Postpartum hemorrhage is a major cause of maternal death and severe morbidity, particularly in low-resource countries.1 Data from the Chinese National Maternal and Child Mortality Surveillance System indicate that postpartum hemorrhage contributed to 31.5% (997/3,164) of all maternal deaths from 1996 to 2006, making it the leading cause of maternal mortality in China.2
The World Health Organization (WHO), together with the International Confederation of Midwives and the International Federation of Gynecology and Obstetrics recommends active management of the third stage of labor to prevent postpartum hemorrhage. The recommended measures include prophylactic use of uterotonics after delivery of the shoulder and controlled cord traction and uterine massage after delivery of the placenta.3,4 The effectiveness of uterotonics in preventing postpartum hemorrhage has been well established.5–7 Recently, Gülmezoglu et al8 conducted a large, multicenter, noninferiority, randomized controlled trial that justified the omission of controlled cord traction from active management of the third stage of labor. However, the effectiveness of uterine massage remains unclear.
Two randomized controlled trials assessed the effectiveness of uterine massage for preventing postpartum hemorrhage after vaginal delivery with conflicting results.9,10 However, one trial was underpowered because of a small sample size and the other investigated uterine massage applied before rather than after delivery of the placenta. Therefore, we performed a multicenter randomized controlled trial to estimate the effectiveness of sustained transabdominal uterine massage, started after delivery of the placenta, in reducing blood loss after vaginal delivery.
MATERIALS AND METHODS
This was a randomized controlled trial conducted in four university hospitals (West China Second University Hospital of Sichuan University, Southwest Hospital of Third Military Medical University, Shanghai First Maternity and Infant Hospital of Tongji University, and Women's Hospital School of Medicine Zhejiang University) in China between March 2012 and November 2012. The protocol was approved by Ethics Committee of West China Second University Hospital of Sichuan University (approval date March 16, 2012, approval number M-2012-005). This trial is registered with one of the primary registries in the WHO Registry Network, Chinese Clinical Trial Registry (ChiCTR.org#ChiCTR-TRC-11001763). All participants provided written informed consent.
Women who were expected to deliver vaginally were evaluated for potential participation. Consent was obtained before full dilation of the cervix if the woman was eligible for recruitment and willing to participate after discussing the study with a midwife or a resident. Inclusion criteria included women giving birth vaginally, aged between 20 and 35 years, with a singleton pregnancy, gestational age of 36 or more weeks, and with fewer than three previous childbirths. Exclusion criteria included fetal malpresentation, emergent cesarean delivery, the third stage of labor lasting more than 30 minutes, a previous cesarean delivery, three or more previous induced abortions, hypertensive disorders of pregnancy, fetal death, an amniotic fluid index of 20 cm or more, hemoglobin of 90 g/L or less before delivery, coagulopathy, and having an abdominal wall too thick to palpate the fundus of the uterus or to perform uterine massage after delivery. Screening for eligibility was accomplished by checking the medical record and directly questioning the woman. Baseline information was collected from each participant.
The participants were randomized into an oxytocin plus uterine massage group and an oxytocin-only group shortly after delivery of the placenta. The random allocation sequence was computer-generated at West China Second University Hospital. Randomization was stratified by each participating hospital according to its annual number of childbirths (ie, a 1.4:1.4:1.4:1.0 ratio) and restricted with a random block of six. Numbered opaque envelopes containing randomized allocations were kept in the labor room in each study site. As a result of the nature of the intervention, neither participants nor research staff could be blinded to group allocation.
In both groups, 10 units of oxytocin were given intramuscularly immediately after delivery of the shoulder and controlled cord traction was performed to assist placental delivery for all participants, per WHO guidelines.3 The umbilical cord was clamped and cut approximately 1 minute after delivery of the neonate in all study sites. The uterine fundus was rubbed and blood clots expelled as quickly as possible after delivery of the placenta. Once the placenta was delivered, the next numbered opaque envelope was opened and the allocation revealed. Those allocated to the oxytocin-only group were managed expectantly, whereas women allocated to the uterine massage group were provided with 30 minutes sustained transabdominal uterine massage starting promptly after placental delivery. At least two research staff stayed with the participant to ensure 30-minute sustained transabdominal uterine massage. The massage was performed as follows: finding the uterine fundus, manually stimulating the fundus and the whole body of the uterus using fingers and palms steadily and repetitively, and trying not to cause discomfort to the woman. Postpartum hemorrhage was treated according to the local policy of each study site, mainly based on WHO guidelines.11 All participants were discharged no earlier than 24 hours after delivery.
