Dr L-L Su, Department of Obstetrics and Gynaecology, Yong Loo Lin School of Medicine, National University of Singapore, National University Health System, 5 Lower Kent Ridge Road, Singapore 119074. Email firstname.lastname@example.org
Objective Prevention of postpartum haemorrhage is essential in the pursuit of improved health care for women. However, limited literature is available for comparing the use of oxytocin agonist carbetocin with syntometrine in women undergoing vaginal deliveries. We aimed to compare intramuscular carbetocin with intramuscular syntometrine for the routine prevention of postpartum haemorrhage in women who deliver vaginally.
Population Pregnant women with no contraindication for vaginal delivery recruited from January 2005 to April 2008.
Methods Participants were randomised to receive either syntometrine or carbetocin during the third stage of labour.
Main outcome measures Primary outcome measure was postpartum haemorrhage requiring additional uterotonics. Secondary outcome measures were the incidence of postpartum haemorrhage (≥500 ml), severe postpartum haemorrhage (≥1000 ml) and adverse effects profile.
Results Women in the carbetocin group (13.5%) and in the syntometrine group (16.8%) had postpartum haemorrhage requiring additional uterotonics (P =0.384). 1.6% of women in each group had postpartum haemorrhage (P =1.0) and the estimated blood loss during the third stage of labour was similar between the two groups (P =0.294). Women who had syntometrine were four times more likely to experience nausea (RR = 4.2; 95% CI 2.2–7.8) and vomiting (RR = 4.3; 95% CI 1.9–9.5) compared with women who had carbetocin. Tremor, sweating, retching and uterine pain were also more likely in the syntometrine group compared with the carbetocin group (P <0.05).
Conclusions Carbetocin has an efficacy similar to syntometrine for prevention of postpartum haemorrhage, but is associated with less adverse effects.
Postpartum haemorrhage is potentially life threatening and is a significant contributor to maternal mortality and morbidity worldwide.1–3 Prevention of postpartum haemorrhage is, therefore, of great importance in the pursuit of improved health care for women. The pharmacologic agents currently used routinely to prevent postpartum haemorrhage are mainly syntometrine (a combination of oxytocin and ergometrine) and oxytocin. Syntometrine is associated with a statistically significant reduction in the risk of postpartum haemorrhage when compared with oxytocin alone.4 However, adverse effects of nausea, vomiting and hypertension are higher in women receiving syntometrine because of the ergometrine component.5 Syntometrine has also been associated with severe complications such as coronary artery spasm6 and intracerebral haemorrhage.7 In addition, syntometrine cannot be used in 10 to 20% of the obstetric population as a result of coexisting medical conditions such as pre-eclampsia and heart disease. These women are then given oxytocin, which, probably because of its short duration of action, is less effective in preventing postpartum haemorrhage.8,9
Over the past two decades, several other alternatives have been explored including the use of prostaglandins such as misoprostol and carboprost. The role of various prostaglandins including misoprostol for postpartum haemorrhage prophylaxis is limited.10 Among the agents that have been studied, oxytocin agonist (carbetocin) appears to be the most promising for this indication.11 Carbetocin is a long-acting synthetic octapeptide analogue of oxytocin with agonist properties. In pharmacokinetic studies, intravenous injections of carbetocin produced tetanic uterine contractions within 2 minutes, lasting 6 minutes, followed by rhythmic contractions for a further hour. Intramuscular injection produced tetanic contractions in <2 minutes, lasting about 11 minutes, and followed by rhythmic contractions for an additional 2 hours. The prolonged duration of activity after intramuscular compared with the intravenous carbetocin was significant. In comparison with oxytocin, carbetocin induces a prolonged uterine response when administered postpartum, in terms of both amplitude and frequency of contractions.12
Carbetocin has been shown to reduce the need for additional oxytocics compared with syntocinon in women who underwent caesarean deliveries.13–15 The need for additional uterotonics is an important outcome variable as it reflects the degree of postpartum vaginal bleeding accounted for by uterine atony. However, very limited published literature is available, which compared the use of carbetocin and syntometrine in low-risk women who delivered vaginally. We, therefore, conducted a double-blind randomised controlled trial comparing carbetocin and syntometrine for the prevention of postpartum haemorrhage in low-risk women who underwent vaginal deliveries. We aimed to test the hypothesis that intramuscular carbetocin is as effective as intramuscular syntometrine for the prevention of postpartum haemorrhage for women who undergo vaginal deliveries with less adverse effects.
