Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Relative effectiveness of medications for opioid-related disorders: A systematic review and network meta-analysis of randomized controlled trials

  • Jihoon Lim,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, QC, Canada

  • Imen Farhat,

    Roles Conceptualization, Data curation, Investigation, Methodology, Project administration, Writing – review & editing

    Affiliation Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, QC, Canada

  • Antonios Douros,

    Roles Conceptualization, Investigation, Methodology, Supervision, Writing – review & editing

    Affiliations Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, QC, Canada, Department of Medicine, McGill University, Montreal, QC, Canada, Centre for Clinical Epidemiology, Lady Davis Institute, Jewish General Hospital, Montreal, QC, Canada, Institute of Clinical Pharmacology and Toxicology, Charité—Universitätsmedizin Berlin, Berlin, Germany

  • Dimitra Panagiotoglou

    Roles Conceptualization, Investigation, Methodology, Supervision, Writing – review & editing

    Affiliation Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, QC, Canada



Several pharmacotherapeutic interventions are available for maintenance treatment for opioid-related disorders. However, previous meta-analyses have been limited to pairwise comparisons of these interventions, and their efficacy relative to all others remains unclear. Our objective was to unify findings from different healthcare practices and generate evidence to strengthen clinical treatment protocols for the most widely prescribed medications for opioid-use disorders.


We searched Medline, EMBASE, PsycINFO, CENTRAL, and for all relevant randomized controlled trials (RCT) from database inception to February 12, 2022. Primary outcome was treatment retention, and secondary outcome was opioid use measured by urinalysis. We calculated risk ratios (RR) and 95% credible interval (CrI) using Bayesian network meta-analysis (NMA) for available evidence. We assessed the credibility of the NMA using the Confidence in Network Meta-Analysis tool.


Seventy-nine RCTs met the inclusion criteria. Due to heterogeneity in measuring opioid use and reporting format between studies, we conducted NMA only for treatment retention. Methadone was the highest ranked intervention (Surface Under the Cumulative Ranking [SUCRA] = 0.901) in the network with control being the lowest (SUCRA = 0.000). Methadone was superior to buprenorphine for treatment retention (RR = 1.22; 95% CrI = 1.06–1.40) and buprenorphine superior to naltrexone (RR = 1.39; 95% CrI = 1.10–1.80). However, due to a limited number of high-quality trials, confidence in the network estimates of other treatment pairs involving naltrexone and slow-release oral morphine (SROM) remains low.


All treatments had higher retention than the non-pharmacotherapeutic control group. However, additional high-quality RCTs are needed to estimate more accurately the extent of efficacy of naltrexone and SROM relative to other medications. For pharmacotherapies with established efficacy profiles, assessment of their long-term comparative effectiveness may be warranted.

Trial Registration

This systematic review has been registered with PROSPERO ( (identifier CRD42021256212).


Opioid use disorder (OUD), including opioid dependence and addiction, is a problematic opioid use commonly characterized by tolerance and withdrawal symptoms. Adverse health outcomes associated with OUD include overdose, infectious diseases (e.g., AIDS; hepatitis C; and skin, soft tissue, and vascular infections), suicide, and death in severe cases [15]. A recent study on OUD burden revealed a global estimate of 40.5 million people who suffer from opioid dependence and 109,500 deaths from opioid overdose (including both accidental and intentional cases) in 2017 [6]. In response to the severity of OUD disease burden, several treatment pathways have emerged, with opioid maintenance therapy (OMT) as the most effective strategy.

OMT involves the use of opioid substitutes (e.g., buprenorphine, methadone, or naltrexone) to treat and manage addiction to opioids such as heroin, fentanyl, hydromorphone, and oxycodone [7]. Compared to other treatment pathways (e.g., detoxification, residential services, and behavioural interventions) or no treatment, OMT has lower risk of overdose hospitalizations [8], mortality [9], and frequency of injection drug use and needle sharing [1012]. With mounting evidence of its effectiveness, OMT has undoubtedly become the gold standard for treating opioid addiction and dependence, and public health stakeholders around the world have sought to expand access to medications for OUD in order to reduce the individual- and population-level burdens of OUD [1315].

Researchers from the United States, Canada, and European nations (e.g., the UK, France, Germany, Spain, and Finland) have adopted guidelines or reached a consensus to prescribe buprenorphine and methadone as first-line OMT treatments against OUD [1518]. At the same time, there have been growing interests in understanding the efficacy of other medications for OUD, such as naltrexone [19, 20] and slow-release oral morphine (SROM) [2123]. For example, recent clinical practice guidelines from Canada and the United States have outlined expectations of professional conduct in relation to prescribing these medications in addition to buprenorphine and methadone [17, 18]. These OMT guidelines reflect advances in clinical practice, and the availability of a wider range of therapeutic options for OMT medications may reduce OUD burden [14]. However, no review to date has compared treatment efficacy between several different OMT medications. Despite the availability of multiple medications, previous meta-analyses have been limited to pairwise comparisons of buprenorphine vs. methadone [24], SROM vs. methadone [21], or OMT medications vs. no maintenance treatment [20, 25]. In the absence of randomized controlled trials (RCT) comparing other combinations of treatment pairs, especially those involving naltrexone or SROM, their treatment efficacy relative to that of buprenorphine or methadone remains poorly understood. We undertook a systematic review with a network meta-analysis to establish which medications out of buprenorphine, methadone, naltrexone, and SROM have superior efficacy profile. This study unifies findings from different healthcare practices around the world and generates evidence to strengthen clinical treatment protocols for the most widely prescribed OMT medications.

Materials and methods

We conducted this systematic review in accordance with a pre-specified protocol (PROSPERO number: CRD42021256212), and we reported our findings following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Network Meta-Analyses (PRISMA-NMA) (S1 Table) [26].

Search strategy

We conducted a systematic literature search of studies comparing medications for OUD among people with opioid-related disorders in MEDLINE, EMBASE, PsycINFO, and Cochrane CENTRAL databases. We also scanned the bibliographies of the included articles and searched through the first 10 pages of Google Scholar (search terms: ‘randomized controlled trials opioid use disorder’) to identify additional references in the grey literature. We tailored our search strategy to each database, and search terms and keywords included those related to buprenorphine, methadone, naltrexone, slow-release oral morphine, and opioid-related disorders (see S2 Table for Search Strategy). There were no restrictions on the language of publication. Initial search was conducted on June 7, 2021, and it was updated on February 12, 2022.

Inclusion and exclusion criteria


This study included adult patients with problematic opioid or heroin use who were receiving pharmacotherapy, including buprenorphine, methadone, naltrexone, and SROM. Problematic use included OUD, opioid dependence, and opioid addiction as well as conditions specific to the use of heroin. These conditions were ascertained by the Diagnostic and Statistical Manual of Mental Disorders (DSM-3, -4, or -5) or the International Classification of Diseases (ICD) criteria. We excluded women who were pregnant in order to mitigate heterogeneity in study population that could arise from scientific and ethical complexities involving these patient population groups.

Interventions and comparators.

The main intervention for this study was pharmacotherapy for OUD including buprenorphine, methadone, naltrexone, and slow-release oral morphine. We included studies that compared any of the described intervention medications with each other or with a control group (e.g., placebo, standard of care, or no treatment). For inclusion, these interventions needed to be administered to treat OUD. Studies with one of the four intervention medications in one arm but another medication that is not one of the four intervention medications or controls (e.g., heroin or clonidine) were excluded. This was because they were not described in the latest best practice guidelines and consensuses for opioid maintenance treatment, and therefore did not reflect the most up-to-date clinical practice around the world (e.g., North America and Europe). In addition, we excluded from our analyses RCTs that compared different treatment regimens of the same medication, including studies where only the dosage, setting, or the route of administration (e.g., sublingual vs. injectable) differed between treatment groups.

For the multi-arm trials, we took the following analytic approach to evidence synthesis. First, in studies with three or more treatment arms, we extracted all intervention information but included only the eligible arms for evidence synthesis, following the approach taken by Rice et al. 2020 [27]. Second, in studies with two or more treatment arms with the same intervention (e.g., two arms with buprenorphine and two arms with methadone), we categorized each arm as ‘low dose’, ‘moderate dose’, or ‘high dose’ and treated two intervention arms with different medications that have the same categorization as a separate study. For example, if the four intervention arms were low dose buprenorphine, low dose methadone, moderate dose buprenorphine, and moderate dose methadone, the two low dose arms were included in the analysis as one study and the two moderate dose arms as another study. Third, in other trials with a small sample size in each arm, we combined multiple treatment arms with the same intervention into a larger single arm, if applicable, as had been done in earlier meta-analyses [20, 24, 25].


