Population Women undergoing caesarean deliveries for labour arrest.
Methods Forty-eight women were randomised to receive infusion of either ergometrine maleate 0.25 mg + oxytocin 20 iu or oxytocin 20 iu alone, diluted in 1 l of lactated Ringer’s Solution, immediately after delivery of the infant. Unsatisfactory uterine contractions after delivery were treated with additional boluses of the study solution or rescue carboprost. Blood loss was estimated based on the haematocrit values before and 48 hours after delivery.
Main outcome measures The primary outcome was the estimated blood loss, while the secondary outcomes included the use of additional uterotonics, need for blood transfusion and the incidence of adverse effects.
Results The estimated blood loss was similar in the oxytocin–ergometrine and oxytocin-only groups; 1218 ± 716 ml and 1299 ± 774 ml, respectively (P= 0.72). Significantly fewer women required additional boluses of the study drug in the oxytocin–ergometrine group (21 and 57%; P= 0.01). Nausea (42 and 9%; P= 0.01) and vomiting (25 and 4%; P= 0.05) were significantly more prevalent in the oxytocin–ergometrine group.
Conclusions In women undergoing caesarean delivery for labour arrest, the co-administration of ergometrine with oxytocin does not reduce intraoperative blood loss, despite apparently superior uterine contraction.
The routine prophylactic administration of uterotonic drugs to reduce the risk of postpartum haemorrhage (PPH) after delivery has become an integral part of the management of labour and delivery.1,2 Previously conducted randomised controlled trials at our institution have shown that the blood loss in women undergoing caesarean deliveries for labour arrest is twice as much as in those undergoing elective caesarean deliveries.3,4 Furthermore, the dose of oxytocin required to produce effective uterine contractions in labouring women during caesarean deliveries is nine-fold greater than that required in nonlabouring women, perhaps due to the phenomenon of desensitisation of the myometrial oxytocin receptors during labour.3–5
Minimizing blood loss in labouring women at emergency caesarean delivery is clearly desirable and requires a re-evaluation of both the pharmacological and the surgical strategies. Since the labouring uterus is less responsive to oxytocin, the evaluation of the efficacy of additional prophylactic uterotonic agents in this setting is desirable. The extensive Cochrane review on the use of uterotonic agents to prevent PPH at vaginal delivery indicates a small, but significant reduction in blood loss with the addition of ergometrine to oxytocin.6
The purpose of our study was to evaluate the effect of the addition of intravenous ergometrine maleate to a standard oxytocin infusion on the blood loss during caesarean deliveries for labour arrest. We hypothesized that this combination of drugs would be more effective in reducing the blood loss in these women than oxytocin alone.
We conducted a randomised, double-blinded study in healthy pregnant women undergoing caesarean deliveries from 12 June 2005 to 18 December 2006. Approval for the study was obtained from the Research Ethics Board at Mount Sinai Hospital and all the study subjects gave written informed consent. Ten staff obstetricians participated in the study. The inclusion criteria were singleton pregnancy at 37–41 weeks of gestation; epidural analgesia; the requirement for caesarean delivery for labour arrest (defined as an arrest of cervical dilation in the first stage, or fetal descent in the second stage) and the augmentation with intravenous oxytocin for a minimum of 4 hours. The exclusion criteria were an allergy or hypersensitivity to oxytocin or ergot derivatives; cardiac disease; hypertension (diastolic blood pressure >90 mmHg, systolic blood pressure >140 mmHg); a requirement for general anaesthesia and any condition predisposing to uterine atony and postpartum haemorrhage, such as placenta praevia, multiple gestation, pre-eclampsia, macrosomia, polyhydramnios, uterine fibroids, bleeding diathesis, chorioamnionitis, or a previous history of uterine atony and postpartum bleeding.