We recruited healthy pregnant women who were attending antenatal clinics at National University Hospital, a tertiary hospital in Singapore. Mothers were considered eligible for participation if they were expected to undergo vaginal delivery at or beyond 34 weeks of gestation. Women scheduled for elective caesarean deliveries and women with risk factors for postpartum haemorrhage, such as multiple pregnancy, past history of postpartum haemorrhage and history of suspected coagulopathy, were excluded from the study. Other exclusion criteria were women with contraindications for the use of syntometrine including history of coronary artery disease or hypertension, as well as women with history of hypersensitivity to syntometrine or carbetocin.
Women who agreed to participate were required to give a written informed consent. The study was approved by the National Healthcare Group Domain Specific Review Board. Women were randomised to receive either intramuscular syntometrine or intramuscular carbetocin. Women in the syntometrine group received one ampoule of syntometrine (Novartis, Basel, Switzerland) intramuscularly, which consisted of 5 iu of oxytocin and 500 microgram of ergometrine. Women in the carbetocin group received one ampoule of carbetocin (100 microgram) (Duratocin®, Ferring, Canada) intramuscularly. Randomisation was blocked and stratified by parity. The randomisation list with the allocation of the mode of intervention was forwarded from the Biostatistics Unit to the Department of Pharmacy at National University Hospital, where the purchased medications were kept. Both these medications were specially packed and coded by the hospital central pharmacy according to the randomisation and allocation list provided by the Biostatistics Unit. The glass ampoules were masked to make the medications look identical. The corresponding opaque package containing the allocated drug was kept at the Delivery Suite. The identities of the medications were not known to the midwives, obstetricians and the participants. The medication codes were only broken following completion of the trial.
At entry to the Delivery Suite, the consent of the subjects to participate in the study was reconfirmed. Just prior to vaginal delivery, the delivery suite staff checked the registration master list to obtain the trial number for the patient. The corresponding opaque package containing the allocated drug was then administered following delivery of the anterior shoulder of the baby. Both medications were administered via the intramuscular route by the nursing staff in the delivery suite.
The primary outcome measure was postpartum haemorrhage requiring additional uterotonic therapy. The standard criteria for the use of additional uterotonics in our delivery suite were (a) suboptimal uterine tone, (b) brisk estimated blood loss exceeding 300 ml with or without hypotension (blood pressure <90/60) or tachycardia (pulse rate >100 beats per minute). The secondary outcome measures were the incidences of postpartum haemorrhage (≥500 ml) and severe postpartum haemorrhage (≥1000 ml) as well as adverse effects of the interventions. Following delivery, the vital signs and uterine tone of the patients were monitored as per routine protocol. Blood pressure, pulse rate and temperature were monitored every 30 minutes for 2 hours after delivery as per protocol.
Visual blood loss estimation was made by both the midwife and the obstetrician. If additional uterotonic therapy was instituted, the midwife or doctor would record the name and dosage of the alternative therapy, as well as the basis for use and effect of the alternative therapy. Any requirement for manual removal of placenta or blood transfusion was also recorded. The length of hospital stay for the participants in the two study groups was also compared.
A standard questionnaire assessing the adverse effects was filled up by the participant. The presence or absence of the adverse symptoms was recorded. For any symptom experienced by the participant, she was required to grade the symptom from very mild to very severe on a five-point visual analogue scale. Assessment of both the primary and secondary outcomes of the study was performed in the delivery suite, within 2 hours after delivery. This was to ensure that any complications related to the delivery or medications resolved prior to the patients being transferred to the postnatal ward.
Statistical analysis was performed using SPSS version 16.0 (SPSS Inc., Chicago, IL, USA) and was based on an intention-to-treat principle. Power calculations were performed. Based on the experience at our centre, the need for the use of additional uterotonic agents was around 10–15%. A value of 13% need for the use of additional uterotonic agents was therefore chosen. To declare an equivalence of 10% with 80% power and a two-sided test of 5%, 180 subjects were needed in each arm. An additional ten subjects were recruited to account for possible attrition. Intention-to-treat analysis was employed. The 95% CI for the difference between the two groups in the use of uterotonic agents was derived using the two-sample t test. Differences in quantitative measures between the two groups were analysed using parametric tests when normality and homogeneity assumptions were satisfied; otherwise, the equivalent non-parametric Mann–Whitney U test was applied. Statistical significance was set at P <0.05.
