Betamimetics for inhibiting preterm labour

James P Neilson; Helen M West; Therese Dowswell

doi:10.1002/14651858.CD004352.pub3

. 2014 Feb 5;2014(2):CD004352. doi: 10.1002/14651858.CD004352.pub3

Betamimetics for inhibiting preterm labour

James P Neilson ^1,^✉, Helen M West ², Therese Dowswell ²

Editor: Cochrane Pregnancy and Childbirth Group

PMCID: PMC10603219 PMID: 24500892

Abstract

Background

Preterm birth is a major contributor to perinatal mortality and morbidity worldwide. Tocolytic agents are drugs used to inhibit uterine contractions. Betamimetics are tocolytic agents that have been widely used, especially in resource‐poor countries.

Objectives

To assess the effects of betamimetics given to women with preterm labour.

Search methods

We searched the Cochrane Pregnancy and Childbirth Group's Trials Register (31 December 2013) and reference lists of retrieved studies.

Selection criteria

Randomised controlled trials of betamimetics, administered by any route or any dose, in the treatment of women in preterm labour where betamimetics were compared with other betamimetics, placebo or no treatment.

Data collection and analysis

Two review authors assessed risk of bias and extracted the data independently.

Main results

Twenty‐eight trials were assessed as eligible for inclusion in the review, but eight did not report any outcome data relevant to the review. Results are based on the 20 trials that contributed data.

Twelve trials, involving 1367 women, compared betamimetics with placebo. Betamimetics decreased the number of women in preterm labour giving birth within 48 hours (average risk ratio (RR) 0.68, 95% confidence interval (CI) 0.53 to 0.88, 10 trials, 1209 women). There was a decrease in the number of births within seven days (average RR 0.80; 95% CI 0.65 to 0.98, five trials, 911 women) but there was no evidence of a reduction in preterm birth (before 37 weeks' gestation) (RR 0.95; 95% CI 0.88 to 1.03, 10 trials, 1212 women). No benefit was demonstrated for betamimetics for perinatal death (RR 0.84; 95% CI 0.46 to 1.55, 11 trials, 1332 infants), or neonatal death (RR 0.90; 95% CI 0.27 to 3.00, six trials, 1174 infants). No significant effect was demonstrated for respiratory distress syndrome (RR 0.87; 95% CI 0.71 to 1.08, eight trials, 1239 infants). A few trials reported on cerebral palsy, infant death and necrotising enterocolitis; no significant differences between groups were identified for any of these outcomes. Betamimetics were significantly associated with the following outcomes: withdrawal from treatment due to adverse effects; maternal chest pain; dyspnoea; palpitation; tremor; headaches; hypokalaemia; hyperglycaemia; nausea or vomiting; nasal stuffiness; and fetal tachycardia.

Nine trials compared different types of betamimetics. Other betamimetics were compared with ritodrine in five trials (n = 948). Other comparisons were examined in single trials: hexoprenaline compared with salbutamol (n = 140), slow versus moderate release salbutamol (n = 52) and salbutamol compared with terbutaline (n = 200). Trials were small, varied, and of insufficient quality to delineate any consistent patterns of effect.

Authors' conclusions

Betamimetics help to delay birth, which may give time to allow women to be transferred to tertiary care or to complete a course of antenatal corticosteroids. However, multiple adverse effects must be considered. The data are too few to support the use of any particular betamimetic.

Plain language summary

Betamimetics for inhibiting preterm labour

Preterm labour is when women go into labour before 37 weeks' gestation. Babies born before term (preterm birth) have poorer outcomes compared with babies who are born at term. The earlier the baby is born, the poorer the outcome. Most preterm births occur in low‐income countries where medical help is less readily available. Preterm birth is a major cause of infant deaths and serious illness worldwide. Preterm birth can result in respiratory distress syndrome and chronic lung disease, bleeding into the fluid spaces (ventricles) within the brain (intraventricular haemorrhage)), generalised infection or infection of the blood stream (sepsis), cerebral palsy and other neuro‐developmental impairments. Even short delays in preterm birth can enable women to reach specialist care and receive 'corticosteroid' drugs that are given to women before birth to improve their babies' lung function.

We assessed the benefits of betamimetics (drugs that inhibit contractions of the uterus) given to women with preterm labour against any adverse effects in randomised controlled trials. The betamimetics were administered, by any route or any dose, and compared with placebo, no treatment or other betamimetics. Twenty trials, testing the effect of betamimetics for inhibiting preterm labour, contributed data.The trials were published over a 44‐year period between 1966 and 2010 and were conducted in tertiary care or university hospitals in high‐income countries.

A total of 1367 women in preterm labour participated in the 12 trials that compared a betamimetic with placebo or no treatment. Betamimetics decreased the number of women giving birth within 48 hours and there was a decrease in the number of births within seven days.

The delay in the timing of birth did not translate into any improvements in neonatal outcomes but most women in the trials were 32 weeks' gestation or more. Betamimetics were not shown to reduce perinatal deaths or respiratory distress syndrome.

The side effects for the woman were considerable. These led to cessation of treatment and symptoms such as palpitations, chest pain, headache, difficulty breathing, nausea and/or vomiting.

There was not enough evidence from the included trials to suggest that one betamimetic agent was superior to another.

Background

Preterm birth is defined as birth occurring before 37 completed weeks of pregnancy (WHO 1977). Of the four million estimated neonatal deaths in the world each year, 99% occur in low‐income countries and around one‐third are directly attributable to premature (preterm) birth (Lawn 2005). Preterm birth is thus the leading cause of neonatal death (Liu 2012). There is considerable regional variation in the incidence of preterm birth; 6.2% in Europe, 10.6% in the US and higher rates in Africa (Beck 2010; Blencowe 2012). Accurate assessment of gestational age is challenging in low‐income countries but accurate ultrasound‐based assessments in Malawi found an incidence as high as 17% (van den Broek 2009). Preterm birth is also a leading cause of perinatal morbidity and of morbidities that persist into childhood. Preterm birth can cause respiratory distress syndrome, bronchopulmonary dysplasia (a chronic lung disease, which can follow respiratory distress syndrome), intraventricular haemorrhage (bleeding into the fluid spaces (ventricles) within the brain), sepsis (generalised infection or infection of the blood stream), cerebral palsy and other forms of neuro‐developmental impairment (Gladstone 2011), blindness and deafness. Preterm infants born in high‐income settings often require intensive care, which is costly (Petrou 2001) and stressful. Intensive care is largely unavailable in those parts of the world with the greatest burden of preterm births, namely Africa and South Asia.    The mechanisms or pathogenesis of preterm labour are not well understood. The causes of spontaneous preterm labour may involve genetic factors, stress, inflammatory response, infection, mechanical factors and haemorrhage (Goldenberg 2008). Antenatal strategies to try to prevent preterm birth have been largely disappointing, and the rate of preterm birth has not decreased (Chang 2013).

Tocolytic drugs (drugs that inhibit contractions of the uterus) have been used to try to prolong pregnancy when preterm labour is diagnosed. Theoretically, tocolysis could improve perinatal outcomes by (a) allowing the fetus to mature further in utero before being born, (b) enhancing lung maturation in the fetus by giving corticosteroid treatment to the mother (Roberts 2006), and (c) allowing time for in‐utero transfer to a tertiary care centre with better facilities for care of the premature newborn, including neonatal intensive care facilities.

Betamimetics (isoxsuprine, hexoprenaline, orciprenaline, ritodrine, terbutaline, and salbutamol) have been used extensively in the past. Beta‐adrenergic agonists activate adenyl cyclase to form cyclic adenosine 3',5' monophosphate (cAMP). The increased cellular levels of cAMP decrease myosin light‐chain kinase activity, both by phosphorylation of the myosin light‐chain kinase itself, and by reducing intracellular calcium through increasing calcium uptake by sarcoplasmic reticulum (Gabor 1982).

The ideal tocolytic agent is one which is effective in prolonging pregnancy but has no side effects for the woman or the baby. Betamimetics stimulate beta‐adrenergic receptors and may cause palpitations, tremor, nausea, vomiting, headaches, nervousness, anxiety, chest pain, shortness of breath, and a range of biochemical disturbances such as hyperglycaemia (high blood sugar level) and hypokalaemia (low serum potassium level). Moreover, pulmonary oedema (fluid accumulation in the lungs) may occur (Mabie 1983) and has even been associated with maternal death. The occurrence of side effects may necessitate some women stopping the medication. Betamimetics cross the placenta and may cause fetal tachycardia (increase fetal heart rate), and hypoglycaemia (low blood sugar level) and hyperinsulinism after birth. When considering whether or not to use tocolytic agents in preterm labour, consideration needs to be given to the risks and benefits for both mother and infant, including the side effects of any medication used.

Although the use of betamimetics to inhibit preterm labour has lost popularity in high‐income settings because of safety concerns, they are still used elsewhere, especially in resource‐poor countries. We aim to review the evidence for benefits and harms of betamimetics in this review. The use of betamimetics to try to reduce the risk of preterm birth in twin pregnancies (Yamasmit 2012) and as maintenance therapy after acute treatment of preterm labour (Dodd 2012) are addressed in other Cochrane reviews.

In addition to betamimetics, several alternative tocolytic agents (that use different physiological mechanisms to inhibit labour) have been examined in trials and used in clinical practice. These include calcium channel blockers, cyclo‐oxygenase (COX) inhibitors (prostaglandin inhibitors), magnesium sulphate, nitric oxide donors, progestational agents, relaxin and oxytocin receptor antagonists. The effectiveness and side effects of these agents have been examined in other Cochrane reviews (Bain 2013; Crowther 2002; Duckitt 2002; King 2003; King 2005; Nardin 2006; Papatsonis 2005; Su 2010), and their relative effects have been explored in a recent network meta‐analysis (Haas 2012).

Objectives

To assess the effects of betamimetics in the treatment of preterm labour.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials (including cluster‐randomised trials) comparing outcomes for women and infants in which betamimetics were used, by any route or any dose, to inhibit preterm labour with outcomes in controls with betamimetics, placebo or no treatment. We did not include trials where the method of allocation to groups was not truly random (e.g. allocation by day of the week).

Types of participants

Pregnant women assessed as being in spontaneous preterm labour and considered suitable for tocolytic agents.

For the purpose of the review, preterm labour was defined by trial authors (with intact or ruptured membranes) unless they contradicted the gestational age criterion.

Types of interventions

Betamimetics, administered by any route or any dose, in the treatment of women in preterm labour where betamimetics were compared with other betamimetics, placebo or no treatment.

Types of outcome measures

Primary outcomes

Primary outcomes were chosen as being most representative of the clinically important measures of effectiveness.

Birth within 48 hours of treatment.
Perinatal death (at seven days).
Respiratory distress syndrome.
Chronic lung disease or bronchopulmonary dysplasia as defined by oxygen dependency at 28 days or 36 weeks postmenstrual age, or both.
Severe neuroradiological abnormality, for example, cystic periventricular leukomalacia (softening of the brain near the ventricles because brain tissue in this area has died) or grade III to IV periventricular haemorrhage, or both.
Abnormal long‐term neurodevelopmental status at more than 12 months corrected age (defined as moderate to severe developmental delay and/or cerebral palsy, and/or sensory impairment, for example, blind and deaf).
Neonatal length of hospital stay.

Secondary outcomes

Time to birth (interval between randomisation and the birth):
1. birth within seven days;
2. birth before 28 completed weeks;
3. birth before 34 completed weeks;
4. birth before 37 completed weeks.
Adverse events:
1. cessation of treatment due to adverse drug reaction;
2. serious maternal outcomes, for example, maternal death, cardiac arrest, respiratory arrest, admission to intensive care unit;
3. maternal adverse drug reactions (all), for example, palpitation, tachycardia, cardiac arrhythmias, pulmonary oedema, myocardial ischaemia, chest pain, dyspnoea, tremor, hypotension, hyperglycaemia, hypokalaemia, nausea, vomiting, nasal stuffiness, headaches, nervousness, anxiety;
4. fetal and neonatal side effects (all), for example, hypoglycaemia, fetal tachycardia.
Neonatal morbidity:
1. neonatal sepsis;
2. necrotising enterocolitis (an inflammation causing injury to the bowel);
3. retinopathy of prematurity (abnormal growth of blood vessels in the baby's eye).
Neonatal death (at 28 days).
Infant death (at 12 months of age).

Search methods for identification of studies

Electronic searches

We searched the Cochrane Pregnancy and Childbirth Group's Trials Register by contacting the Trials Search Co‐ordinator (31 December 2013).

The Cochrane Pregnancy and Childbirth Group’s Trials Register is maintained by the Trials Search Co‐ordinator and contains trials identified from:

monthly searches of the Cochrane Central Register of Controlled Trials (CENTRAL);
weekly searches of MEDLINE;
weekly searches of Embase;
handsearches of 30 journals and the proceedings of major conferences;
weekly current awareness alerts for a further 44 journals plus monthly BioMed Central email alerts.

Details of the search strategies for CENTRAL, MEDLINE and Embase, the list of handsearched journals and conference proceedings, and the list of journals reviewed via the current awareness service can be found in the ‘Specialized Register’ section within the editorial information about the Cochrane Pregnancy and Childbirth Group.

Trials identified through the searching activities described above are each assigned to a review topic (or topics). The Trials Search Co‐ordinator searches the register for each review using the topic list rather than keywords.