The primary outcome of interest was treatment retention at the end of the study. We defined treatment retention for each intervention or control arm as the number of participants who completed or remained in the study (did not withdraw) divided by the total number of participants who were randomized to a specific treatment in the beginning of the study. The secondary outcome of interest was opioid use based on urinalysis. Opioid use could be reported as either abstinence from opioids or illicit opioid use (e.g., heroin), and the definition of abstinence could vary by study duration or the number of times of opioid use. Therefore, for each intervention or control arm, we defined opioid use in two different ways: (1) Percentage of urine samples that were positive for opiates at the end of the study, and (2) Number of patients who had at least one urine sample that was positive for opiates at the end of the study out of all those randomized to each arm in the beginning of the study.

Study design.

We included RCTs that compared any of the four medications for OUD, namely buprenorphine, methadone, naltrexone, and SROM against each other for any treatment regimen (dose, frequency, timing, duration, and route of administration), placebo, standard of care, or no treatment. We restricted our study design to RCTs because they were best suited to address the relative efficacy of the pharmacotherapies on retention and opioid use, while eliminating confounding present in other study designs. Consequently, observational studies, case series, and case reports were excluded from our analysis. To identify additional eligible studies, we inspected earlier systematic reviews and meta-analyses, but these systematic reviews and meta-analyses themselves were not included in the qualitative and quantitative evidence syntheses.


Two independent reviewers (JL and IF) screened titles and abstracts, and then conducted a full-text review of all articles retrieved from the databases for eligibility (study selection) based on specified inclusion/exclusion criteria. We conducted title and abstract screening using a liberal accelerated approach, in which only one reviewer was needed to include a citation, while two reviewers were needed to exclude a citation [28]. At this stage, articles that were judged to be potentially relevant underwent a full-text review, which two reviewers performed independently and in duplicate. Disagreements between the two reviewers were resolved through consensus or adjudication by a third independent reviewer (DP), if necessary.

Data extraction

Two reviewers (JL and IF) conducted data extraction using a pre-piloted, standardized data extraction form (S3 Table). Data were independently extracted by a single reviewer (JL) and were verified by a second reviewer (IF). Disagreements were resolved through consensus or adjudication by a third independent reviewer (DP), if necessary. Extracted information included publication traits (year, setting/country, and funding source), study characteristics (trial design, trial duration, intervention type, comparator type, treatment regimen for each arm, number of participants randomized [total and for each arm], inclusion/exclusion criteria), basic participant characteristics (age and sex), and outcome data (number of participants retained and results from urinalysis for detecting opiates, including morphine and heroin). If multiple studies reported on the same cohort, we included only the most recent article with the most up-to-date study information.

Risk of bias assessment

Two reviewers (JL and IF) independently assessed the risk of bias for all included studies. We used Version 2 of the Cochrane Risk of Bias (RoB 2) tool to examine potential biases in five domains [29]. These include biases in randomization process, deviations from intended intervention, missing outcome data, measurement of the outcome, and selection of the reported results. Based on the assessment of each domain, we determined the overall risk of bias as one of ‘low risk of bias’, ‘some concerns’, and ‘high risk of bias’, where the overall risk was determined by the highest risk assigned in any individual domain.

Review of network geometry

We constructed network graphs to visualize the overall structure of the comparisons in our network. The nodes of the network graph represent the competing treatments, and an edge connects them if at least one RCT compared these corresponding treatments. We examined which treatments were compared directly (head-to-head comparisons) or indirectly (through one or more common comparators) and the amount of evidence generated from each comparison.

Strategies for evidence synthesis

We conducted a network meta-analysis of available direct and indirect evidence using a Bayesian framework to account for the correlation between treatment effects by different comparisons in multi-arm trials [30]. We chose to conduct network meta-analysis using the Bayesian framework, as this paradigm allows for probability statements [31], such as “There is X% probability that treatment A is the most efficacious out of all treatments”.

For treatment retention and opioid use, the treatment effect measure was the risk ratio (RR). To estimate the posterior distribution of the treatment effects, we conducted a Markov Chain Monte Carlo (MCMC) simulation with 100,000 iterations total and 5,000 burn-in iterations. To derive the posterior distribution, we ran both random-effects and fixed-effects models with an uninformative prior distribution of treatment effects. Next, we determined model fit using the deviance information criterion (DIC), where smaller DIC values correspond to better fit [32]. We then assessed the convergence of the MCMC simulations using the Gelman-Rubin-Brooks plot and the potential scale reduction factor (PSRF) [33, 34], where we determined that convergence was reached if the PSRF < 1.05.

We analyzed the outcomes by constructing the model for binary endpoints for network meta-analysis, and we calculated the RR with a 95% credible interval (CrI). To rank the preference of treatment options in the study, we plotted a rankogram and calculate the Surface Under the Cumulative Ranking (SUCRA) score [35]. The SUCRA score ranges from 0 to 1, where values closer to 1 indicate the more preferred treatments. Finally, we used the node split method to evaluate consistency of our network model, where consistency refers to the concordance of results between direct and indirect estimates within the network meta-analytic model. All analyses were conducted using R Studio version 4.0.5.

Sensitivity analysis

To assess whether study characteristics are associated with effect size differences, we conducted univariate network meta-regression analyses by the overall risk of bias (Low Risk vs. Some Concerns or High Risk from RoB 2) and by the year of study publication (Before 2010 vs. On or After 2010). We conducted an MCMC simulation with the same conditions as above and compared model fit with the main analysis model using the DIC. We also examined the credibility of the network meta-analysis results using the Confidence in Network Meta-Analysis (CINeMA) tool to improve transparency and limit subjectivity in the evidence synthesis process [36]. For each pairwise comparison, both direct and indirect, we specifically assessed six domains: (1) within-study bias, (2) reporting bias, (3) indirectness, (4) imprecision, (5) heterogeneity, and (6) incoherence. We evaluated each domain and rated it as ‘No concerns’, ‘Some concerns’, or ‘Major concerns’ based on a set of criteria described below. For all possible pairwise comparisons across the six domains, we took the average of the level of concerns to assign a rating. For additional details, please refer to S1 Text.


Search results

The database searches identified 13,262 publications (Fig 1). We removed 4,915 duplicates and an additional 7,915 studies upon review of the title and abstract. These studies were excluded because they were irrelevant to the research question or had observational study designs. Two reviewers independently examined the remaining 432 articles and excluded 360 of them (see S4 Table for the list of articles for full-text review and reasons for exclusion). In addition to the 72 RCTs that met our inclusion criteria, we identified 7 additional studies through grey literature search and review of earlier meta-analyses [3743]. In total, 79 studies were included for the qualitative and quantitative evidence syntheses [37115].

Study characteristics

Table 1 summarizes the key demographic and study characteristics of the RCTs included whose duration ranged from 2 weeks to 1 year. Additional details of each study can be found in the S5 Table. The included studies were published between 1976 and 2021, and the size of the study ranged from 19 to 1,269 participants. These trials were conducted in the United States (N = 37) [38, 39, 42, 48, 5054, 58, 6065, 68, 69, 71, 7880, 82, 83, 85, 89, 90, 92, 97, 99, 101103, 107109, 114], Europe including Russia (N = 24) [40, 43, 47, 49, 56, 57, 66, 67, 7277, 84, 88, 95, 96, 98, 104, 106, 110, 111, 113], Asia including the Middle East and the Caucasus (N = 14) [4446, 59, 70, 81, 87, 91, 93, 94, 100, 105, 112, 115], and Australia (N = 4) [37, 41, 55, 86]. Most studies were financially supported by the government or non-profit organization grants. However, 10 studies did not mention their funding source [37, 41, 43, 45, 46, 48, 59, 81, 88, 104], and seven were industry-sponsored trials [47, 56, 60, 74, 82, 84, 112]. The analysis included 70 trials with a parallel design [3846, 4855, 5761, 6388, 9199, 101108, 110114], three with a cross-over design [37, 47, 56], and six with a factorial design [62, 89, 90, 100, 109, 115]. There were 46 double- or triple-blind trials [3840, 44, 48, 50, 53, 54, 56, 5961, 6365, 67, 68, 71, 7377, 8184, 8688, 91, 92, 9598, 100102, 104, 105, 107109, 111, 112], 22 open-label trials [37, 41, 43, 47, 49, 51, 52, 55, 57, 66, 69, 70, 7880, 89, 90, 99, 103, 115], and 11 trials that did not specify blinding mechanism used [42, 45, 46, 58, 62, 72, 85, 93, 94, 106, 114].

Most of the patients in the included studies were male (78% among those that reported it for the entire study population before randomization), with 12 trials including only male participants [4446, 48, 61, 69, 79, 85, 91, 93, 105, 115]. The mean age of the participants in the study was 34 among those that reported it for the entire study population before randomization (range of mean age: 20 to 47). All of the included studies focused on patients with OUD, with 13 studies focusing specifically on patients with heroin-related disorders [37, 41, 44, 50, 55, 6567, 69, 77, 91, 103, 114]. Two studies recruited patients with co-occurring opioid use disorder and chronic non-cancer pain [89, 90], while 61 studies excluded patients with psychiatric disorders or concurrent dependence on alcohol, sedative, or hypnotic medications [37, 39, 40, 42, 43, 4850, 52, 5457, 59, 60, 6268, 70, 7480, 8288, 9093, 95113, 115].

