Maintenance agonist treatments for opiate-dependent pregnant women

  • Conclusions changed
  • Review
  • Intervention

Authors


Abstract

Background

The prevalence of opiate use among pregnant women can range from 1% to 2% to as high as 21%. Heroin crosses the placenta and pregnant, opiate-dependent women experience a six-fold increase in maternal obstetric complications such as low birth weight, toxaemia, third trimester bleeding, malpresentation, puerperal morbidity, fetal distress and meconium aspiration. Neonatal complications include narcotic withdrawal, postnatal growth deficiency, microcephaly, neuro-behavioural problems, increased neonatal mortality and a 74-fold increase in sudden infant death syndrome.

Objectives

To assess the effectiveness of any maintenance treatment alone or in combination with psychosocial intervention compared to no intervention, other pharmacological intervention or psychosocial interventions for child health status, neonatal mortality, retaining pregnant women in treatment and reducing the use of substances.

Search methods

We searched the Cochrane Drugs and Alcohol Group Trials Register (September 2013), PubMed (1966 to September 2013), CINAHL (1982 to September 2013), reference lists of relevant papers, sources of ongoing trials, conference proceedings and national focal points for drug research. We contacted authors of included studies and experts in the field.

Selection criteria

Randomised controlled trials assessing the efficacy of any maintenance pharmacological treatment for opiate-dependent pregnant women.

Data collection and analysis

We used the standard methodological procedures expected by The Cochrane Collaboration.

Main results

We found four trials with 271 pregnant women. Three compared methadone with buprenorphine and one methadone with oral slow-release morphine. Three out of four studies had adequate allocation concealment and were double-blind. The major flaw in the included studies was attrition bias: three out of four had a high drop-out rate (30% to 40%) and this was unbalanced between groups.

Methadone versus buprenorphine: the drop-out rate from treatment was lower in the methadone group (risk ratio (RR) 0.64, 95% confidence interval (CI) 0.41 to 1.01, three studies, 223 participants). There was no statistically significant difference in the use of primary substance between methadone and buprenorphine (RR 1.81, 95% CI 0.70 to 4.69, two studies, 151 participants). For both, we judged the quality of evidence as low. Birth weight was higher in the buprenorphine group in the two trials that could be pooled (mean difference (MD) -365.45 g (95% CI -673.84 to -57.07), two studies, 150 participants). The third study reported that there was no statistically significant difference. For APGAR score neither of the studies which compared methadone with buprenorphine found a significant difference. For both, we judged the quality of evidence as low. Many measures were used in the studies to assess neonatal abstinence syndrome. The number of newborns treated for neonatal abstinence syndrome, which is the most critical outcome, did not differ significantly between groups. We judged the quality of evidence as very low.

Methadone versus slow-release morphine: there was no drop-out in either treatment group. Oral slow-release morphine seemed superior to methadone for abstinence from heroin use during pregnancy (RR 2.40, 95% CI 1.00 to 5.77, one study, 48 participants). We judged the quality of evidence as moderate.

Only one study which compared methadone with buprenorphine reported side effects. For the mother there was no statistically significant difference; for the newborns in the buprenorphine group there were significantly fewer serious side effects.

In the comparison between methadone and slow-release morphine no side effects were reported for the mother, whereas one child in the methadone group had central apnoea and one child in the morphine group had obstructive apnoea.

Authors' conclusions

We did not find sufficient significant differences between methadone and buprenorphine or slow-release morphineto allow us to conclude that one treatment is superior to another for all relevant outcomes. While methadone seems superior in terms of retaining patients in treatment, buprenorphine seems to lead to less severe neonatal abstinence syndrome. Additionally, even though a multi-centre, international trial with 175 pregnant women has recently been completed and its results published and included in this review, the body of evidence is still too small to draw firm conclusions about the equivalence of the treatments compared. There is still a need for randomised controlled trials of adequate sample size comparing different maintenance treatments.

Plain language summary

Maintenance treatments for opiate-dependent pregnant women

Some women continue to use opiates when they are pregnant, yet heroin readily crosses the placenta. Opiate-dependent women experience a six-fold increase in maternal obstetric complications and give birth to low-weight babies. The newborn may experience narcotic withdrawal (neonatal abstinence syndrome) and have development problems. There is also increased neonatal mortality and a 74-fold increase in the risk of sudden infant death syndrome. Maintenance treatment with methadone provides a steady concentration of opiate in the pregnant woman's blood and so prevents the adverse effects on the fetus of repeated withdrawals. Buprenorphine is also used. These treatments reduce illicit drug use, improve compliance with obstetric care and improve neonatal birth weight but they are still associated with neonatal abstinence syndrome.

This review found few differences in newborn or maternal outcomes for pregnant, opiate-addicted women who were maintained on methadone, buprenorphine or oral slow-release morphine from a mean gestational age of 23 weeks to delivery. Only four randomised controlled trials with 271 participants trials satisfied the inclusion criteria for the review: two from Austria (outpatients), one from the USA (inpatients) and the fourth a multi-centre, international study conducted in Austria, Canada and the USA. The trials continued for 15 to 18 weeks. Three compared methadone with buprenorphine (223 participants) and one compared methadone with oral slow-release morphine (48 participants). The number of women who dropped out from treatment was lower in the methadone group. However, there was no difference in the use of primary substance between the methadone and buprenorphine groups. The number of newborns treated for neonatal abstinence syndrome did not differ significantly between groups. Birth weight was higher in the buprenorphine group in two trials and no different in the third. Oral slow-release morphine seemed superior to methadone in terms of the number of women who used heroin in their third trimester. However, there was no clear improvement in infant birth weight or duration of neonatal abstinence syndrome. The number of participants in the trials was small and may not be sufficient to draw firm conclusions. All the included studies ended immediately after the baby was born. No severe complications were noted.

Summary of findings(Explanation)

Summary of findings for the main comparison. Methadone compared to buprenorphine for opiate-dependent pregnant women
  1. 1For incomplete outcome data, we judged the studies at high risk of attrition bias because the attrition rate was high and unbalanced between groups.
    2Small sample size.
    3Statistically significant heterogeneity.
    4No explanation was provided.
    5Variability in results.

Methadone compared to buprenorphine for opiate-dependent pregnant women
Patient or population: opiate-dependent pregnant women
Settings:
Intervention: methadone
Comparison: buprenorphine
OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of participants
(studies)
Quality of the evidence
(GRADE)
Comments
Assumed riskCorresponding risk
Buprenorphine Methadone
Drop-out
Objective
Follow-up: 15 to 18 weeks
Study population RR 0.64
(0.41 to 1.01)
223
(3 studies)
⊕⊕⊝⊝
low 1,2
 
318 per 1000 204 per 1000
(130 to 321)
Moderate
326 per 1000 209 per 1000
(134 to 329)
Use of primary substance
Objective
Follow-up: 15 to 18 weeks
Study population RR 1.81
(0.7 to 4.69)
151
(2 studies)
⊕⊕⊝⊝
low 1,2
 
75 per 1000 135 per 1000
(52 to 350)
Moderate
43 per 1000 78 per 1000
(30 to 202)
Birth weight
Objective
Follow-up: mean 18 weeks
The mean birth weight ranged across control groups from
3.53 to 3.09 grams
The mean birth weight in the intervention groups was
224.91 lower
(248.46 to 201.36 lower)
 150
(2 studies)
⊕⊕⊝⊝
low 1,2,3,4
 
APGAR score
Objective. Scale from: 0 to 10
Follow-up: mean 18 weeks
The mean APGAR score ranged across control groups from
8.9 to 9.0
The mean APGAR score in the intervention groups was
0 higher
(0.03 lower to 0.03 higher)
 163
(2 studies)
⊕⊕⊝⊝
low 1,2
 
Number treated for NAS
Objective
Follow-up: 15 to 18 weeks
Study population RR 1.22
(0.89 to 1.67)
166
(3 studies)
⊕⊝⊝⊝
very low 1,2,5
 
447 per 1000 546 per 1000
(398 to 747)
Moderate
466 per 1000 569 per 1000
(415 to 778)
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
APGAR score: Activity, Pulse, Grimace, Appearance and Respiration score; CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: 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 quality: We are very uncertain about the estimate.

Summary of findings 2 Methadone compared to oral slow-release morphine for opiate-dependent pregnant women

Summary of findings 2. Methadone compared to oral slow-release morphine for opiate-dependent pregnant women
  1. 1Only one study with small sample size.

Methadone compared to oral slow-release morphine for opiate-dependent pregnant women
Patient or population: opiate-dependent pregnant women
Settings:
Intervention: methadone
Comparison: oral slow-release morphine
OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of participants
(studies)
Quality of the evidence
(GRADE)
Comments
Assumed riskCorresponding risk
Oral slow-release morphine Methadone
Use of substance
Objective
Follow-up: 15 weeks
Study population RR 2.4
(1 to 5.77)
48
(1 study)
⊕⊕⊕⊝
moderate 1
 
208 per 1000 500 per 1000
(208 to 1000)
Moderate
208 per 1000 499 per 1000
(208 to 1000)
Birth weight
Objective
Follow-up: 15 weeks
The mean birth weight in the control groups was
2.912 grams
The mean birth weight in the intervention groups was
124 higher
(186.94 lower to 434.94 higher)
 48
(1 study)
⊕⊕⊕⊝
moderate 1
 
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: 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 quality: We are very uncertain about the estimate.

