Slow‐release oral morphine as maintenance therapy for opioid dependence

Types of studies

Randomised controlled studies (RCTs) and controlled clinical trials (CCTs).

Types of participants

Adults (aged ≥ 18 years) dependent on heroin according to Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM‐IV) (or International Classification of Diseases (ICD)‐10) criteria seeking for maintenance treatment. Pregnant women were excluded.

Types of interventions

Types of outcome measures

Electronic searches

The following electronic databases were searched for relevant trials:

1. Cochrane Drugs and Alcohol Group's Register of Trials.

2. The Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 3, 2013).

3. PubMed (January 1966 to April 2013).

4. EMBASE (January 1974 to April 2013).

Databases were searched using MeSH terms and free‐text terms relating to opioid dependence and SROM as shown in Appendix 1. For the MEDLINE search, the Cochrane Highly Sensitive Search Strategy (sensitivity maximising version) was used to filter for RCTs (Higgins 2011). This strategy was revised appropriately for each database to take into account the differences in controlled vocabulary and syntax rules.

We also searched the following main electronic sources of ongoing trials:

• ClinicalTrials.gov (www.clinicaltrials.gov);

• World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (apps.who.int/trialsearch/);

• Current Controlled Trials (www.controlled‐trials.com/);

• EU Clinical Trials Register (www.clinicaltrialsregister.eu);

• Osservatorio Nazionale sulla Sperimentazione Clinica dei Medicinali (oss‐sper‐clin.agenziafarmaco.it);

• Trials (www.trialsjournal.com).

Searching other resources

• Reference lists of all relevant papers to identify further studies.

• Conference proceedings likely to contain trials relevant to the review (e.g. the College on Problems of Drug Dependence).

• We contacted investigators seeking information about unpublished or incomplete trials.

All searches included non‐English language literature and studies with English abstracts were assessed for inclusion. When considered likely to meet inclusion criteria, studies were translated.

Selection of studies

Two review authors (MF, SM) independently screened titles and abstracts of studies obtained by the search strategy. Each potentially relevant study located in the search was obtained in full text and assessed for inclusion independently by two review authors. In case of disagreement, a third review author (MD) was consulted.

Data extraction and management

Data were extracted independently by two review authors (MF, SM). Any disagreements were discussed and solved by consensus.

Assessment of risk of bias in included studies

The risk of bias assessment for RCTs and CCTs in this review was performed using the criteria recommended by the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The recommended approach for assessing risk of bias is a two‐part tool, addressing seven specific domains, namely sequence generation and allocation concealment (selection bias), blinding of participants and providers (performance bias), blinding of outcome assessor (detection bias), incomplete outcome data (attrition bias), selective outcome reporting (reporting bias) and other source of bias. The first part of the tool involves describing what was reported to have happened in the study. The second part of the tool involves assigning a judgement relating to the risk of bias for that entry, in terms of low, high or unclear risk. To make these judgements, we used the criteria indicated by the Cochrane Handbook for Systematic Reviews of Interventions adapted to the addiction field. See Table 1 for details.

The domains of sequence generation and allocation concealment (avoidance of selection bias) were addressed in the tool by a single entry for each study.

Blinding of participants, personnel and outcome assessor (avoidance of performance bias and detection bias) were considered separately for objective outcomes (e.g. dropout, use of substance of abuse measured by urine analysis, subjects relapsed at the end of follow‐up, subjects engaged in further treatments) and subjective outcomes (e.g. duration and severity of signs and symptoms of withdrawal, participant self reported use of substance, side effects, social functioning as integration at school or at work, family relationship).

Incomplete outcome data (avoidance of attrition bias) were considered for all outcomes except for the dropout from the treatment, which is very often the primary outcome measure in trials on addiction.

Two review authors (MF, SM) independently applied the 'Risk of bias' tool to included studies.