Risk of bias of included RCTs

Overall, the risk of bias was judged ‘Low’ for 38 studies [39, 40, 50, 5357, 60, 6367, 70, 7380, 82, 83, 8587, 89, 90, 97, 100102, 109112], ‘Some Concerns’ for 37 studies [37, 38, 4143, 4549, 51, 52, 58, 59, 62, 68, 71, 72, 81, 84, 88, 9196, 98, 99, 104108, 113, 115], and ‘High’ for 4 studies [44, 69, 103, 114] (Table 2; see S6 Table for a more detailed assessment of risk of bias). For randomization and allocation concealment, 49 trials were judged ‘Low Risk’ [3941, 48, 50, 5357, 5961, 6368, 70, 71, 7380, 82, 83, 8587, 89, 90, 92, 9698, 100102, 104, 109113], and 30 trials were judged ‘Some Concerns’ [37, 38, 4247, 49, 51, 52, 58, 62, 69, 72, 81, 84, 88, 91, 9395, 99, 103, 105108, 114, 115]. The trials in the latter group were rated as such due to unclear reporting of whether the allocation sequence was concealed until participants were enrolled and assigned to interventions. That is, these studies provided no information on whether treatment allocation was remotely or centrally administered. For the three cross-over studies [37, 47, 56], we also examined the risk of bias associated with period or carryover effects. Two studies [47, 56] were judged to have ‘Low Risk’ of carryover effects, but one study [37] was judged to have ‘Some Concerns’ because it had provided no information on whether there was sufficient time between the two study periods for the carryover effects to have disappeared. For most studies, the risk of bias due to deviations from the intended interventions and the risk of bias due to missing outcome data was ‘Low’. However, two trials were judged to have higher risk of bias due to deviations from the intended interventions because a large number of patients failed to remain and adhere to the originally assigned treatment shortly after randomization [44, 114]. Relatedly, four studies that examined opioid use as their main endpoint were at higher risk of bias due to missing outcome data because they employed per-protocol analyses that excluded participants who were lost to follow-up before the end of the study period [43, 69, 103, 114]. Since both treatment retention and opioid use were measured objectively, the risk of bias in the measurement of outcome was judged ‘Low Risk’ for all studies, and the risk of bias from selective reporting remains low. However, due to the lack of available protocol in the grant (e.g., NIH REPORTER) or trial registry (e.g.,, the risk of reporting bias was judged ‘Some Concerns’ for 31 studies [37, 38, 41, 4346, 48, 49, 52, 58, 59, 61, 68, 71, 72, 81, 84, 88, 91, 92, 95, 96, 98, 99, 103105, 107, 108, 113].

Evidence synthesis

Due to differences in approach to measurement of opioid use (e.g., differences in defining abstinence) even between studies using urinalysis to ascertain it, we judged that the network meta-analysis (NMA) would not produce reliable estimates for the secondary outcome. Therefore, we only conducted NMA for treatment retention (primary outcome).

Synthesis of primary outcome.

In our network meta-analysis, 73 studies, representing 5 interventions, presented data on treatment retention [3742, 4461, 6368, 70102, 104113], measured as the number of patients who remained in treatment at the end of the study. Fig 2 illustrates a network diagram that summarizes the overall structure of the competing treatments in our network of interventions. The most frequently analyzed pairwise comparison has been the buprenorphine-methadone comparison (N = 35 from 32 studies) [41, 4446, 48, 49, 57, 63, 65, 67, 68, 7072, 83, 85, 86, 8890, 92, 9496, 99, 101, 102, 104, 106108, 113], followed by the naltrexone-control (N = 21) [38, 40, 5052, 59, 61, 62, 7479, 81, 93, 98, 100, 105, 109, 111] and buprenorphine-control (N = 13) [39, 5355, 60, 64, 66, 73, 82, 97, 100, 112, 115] comparisons. While treatment retention for methadone has been compared with that for all other treatments and controls in our network, the efficacy of SROM has only been compared with that of methadone [37, 47, 56, 84].

Fig 2. Network geometry for the treatment retention outcome for each medication or control group.

For the primary analysis, we initially ran both fixed effects (FE) and random effects (RE) model. Upon comparison of the DIC values, both models demonstrated similar model fit (FE: 294.28 vs. RE: 294.58), but we chose to conduct all subsequent analyses with the random effects model (see S7 Table for the explanation and model fit comparison). The MCMC simulation with the random effects model resulted in the PSRF of 1.0037, suggesting convergence of the MCMC algorithm (i.e., the simulation resulted in an accurate estimate of our parameters; see S1 and S2 Figs for detailed results of the simulation).

All pharmacotherapy options were more efficacious with respect to treatment retention than the control group. Compared to the control group, the likelihood of retention was 2.62 (95% CrI = 2.09–3.33), 2.52 (95% CrI = 1.62–3.94), 2.15 (95% CrI = 1.76–2.69), and 1.54 (95% CrI = 1.26–1.90) times higher for methadone, SROM, buprenorphine, and naltrexone, respectively (Fig 3). Further, the average percentage of treatment retention across all studies was 77.6% for SROM, 64.1% for methadone, 54.3% for buprenorphine, 41.0% for naltrexone, and 30.1% for control (see S8 Table and S3 Fig). Analysis using the SUCRA score also indicated similar findings (see S8 Table and S4 Fig for additional details). Methadone was the highest ranked treatment based on the SUCRA value of 0.901. The next highest ranked treatment was SROM (SUCRA = 0.784), followed by buprenorphine (SUCRA = 0.559), naltrexone (SUCRA = 0.257), and control (SUCRA = 0.000). While the SUCRA values demonstrated which treatment in the network may be the most efficacious, pairwise comparisons revealed the relative effectiveness of one treatment versus the other in retaining patients in the initially assigned treatment (Table 3). Methadone had a higher likelihood of retention than buprenorphine (RR = 1.22; 95% CrI = 1.06–1.40) and naltrexone (RR = 1.69; 95% CrI = 1.30–2.24) but remained statistically equivalent to SROM (RR = 1.04; 95% CrI = 0.71–1.52). Similarly, buprenorphine (RR = 1.39; 95% CrI = 1.10–1.80) and SROM (RR = 1.63; 95% CrI = 1.01–2.63) both had a higher likelihood of retention than naltrexone, but the two medications remained equivalent (RR = 0.86; 95% CrI = 0.57–1.28 with SROM as the reference group).

Fig 3. Forest plot of the risk ratio for treatment retention for each pharmacotherapy compared to the control group.

Table 3. Risk ratio estimates for treatment retention and 95% credible interval (CrI)1.

Fig 4 illustrates forest plots of direct and indirect estimates of likelihood of treatment retention from the node split method, which was used to evaluate consistency of our network model. Results from six pairwise comparisons involving buprenorphine, methadone, naltrexone, and control group were presented. Pairwise comparisons involving SROM were not included in this node split approach because methadone was its only comparator, and SROM was disconnected from all other interventions including the control group in the network graph. For all pairwise comparisons except naltrexone versus methadone, the estimates from direct and indirect comparisons were statistically equivalent (p-value > 0.05). However, for the naltrexone versus methadone pair, there was some evidence of inconsistency (p-value = 0.027), as the magnitude of the RR was larger for the direct estimate than for the indirect estimate.

Fig 4. Direct and indirect estimate of the likelihood of retention between treatments using the node-splitting method to assess consistency of the network model.

As for sensitivity analysis, model fit from univariate network meta-regression analysis adjusting for risk of bias and publication year remained similar to that from the unadjusted analysis (DIC: Unadjusted = 294.58; Risk of bias = 293.46; Publication year = 294.66). Thus, it remains unlikely that controlling for these study characteristics and methodological differences between studies influenced the magnitude of the effect sizes in our network. Using the CINeMA tool (Table 4, S5 and S6 Figs), we judged with ‘High’ level of confidence the estimates of the likelihood of treatment retention comparing buprenorphine vs. control and methadone vs. control. For the buprenorphine vs. methadone and buprenorphine vs. naltrexone comparisons, we assigned a ‘Moderate’ confidence rating due to the presence of heterogeneity. For the methadone vs. naltrexone comparison, we assigned a confidence rating of ‘Low’ due to having a small number of studies as well as concerns related to the ‘Heterogeneity’ and ‘Incoherence’ domains for the estimates derived from the network. Finally, all pairwise comparisons involving SROM received the confidence rating of ‘Very Low’ due to high degree of incoherence that resulted from insufficient evidence in the network.

Table 4. Summary of credibility of the network estimates using the Confidence in Network Meta-Analysis (CINeMA) tool.

Synthesis of secondary outcomes.