Searching other resources

We scanned the reference lists of all articles found.

We did not apply any language restrictions.

Data collection and analysis

For the methods used when assessing the trials identified in the previous version of this review, seeAppendix 1.

For this update, we used the following methods when assessing the reports identified by the updated searches.

Selection of studies

Two review authors (T Dowswell (TD) and H West (HW)) independently assessed eligibility for inclusion for studies identified by the search strategy. We resolved any disagreement through discussion or, if required, we consulted the third author (J Neilson (JPN)).

Data extraction and management

We designed a form to extract data. For eligible studies, two review authors (TD and HW) extracted data independently using the agreed form. We resolved discrepancies through discussion or, if required, we consulted the third review author (JPN) We entered data into Review Manager software (RevMan 2012) and checked for accuracy.

When information regarding any of the above was unclear, we attempted to contact authors of the original reports to provide further details.

Assessment of risk of bias in included studies

Two review authors independently assessed risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We resolved any disagreement by discussion or by involving a third assessor.

(1) Sequence generation (checking for possible selection bias)

We have described for each included study the method used to generate the allocation sequence in sufficient detail to allow an assessment of whether it should produce comparable groups.

We assessed the method as:

low risk of bias (any truly random process, e.g. random number table; computer random number generator);
high risk of bias (any non‐random process, e.g. odd or even date of birth; hospital or clinic record number);
unclear.

(2) Allocation concealment (checking for possible selection bias)

We have described for each included study the method used to conceal the allocation sequence and determined whether intervention allocation could have been foreseen in advance of, or during recruitment, or changed after assignment.

We assessed the methods as:

low risk of bias (e.g. telephone or central randomisation; consecutively numbered sealed opaque envelopes);
high risk of bias (open random allocation; unsealed or non‐opaque envelopes, alternation; date of birth);
unclear.

(3) Blinding (checking for possible performance and detection bias)

We have described for each included study the methods used, if any, to blind study participants and personnel from knowledge of which intervention a participant received. We considered that studies were at low risk of bias if they were blinded, or if we judged that the lack of blinding could not have affected the results. We assessed blinding separately for different outcomes or classes of outcomes.

We assessed the methods as:

high, low or unclear risk of bias for participants;
high, low or unclear risk of bias for personnel;
high, low or unclear risk of bias for outcome assessors.

(4) Incomplete outcome data (checking for possible attrition bias through withdrawals, dropouts, protocol deviations)

We have described for each included study, and for each outcome or class of outcomes, the completeness of data including attrition and exclusions from the analysis. We have stated whether attrition and exclusions were reported, the numbers included in the analysis at each stage (compared with the total randomised participants), reasons for attrition or exclusion where reported, and whether missing data were balanced across groups or were related to outcomes. Where sufficient information was reported, or supplied by the trial authors, we re‐included missing data in the analyses. We assessed methods as:

low risk of bias;
high risk of bias;
unclear.

Studies where more than 20% of the participants were lost to follow‐up have been excluded. We also excluded outcomes where there was more than 20% missing data. We considered that serious sample attrition meant that studies were at high risk of bias and results would be difficult to interpret.

(5) Selective reporting bias

We have described for each included study how we investigated the possibility of selective outcome reporting bias and what we found.

We assessed the methods as:

low risk of bias (where it was clear that all of the study’s pre‐specified outcomes and all expected outcomes of interest to the review have been reported);
high risk of bias (where not all the study’s pre‐specified outcomes have been reported; one or more reported primary outcomes were not pre‐specified; outcomes of interest were reported incompletely and so could not be used; study failed to include results of a key outcome that would have been expected to have been reported);
unclear.

(6) Other sources of bias

We have described for each included study any important concerns we had about other possible sources of bias, such as baseline imbalance.

We assessed whether each study was free of other problems that could put it at risk of bias:

high risk of other bias;
low risk of other bias;
unclear.

(7) Overall risk of bias

We made explicit judgements about whether studies are at high risk of bias, according to the criteria given in the Handbook (Higgins 2011). With reference to (1) to (6) above, we assessed the likely magnitude and direction of the bias and whether we considered it was likely to impact on the findings. We explored the impact of the level of bias through undertaking sensitivity analyses ‐ seeSensitivity analysis.

Measures of treatment effect

Dichotomous data

For dichotomous data, we have presented results as summary risk ratio with 95% confidence intervals.

Continuous data

For continuous data, we used the mean difference if outcomes were measured in the same way between trials. Where appropriate, we planned to use the standardised mean difference to combine trials that measured the same outcome, but used different methods.

Unit of analysis issues

Cluster‐randomised trials

None of the trials in this version of the review randomised clusters. In future updates if such trials are identified we will include them in the analyses along with individually‐randomised trials. We will adjust their sample sizes using the methods described in the Handbook (Higgins 2011) using an estimate of the intra‐cluster correlation co‐efficient (ICC) derived from the trial (if possible), from a similar trial or from a study of a similar population. If we use ICCs from other sources, we will report this and conduct sensitivity analyses to investigate the effect of variation in the ICC. If we identify both cluster‐randomised trials and individually‐randomised trials, we plan to synthesise the relevant information. We will consider it reasonable to combine the results from both if there is little heterogeneity between the study designs and the interaction between the effect of intervention and the choice of randomisation unit is considered to be unlikely.

We will also acknowledge heterogeneity in the randomisation unit and perform subgroup analysis to investigate the effects of the randomisation unit.

Cross‐over trials

We have not included cross‐over trials; this type of study design was not considered suitable for this type of intervention.

Other unit of analysis issues

Where appropriate, in multi‐arm trials with more than one arm examining the same betamimetic (e.g. different doses), we combined results for the betamimetics group to form a single pair‐wise comparison.

If we had identified any multi‐arm studies where two different betamimetics were compared with a control group (placebo or no treatment), we planned to divide the control group between the two experimental arms (dividing control event rate and sample size by two for dichotomous outcomes, and dividing the sample size by two and assuming the same mean and standard deviation for continuous outcomes). No such trials were identified for inclusion in this version of the review.

Dealing with missing data

For included studies, we noted levels of attrition. We explored the impact of including studies with high levels of missing data in the overall assessment of treatment effect by using sensitivity analysis.

For all outcomes, we carried out analyses, as far as possible, on an intention‐to‐treat basis, i.e. we attempted to include all participants randomised to each group in the analyses, and all participants were analysed in the group to which they were allocated, regardless of whether or not they received the allocated intervention. The denominator for each outcome in each trial was the number randomised minus any participants whose outcomes were known to be missing.

Assessment of heterogeneity

We assessed statistical heterogeneity in each meta‐analysis using the Tau², I² and Chi² statistics. We regarded heterogeneity as substantial if a Tau² was greater than zero and either an I² was greater than 30% or there was a low P value (less than 0.10) in the Chi² test for heterogeneity.

Assessment of reporting biases

Where there were 10 or more studies in the meta‐analysis, we investigated reporting biases (such as publication bias) using funnel plots. We assessed funnel plot asymmetry visually, and if there was any obvious asymmetry identified we planned to use formal tests for funnel plot asymmetry (Higgins 2011).

Data synthesis

We carried out statistical analysis using the Review Manager software (RevMan 2012). We used fixed‐effect meta‐analysis for combining data where it was reasonable to assume that studies were estimating the same underlying treatment effect: i.e. where trials examined the same intervention, and the trials’ populations and methods were judged sufficiently similar. If there was clinical heterogeneity sufficient to expect that the underlying treatment effects differed between trials, or if substantial statistical heterogeneity was detected, we used random‐effects meta‐analysis to produce an overall summary, if an average treatment effect across trials was considered clinically meaningful. The random‐effects summary was treated as the average range of possible treatment effects and we have discussed the clinical implications of treatment effects differing between trials. If the average treatment effect was not considered clinically meaningful, we did not combine trials.

For random‐effects analyses, the results are presented as the average treatment effect with its 95% confidence interval, and the estimates of Tau² and I².

Subgroup analysis and investigation of heterogeneity

We planned to carry out subgroup analysis of trials where antenatal corticosteroids were given routinely to see if this had any impact on the treatment effect.

Studies that explicitly stated that both the intervention and control participants received corticosteroids concurrently with the betamimetics, placebo or no treatment compared with studies that did not mention the use of steroids routinely or where steroid administration rate was reported as 50% or less.

Subgroup analysis was restricted to the review’s primary outcomes.

We planned to assess differences between subgroups by interaction tests available in RevMan.

Sensitivity analysis

For primary outcomes we carried out sensitivity analysis by risk of bias, temporarily excluding trials with unclear or high risk of bias for allocation concealment from the analysis to examine the impact on results.

Results

Description of studies

One hundred and five reports were identified by the search strategy, equating to 93 separate studies (some trials resulted in multiple publications). Of these 93 studies, 28 were eligible for inclusion, although findings are based on 20 trials which contributed data. Eight trials that were otherwise eligible for inclusion did not report outcomes relevant to the review, and did not contribute data. Further information on these trials is set out in Characteristics of included studies tables but they are not otherwise discussed in the results below (Kovacs 1987; Kullander 1985; Pasargiklian 1983; Penney 1980; Philipsen 1981; Spellacy 1974; Spellacy 1978; Thoulon 1982).

Sixty‐four trials were excluded and one is awaiting classification pending further information from the trial authors.

Excluded studies

Trials were excluded for the following reasons: method of generation of allocation was not truly randomised or inadequate (Calder 1985; Caritis 1991; Castillo 1988; Castren 1975; Chhabra 1998; Csapo 1977; Das 1969; Gonik 1988; Gummerus 1981; Karlsson 1980; Kim 1983; Kosasa 1985; Lenz 1985; Park 1982; Ritcher 1975; Ryden 1977; Sciscione 1993; Sirohiwal 2001; Sivasamboo 1972; Spatling 1989; Weisbach 1986; Zeller 1986); it was not clear that all participants were in preterm labour (Dunlop 1986; Guinn 1997; Hallak 1992; Hallak 1993; Leake 1983; Levy 1985; Lipshitz 1976; Muller‐Holve 1987; Sanchez 1972; Tarnow‐Mordi 1988); the interventions were not consistent with the protocol (Ally 1992; Bedoya 1972; Besinger 1991; Bulgay‐Moerschel 2008; Caballero 1979; Cararach 2006; Dellenbach 1971; Ferguson 1987; Francioli 1988; Gamissans 1978; Gamissans 1982; Goel 2011; Hatjis 1987; Herzog 1995; Ieda 1991; Kanayama 1996; Katz 1983; Lyell 2005; Neri 2009; Raymajhi 2003; Reynolds 1978; Rios‐Anez 2001; Ritcher 1979; Ross 1983; Trabelsi 2008); some of the comparisons examined in these trials are examined in other Cochrane reviews. Loss to follow‐up exceeded 20% in one study (Beall 1985), and one paper reported results from a series of trials which were reported individually in included studies (Merkatz 1980).

In addition, some studies, which were included in the previous versions of the review, were excluded from this updated version due to several reasons: loss to follow‐up 20% or more (Garite 1987; Howard 1982; Larsen 1986); conducted in preterm premature rupture of membranes without preterm labour (Christensen 1980); and data were reported by episode of preterm labour (Wesselius‐De 1971).

Please see table of Characteristics of excluded studies for further details.

Included studies

Twenty randomised controlled trials examining the effect of betamimetics for inhibiting preterm labour contributed data to the review.

A total of 1367 women participated in the 12 included trials comparing betamimetics with placebo or no treatment (Adam 1966; Barden 1980; Cotton 1984; CPLG 1992; Hobel 1980; Ingemarsson 1976; Larsen 1980; Leveno 1986; Mariona 1980; Scommegna 1980; Spellacy 1979; Tohoku 1984). Data for four trials (Barden 1980; Hobel 1980; Mariona 1980; Scommegna 1980) were included in a previous version of the review (King 1988) and were extracted from personal communications from trial authors.

A further 948 women participated in the five trials of comparing other betamimetics with ritodrine (Caritis 1984; Essed 1978; Holleboom 1996; Lipshitz 1988; Von Oeyen 1990); in two of these trials results were reported in brief abstracts only (Lipshitz 1988; Von Oeyen 1990).

One trial with 140 women compared hexoprenaline and salbutamol (Gummerus 1983), Motazedian 2010 compared terbutaline with salbutamol in a trial with 200 women, and Crepin 1989 (60 women) examined slow versus normal release salbutamol.

(1) Study location and settings

Trials were published over a 44‐year period between 1966 (Adam 1966) and 2010 (Motazedian 2010). Included studies were conducted in tertiary care or University hospitals in high‐income settings: trials were conducted in the USA (Barden 1980; Caritis 1984; Cotton 1984; Hobel 1980; Leveno 1986; Lipshitz 1988; Mariona 1980; Scommegna 1980; Spellacy 1979; Von Oeyen 1990), Europe (Crepin 1989; Essed 1978; Gummerus 1983; Holleboom 1996; Ingemarsson 1976; Larsen 1980), Australia (Adam 1966), Canada (CPLG 1992), Iran (Motazedian 2010) and Japan (Tohoku 1984). Eight trials were multicentre studies.

