Please cite this paper as: Elati A, Elmahaishi M, Elmahaishi M, Elsraiti O, Weeks A. The effect of misoprostol on postpartum contractions: a randomised comparison of three sublingual doses. BJOG 2011;118:466–473.
Objective To compare the postpartum uterine activity and side effects of various doses of sublingual misoprostol and intramuscular oxytocin.
Design Single centre, randomised trial.
Setting Zliten Teaching Hospital in Libya.
Population Forty-nine women who did not receive oxytocics in labour and who delivered vaginally.
Methods Thirty-five women were randomised to receive 200, 400 or 600 mcg of sublingual misoprostol PPH prophylaxis immediately following delivery. These were compared with 14 consecutive women given 10 IU of intramuscular oxytocin. Immediately after placental delivery, a Koala intra uterine pressure catheter was inserted transcervically into the uterine cavity.
Main outcomes measures Main outcomes measures are the uterine pressure (in Montevideo units) measured over 120 minutes. Other outcomes included temperature and measured blood loss.
Results Women’s age, parity, gestational age and neonatal birth weight were not significantly different between the four groups. There was no difference in intrauterine pressure between the three misoprostol doses. However, the uterine pressure was significantly lower than oxytocin with all three doses for the first 10 minutes (P < 0.008) and significantly higher than oxytocin from 50 to 120 minutes (P < 0.008). A dose-related rise in the body temperature and chills was observed in the misoprostol groups, with 8.3%, 8.3% and 45% of women experiencing a fever >39 °C with the 200, 400, and 600 mcg doses respectively.
Conclusion Intramuscular oxytocin has the highest immediate post partum uterine activity. Lower doses of misoprostol may be as effective as high doses and with fewer side effects. Clinical outcomes with low-dose misoprostol should be further explored (ISRCTN97277056).
Postpartum haemorrhage (PPH) is the leading cause of maternal mortality in the developing world and is responsible for about 25% of all maternal deaths worldwide.1 The most common cause of PPH is the failure of the uterus to contract after childbirth (atonic PPH) and active management of the third stage of labour is recommended to prevent this.2,3 This involves, as a minimum, the administration of a uterotonic drug at delivery and controlled cord traction. Oxytocin is recommended as the first-line oxytocic for the prevention of PPH.3,4 However, oxytocin requires refrigeration because it is unstable when exposed to high ambient temperatures. Furthermore, this drug must be given parenterally, which requires a skilled birth attendant and a continuous supply of sterile syringes and needles. Both of these are frequently unavailable in low-resource settings. About 99% of maternal deaths occur in low-resource settings where there are poor transportation systems and a lack of skilled birth attendants and emergency obstetrics services.5 Hence, a major objective for reducing maternal deaths in poor areas is to find low-cost, effective ways to prevent and control PPH.
Misoprostol is an orally active prostaglandin analogue with uterotonic effects, and is an option for PPH prevention in low-resource settings because of its thermostability, cost-effectiveness and ease of administration. There have been at least 36 trials that have studied misoprostol for PPH prevention using doses between 200 and 1000 μg and a variety of routes including oral, vaginal, sublingual and rectal.6 It is, however, as yet unclear which gives the best balance of efficacy and safety. Given the relative ineffectiveness of misoprostol for PPH, the tendency has been to use high doses. There are, however, potential dangers in this. Shivering and pyrexia are commonly reported adverse effects, and hyperpyrexia of over 40°C has been reported, reaching an incidence of 36% in some population groups.7 A recent systematic review recommends further research to find the optimal route and minimum effective dose of misoprostol for routine use for the prevention of PPH.8 The authors suggest that the sublingual route is likely to be the most suitable because of rapid uptake, prolonged duration of action and greater bioavailability.
This study was devised to compare the effects of three sublingual misoprostol doses using measurement of postpartum intrauterine pressure as a surrogate endpoint to evaluate the uterine activity of these uterotonics. We also compared the adverse effects associated with each treatment. Data from a small cohort of women given intramuscular oxytocin prophylaxis are also shown for comparison with the misoprostol data. We hypothesised that low doses of misoprostol produce a similar strength of myometrial contraction to high doses.