Due to both small number of studies reporting secondary outcomes and inconsistent reporting formats, we were unable to meta-analyze studies reporting these outcomes. We present below details of the post-hoc analysis of the secondary outcomes.

Thirty-five studies [47, 5053, 57, 5963, 65, 71, 72, 76, 78, 79, 82, 83, 88, 9398, 101, 102, 107, 108, 110112, 115] assessed opioid use as the percentage of opioid positive urine samples; twelve studies [43, 52, 53, 61, 62, 69, 74, 87, 98, 103, 105, 114] assessed opioid use as the number of patients who had at least one opioid positive urine sample at the end of the study out of all those randomized to each arm in the beginning of the study; and three studies [61, 62, 98] applied both definitions. In addition, there were heterogeneities in reporting opioid use across studies. Among the studies that reported percentage of opioid positive urine samples, nine reported the total number of tests and number of positive urine samples [50, 59, 61, 62, 76, 79, 82, 88, 94], eight reported percentages with a 95% CI or standard error [47, 50, 54, 63, 76, 82, 97, 115], and six reported percentages with a standard deviation [60, 62, 83, 95, 101, 110]. However, 13 studies only reported the percentage value without other parameters to assess variability of the sample [51, 57, 65, 71, 72, 78, 96, 98, 102, 107, 108, 111, 112].


All pharmacotherapeutic strategies were associated with a higher likelihood of treatment retention compared to the control arm, which consisted of placebo, standard of care, or no treatment. Based on SUCRA rankings, methadone appeared to be the most effective pharmacotherapy for treatment retention, followed by SROM, buprenorphine, and naltrexone. Buprenorphine was superior to naltrexone, and naltrexone was superior to non-pharmacotherapeutic interventions in treatment retention. SROM was compared only with methadone, and all other pairwise comparisons involving SROM were based on indirect evidence. The lack of available evidence on direct comparison also raised major concerns with respect to the ‘Incoherence’ domain in the CINeMA, which in turn lowered confidence in the estimates related to SROM generated by the NMA.

To our knowledge, this is the first network meta-analysis of the medications for opioid use disorder to examine their relative efficacy against each other. The findings from this study remain consistent with those from earlier meta-analyses by Mattick et al. (2014) and Mattick et al. (2009) [24, 25], which concluded that buprenorphine and methadone were associated with higher treatment retention than non-pharmacotherapeutic controls and that methadone resulted in superior treatment retention to buprenorphine. At the same time, our study provides an important update to an earlier Cochrane review by Minozzi et al. (2011) [20]. This review concluded that naltrexone did not result in higher retention than placebo or non-pharmacotherapeutic agents, and that naltrexone was non-inferior to buprenorphine in treatment retention from only one study [100]. The authors’ conclusions were based on a limited number of studies that were published before June 2010. With additional evidence, we found that naltrexone is an effective treatment strategy for retention compared to non-pharmacotherapeutic interventions, which contrasts findings by Minozzi et al. (2011). Further, our NMA generated preliminary evidence on the superiority of buprenorphine to naltrexone in treatment retention, supported by both the risk ratio estimates and the average proportion of retention calculated for each medication. However, additional RCTs will need to be conducted to better ascertain the efficacy of naltrexone in relation to methadone or SROM, since the network estimates were largely based on indirect comparisons.

Evidence on naltrexone from this network meta-analysis may also have important clinical implications in relation to the current guidelines that recommend buprenorphine and methadone as first-line treatments [1517]. A 2015 guideline by the American Society of Addiction Medicine (ASAM) described buprenorphine, methadone, and naltrexone as more effective than non-pharmacotherapeutic strategies in treating OUD but concluded that their relative advantages over each other remain unknown [18]. Our NMA partially addresses this important gap raised in the ASAM guideline by generating evidence that demonstrates greater likelihood of treatment retention associated with buprenorphine than with naltrexone. In other words, should the clinicians prioritize retention while treating patients with OUD, buprenorphine may be the preferred treatment to naltrexone. On the other hand, although methadone was ranked higher than both buprenorphine and naltrexone using the SUCRA scores, the network estimates from the methadone-naltrexone pair would need to be interpreted more cautiously. Only one trial conducted a head-to-head comparison of these two medications, which subsequently raised some concerns around heterogeneity and incoherence of the NMA results. While methadone may have higher likelihood of treatment retention than naltrexone, the extent to which this is true may require further validation through additional RCTs.

Evidence on SROM from this network meta-analysis reveals the potential methodological issues surrounding earlier trials that contributed to the current clinical practice guidelines. In Canada, for example, clinical practice guidelines have stated that SROM could be a safe and effective alternative to treating OUD based on a small number of studies comparing SROM to methadone [17, 116]. Two review studies reached similar conclusions by stating that treatment retention may be similar between SROM and methadone, but the authors of these studies also acknowledged that the methodological quality of the included studies may be low to moderate [21, 22]. Our NMA results were consistent with these reviews and guidelines, as the number of high-quality RCTs comparing SROM to other medications for OUD was limited. However, through extensive examination of the risk of bias and credibility of the NMA estimates, we also observed that most of the trials comparing SROM to methadone had been industry-sponsored by the manufacturers of SROM, namely the Mundipharma Medical Company [47, 56, 84], which had not been detected in earlier reviews. Relatedly, the recommendations in the above guidelines may be based on studies with a potential for reporting bias, further lowering the confidence in the estimates from pairwise comparisons in the NMA involving SROM. Therefore, additional RCTs comparing SROM with other pharmacotherapeutic interventions are needed to better determine comparative effectiveness and facilitate treatment decisions for opioid maintenance treatments.

Our study has several strengths. First, we conducted a rigorous and extensive literature search across five databases and grey literature. Second, we included only the trials whose participants had opioid-related disorders and whose outcomes were measured objectively, which include treatment retention and opioid use measured by urinalysis. Although this may have reduced the number of studies in the network, we applied such eligibility criteria to ensure comparable study populations across studies. Third, the network meta-analysis included all pharmacotherapeutic strategies in the model. This allowed us to derive risk ratio estimates from both direct and indirect comparisons of all treatments in the network, which was not possible in the previous pairwise meta-analyses. Fourth, we examined the credibility of our network meta-analytic estimates using the CINeMA tool, a state-of-the-art platform that enabled a thorough assessment of the NMA results across six methodological domains.

Our study also has a few limitations. First, as with all network meta-analyses, randomization holds within each study but not between different studies [117], which could lead to heterogeneity in the study population across studies and subsequently yield biased results. However, we followed a pre-specified protocol with pre-determined eligibility criteria, which allowed us to mitigate between-study population differences. Second, due to heterogeneities between studies in outcome measurement and reporting formats, we were unable to meta-analyze trials on secondary outcomes. Although we conducted NMA only on the treatment retention outcome, findings on this endpoint remain clinically relevant because improvements in retention can reduce illicit opioid use as well as psychiatric, medical, and legal setbacks, while enhancing quality of life for people with OUD [118, 119]. Third, evidence on comparisons involving naltrexone or SROM remained sparse, and confidence in the network meta-analytic estimates was low. This hindered us from drawing substantive conclusions on their efficacy and advantages in relation to the more widely prescribed buprenorphine and methadone. Fourth, about 78% of all participants across the studies in the network were male, which could compromise the generalizability of the study findings. We note, however, that the large majority of those who are treated for opioid use disorders are also male patients[13, 120122]. Therefore, the generalizability of our study findings likely still holds true. Finally, long-term comparative effectiveness of the included pharmacotherapies remains uncertain because only RCTs were included in the network. However, the findings from our NMA serve as evidence that could pave the way for future observational studies, which are accepted as the best evidence for clinical and policy decision-making surrounding these medications [123].


For treating opioid-related disorders, maintenance treatment through buprenorphine, methadone, naltrexone, and SROM is more effective than non-pharmacotherapeutic interventions. Among the medications included in the network, methadone appears to be the most efficacious pharmacotherapy for treatment retention. Due to limitations in reporting and heterogeneity in outcome measurement formats, the relative efficacy of these interventions for other clinical endpoints remains unclear. Buprenorphine and methadone appear to have superior retention to naltrexone based on a small number of studies. Upon comparison with methadone, the efficacy of slow-release oral morphine was not statistically different. However, the lack of comparison with other pharmacotherapeutic options and the potential presence of reporting bias may hinder accurate conclusions about the efficacy of SROM. Finally, our study revealed directions for future research, which include (1) further RCTs involving naltrexone or SROM to assess their relative efficacy in relation to buprenorphine and methadone, and (2) observational studies to examine long-term comparative effectiveness of these medications.

Supporting information

S1 Text. Analysis plan for using the CINeMA framework.


S1 Table. PRISMA NMA checklist of items to include when reporting a systematic review involving a network meta-analysis of randomized controlled trials’.


S4 Table. List of excluded studies (full-test review) along with reasons for exclusion.


