SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Objective To compare intravenous oxytocin administration (Partocon® 10 IU) with saline solution in the management of postpartum haemorrhage in the third stage of labour.

Design A double-blind, randomised controlled trial involving 1000 parturients with singleton fetuses in cephalic presentation and undergoing vaginal delivery, randomly allocated to treatment with oxytocin (n=513) or 0.9% saline solution (n=487).

Setting Labour ward at a central county hospital.

Main outcome measures Mean blood loss (total, and before and after placenta delivery); frequencies of blood loss > 800 mL, need of additional oxytocic treatment, postpartum haemoglobin < 10 g/dL; and duration of postpartum hospitalisation.

Results As compared with saline solution, oxytocin administration was associated with significant reduction in mean total blood loss (407 versus 527 mL), and in frequencies of postpartum haemorrhage > 800 mL (8.8%versus 15.2%), additional treatment with metylergometrine (7.8%versus 13.8%), and postpartum Hb < 10 g/dL (9.7%versus 15.2%), and a nonsignificant increase in the frequency of manual placenta removal (3.5%versus 2.3%). There was no group difference in the mean duration of postpartum hospitalisation (4.6 versus 4.5 days, respectively).

Conclusions Administration of intravenous oxytocin in the third stage of labour is associated with an approximately 22% reduction in mean blood loss, and approximately 40% reductions in frequencies of postpartum haemorrhage (> 500 mL or >800 mL) and of postpartum haemoglobin < 10 g/dL. Identification of risk groups for oxytocin treatment does not seem worthwhile. Oxytocin is a cheap atoxic drug and should be given routinely after vaginal delivery.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Postpartum haemorrhage remains an important cause of maternal mortality in developing countries, though the mortality rate has dramatically decreased in Western countries during the twentieth century. This change has coincided with the introduction of oxytocic drugs in the management of the third stage of labour1,2, though other factors may also have contributed to this improvement (e.g. improved nutritional status and socio-economic standards, and the availability of blood transfusion services). Nonetheless, postpartum haemorrhage is still associated with substantial maternal morbidity, mainly due to anaemia and infections. Meta-analysis of controlled trials has shown the frequency of postpartum haemorrhage to be reduced by 40% when routine oxytocic treatment is used3.

The clinical significance of the diagnosis postpartum haemorrhage is still open to question. Previously reported figures for the frequency of postpartum haemorrhage have been high: 5% to 16% with no treatment and 2.5% to 14.8% with active treatment4–8. Another question which arises is whether the cutoff limit used (postpartum haemorrhage > 500 mL) really identifies increased maternal morbidity in a healthy population of well-nourished gravidae, where chronic infection and concomitant anaemia are rare. In Sweden there has been a trend towards less intervention during labour. During the Spring of 1994 a nationwide questionnaire was sent to all 55 labour wards in Sweden asking how they managed third stage of labour, and all units replied. A total of 34 labour wards did not routinely give oxytocic drugs (L. Nordstrom, unpublished data). The only support in the literature for such a development is findings in a study from Thilaganathan et al.9 who concluded that routine oxytocin with ergometrine (Syntometrine®, Sandoz Pharmaceuticals, Canberra, Surrey) administration in the third stage of labour did not reduce blood loss, as compared with physiological management.

The aims of the present controlled study were to ascertain whether the use of oxytocin (Partocon®, Ferring, Malmö, Sweden) in the third stage of labour significantly decreased the incidence of postpartum haemorrhage (with an arbitrary cutoff level of > 800 mL), the need of additional oxytocic treatment or of puerperal blood transfusion, and the duration of puerperal hospitalisation in a series of vaginally delivered mothers. A further aim was to ascertain whether any particular subgroups of parturients, divided according to parity, delivery of macrosomic babies, or a history of postpartum haemorrhage in previous deliveries, would benefit more than others from routine treatment with oxytocin.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Sample size and randomisation

In a pilot survey at our hospital during the first three months of 1992, when intravenous administration of 10 IU oxytocin was given routinely, we found 24% (106/451) of the parturients to have postpartum haemorrhage > 500 mL, and 12% (55/451) to have postpartum haemorrhage of > 800 mL. To test the hypothesis that the number of parturients with postpartum haemorrhage of more than 800 mL is increased by 50% if no active treatment is given (two-sided testing at a significance level of 0.05, and with a power of 0.80), the sample size needed would be some 450 cases in each group. We therefore enrolled a total of 1000 parturients in the present study.

Two sets of ampullas were prepared at the hospital pharmacy, one set with 1 mL of oxytocin 10 IU, the other with 1 mL of 0.9% saline solution, and numbered consecutively according to a computer-generated random list. The content of the ampullas was unknown to mothers, midwives and doctors until the study was completed.

Subjects

Between 16 December 1993 and 6 October 1994, all parturients with a singleton fetus in cephalic presentation and undergoing vaginal delivery were asked to participate. Of the 1146 gravidae eligible for the study, 146 preferred not to participate. During the same period, a total of 1281 gravidae were delivered at our department. Of the 1000 ampullas prepared, 48 had to be redone due to leakage, or forgotten/incorrect notation of randomisation number labelling.