Measures of treatment effect

Dichotomous outcomes were analysed calculating the risk ratio (RR) for each trial with the uncertainty in each result being expressed by their 95% confidence intervals (CIs). Continuous outcomes were analysed calculating the mean difference (MD) or the standardised mean difference (SMD) with 95% CIs. For outcomes assessed by scales we compared and pooled the mean score differences from the end of treatment to baseline (post minus pre) in the experimental and control group. In case of missing data about the standard deviation of the change, we imputed this measure using the standard deviation at the end of treatment for each group.

Unit of analysis issues

We did not used data presented as number of positive urine tests over total number of tests in the experimental and control group as a measure of substance abuse. This is because using the number of tests instead of the number of participants as the unit of analysis violates the hypothesis of independence among observations. In fact, the results of tests done in each participant are not independent.

If all arms in a multi‐arm trial were included in the meta‐analysis and one treatment arm was to be included in more than one of the treatment comparisons, we divided the number of events and the number of participants in that arm by the number of treatment comparisons made. This method avoids the multiple use of participants in the pooled estimate of treatment effect while retaining information from each arm of the trial. It compromises the precision of the pooled estimate slightly.

Dealing with missing data

Study authors were contacted to request any data missing from included studies.

Assessment of heterogeneity

Statistically significant heterogeneity among primary outcome studies was assessed using the Chi² test and I² statistic (Higgins 2011). A significant Chi² (P value < 0.10) and I² statistic of at least 50% were considered as statistical heterogeneity.

Assessment of reporting biases

Funnel plots (plot of the effect estimate from each study against the sample size or effect standard error) were planned to be used to assess the potential for bias related to the size of the trials, which could indicate possible publication bias. Only four studies were retrieved so the funnel plot could not be used.

In order to grade the quality of the evidence, the Grading of Recommendation, Assessment, Development, and Evaluation Working Group (GRADE) (a system for grading the quality of evidence (GRADE Working Group 2004; Guyatt 2008; Guyatt 2010; Schünemann 2006) that takes into account issues not only related to internal validity but also to external validity such as directness of results) was planned to be used. The overall quality of the evidence for each outcome of this review was assessed using the GRADE system. The 'Summary of findings' tables present the main findings of a review in a transparent and simple tabular format. In particular, they provide key information concerning the quality of evidence, the magnitude of effect of the interventions examined and the sum of available data on the main outcomes.

The GRADE system uses the following criteria for assigning grade of evidence:

• high = further research is very unlikely to change our confidence in the estimate of effect;

• moderate = further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate;

• low = further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate;

• very low = any estimate of effect is very uncertain.

Decrease grade if:

• serious (‐1) or very serious (‐2) limitation to study quality;

• important inconsistency (‐1);

• some (‐1) or major (‐2) uncertainty about directness;

• imprecise or sparse data (‐1);

• high probability of reporting bias (‐1).

Increase grade if:

• strong evidence of association:‐ significant RR of greater than 2 (< 0.5) based on consistent evidence from two or more observational studies, with no plausible confounders (+1);

• very strong evidence of association: significant RR of greater than 5 (< 0.2) based on direct evidence with no major threats to validity (+2);

• evidence of a dose‐response gradient (+1);

• all plausible confounders would have reduced the effect (+1).

Because of the paucity of retrieved studies and the fact that results were not pooled, 'Summary of finding' table was not done.

Data synthesis

Given the expected heterogeneity of the results among studies due to differences in the populations, types of control interventions, and setting dosages and duration of treatment intervention, the outcome measures from the individual trials were planned to be combined through meta‐analysis when possible (clinical comparability of intervention and outcomes between trials) using a random‐effects model. Only four studies with different outcomes were found so meta‐analyses were not judged to be appropriate.