S5 Table. Detailed description of included studies.


S6 Table. Detailed assessment of risk of bias using RoB 2.0.


S7 Table. Deviance information criterion (DIC) for fixed effects model versus random effects model0 and meta-regression analysis.

0 This was because the fixed effects model assumes that all studies share the same common effect, but it may not be reasonable to assume that there is one common effect size. On the other hand, the random effects model assumes that the observed estimates of the treatment effect can vary across studies due to systematic differences in the treatment effect as well as random variations due to chance [Riley, R.D., J.P. Higgins, and J.J. Deeks, Interpretation of random effects meta-analyses. BMJ, 2011. 342: p. d549.]. 1 Comparison with the random effects model. 2 Categorization of risk of bias (RoB) was based on whether the overall RoB was ‘Low risk’ or ‘Some or High risk’ using the RoB 2.0 tool. 3 Publication year was stratified by whether the article was published ‘Before 2010’ or ‘On or after 2010’.


S8 Table. Average percentage of treatment retention and Surface under the cumulative ranking (SUCRA) score for each treatment in the network.


S1 Fig. Convergence of the Monte Carlo Markov Chain (MCMC) simulation with treatment retention as outcome.


S2 Fig. Gelman-Rubin plot to visualize convergence of the MCMC simulation with treatment retention as outcome.


S3 Fig. Rankogram of the preference of treatment options with respect to the retention outcome.


S4 Fig. Surface under the cumulative ranking (SUCRA) score diagram for the retention outcome.


S5 Fig. Confidence in network meta-analysis (CINeMA) risk of bias contributions for each pairwise comparison based on RoB 2.


S6 Fig. Confidence in network meta-analysis (CINeMA) indirectness contributions for each pairwise comparison.



DP and AD are recipients of Chercheur-Boursier Junior 1 Award from the Fonds de recherche du Québec–Santé (FRQS). IF is a recipient of the FRQS Doctoral Training Award.