We recruited a total of 370 subjects, between January 2005 and April 2008, out of whom, 185 were randomised to receive carbetocin and 185 were randomised to receive syntometrine. Informed consent was obtained from 420 participants, but 50 women were withdrawn from the study after informed consent, but before randomisation. Thirty-four of these cases were withdrawn because of the need of the participants to undergo emergency caesarean sections. The other reasons for withdrawal included intrapartum hypertension and withdrawal of consent. The trial profile is shown in Figure 1.
Baseline characteristics between the two randomised groups were similar (Table 1). The study groups were also similar in the birth-related variables including the mode of delivery and the mean birthweight (Table 1). There were no differences in factors, which could predispose to postpartum haemorrhage, such as retained placenta and third- or fourth-degree perineal tears.
Table 1. Baseline and intrapartum characteristics of the study population
Carbetocin (n = 185)
Syntometrine (n =185)
Total (n =370)
Parity, n (%)
Gestational age (weeks)
Duration of third stage of labour (minutes)
Mode of delivery, n (%)
Manual removal of placenta, n (%)
Third-/Fourth-degree perineal test, n (%)
No significant difference was demonstrated between the carbetocin group and syntometrine group in terms of the need for the use of additional uterotonic agents. 13.5% (n =25) of women in the carbetocin group and 16.8% (n =31) of women in the syntometrine group required additional uterotonic agents (P =0.384). The estimated mean blood loss was 217.4 ml for the carbetocin group and 223.1 ml for the syntometrine group (P =0.294). Subgroup analyses were performed to determine the need for additional uterotonic agents between carbetocin and syntometrine groups among primiparous women and multiparous women. The need for additional uterotonics was similar between the two interventions for both primiparous and multiparous women (Table 2). The most common therapeutic uterotonic agents used were syntocinon infusion, a repeat dose of syntometrine or a bolus dose of ergometrine.
Table 2. Primary and secondary outcome measures
Carbetocin (n =185)
Syntometrine (n =185)
Total (n =370)
Need for additional uterotonics, n (%)
PPH (≥500 ml), n (%)
Severe PPH (≥1000 ml), n (%)
Need for blood transfusion, n (%)
Mean blood loss (ml)
Length of hospital stay (days)
Three women (1.6%) in each of the carbetocin group and syntometrine group had postpartum haemorrhage (estimated blood loss equal to or greater than 500 ml) (P =1.0). In addition, one woman in the syntometrine group experienced severe postpartum haemorrhage. No difference in the need for blood transfusion was demonstrated between the carbetocin and syntometrine groups (P value = 1.0).
For adverse effect profile, women in the carbetocin group were less likely to experience nausea and vomiting. For the syntometrine group, 24.9% of women (n =46) experienced nausea and 16.2% of women (n =30) experienced vomiting. Out of these, 54.2% reported the nausea to be moderate to very severe and 59.9% reported the vomiting to be moderate to very severe. In contrast, in the carbetocin group, only 5.9% (n =11) of women reported nausea and 3.8% (n =7) of women reported vomiting. Of these, 18.6 and 13.2% reported the nausea and vomiting to be moderate to very severe respectively. Therefore, women who had syntometrine were 4.2 times (RR = 4.2; 95% CI 2.2–7.8) more likely to experience nausea and 4.3 times (RR = 4.3; 95% CI 1.9–9.5) more likely to experience vomiting compared with women who had carbetocin. In addition, women who received syntometrine were 12.5 times (RR 12.5; 95% CI 3.0–52.6) more likely to experience moderate to very severe nausea and 17.8 times (RR 17.8; 95% CI 2.4–142.8) more likely to experience moderate to very severe vomiting compared to women who received carbetocin (Table 3).
Table 3. Secondary outcomes of adverse effects
Carbetocin (n =185)
Syntometrine (n =185)
RR (95% CI)
Moderate to very severe
Moderate to very severe
Women in the syntometrine group were also significantly more likely to experience tremor, sweating, retching and uterine pain compared with women in the carbetocin group (Table 3). The duration of hospitalisation stay was also similar for the participants in the two groups with mean hospitalisation stay of 1.82 days in the carbetocin group and 1.81 days in the syntometrine group (P value = 0.936).
Our study showed that intramuscular carbetocin is comparable with intramuscular syntometrine in terms of the need for additional uterotonic agents and the incidence of postpartum haemorrhage. Although there was a trend for carbetocin to be associated with reduced need for additional uterotonic agent, the difference was not statistically significant. The incidence of several adverse effects, such as nausea, vomiting, retching, tremor and sweating, was significantly lower for women who received carbetocin compared with women who received syntometrine.
