on illicit heroin use	N/A	N/A	Pooled analysis showed significant difference favouring SIH compared to oral methadone maintenance treatment on treatment retention
Thomas 2014 [60]	CRITICALLY LOW	To describe buprenorphine maintenance therapy and review available research on its efficacy	n = 19	Individuals with opioid dependence	N/A	N/A	The majority of studies reported a significant positive impact of buprenorphine maintenance treatment on illicit opioid use compared to placebo, lower dosage BMT or methadone maintenance treatment. However, when dosing the medication adequately, both buprenorphine and methadone showed comparable reduction in illicit opioid use.	N/A	N/A	Results indicate better treatment retention associated with methadone maintenance treatment. Rates of neonatal abstinence syndrome were similar for mothers treated with either buprenorphine or methadone. However, infant symptoms were less among the buprenorphine group
Weinmann 2004 [61]	CRITICALLY LOW	To assess the national and international literature on the effectiveness of substitution-based treatment	n = 13	Opiate addicts	N/A	N/A	Substitution therapy with methadone was found to be an effective strategy to reduce illicit drug use and improve the rehabilitation of opiate addicts	N/A	N/A	Speedy inclusion in the substitution therapy improved success of treatment
Wilder 2015 [62]	CRITICALLY LOW	To evaluate the discontinuation rates of MAT during pregnancy and in the immediate postpartum period. To examine what interventions (methadone or buprenorphine) improve treatment retention compared to standard care	n = 15	Pregnant and postpartum women with opioid use disorder	N/A	N/A	N/A	N/A	N/A	There is a scarce information on the range of prenatal and postnatal discontinuation rates, which limits generalizable findings. Duration of methadone treatment prior to delivery was inversely associated with risk for postpartum discontinuation of treatment
Pharmacologic agents for reversal of opioid poisoning
Bahji 2018 [41]	CRITICALLY LOW	To identify pharmacological interventions for the prevention and treatment of opioid overdose	n = 4	Patients with an established opioid use disorder	N/A	N/A	There was no significant difference in efficacy between naloxone and physostigmine for in-hospital treatment of heroin overdose. There was no significant difference in efficacy between intranasal and intramuscular naloxone. There was no significant difference in efficacy between naloxone and nalmefene (1 or 2 mg).	N/A	N/A	N/A