Background

Description of the condition

In Australia , a national survey conducted in 2010 found that, less than one in 20 women (4.2%) who were pregnant and/or breastfeeding in the past 12 months used any illicit drug while they were pregnant. This included those who used illicit drugs in the time before they knew they were pregnant (AIHW 2011). The estimated prevalence of opiate use among pregnant women ranges from 1% to 2% with a peak of 21% (Brown 1998). The availability of more recent data on prevalence is extremely limited(sources searched: European Monitoring Centre for Drug & Drug Addiction (EMCDDA 2013), Office of National Drug Control Policy (ONDCP 2013) (USA), United Nations Office on Drug and Crime - World Drug Report 2007 (UNODC 2007) and Monitoring the Future 2012). Details of opiate use by subgroups of the population are not reported. The most recent data are from the 2010 National Survey on Drug Use and Health (NSDUH 2010), reporting that in the USA 5% of pregnant women (aged 15 to 44) reported the use of illicit drugs (not specifying the type of drug). The rate of current illicit drug use in the combined 2010-2011 data was 20.9% among pregnant women aged 15 to 17, 8.2% among pregnant women aged 18 to 25, and 2.2% among pregnant women aged 26 to 44. None of these rates were significantly different from those in the combined 2008-2009 data (15.8% among pregnant women aged 15 to 17, 7.1% among pregnant women aged 18 to 25 and 2.3% among pregnant women aged 26 to 44) (SAMHA 2012). In Australia, a national survey conducted in 2010 found that fewer than one in 20 women (4.2%) who were pregnant or breastfeeding (or both) in the past 12 months used any illicit drug while they were pregnant. This included those who used illicit drugs in the time before they knew they were pregnant (AIHW 2011).

Heroin readily crosses the placenta and untreated opiate dependence in pregnant women is associated with many environmental and medical factors that contribute to poor maternal and child outcomes and it causes a six-fold increase in obstetric complications, such as low birth weight, toxaemia, third trimester bleeding, malpresentation, puerperal morbidity, fetal distress and meconium aspiration. Neonatal complications include narcotic withdrawal, postnatal growth deficiency, microcephaly, neurobehavioural problems, increased neonatal mortality and a 74-fold increase in sudden infant death syndrome (Dattel 1990; Fajemirokun 2006; Ludlow 2004).

All of the commonly used opioids, including heroin and methadone, can produce neonatal abstinence syndrome in infants born to opiate-dependent mothers. Neonatal abstinence syndrome combines all the symptoms of adult withdrawal syndrome with irritability, poorly co-ordinated sucking and, in the most severe cases, seizures and death (Kaltenbach 1998).

Description of the intervention

Since the early 1970s, treatment with methadone has been the standard of care for pregnant women addicted to opiates. Despite its ability to induce fetal dependence and withdrawal, maintenance treatment provides a steady concentration in the maternal blood plasma and thus prevents the adverse effects of repeated withdrawal on the fetus (Jarvis 1994).

Buprenorphine has also been administered to opioid-dependent pregnant women as a maintenance treatment. Placental transfer of buprenorphine may be less than with methadone, reducing fetal exposure and development of neonatal abstinence syndrome (Rayburn 2004).

How the intervention might work

Methadone maintenance treatment given during pregnancy reduces maternal illicit opiate use and fetal exposure, enhances compliance with obstetric care and is associated with improved neonatal outcomes such as heavier birth weight (Fajemirokun 2006; Kaltenbach 1998). Additional benefits include a potential reduction in drug-seeking behaviours, including commercial sex to gain money for drugs. This reduction may decrease a woman's chance of acquiring sexually transmitted diseases such as human immunodeficiency virus (HIV) and hepatitis. For all these reasons, methadone treatment has become the 'gold standard' for the management of pregnant heroin users (NIH 1998) and many guidelines, in the UK (UK Guidelines 2007), USA (CSAT 2005) and Australia (Dunlop 2003), support the use of methadone during pregnancy.

Previous studies have been performed in centres which offered methadone and comprehensive services, including obstetric, health, psychiatric care, individual, group and family therapy. Consequently, it is difficult to evaluate the results of these studies and to distinguish the benefits of methadone itself from other measures of psychosocial and obstetric care (Wang 1999).

The available clinical literature suggests that buprenorphine maintenance is also associated with reduced maternal illicit opiate use and fetal exposure, enhanced compliance with obstetric care and improved neonatal outcomes such as heavier birth weight (Johnson 2003; Lejuene 2006).

Why it is important to do this review

The Cochrane Drugs and Alcohol Group has conducted several systematic reviews on maintenance treatment: methadone (Faggiano 2003; Mattick 2009), heroin (Ferri 2011), levomethadyl acetate hydrochloride (LAAM) (Clark 2002), buprenorphine (Mattick 2008), naltrexone (Minozzi 2011), psychosocial treatment alone (Mayet 2004) and psychosocial treatment combined with maintenance treatment (Amato 2011), however none of these reviews include studies on pregnant, opiate-dependent women.

Two narrative reviews have discussed the risks and benefits of maintenance treatment (Rayburn 2004; Wang 1999), however neither specified the inclusion criteria for the studies. They also described the studies and the results in a very generic way and did not draw firm conclusions about the superiority of one drug treatment over another. Two other out of date reviews also included observational studies and assessed the relationship between maternal opiate use and infant birth weight (Hulse 1997) and neonatal mortality (Hulse 1998). They found better results in patients taking methadone in terms of birth weight but no reduction in mortality.

Objectives

To assess the effectiveness of any opioid agonist maintenance treatment alone or in combination with psychosocial intervention compared to no intervention, other pharmacological intervention or psychosocial interventions for child health status, neonatal mortality, retaining pregnant women in treatment and reducing the use of substances.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials enrolling pregnant women. Studies starting after the delivery were excluded.

Types of participants

Opiate-addicted pregnant women of any age irrespective of duration of pregnancy. There was no restriction with respect to physical or psychological illness.

Types of interventions

Experimental intervention
  • Any pharmacological intervention (methadone, buprenorphine, LAAM, heroin, morphine, codeine) alone or combined with psychosocial intervention for maintenance treatment.

Control intervention
  • No intervention.

  • Other pharmacological intervention.

  • Psychosocial intervention alone.

Types of outcome measures

Primary outcomes
For the woman

1. Drop-out from treatment, as measured by the number of women who had dropped out at the end of the intervention

2. Use of primary substance of abuse

2.1 Use of primary substance as measured by the number of women using heroin during or at the end of treatment (self report or urine analysis results)

2.2. Use of primary substance at follow-up as measured by the number of women using heroin at the end of follow-up (after childbirth)

3. Obstetric outcomes

3.1 Third trimester bleeding

3.2 Fetal distress and meconium aspiration

3.3 Caesarean section

3.4 Abnormal presentation

3.5 Medical complications at delivery

3.6 Breastfeeding following delivery

3.7 Puerperal morbidity

For the child

4. Health status measured as:

4.1 Birth weight

4.2 APGAR score (Activity, Pulse, Grimace, Appearance and Respiration score)

4.3 Neonatal abstinence syndrome

4.4 Prenatal and neonatal mortality 

Secondary outcomes

5. Nicotine consumption
6. Use of other substances
7. Side effects for the mother
8. Side effects for the child

Search methods for identification of studies

Electronic searches

We searched the following electronic databases (search date 30 September 2013):

  • Cochrane Drugs and Alcohol Group (CDAG) Specialised Register;

  • Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2013, Issue 3) (September 2013);

  • MEDLINE (PubMed) (1966 to September 2013);

  • EMBASE (Elsevier, EMBASE.com) (1974 to September 2013);

  • CINAHL (EBSCO Host) (1982 to September 2013);

  • Web of Science (June 2007 to September 2013).

Searching other resources

We searched:

  1. the reference lists of all relevant papers to identify further studies;

  2. some of the main electronic sources of ongoing trials (National Research Register, metaRegister of Controlled Trials; ClinicalTrials.gov, Agenzia Italiana del Farmaco);

  3. conference proceedings likely to contain trials relevant to the review (College on Problems of Drug Dependence - CPDD);

  4. national focal points for drug research (e.g. National Institute of Drug Abuse (NIDA), National Drug & Alcohol Research Centre (NDARC).

We contacted the authors of included studies and experts in the field in various countries to find out if they knew any other published or unpublished controlled trials.

We did not apply any language restrictions.

Data collection and analysis

Selection of studies

Two authors (Minozzi, Bellisario) independently inspected the search hits by reading titles and abstracts. We obtained each potentially relevant study located by the search in full text and the two authors assessed each for inclusion independently. Doubts were resolved by discussion between the authors.

Data extraction and management

Two authors (Minozzi, Bellisario) independently extracted data from studies using a standardised checklist. Any disagreement was resolved by discussion.

Assessment of risk of bias in included studies

Two authors (Minozzi, Bellisario) independently performed the 'Risk of bias' assessment for RCTs and CCTs using the criteria recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The recommended approach for assessing risk of bias in studies included in Cochrane Reviews 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 sources 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 in the Cochrane Handbook for Systematic Reviews of Interventions adapted to the addiction field. See Appendix 7 for details.

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

We considered blinding of participants, personnel and outcome assessor (avoidance of performance bias and detection bias) separately for objective outcomes (e.g. drop-out, use of substance of abuse measured by urine analysis, participants relapsed at the end of follow-up, participants engaged in further treatment) and subjective outcomes (e.g. duration and severity of signs and symptoms of withdrawal, patient self reported use of substance, side effects).