The study population was made up of women who gave birth at Zliten Teaching Hospital (in Zliten, Libya) between July and December 2009. This is a government-funded, secondary referral hospital serving a population of 200 000 people with 4500 deliveries per year. Before the study, a small pilot study was conducted in a nearby teaching hospital (in Misurata) to test and perfect the study equipment and data acquisition methodology. The data from this pilot study were also used to calculate the sample size for the main study. At the end of this small pilot study, the researcher moved to Zliten Teaching Hospital and the ethical approval was amended accordingly. During the trial set-up period for the main study at this new site (pilot data analysis, recalculation of samples sizes and development of the randomisation schedule and envelopes), 14 women were recruited to an observational study at Zliten where women were treated with oxytocin alone. Once all the study instruments were in place, the main randomised study commenced and women entering the delivery suite when the researcher was present (usual working hours during the week) were invited to participate.
Women who were 18 years old or over, who had a spontaneous onset of labour and no risk factors for PPH were approached upon arrival at hospital in early labour and invited to participate in the randomised study. If they agreed, informed consent was obtained. Risk factors for PPH included a history of PPH in a previous pregnancy, a history of antepartum haemorrhage in the current pregnancy, a previous caesarean delivery, multiple pregnancy and polyhydraminos, Women with anaemia (haemoglobin < 10.0 g/dl) or maternal infection were also excluded (Figure 1). Intrapartum exclusions were augmentation with oxytocin, instrumental delivery or caesarean section.
Women who gave birth to babies weighing >4 kg were not randomised. To enable this, the baby was weighed immediately after delivery, and the study drug was given slightly later than originally described in the protocol so as to allow the neonate’s weight to be obtained. If the birthweight was ≤4 kg then the randomisation envelope was opened and the study drug was administered. In all women the drug was administered within a minute of birth. This change to the original protocol, made at the request of the local committee, resulted in an inconsistency that required both delivery of the study drug at the point of delivery and knowledge of the birthweight before randomisation. To correct this inconsistency, it was decided to delay the drug administration for a maximum of 60 seconds to allow rapid weighing and randomisation as described. This was done in the knowledge that the main effect of misoprostol does not occur until 20–30 minutes after administration so the effect of a 1-minute delay was likely to be minimal. Furthermore, as the delay occurred in all randomised women as well as in the cohort given oxytocin, it would have no effect on between-group comparisons.
A commercial randomisation programme was used to produce a random list of allocations to three doses of misoprostol (http://www.sealedenvelope.com). The allocations were written on cards and placed in consecutively numbered sealed opaque envelopes by staff not involved in the study. Blinding was not possible because the treatment was provided by the researcher who was available at the time of the delivery to carry out the final selection of women and collection of data.
The delivery of the baby was left entirely to the midwives. Routine practice is to give the uterotonic at the delivery of the anterior shoulder and deliver the placenta using controlled cord traction. Once the baby is delivered, the cord is cut and the baby is weighed immediately. All babies are then transferred to the neonatal unit for 2 hours of observation, even if the birth and immediate neonatal period were completely normal. Once the woman is transferred to the postnatal ward (after 2 hours of observation on the labour ward), the baby rejoins her and she is then able to start breastfeeding. This routine was not changed for women in the study except that the study drug was given by the researcher immediately after the baby was weighed (which occurred within seconds of delivery) rather than at delivery. For women receiving misoprostol, the oxytocin was omitted.
Once the woman was determined as eligible, the next successive treatment envelope was opened and the designated treatment was given. We randomly allocated eligible women to 200, 400 or 600 μg sublingual misoprostol (Cytotec, Pfizer, Italy). The tablets were moistened with tap water before being placed under the tongue. At the same time, an ‘under-buttocks drape with fluid-collecting pouch’ (Kimberly-Clark, Kent, UK) was placed under the woman’s buttocks for collection of any blood over the next 120 minutes. Immediately after placental delivery, an intrauterine catheter (Koala External Balloon Catheter IPC-5000E; Clinical Innovations, Salt Lake City, UT, USA; Figure 2) was inserted manually through the cervix into the uterine cavity until the tip of the catheter could be felt to touch the fundus. The catheter was secured in place with tape to the mother’s thigh and connected to a Corometrics 118 maternal/fetal monitor (Corometrics Medical Systems Inc., Wallingford, CT, USA). The uterine activity was recorded and saved over the next 120 minutes. A researcher was with the women throughout the 2 hours of observation to record maternal temperature, pulse and blood pressure before labour, immediately after the delivery and then at 30, 60, 90 and 120 minutes. The women were closely observed for any adverse effects experienced. None of the women initiated breastfeeding until 2 hours after delivery, when they transferred to the postnatal ward.