The procedure for collection and weighing of lost blood was similar to that of a previous study.12 Collection of lost blood was initiated immediately after delivery of the neonate by putting a drape under the woman's buttocks. Blood collected was weighed on an electronic scale together with the drape, and the amount was recorded in grams after subtracting the weight of the drape. The amount of blood loss was recorded 2 hours after delivery of the neonate. An additional drape was placed under the woman's buttocks whenever needed. The amount of lost blood in grams was converted to volume by dividing by 1.06 (blood density in g/mL) for the analysis.13
The primary outcome was blood loss of 400 mL or more in the 2 hours after delivery of the neonate. Secondary outcomes were blood loss of 1,000 mL or more in the 2 hours after delivery, blood loss in milliliters in the 2 hours after delivery, use of therapeutic uterotonics, use of other hemostatic procedures (eg, uterine tamponade, bimanual compression of the uterus, surgical procedures, uterine artery embolization) for postpartum hemorrhage, hemoglobin lower than 80 g/L before discharge, and need for blood transfusion.
The investigators collected the data and entered them into a database in EpiData 3.1 in a double-entry and validation manner, ie, two investigators entered the data independently and all inconsistent data were resolved by checking the participants' obstetric and nursing records. Accuracy of information was ensured by checking each participant's medical records.
The sample size was determined according to the results of our previous study, which showed a 9.6% incidence of blood loss of 500 mL or more within 24 hours after a vaginal delivery.14 We assumed the incidence of blood loss of 400 mL or more within 2 hours after vaginal delivery to be approximately this high. On the basis of this, inclusion of at least 1,061 women in each group would be required to detect a reduction from 9.6% to 6.6% in the primary outcome in the uterine massage group with a one-sided α of 0.05 and a power of 0.8. Factoring in a 10% dropout rate, a total of 1,170 women would be required in each arm for adequate power.
Data in the EpiData database were transferred into SPSS 21.0 for analysis. Analysis was by intent to treat; participants were analyzed in the groups to which they were initially randomized no matter whether they received the assigned intervention or not. Per-protocol analysis was also performed to compare outcomes between the two groups. Categorical data were analyzed with Pearson’s χ2 test or Fisher’s exact test where appropriate and continuous data were analyzed with Student’s t test. Relative risks (RRs) and 95% confidence intervals (CIs) of all outcomes are reported. The threshold for statistical significance was set at P=.05.
Between March 2012 and November 2012, a total of 3,829 women were assessed for eligibility, and 2,340 were included in our study. They were randomized into the oxytocin plus uterine massage group (1,170 women) or the oxytocin-only group (1,170 women). Figure 1 shows participant flow through this trial. Overall, the adherence to the allocated intervention was high in both groups: 98.6% (1,154/1,170) in the uterine massage group and 99.4% (1,163/1,170) in the oxytocin-only group, respectively. Massage was discontinued in 16 women in the uterine massage group because of discomfort when they were receiving uterine massage. In the oxytocin-only group, seven women received sustained uterine massage and bimanual uterine compression for the treatment of postpartum hemorrhage. No woman was lost to follow-up in the study.
There were no significant differences between the two groups with regard to baseline characteristics including maternal age, height, weight at admission, gestational age at delivery, weight of the newborn or use of labor induction, operative vaginal delivery, episiotomy or perineal tear, length of labor, manual removal of placenta, use of curettage, or hemoglobin before delivery (Table 1). Most of the participants were primiparous women and received episiotomy, and approximately half received labor analgesia.
Intent-to-treat analysis showed that the incidence of blood loss of 400 mL or more in the 2 hours after delivery of the neonate was not significantly different between the two groups (12.2% compared with 12.3%; RR 0.99, 95% CI 0.88–1.13, P=.95) (Table 2). The average amount of blood loss in the 2 hours after delivery was higher in the uterine massage group than in the oxytocin-only group, but the difference was not statistically significant. The incidences of blood loss of 1,000 mL or more in the 2 hours after delivery, administration of therapeutic uterotonics, other hemostatic procedures for postpartum hemorrhage, hemoglobin of lower than 80 g/L before discharge, or blood transfusion also did not differ significantly between the two groups. Per-protocol analysis also showed that neither the primary nor any of the secondary outcomes differed significantly between the two groups.
In the uterine massage group, 378 (32.3%) women reported pain or discomfort when receiving uterine massage and 16 (1.4%) asked to have massage discontinued as a result of discomfort. No severe maternal morbidity such as admission to the intensive care unit, sepsis or hysterectomy, or maternal death occurred in this trial.