We considered incomplete outcome data (avoidance of attrition bias) for all outcomes except for drop-out from the treatment, which is very often the primary outcome measure in trials on addiction.

Grading of evidence

We assessed the overall quality of the evidence for the primary outcome using the GRADE system. The Grading of Recommendation, Assessment, Development and Evaluation Working Group (GRADE) developed a system for grading the quality of evidence (GRADE 2004; Guyatt 2008; Guyatt 2011; Schünemann 2006) which takes into account issues not only related to internal validity but also to external validity, such as directness of results. 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 grades 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.

Grading is decreased for the following reasons:

  • 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).

Grading is increased for the following reasons:

  • Strong evidence of association - significant relative risk of > 2 (< 0.5) based on consistent evidence from two or more observational studies, with no plausible confounders (+1).

  • Very strong evidence of association - significant relative risk of > 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).

Measures of treatment effect

We analysed dichotomous outcomes by calculating the risk ratio (RR) for each trial with the uncertainty in each result being expressed by its confidence interval. We analysed continuous outcomes by calculating the mean difference (MD) or the standardised mean difference (SMD) with 95% CI. For nicotine use we compared the difference of the mean number of cigarettes smoked from baseline to end of treatment in the experimental and control group. In case of missing standard deviations (SDs) for the difference from baseline to the end of treatment, we imputed the SD using the SD of the mean at the end of treatment for each group.

We did not use data presented as the number of positive urine tests over the total number of tests in the experimental and control group as a measure of substance abuse. This is because using tests instead 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.

Assessment of heterogeneity

We analysed heterogeneity by means of the I2 statistic and Chi2 test for heterogeneity. The cut points were an I2 value > 50% and a P value of the Chi2 test of < 0.1.

Assessment of reporting biases

We intended to use funnel plots (plots of the effect estimate from each study against the standard error) to assess the potential for bias related to the size of the trials, which could indicate possible publication bias, if a sufficient number of studies (i.e. at least 10 studies) were included. If asymmetry had been found we would have conducted the Egger test (Egger 1997).

Data synthesis

We combined the outcome measures from the individual trials through meta-analysis where possible (comparability of intervention and outcomes between trials). We used the fixed-effect model because studies were expected to be similar enough for types of participants, setting and treatments administered. For three outcomes, were an high heterogeneity was found a random effect model was used.

Sensitivity analysis

To incorporate the assessment of risk of bias into the review process we first plotted the intervention effect estimates against the assessment of risk of bias. We inspected the results stratified for risk of bias and if we found significant associations between measure of effect and risk of bias we excluded studies with high risk of bias from the analysis. The items considered in the sensitivity analysis were: random sequence generation, allocation concealment and blinding of personnel and outcome assessors.

Results

Description of studies

Results of the search

We identified 968 records and discarded 933 on basis of title and abstract. We retrieved 35 articles in full text for more detailed evaluation, 20 of which were excluded. Fifteen articles related to four studies satisfied all the criteria to be included in the review. See Figure 1 for a flow chart of the search process.

Figure 1.

Study flow diagram.

We did not find any unpublished studies. We wrote to the first authors of published studies and one replied. He confirmed that to his knowledge there are no unpublished trials and that the four studies included in this review are the only existing randomised trials on pharmacotherapy for pregnant women.

For substantive descriptions of studies see the Characteristics of included studies and Characteristics of excluded studies tables.

Included studies

Four studies with 271 participants met the inclusion criteria for this review.

Duration of trials

The mean duration of the trials was 16.3 weeks (range 15 to 18 weeks).

Treatment regimes and settings

Two trials were conducted in Austria (Fischer 1999; Fischer 2006) and one in the USA (Jones 2005). The MOTHER Study is a multicentre, international study conducted in Austria, Canada and the USA.

Three trials compared a methadone dose of between 20 and 140 mg/day with a buprenorphine dose of between 2 and 32 mg/day (Fischer 2006; Jones 2005; MOTHER Study). One trial compared methadone (mean dose at delivery 53.48 mg) with oral slow-release morphine (mean dose at delivery 300.43 mg) (Fischer 1999).
Three studies were conducted in an outpatient setting (Fischer 1999; Fischer 2006; MOTHER Study) and one in an inpatient setting (Jones 2005).

Participants

271 opiate-dependent pregnant women meeting the DSM-IV criteria were included in the studies. The mean age of participants was 27.3 years. The mean gestational agewas 22 weeks. Nicotine use during pregnancy was reported only in one trial (Fischer 1999) as mean number of cigarettes per day: it was 27.56 (SD 26.28) for the methadone group and 31.30 (SD 22.56) for the morphine group.

Rating instruments utilised in the studies

All the included studies measured neonatal abstinence syndrome using the Finnegan scale (Finnegan 1992), but the MOTHER Study used a modified Finnegan scale (called the MOTHER neonatal abstinence syndrome scale) which includes 28 items: 19 items were used for scoring and medication decisions. Scores on the modified scale range from 0 to 42, with higher scores indicating more severe withdrawal. Original neonatal abstinence syndrome item definitions, as well as the morphine medication protocol, were refined before data collection.

Comparisons
Comparison 1

Methadone versus buprenorphine was compared in three trials with 223 participants (Fischer 2006; Jones 2005; MOTHER Study).

Comparison 2

Methadone versus slow-release morphine was compared in one trial with 48 participants (Fischer 1999).

Excluded studies

Twenty studies did not meet the criteria for inclusion in this review. The grounds for exclusion were: study design (13 studies) (Bandstra 2012; Binder 2008; Ebner 2007; Fisher 1998; Gordon 2004; Hulse 2004; Keyser-Marcus 2002; Jones 2008; Lacroix 2011; Laken 1997; Martin 2011; Newman 2009; Stine 2009), type of participants (four studies) (Bell 2007; Dawe 2007; Jackson 2004; Suchman 2007), type of experimental intervention (one study) (Carroll 1995) and type of control intervention (two studies) (Jones 2011; Tuten 2012).

Risk of bias in included studies

See Figure 2 and Figure 3. All of the studies were randomised controlled trials.

Figure 2.

'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 3.

'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Random sequence generation

Only one study (Jones 2005) used a random sequence generation method and was at low risk of selection bias. We judged all the other studies to be at unclear risk of bias.

Allocation concealment

There was a low risk of bias for allocation concealment in three studies (Fischer 2006; Jones 2005; MOTHER Study) and an unclear risk in the fourth study (Fischer 1999).

Blinding

Subjective outcomes

We judged one study (Fischer 1999) to be at high risk of performance and detection bias because it was an open study. All the other studies were double-blind and we judged them to be at low risk of bias.

Objective outcomes

We judged all four studies to be at low risk of performance and detection bias.

Incomplete outcome data

Only one study had no attrition (Fischer 1999). We judged the other three to be at high risk of attrition bias because the attrition rate was high and unbalanced between groups.

Effects of interventions

See: Summary of findings for the main comparison Methadone compared to buprenorphine for opiate-dependent pregnant women; Summary of findings 2 Methadone compared to oral slow-release morphine for opiate-dependent pregnant women

Comparison 1: Methadone versus buprenorphine

See Summary of findings for the main comparison.

Primary outcomes: for the woman
1. Drop-out from treatment

We analysed the number of participants who did not complete the treatment (Fischer 2006; Jones 2005; MOTHER Study): the risk ratio (RR) was 0.64 (95% confidence interval (CI) 0.41 to 1.01, 223 participants, three studies). The result is in favour of methadone (Analysis 1.1; Figure 4).

Figure 4.

Forest plot of comparison: 1 Methadone versus buprenorphine, outcome: 1.1 Drop-out.

2. Use of primary substance of abuse
2.1 During or at the end of treatment

We pooled two trials (Jones 2005, MOTHER Study) with 151 participants. The result was not statistically significant (RR 1.81, 95% CI 0.70 to 4.69) (Analysis 1.2).

2.2 At follow-up

None of the studies considered this outcome.

3. Obstetric outcomes
3.1 Third trimester bleeding

Data for this outcome were not reported in any of the included studies.

3.2 Preterm delivery

In Fischer 2006, three children were delivered prematurely in the methadone group and two in the buprenorphine group (one at week 34, one at week 35 and three at week 36). In Jones 2005 there was one preterm birth in the methadone group (week not reported). In the MOTHER Study there were 19% preterm deliveries in the methadone group and 7% in the buprenorphine group. The difference was not statistically significant.

3.3 Fetal distress and meconium aspiration

Data for this outcome were not reported in Jones 2005 and Fischer 2006. In the MOTHER Study there was one case of meconium aspiration in the buprenorphine group.

3.4 Caesarean section

In Jones 2005, all but one birth in each group were vaginal. In Fischer 2006, two women maintained on buprenorphine were delivered by planned caesarean section at week 40. In the MOTHER Study there were 37% caesarean sections in the methadone group and 29% in the buprenorphine group. The difference was not statistically significant.

3.5 Abnormal presentation

In Jones 2005, all births were normal presentation. Data were not reported in Fischer 2006. In the MOTHER Study there were 14% abnormal fetal presentations in the methadone group and 5% in the buprenorphine group. The difference was not statistically significant.