Based on a random allocation scheme derived from a computer-generated list of numbers, sealed and consecutively numbered opaque envelopes were prepared by a research assistant not involved in the study. The women were randomly allocated to one of the two study groups by opening the next available envelope just before surgery. After the delivery of the infant, one group received a combination of oxytocin and ergometrine, while the other received oxytocin alone. The study drugs were prepared in 3-ml volume by the research assistant not involved in the study. The obstetrician, the anaesthetist, nursing staff as well as the women, were all blinded to the study groups.
Obstetric data, including the onset of active labour, the duration of the first and/or second stage of labour and the cervical dilation at the time of the diagnosis of labour arrest were recorded, before proceeding for caesarean delivery. The dose, duration and type of the agents used for the induction and augmentation of labour were also recorded. The oxytocin was stopped once the obstetrician made the decision for caesarean delivery, and the operation proceeded within 30 minutes. On arrival in the operating room, the woman was placed in the supine position with left uterine displacement using a wedge under the right buttock. Lactated Ringer’s solution was infused through the existing intravenous line at the rate of 150 ml/hour. The standard monitoring included an electrocardiogram, noninvasive blood pressure (BP), heart rate (HR) and pulse oximetry (SpO2). The baseline BP and HR were calculated as the mean of three readings, recorded 1 minute apart, prior to the caesarean delivery using an automated noninvasive BP device. To provide surgical anaesthesia, the existing epidural was topped up with 15–20 ml of 2% lidocaine with epinephrine 1:200000, titrated to achieve a block height to the T4 dermatome. Any pain or discomfort during the operation was treated with the epidural administration of 5-ml aliquots of 2% lidocaine or 50–100 μg fentanyl. After the delivery, 2.5 mg of morphine was injected via the epidural catheter to provide postoperative pain relief. The systolic BP and HR were recorded every minute until the delivery and then every 2.5 minutes until the end of the operation. The systolic BP was maintained within 10% of the baseline values with aliquots of phenylephrine 0.1 mg. If the woman developed sustained hypertension (systolic BP > 20% of the baseline), the study solution was discontinued and replaced by standard oxytocin solution. All women received supplemental oxygen via nasal prongs at the rate of 4 l/minute until the delivery of the infant.
The study drugs, either 0.25 mg of ergometrine maleate and 20 iu of oxytocin (oxytocin–ergometrine group), or 20 iu of oxytocin (oxytocin-only group), diluted in 1 l of lactated Ringer’s solution, were administered immediately following the delivery of the infant. Participants received an initial bolus of 150 ml of the study solution, followed by a maintenance infusion of 120 ml/h for 6 hours.
The obstetrician was asked to assist the spontaneous delivery of the placenta using cord traction without uterine massage, rather than manual extraction. The obstetrician assessed uterine tone by palpation every minute until satisfactory contractions were obtained, similar to a method described earlier.3,4 If the uterine tone remained unsatisfactory following the initial study solution bolus, the obstetrician massaged the uterus and requested an additional bolus of 25 ml of the study solution. Intramuscular carboprost tromethamine (Hemabate; Pfizer, Arnprior, ON, Canada) 250 μg was used as a rescue medication if needed. The administration of any additional uterotonic agents in the first 24 hours postpartum was also recorded. After the delivery of the infant, cefazolin (1 g diluted in 20 ml normal saline), granisetron 1 mg and ketorolac 30 mg were administered intravenously. The standard approach to uterine incision in our unit is to perform a transverse incision with a scalpel and digitally extend the incision bilaterally. Following delivery, the uterus is repaired in situ, or exteriorised, according to the operator’s preference. The edges of the uterine incision are grasped with Green-Armytage clamps and a double-layer uterine closure performed, with additional figure-of-eight haemostatic sutures as necessary.