Uterine massage after delivery of the placenta is a part of active management of the third stage of labor according to the guidelines of WHO and the International Confederation of Midwives and the International Federation of Gynecology and Obstetrics.3,4 It may not be routinely used in developed countries where effective preventive and therapeutic uterotonics are readily available, whereas in many underresourced areas (eg, Egypt, Thailand, and Uganda), it remains a routine practice.8 The results of our study indicate there may be no need for routine uterine massage after vaginal delivery, because it neither reduced the amount of blood loss nor decreased the incidence of the primary and secondary outcomes. Instead, both the incidence of use of therapeutic uterotonics and the average amount of blood loss in the 2 hours after delivery were higher in the uterine massage group, although the differences between the groups were not statistically significant.
Only two previous randomized controlled trials assessed the effectiveness of uterine massage in preventing postpartum hemorrhage, and they were conducted by the same research team. The earlier trial reported that mean blood loss was reduced at 30 minutes (168.8 mL compared with 210.4 mL, P=.02) and 60 minutes (204.3 mL compared with 281.7 mL, P<.001) after delivery, and fewer women needed additional uterotonics (6% compared with 25%; RR 0.20, 95% CI 0.08–0.50, P<.001), but the incidence of blood loss of over 500 mL was not reduced (5% compared with 7%; RR 0.52, 95% CI 0.16–1.67) in the uterine massage group.9 However, this trial only included 200 women and therefore was underpowered to detect a difference between the two groups with regard to incidence of blood loss of over 500 mL. For this reason, the same team conducted another randomized controlled trial with a total of 1,964 women and found that uterine massage was inferior to oxytocin in reducing blood loss after delivery and that massage did not bring about additional benefit when oxytocin was administered.10 Our findings are partly in accordance with those of the latter trial. The difference is that we started uterine massage after delivery of the placenta in accordance with international guidelines, whereas in the previous trial, massage was performed before delivery of the placenta. The frequency of the use of therapeutic uterotonics in our trial (approximate 20%) was much higher than that of either the former (15.5%) or the latter trial (4.9%), but we were unable to discern the underlying reason for this. Nonetheless, the frequency of use of therapeutic uterotonics in our study was similar to that in another large randomized controlled trial (20.4%).8
In our study, the incidence of manual removal of placenta (8.3% and 8.8% in each group) and curettage (5.3% and 3.8%) was higher than those of many previous studies, which ranged from 0.008% to 8.0%5,7,15,16 and from 0.8% to 1.4%,8 respectively. A possible reason for the high incidences of manual removal of placenta and curettage was the high rate of previous induced abortion (31.5% and 34.0%), because previous abortion has been recognized as one of the major independent risk factors for retained placenta.17,18 Although manual removal of placenta and curettage might result in more blood loss, we believe this would be unlikely to affect our conclusion because the incidence of manual removal of placenta and curettage was similar in the two groups.
The strengths of our study include the fact it is a multicenter randomized controlled trial and the relatively large sample size. Each study site is a university hospital, and some of them have cooperated in previous randomized controlled trials.14,19 The incidence of the primary outcome (12.3%) in this trial was somewhat higher than that used for sample size calculation (9.6%), increasing our power to detect differences between groups.
Despite these strengths, our study has several limitations. First, our study would be underpowered if we had used blood loss of 1,000 mL or more (ie, severe postpartum hemorrhage) after delivery as the primary outcome as other studies had.8,20,21 Second, there was potential bias because blinding was impossible. Third, this trial was carried out in four university hospitals where effective uterotonics and other hemostatic methods for treating postpartum hemorrhage were readily available, so clinical practitioners should be cautious when applying our findings in resource-poor areas.
Our findings indicate that routine uterine massage provides no additional benefit when combined with oxytocin administration and therefore may not be a necessary step for the prevention of postpartum hemorrhage after vaginal delivery. Performing uterine massage for every woman is a time-consuming and strenuous procedure. Elimination of routine uterine massage from active management of the third stage of labor will help obstetric staff spare more time and efforts for other work.
1. Khan KS, Wojdyla D, Say L, Gülmezoglu AM, Van Look PFA. WHO analysis of causes of maternal death: a systematic review. Lancet 2006;367:1066–74.
2. Feng XL, Zhu J, Zhang L, Song L, Hipgrave D, Guo S, et al.. Socio-economic disparities in maternal mortality in China between 1996 and 2006. BJOG 2010;117:1527–36.
3. World Health Organization. WHO recommendations for the prevention of postpartum haemorrhage. Geneva (Switzerland): World Health Organization; 2007.