3.6 Medical complications at delivery

In Fischer 2006, one woman in the methadone group required vacuum extraction due to a prolonged delivery. No medical complications occurred in Jones 2005. In the MOTHER Study there were 51% medical complications at delivery in the methadone group and 31% in the buprenorphine group(P = 0.03).

3.7 Breastfeeding following delivery

Data were not reported in any of the included studies.

3.8 Puerperal morbidity

No cases of puerperal morbidity were observed in Jones 2005 or in the MOTHER Study. Data were not reported in Fischer 2006.

Primary outcomes: for the child
4. Health status
4.1 Birth weight

We pooled two studies (Jones 2005; MOTHER Study) with 150 participants (one mother in the buprenorphine group delivered twins; twin data were not considered for this outcome because it could be altered by twin status). The mean difference (MD) was -365.45 g (95% CI -673.84 to -57.07) ; the results are in favour of buprenorphine (Analysis 1.3; Figure 5) but there was high heterogeneity ( P 0.00001, I2 95%) The third study (Fischer 2006) did not report data but states that there was no statistically significant difference in birth weight between groups (mean 2820 g).

Figure 5.

Forest plot of comparison: 1 Methadone versus buprenorphine, outcome: 1.3 Birth weight.

4.2 APGAR score (Activity, Pulse, Grimace, Appearance and Respiration score)at five minutes

We pooled two studies (Jones 2005; MOTHER Study) with 163 participants. The MD was 0.00 (95% CI -0.03 to 0.03), which is not statistically significant (Analysis 1.4; Figure 6). The third study (Fischer 2006) did not report data but states that there was no statistically significant difference in the APGAR score between groups.

Figure 6.

Forest plot of comparison: 1 Methadone versus buprenorphine, outcome: 1.4 APGAR score.

4.3 Neonatal abstinence syndrome

We pooled the number of newborns treated for neonatal abstinence syndrome from three studies with 166 participants (Fischer 2006; Jones 2005; MOTHER Study). The RR was 1.22 (95% CI 0.89 to 1.67), which is not statistically significant (Analysis 1.5).

Neonatal abstinence syndrome peak score over all observation days was measured in two studies (Jones 2005; MOTHER Study). In Jones 2005 (21 participants) the results were methadone: 4.9, buprenorphine: 6.8(SDs not provided), which is not statistically significant. In the MOTHER Study (131 participants) the results were methadone: 12.8 ± 0.6, buprenorphine: 11.0 ± 0.6 (P = 0.04), which is in favour of buprenorphine.

We pooled the mean duration of treatment for neonatal abstinence syndrome results from two studies (Fischer 2006; MOTHER Study) (145 participants). The MD was 0.00 (95% CI -0.03 to 0.03), which is not statistically significant (Analysis 1.6).

The total number of morphine drops administered was measured in one study with 21 participants (Jones 2005). The results were methadone: 93.1, buprenorphine: 23.6, which is not statistically significant.

We pooled the results for length of hospital stay from two studies (Jones 2005; MOTHER Study) (152 participants). The MD was 4.01 (95%CI 1.29 to 9.30] which is not statistically significant, but there was high heterogeneity (P<0.00001, I2: 99%). (Analysis 1.7).

We pooled the results for total amount of morphine required to manage neonatal abstinence syndrome from two studies (Fischer 2006; MOTHER Study) (145 participants). The MD was 5.06 95%CI -3.36 to 13.47) which is not statistically significant, but there was high heterogeneity (P<0.00001, I2: 99%), (Analysis 1.8).

4.4 Prenatal and neonatal mortality

In one study (Fischer 2006) there was one sudden intrauterine death at 38 weeks of pregnancy and one late abortion at 28 weeks of pregnancy, both in the methadone group. In the first woman urine toxicology revealed 66% opioid-positive results, 48% cocaine-positive results and 16% benzodiazepine-positive results over the study period. Cigarette consumption was a mean of 35 per day. In the second woman all urine toxicology results were negative.

Secondary outcomes
5. Nicotine consumption

Smoking data were available from 124 (methadone n = 67 and buprenorphine n = 57) of the patients enrolled in the MOTHER Study. Among the sample 95% reported cigarette smoking at treatment entry. Participants in the two medication conditions were similar for pretreatment characteristics including smoking rates and daily cigarette amounts. Over the course of the pregnancy, no meaningful changes in cigarette smoking were observed for either medication condition. The fitted difference in
change in adjusted cigarettes per day between the two conditions was small and non-significant (â = 0.08, standard error (SE) = 0.05, P = 0.132).

6. Use of other substances
Jones 2005 reported the percentage of urine positive for each substance during the study period for the methadone and buprenorphine group respectively. The results were cocaine: 15.6% and 16.7%; benzodiazepines: 0.4% and 2.5%; amphetamine: 0% and 0%; and marijuana 7.5% and 0%. The Fischer 2006 study reported the median number of urine samples positive for methadone and buprenorphine respectively: cocaine: 0.00 and 0.00; benzodiazepines: 7.82 and 5.36. No data were reported in the MOTHER Study.

7. Side effects for the mother
No side effects for the mothers were reported in the Jones 2005 and Fischer 2006 studies. In the MOTHER Study there were 14/89 (16%) serious adverse events in the methadone group and 8/86 (9%) in the buprenorphine group(RR 1.69, 95% CI 0.75 to 3.87) (Analysis 1.9). There were also 83/89 (93%) non-serious adverse events in the methadone group and 66/86 (77%) in the buprenorphine group (RR 4.77, 95% CI 0.59 to 38.49)(Analysis 1.10). The results were not statistically significant.

8. Side effects for the child
No side effects for the child were reported in the Jones 2005 and Fischer 2006 studies. In the MOTHER Study there were 6/73 (8%) serious adverse events in the methadone group and 1/58 (2%) in the buprenorphine group(RR 1.22, 95% CI 1.07 to 1.38), which was in favour of buprenorphine(Analysis 1.11). There were also 34/73 (47%) non-serious adverse events in the methadone group and 29/58 (50%) in the buprenorphine group(RR 1.08, 95% CI 0.74 to 1.59). The results were not statistically significant.

Comparison 2: Methadone versus oral slow-release morphine

See Summary of findings 2. This comparison includes one study (Fischer 1999) with 48 participants.

Primary outcomes: for the woman
1. Drop-out from treatment

No participants dropped out from either group.

2. Use of primary substance of abuse
2.1 During or at the end of treatment

The result for the number of participants who used heroin in the third trimester was RR 2.40 (95% CI 1.00 to 5.77); this is in favour of oral slow-release morphine (Analysis 2.1).

2.2 At follow-up

The study did not consider this outcome.

3. Obstetric outcomes

3.1 Third trimester bleeding

Data were not reported.

3.2 Preterm delivery

One woman delivered at 31 weeks due to early amniotic rupture, but it is not reported to which group she was allocated. No other gynaecological problems occurred during the study period. The mean week of delivery was as follows: methadone: 38.92 (SD 1.74), morphine: 37.79 (SD 2.55); the difference was not significant.

3.3 Fetal distress and meconium aspiration

Data were not reported.

3.4 Caesarean section

The percentage of caesarean sections was 25% in both groups.

3.5 Abnormal presentation

Datawere not reported.

3.6 Medical complications at delivery

The percentage of vacuum extractions was 8.3% in both groups.

3.7 Breastfeeding following delivery

Data were not reported.

3.8 Puerperal morbidity

Data were not reported.

Primary outcomes: for the child

4. Health status

4.1 Birth weight

The MD was 124 g (95% CI -186 to 434); the result is not statistically significant (Analysis 2.2).

4.2 APGAR score

The study did not consider this outcome.

4.3 Neonatal abstinence syndrome

The mean duration of neonatal abstinence syndrome was MD -5.00 (95% CI -10.97 to 0.97); the result is not statistically significant (Analysis 2.3).

4.4 Prenatal and neonatal mortality

There was no prenatal or neonatal mortality in either group.

Secondary outcomes
5. Nicotine consumption

Fischer 1999 measured the difference in the mean number of cigarettes smoked per day before and at the end of treatment. At the start of the trial the mean number of cigarettes smoked per day was 27.56 (SD 16.28) and 31.30 (SD 22.56) for the methadone and morphine group respectively. At delivery it was 15.89 (SD 12.24) and 15.20 (SD 8.24) respectively (MD -4.43, 95% CI -1.47 to 10.33); the result is not statistically significant, but there is a trend in favour of morphine (Analysis 2.4).

6. Use of other substances

The study only reported the percentage of urine testing negative during each week of treatment for methadone and slow-release morphine in a graph: the mean percentages for the whole study period were about 95% and 90% respectively for cocaine and 54% and 89% for benzodiazepines.

7. Side effects for the mother

No side effects for the mother were reported.

8. Side effect for the child

No severe complications were noted in any of the neonates as monitored by polysomnography, although one child in the methadone group had central apnoea and one child in the morphine group had obstructive apnoea.

Sensitivity analysis

The only difference among the trials in terms of risk of bias was in random sequence generation (we judged one as at low risk and two as at unclear risk of bias). No differences were found in the results when studies were stratified for risk of bias for random sequence generation, therefore we did not perform sensitivity analysis.

Assessment of reporting bias

We did not use a funnel plot to explore publication bias because only four studies are included, all with small sample sizes. In fact the power of this method is at most limited when meta-analysis comprises only a few small studies (Higgins 2011).