The primary outcome was intrauterine pressure over the first 10 minutes after delivery. The slight differences in time during which the pressure was measured in the first 10 postnatal minutes (because the catheter was only inserted after placental expulsion) meant that the Montevideo units (MVU) measurement had to be adjusted for the first reading. This was done by adding together the uterine pressures collected in the first 10 minutes after delivery, dividing this by the length of time during which the measurements were taken (to get MVU per minute) and then multiplying by 10. The difference in means of the primary outcome was calculated using one way analysis of variance. The intrauterine pressures from 20 to 120 minutes postpartum were secondary outcomes.
Repeated measures of longitudinal data analysis (repeated measures analysis of variance) were used to compare the effect of the study treatments over 120 minutes of observation. Other mean comparisons used the one-way analysis of variance with multiple comparisons Bonferroni test. The incidence of shivering and fever and other adverse effects among the three misoprostol groups were compared using chi-square test with Yates’ correction and Fisher exact test as appropriate. Statistical analysis was performed using Excel and PASW Statistics 17 (Excel, Microsoft Corp., Redmond, WA, USA; PASW 17, IBM Corp., Somers, New York, NY, USA). The difference between the four groups’ parameters was taken as statistically significant when P values were <0.05. Based on the primary outcomes (mean intrauterine pressure over the first 10 minutes) and to detect a significant difference (P = 0.05, two-sided) at 0.8 power with 50% effect size, we needed a sample size of 12 in each group. The calculation was performed by G*Power 3.0.10 software (G*Power, Dusseldorf, Germany).
The study was approved by the University of Liverpool Ethics Committee (RETH000237) and accepted by the local hospital committees in Zliten and Misurata Teaching Hospitals.
A total of 35 women were randomised out of a planned sample of 36. The study was curtailed before the planned finish when the investigator had to return to the UK. The randomised women were compared with a cohort of 14 women treated with oxytocin before the start of the randomised study. The age, parity, gestational age and baby’s birthweight were similar in the four groups (Table 1).
Table 1. Basic characteristics of four treatment groups
Oxytocin 10 IU IM
Date are mean ± SD unless otherwise stated.
No. of women recruited
Parity, n (%)
28.5 ± 6.2
28 .1 ± 5.4
27 ± 3.3
26.3 ± 3.7
Gestational age (weeks)
40.2 ± 1.4
40.1 ± 1.1
40.3 ± 0.8
41 ± 1.1
3350 ± 0.364
3458 ± 0.262
3595 ± 0.397
3522 ± 0.337
All three doses of sublingual misoprostol produced a rapid increase in uterine activity with a peak at 40 minutes followed by a gradual decrease over the following 80 minutes (Figure 3). Throughout the observations, there was no significant difference between the intrauterine pressure in the three misoprostol dosage groups (P = 0.8).
The uterine pressure in those receiving oxytocin was highest in the immediate postpartum period and declined gradually after 40 minutes of administration. In the first 10 minutes, intrauterine pressures of the three misoprostol groups were significantly lower than those of the oxytocin group (P = 0.008). Conversely, the uterine pressure over the period from 50 to 120 minutes was significantly higher in the three misoprostol groups than in the oxytocin group (P = 0.008).
All the women were closely observed for adverse effects. Women who received intramuscular oxytocin reported no adverse effects (Table 2). In the misoprostol groups, the two most commonly reported adverse effects were shivering and hyperthermia, and a dose-related rise in the body temperature was observed. The incidence of severe hyperthermia (>39°C) was higher in the 600-μg misoprostol group than the 200- and 400-μg misoprostol groups, but this was not statistically significant (P = 0.1; Figure 4). Most of the women were not aware of the hyperthermia, but complained of coldness and marked shivering. All women in the 400- and 600-μg groups had shivering compared with 75% of the women in the 200-μg group. The differences were statistically significant (P = 0.04).