The need for additional uterotonics was used as the primary outcome variable as it is the most important clinical indicator of postpartum blood loss after delivery. Clinical estimation of blood loss was not used as the primary outcome as quantitative measurement of postpartum blood loss is impractical and difficult to achieve with precision.16,17 Larsson et al.18 reported in their study that for clinical purposes, estimation of blood loss and measurement of postpartum haemoglobin is of low value and may lead to the wrong conclusions. Besides, blood loss after delivery does not only come from the placental bed related to inefficient uterine contractions, but it is also lost from episiotomy wounds, lacerations and other trauma to the birth canal. Decision for the use of additional uterotonics, on the other hand, is decided by the obstetrician based not only on clinical estimation of blood loss but also on the diagnosis of uterine atony based on palpated uterine tone after delivery.
The need for additional uterotonics among the women in our study was 15.1% even though the incidence of postpartum haemorrhage in the group was 1.6%. The discrepancy in the incidences of these two outcomes was because obstetricians generally did not wait till blood loss exceeded 500 mls (definition of postpartum haemorrhage) before commencing additional uterotonics. In our centre, additional uterotonics is started upon making the diagnosis of uterine atony. The active use of additional uterotonics in our centre could account not only for the low incidence of postpartum haemorrhage but also for the relatively low estimated blood loss volume, with mean of 220.2 ml in the study cohort.
Minimising unpleasant adverse effects in the third stage of labour is important to improve the childbirth experience of a woman. Our data showed that the incidence of many adverse effects, such as nausea, vomiting, retching, tremor and sweating, was significantly lower for women who received carbetocin compared with women who received syntometrine. The difference was particularly significant for moderate to very severe nausea or vomiting. There was no difference in the other adverse effects assessed in our study.
Thus far, three trials have compared the use of carbetocin and oxytocin for the prevention of postpartum haemorrhage, two for women who underwent caesarean deliveries13,14 and one for women with risk factor(s) for postpartum haemorrhage, who underwent vaginal deliveries.19 Pooled data from these trials did not reveal any statistically significant differences in terms of the adverse effects between carbetocin and oxytocin. However, the ergometrine component of syntrometrine is known to be associated with higher incidence of unpleasant effects. In the only trial conducted so far to compare carbetocin with syntometrine, Leung et al.20 showed that carbetocin was associated with a lower incidence of adverse effects such as nausea and vomiting compared with syntometrine. Our data further confirm this favourable adverse effect profile of carbetocin compared with syntometrine.
The major strength of our study was the rigorous process by which the trial was conducted. Despite following Good Clinical Practice guidelines, the study was conducted in a pragmatic fashion and simulated a non-research setting in a busy tertiary hospital. Robust randomisation, allocation concealment and blinding were conducted to eliminate any selection, performance and detection biases.
One limitation of the study is the relatively small sample size for looking at parameters such as postpartum haemorrhage. Thus far, there was only one randomised controlled trial on the comparison between carbetocin and syntometrine for the third stage of labour following vaginal delivery. For parameters such as postpartum haemorrhage, it is important to have pooled data from various trials to have a powerful conclusion. Another limitation of our study is that proper cost-effectiveness analysis was not performed. The cost of carbetocin (S$39) is around ten times higher than the cost of syntometrine (S$4). However, as carbetocin is associated with reduced adverse effects, this could translate into cost savings in terms of less medications needed for adverse effects and less monitoring required of the nursing personnel.
Carbetocin has an efficacy similar to ergometrine for the prevention of postpartum haemorrhage during the third stage of labour. The need for additional uterotonic agents was similar between women who received carbetocin or syntometrine. However, carbetocin is associated with a lower risk of various adverse effects.
Disclosure of interest
Contribution to authorship
L.L.S. and Y.S.C. were involved in conception and design of the project, supervising the conduct of the trial as well as preparation of the manuscript. M.R. and A.B. were involved in conception and design of the project as well as recruitment of participants. Y.H.C. was the biostatistician who helped with the design of the trial and analysis of the results. N.B.M.S. and T.P.L. were involved in recruitment of participants and other critical aspects in the conduct of the trial.
Details of ethics approval
The study was approved by the National Healthcare Group Domain Specific Review Board.
This study was supported by a grant from the National Healthcare Group, Singapore (NHG-SIG/07059).
We thank Ms Janie Foo who helped us with the conduct of the trial and NUHS Medical Publications Support Unit for the help in preparing the manuscript.