Discussion

Summary of main results

Only four trials with 271 pregnant women satisfied the criteria for inclusion in the review. Three (Fischer 2006; Jones 2005; MOTHER Study) compared methadone with buprenorphine (223 participants) and one (Fischer 1999) compared methadone with oral slow-release morphine (48 participants).

In the comparison of methadone versus buprenorphine the drop-out rate of the mothers was lower in the methadone group. However, there was no difference in the use of the primary substance between the methadone and buprenorphine groups. We judged the quality of evidence for both to be low. Birth weight was higher in the buprenorphine group in the two trials that could be pooled. The third study reported that there was no statistically significant difference. For APGAR score none of the studies found a significant difference. We judged the quality of evidence for both to be low. Many measures were used in the studies to assess neonatal abstinence syndrome. The number of newborns treated for neonatal abstinence syndrome, which is the most critical outcome, did not differ significantly between groups. We judged the quality of evidence to be very low.

In the comparison of methadone versus oral slow-release morphine we found no difference in the drop-out rate, whereas slow-release morphine seemed to be superior to methadone for abstinence of women from the use of heroin. The study did not find any statistically significant differences in birth weight or mean duration of neonatal abstinence syndrome. The APGAR score was not considered in the study. We judged the quality of evidence to be moderate for all outcomes.

Only one study which compared methadone with buprenorphine reported side effects: for the mother there was no statistically significant difference between groups. For the newborns in the buprenorphine group there were significantly fewer serious side effects.

In the comparison of methadone versus slow-release morphine no side effects were reported for the mother, whereas one child in the methadone group had central apnoea and one child in the morphine group had obstructive apnoea.

The difficulty of determining whether opoid substitution is associated with better outcomes for newborns always needs to be considered in relation to the direct effects of cigarette smoking. Only one study (Fischer 1999) reported data on cigarette consumption at the start of the study and at delivery. Women smoked a mean of 29 cigarettes per day at enrolment in the study and a mean of 14 cigarettes per day at delivery. There was no statistically significant difference between groups in the reduction of cigarettes smoked. This seems to be a relevant outcome and was not considered by the majority of the included studies. The level of nicotine exposure during pregnancy does affect birth weight and could affect neonatal abstinence syndrome.

Overall completeness and applicability of evidence

The number of included studies and participants is too small to allow definite conclusions. Although, from a clinical point of view, it is well recognised that for both women opiate addicts and their infants agonist treatment during pregnancy is protective, many questions still remain unanswered. Which is the most effective drug treatment and at what dosage? What is the most appropriate type of setting? In particular, is it useful or not to associate any type of psychosocial intervention with pharmacological treatment?

Furthermore, all of the included studies ended their follow-up immediately after the delivery. It would be useful to know, both for maternal and child health, whether the mothers continued to use maintenance treatments after their delivery.

Quality of the evidence

Three out of four studies had adequate allocation concealment and were double-blind. The major flaw of the studies was attrition bias: three out of four had a high drop-out rate (30% to 40%) and this was unbalanced between groups.

Potential biases in the review process

We did not find any unpublished studies, despite contacting all of the first authors of the included studies and searching conference proceedings

Agreements and disagreements with other studies or reviews

We did not find any other systematic reviews on this topic. One narrative review published in 2008 (Winklbaur 2008) considered all types of treatment for pregnant opioid-dependent women and included both randomised and non-randomised studies. This review cited the three randomised controlled trials (RCTs) published before 2006 and concluded that buprenorphine and methadone are substantially equivalent but may attenuate neonatal abstinence syndrome.

A recently published narrative review (Jones 2012) compared buprenorphine with methadone and included the same three RCTs together with observational studies but did not pool the data in a meta-analysis. This review reached the same conclusions, reporting that there is no significant difference between methadone and buprenorphine for maternal outcomes whereas buprenorphine "...results in a clinically significant less severe neonatal abstinence syndrome (NAS) than treatment with methadone". They also stated "...concluding that buprenorphine is an effective treatment for opioid dependence during pregnancy does not mean that methadone should no longer be considered a useful and effective medication for opioid dependence, nor does it mean that all opioid-dependent pregnant women should be treated with buprenorphine without regard to their preferences and life circumstances."

Authors' conclusions

Implications for practice

We did not find sufficient significant differences between methadone and buprenorphine or slow-release morphine to allow us to conclude that one treatment is superior to another for all relevant outcomes. While methadone seems superior in terms of retaining patients in treatment, buprenorphine seems to lead to less severe neonatal abstinence syndrome. Additionally, even though a multicentre, international trial with 175 pregnant women has recently been completed, its results published and included in this review, the body of evidence is still too small to draw firm conclusions about the equivalence of the treatments compared. Many questions remain unanswered. Which is the most effective drug treatment and at what dosage? What is the most appropriate type of setting? In particular, is it useful or not to associate any type of psychosocial intervention with pharmacological treatment?

Implications for research

Large randomised controlled trials which compare different pharmacological maintenance treatments are still needed, with longer follow-up periods (ideally up to one year) and which also consider as relevant outcomes the level of nicotine exposure, concomitant use during pregnancy of other prescribed medications (such as selective serotonin reuptake inhibitors and benzodiazepines) and use of non-prescribed drugs such as cocaine, alcohol and marijuana. Moreover, studies assessing the effectiveness of psychosocial treatments in addition to pharmacological treatments versus pharmacological treatments alone should be conducted.

Acknowledgements

We would like to thank Zuzana Mitrova for developing and running the search strategies and for her help in the editorial process.

Data and analyses

Download statistical data

Comparison 1. Methadone versus buprenorphine
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Drop-out3223Risk Ratio (M-H, Fixed, 95% CI)0.64 [0.41, 1.01]
2 Use of primary substance2151Risk Ratio (M-H, Fixed, 95% CI)1.81 [0.70, 4.69]
3 Birth weight2150Mean Difference (IV, Random, 95% CI)-365.45 [-673.84, -57.07]
4 APGAR score2163Mean Difference (IV, Fixed, 95% CI)0.0 [-0.03, 0.03]
5 Number treated for NAS3166Risk Ratio (M-H, Fixed, 95% CI)1.22 [0.89, 1.67]
6 Mean duration of NAS treatment2145Mean Difference (IV, Fixed, 95% CI)0.00 [-0.03, 0.03]
7 Length of hospital stay2152Mean Difference (IV, Random, 95% CI)4.01 [-1.29, 9.30]
8 Total amount of morphine for NAS2145Mean Difference (IV, Random, 95% CI)5.06 [-3.36, 13.47]
9 Serious AE for the mother1175Risk Ratio (M-H, Fixed, 95% CI)1.69 [0.75, 3.83]
10 Serious AE for the child1131Risk Ratio (M-H, Fixed, 95% CI)4.77 [0.59, 38.49]
11 Non-serious AE for the mother1175Risk Ratio (M-H, Fixed, 95% CI)1.22 [1.07, 1.38]
12 Non-serious AE for the child1131Risk Ratio (M-H, Fixed, 95% CI)1.08 [0.74, 1.59]
Analysis 1.1.

Comparison 1 Methadone versus buprenorphine, Outcome 1 Drop-out.

Analysis 1.2.

Comparison 1 Methadone versus buprenorphine, Outcome 2 Use of primary substance.

Analysis 1.3.

Comparison 1 Methadone versus buprenorphine, Outcome 3 Birth weight.

Analysis 1.4.

Comparison 1 Methadone versus buprenorphine, Outcome 4 APGAR score.

Analysis 1.5.

Comparison 1 Methadone versus buprenorphine, Outcome 5 Number treated for NAS.

Analysis 1.6.

Comparison 1 Methadone versus buprenorphine, Outcome 6 Mean duration of NAS treatment.

Analysis 1.7.

Comparison 1 Methadone versus buprenorphine, Outcome 7 Length of hospital stay.

Analysis 1.8.

Comparison 1 Methadone versus buprenorphine, Outcome 8 Total amount of morphine for NAS.

Analysis 1.9.

Comparison 1 Methadone versus buprenorphine, Outcome 9 Serious AE for the mother.

Analysis 1.10.

Comparison 1 Methadone versus buprenorphine, Outcome 10 Serious AE for the child.

Analysis 1.11.

Comparison 1 Methadone versus buprenorphine, Outcome 11 Non-serious AE for the mother.

Analysis 1.12.

Comparison 1 Methadone versus buprenorphine, Outcome 12 Non-serious AE for the child.

Comparison 2. Methadone versus oral slow-release morphine
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Use of substance148Risk Ratio (M-H, Fixed, 95% CI)2.4 [1.00, 5.77]
2 Birth weight148Mean Difference (IV, Fixed, 95% CI)124.00 [-186.94, 434.94]
3 NAS mean duration148Mean Difference (IV, Fixed, 95% CI)-5.0 [-10.97, 0.97]
4 Nicotine consumption148Mean Difference (IV, Fixed, 95% CI)4.43 [-1.47, 10.33]
Analysis 2.1.

Comparison 2 Methadone versus oral slow-release morphine, Outcome 1 Use of substance.

Analysis 2.2.

Comparison 2 Methadone versus oral slow-release morphine, Outcome 2 Birth weight.

Analysis 2.3.

Comparison 2 Methadone versus oral slow-release morphine, Outcome 3 NAS mean duration.

Analysis 2.4.

Comparison 2 Methadone versus oral slow-release morphine, Outcome 4 Nicotine consumption.