Sensitivity analysis

To incorporate assessment in the review process we planned to first plot intervention effects estimates stratified for risk of bias for each relevant domain. If differences in results among studies were identified at different risk of bias, we planned to perform sensitivity analysis excluding from the analysis studies with high risk of bias. We also planned to perform subgroup analysis for studies with low and unclear risk of bias. Meta‐analyses were not performed for the reason explained above, and consequently the planned sensitivity analysis was not possible.

Results of the search

After removing duplicates, the electronic search identified 1702 unique records. Through the screening of titles and abstracts, 1692 records were excluded as obviously irrelevant. The full‐text reports of the remaining 10 records were examined. Two further titles (Fischer 1996; Sherman 1996) were found on already published reviews and full text of these articles were searched but not yet found, we classified them as awaiting assessment. Twelve titles relating to eight original studies were assessed for inclusion. Five studies were excluded after full‐text examination. Three studies (five articles) were finally included. The study flow diagram of the search is shown in Figure 1.

Included studies

Three studies were included with 195 people (Clark 2002; Eder 2005; Giacomuzzi 2006).

For one study, only a conference abstract was available (Clark 2002). The study was an open‐label cross‐over study enrolling 11 people administered methadone or SROM for six week each.

Giacomuzzi 2006 compared physical symptoms, quality of life and urine analysis of participants at admission for maintenance treatment with people already in treatment. The three treatments compared were: oral methadone, sublingual buprenorphine and SROM given for six months. A total of 120 people were randomised to the three treatments.

Eder 2005 was a double‐blind, double‐dummy cross‐over trial comparing SROM with methadone for seven weeks in 64 people.

Two studies were conducted in Austria (Eder 2005; Giacomuzzi 2006) and the third in Australia (Clark 2002).

Excluded studies

Five studies were excluded after full‐text examination. Reasons for exclusions were: the type of intervention not in the inclusion criteria (i.e. SROM used for detoxification rather than maintenance treatment) (one study) (Madlung‐Kratzer 2009); study design not in the inclusion criteria, as four studies were non‐randomised or uncontrolled studies (Mitchell 2003; Mitchell 2004; Mitchell 2006; Moldovanyi 1996).

Allocation

Only one study was judged as a low risk of bias for selection bias (adequate random sequence generation and allocation concealment) (Eder 2005). The other two studies did not provide a description for this item and were judged to be at unclear risk of bias.

Blinding

Only one study was double‐blinded, double‐dummy (Eder 2005). The other two were open label (Clark 2002; Giacomuzzi 2006)

Incomplete outcome data

All the studies were judged at low risk of bias for this outcome.

Selective reporting

All the studies were judged at low risk of bias for this outcome.

Retention in treatment

In Clark 2002, nine participants out of 11 (81.8%) completed the study. It was not reported whether dropout occurred during the methadone or the SROM phase.

In Eder 2005, five out of 32 participants dropped out during the SROM phase and four out of 32 during the methadone phase (14%).

In Giacomuzzi 2006, no information was reported on dropout but it seems that all participants remained in treatment until the end of the study.

Use of primary substance of abuse

In Clark 2002, it was reported that among the participants completing the study, there were no clinically or statistically significant differences in heroin use.

In Eder 2005, six participants (10%) during treatment with SROM and 11 participants (19%) during treatment with methadone had evidence of
 fresh needle marks to indicate concomitant intravenous drug abuse.

In Giacomuzzi 2006, no data were reported on use of heroin. The only available information was that "post hoc paired group comparisons showed statistically significant differences between clients at admission and SROM maintenance treatment in opioid consumption p<0.001) in which participants in the SROM group showed the more favourable values".

Use of other substances

In Clark 2002, it was reported that of those who completed, there were no clinically or statistically significant differences in other drug use.

In Eder 2005, cocaine usage increased significantly over time in both treatment groups (P value < 0.001). Benzodiazepine urinalysis results remained stable throughout the study. Concomitant amphetamine consumption never exceeded 2%. There was no significant difference between treatment groups or medications (data shown only in figure).