  1. 1. Wilcox HC, Conner KR, Caine ED. Association of alcohol and drug use disorders and completed suicide: an empirical review of cohort studies. Drug Alcohol Depend. 2004;76 Suppl:S11–9. pmid:15555812
  2. 2. Mathers BM, Degenhardt L, Bucello C, Lemon J, Wiessing L, Hickman M. Mortality among people who inject drugs: a systematic review and meta-analysis. Bull World Health Organ. 2013;91(2):102–23. pmid:23554523
  3. 3. Mathers BM, Degenhardt L, Phillips B, Wiessing L, Hickman M, Strathdee SA, et al. Global epidemiology of injecting drug use and HIV among people who inject drugs: a systematic review. Lancet. 2008;372(9651):1733–45. pmid:18817968
  4. 4. Degenhardt L, Peacock A, Colledge S, Leung J, Grebely J, Vickerman P, et al. Global prevalence of injecting drug use and sociodemographic characteristics and prevalence of HIV, HBV, and HCV in people who inject drugs: a multistage systematic review. Lancet Glob Health. 2017;5(12):e1192–e207. pmid:29074409
  5. 5. Trickey A, Fraser H, Lim AG, Peacock A, Colledge S, Walker JG, et al. The contribution of injection drug use to hepatitis C virus transmission globally, regionally, and at country level: a modelling study. Lancet Gastroenterol Hepatol. 2019;4(6):435–44. pmid:30981685
  6. 6. Degenhardt L, Grebely J, Stone J, Hickman M, Vickerman P, Marshall BDL, et al. Global patterns of opioid use and dependence: harms to populations, interventions, and future action. Lancet. 2019;394(10208):1560–79. pmid:31657732
  7. 7. Centre for Addiction and Mental Health. Opioid Agonist Therapy Toronto, ON 2016 [Available from:
  8. 8. Wakeman SE, Larochelle MR, Ameli O, Chaisson CE, McPheeters JT, Crown WH, et al. Comparative Effectiveness of Different Treatment Pathways for Opioid Use Disorder. JAMA Netw Open. 2020;3(2):e1920622. pmid:32022884
  9. 9. Sordo L, Barrio G, Bravo MJ, Indave BI, Degenhardt L, Wiessing L, et al. Mortality risk during and after opioid substitution treatment: systematic review and meta-analysis of cohort studies. BMJ. 2017;357:j1550. pmid:28446428
  10. 10. Karki P, Shrestha R, Huedo-Medina TB, Copenhaver M. The Impact of Methadone Maintenance Treatment on HIV Risk Behaviors among High-Risk Injection Drug Users: A Systematic Review. Evid Based Med Public Health. 2016;2. pmid:27066590
  11. 11. Mittal ML, Jain S, Sun S, DeBeck K, Milloy MJ, Hayashi K, et al. Opioid agonist treatment and the process of injection drug use initiation. Drug Alcohol Depend. 2019;197:354–60. pmid:30922483
  12. 12. Mittal ML, Vashishtha D, Sun S, Jain S, Cuevas-Mota J, Garfein R, et al. History of medication-assisted treatment and its association with initiating others into injection drug use in San Diego, CA. Subst Abuse Treat Prev Policy. 2017;12(1):42. pmid:28974239
  13. 13. Pearce LA, Min JE, Piske M, Zhou H, Homayra F, Slaunwhite A, et al. Opioid agonist treatment and risk of mortality during opioid overdose public health emergency: population based retrospective cohort study. BMJ. 2020;368:m772. pmid:32234712
  14. 14. Priest KC, Gorfinkel L, Klimas J, Jones AA, Fairbairn N, McCarty D. Comparing Canadian and United States opioid agonist therapy policies. Int J Drug Policy. 2019;74:257–65. pmid:30765118
  15. 15. Dematteis M, Auriacombe M, D’Agnone O, Somaini L, Szerman N, Littlewood R, et al. Recommendations for buprenorphine and methadone therapy in opioid use disorder: a European consensus. Expert Opin Pharmacother. 2017;18(18):1987–99. pmid:29183228
  16. 16. World Health Organization. Guidelines for the psychosocially assisted pharmacological treatment of opioid dependence: World Health Organization; 2009.
  17. 17. Bruneau J, Ahamad K, Goyer ME, Poulin G, Selby P, Fischer B, et al. Management of opioid use disorders: a national clinical practice guideline. CMAJ. 2018;190(9):E247–E57. pmid:29507156
  18. 18. Kampman K, Jarvis M. American Society of Addiction Medicine (ASAM) National Practice Guideline for the Use of Medications in the Treatment of Addiction Involving Opioid Use. J Addict Med. 2015;9(5):358–67. pmid:26406300
  19. 19. Jarvis BP, Holtyn AF, Subramaniam S, Tompkins DA, Oga EA, Bigelow GE, et al. Extended-release injectable naltrexone for opioid use disorder: a systematic review. Addiction. 2018;113(7):1188–209.
  20. 20. Minozzi S, Amato L, Vecchi S, Davoli M, Kirchmayer U, Verster A. Oral naltrexone maintenance treatment for opioid dependence. Cochrane Database Syst Rev. 2011(4):CD001333. pmid:21491383
  21. 21. Klimas J, Gorfinkel L, Giacomuzzi SM, Ruckes C, Socias ME, Fairbairn N, et al. Slow release oral morphine versus methadone for the treatment of opioid use disorder. BMJ Open. 2019;9(4):e025799. pmid:30944135
  22. 22. Ferri M, Minozzi S, Bo A, Amato L. Slow-release oral morphine as maintenance therapy for opioid dependence. Cochrane Database Syst Rev. 2013(6):CD009879. pmid:23740540
  23. 23. Jegu J, Gallini A, Soler P, Montastruc JL, Lapeyre-Mestre M. Slow-release oral morphine for opioid maintenance treatment: a systematic review. Br J Clin Pharmacol. 2011;71(6):832–43. pmid:21265874
  24. 24. Mattick RP, Breen C, Kimber J, Davoli M. Buprenorphine maintenance versus placebo or methadone maintenance for opioid dependence. Cochrane Database Syst Rev. 2014(2):CD002207. pmid:24500948
  25. 25. Mattick RP, Breen C, Kimber J, Davoli M. Methadone maintenance therapy versus no opioid replacement therapy for opioid dependence. Cochrane Database Syst Rev. 2009(3):CD002209. pmid:19588333
  26. 26. Hutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann Intern Med. 2015;162(11):777–84. pmid:26030634
  27. 27. Rice D, Corace K, Wolfe D, Esmaeilisaraji L, Michaud A, Grima A, et al. Evaluating comparative effectiveness of psychosocial interventions adjunctive to opioid agonist therapy for opioid use disorder: A systematic review with network meta-analyses. PLoS One. 2020;15(12):e0244401. pmid:33370393
  28. 28. Khangura S, Konnyu K, Cushman R, Grimshaw J, Moher D. Evidence summaries: the evolution of a rapid review approach. Syst Rev. 2012;1:10. pmid:22587960
  29. 29. Sterne JAC, Savovic J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366:l4898. pmid:31462531
  30. 30. Franchini AJ, Dias S, Ades AE, Jansen JP, Welton NJ. Accounting for correlation in network meta-analysis with multi-arm trials. Res Synth Methods. 2012;3(2):142–60. pmid:26062087
  31. 31. Jansen JP, Trikalinos T, Cappelleri JC, Daw J, Andes S, Eldessouki R, et al. Indirect treatment comparison/network meta-analysis study questionnaire to assess relevance and credibility to inform health care decision making: an ISPOR-AMCP-NPC Good Practice Task Force report. Value Health. 2014;17(2):157–73. pmid:24636374
  32. 32. Spiegelhalter DJ, Best NG, Carlin BP, Van Der Linde A. Bayesian measures of model complexity and fit. Journal of the royal statistical society: Series b (statistical methodology). 2002;64(4):583–639.
  33. 33. Brooks SP, Gelman A. General methods for monitoring convergence of iterative simulations. Journal of computational and graphical statistics. 1998;7(4):434–55.
  34. 34. Gelman A, Rubin DB. Inference from iterative simulation using multiple sequences. Statistical science. 1992;7(4):457–72.
  35. 35. Salanti G, Ades AE, Ioannidis JP. Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. J Clin Epidemiol. 2011;64(2):163–71. pmid:20688472
  36. 36. Nikolakopoulou A, Higgins JPT, Papakonstantinou T, Chaimani A, Del Giovane C, Egger M, et al. CINeMA: An approach for assessing confidence in the results of a network meta-analysis. PLoS Med. 2020;17(4):e1003082. pmid:32243458
  37. 37. Clark NC. Expanding the Options for the Treatment of Heroin Dependence with Oral Supervised Opioid Substitution Therapy: LAAM, Buprenorphine and Slow Release Oral Morphine as Alternatives to Methadone: University of Melbourne, Department of Medicine, St Vincent’s Hospital; 2006.
  38. 38. Curran S, Savage C. Patient response to naltrexone: issues of acceptance, treatment effects, and frequency of administration. NIDA Research Monograph. 1976;9:67–9.
  39. 39. Fudala PJ, Bridge TP, Herbert S, Williford WO, Chiang CN, Jones K, et al. Office-based treatment of opiate addiction with a sublingual-tablet formulation of buprenorphine and naloxone. New England Journal of Medicine. 2003;349(10):949–58. pmid:12954743
  40. 40. Krupitsky EM, Zvartau EE, Masalov DV, Tsoy MV, Burakov AM, Egorova VY, et al. Naltrexone with or without fluoxetine for preventing relapse to heroin addiction in St. Petersburg, Russia. Journal of Substance Abuse Treatment. 2006;31(4):319–28. pmid:17084785
  41. 41. Lintzeris N, Ritter A, Panjari M, Clark N, Kutin J, Bammer G. Implementing Buprenorphine Treatment in Community Settings in Australia: Experiences from the Buprenorphine Implementation Trial. American Journal on Addictions. 2004;13(SUPPL. 1):S29–S41. pmid:15204674
  42. 42. Schwartz RP, Kelly SM, Mitchell SG, O’Grady KE, Sharma A, Jaffe JH. Methadone treatment of arrestees: A randomized clinical trial. Drug and Alcohol Dependence. 2020;206 (no pagination). pmid:31753737
  43. 43. Stella L, D’Ambra C, Mazzeo F, Capuano A, Del Franco F, Avolio A, et al. Naltrexone plus benzodiazepine aids abstinence in opioid-dependent patients. Life Sciences. 2005;77(21):2717–22. pmid:15979652
  44. 44. Ahmadi J. Methadone versus buprenorphine maintenance for the treatment of heroin-dependent outpatients. Journal of Substance Abuse Treatment. 2003;24(3):217–20. pmid:12810142
  45. 45. Ahmadi J, Ahmadi K, Ohaeri J. Controlled, randomized trial in maintenance treatment of intravenous buprenorphine dependence with naltrexone, methadone or buprenorphine: A novel study. European Journal of Clinical Investigation. 2003;33(9):824–9. pmid:12925043