Appendices

Appendix 1. Cochrane Drugs and Alcohol Group Trials Register search strategy

diagnosis=opioid or opiate* and Pregnan* [TI, AB]

Appendix 2. CENTRAL search strategy

  1. MeSH descriptor: [Opioid-Related Disorders] explode all trees  

  2. ((drug or substance) near (abuse* or addict* or dependen* or disorder*)):ti,ab,kw  (Word variations have been searched)

  3. ((opioid* or opiate*) near (abuse* or addict* or dependen*)):ti,ab,kw  (Word variations have been searched)  1133

  4. #1 or #2 or #3

  5. MeSH descriptor: [Heroin] explode all trees   

  6. (opioid* or opiate* or opium or heroin):ti,ab,kw  (Word variations have been searched)    

  7. MeSH descriptor: [Methadone] explode all trees

  8. "methadone":ti,ab,kw  (Word variations have been searched)    

  9. MeSH descriptor: [Buprenorphine] explode all trees         

  10. "buprenorphine":ti,ab,kw  (Word variations have been searched)

  11. "codeine":ti,ab,kw  (Word variations have been searched)            

  12. "morphine":ti,ab,kw  (Word variations have been searched)       

  13. "LAAM":ti,ab,kw  (Word variations have been searched) 

  14. #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13     

  15. MeSH descriptor: [Pregnancy] explode all trees  

  16. pregnant:ti,ab,kw  (Word variations have been searched)           

  17. "mother":ti,ab,kw  (Word variations have been searched)      

  18. #15 or #16 or #17

  19. #4 and #14 and #18

Appendix 3. PubMed search strategy

  1. "Opioid-Related Disorders"[MeSh]

  2. ((opioid* OR opiate*) AND (abuse* OR addict* OR dependen*))

  3. ((drug OR substance) AND (abuse* OR addict* OR dependen* OR disorder*))

  4. #1 OR #2 OR #3

  5. Heroin[MeSH]

  6. heroin[tiab]

  7. (opioid* OR opiate* OR opium)

  8. methadone[MeSH] OR methadone[tiab]

  9. #5 OR #6 OR #7 OR #8

  10. pregnan*[tiab]

  11. "Pregnancy"[Mesh]

  12. "Pregnancy Complications"[Mesh]

  13. mother*[tiab]

  14. #10 OR #11 OR #12 OR #13

  15. randomized controlled trial [pt]

  16. controlled clinical trial [pt]

  17. randomized [tiab]

  18. placebo [tiab]

  19. drug therapy [sh]

  20. randomly [tiab]

  21. trial [tiab]

  22. groups [tiab]

  23. #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22

  24. animals [mh] NOT humans [mh]

  25. #23 NOT #24

  26. #4 AND #9 AND #14 AND #25

Appendix 4. CINAHL search strategy

  1. (MH "Substance Use Disorders+")

  2. TX(drug N3 addict*) or TX(drug N3 dependen*) or TX(drug N3 abuse*) or TX(drug N3 misus*) or TX(drug N3 use*)

  3. TX(substance N3 addict*) or TX(substance N3 dependen*) or TX(substance N3 abuse*) or TX(substance N3 misus*)

  4. TX(opioid* N3 addict*) or TX(opioid* N3 dependen*) or TX(opioid* N3 abuse*) orTX(opiate* N3 addict*) or TX(opiate* N3 dependen*) or TX(opiate* N3 abuse*)

  5. S1 or S2 or S3 or S4

  6. MH "Heroin"

  7. TX heroin

  8. TX (opioid* or opiate*)

  9. opium

  10. (MH "Methadone")

  11. TX methadone

  12. S6 or S7 or S8 or S9 or S10

  13. (MH "Pregnancy+")

  14. TI pregnan* or AB pregnan* or TI mother* or AB mother*

  15. (MH "Pregnancy Complications+")

  16. S13 or S14 or S15

  17. MH "Clinical Trials+"

  18. PT Clinical trial

  19. TI clinic* N1 trial* or AB clinic* N1 trial*

  20. TI ( singl* or doubl* or trebl* or tripl* ) and TI ( blind* or mask* )

  21. AB ( singl* or doubl* or trebl* or tripl* ) and AB ( blind* or mask* )

  22. TI randomi?ed control* trial* or AB randomi?ed control* trial*

  23. MH "Random Assignment"

  24. TI random* allocat* or AB random* allocat*

  25. MH "Placebos"

  26. TI placebo* or AB placebo*

  27. MH "Quantitative Studies"

  28. S16 or S17 or S18 or S19 or S20 or S21 or S22 or S23 or S24 or S25 or S26 or S27

  29. S5 AND S12 AND S16 AND S28

Appendix 5. EMBASE search strategy

  1. 'addiction'/exp

  2. 'drug abuse'/exp

  3. ((drug OR substance OR opioid* OR opiat*) NEXT/5 (abuse* OR addict* OR depend* OR disorder*)):ab,ti

  4. #1 OR #2 OR #3

  5. opioid*:ab,ti OR opiat*:ab,ti OR opium:ab,ti OR heroin*:ab,ti OR narcot*:ab,ti

  6. 'methadone'/exp OR methadone:ab,ti OR 'buprenorphine'/exp OR buprenorphine:ab,ti OR 'codeine'/exp OR codeine:ab,ti OR 'diamorphine'/exp OR morphine:ab,ti OR laam:ab,ti

  7. #5 OR #6

  8. 'pregnancy'/exp OR 'pregnancy complication'/exp OR pregnan*:ab,ti

  9. mother*:ab,ti

  10. #8 OR #9

  11. 'crossover procedure'/exp OR 'double blind procedure'/exp OR 'single blind procedure'/exp OR 'controlled clinical trial'/exp OR 'clinical trial'/exp OR placebo:ab,ti OR 'double blind':ab,ti OR 'single blind':ab,ti OR assign*:ab,ti OR allocat*:ab,ti OR volunteer*:ab,ti OR random*:ab,ti OR factorial*:ab,ti ORcrossover:ab,ti OR (cross:ab,ti AND over:ab,ti) OR 'randomized controlled trial'/exp

  12. #4 AND #7 AND #10 AND #11

Appendix 6. Web of Science search strategy

Timespan=2007-06-01 - 2013-03-18. Databases=SCI-EXPANDED, SSCI, A&HCI.

Topic=(((opioid* OR opiate* OR opium OR heroin OR methadone) same (abuse* or addict* or dependen* or disorder*))) AND Topic=((pregnan* OR mother*)) AND Topic=((randomi* OR randomly OR placebo* OR trial*))

Appendix 7. 'Risk of bias' assessment criteria

 Item Judgement  Description
1. Random sequence generation (selection bias)Low riskThe investigators describe a random component in the sequence generation process such as: random number table; computer random number generator; coin tossing; shuffling cards or envelopes; throwing dice; drawing of lots; minimisation
 High riskThe investigators describe a non-random component in the sequence generation process such as: odd or even date of birth; date (or day) of admission; hospital or clinic record number; alternation; judgement of the clinician; results of a laboratory test or a series of tests; availability of the intervention
 Unclear riskInsufficient information about the sequence generation process to permit judgement of low or high risk
2. Allocation concealment (selection bias)Low riskInvestigators enrolling participants could not foresee assignment because one of the following, or an equivalent method, was used to conceal allocation: central allocation (including telephone, web-based and pharmacy-controlled randomisation); sequentially numbered drug containers of identical appearance; sequentially numbered, opaque, sealed envelopes
 High riskInvestigators enrolling participants could possibly foresee assignments because one of the following methods was used: open random allocation schedule (e.g. a list of random numbers); assignment envelopes without appropriate safeguards (e.g. if envelopes were unsealed or non­opaque or not sequentially numbered); alternation or rotation; date of birth; case record number; any other explicitly unconcealed procedure
 Unclear riskInsufficient information to permit judgement of low or high risk. This is usually the case if the method of concealment is not described or not described in sufficient detail to allow a definite judgement

3. Blinding of participants and providers (performance bias)

Objective outcomes 

Low risk

 

 

No blinding or incomplete blinding, but the review authors judge that the outcome is not likely to be influenced by lack of blinding
Blinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken

 

 High riskNo blinding or incomplete blinding, and the outcome is likely to be influenced by lack of blinding
Blinding of key study participants and personnel attempted, but likely that the blinding could have been broken, and the outcome is likely to be influenced by lack of blinding
 Unclear riskInsufficient information to permit judgement of low or high risk

4. Blinding of participants and providers (performance bias)

Subjective outcomes

Low risk

 

Blinding of participants and providers and unlikely that the blinding could have been broken

 

 High riskNo blinding or incomplete blinding, and the outcome is likely to be influenced by lack of blinding
Blinding of key study participants and personnel attempted, but likely that the blinding could have been broken, and the outcome is likely to be influenced by lack of blinding
 Unclear riskInsufficient information to permit judgement of low or high risk

5. Blinding of outcome assessor (detection bias)

Objective outcomes 

Low risk

 

 

No blinding of outcome assessment, but the review authors judge that the outcome measurement is not likely to be influenced by lack of blinding

Blinding of outcome assessment ensured, and unlikely that the blinding could have been broken

 High risk

No blinding of outcome assessment, and the outcome measurement is likely to be influenced by lack of blinding

Blinding of outcome assessment, but likely that the blinding could have been broken, and the outcome measurement is likely to be influenced by lack of blinding