The primary outcome was the estimated blood loss within the 48 hours postpartum. The blood loss was calculated through the difference in the haematocrit values assessed prior to and at 48 hours after the caesarean delivery, according to the following formula:
where EBV (estimated blood volume) in millilitres = woman’s weight in kilograms × 85.7 The secondary outcome measures included the adequacy of the uterine tone based on the requirement for additional uterotonics, the need for blood transfusion as well as adverse effects such as nausea, vomiting, headache, shortness of breath, chest pain, palpitations, hypotension and hypertension.
Based on the results of a randomised controlled trial carried out at our institution, we estimated a blood loss of 1178 ml, with a common SD of 716 ml, in the oxytocin-only group.4 Assuming a 50% reduction in the amount of blood loss in the oxytocin–ergometrine group, and an α error level or confidence level of 5%, and a β error level or statistical power (1 −β) of 80%, we concluded that the sample size required to achieve a statistically significant difference between these groups was 24 women per group. The data for the continuous outcome variables were analysed using the Student’s t test, and for the discrete variables, the chi-square test was used. A P value of <0.05 was considered statistically significant.
During the recruitment period, 97 women were approached of whom 39 refused to participate and 10 women did not meet the inclusion criteria. Forty-eight pregnant women requiring caesarean deliveries for arrest of labour were enrolled and completed the study protocol.
The women were randomised into two equal groups. All the women received oxytocin for a minimum of 4 hours for either induction and/or augmentation of labour. The participant demographics and other labour details were comparable in the two groups and are summarised in Table 1.
Table 1. Demographics and labour details
Oxytocin–ergometrine group (n= 24)
Oxytocin group (n= 24)
Values are expressed as mean ± SD or n (%).
32 ± 6
33 ± 5
83 ± 13
82 ± 24
163 ± 9
165 ± 7
Gestational age (weeks)
40 ± 2
40 ± 1
Maximum oxytocin dose (mU/minute)
13 ± 6
15 ± 8
Duration of oxytocin infusion (hours)
11 ± 5
10 ± 6
Arrest of cervical dilation
Arrest of fetal descent
The primary outcome measure, estimated blood loss at delivery, did not differ significantly between the oxytocin–ergometrine and the oxytocin-only groups (1218 ± 716 ml and 1299 ± 774 ml, respectively, P= 0.72) and is illustrated in Figure 1. However, significantly fewer women in the oxytocin–ergometrine group required additional boluses of the study drug compared with the women receiving oxytocin alone (21 and 57%; P= 0.01). The proportion of women requiring intraoperative rescue injections of carboprost was similar in both groups. None of the women required supplemental uterotonics or transfusions of blood or blood products in the 24-hour postpartum period.
The adverse effects in both groups are summarised in Table 2. The rates of hypotension (21 and 26%; P= 0.67) and hypertension (4 and 17%; P= 0.14) were similar in groups oxytocin–ergometrine and oxytocin only. Hypertension, seen in both groups, was transient and required no change in the management. Nausea (42 and 9%; P= 0.01) was significantly more prevalent in the oxytocin–ergometrine group as compared with the oxytocin-only group and the difference in the proportion of women experiencing vomiting in the oxytocin–ergometrine group as compared with oxytocin-only group approached significance (25 and 4%; P= 0.05).
Table 2. Secondary outcomes
Oxytocin–ergometrine group (n= 24)
Oxytocin group (n= 24)
Values are expressed as n (%).
Additional boluses of study solution
Despite marked improvements in the obstetric management, primary PPH (haemorrhage occurring within first 24 hours postpartum) remains a significant contributor of maternal morbidity and mortality worldwide. In a recent population-based prospective study of major obstetric morbidity in Scotland, haemorrhage was by far the most important cause, especially following caesarean delivery and accounted for 50% of the total.8 Given the rise in caesarean deliveries following augmented and/or prolonged labours, attention to methods that may reduce the risk of major haemorrhage in this setting is timely and important.