In Giacomuzzi 2006, the only available information was that "post hoc paired group comparisons showed statistically significant differences between clients at admission and SROM maintenance treatment in cocaine consumption p<0.001) in which participants in the SROM group showed the more favourable values".

Social functioning and participant satisfaction

In Clark 2002, it was reported that among completing participants, there were no clinically or statistically significant differences in severity of dependence or mental health/social functioning. There was a trend for less severe opiate withdrawal symptoms between doses among people taking morphine than those taking methadone (withdrawal score 2.2 vs. 4.8, P value = 0.06). Morphine was generally well tolerated and was preferred by seven out of nine subjects. No participants complained that morphine did not "hold" for the 24‐hour dosing interval.

In Eder 2005, a statistically significant interaction between time and group was apparent for heroin (P value = 0.037) and alcohol (P value = 0.013) craving with craving reduced more during SROM treatment (data shown only in figure). A decrease in depressive symptoms (measured by Beck Depression Inventory (BDI)) and physical
 complaints (measured using the Beschwerde‐Liste (BL)) was achieved at the time of cross‐over with a significant difference (P value < 0.001) in the second period in favour of SROM and a significant increase for people treated with methadone in the second study period. Similarly, there was a statistically significant time effect for anxiety scores (State‐Trait Anxiety Inventory (STAI)) (P value = 0.008) and a statistically significant interaction for time and group (P value = 0.003) in favour of morphine (data shown only in figure). For quality of life at cross‐over (week 7) and at the end of the study phase (week 14), no significant differences between groups were found in any of the six QoL domains.

In Giacomuzzi 2006, people on SROM generally showed less favourable values in QoL score than people on methadone or sublingual buprenorphine. Methadone and buprenorphine maintained group showed significantly more favourable values for finances (4.4 and 4.2 vs. 2.6, P value < 0.010), family (5.8 and 5.1 vs. 3.6, P value < 0.044), and overall satisfaction (5.3 and 4.9 vs. 4.1, P value < 0.031) compared to people maintained on SROM. The methadone group showed more favourable value for leisure time (5.4 vs. 3.7, P value < 0.001), housing (6.1 vs. 4.7, P value < 0.023), partnership (5.7 vs. 4.2 , P value = 0.034), friends and acquaintances (5.6 vs. 4.4, P value = 0.003), mental health (5.0 vs. 3.4, P value = 0.002) and self esteem (8.2 vs. 5.7, P value = 0.002) compared to people on SROM. A statistically significant difference between buprenorphine and SROM was found for physical health in favour of buprenorphine (4.8 vs. 3.3, P value = 0.043).

Medical adverse events

In Eder 2005, at least one side effect was reported by 82% of participants receiving morphine and 76% of those receiving methadone. The most commonly reported adverse event for morphine capsules was toothache (26%), followed by headache (23%), constipation (11%) and influenza (11%). The most commonly reported adverse events for methadone solution were: toothache (22%), vomiting (17%), headache (14%) and stomach ache (12%). Sleep disturbance and insomnia were commonly reported with methadone (15% of people) but not with morphine. None of these adverse events were rated as serious or considered to be treatment‐related. One person receiving SROM was withdrawn from the study because of aggravation of depression and required hospital treatment a few days after discontinuing the study.

In Giacomuzzi 2006, it was reported that buprenorphine‐ and methadone‐treated participants showed significantly more favourable values than the group maintained on SROM for stomach cramps (17% and 13% vs. 47%, P value < 0.013), fatigue or tiredness (50% and 30% vs. 80%, P value < 0.015), general aches and pain (13% and 20% vs. 33%, P value < 0.001) and insomnia (40% and 40% vs. 68%, P value < 0.023). People treated with methadone had fewer problems in falling asleep compared with people treated with SROM (36% vs. 70%, P value = 0.009) and people treated with sublingual buprenorphine experienced less depression than people treated with SROM (37% vs. 68%, P value = 0.023).