 Unclear riskInsufficient information to permit judgement of low or high risk

6. Blinding of outcome assessor (detection bias)

Subjective outcomes

Low risk

 

No blinding of outcome assessment, but the review authors judge that the outcome measurement is not likely to be influenced by lack of blinding

Blinding of outcome assessment ensured, and unlikely that the blinding could have been broken

 High risk

No blinding of outcome assessment, and the outcome measurement is likely to be influenced by lack of blinding

Blinding of outcome assessment, but likely that the blinding could have been broken, and the outcome measurement is likely to be influenced by lack of blinding

 Unclear riskInsufficient information to permit judgement of low or high risk

7. Incomplete outcome data (attrition bias)

For all outcomes except retention in treatment or drop-out

Low risk

 

 

 

No missing outcome data

Reasons for missing outcome data unlikely to be related to true outcome (for survival data, censoring unlikely to be introducing bias)

Missing outcome data balanced in numbers across intervention groups, with similar reasons for missing data across groups

For dichotomous outcome data, the proportion of missing outcomes compared with observed event risk not enough to have a clinically relevant impact on the intervention effect estimate

For continuous outcome data, plausible effect size (difference in means or standardised difference in means) among missing outcomes not enough to have a clinically relevant impact on observed effect size

Missing data have been imputed using appropriate methods

All randomised patients are reported/analysed in the group they were allocated to by randomisation irrespective of non-compliance and co-interventions (intention-to-treat)

 High risk

Reason for missing outcome data likely to be related to true outcome, with either imbalance in numbers or reasons for missing data across intervention groups

For dichotomous outcome data, the proportion of missing outcomes compared with observed event risk enough to induce clinically relevant bias in intervention effect estimate

For continuous outcome data, plausible effect size (difference in means or standardised difference in means) among missing outcomes enough to induce clinically relevant bias in observed effect size

'As-treated' analysis done with substantial departure of the intervention received from that assigned at randomisation

 Unclear riskInsufficient information to permit judgement of low or high risk (e.g. number randomised not stated, no reasons for missing data provided; number of drop-outs not reported for each group)

What's new

Last assessed as up-to-date: 7 October 2013.

DateEventDescription
10 December 2013New search has been performedNew search, new trials
10 December 2013New citation required and conclusions have changedconclusions changed

History

Protocol first published: Issue 1, 2007
Review first published: Issue 2, 2008

DateEventDescription
20 October 2008AmendedContact details updated.
26 March 2008AmendedConverted to new review format.
8 January 2008New citation required and conclusions have changedSubstantive amendment.

Contributions of authors

Minozzi and Bellisario inspected the search hits by reading titles and abstracts; Minozzi and Bellisario independently assessed for inclusion each potentially relevant study located in the search; Minozzi and Bellisario extracted data independently. Minozzi performed meta-analysis and wrote the text of the review. Amato commented on the draft and contributed with Ferri to the discussion and the conclusions of the review, and Davoli supervised.

Declarations of interest

None.

Sources of support

Internal sources

  • Department of Epidemiology, Lazio Regional Health Service, Italy.

External sources

  • No sources of support supplied

Differences between protocol and review

We reassessed the risk of bias of the included studies in the updated version of the review according to the new recommendations of The Cochrane Collaboration.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Fischer 1999

MethodsRandomised controlled trial
Open
ParticipantsParticipants: 48 pregnant adults; mean age 26 years; 39.6% unmarried; duration of opioid dependence: mean 57.2 months
Mean gestational age at entry: 22 weeks
Inclusion criteria: opioid-dependent and polysubstance abusing pregnant females meeting DSM-IV criteria
Interventions(1) oral methadone (24 participants) versus (2) oral slow-release morphine (24 participants) after an induction period of 10 days. At delivery mean methadone dose was 53.48 mg, mean morphine dose was 300.43 mg
Outpatients
Follow-up mean: 15 weeks
Outcomes

Neonatal outcomes: fetal distress; birth weight; NAS (Finnegan scale)

Maternal outcomes: retention; cocaine and benzodiazepine consumption (urinalysis); opioid use (identification of injection sites for morphine-maintained group and urinalysis for methadone-maintained group)

NotesCountry of origin: Austria
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskQuote: "The women were randomly assigned to receive either oral methadone or oral slow-release morphine, and induction occurred over a period of 10 days"
Allocation concealment (selection bias)Unclear riskNo information provided concerning allocation concealment
Blinding of participants and personnel (performance bias)
subjective outcomes
High riskOpen-label
Blinding of participants and personnel (performance bias)
objective outcomes
Low riskCOMMENT: open trial but objective outcomes unlikely to be biased by lack of blinding
Blinding of outcome assessment (detection bias)
subjective outcomes
High riskOpen-label
Blinding of outcome assessment (detection bias)
objective outcomes
Low riskCOMMENT: open trial but objective outcomes unlikely to be biased by lack of blinding
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo missing outcome data

Fischer 2006

MethodsRandomised controlled trial
Double-blind
ParticipantsParticipants: 18 pregnant adults; mean age 25.9 years; 66.6% single; 61.1% completed 9 years of education; duration of heroin consumption: mean 20.6 months
Mean gestational age at entry: 24 weeks
Inclusion criteria: opioid-dependent pregnant females meeting DSM-IV criteria
Exclusion criteria: severe somatic or other severe psychiatric diseases, high-risk pregnancy
Interventions(1) oral methadone (9 participants) versus (2) oral buprenorphine (9 participants). Dose of methadone between 40 and 100 mg/day; dose of buprenorphine between 8 and 24 mg/day
Outpatients
Follow-up: mean 16 weeks
Outcomes

Neonatal outcomes: birth weight; NAS (Finnegan scale); child health status APGAR score

Maternal outcomes: retention; maternal withdrawal symptoms (Wang Withdrawal Questionnaire); illicit drugs use: opioid, cocaine, benzodiazepine (urinalysis)

NotesCountry of origin: Austria
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskInformation concerning method of random sequence generation not provided
Allocation concealment (selection bias)Low riskQuote: "Subjects were randomised externally by the hospital pharmacy using a double-blind, double-dummy design and received supervised medication daily from the study pharmacist at the addiction clinic"
Blinding of participants and personnel (performance bias)
subjective outcomes
Low riskQuote: "To maintain study blinding, placebo treatment matched the active treatment. Methadone (EBEWE Arzneimittel GesmbH and Gatt/Koller GesmbH KG, Austria) was prepared at the hospital pharmacy as an oral solution at the required concentration. Placebo for the methadone arm consisted of 30 mg dextromethorphan oral solution (Nowak GmbH, Germany) to simulate the taste of methadone, as used previously in double-dummy trials"
Blinding of participants and personnel (performance bias)
objective outcomes
Low riskQuote: "To maintain study blinding, placebo treatment matched the active treatment. Methadone (EBEWE Arzneimittel GesmbH and Gatt/Koller GesmbH KG, Austria) was prepared at the hospital pharmacy as an oral solution at the required concentration. Placebo for the methadone arm consisted of 30 mg dextromethorphan oral solution (Nowak GmbH, Germany) to simulate the taste of methadone, as used previously in double-dummy trials"
Blinding of outcome assessment (detection bias)
subjective outcomes
Low risk

Quote: "Neonates were observed for a minimum of 10 days on an in-patient basis under blinded conditions for the mothers’ treatment condition and scored every 4 hours using the Finnegan scale (range 0–45)"

Quote: "The double-blind, double-dummy code remained blinded until 30 days after delivery"

Blinding of outcome assessment (detection bias)
objective outcomes
Low risk

Quote: "Neonates were observed for a minimum of 10 days on an in-patient basis under blinded conditions for the mothers’ treatment condition and scored every 4 hours using the Finnegan scale (range 0–45)"

Quote: "The double-blind, double-dummy code remained blinded until 30 days after delivery"

Incomplete outcome data (attrition bias)
All outcomes
High risk4 drop-outs occurred during the study period. One participant maintained on methadone (70 mg/day) experienced a stillbirth caused by a sudden intrauterine death during week 38 of pregnancy. A second woman maintained on methadone had a late abortion during week 28. 2 other participants were withdrawn from the study due to lack of compliance with the scheduled visits. 1 patient was taking buprenorphine and the other methadone. This attrition left 14 participants who completed the study, 6 in the methadone group and 8 in the buprenorphine group. The 4 women who failed to complete the study were older and had a longer duration of opioid dependency than those who completed the study, whereas no significant differences in socio-demographic characteristics were seen at enrolment (see Table 1)
COMMENT: high drop-out (33.3%) in the methadone group and unbalanced (11.1% in the buprenorphine group) Reason for drop-out given and likely to be linked to treatments received

Jones 2005

MethodsRandomised controlled trial
Double-blind
ParticipantsParticipants: 30 pregnant adults; mean age 30.1 years; 75% African-American, 20% white, 5% other; 55% unemployed seeking, 40% unemployed not seeking, 5% homemaker; mean years of education 10.2; use cocaine: past 30 days 75%; opioid use: > 4 x day 55%
Mean gestational age at entry: 23 weeks
Inclusion criteria: estimated gestational age of 16 to 30 weeks; opioid-dependent pregnant females meeting DSM-IV criteria Exclusion criteria: current diagnosis of alcohol abuse or dependence; self reported use of benzodiazepines; serious medical illness; diagnosis of preterm labour; evidence of fetal malformation; positive HIV test
Interventions(1) Oral methadone (15 participants) versus (2) oral buprenorphine (15 participants). Dose of methadone: mean 60 mg/day; dose of buprenorphine: mean 12 mg /day
Setting: inpatients
Follow-up mean: 18 weeks
Outcomes