(2) Participants

The participants included in these trials were reasonably homogeneous. The gestational age at inclusion ranged from 20 to up to 37 weeks. Preterm labour was reasonably consistently defined across the trials as the presence of uterine contractions documented by external tocography or cervical dilatation up to 4 to 5 cm, or both, except for three trials, which did not state the definition (Gummerus 1983; Lipshitz 1988; Von Oeyen 1990). Three trials included women admitted for preterm labour with preterm premature rupture of membranes (Caritis 1984; Cotton 1984; CPLG 1992) and three trials included twins (Adam 1966; Cotton 1984; CPLG 1992). Four trials did not state the exclusion criteria (Essed 1978; Gummerus 1983; Lipshitz 1988; Von Oeyen 1990). Most trials excluded those women with chorioamnionitis, pregnancy‐induced hypertension, vaginal bleeding, intrauterine compromise or fetal death, and contraindications to betamimetics.

(3) Interventions

Nine trials compared ritodrine with placebo (Barden 1980; CPLG 1992; Hobel 1980; Larsen 1980; Leveno 1986; Mariona 1980; Scommegna 1980; Spellacy 1979, Tohoku 1984). Intravenous ritodrine was used initially and maintained orally in all trials except one (Larsen 1980), which used intramuscular ritodrine initially. The dosage was consistent both intravenously and orally. Ritodrine was usually started at 100 mcg/minute and increased by 50 mcg/minute up to a maximum of 350 mcg/minute, or until contractions ceased, or there were intolerable adverse effects. Two trials compared terbutaline with placebo (Cotton 1984; Ingemarsson 1976); the same regimen for intravenous infusion was used but oral terbutaline was only given in one trial (Ingemarsson 1976). One trial compared isoxsuprine with placebo (Adam 1966). Two trials compared terbutaline with ritodrine (Caritis 1984; Von Oeyen 1990). Only one trial compared other betamimetics with ritodrine: fenoterol with ritodrine (Essed 1978); ritodrine loading dose with incremental dose (Holleboom 1996); and hexoprenaline with ritodrine (Lipshitz 1988). An additional study compared hexoprenaline with salbutamol (Gummerus 1983) while Motazedian 2010 compared terbutaline with salbutamol, and Crepin 1989 compared slow with normal release salbutamol.

(4) Outcomes

There was some inconsistency across the trials with regards to the way in which outcomes were reported. Only three common primary outcomes were consistently reported: birth within 48 hours, perinatal death, and respiratory distress syndrome. However, the definition of respiratory distress syndrome was not stated. Only the biggest trial reported neonatal outcomes both in the short‐ and long‐term (CPLG 1992). Secondary outcomes were less consistently reported and definitions were not stated. Mainly cardiovascular adverse effects from betamimetics were reported, particularly when treatment was stopped due to adverse reactions. Other adverse effects were less consistently reported.

Please see tables of Characteristics of included studies for further details.

Risk of bias in included studies

Summaries of 'Risk of bias' assessments for individual included studies, and for all studies are set out in Figure 1 and Figure 2.

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

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

Allocation

The method of generating the randomisation sequence was not described or was not clear in eight of these studies; 12 studies were assessed to be at low risk of bias for sequence generation. In these studies the sequence was determined by random number tables, by computer, or by an external randomisation service (Barden 1980; Caritis 1984; CPLG 1992; Hobel 1980; Kovacs 1987; Leveno 1986; Mariona 1980; Motazedian 2010; Scommegna 1980; Spellacy 1978; Spellacy 1979; Tohoku 1984).

Seven of the included trials reported adequate concealment of allocation (Caritis 1984; CPLG 1992; Holleboom 1996; Ingemarsson 1976; Leveno 1986; Motazedian 2010; Tohoku 1984). The method of concealment of allocation was unclear in the remaining trials (Adam 1966; Barden 1980; Cotton 1984; Crepin 1989; Essed 1978; Gummerus 1983; Hobel 1980; Kovacs 1987; Kullander 1985; Larsen 1980; Lipshitz 1988; Mariona 1980; Pasargiklian 1983; Penney 1980; Philipsen 1981; Scommegna 1980; Spellacy 1974; Spellacy 1978; Spellacy 1979; Thoulon 1982; Von Oeyen 1990). All these trials were conducted before 1990, the majority more than 25 years ago. More information on allocation of concealment could not be obtained from the authors. We decided to include these papers.

Blinding

Method of blinding (double‐blind and single‐bind) was reported in 15 trials (Adam 1966; Barden 1980; Caritis 1984; CPLG 1992; Crepin 1989; Hobel 1980; Ingemarsson 1976; Mariona 1980; Penney 1980; Philipsen 1981; Scommegna 1980; Spellacy 1974; Spellacy 1978; Spellacy 1979; Tohoku 1984), In some trials women were described as blinded, but it is unlikely that staff would have been unaware of treatment allocation as regimens in different arms of trials were not identical; other trials did not discuss blinding at all.

Incomplete outcome data

In this review, we have attempted to conduct an intention‐to‐treat analysis for all outcomes. All or most participants in 15 trials completed the study and most participants were accounted for in the analysis (Barden 1980; Caritis 1984; Cotton 1984; CPLG 1992; Hobel 1980; Holleboom 1996; Ingemarsson 1976; Leveno 1986; Mariona 1980; Motazedian 2010; Pasargiklian 1983; Philipsen 1981; Scommegna 1980; Spellacy 1979; Tohoku 1984). In two trials attrition may have related to treatment group and these trials were judged to be at high risk of bias for this domain (Crepin 1989; Essed 1978). In the remaining trials loss to follow‐up was not mentioned or not clear.

Selective reporting

Assessing selective reporting bias without access to trial protocols is difficult, and most studies were assessed as unclear. In the study by Motazedian 2010, outcomes for infants were only reported for those born after 37 weeks' gestation; data for infant outcomes were therefore at high risk of bias and have not been included in the data and analysis tables in the review.

Other potential sources of bias

In two studies there appeared to be baseline differences between groups (CPLG 1992; Essed 1978), and in one study there was considerable deviation from the protocol (Holleboom 1996).

Effects of interventions

1. Betamimetics compared with placebo (comparison 1)

Twelve trials compared betamimetics with placebo; trials included a total of 732 women in placebo groups and 635 women in betamimetics group. One trial (Larsen 1980) randomised the participants into four groups: long ritodrine infusion group, short ritodrine infusion group, intramuscular group and placebo. Data from all women on ritodrine regimens (three groups) were pooled together to compare with placebo‐treated women.

Primary outcomes

Birth within 48 hours of treatment

The use of betamimetics significantly decreased the number of women giving birth within 48 hours (average risk ratio (RR) 0.68; 95% confidence interval (CI) 0.53 to 0.88, 10 trials, 1209 women) (Analysis 1.1). There was moderate heterogeneity for this outcome (I² = 32%) and we used a random‐effects model. The effect of betamimetics on reduction of birth within 48 hours was not changed after sensitivity analysis, retaining only studies with adequate allocation concealment (CPLG 1992; Ingemarsson 1976; Leveno 1986) (average RR 0.55; 95% CI 0.38 to 0.80) (data not shown). There was no obvious funnel plot asymmetry for this outcome (Figure 3).

1.1 — Comparison 1 All betamimetics versus placebo, Outcome 1 Birth within 48 hours of treatment.

Funnel plot of comparison: 1 All betamimetics versus placebo, outcome: 1.1 Birth within 48 hours of treatment.

Perinatal death (up to seven days) and serious neonatal morbidity

Betamimetics were not shown to reduce perinatal deaths (RR 0.84; 95% CI 0.46 to 1.55, 11 trials, 1332 infants) (Analysis 1.2), respiratory distress syndrome (RR 0.87; 95% CI 0.71 to 1.08, eight trials, 1239 infants) (Analysis 1.3), or cerebral palsy (RR 0.19; 95% CI 0.02 to 1.63, one trial, 246 infants) (Analysis 1.4), Differences between groups remained non‐significant when we carried out sensitivity analysis retaining only those studies at low risk of bias for allocation concealment (data not shown).Visual assessment suggested no obvious funnel plot asymmetry for perinatal death (Figure 4).

1.2 — Comparison 1 All betamimetics versus placebo, Outcome 2 Perinatal death (7 days).

1.3 — Comparison 1 All betamimetics versus placebo, Outcome 3 Respiratory distress syndrome.

1.4 — Comparison 1 All betamimetics versus placebo, Outcome 4 Cerebral palsy.

Funnel plot of comparison: 1 All betamimetics versus placebo, outcome: 1.2 Perinatal death (7 days).

Neonatal length of hospital stay

Only one trial reported the mean neonatal length of hospital stay (Leveno 1986) with the total length of hospital stay being 24.5 days in the ritodrine group and 24.8 days in the placebo group. The mean neonatal length of intensive care stay was 14.3 days in the ritodrine group and 12.4 days in the placebo group. (Standard deviations were not provided in the trial report and data for these outcomes are not shown in data and analyses tables.)

Maternal secondary outcomes

Betamimetics decreased the frequency of birth within seven days (average RR 0.80; 95% CI 0.65 to 0.98, five trials, 911 women) (Analysis 1.5), but did not significantly reduce the risk of delivering prior to 37 weeks' gestation (RR 0.95; 95% CI 0.88 to 1.03, 10 trials, 1212 women) (Analysis 1.6).

1.5 — Comparison 1 All betamimetics versus placebo, Outcome 5 Birth within 7 days.

1.6 — Comparison 1 All betamimetics versus placebo, Outcome 6 Birth less than 37 weeks' gestation.

No subgroup data based on birth before 28 and before 34 completed weeks were reported because most participants were 32 weeks' gestation or more.

No maternal deaths were reported in two trials that explicitly reported this (CPLG 1992; Larsen 1980). Other serious maternal outcomes, for example, cardiac arrest, respiratory arrest and admission to intensive care unit, were not reported in any trial. Maternal pulmonary oedema was reported in three trials; only one case was identified (Cotton 1984; CPLG 1992; Leveno 1986) (RR 3.03; 95% CI 0.12 to 74.23) (Analysis 1.8). There were no significant differences between groups for maternal cardiac arrhythmias (RR 3.54; 95% CI 0.74 to 16.92, one trial, 708 women) (Analysis 1.9); myocardial ischaemia (RR 12.53; 95% CI 0.72 to 216.91, one trial, 106 women) (Analysis 1.10), or maternal hypotension (average RR 1.56; 95% 0.12 to 20.86, two trials, 136 women) (Analysis 1.11).

1.8 — Comparison 1 All betamimetics versus placebo, Outcome 8 Pulmonary oedema.

1.9 — Comparison 1 All betamimetics versus placebo, Outcome 9 Cardiac arrhythmias.

1.10 — Comparison 1 All betamimetics versus placebo, Outcome 10 Myocardial ischemia.

1.11 — Comparison 1 All betamimetics versus placebo, Outcome 11 Hypotension.

The frequency of other maternal adverse effects varied in different trials but overall, were significantly increased with betamimetics compared to placebo: cessation of treatment due to adverse drug reaction (RR 11.38; 95% CI 5.21 to 24.86, five trials, 1081 women) (Analysis 1.12); chest pain (RR 11.29; 95% CI 3.81 to 33.46, two trials, 814 women) (Analysis 1.15); dyspnoea (RR 3.86; 95% CI 2.21 to 6.77, two trials, 814 women) (Analysis 1.16); palpitation (RR 9.91; 95% CI 6.46 to 15.20, five trials, 1089 women) (Analysis 1.13); tremor (RR 10.74; 95% CI 6.20 to 18.59, one trial 708 women) (Analysis 1.17); headaches (RR 4.07; 95% CI 2.60 to 6.35, three trials, 936 women) (Analysis 1.22); hypokalaemia (RR 6.07; 95% CI 4.00 to 9.20, one trial, 708 women) (Analysis 1.19); hyperglycaemia (RR 2.90; 95% CI 2.05 to 4.09, one trial, 708 women) (Analysis 1.18); nausea or vomiting (RR 1.76; 95% CI 1.29 to 2.42, three trials, 932 women) (Analysis 1.20); and nasal stuffiness (RR 2.90; 95% CI 1.64 to 5.12, one trial, 708 women) (Analysis 1.21). Two trials reported tachycardia; in one trial there was a significant difference between groups, while in the second trial, similar numbers of women in both groups reported this side effect and we have not pooled results for this outcome (Analysis 1.14).

1.12 — Comparison 1 All betamimetics versus placebo, Outcome 12 Cessation of treatment due to adverse drug reaction.

1.15 — Comparison 1 All betamimetics versus placebo, Outcome 15 Chest pain.

1.16 — Comparison 1 All betamimetics versus placebo, Outcome 16 Dyspnoea.

1.13 — Comparison 1 All betamimetics versus placebo, Outcome 13 Palpitation.

1.17 — Comparison 1 All betamimetics versus placebo, Outcome 17 Tremor.

1.22 — Comparison 1 All betamimetics versus placebo, Outcome 22 Headaches.

1.19 — Comparison 1 All betamimetics versus placebo, Outcome 19 Hypokalaemia.

1.18 — Comparison 1 All betamimetics versus placebo, Outcome 18 Hyperglycaemia.

1.20 — Comparison 1 All betamimetics versus placebo, Outcome 20 Nausea or vomiting.

1.21 — Comparison 1 All betamimetics versus placebo, Outcome 21 Nasal stuffiness.

1.14 — Comparison 1 All betamimetics versus placebo, Outcome 14 Tachycardia.