The prophylactic administration of oxytocin after delivery is the current standard for prevention of PPH.1 The main advantages of this drug are its rapid onset of action and the lack of serious adverse effects, if administered appropriately. Effective protocols include 10 iu intramuscular, 5 iu intravenous bolus or 10–20 iu/l intravenous drip at 100–150 ml/h.9–11 Nevertheless, in clinical practice this drug is often administered in excessive amounts, especially in the event of postpartum bleeding, which can lead to several undesirable haemodynamic effects.12–14 In previous work, our group determined the minimum effective intravenous bolus dose of oxytocin (ED90) required for adequate uterine contraction at elective caesarean deliveries in nonlabouring women to be only 0.35 iu.3 Our group has also determined the ED90 for oxytocin at caesarean deliveries for labour arrest to be 2.99 iu, with the amount of blood loss approximately twice that of elective caesarean deliveries.4 These data support the concept of attenuation of the oxytocin receptor signalling and desensitisation of receptors in labouring women from exposure to oxytocin during labour, resulting in reduced uterine responsiveness.5 By contrast, Munn et al. used extremely high doses of oxytocin at the rate of 2667 mU/minute and 333 mU/minute for 30 minutes after caesarean delivery in labouring women, and yet required additional uterotonic agents in 19 and 39% of women, respectively.15 Clearly, the optimum dose of oxytocin to be used in this setting remains controversial, but to be safe the dose used must balance both the desire to maximize uterine tone with the desire to avoid the vasodilatory hypotensive adverse effects.
We hypothesized that the addition of a second uterotonic agent, acting independently of desensitised oxytocin receptors, would be effective in reducing the estimated blood loss at emergency caesarean delivery following augmentation of labour. Of the currently-licensed options (ergot derivative or prostaglandins [carboprost and misoprostol]), we chose to evaluate an ergot derivative, since prostaglandins are known to be associated with more adverse effects.16,17 Ergometrine maleate is a potent uterotonic agent acting via the α-receptors in the myometrium; however, it is not used routinely for prophylaxis of PPH due to its associated adverse effects. Hypertension is a relative contraindication for this agent due to the potential for severe hypertension and tissue (myocardial) ischaemia.18,19
The Cochrane systematic review evaluated six trials on the pharmacological prevention of PPH, comparing ergot alkaloids with oxytocin in women in the third stage of labour.1 The trials showed that the agents were equally effective in preventing PPH, but the ergot alkaloids were associated with an increased requirement for manual removal of the placenta (number needed to harm [NNH] = 92) with the suggestion of more raised blood pressure than with oxytocin. The incidence of retained placenta was perhaps attributable to the spasm of the lower uterine segment induced by the drug.
The Cochrane systematic review also evaluated six trials, totalling more than 9000 women that compared the prophylactic use of intramuscular ergometrine–oxytocin with oxytocin alone.6 The combination of uterotonic agents was found to be more effective than oxytocin alone for preventing PPH (blood loss of 500 ml or more) (number needed to treat; NNT = 61). No difference was seen for the prevention of severe PPH (blood loss of 1000 ml or more), and there was significantly more nausea and vomiting (NNH = 61), and hypertension (NNH = 96), in the women receiving ergometrine–oxytocin. However, these reviews included only vaginal deliveries, and there was a great variability in the doses and preparations of ergot alkaloids used in these trials. Our study is the first to specifically examine the efficacy of the oxytocin–ergometrine combination at caesarean deliveries.
The dose of oxytocin used in our study was based on a previously conducted study at our institute.4 So far no dose–response studies have been conducted for ergometrine, since this drug is usually reserved for therapeutic purposes in the event of bleeding. We chose a proportionate dose of ergometrine, since it was administered prophylactically in combination with oxytocin. The dose of ergometrine was an initial bolus of 37.5 μg followed by an intravenous infusion of 0.5 μg/minute. In spite of this small dose, we observed a significantly higher incidence of nausea and vomiting in the oxytocin–ergometrine group. Although all women had received an anti-emetic (granisetron), the presence of nausea and vomiting demanded an additional anti-emetic and also delayed oral intake. However, other haemodynamic adverse effects seen in other studies6,18,19 were not observed in our study, thus supporting the safety of this regimen. Moreover, the administration by intravenous infusion allows appropriate titration of the drug in relation to the clinical effect.