Neonatal outcomes: number of neonates treated for NAS; peaks of NAS score; length of neonatal hospitalisation; birth weight; child health status; APGAR score

Maternal outcomes: retention; illicit drug use (urinalysis)

NotesCountry of origin: USA
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskQuote: "They were assigned to one of the two treatment groups using a computerized dynamic balanced randomisation"
Allocation concealment (selection bias)Low riskQuote: "Research staff, with no other study involvement, generated the randomised allocation sequence"
Blinding of participants and personnel (performance bias)
subjective outcomes
Low risk

Once medically cleared and randomised, participants were switched from their individualised dose of immediate-release morphine onto an equivalent dose of double-blind study medication

To maintain the double-blind, the actual dose changes were known only to pharmacy staff

Blinding of participants and personnel (performance bias)
objective outcomes
Low risk

Once medically cleared and randomised, participants were switched from their individualised dose of immediate-release morphine onto an equivalent dose of double-blind study medication

To maintain the double-blind, the actual dose changes were known only to pharmacy staff

Blinding of outcome assessment (detection bias)
subjective outcomes
Low risk

Once medically cleared and randomised, participants were switched from their individualised dose of immediate-release morphine onto an equivalent dose of double-blind study medication

To maintain the double-blind, the actual dose changes were known only to pharmacy staff

Blinding of outcome assessment (detection bias)
objective outcomes
Low risk

Once medically cleared and randomised, participants were switched from their individualised dose of immediate-release morphine onto an equivalent dose of double-blind study medication

To maintain the double-blind, the actual dose changes were known only to pharmacy staff

Incomplete outcome data (attrition bias)
All outcomes
High risk

Of the 30 randomised patients, 20 delivered while enrolled in the study; the remaining 10 dropped out during
the study. Of those randomised to buprenorphine reasons for drop-out included: discharged for medical condition (n = 1), missed consecutive dosing days (n = 4) and elected to withdraw (n = 1). Of those randomised to methadone reasons for discharge included: missed consecutive dosing days (n = 3) and elected to withdraw (n = 1)

COMMENT: high drop-out (40%) in the methadone group and unbalanced (26% in the buprenorphine group)

MOTHER Study

  1. a

    APGAR score: Activity, Pulse, Grimace, Appearance and Respiration score
    DSM: Diagnostic and Statistical Manual of Mental Disorders
    NAS: neonatal abstinence syndrome
    VAS: visual analogue scale
    HAD: Hamilton Depression Scale
    WWQ: Wang Withdrawal Questionnaire

Methods

Multicentre, randomised controlled trial

Double-blind, double-dummy

Participants

175 pregnant adults; mean age 27.3 years; 83.2% white; employed 13.3%; mean years of education 11.3; substance use: heroin previous 30 days 10.2, cocaine previous 30 days 4.8, any alcohol previous 30 days 0.4, benzodiazepines previous 30 days 0.8
Mean estimated gestational age of fetus: 18.6 weeks

Inclusion criteria: opioid-dependent women between the ages of 18 and 41 years with a singleton pregnancy between 6 and 30 weeks of gestation

Exclusion criteria: no medical or other conditions contraindicating participation; pending legal action; disorders related to the use of benzodiazepines or alcohol; planning to give birth outside the hospital at the study site

Interventions(1) Sublingual tablets of buprenorphine (58 participants) versus (2) oral methadone (73 participants). Flexible dose range of 2 to 32 mg of buprenorphine; dose range of 20 to 140 mg of methadone
Setting: outpatients
Follow-up: up to 10 days after delivery
Outcomes

Neonatal outcome measures: neonates requiring treatment for NAS, peak NAS score, total amount of morphine needed for treatment of NAS, length of hospital stay and head circumference, number of days during which medication was given for NAS, weight and length at birth, preterm birth, gestational age at delivery, and 1-minute and 5-minute APGAR scores

Maternal outcomes measures: caesarean section, weight gain, abnormal fetal presentation during delivery, anaesthesia during delivery, the results of drug screening at delivery, medical complications at delivery, study discontinuation, amount of voucher money earned for drug-negative tests and number of prenatal obstetric visits

Adverse events for children and mothers

NotesCountry of origin: United States, Austria, Canada
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "Patients were randomised within three stratification factors: (i) site; (ii) EGA (6 weeks and 0/7 days–18 weeks and 6/7 days; 19 weeks and 0/7–30 weeks and 6/7 days) to control for differences in duration of medication exposure; and (iii) cocaine use in the past 30 days (yes/no)"

COMMENT: no further detail provided on the randomisation procedure

Allocation concealment (selection bias)Low riskQuote: "The Center for Substance Abuse Research (CESAR) at the University of Maryland, College Park, was the coordinating center responsible for across-site randomisation, data collection, coordination and protocol adherence monitoring. Randomization assignment was communicated from CESAR directly to the site’s pharmacy."
Blinding of participants and personnel (performance bias)
subjective outcomes
Low riskQuote: "A blinded, individualized dosing schedule was used for the study medications, and a double blind
method was used to implement dose-unit increases or decreases (with dose adjustments of 2 mg for buprenorphine and 5 or 10 mg for methadone). Patients received seven tablets (three in the size of an 8-mg tablet and four in the size of a 2-mg tablet) to place under the tongue for 5 minutes, or until the tablets dissolved. Each tablet contained buprenorphine or placebo. After receiving these tablets, participants received liquid containing methadone or placebo. Oral methadone and flavour-masking concentrates were diluted
to provide the dose in a fixed volume (e.g., 40 ml at U.S. sites and 50 ml in Vienna). Methadone placebo was given in the same fixed volume and included the same flavour-masking concentrates as the active drug concentrate."
Blinding of participants and personnel (performance bias)
objective outcomes
Low riskQuote: "A blinded, individualized dosing schedule was used for the study medications, and a double blind
method was used to implement dose-unit increases or decreases (with dose adjustments of 2 mg for buprenorphine and 5 or 10 mg for methadone). Patients received seven tablets (three in the size of an 8-mg tablet and four in the size of a 2-mg tablet) to place under the tongue for 5 minutes, or until the tablets dissolved. Each tablet contained buprenorphine or placebo. After receiving these tablets, participants received liquid containing methadone or placebo. Oral methadone and flavour-masking concentrates were diluted
to provide the dose in a fixed volume (e.g., 40 ml at U.S. sites and 50 ml in Vienna). Methadone placebo was given in the same fixed volume and included the same flavour-masking concentrates as the active drug concentrate."
Blinding of outcome assessment (detection bias)
subjective outcomes
Low risk

Quote: "With limited exceptions, data were submitted to CESAR using a paper document capture and processing methodology. The Center for Substance Abuse Research (CESAR) at the University of Maryland, College Park, was the coordinating centre responsible for data collection"

COMMENT: not specified whether the trained personnel who collected data were blinded, but given the double-blind condition for personnel who provided dugs and participants, we judged that they were

Blinding of outcome assessment (detection bias)
objective outcomes
Low riskThe outcomes are not likely to be influenced by lack of blinding
Incomplete outcome data (attrition bias)
All outcomes
High risk

Quote: "A total of 16 of the 89 women in the methadone group (18%) and 28 of the 86 women in the buprenorphine group (33%) discontinued treatment before delivery (P = 0.02). Dissatisfaction with the study medication was reported as the reason for discontinuation by 71% of participants in the buprenorphine
group, as compared with only 13% of those in the methadone group"

COMMENT: high drop-out (33.%) in the buprenorphine group and unbalanced (18% in the methadone group) Reason for drop-out given and likely to be linked to treatments received

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Bandstra 2012Editorial
Bell 2007Commentary on a randomised controlled trial where pregnant women were not included
Binder 2008Study defined as randomised but not clear from the text if women were randomised to different maintenance treatments; it seems that they were not
Carroll 1995Excluded as the type of control intervention was not among our inclusion criteria: both groups received methadone at the same dose
Dawe 2007Participants did not meet the inclusion criteria: not pregnant women
Ebner 2007Excluded as the study design did not meet the inclusion criteria: not a randomised controlled trial
Fisher 1998Excluded as the study design did not meet the inclusion criteria: not a randomised controlled trial
Gordon 2004Excluded as the study design did not meet the inclusion criteria: not a randomised controlled trial
Hulse 2004Excluded as the study design did not meet the inclusion criteria: not a randomised controlled trial
Jackson 2004Excluded as the type of participants did not meet the inclusion criteria: newborns randomised to receive different treatments after the delivery
Jones 2008Narrative review
Jones 2011Intervention did not meet the inclusion criteria: psychosocial intervention
Keyser-Marcus 2002Excluded as the study design did not meet the inclusion criteria: not a randomised controlled trial
Lacroix 2011Study design did not meet the inclusion criteria: not a randomised study
Laken 1997Excluded as the study design did not meet the inclusion criteria: not a randomised controlled trial
Martin 2011Narrative review
Newman 2009Comment, letter
Stine 2009Study design did not meet the inclusion criteria: comparison of women who did and did not give consent to participate in the MOTHER study
Suchman 2007Participants did not meet the inclusion criteria: non-pregnant women
Tuten 2012Intervention did not meet the inclusion criteria: psychosocial treatment

Ancillary