Neonatal secondary outcomes

There were no statistically significant differences between groups for neonatal death (average RR 0.90; 95% CI 0.27 to 3.00, six trials, 1174 infants) (Analysis 1.23), infant death (RR 0.51; 95% CI 0.05 to 5.64, one trial, 750 women) (Analysis 1.24) and necrotising enterocolitis (RR 0.42; 95% CI 0.06 to 2.78, two trials, 149 infants) (Analysis 1.25). Significant heterogeneity was identified for two outcomes: sepsis and fetal hypoglycaemia, and results were not pooled for these outcomes (Analysis 1.26; Analysis 1.27).Fetal tachycardia associated with betamimetics was significantly increased compared with placebo in one trial with data for 30 infants (RR 2.40; 95% CI 1.12 to 5.13) (Ingemarsson 1976) (Analysis 1.28).

1.23 — Comparison 1 All betamimetics versus placebo, Outcome 23 Neonatal death.

1.24 — Comparison 1 All betamimetics versus placebo, Outcome 24 Infant death.

1.25 — Comparison 1 All betamimetics versus placebo, Outcome 25 Necrotising enterocolitis.

1.26 — Comparison 1 All betamimetics versus placebo, Outcome 26 Sepsis or infection.

1.27 — Comparison 1 All betamimetics versus placebo, Outcome 27 Fetal hypoglycaemia.

1.28 — Comparison 1 All betamimetics versus placebo, Outcome 28 Fetal tachycardia.

No data on chronic lung disease, bronchopulmonary dysplasia, severe neuroradiological abnormality, and retinopathy of prematurity were reported. Only one trial reported abnormal long‐term neurodevelopmental status after follow‐up of 246 infants at 18 months (CPLG 1992). There was no significant difference between groups on the Bayley Psychomotor Development Index (mean difference (MD) 1.30; 95% CI ‐2.74 to 5.34) (Analysis 1.29). The average score on the Bayley Mental Developmental Index was slightly higher in the ritodrine group (MD 5.20; 95% CI 0.56 to 9.84) (Analysis 1.30).

1.29 — Comparison 1 All betamimetics versus placebo, Outcome 29 Infant long‐term neurological development (Bayley score: Psychomotor development).

1.30 — Comparison 1 All betamimetics versus placebo, Outcome 30 Infant long‐term neurological development (Bayley score: Mental development).

2. Other betamimetics compared with ritodrine

Terbutaline compared with ritodrine (comparison 2)

There were only two trials in this group involving 183 women: 90 in the terbutaline group and 93 in the ritodrine group (Caritis 1984; Von Oeyen 1990). Only 19 outcomes were reported.

Primary outcomes

Birth within 48 hours of treatment

There was no significant difference between terbutaline and ritodrine for the number of women giving birth within 48 hours in one trial with 83 women (RR 2.05; 95% confidence interval (CI) 0.77 to 5.48, one trial) (Analysis 2.1).

2.1 — Comparison 2 Terbutaline versus ritodrine, Outcome 1 Birth within 48 hours.

Perinatal death (up to seven days) and serious neonatal morbidity

There were no significant differences between the two betamimetics for perinatal deaths (no deaths in one trial with 83 infants) (Analysis 2.2) or respiratory distress syndrome (RR 1.99; 95% CI 0.93 to 4.27, one trial, 101 infants) (Analysis 2.3). Cerebral palsy was not reported.

2.2 — Comparison 2 Terbutaline versus ritodrine, Outcome 2 Perinatal death.

2.3 — Comparison 2 Terbutaline versus ritodrine, Outcome 3 Respiratory distress syndrome.

Neonatal length of hospital stay

Length of hospital stay was not reported in either trial contributing data to this comparison.

Secondary outcomes

Most of our secondary maternal and infant outcomes were not reported.

Hyperglycaemia was statistically significantly increased in the terbutaline group (RR 1.78; 95% CI 1.05 to 3.03) (Analysis 2.10). However, this outcome was reported in one trial with 100 women only (Caritis 1984). Other maternal and neonatal outcomes (measured in only one trial) were not statistically significantly different between the two groups: birth within seven days (RR 0.80; 95% CI 0.57 to 1.10) (Analysis 2.4); birth before 28 completed weeks (RR 2.08; 95% CI 0.55 to 7.87) (Analysis 2.5); or cessation of treatment due to adverse drug reactions (RR 0.83; 95% CI 0.24 to 2.92) (Analysis 2.6). Both trials reported chest pain, shortness of breath and hypotension; there were no significant differences between groups for any of these outcomes (RR 1.11; 95% CI 0.55 to 2.25; RR 0.83; 95% CI 0.41 to 1.67, and RR 1.00; 95% CI 0.67 to 1.49 respectively). Dyspnoea, arrhythmia, tachycardia, palpitation, and headache were reported in only one trial and there were no clear differences between terbutaline and ritodrine for these side effects.

2.10 — Comparison 2 Terbutaline versus ritodrine, Outcome 10 Hyperglycaemia or abnormal glucose tolerance test.

2.4 — Comparison 2 Terbutaline versus ritodrine, Outcome 4 Birth within 7 days.

2.5 — Comparison 2 Terbutaline versus ritodrine, Outcome 5 Birth less than 28 weeks' gestation.

2.6 — Comparison 2 Terbutaline versus ritodrine, Outcome 6 Cessation of treatment due to adverse drug reactions.

Fenoterol compared with ritodrine (comparison 3)

There was only one trial in this group, involving 48 women in each group (Essed 1978). Only six outcomes were reported.

Maternal and neonatal outcomes

Neonatal death was statistically significantly decreased in the fenoterol group (RR 0.13; 95% CI 0.02 to 0.96) (Analysis 3.6). However, the sample size was small and the methodological quality of the single study reporting this outcome was poor. There were no statistically significant differences between the groups for other maternal and neonatal outcomes reported: perinatal death; respiratory distress syndrome; tachycardia; fetal bradycardia; and hypoglycaemia.

3.6 — Comparison 3 Fenoterol versus ritodrine, Outcome 6 Neonatal death.

Ritodrine loading dose with conventional incremental dose (comparison 4)

There was only one trial in this group involving 203 women: 101 in the ritodrine loading dose group and 102 in the incremental dose group (Holleboom 1996). Only 12 outcomes were reported.

There were no statistically significant differences between the groups for maternal and neonatal outcomes: birth within 48 hours; birth before 34 completed weeks; birth before 37 completed weeks; respiratory distress syndrome; any maternal adverse effects; tachycardia; nausea or vomiting; palpitation; headache; neonatal death; periventricular haemorrhage grade three to four; and sepsis.

Hexoprenaline compared with ritodrine (comparison 5)

There was only one trial in this group involving 466 women: 314 in the hexoprenaline group and 152 in the ritodrine group (Lipshitz 1988). Only six outcomes were reported.

Maternal and neonatal adverse effects were statistically significantly decreased in the hexoprenaline group: cessation of treatment due to adverse reaction (RR 0.28; 95% CI 0.08 to 0.93) (Analysis 5.1); any maternal adverse effects (RR 0.83; 95% CI 0.76 to 0.91) (Analysis 5.2); nausea or vomiting (RR 0.63; 95% CI 0.45 to 0.89) (Analysis 5.5); palpitation (RR 0.75; 95% CI 0.60 to 0.94) (Analysis 5.3); hypotension (RR 0.77; 95% CI 0.61 to 0.96) (Analysis 5.4); increased fetal heart rate (RR 0.74; 95% CI 0.56 to 0.98) (Analysis 5.6). However, the quality of this trial was poor (abstract only).

5.1 — Comparison 5 Hexoprenaline versus ritodrine, Outcome 1 Cessation of treatment due to adverse drug reactions.

5.2 — Comparison 5 Hexoprenaline versus ritodrine, Outcome 2 Any maternal adverse effects.

5.5 — Comparison 5 Hexoprenaline versus ritodrine, Outcome 5 Nausea or vomiting.

5.3 — Comparison 5 Hexoprenaline versus ritodrine, Outcome 3 Palpitations.

5.4 — Comparison 5 Hexoprenaline versus ritodrine, Outcome 4 Hypotension.

5.6 — Comparison 5 Hexoprenaline versus ritodrine, Outcome 6 Increase in fetal heart rate.

Hexoprenaline compared with salbutamol (comparison 6)

There was only one trial in this group involving 140 women: 70 in the hexoprenaline group and 70 in the salbutamol group (Gummerus 1983). Only six outcomes were reported.

Maternal adverse effects were statistically significantly decreased in the hexoprenaline group: any maternal adverse effects (RR 0.38; 95% CI 0.18 to 0.80) ( Analysis 6.3) and tachycardia (RR 0.11; 95% CI 0.01 to 0.85) (Analysis 6.4). However, the methodological quality of this trial was poor. There were no statistically significant differences between the groups for respiratory distress syndrome, cessation of treatment due to adverse reactions, nausea or vomiting, headache, and tremor.

6.3 — Comparison 6 Hexoprenaline versus salbutamol, Outcome 3 Any maternal adverse effects.

6.4 — Comparison 6 Hexoprenaline versus salbutamol, Outcome 4 Tachycardia.

Subgroup analysis based on adjunctive steroid treatment was not done because of the low incidence of corticosteroid usage in all trials.

Terbutaline compared with salbutamol (comparison 7)

Only one trial with 200 women contributed data to this comparison (Motazedian 2010) and few of our primary and secondary outcomes were reported.

There was no significant difference between women receiving terbutaline and salbutamol for birth within 48 hours (RR 0.63; 95% CI 0.21 to 1.84).

Side effects were more likely in the salbutamol group. Women in the terbutaline group were at reduced risk of side effects overall (RR 0.54; 95% CI 0.42 to 0.70) (Analysis 7.2); and of tachycardia (RR 0.48; 95% CI 0.30 to 0.75) (Analysis 7.3); anxiety (RR 0.48; 95% CI 0.28 to 0.84) (Analysis 7.6); and chills (RR 0.48; 95% CI 0.34 to 0.69) (Analysis 7.7). There were no statistically significant differences between groups for dyspnoea (RR 0.42; 95% CI 0.15 to 1.14) (Analysis 7.4), nausea (RR 0.66; 95% CI 0.39 to 1.09) (Analysis 7.5), or oedema (RR 0.75; 95% CI 0.17 to 3.27) (Analysis 7.8).

7.2 — Comparison 7 Terbutaline versus salbutamol, Outcome 2 Side effects (overall).

7.3 — Comparison 7 Terbutaline versus salbutamol, Outcome 3 Tachycardia.

7.6 — Comparison 7 Terbutaline versus salbutamol, Outcome 6 Anxiety.

7.7 — Comparison 7 Terbutaline versus salbutamol, Outcome 7 Chills.

7.4 — Comparison 7 Terbutaline versus salbutamol, Outcome 4 Dyspnea.

7.5 — Comparison 7 Terbutaline versus salbutamol, Outcome 5 Nausea.

7.8 — Comparison 7 Terbutaline versus salbutamol, Outcome 8 Oedema.

The length of maternal hospital stay was reduced in the terbutaline group (MD ‐0.80 days; 95% CI ‐1.19 to ‐0.41) (Analysis 7.9).

7.9 — Comparison 7 Terbutaline versus salbutamol, Outcome 9 Duration of hospital admission (days).

Women who delivered during the treatment period, whose contractions did not cease within 48 hours, or who developed complications were not included in analysis for pregnancy outcomes or complications (44/200 – 22%) data for these outcomes have not been included in the review because of high risk of bias.

Comparison 8: Slow release salbutamol compared with normal release salbutamol

A single trial with data for 56 women was included in this comparison (Crepin 1989). None of our primary outcomes were reported.

For maternal secondary outcomes, there was no difference between the two types of salbutamol for preterm birth (before 37 weeks) (RR 2.57; 95% CI 0.29 to 23.13) (Analysis 8.1) or caesarean birth (RR 1.29; 95% CI 0.23 to 7.07) (Analysis 8.2). None of the women in either group had severe side effects leading to discontinuation of treatment, and there was no evidence of difference in the frequency of nausea (RR 0.86; 95% CI 0.06 to 12.98) (Analysis 8.4). For infant Apgar score, no infants (in either group) were reported to have a score less than seven at five minutes. Other outcomes were not reported.

8.1 — Comparison 8 Slow release versus normal release salbutamol, Outcome 1 Preterm birth (before 37 weeks).

8.2 — Comparison 8 Slow release versus normal release salbutamol, Outcome 2 Caesarean section.

8.4 — Comparison 8 Slow release versus normal release salbutamol, Outcome 4 Nausea.

Discussion

Summary of main results

Betamimetics were shown to be effective in delaying birth for 48 hours and up to seven days. This is sufficient time to allow transfer of a woman to a tertiary care unit or to complete a course of antenatal corticosteroid. However, there was no evidence that this delay in the timing of birth translated into any improvements in neonatal outcomes, and the side effects for the woman are considerable.

It may be that health gains have not been demonstrated for several reasons.