The estimation of blood loss during caesarean deliveries is known to be inaccurate and subjective due to the admixture with amniotic fluid. We prespecified our primary outcome based on the difference in preoperative and postoperative haematocrit levels. We did not observe any reduction in our primary outcome, estimated blood loss, from the co-administration of ergometrine with oxytocin. This lack of difference was noted, despite the observation that uterine contractility was considered by the operating obstetrician to be more effective in the oxytocin–ergometrine group, as reflected by fewer requirements for supplemental uterotonics. These data suggest a limited role for additional uterotonic agents in the prevention of major blood loss at caesarean delivery for labour arrest after oxytocin augmentation.
Several aspects of the surgical technique therefore warrant re-evaluation in the light of these findings. These include the method of opening the lower segment of the uterus, the timing and type of the clamps placed on the bleeding lower edges and/or the angles, the timing and the method of the removal of the placenta and the suture techniques to close the lower segment. In a recent comparison of digital tearing to open the lower segment with the extension of a central incision with scissors, intraoperative blood loss was greater with the more commonly used digital method in part due to greater angle damage.20 Following delivery, the timing and the placement of one versus several Green-Armytage clamps across the incised lower segment, and especially across bleeding venous sinusoids, are likely to influence the total blood loss, as would their placement before or following removal of the placenta. We believe that the uterus of labouring women is more vascular with soft and thin lower segment. Also the incised lower segment stays soft and floppy, whereas in elective caesarean section in nonlabouring women, incision edges often contract or retract. Hence, Green-Armytage clamps could be more beneficial particularly in labouring women undergoing caesarean section. Since manual removal of the placenta at caesarean delivery increases total blood loss,21 we were careful to standardize our approach to removing the placenta, and waiting for the uterine fundus to contract before beginning the placental expulsion. Finally, since the majority of blood loss will be arrested following a single-layer closure of the uterus, the speed of completing this suture is likely to influence the total blood loss. The exteriorisation of the uterus immediately following the removal of the placenta is a popular manoeuvre that is associated with reduced blood loss at caesarean delivery.22 This approach may facilitate uterine fundal massage and improve surgical visualisation, but increases maternal discomfort during regional anaesthesia. In September 2007, we reported greater maternal comfort, with no increase in estimated blood loss at elective caesarean deliveries with in situ repair as compared with exteriorisation,23 although the apparent safety of this approach at emergency caesarean delivery remains unevaluated.
The lack of benefit observed in this study from the co-administration of ergometrine and oxytocin, despite superior uterine contraction in this group, suggests a limited role for pharmacological prevention of blood loss in this setting. In addition, this co-administration was associated with adverse effects such as nausea and vomiting. These findings have stimulated discussions in our research group regarding the importance of optimising obstetric surgical techniques to reduce operative blood loss. Further clinical trials to assess the role of intraoperative ergometrine in reducing the overall incidence of PPH in the general obstetric population, and in subsets of women at high risk of PPH due to uterine distension (macrosomia, multiple gestation etc.), are warranted, but to be meaningful will need to be conducted in a setting that optimises the surgical techniques at caesarean delivery.
This work was supported by departmental funds only. None of the authors has any conflict of interest relative to this work.
Details of ethics approval
This project was initially approved by the Research Ethics Board at Mount Sinai Hospital # 05-0042-A dated 5 May 2005 for a period of 1 year and was again re-approved for continuation until 5 May 2007. The study is registered under National Clinical Trials Registry: www.clinicaltrials.gov, registration number NCT00481533.
The authors acknowledge Leda Ivic Weiss (Research Assistant) for organizing the database and providing statistical assistance.