Table 2. Incidence of adverse effects in the different treatment groups, n (%)
Oxytocin 10 IU IM
No. of women recruited
Abdominal pain requiring analgesia
Nausea and vomiting
The blood loss was <500 ml in all women and was normally distributed. The lowest mean blood loss (±SD) was observed in women who received 10 IU oxytocin (193 ± 35 ml), followed by those who received 200 μg sublingual misoprostol (239 ± 46 ml), 600 μg sublingual misoprostol (275 ± 62 ml) and 400 μg sublingual misoprostol (299 ± 60 ml). The difference was statistically significant between oxytocin and 400 μg misoprostol (P < 0.001), oxytocin and 600 μg misoprostol (P < 0.001) and also between the three misoprostol groups (P = 0.04). None of the women received additional uterotonics or blood transfusion.
This randomised trial was conducted to find out whether high-dose misoprostol offered any significant improvements in postpartum uterine pressure over low-dose misoprostol, and to compare the adverse drug reactions associated with each treatment. We also had concurrently collected data on intrauterine pressures following oxytocin administration, and this allowed us to compare the efficacy of these two types of oxytocics.
We used the Koala External Balloon sensor system9,10 to measure uterine activity (Figure 2). All previous postpartum intrauterine pressure measurements have been conducted using a Galtec tipped catheter where a small pressure sensor is mounted into the tip of the device.11–14 The Galtec tipped catheter was designed to measure intrauterine pressure in the first stage of labour. The pressure sensor is located in a recess within the head of the catheter, and it detects the myometrial contraction through an increase in the pressure of the surrounding amniotic fluid within the almost closed cavity. In the third stage of labour, however, the uterus is open and the amount of fluid is limited because it is squeezed out during each contraction. The Galtec tipped catheter therefore functions poorly in this situation. For accurate measurements it is logical for any measurement device to have a large surface area in direct contact with the adjacent uterine wall. The Koala catheter has a 3-cm air-filled balloon near its tip that connects directly to an external reusable transducer mounted into the connecting cable. The balloon sits in direct contact with the uterine wall and is compressed directly during the contraction.
The mean intrauterine pressure over 120 minutes for the intramuscular oxytocin and the three doses of sublingual misoprostol are shown in Figure 3. The difference seen in speed of onset of action can be explained both by the different pharmacokinetics of these drugs and by the different routes of administration. The absorption of misoprostol is affected by mouth dryness and any surrounding fluid;15,16 we standardised for this by moistening the tablets with tap water before placing them under the tongue. The pharmacokinetics may explain our observation of stronger and more frequent initial uterine contractions with intramuscular oxytocin than with sublingual misoprostol. The peak of uterine contraction for oxytocin was within the first 10 minutes after administration whereas for sublingual misoprostol it was only achieved after 30–40 minutes. This mirrors their plasma concentrations.15,17
Most uterine bleeding occurs immediately after placental separation.18 At this time, the mean intrauterine pressure for intramuscular oxytocin was significantly higher than with the three sublingual misoprostol doses. Even though the effective action of sublingual misoprostol started late, it caused high uterine contractions maintained over a considerable period of time. This may help to prevent steady moderate bleeding, which may be unobserved until serious hypodynamic manifestations occur.
Very few studies have examined the effect of misoprostol on uterine activity during the third stage of labour using measurement of intrauterine pressure as an indicator. Chong et al., using a Galtec tipped catheter, found no difference in postpartum uterine pressures between women given oral misoprostol and intramuscular syntometrine.11,12 The readings with this catheter, however, may not be reliable for the reasons outlined above.
The mean intrauterine pressure of the three different doses of sublingual misoprostol was not significantly different over the 2-hour observation period. This is consistent with previous research findings. For example, in one previous study there was no difference in intrauterine pressure measurements with five different doses of oral misoprostol, although again this used a Galtec tipped catheter.11 Furthermore, randomised control trials found that both 600 μg sublingual and 400 μg oral misoprostol were more effective than placebo for the prevention of PPH,19,20 and a meta-analysis in which an indirect comparison was made between the 400 and 600 μg oral doses concluded that the two were likely to be equally effective.21 As a result of adverse effects associated with high doses of sublingual misoprostol, some researchers have argued that the dosage of misoprostol should be reduced from the 600 and 800 μg doses in common usage today. Our research findings give further weight to that argument.