Time gain in‐utero: most participants were 32 weeks' gestation or more. Therefore, increased gestational age might have little or no effect on perinatal outcomes leading to no statistical significance in the pooled estimate of important outcomes such as respiratory distress syndrome, perinatal death, neonatal death and infant death. Also, prolongation of pregnancy itself may be disadvantage to the baby. If the eligibility criteria had been restricted to lower gestation groups, and had the number of participants been greater, an effect on clinical outcomes might have been detected.
Referral to higher care level: all trials in this review took place in tertiary care or University hospitals, that were likely to have neonatal intensive care facilities. If the trials had involved women who needed to be transferred to distant referral hospitals, again a difference in outcome might have been seen.
Antenatal corticosteroids administration: the majority of trials were conducted before 1990, when antenatal corticosteroids were not widely used. Even in CPLG 1992, antenatal corticosteroids were prescribed only to one third of all women in each arm. Therefore, subgroup analysis to evaluate the effect of betamimetics to allow a complete course of antenatal corticosteroids (betamimetics alone and betamimetics combined with corticosteroids) on neonatal outcomes could not be assessed.

There is not enough evidence to suggest that one betamimetic agent is superior to another.

Agreements and disagreements with other studies or reviews

Betamimetics have been used for primary tocolysis and maintenance therapy for many years, and continue to be used throughout the world. However, alternative agents have been shown to be more effective in terms of pregnancy prolongation and have been associated with fewer and less severe side effects compared with betamimetics. The effects of different classes of tocolytics compared with placebo, and with each other (where different types of tocolytics have been compared directly in trials) have been set out in a series of Cochrane Reviews (Bain 2013; Crowther 2002; Duckitt 2002; King 2003; King 2005; Papatsonis 2005; Su 2010), and the effects of combinations of different tocolytic drugs to treat preterm labour is the subject of a review currently in development (Nardin 2006). A recent review, which included both network and pair‐wise meta‐analysis, incorporated evidence from 95 trials examining a range of tocolytics (betamimetics, magnesium sulphate, calcium channel blockers, COX inhibitors, oxytocin receptor antagonists, nitric oxide donors and other drugs) (Haas 2012). The results of this review suggest that all types of tocolytics were better than placebo in terms of prolonging pregnancy beyond 48 hours, but that betamimetics performed less well than other types of tocolytics (e.g. COX inhibitors, and calcium channel blockers). It was not clear that any class of tocolytic was effective in reducing respiratory distress syndrome compared with placebo. The evidence regarding side effects suggested that COX inhibitors and oxytocin receptor antagonists were associated with fewer maternal side effects compared with other types of tocolytics including betamimetics.

Authors' conclusions

Implications for practice.

Betamimetics are widely used especially in resource‐poor countries, and take effect rapidly. Betamimetics are effective in delaying birth for 48 hours, sufficient time to allow the transfer of a woman to a higher level of care and to allow completion of an antenatal corticosteroid to facilitate fetal lung maturation. However, this benefit is balanced by more frequent unpleasant and sometimes potentially serious adverse effects which significantly increase the likelihood of having to discontinue betamimetic treatment.

Implications for research.

Any future studies should aim to improve diagnostic criteria for preterm labour and have sufficient power to detect clinically meaningful differences in important neonatal outcomes. However, given the less favourable adverse effect profile of betamimetics in comparison to other tocolytic drugs, it seems unlikely that further trials involving betamimetics will be conducted.

What's new

Date	Event	Description
31 December 2013	New citation required but conclusions have not changed	Review updated.
31 December 2013	New search has been performed	The review has been updated by a new review team. The review now includes data from 20 randomised trials, with data for three additional studies in this update (Crepin 1989; Motazedian 2010; Tohoku 1984). All included studies have been assessed for risk of bias using updated methods.

Date	Event	Description
1 October 2009	Amended	Search updated. Nine reports added to Studies awaiting classification (Bulgay‐Moerschel 2008a; Caballero 1979a; Cararach 2006a; Kanayama 1996a; Lyell 2005a; Neri 2009a; Roy 2006; Tohoku 1984a; Trabelsi 2008a).
1 September 2008	Amended	Converted to new review format.
20 May 2006	New search has been performed	We updated the search and identified four new studies for assessment. We included one new study (Gummerus 1983) and excluded three (Kovacs 1987a; Raymajhi 2003; Sirohiwal 2001). The conclusions remain unchanged.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Birth within 48 hours of treatment	10	1209	Risk Ratio (M‐H, Random, 95% CI)	0.68 [0.53, 0.88]
2 Perinatal death (7 days)	11	1332	Risk Ratio (M‐H, Fixed, 95% CI)	0.84 [0.46, 1.55]
3 Respiratory distress syndrome	8	1239	Risk Ratio (M‐H, Fixed, 95% CI)	0.87 [0.71, 1.08]
4 Cerebral palsy	1	246	Risk Ratio (M‐H, Fixed, 95% CI)	0.19 [0.02, 1.63]
5 Birth within 7 days	5	911	Risk Ratio (M‐H, Random, 95% CI)	0.80 [0.65, 0.98]
6 Birth less than 37 weeks' gestation	10	1212	Risk Ratio (M‐H, Fixed, 95% CI)	0.95 [0.88, 1.03]
7 Maternal death	2	907	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
8 Pulmonary oedema	3	852	Risk Ratio (M‐H, Fixed, 95% CI)	3.03 [0.12, 74.23]
9 Cardiac arrhythmias	1	708	Risk Ratio (M‐H, Fixed, 95% CI)	3.54 [0.74, 16.92]
10 Myocardial ischemia	1	106	Risk Ratio (M‐H, Fixed, 95% CI)	12.53 [0.72, 216.91]
11 Hypotension	2	136	Risk Ratio (M‐H, Random, 95% CI)	1.56 [0.12, 20.86]
12 Cessation of treatment due to adverse drug reaction	5	1081	Risk Ratio (M‐H, Fixed, 95% CI)	11.38 [5.21, 24.86]
13 Palpitation	5	1089	Risk Ratio (M‐H, Random, 95% CI)	9.91 [6.46, 15.20]
14 Tachycardia	2		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
15 Chest pain	2	814	Risk Ratio (M‐H, Fixed, 95% CI)	11.29 [3.81, 33.46]
16 Dyspnoea	2	814	Risk Ratio (M‐H, Fixed, 95% CI)	3.86 [2.21, 6.77]
17 Tremor	1	708	Risk Ratio (M‐H, Fixed, 95% CI)	10.74 [6.20, 18.59]
18 Hyperglycaemia	1	708	Risk Ratio (M‐H, Fixed, 95% CI)	2.90 [2.05, 4.09]
19 Hypokalaemia	1	708	Risk Ratio (M‐H, Fixed, 95% CI)	6.07 [4.00, 9.20]
20 Nausea or vomiting	3	932	Risk Ratio (M‐H, Fixed, 95% CI)	1.76 [1.29, 2.42]
21 Nasal stuffiness	1	708	Risk Ratio (M‐H, Fixed, 95% CI)	2.90 [1.64, 5.12]
22 Headaches	3	936	Risk Ratio (M‐H, Fixed, 95% CI)	4.07 [2.60, 6.35]
23 Neonatal death	6	1174	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.27, 3.00]
24 Infant death	1	750	Risk Ratio (M‐H, Fixed, 95% CI)	0.51 [0.05, 5.64]
25 Necrotising enterocolitis	2	149	Risk Ratio (M‐H, Fixed, 95% CI)	0.42 [0.06, 2.78]
26 Sepsis or infection	2		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
27 Fetal hypoglycaemia	3		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
28 Fetal tachycardia	1	30	Risk Ratio (M‐H, Fixed, 95% CI)	2.4 [1.12, 5.13]
29 Infant long‐term neurological development (Bayley score: Psychomotor development)	1	246	Mean Difference (IV, Fixed, 95% CI)	1.30 [‐2.74, 5.34]
30 Infant long‐term neurological development (Bayley score: Mental development)	1	246	Mean Difference (IV, Fixed, 95% CI)	5.20 [0.56, 9.84]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Birth within 48 hours	1	83	Risk Ratio (M‐H, Fixed, 95% CI)	2.05 [0.77, 5.48]
2 Perinatal death	1	83	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Respiratory distress syndrome	1	101	Risk Ratio (M‐H, Fixed, 95% CI)	1.99 [0.93, 4.27]
4 Birth within 7 days	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	0.80 [0.57, 1.10]
5 Birth less than 28 weeks' gestation	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	2.08 [0.55, 7.87]
6 Cessation of treatment due to adverse drug reactions	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	0.83 [0.24, 2.92]
7 Any maternal adverse effects	1	83	Risk Ratio (M‐H, Fixed, 95% CI)	0.95 [0.84, 1.07]
8 Chest pain	2	183	Risk Ratio (M‐H, Fixed, 95% CI)	1.11 [0.55, 2.25]
9 Shortness of breath or dyspnea	2	183	Risk Ratio (M‐H, Fixed, 95% CI)	0.83 [0.41, 1.67]
10 Hyperglycaemia or abnormal glucose tolerance test	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	1.78 [1.05, 3.03]
11 Palpitations	1	83	Risk Ratio (M‐H, Fixed, 95% CI)	1.18 [0.78, 1.79]
12 Tachycardia	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	0.66 [0.43, 1.00]
13 Arrhythmia	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	0.35 [0.04, 3.22]
14 Hypotension	2	183	Risk Ratio (M‐H, Fixed, 95% CI)	1.00 [0.67, 1.49]
15 Nausea or vomiting	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	1.50 [0.71, 3.20]
16 Headache	1	83	Risk Ratio (M‐H, Fixed, 95% CI)	0.48 [0.23, 0.99]
17 Anxiety	1	83	Risk Ratio (M‐H, Fixed, 95% CI)	1.08 [0.67, 1.75]
18 Necrotising enterocolitis	1	101	Risk Ratio (M‐H, Fixed, 95% CI)	0.53 [0.05, 5.67]
19 Neonatal death	2	184	Risk Ratio (M‐H, Fixed, 95% CI)	1.27 [0.42, 3.91]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Perinatal death	1	98	Risk Ratio (M‐H, Fixed, 95% CI)	0.11 [0.01, 2.01]
2 Respiratory distress syndrome	1	98	Risk Ratio (M‐H, Fixed, 95% CI)	2.0 [0.38, 10.42]
3 Tachycardia	1	96	Risk Ratio (M‐H, Fixed, 95% CI)	0.71 [0.35, 1.45]
4 Hypoglycaemia	1	98	Risk Ratio (M‐H, Fixed, 95% CI)	1.33 [0.31, 5.65]
5 Fetal bradycardia	1	98	Risk Ratio (M‐H, Fixed, 95% CI)	1.0 [0.06, 15.54]
6 Neonatal death	1	98	Risk Ratio (M‐H, Fixed, 95% CI)	0.13 [0.02, 0.96]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Birth within 48 hours	1	203	Risk Ratio (M‐H, Fixed, 95% CI)	1.07 [0.60, 1.91]
2 Respiratory distress syndrome	1	222	Risk Ratio (M‐H, Fixed, 95% CI)	0.71 [0.35, 1.41]
3 Periventricular haemorrhage grade 3‐4	1	222	Risk Ratio (M‐H, Fixed, 95% CI)	0.14 [0.01, 2.73]
4 Birth less than 34 weeks	1	203	Risk Ratio (M‐H, Fixed, 95% CI)	1.01 [0.70, 1.45]
5 Birth less than 37 weeks	1	203	Risk Ratio (M‐H, Fixed, 95% CI)	0.82 [0.60, 1.13]
6 Any maternal adverse effects	1	203	Risk Ratio (M‐H, Fixed, 95% CI)	0.69 [0.43, 1.11]
7 Palpitations	1	203	Risk Ratio (M‐H, Fixed, 95% CI)	0.50 [0.23, 1.13]
8 Tachycardia	1	203	Risk Ratio (M‐H, Fixed, 95% CI)	0.88 [0.33, 2.35]
9 Nausea or vomiting	1	203	Risk Ratio (M‐H, Fixed, 95% CI)	1.21 [0.38, 3.84]
10 Headache	1	203	Risk Ratio (M‐H, Fixed, 95% CI)	1.01 [0.06, 15.93]
11 Sepsis	1	222	Risk Ratio (M‐H, Fixed, 95% CI)	0.71 [0.23, 2.18]
12 Neonatal death	1	222	Risk Ratio (M‐H, Fixed, 95% CI)	0.11 [0.01, 2.04]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Cessation of treatment due to adverse drug reactions	1	466	Risk Ratio (M‐H, Fixed, 95% CI)	0.28 [0.08, 0.93]
2 Any maternal adverse effects	1	466	Risk Ratio (M‐H, Fixed, 95% CI)	0.83 [0.76, 0.91]
3 Palpitations	1	466	Risk Ratio (M‐H, Fixed, 95% CI)	0.75 [0.60, 0.94]
4 Hypotension	1	466	Risk Ratio (M‐H, Fixed, 95% CI)	0.77 [0.61, 0.96]
5 Nausea or vomiting	1	466	Risk Ratio (M‐H, Fixed, 95% CI)	0.63 [0.45, 0.89]
6 Increase in fetal heart rate	1	466	Risk Ratio (M‐H, Fixed, 95% CI)	0.74 [0.56, 0.98]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Respiratory distress syndrome	1	140	Risk Ratio (M‐H, Fixed, 95% CI)	1.75 [0.54, 5.71]
2 Cessation of treatment due to adverse drug reactions	1	140	Risk Ratio (M‐H, Fixed, 95% CI)	1.0 [0.06, 15.67]
3 Any maternal adverse effects	1	140	Risk Ratio (M‐H, Fixed, 95% CI)	0.38 [0.18, 0.80]
4 Tachycardia	1	140	Risk Ratio (M‐H, Fixed, 95% CI)	0.11 [0.01, 0.85]
5 Nausea or vomiting	1	140	Risk Ratio (M‐H, Fixed, 95% CI)	1.0 [0.34, 2.95]
6 Headache	1	140	Risk Ratio (M‐H, Fixed, 95% CI)	0.33 [0.04, 3.13]
7 Tremor	1	140	Risk Ratio (M‐H, Fixed, 95% CI)	0.14 [0.01, 2.72]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 BIrth within 48 hours	1	200	Risk Ratio (M‐H, Fixed, 95% CI)	0.63 [0.21, 1.84]
2 Side effects (overall)	1	200	Risk Ratio (M‐H, Fixed, 95% CI)	0.54 [0.42, 0.70]
3 Tachycardia	1	200	Risk Ratio (M‐H, Fixed, 95% CI)	0.48 [0.30, 0.75]
4 Dyspnea	1	200	Risk Ratio (M‐H, Fixed, 95% CI)	0.42 [0.15, 1.14]
5 Nausea	1	200	Risk Ratio (M‐H, Fixed, 95% CI)	0.66 [0.39, 1.09]
6 Anxiety	1	200	Risk Ratio (M‐H, Fixed, 95% CI)	0.48 [0.28, 0.84]
7 Chills	1	200	Risk Ratio (M‐H, Fixed, 95% CI)	0.48 [0.34, 0.69]
8 Oedema	1	200	Risk Ratio (M‐H, Fixed, 95% CI)	0.75 [0.17, 3.27]
9 Duration of hospital admission (days)	1	200	Mean Difference (IV, Fixed, 95% CI)	‐0.80 [‐1.19, ‐0.41]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Preterm birth (before 37 weeks)	1	52	Risk Ratio (M‐H, Fixed, 95% CI)	2.57 [0.29, 23.13]
2 Caesarean section	1	52	Risk Ratio (M‐H, Fixed, 95% CI)	1.29 [0.23, 7.07]
3 Side effects leading to discontinuation of treatment	1	52	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Nausea	1	52	Risk Ratio (M‐H, Fixed, 95% CI)	0.86 [0.06, 12.98]
5 Apgar score less than 7 at 5 minutes	1	52	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]