  46. 46. Ahmadi J, Maany I, Ahmadi M. Treatment of intravenous buprenorphine dependence. A randomized open clinical trial. German Journal of Psychiatry. 2003;6(1):23–9.
  47. 47. Beck T, Haasen C, Verthein U, Walcher S, Schuler C, Backmund M, et al. Maintenance treatment for opioid dependence with slow-release oral morphine: a randomized cross-over, non-inferiority study versus methadone. Addiction (Abingdon, England). 2014;109(4):617–26. pmid:24304412
  48. 48. Bickel WK, Stitzer ML, Bigelow GE, Liebson IA, Jasinski DR, Johnson RE. A clinical trial of buprenorphine: Comparison with methadone in the detoxification of heroin addicts. Clinical Pharmacology and Therapeutics. 1988;43(1):72–8. pmid:3275523
  49. 49. Cameron IM, Matheson CI, Bond CM, McNamee P, Lawrie T, Robinson A, et al. Pilot randomised controlled trial of community pharmacy administration of buprenorphine versus methadone. International Journal of Pharmacy Practice. 2006;14(4):243–8.
  50. 50. Comer SD, Sullivan MA, Yu E, Rothenberg JL, Kleber HD, Kampman K, et al. Injectable, sustained-release naltrexone for the treatment of opioid dependence: A randomized, placebo-controlled trial. Archives of General Psychiatry. 2006;63(2):210–8. pmid:16461865
  51. 51. Cornish JW, Metzger D, Woody GE, Wilson D, McLellan AT, Vandergrift B, et al. Naltrexone pharmacotherapy for opioid dependent federal probationers. Journal of Substance Abuse Treatment. 1997;14(6):529–34. pmid:9437624
  52. 52. Coviello DM, Cornish JW, Lynch KG, Alterman AI, O’Brien CP. A randomized trial of oral naltrexone for treating opioid-dependent offenders. The American Journal on Addictions. 2010;19(5):422–32. pmid:20716305
  53. 53. Cropsey KL, Lane PS, Hale GJ, Jackson DO, Clark C, Ingersoll KS, et al. Results of a pilot randomized controlled trial of buprenorphine for opioid dependent women in the criminal justice system. Drug and Alcohol Dependence. 2011;119(3):172–8. pmid:21782352
  54. 54. D’Onofrio G, O’Connor PG, Pantalon MV, Chawarski MC, Busch SH, Owens PH, et al. Emergency department-initiated buprenorphine/naloxone treatment for opioid dependence: A randomized clinical trial. JAMA—Journal of the American Medical Association. 2015;313(16):1636–44. pmid:25919527
  55. 55. Dunlop AJ, Brown AL, Oldmeadow C, Harris A, Gill A, Sadler C, et al. Effectiveness and cost-effectiveness of unsupervised buprenorphine-naloxone for the treatment of heroin dependence in a randomized waitlist controlled trial. Drug and Alcohol Dependence. 2017;174:181–91. pmid:28371689
  56. 56. Eder H, Jagsch R, Kraigher D, Primorac A, Ebner N, Fischer G. Comparative study of the effectiveness of slow-release morphine and methadone for opioid maintenance therapy. Addiction. 2005;100(8):1101–9. pmid:16042640
  57. 57. Fischer G, Gombas W, Eder H, Jagsch R, Peternell A, Stuhlinger G, et al. Buprenorphine versus methadone maintenance for the treatment of opioid dependence. Addiction. 1999;94(9):1337–47. pmid:10615719
  58. 58. Gruber VA, Delucchi KL, Kielstein A, Batki SL. A randomized trial of 6-month methadone maintenance with standard or minimal counseling versus 21-day methadone detoxification. Drug and Alcohol Dependence. 2008;94(1–3):199–206. pmid:18243585
  59. 59. Guo S, Jiang Z, Wu Y. Efficacy of naltrexone hydrochloride for preventing relapse among opiate-dependent patients after detoxification. Hong Kong Journal of Psychiatry. 2001;11(4):2–8+28.
  60. 60. Haight BR, Learned SM, Laffont CM, Fudala PJ, Zhao Y, Garofalo AS, et al. Efficacy and safety of a monthly buprenorphine depot injection for opioid use disorder: a multicentre, randomised, double-blind, placebo-controlled, phase 3 trial. The Lancet. 2019;393(10173):778–90. pmid:30792007
  61. 61. Hollister LE, Bearman JE, Duster TS. Clinical evaluation of naltrexone treatment of opiate-dependent individuals. Archives of General Psychiatry. 1978;35(3):335–40. pmid:365122
  62. 62. Jarvis BP, Holtyn AF, DeFulio A, Koffarnus MN, Leoutsakos JMS, Umbricht A, et al. The effects of extended-release injectable naltrexone and incentives for opiate abstinence in heroin-dependent adults in a model therapeutic workplace: A randomized trial. Drug and Alcohol Dependence. 2019;197:220–7. pmid:30852374
  63. 63. Johnson RE, Chutuape MA, Strain EC, Walsh SL, Stitzer ML, Bigelow GE. A comparison of levomethadyl acetate, buprenorphine, and methadone for opioid dependence. New England Journal of Medicine. 2000;343(18):1290–7.
  64. 64. Johnson RE, Eissenberg T, Stitzer ML, Strain EC, Liebson IA, Bigelow GE. A placebo controlled clinical trial of buprenorphine as a treatment for opioid dependence. Drug and Alcohol Dependence. 1995;40(1):17–25. pmid:8746920
  65. 65. Johnson RE, Jaffe JH, Fudala PJ. A controlled trial of buprenorphine treatment for opioid dependence. Journal of the American Medical Association. 1992;267(20):2750–5. pmid:1578593
  66. 66. Kakko J, Dybrandt Svanborg K, Kreek MJ, Heilig M. 1-year retention and social function after buprenorphine-assisted relapse prevention treatment for heroin dependence in Sweden: A randomised, placebo-controlled trial. Lancet. 2003;361(9358):662–8. pmid:12606177
  67. 67. Kakko J, Gronbladh L, Svanborg KD, Von Wachenfeldt J, Ruck C, Rawlings B, et al. A stepped care strategy using buprenorphine and methadone versus conventional methadone maintenance in heroin dependence: A randomized controlled trial. American Journal of Psychiatry. 2007;164(5):797–803. pmid:17475739
  68. 68. Kamien JB, Branstetter SA, Amass L. Buprenorphine-naloxone versus methadone maintenance therapy: A randomised double-blind trial with opioid-dependent patients. Heroin Addiction and Related Clinical Problems. 2008;10(4):5–18.
  69. 69. Kinlock TW, Gordon MS, Schwartz RP, Fitzgerald TT, O’Grady KE. A randomized clinical trial of methadone maintenance for prisoners: Results at 12 months postrelease. Journal of Substance Abuse Treatment. 2009;37(3):277–85. pmid:19339140
  70. 70. Korthuis PT, King C, Cook RR, Khuyen TT, Kunkel LE, Bart G, et al. HIV clinic-based buprenorphine plus naloxone versus referral for methadone maintenance therapy for treatment of opioid use disorder in HIV clinics in Vietnam (BRAVO): an open-label, randomised, non-inferiority trial. The Lancet HIV. 2021;8(2):e67–e76. pmid:33539760
  71. 71. Kosten TR, Schottenfeld R, Ziedonis D, Falcioni J. Buprenorphine versus methadone maintenance for opioid dependence. Journal of Nervous and Mental Disease. 1993;181(6):358–64. pmid:8501457
  72. 72. Kristensen Ø, Espegren O, Asland R, Jakobsen E, Lie Ø, Seiler S. Buprenorphine and methadone to opiate addicts—a randomized trial. Tidsskrift for den Norske laegeforening. 2005;125(2):148–51. pmid:15665884
  73. 73. Krook AL, Brors O, Dahlberg J, Grouff K, Magnus P, Roysamb E, et al. A placebo-controlled study of high dose buprenorphine in opiate dependents waiting for medication-assisted rehabilitation in Oslo, Norway. Addiction. 2002;97(5):533–42. pmid:12033654
  74. 74. Krupitsky E, Nunes EV, Ling W, Illeperuma A, Gastfriend DR, Silverman BL. Injectable extended-release naltrexone for opioid dependence: A double-blind, placebo-controlled, multicentre randomised trial. The Lancet. 2011;377(9776):1506–13. pmid:21529928
  75. 75. Krupitsky E, Zvartau E, Blokhina E, Verbitskaya E, Tsoy M, Wahlgren V, et al. Naltrexone with or without guanfacine for preventing relapse to opiate addiction in St.-Petersburg, Russia. Drug and Alcohol Dependence. 2013;132(3):674–80. pmid:23683793
  76. 76. Krupitsky E, Zvartau E, Blokhina E, Verbitskaya E, Wahlgren V, Tsoy-Podosenin M, et al. Randomized trial of long-acting sustained-release naltrexone implant vs oral naltrexone or placebo for preventing relapse to opioid dependence. Archives of General Psychiatry. 2012;69(9):973–81. pmid:22945623
  77. 77. Krupitsky EM, Zvartau EE, Masalov DV, Tsoi MV, Burakov AM, Egorova VY, et al. Naltrexone for heroin dependence treatment in St. Petersburg, Russia. Journal of Substance Abuse Treatment. 2004;26(4):285–94. pmid:15182893
  78. 78. Lee JD, Friedmann PD, Kinlock TW, Nunes EV, Boney TY, Hoskinson RA, et al. Extended-release naltrexone to prevent opioid relapse in criminal justice offenders. New England Journal of Medicine. 2016;374(13):1232–42. pmid:27028913
  79. 79. Lee JD, McDonald R, Grossman E, McNeely J, Laska E, Rotrosen J, et al. Opioid treatment at release from jail using extended-release naltrexone: a pilot proof-of-concept randomized effectiveness trial. Addiction (Abingdon, England). 2015;110(6):1008–14. pmid:25703440
  80. 80. Lee JD, Nunes EV, Novo P, Bachrach K, Bailey GL, Bhatt S, et al. Comparative effectiveness of extended-release naltrexone versus buprenorphine-naloxone for opioid relapse prevention (X:BOT): a multicentre, open-label, randomised controlled trial. The Lancet. 2018;391(10118):309–18. pmid:29150198
  81. 81. Lerner A, Sigal M, Bacalu A, Shiff R, Burganski I, Gelkopf M. A Naltrexone double blind placebo controlled study in Israel. Israel Journal of Psychiatry and Related Sciences. 1992;29(1):36–43. pmid:1568861
  82. 82. Ling W, Casadonte P, Bigelow G, Kampman KM, Patkar A, Bailey GL, et al. Buprenorphine implants for treatment of opioid dependence: A randomized controlled trial. JAMA—Journal of the American Medical Association. 2010;304(14):1576–83. pmid:20940383
  83. 83. Ling W, Wesson DR, Charuvastra C, Klett CJ. A controlled trial comparing buprenorphine and methadone maintenance in opioid dependence. Archives of General Psychiatry. 1996;53(5):401–7. pmid:8624183
  84. 84. Madlung-Kratzer E, Spitzer B, Brosch R, Dunkel D, Haring C. A double-blind, randomized, parallel group study to compare the efficacy, safety and tolerability of slow-release oral morphine versus methadone in opioid-dependent in-patients willing to undergo detoxification. Addiction. 2009;104(9):1549–57. pmid:19686525