The most commonly observed adverse effects in this study were shivering and hyperthermia. Around half of the women who received 600 μg had a temperature >39°C whereas the incidence for women who received 400 and 200 μg was around 8%. Most of the women were not aware of the hyperthermia, but complained of coldness and intolerable shivering (all women in the 400 and 600 μg groups had shivering). Although there appeared to be clinical differences in the incidence of hyperthermia between high and low doses, the difference was not statistically significant because the study was not powered enough to detect differences in incidence of adverse effects. Shivering and hyperthermia have been reported in most of the studies using misoprostol for different indications and with variable routes and doses. In a recent multicentre randomised control trial using 800 μg sublingual misoprostol for treatment of PPH, the overall incidence of fever (>40°C) was 14%. However, the incidence of fever varied between different populations with the highest incidence in Ecuador (36%) and the lowest (0%) in Egypt.7,22 In our study, the incidence of adverse effects appeared to be dose-related. Prostaglandins are the principal mediator of fever in the brain and can pass the blood–brain barrier to the thermoregulation centres in the hypothalamus, causing elevation of the thermoregulatory set point. To increase the body temperature to the new set point, there are increases in heart rate, muscle tone and shivering.23–25
Clinical trials have shown that oxytocin prophylaxis is more effective than oral misoprostol for the prevention of blood loss >1000 ml.26 Given that most blood is lost around the time of placental expulsion during the first 10 postpartum minutes, it is not surprising that oxytocin is the more effective prophylactic.17 It is only after the first 50 minutes that the uterine contraction strength was higher in the misoprostol groups. However, although the effect of misoprostol is delayed, it should be eventually as effective as oxytocin in preventing massive blood loss through atony. There is evidence for this from a Cochrane Review in which there are significant differences between oxytocin and misoprostol in blood loss >500 and >1000 ml, but no difference in need for blood transfusion.6
We acknowledge that this study recruited only low-risk women in whom the chance of developing an atonic PPH was small. However, even in a study that included high-risk women, the majority of women would not develop an atonic PPH—and this is the very group for which the intervention is designed. However, the area of interest in this study is the comparison between drug dosages and, as such, it is important to have as homogeneous a group of women as possible. This was best achieved through the use of a low-risk group. Furthermore, it was considered by the ethical committee to be inappropriate to expose women at risk of PPH to misoprostol, which has been shown to be less effective for prophylaxis than oxytocin.
In conclusion, our results showed that 200, 400 and 600-μg doses of sublingual misoprostol produced similar levels of uterine activity, but that the severity of adverse effects was dose-related. These findings suggest that lower doses of misoprostol may be as effective as high doses. Clinical applications of low doses of sublingual misoprostol for the prevention of PPH should be further explored by large randomised trials comparing the effectiveness and the safety of low doses of sublingual misoprostol.
Disclosure of interests
The authors have no conflicts of interest to declare. AW runs the independent http://www.misoprostol.org website to promote the appropriate use of misoprostol, but this does not receive any support from the pharmaceutical industry.
Contribution to authorship
AE participated in the study design, carried out the study procedure and wrote the manuscript. MSE assisted with running the study in Libya and the final approval for publication. MOE and OAE assisted with the recruitment of the participants and data acquisition of the study in Libya. AW designed the study, revised the manuscript providing comments on its intellectual content and helped draft the manuscript and the final approval for publication.
Details of ethics approval
The study was granted ethical approval from the University of Liverpool ethics committee (RETH000237) and was accepted by the local hospitals in Libya on 14 July 2009 and 24 August 2009.
Funding was received from the University of Liverpool, UK.
We acknowledge the help and support received from the doctors and midwives at Misurata and Zliten Teaching Hospitals, Libya. Our thanks also go to Mr David Cordon from the Bioengineering Department in Liverpool Women’s Hospital for his help with adapting the Corometrics 118 for use in Libya. We are grateful to all the women who participated in this study.