Methods	Single‐centred randomised controlled trial, 2 arms with individual randomisation.
Participants	Location: Australia, University hospital.  Timeframe: 1963‐1964.  Eligible criteria: pregnant women less than 37 weeks' gestation with preterm labour and; moderate to strong uterine contractions; cervical dilatation not more than 4‐5 cm; intact membranes; minimal antepartum haemorrhage, if present; twins.  Exclusion criteria: advanced cervical dilatation; rupture membranes; accidental antepartum haemorrhage.  Total recruited: 44 women, 22 to isoxsuprine (vs) 22 to placebo.
Interventions	1) Isoxsuprine.  Dose: intramuscular injection every 3 hrs (average 80 mg in first 24 hrs) until contraction ceased for 36 hrs.  Maintenance: 40‐60 mg oral/day.  2) Placebo.
Outcomes	Primary outcomes: perinatal death (7 days).  Secondary outcomes: neonatal death.
Notes	Prenatal corticosteroid use: not stated.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	This was described as a double‐blind trial but there was no information on randomisation methods.
Allocation concealment (selection bias)	Unclear risk	This was a placebo‐controlled double‐blind trial. It was not clear whether study medications were identical. Methods of allocation at the point of randomisation were not described.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	This was a described as a double‐blind trial.
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not clear whether or not those collecting outcome data were aware of treatment allocation.
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	It was not entirely clear how many women were randomised to each group. One table suggests that there were 27 in the intervention group and 24 in the control group, later it said that 48 women were included in the analysis. Women who had not given birth at the end of the trial period were excluded from the analysis (2 in each group).
Selective reporting (reporting bias)	Unclear risk	It appeared that some analysis may have been carried out retrospectively as outcomes were not set out in the methods. The main outcome (prolongation of pregnancy) was reported as a mean without SD.
Other bias	Unclear risk	There was very little information on study methods. There were 5 multiple pregnancies in the intervention group and 2 in the control group.

Methods	Randomised controlled trial (part of multicentre study described in Merkatz 1980).
Participants	Location: USA.  Timeframe: 1972‐1977.  Eligible criteria: pregnant women between 20‐36 weeks' gestation with preterm labour (defined as regular contraction at least 1/10 min for 30 mins with cervical change).  Exclusion criteria: active vaginal bleeding; dilatation > 5 cm; fever or severe maternal or fetal diseases.  Total recruited: 25 women, 12 to ritodrine (vs) 13 to control.
Interventions	1) Ritodrine.  Dose: 100 mcg/min iv, increased 50 mcg/min (maximum 350 mcg/min).  Duration: 12 hrs then 5‐10 mg im every 3‐8 hrs for 24 hrs.  Maintenance: 10‐20 mg oral 3‐8 times/day (maximum 120 mg/day) until 38 weeks.  2) Placebo (identical).
Outcomes	Primary outcomes: birth within 48 hrs; RDS; perinatal death (7 days).  Secondary outcomes: birth < 37 weeks.
Notes	Prenatal corticosteroid use: not stated.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random number table.
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Described as double‐blind.
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not described.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No loss to follow‐up.
Selective reporting (reporting bias)	Unclear risk	Risk of bias was assessed by the previous review team from a personal communication from the trials authors. This was not available to the current review team and so we have used the previous assessments.
Other bias	Unclear risk	Not able to assess.

Methods	Single‐centre randomised trial, 2 arms with individual randomisation.
Participants	Location: USA.  Timeframe: not stated.  Eligible criteria: pregnant women from 23 to 36 weeks' gestation with preterm labour with intact or ruptured membranes.  Total recruited: 100 women, 49 to terbutaline (vs) 51 to ritodrine.
Interventions	1) Terbutaline (in physiologic saline with concentration of terbutaline 20 mcg/mL).  Dose: intravenous infusion 0.125 mL/min (2.5 mcg/min) and increased by 0.125 mL/min every 20 mins (maximum infusion rate 17.5 mcg/min).  Maintenance: maintained infusion rate for 1 hr then reduced 0.125 mL/min 10 mg every 20 mins until maintained adequate uterine inhibition, 1.5 hrs before intravenous infusion stopped, 1 capsule of 5 mg terbutaline oral every 4 hrs for 5 days in women with intact membranes (maximum 30 mg).  2) Ritodrine (in physiologic saline with concentration of ritodrine 400 mcg/mL.  Dose: intravenous infusion 0.125 mL/min (50 mcg/min) and increased by 0.125 mL/min every 20 mins (maximum infusion rate 350 mcg/min. After labour stopped.  Maintenance: maintain infusion rate for 1 hr then reduced 0.125 mL/min 10 mg every 20 mins until maintained adequate uterine inhibition, 1.5 hrs before intravenous infusion stopped, 1 capsule of 20 mg ritodrine oral every 4 hrs for 5 days in women with intact membranes (maximum 120 mg).
Outcomes	Primary outcomes: RDS.  Secondary outcomes: birth within 7 days; birth < 28 weeks; adverse effects; NEC; neonatal death.
Notes	Prenatal corticosteroid use: not stated.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Table of random numbers.
Allocation concealment (selection bias)	Low risk	This was described as a double‐blind trial. Study drugs were prepared in pharmacy and the same volume and type of diluent (saline) was used so would have looked similar.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	This was described as a double‐blind trial. It was stated that investigators and staff were not aware of which treatment arm women were in.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	“all outcomes were ascertained by personnel blinded to the women’s treatment assignment”.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	“All the enrolled patients were included in the analysis”. 5 women (0.7% did not receive the assigned treatment, but were included in an intention‐to‐treat analysis. Unclear for length of prolongation. 8/100 women randomised were withdrawn from the treatment due to side effects, and the length of prolongation was up to the time of treatment withdrawal rather than the birth (so this outcome may be more susceptible to bias – similar numbers of women in the two arms were withdrawn for side effects).
Selective reporting (reporting bias)	Unclear risk	Assessment from published study report.
Other bias	Low risk	Not apparent. Groups were reported to be similar at baseline.

Methods	Single‐centre randomised controlled trial, individual randomisation.
Participants	Location: USA, University Hospital.  Timeframe: not stated.  Eligible criteria: pregnant women between 26‐34 weeks with preterm labour (defined as persistent 3 contractions/10 mins and after 1 L of 5% dextrose in lactated Ringer's solution over one‐half hr showed progressive cervical dilatation or dilatation 2 cm or more or effacement 80% or more or spontaneous rupture of membranes).  Exclusion criteria: cervical dilatation more than 4 cm.  Total recruited: 54 women, 19 to terbutaline (vs) 19 to placebo (vs) 16 to magnesium sulphate.
Interventions	1) Terbutaline.  Dose: start with 9.2 mcg/min iv increased 5 mcg/min (maximum 25.3 mcg/min) until contraction ceased for 12 hrs (for women with ruptured membranes, the medication continued at least 48 hrs).  2) Placebo (5% dextrose in lactated Ringer's solution) iv 125 mL/hr.
Outcomes	Primary outcomes: birth within 48 hrs; RDS.  Secondary outcomes: neonatal infection (defined as positive cultures); intracranial haemorrhage; NEC; hypoglycaemia.
Notes	Prenatal corticosteroid use: 14 participants, distributed evenly among three groups.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Women were “randomized” method not described.
Allocation concealment (selection bias)	Unclear risk	Women were “randomized” method not described.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Blinding not mentioned. Staff would have been aware of study treatment allocations as iv regimens were different in the different arms of the trial.
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not mentioned but staff aware of treatment allocation.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	57 women randomised, 54 analysed, 1 women from each group required more than one tocolysis treatment and both were excluded.
Selective reporting (reporting bias)	Unclear risk	Little information on study methods. It was stated that 14 women received steroids for fetal lung maturation. It was not clear which groups these were in (reported as evenly distributed among the groups) or whether this related to fetal survival. assessment from published study report.
Other bias	Unclear risk	Very little information on methods. Groups appeared similar at baseline.

Methods	Type of study: multicentre randomised controlled trial (6 University hospitals).  Method of generation of allocation: stratified by gestational age (20‐23 wks, 24‐27 wks, 28‐31 wks, and 32‐35 wks) then randomisation for each gestational age by computerised random schedule controlled centrally by trial co‐ordinator.  Method of concealment of allocation: centrally by pharmacist.  Blinding: double‐blind.  Sample size calculation: not stated.  Intention‐to‐treat analyses: yes.  Losses to follow‐up: 5 (0.7%) from rapid progress of labour.
Participants	Location: Canada.  Timeframe: 1985‐1988 and follow‐up of infants until 1990.  Eligible criteria: pregnant women from 20 to 35 weeks' gestation with preterm labour (defined as presence of 4 uterine contraction per 20 mins or six per 60 mins documented by external tocography, or any uterine activity with either ruptured membranes or a cervical dilatation 2 cm or more, effacement at least 50%).  Exclusion criteria: suspected chorioamnionitis; fetal distress; serious vaginal bleeding; severe pre‐eclampsia; any condition necessitating immediate birth; lethally malformed or dead fetus; or any contraindication to use betamimetics.  Total recruited: 708 women, 352 to ritodrine (vs) 356 to placebo.
Interventions	1) Ritodrine (in physiologic saline with concentration of ritodrine 400 mcg/mL).  Dose: intravenous infusion was adjusted every 20 mins from 10 to 70 mL. After labour stopped, the effective dose was maintained for at least 6 hrs, then progressively decreased until the infusion was ended.  Maintenance: up to 12 tablets (10 mg/tablet) /day of ritodrine orally.  2)The dextrose solution alone and placebo tablet for maintenance.
Outcomes	Primary outcomes: birth within 48 hrs; perinatal death; abnormal long‐term neurodevelopmental delay; RDS.  Secondary outcomes: birth < 7 days; birth < 37 weeks; neonatal death; adverse effects.
Notes	Prenatal corticosteroid use: 34.6% in ritodrine group (vs) 36.0% in the placebo group.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computerised randomisation lists in each participating centre (randomisation stratified by gestational age).
Allocation concealment (selection bias)	Low risk	Placebo‐controlled trial with external randomisation by pharmacy at each centre.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Women, and staff were reported to be blind to treatment allocation.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Outcome assessment was reported to have been carried out by staff who were blind to women’s treatment group,
Incomplete outcome data (attrition bias)   All outcomes	Low risk	A small number of women did not receive the therapy assigned due to rapid labour progression (0.7%). All women randomised were included in the analysis.
Selective reporting (reporting bias)	Low risk	All outcomes mentioned in the trial registration were reported in the paper. There was no evidence of outcome reporting bias.
Other bias	Unclear risk	More women in the placebo group (19.2%) had cervical dilatation >=4 at randomisation compared with women in the treatment group (12.6%). Other characteristics at baseline appeared similar.