4. Lalonde A, Daviss BA, Acosta A, Herschderfer K. Postpartum hemorrhage today: ICM/FIGO initiative 2004–2006. Int J Gynaecol Obstet 2006;94:243–53.
5. Etbourne DR, Prendiville WJ, Carroli G, Wood J, McDonald S. Prophylactic oxytocin for the third stage of labour. The Cochrane Database of Systematic Reviews 2001, Issue 4. Art. No.: CD001808. DOI: 10.1002/14651858.CD001808.
6. Liabsuetrakul T, Choobun T, Peeyananjarassri K, Islam QM. Prophylactic use of ergot alkaloids in the third stage of labour. The Cochrane Database of Systematic Reviews 2007, Issue 18. Art. No.: CD005456. DOI: 10.1002/14651858.CD005456.pub2.
7. Tunçalp Ö, Hofmeyr GJ, Gülmezoglu AM. P rostaglandins for preventing postpartum haemorrhage. The Cochrane Database of Systematic Reviews 2012, Issue 8. Art. No.: CD000494. DOI: 10.1002/14651858.CD000494.pub4.
8. Gülmezoglu AM, Lumbiganon P, Landoulsi S, Widmer M, Abdel-Aleem H, Festin M, et al.. Active management of the third stage of labour with and without controlled cord traction: a randomised, controlled, non-inferiority trial. Lancet 2012;379:1721–7.
9. Abdel-Aleem H, Hofmeyr GJ, Shokry M, El-Sonoosy E. Uterine massage and postpartum blood loss. Int J Gynaecol Obstet 2006;93:238–9.
10. Abdel-Aleem H, Singata M, Abdel-Aleem M, Mshweshwe N, Williams X, Hofmeyr GJ. Uterine massage to reduce postpartum hemorrhage after vaginal delivery. Int J Gynaecol Obstet 2010;111:32–6.
11. World Health Organization. WHO guidelines for the management of postpartum haemorrhage and retained placenta. Geneva (Switzerland): World Health Organization; 2009.
12. Patel A, Goudar SS, Geller SE, Kodkany BS, Edlavitch SA, Wagh K, et al.. Drape estimation vs. visual assessment for estimating postpartum hemorrhage. Int J Gynaecol Obstet 2006;93:220–24.
13. Althabe F, Alemán A, Tomasso G, Gibbons L, Vitureira G, Belizán JM, et al.. A pilot randomized controlled trial of controlled cord traction to reduce postpartum blood loss. Int J Gynaecol Obstet 2009;107:4–7.
14. Lin JH, Lin QD, Liu XH, Yan JY, He J, Li L, et al.. The multi-center clinical observation of the motherwort injection (Yimucao Zhusheye) for preventing postpartum hemorrhage and promoting the uterine contraction after vaginal delivery [in Chinese]. J Pract Obstet Gynecol 2009;44:175–8.
15. Cheung WM, Hawkes A, Ibish S, Weeks AD. The retained placenta: historical and geographical rate variations. J Obstet Gynaecol 2011;31:37–42.
16. McDonald S, Abbott JM, Higgins SP. Prophylactic ergometrine-oxytocin versus oxytocin for the third stage of labour. The Cochrane Database of Systematic Reviews 2004, Issue 1. Art. No.: CD000201. DOI: 10.1002/14651858.CD000201.pub2.
17. Endler M, Grünewald C, Saltvedt S. Epidemiology of retained placenta: oxytocin as an independent risk factor. Obstet Gynecol 2012;119:801–9.
18. Owolabi AT, Dare FO, Fasubaa OB, Ogunlola IO, Kuti O, Bisiriyu LA. Risk factors for retained placenta in southwestern Nigeria. Singapore Med J 2008;49:532–7.
19. Lin JH, Lin QD, Liu XH, Yan JY, He J, Li L, et al.. Multi-center study of motherwort injection to prevent postpartum hemorrhage after caesarian section [in Chinese]. Zhonghua Fu Chan Ke Za Zhi 2009;25:175–8.
20. Zhang WH, Deneux-Tharaux C, Brocklehurst P, Juszczak E, Joslin M, Alexander S, et al.. Effect of a collector bag for measurement of postpartum blood loss after vaginal delivery: cluster randomised trial in 13 European countries. BMJ 2010;340:c293.
21. Begley CM, Gyte GM, Devane D, McGuire W, Weeks A. Active versus expectant management for women in the third stage of labour. The Cochrane Database of Systematic Reviews 2011, Issue 11. Art. No.: CD007412. DOI: 10.1002/14651858.CD007412.pub3.