  85. 85. Magura S, Lee JD, Hershberger J, Joseph H, Marsch L, Shropshire C, et al. Buprenorphine and methadone maintenance in jail and post-release: A randomized clinical trial. Drug and Alcohol Dependence. 2009;99(1–3):222–30. pmid:18930603
  86. 86. Mattick RP, Ali R, White JM, O’Brien S, Wolk S, Danz C. Buprenorphine versus methadone maintenance therapy: A randomized double-blind trial with 405 opioid-dependent patients. Addiction. 2003;98(4):441–52. pmid:12653814
  87. 87. Mokri A, Chawarski MC, Taherinakhost H, Schottenfeld RS. Medical treatments for opioid use disorder in Iran: a randomized, double-blind placebo-controlled comparison of buprenorphine/naloxone and naltrexone maintenance treatment. Addiction (Abingdon, England). 2016;111(5):874–82. pmid:26639678
  88. 88. Neri S, Bruno CM, Pulvirenti D, Malaguarnera M, Italiano C, Mauceri B, et al. Randomized clinical trial to compare the effects of methadone and buprenorphine on the immune system in drug abusers. Psychopharmacology. 2005;179(3):700–4. pmid:15806416
  89. 89. Neumann AM, Blondell RD, Hoopsick RA, Homish GG. Randomized clinical trial comparing buprenorphine/naloxone and methadone for the treatment of patients with failed back surgery syndrome and opioid addiction. Journal of Addictive Diseases. 2020;38(1):33–41. pmid:31774028
  90. 90. Neumann AM, Blondell RD, Jaanimagi U, Giambrone AK, Homish GG, Lozano JR, et al. A preliminary study comparing methadone and buprenorphine in patients with chronic pain and coexistent opioid addiction. Journal of Addictive Diseases. 2013;32(1):68–78. pmid:23480249
  91. 91. Newman RG, Whitehill WB. Double-blind comparison of methadone and placebo maintenance treatments of narcotic addicts in Hong Kong. Lancet (london, england). 1979;2(8141):485–8. pmid:90214
  92. 92. Oliveto AH, Feingold A, Schottenfeld R, Jatlow P, Kosten TR. Desipramine in opioid-dependent cocaine abusers maintained on buprenorphine vs methadone. Archives of General Psychiatry. 1999;56(9):812–20. pmid:12884887
  93. 93. Otiashvili D, Kirtadze I, O’Grady KE, Jones HE. Drug use and HIV risk outcomes in opioid-injecting men in the Republic of Georgia: Behavioral treatment+naltrexone compared to usual care. Drug and Alcohol Dependence. 2012;120(1–3):14–21. pmid:21742445
  94. 94. Otiashvili D, Piralishvili G, Sikharulidze Z, Kamkamidze G, Poole S, Woody GE. Methadone and buprenorphine-naloxone are effective in reducing illicit buprenorphine and other opioid use, and reducing HIV risk behavior—outcomes of a randomized trial. Drug and alcohol dependence. 2013;133(2):376–82. pmid:23916321
  95. 95. Pani PP, Maremmani I, Pirastu R, Tagliamonte A, Gessa GL. Buprenorphine: A controlled clinical trial in the treatment of opioid dependence. Drug and Alcohol Dependence. 2000;60(1):39–50. pmid:10821988
  96. 96. Petitjean S, Stohler R, Deglon JJ, Livoti S, Waldvogel D, Uehlinger C, et al. Double-blind randomized trial of buprenorphine and methadone in opiate dependence. Drug and Alcohol Dependence. 2001;62(1):97–104. pmid:11173173
  97. 97. Rosenthal RN, Ling W, Casadonte P, Vocci F, Bailey GL, Kampman K, et al. Buprenorphine implants for treatment of opioid dependence: randomized comparison to placebo and sublingual buprenorphine/naloxone. Addiction (Abingdon, England). 2013;108(12):2141–9. pmid:23919595
  98. 98. San L, Pomarol G, Peri JM, Olle JM, Cami J. Follow-up after a six-month maintenance period on naltrexone versus placebo in heroin addicts. British Journal of Addiction. 1991;86(8):983–90. pmid:1912753
  99. 99. Saxon AJ, Ling W, Hillhouse M, Thomas C, Hasson A, Ang A, et al. Buprenorphine/Naloxone and methadone effects on laboratory indices of liver health: A randomized trial. Drug and Alcohol Dependence. 2013;128(1–2):71–6. pmid:22921476
  100. 100. Schottenfeld RS, Chawarski MC, Mazlan M. Maintenance treatment with buprenorphine and naltrexone for heroin dependence in Malaysia: a randomised, double-blind, placebo-controlled trial. The Lancet. 2008;371(9631):2192–200. pmid:18586174
  101. 101. Schottenfeld RS, Chawarski MC, Pakes JR, Pantalon MV, Carroll KM, Kosten TR. Methadone versus buprenorphine with contingency management or performance feedback for cocaine and opioid dependence. American Journal of Psychiatry. 2005;162(2):340–9. pmid:15677600
  102. 102. Schottenfeld RS, Pakes JR, Oliveto A, Ziedonis D, Kosten TR. Buprenorphine vs methadone maintenance treatment for concurrent opioid dependence and cocaine abuse. Archives of General Psychiatry. 1997;54(8):713–20. pmid:9283506
  103. 103. Schwartz RP, Highfield DA, Jaffe JH, Brady JV, Butler CB, Rouse CO, et al. A randomized controlled trial of interim methadone maintenance. Archives of General Psychiatry. 2006;63(1):102–9. pmid:16389204
  104. 104. Seifert J, Metzner C, Paetzold W, Borsutzky M, Passie T, Rollnik J, et al. Detoxification of opiate addicts with multiple drug abuse: A comparison of buprenorphine vs. methadone. Pharmacopsychiatry. 2002;35(5):159–64. pmid:12237786
  105. 105. Shufman EN, Porat S, Witztum E, Gandacu D, Bar-Hamburger R, Ginath Y. The efficacy of Naltrexone in preventing reabuse of heroin after detoxification. Biological Psychiatry. 1994;35(12):935–45. pmid:8080893
  106. 106. Soyka M, Zingg C, Koller G, Kuefner H. Retention rate and substance use in methadone and buprenorphine maintenance therapy and predictors of outcome: Results from a randomized study. International Journal of Neuropsychopharmacology. 2008;11(5):641–53. pmid:18205978
  107. 107. Strain EC, Stitzer ML, Liebson IA, Bigelow GE. Buprenorphine versus methadone in the treatment of opioid-dependent cocaine users. Psychopharmacology. 1994;116(4):401–6. pmid:7701040
  108. 108. Strain EC, Stitzer ML, Liebson IA, Bigelow GE. Comparison of buprenorphine and methadone in the treatment of opioid dependence. American Journal of Psychiatry. 1994;151(7):1025–30. pmid:8010359
  109. 109. Sullivan MA, Bisaga A, Glass A, Mishlen K, Pavlicova M, Carpenter KM, et al. Opioid use and dropout in patients receiving oral naltrexone with or without single administration of injection naltrexone. Drug and Alcohol Dependence. 2015;147:122–9. pmid:25555621
  110. 110. Tanum L, Solli KK, Latif ZEH, Benth JS, Opheim A, Sharma-Haase K, et al. Effectiveness of injectable extended-release naltrexone vs daily buprenorphine-naloxone for opioid dependence: A randomized clinical noninferiority trial. JAMA Psychiatry. 2017;74(12):1197–205. pmid:29049469
  111. 111. Tiihonen J, Krupitsky E, Verbitskaya E, Blokhina E, Mamontova O, Fohr J, et al. Naltrexone implant for the treatment of polydrug dependence: A randomized controlled trial. American Journal of Psychiatry. 2012;169(5):531–6. pmid:22764364
  112. 112. Wang X, Jiang H, Zhao M, Li J, Gray F, Sheng L, et al. Treatment of opioid dependence with buprenorphine/naloxone sublingual tablets: a phase 3 randomized, double-blind, placebo-controlled trial. Asia-pacific psychiatry. 2018;(no pagination). pmid:30460781
  113. 113. Wright NM, Sheard L, Adams CE, Rushforth BJ, Harrison W, Bound N, et al. Comparison of methadone and buprenorphine for opiate detoxification (LEEDS trial): a randomised controlled trial. British journal of general practice. 2011;61(593):e772–80. pmid:22137413
  114. 114. Yancovitz SR, Des Jarlais DC, Peyser NP, Drew E, Friedmann P, Trigg HL, et al. A randomized trial of an interim methadone maintenance clinic. American Journal of Public Health. 1991;81(9):1185–91. pmid:1659236
  115. 115. Schottenfeld RS, Chawarski MC, Mazlan M. Behavioral counseling and abstinence-contingent take-home buprenorphine in general practitioners’ offices in Malaysia: a randomized, open-label clinical trial. Addiction. 2021;116(8):2135–49. pmid:33404150
  116. 116. British Columbia Centre on Substance Use. A Guideline for the Clinical Management of Opioid Use Disorder: British Columbia Ministry of Health; 2017 [Available from:
  117. 117. Jansen JP, Fleurence R, Devine B, Itzler R, Barrett A, Hawkins N, et al. Interpreting indirect treatment comparisons and network meta-analysis for health-care decision making: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 1. Value Health. 2011;14(4):417–28. pmid:21669366
  118. 118. Mitchell SG, Gryczynski J, Schwartz RP, Myers CP, O’Grady KE, Olsen YK, et al. Changes in Quality of Life following Buprenorphine Treatment: Relationship with Treatment Retention and Illicit Opioid Use. J Psychoactive Drugs. 2015;47(2):149–57. pmid:25950595
  119. 119. Fareed A, Casarella J, Amar R, Vayalapalli S, Drexler K. Benefits of retention in methadone maintenance and chronic medical conditions as risk factors for premature death among older heroin addicts. J Psychiatr Pract. 2009;15(3):227–34. pmid:19461397
  120. 120. Kimber J, Larney S, Hickman M, Randall D, Degenhardt L. Mortality risk of opioid substitution therapy with methadone versus buprenorphine: a retrospective cohort study. Lancet Psychiatry. 2015;2(10):901–8. pmid:26384619
  121. 121. Hickman M, Steer C, Tilling K, Lim AG, Marsden J, Millar T, et al. The impact of buprenorphine and methadone on mortality: a primary care cohort study in the United Kingdom. Addiction. 2018;113(8):1461–76. pmid:29672985
  122. 122. Larochelle MR, Bernson D, Land T, Stopka TJ, Wang N, Xuan Z, et al. Medication for Opioid Use Disorder After Nonfatal Opioid Overdose and Association With Mortality: A Cohort Study. Ann Intern Med. 2018;169(3):137–45. pmid:29913516
  123. 123. Amato L, Davoli M, Perucci CA, Ferri M, Faggiano F, Mattick RP. An overview of systematic reviews of the effectiveness of opiate maintenance therapies: available evidence to inform clinical practice and research. J Subst Abuse Treat. 2005;28(4):321–9. pmid:15925266