Methods	Described as a double‐blind, double‐placebo study, in 2 centres. RCT, 2 arms with individual randomisation
Participants	Setting: 2 hospitals in France. Inclusion criteria: 60 women recruited with threatened preterm labour (confirmed by observation and external tocography). Women were between 24‐37 weeks’ gestation. Exclusion criteria: women for whom betamimetic therapy was not suitable, or already receiving tocolytic treatment.
Interventions	60 women were randomised, 8 were excluded after randomisation. It was not clear how many were randomised to each group Experimental intervention: (28 analysed) Slow release salbutamol (iv over 48 hrs, oral after 24 hrs: slow release (2/days + 2 placebo) for further 4 days). 16 mg per day in 2 oral capsules. Comparison intervention: (24 analysed). Normal release salbutamol (iv over 48 hrs, oral after 24 hrs: with normal release (4/days) for further 4 days). 16 mg/day in 4 oral capsules
Outcomes	Birth within 4 days, birth before 36 completed weeks, CS, Apgar scores, side effects, pregnancy continuation (mean but no SDs reported).
Notes	Before the oral salbutamol was administered women received an average of 31.4mg (slow release) vs 38.9 mg (normal release). So total dose greater in the normal release group.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Original report in French. Sequence generation not described. This was described as a randomised study, but it was stated that women were allocated to groups after being matched (in the abstract and in translated notes).
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Described as double blind‐double placebo.
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not clear (translated notes).
Incomplete outcome data (attrition bias)   All outcomes	High risk	60 women were randomised, 8 were excluded after randomisation. It was not clear how many were randomised to each group. 52 included in the analysis. Others were not “ready to come off the intravenous drip within the specified timeframe”, possibly indicating that women who continued in preterm labour were systematically excluded.
Selective reporting (reporting bias)	Unclear risk	Assessment from published report and translated notes.
Other bias	Unclear risk	Little information on methods and assessment of risk of bias from translated notes.

Methods	Multicentre RCT in 5 hospitals. 2 arms with individual randomisation.
Participants	Location: The Netherlands, 5 teaching hospitals.  Timeframe: 4 years.  Eligible criteria: pregnant women with gestational age less than 34 weeks with preterm labour (defined as 3 or more regular contractions/15 mins and/or an increase in cervical dilatation).  Exclusion criteria: earlier tocolytic treatment in the current pregnancy; diagnosis of intrauterine infection; fetal anomaly.  Total recruited: 203 women 101 to loading dose (vs) 102 to conventional treatment.
Interventions	1) Loading dose ritodrine.  Dose: loading dose with 200 mg/min, no tocolysis in 2 hrs increased to 400 mcg/min, then no tocolysis within 4 hrs, increased to 600 mcg/min.  2) Conventional incremental dose ritodrine.  Dose: 50 mcg/min increase 50 mcg/min every 15 min.  Duration: 24‐48 hrs after contraction ceased.  Maintenance: 50 mcg/min for 12‐24 hrs whereupon placebo or sustained release capsules of 240 mg ritodrine daily for seven days.
Outcomes	Primary outcomes: birth within 48 hrs; RDS; periventricular haemorrhage grade 3‐4.  Secondary outcomes: birth < 34; birth < 37; neonatal death; sepsis; adverse effects.
Notes	Prenatal corticosteroid: yes (do not state exact number).
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	The method for generating the randomisation sequence was not stated.
Allocation concealment (selection bias)	Low risk	Allocations were in numbered, opaque, sealed envelopes. All envelopes were accounted for.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	It was not clear if women were aware of treatment group. Staff would have been aware of treatment as dosing regimens were different in the two groups. Clinical decisions appear to have been affected by treatment allocation.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	It was stated that data entry sheets were assessed by blind outcome assessors.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Nine women were excluded post randomisation as they did not meet inclusion criteria. Of the remaining 203 women 2 women (1 in each group) were not included in the analysis.
Selective reporting (reporting bias)	Unclear risk	Not apparent (protocol not available).
Other bias	High risk	There was considerable deviation from protocol in the control group. It was stated that the strict interval for dose augmentation was only observed for 30% of the control group “This was because the participating clinicians were used to being more cautious with the conventional infusion scheme, often applying it in a more gradual way than officially prescribed.”

Methods	(Brief abstract) randomised trial.
Participants	61 women in premature labour. No other details provided.
Interventions	32 women were given clenbuterol tablets. 29 women were given fenoterol perlongettes.
Outcomes	Electrocadiographic changes.
Notes	Although this study met the inclusion criteria, no relevant outcome data were reported, and this study does not contribute any data to the review.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"randomized by computer".
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Different regimens. Staff would be aware of treatment group
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Different regimens. Staff would be aware of treatment group
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Not described
Selective reporting (reporting bias)	Unclear risk	Assessment from published abstract.
Other bias	Unclear risk	Very little information on methods.

Methods	Randomised trial; 3 arms with individual randomisation.
Participants	Study in hospital setting in Sweden. 13 women with established premature contractions, 27‐35 weeks' gestation with intact membranes.
Interventions	1. Terbutaline impregnated vaginal polymer ring (5 g) (5 women). 2. Terbutaline vaginal gel (4 women). 3. Control, placebo vaginal polymer ring (5 women).
Outcomes	Percentage reduction in uterine contractions over a 2‐ hr period; maternal pulse and blood pressure over a 2‐hr period, and peripheral blood mean terbutaline concentration over 2 hrs of treatment.
Notes	Although this study met the inclusion criteria, no relevant outcome data were reported, and this study does not contribute any data to the review.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described.
Allocation concealment (selection bias)	Unclear risk	"randomized". Not described.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Different regimens. Staff would be aware of treatment group
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Different regimens. Staff would be aware of treatment group
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Not mentioned.
Selective reporting (reporting bias)	Unclear risk	Assessment from published study report.
Other bias	Unclear risk	Not apparent. Small sample size.

Methods	Multicentre randomised controlled trial, 4‐arm trial, (3 different doses/routes of ritodrine and 1 placebo group)
Participants	Location: Denmark.  Timeframe: not stated.  Eligible criteria: pregnant women between 20 and 36 weeks' gestation with preterm labour (defined as regular uterine contractions accompanied by effacement and/or dilation of the cervix) and the women who were in labour and had a fetus that was thought to weigh less than 2500 g.  Exclusion criteria: antepartum haemorrhage or signs of abruptio placentae; Rh negative with previously affected babies or a history of ABO incompatibility; cardiac diseases; rupture membranes; cervical dilatation 5 cm or more; intrauterine infection; eclampsia or severe pre‐eclampsia; diabetes; twins.  Total recruited: 199 women, 48 to long ritodrine infusion plus 52 to short ritodrine infusion and 50 to intramuscular ritodrine (vs) 49 to placebo.
Interventions	1) Long ritodrine infusion group (ritodrine plus 5.5% glucose).  Dose: 100 mcg/min iv increase 50 mcg/min every 5‐10 mins until contraction ceased or unacceptable side effects (maximum 350 mcg/min) then continued for 24 hrs.  Maintenance: 1 tablet (10 mg) orally four times a day 30 mins before discontinuing iv form, maximum at 12 tablets, continued until 37 weeks' gestation.  2) Short ritodrine infusion.  Dose: same as long ritodrine infusion but iv treatment was discontinued at 30 mins after contraction ceased.  Maintenance: same dose but started when contraction ceased.  3) Intramuscular ritodrine.  Dose: 10 mg im every 4 hrs for first 12 hrs, and every 6 hrs for the second 12 hrs.  Maintenance: same dose but started 3 hr after last injection.  4) Placebo.  Dose: 5.5% glucose for 24 hrs.  Maintenance: 1 placebo tablet 4 times a day until 37 weeks' gestation or birth.
Outcomes	Primary outcomes: birth within 48 hrs; RDS; perinatal deaths.  Secondary outcomes: neonatal deaths; maternal adverse effects.
Notes	Prenatal corticosteroid: betamethasone 12 + 12 mg, number of participants treated by steroids not stated.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described.
Allocation concealment (selection bias)	Unclear risk	“allocated randomly”, no other description.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	It was stated that women in the control group were given an iv 5% glucose infusion to “blind” them As the other arms had different regimens and routes (one group had im) it was not clear whether blinding was effective. Staff would have been aware of allocation due to different treatment regimens.
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Not mentioned. Staff would have been aware of the treatment regimens.
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	176/199 women were included in the analysis. 23 women were excluded after randomisation. There were some missing data for some variables.
Selective reporting (reporting bias)	Unclear risk	Assessed from published reports.
Other bias	Low risk	Not apparent. Groups appeared comparable at baseline.

Methods	Multicentre randomised trials, 2 arms with individual randomisation.
Participants	Location: USA.  Timeframe: not stated.  Eligible criteria: preterm labour.  Exclusion criteria: not stated.  Total recruited: 466 women, 314 to hexoprenaline (vs) 152 to ritodrine.
Interventions	1) Hexoprenaline.  Dose: not stated.  2) Ritodrine.  Dose: not stated.
Outcomes	Secondary outcomes: adverse effects.
Notes	Prenatal corticosteroid use: not stated.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Described as “randomized” no description of method.
Allocation concealment (selection bias)	Unclear risk	Described as “randomized” no description of allocation concealment method.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Not mentioned (likely to be high risk of bias).
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Not mentioned (likely to be high risk of bias).
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Loss to follow‐up was not described.
Selective reporting (reporting bias)	Unclear risk	Assessed from published study report.
Other bias	Unclear risk	Very little information on study methods.

PERMALINK

Betamimetics for inhibiting preterm labour

James P Neilson

Helen M West

Therese Dowswell

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Background

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

(1) Sequence generation (checking for possible selection bias)

(2) Allocation concealment (checking for possible selection bias)

(3) Blinding (checking for possible performance and detection bias)

(4) Incomplete outcome data (checking for possible attrition bias through withdrawals, dropouts, protocol deviations)

(5) Selective reporting bias

(6) Other sources of bias

(7) Overall risk of bias

Measures of treatment effect

Dichotomous data

Continuous data

Unit of analysis issues

Cluster‐randomised trials

Cross‐over trials

Other unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

Results

Description of studies

Excluded studies

Included studies

(1) Study location and settings

(2) Participants

(3) Interventions

(4) Outcomes

Risk of bias in included studies

1.

2.

Allocation

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

Effects of interventions

1. Betamimetics compared with placebo (comparison 1)

Primary outcomes

Birth within 48 hours of treatment

1.1. Analysis.

3.

Perinatal death (up to seven days) and serious neonatal morbidity

1.2. Analysis.

1.3. Analysis.

1.4. Analysis.

4.

Neonatal length of hospital stay

Maternal secondary outcomes

Methods	RCT, 2 arms with individual randomisation.
Participants	Setting: hospital (tertiary care) in Iran, recruitment between April‐September 2009 Inclusion criteria: 287 assessed and 200 eligible. Women between 20 and 37 weeks were monitored over 1 hr and were in threatened preterm labour (2 contractions within a 10‐min period) with cervical dilation 0‐3cm for primiparous and 1‐3cm for multiparous women, and less than 50% cervical effacement. Exclusion criteria: Cervix dilated > 5cm, polyhydramnios, oligohydramnios, macrosomia, suspected infection, IUGR, antepartum haemorrhage, ruptured membranes, medical disorder, previous CS, multiple pregnancy, pregnancy hypertension,low BP, fetal distress or abnormality, previous treatment with or contraindication to betamimetics.
Interventions	Group 1: (100 women) subcutaneous terbutaline 250 mcg loading dose, same dose every 45 mins until contractions ceased. If contractions ceased maintenance therapy of 20 mg daily oral terbutaline. Group 2: (100 women) iv salbutamol, bolus 0.1 mg with same boluses every 5 mins, if contractions stopped maintenance therapy of 24 mg/day of oral salbutamol.
Outcomes	Prolongation of pregnancy beyond 48 hrs, mean prolongation of pregnancy, pregnancy outcomes (for those women delivering after 37 weeks only) side effects, adverse events, neonatal weight, Apgar score, umbilical arterial and venous pH values, hyperbilirubinaemia. Neonatal complications such as haemorrhage or infections.
Notes	Women who delivered during the treatment period, whose contractions did not cease within 48 hrs or who developed complications were not included in analysis for pregnancy outcomes or complications (44/200 – 22%).Data for these outcomes have not been included in the review.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer random number generator.
Allocation concealment (selection bias)	Low risk	Sealed, opaque numbered envelopes.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Different regimens, not blinded.
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Different regimens, not blinded.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Low risk of bias for primary outcome but high risk of bias for all secondary outcomes (> 20% no data for pregnancy outcomes). We have not included outcome data for these secondary outcomes as there was > 20% missing data and the data are likely to be at high risk of bias.
Selective reporting (reporting bias)	High risk	Outcomes for babies delivered before 37 weeks were not reported. Results are therefore difficult to interpret.
Other bias	Low risk	Other bias not apparent.

Methods	RCT in one centre, 2 treatment arms (with stratification by gestational age).
Participants	Setting: Hospital in Milan, Italy. Inclusion criteria: 82 randomised. Inclusion criteria not clear, women for whom “tocolytic treatment was considered necessary” (not clear that all women had threatened preterm labour). All women had intact membranes. (Only two of the three stratified groups are relevant to this review – 16/82 randomised were 14‐20 weeks gestation) Exclusion criteria: Blood loss, hypertension, hyperthyroidism or heart problems.
Interventions	Group 1: Clenbuterol (a beta‐2 mimetic): vials: 100 mcg in 5 mL doses (equal to 20 mcg/mL), tablets: 20 mcg Group 2: Isoxsuprine: vials: 100 mg in 2 mL doses, tablets 10 mg. (In both groups 41 women were randomised, 8 had a gestational age of < 21 wks; 33 women in each group are relevant to this review). “The administration plan was the same for both drugs. The tocolytic treatment was composed of a first, more intense, phase carried out intravenously and maintained until a valid tocolysis was obtained. In second phase, the tocolytic treatment was administered orally until the 37^th week of pregnancy.”
Outcomes	Maternal and fetal heart rate during treatment (reported as average) and minimum and maximum blood pressure during iv treatment. There were also figures reported for birth before 37 weeks (in graphs) but denominators are not clear and data includes those women recruited before 21 weeks' gestation.
Notes	We were unable to include any outcome data from this study as separate data were not available for women whose gestational age was greater than 20 weeks.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described, sample stratified by gestational age.
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Not described (different regimens) Staff likely to have been aware of treatment group.
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Not described (different regimens) Staff likely to have been aware of treatment group.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	All women appear to have been accounted for.
Selective reporting (reporting bias)	Unclear risk	Assessment from translated notes.
Other bias	Unclear risk	Little information on study methods.

Methods	Double‐blind RCT.
Participants	33 women in premature labour (women with ruptured membranes, placental abruption or cervical cerclage were excluded).
Interventions	iv terbutaline 15 women) versus placebo (18 women).
Outcomes	"prolongation index". Outcomes were reported in graphs and results were not reported by randomisation group.
Notes	Although this study met the inclusion criteria, no relevant outcome data were reported, and this study does not contribute any data to the review.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described.
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Described as a double‐blind study.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Described as a double‐blind study.
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Results were not reported by randomisation group. It was not clear if there was any loss to follow‐up.
Selective reporting (reporting bias)	Unclear risk	Results from published report.
Other bias	Unclear risk	Little information about methods.

Methods	Randomised trial with 2 arms.
Participants	23 women with premature labour between 27‐35 weeks' gestation.
Interventions	iv ritodrine (12 women) versus iv isotonic glucose (11 women).
Outcomes	Water and salt metabolism and blood chemistry.
Notes	Although this study met the inclusion criteria, no relevant outcome data were reported, and this study does not contribute any data to the review.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described.
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Placebo‐controlled trial.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Placebo‐controlled trial.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No loss to follow‐up apparent.
Selective reporting (reporting bias)	Unclear risk	Assessment from published study report.
Other bias	Low risk	Not apparent.

Methods	Randomised double‐blind trial with 2 arms.
Participants	10 women in premature labour (no further details of participants provided).
Interventions	Not clear how many women randomised to the active treatment and placebo groups. Interventions group: Ritodrine infusion over 12 hrs. Control: Placebo iv infusion (saline).
Outcomes	Maternal blood analyses (including glucose, insulin, glucagon) at half‐hourly intervals.
Notes	Although this study met the inclusion criteria, no relevant outcome data were reported, and this study does not contribute any data to the review.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described.
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Double‐blind trial.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Double‐blind trial.
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Not mentioned.
Selective reporting (reporting bias)	Unclear risk	Assessment from brief conference abstract.
Other bias	Unclear risk	Little information on study methods.

Methods	Randomised trial (2 arms).
Participants	29 women in premature labour, 24 to 35 weeks' gestation.
Interventions	Ritodrine (15 women) versus placebo (14 women).
Outcomes	Maternal metabolism (glucose, insulin, glucagon, triglycerides, cholesterol, lactogen, chorionic gonadotropin).
Notes	Although this study met the inclusion criteria, no relevant outcome data were reported, and this study does not contribute any data to the review.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random number tables.
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Placebo‐controlled trial.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Placebo‐controlled trial.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No loss to follow‐up apparent (some missing data).
Selective reporting (reporting bias)	Unclear risk	Assessment from published study report.
Other bias	Low risk	Groups described as similar at baseline.

Methods	Multicentre randomised study.
Participants	No information on inclusion or exclusion criteria. 82 women in premature labour.
Interventions	Two iv doses or ritodrine were compared (fixed versus variable dose). Group 1: (35 women) starting dose of 0.050 mg/min augmented every 5 mins to a maximum of 0.300 mg/min. Group 2: (37 women) starting dose between 0.200 to 0.300 mg/min adapted to clinical effects (variable dose).
Outcomes	Mean average dosage and mean average duration of perfusion, "delay to obtain tocolysis". Maternal tolerance (results not reported).
Notes	Although this study met the inclusion criteria, no relevant outcome data were reported, and this study does not contribute any data to the review.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described.
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Different doses. No blinding.
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Not mentioned.
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Not mentioned.
Selective reporting (reporting bias)	Unclear risk	Assessment from brief abstract.
Other bias	Unclear risk	Very little information on study methods.

Methods	Multicentre RCT in 10 centres, 2 arms, placebo‐controlled, individual randomisation.
Participants	10 centres between 1981 to 1982. Inclusion criteria: 47 women between 24‐36 weeks’ gestation with threatened preterm labour (2 or more uterine contractions during 40 mins monitoring period), estimated fetal body weight of less than 2500 g, dilation of the external os of uterus at no more than 5 cm, Exclusion criteria: Ruptured membranes, fetal malformation, infection or non‐viable fetus, fetal distress, maternal pregnancy complication (toxaemia, placental abruption) or pre‐existing illness, maternal infection.
Interventions	Experimental intervention: (22 women) 50 mg of ritodrine hydrochloride in a 5ml ampoule in 500 mL saline iv (initial dose 100 mcg/min) increased to 150 mcg/min after 30 mins if no effect tocolytic effect observed and increased to 200 mcg/min after 30 mins. After one hr if no effect observed other interventions could be introduced at the discretion of the doctor. Control intervention: (25 women) placebo, same regimen. After one hr if no effect observed other interventions could be introduced at the discretion of the doctor. 6 women in the placebo group had treatment stopped after 1 hr as no effect was observed and they received other treatment. These women were included in the analysis.
Outcomes	Monitoring of uterine contractions and women’s assessment of labour pains over 2 hrs of treatment. Side effects.
Notes	We were unable to include most of the data included in this study in the data and analysis in the review. The primary outcome was uterine contractions during treatment period. It was stated that there were no significant differences in pregnancy outcome attributable to the drugs but data were not reported.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	External randomisation.
Allocation concealment (selection bias)	Low risk	External randomisation by external trial controller. “the allocation table was sealed and stored by the controller during the evaluation period”.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Placebo‐controlled. Placebo had the same appearance and volume as active drug. Doctor could intervene if no progress. In 6 out of 25 cases in the placebo group there was some discretionary intervention.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	It was stated that randomisation code was not broken during the randomisation period.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	All women appeared to be accounted for in the analysis. 6 women in the placebo group were withdrawn and given alternative treatment after 1 hr, but there was an ITT analysis.
Selective reporting (reporting bias)	Unclear risk	Assessment from published report.
Other bias	Low risk	Not apparent. No significant differences in baseline characteristics between groups.

Methods	Single‐centre randomised trial.
Participants	Location: USA.  Timeframe: unclear.  Eligible criteria: strict preterm labour but unclear.  Exclusion criteria: unclear.  Total recruited: 83 women, 42 to terbutaline (vs) 41 to ritodrine.
Interventions	1) Terbutaline.  Dose: 10 mcg/min stepwise (maximum 35 mcg/min).  2) Ritodrine.  100 mcg/min iv stepwise (maximum at 350 mcg/min).  Both interventions were increased until contraction ceased or intolerable side effects. Also, maternal heart rate was not more than 140 beats/min.  Maintenance for both: oral therapy after 12 hr until 36‐37 weeks.
Outcomes	Primary outcomes: birth within 48 hrs; perinatal death.  Secondary outcomes: neonatal death; maternal adverse effects.
Notes	Prenatal corticosteroid use: not stated.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described.
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Not mentioned.
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Not mentioned.
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	83 women. Not clear if any women were lost to follow‐up. Some missing data for some outcomes.
Selective reporting (reporting bias)	Unclear risk	Assessment from brief conference abstract.
Other bias	Unclear risk	Assessment from brief conference abstract. Very little information on study methods.

Study	Reason for exclusion
Ally 1992	The interventions were ritodrine compared with ritodrine plus magnesium gluconate.
Beall 1985	Loss to follow‐up was more than 20% (26%).
Bedoya 1972	The interventions were orciprenaline compared with other uterine sedatives (trasentine, hyoscine bromide, benzodiazepine and progesterone).
Besinger 1991	The interventions were ritodrine compared with indomethacin.
Bulgay‐Moerschel 2008	The interventions were fenoterol compared with nitroglycerin patches.
Caballero 1979	The interventions were ritodrine compared with indomethacin.
Calder 1985	The randomisation was inadequate.
Cararach 2006	The interventions were ritodrine compared with nifedipine.
Caritis 1991	The participants were not randomised.
Castillo 1988	Study reported in brief abstract. Not clear that this was a randomised trial.
Castren 1975	Method of generation of allocation was inadequate.
Chhabra 1998	The participants were not randomised.
Christensen 1980	The study was conducted in preterm premature rupture of membranes without preterm labour.
Csapo 1977	The participants were not randomised.
Das 1969	The participants were not randomised and the interventions were isoxsuprine compared with sedatives, analgesics and antispasmodics.
Dellenbach 1971	The interventions compared were dysmalgine with placebo. It was not clear that the study drug was a betamimetic (described as antispasmodic). Results were not reported for both randomised groups.
Dunlop 1986	The study was conducted in preterm premature rupture of membranes without uterine contractions.
Ferguson 1987	The interventions were ritodrine plus magnesium sulfate compared with ritodrine plus placebo.
Francioli 1988	The interventions were betamimetics (hexoprenaline) compared with hexoprenaline plus magnesium sulfate.
Gamissans 1978	The interventions were ritodrine plus placebo compared with ritodrine plus indomethacin or placebo plus indomethacin.
Gamissans 1982	The interventions were ritodrine plus placebo compared with ritodrine plus indomethacin or placebo plus indomethacin.
Garite 1987	Loss to follow‐up more than 20% after randomisation.
Goel 2011	The interventions were isoxuprine with micronised progesterone.
Gonik 1988	Method of generation of allocation was inadequate.
Guinn 1997	The study was conducted in participants with premature uterine contraction.
Gummerus 1981	Method of generation of allocation was inadequate.
Hallak 1992	The study was conducted in participants without signs of preterm labour.
Hallak 1993	The study was conducted in participants without signs of preterm labour.
Hatjis 1987	The interventions were ritodrine compared with ritodrine plus magnesium sulfate.
Herzog 1995	The interventions were fenoterol plus oral magnesium sulfate compared with other regimens of magnesium sulfate.
Howard 1982	Loss to follow‐up 35% after randomisation.
Ieda 1991	The interventions were magnesium sulfate plus ritodrine compared with control group (unclear).
Kanayama 1996	The interventions were ritodrine compared with a urinary trypsin inhibitor.
Karlsson 1980	The participants were not randomised.
Katz 1983	The interventions were betamimetics (ritodrine) compared with ritodrine plus indomethacin.
Kim 1983	The participants were not randomised.
Kosasa 1985	Method of generation of allocation was inadequate.
Larsen 1986	Loss to follow‐up was more than 20%.
Leake 1983	The participants had preterm premature rupture of membranes without labour.
Lenz 1985	The participants were not randomised.
Levy 1985	The study was conducted in participants with preterm premature rupture of membranes without evidence of true labour.
Lipshitz 1976	The study was conducted in participants with elective inductions of labour.
Lyell 2005	The interventions were magnesium sulfate compared with nifedipine.
Merkatz 1980	All trials in this multi‐centred study have been reported individually.
Muller‐Holve 1987	Participants were premature uterine contractions without mention of cervical dilatations or inclusion or exclusion criteria. Also, the outcomes did not contribute to the review.
Neri 2009	The interventions were ritodrine compared with Atosiban (oxytocin receptor antagonist).
Park 1982	The participants were not randomised.
Raymajhi 2003	The interventions were nifedipine compared with isoxsuprine.
Reynolds 1978	The interventions were salbutamol compared with salbutamol plus ethanol.
Rios‐Anez 2001	The interventions were betamimetics (fenoterol) compared with fenoterol plus naproxen.
Ritcher 1975	The participants were not randomised.
Ritcher 1979	The interventions were betamimetics (ritodrine) compared with ritodrine plus calcium antagonist or buphenine plus calcium antagonists.
Ross 1983	The interventions were terbutaline plus metoprolol compared with terbutaline plus placebo.
Ryden 1977	The participants were not randomised.
Sanchez 1972	The study was conducted in participants with normal labour.
Sciscione 1993	The interventions were crossed‐over between betamimetics (iv ritodrine, subcutaneous terbutaline) and magnesium sulfate.
Sirohiwal 2001	The randomisation was inadequate.
Sivasamboo 1972	Method of generation of allocation was not truly randomisation.
Spatling 1989	Method of generation of allocation was inadequate.
Tarnow‐Mordi 1988	The study was conducted in participants before cesarean section without labour.
Trabelsi 2008	The interventions were salbutamol compared with nicardipine.
Weisbach 1986	The participants were not randomised.
Wesselius‐De 1971	Recurrent preterm was considered as new case. The outcomes did not contribute to the review.
Zeller 1986	This paper did not report the findings of a randomised trial, rather there is secondary analysis of data for women who had undergone long‐term tocolysis (it was stated that this was "mainly with Fenoterol"). Women were divided into groups retrospectively according to the interval between the end of tocolysis and delivery.