Clinical management

Each woman with a singleton pregnancy, and undergoing vaginal delivery of a fetus in cephalic presentation, was given a 1 mL intravenous injection of the content of the consecutive numbered ampulla immediately after delivery. The management of placental delivery was passive. Immediately after the birth of the baby an auxilliary nurse punctured the cord artery to collect blood for acid-base balance or lactate determinations. The cord was clamped only if excessive bleeding occurred from the puncture site. After a majority of all deliveries the baby was placed on the abdomen of the mother for first skin to skin contact. Clamping and cutting of the cord was carried out after some minutes, earlier only if the baby was not vigorous at birth. No routine cord traction was applied, and the parturient was asked to first push when she felt uterine contractions again after the baby was bom. Manual placenta removal was carried out immediately only in the event of profuse postpartum bleeding. With normal bleeding, developments were awaited for 60 minutes before manual removal was undertaken.

In the event of excessive bleeding, first line additional treatment was intravenous injection of 0.2 mg methylergometrine, second line treatment being intravenous infusion of 500 mL glucose 5%+ 30 IU oxytocin, while third line therapy was intravenous infusion of 500 mL glucose 5%+ 0.25 mg prostaglandin F.

As oxytocin infusion for augmentation of labour is common practice, we standardised the postpartum drip rate in augmented cases to 30 mL/hour (1000 mL glucose 5%+ 10 IU oxytocin). A bolus dose of 10 IU oxytocin was sometimes also given as intranasal inhalation during the second stage of labour.

Blood loss volume was calculated by measuring collected blood and adding what was estimated to have been absorbed by surgical cloths and tissues. Main bleeding source from episiotomies or tears was also recorded. The venous maternal haemoglobin value was routinely determined on day two after delivery.

Statistical analyses

Statistical analyses were performed with the statistical package Statistica for Windows(tm), version 4.5 (Stat-Soft, Tulsa, Oklahoma, USA). Data are presented as means and standard deviations (SD) and, when applicable, as odds ratios (OR) with 95% confidence intervals (CI) calculated as previously described10. Group differences in values were assessed with Student's t test for independent samples, and differences in frequencies with the χ2 test. Number needed to treat was calculated according to Chatellier et al.11. AP value of < 0.05 was considered significant.

The study design was approved by the Regional Ethics Committee.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

One thousand parturients were randomly assigned to routine intravenous treatment with oxytocin (n=513) or saline solution (n= 487) directly after delivery. There were no significant group differences in maternal age, parity, gestational age, oxytocin augmentation of labour, instrumental delivery, bleeding mainly from episiotomy/tear, or fetal weight (Table 1). Information on recent maternal usage of acetylsalicylic acid (ASA) was available in 839 cases, no significant difference being found in the usage of ASA between the oxytocin and the saline treatment groups (18/431 and 9/408, respectively; OR 1.93, 95% CI 0.87–4.28). Mean total blood loss (mainly consisting in blood loss before placental delivery) was significantly lower in the oxytocin group than in the saline group (409 mL versus 527 mL, P < 0.001) (Table 2). This difference persisted when cases of intravenous or intranasal oxytocin augmentation during labour were excluded (total mean blood loss 373 mL versus 486 mL,P= 0.001; postpartum haemorrhage > 800 mL 7.7%versus 13.0%; OR 0.55, 95% CI 0.30–0.99). The same was accounted for when instrumental deliveries and parturients mainly bleeding from episiotomy/tear were excluded (mean blood loss 399 mL versus 499 mL, P= 0.001; postpartum haemorrhage > 800 mL8.6%versus 13.6%; OR 0.60, 95% CI 0.38–0.495). Time between delivery of baby and placenta did not differ between the two groups (15 min versus 14 min; P= 0.30). Further data on blood loss in the third stage of labour are presented in Table 2.

Table 1.  Maternal and fetal data presented for the two groups. Values are shown as mean [±1 SD], n / n or n (%).
 NaCl n= 487)Oxytocin (n= 513)t test or OR (95% CI)
Maternal age (years)30 [5]29 [5]P = 0-29
Gestational age (days)279 [16]279 [10]P = 0-42
Primipara / multipara157/330173/3401-07 (0.82–1.39)
Oxytocin augmentation 1st stage147 (30-2)160 (31-2)1-05 (0.80–1.37)
Oxytocin intranasal 2nd stage90 (18-5)119 (23-2)1-33 (0.98–1.81)
Instrumental delivery42 (8-6)33 (6-4)0-73 (0.54–1.17)
Bleeding mainly from episiotomy/tear81 (16-6)97 (18-9)1-17 (0.84–1.62)
Fetal weight (g)]3647 [493]3613 [517]P = 0-28
Fetal macrosomia (> 4500 g)19 (3-9)20 (3-9)1-00 (0.51–1.93)
Table 2.  Comparison of the two treatment groups (oxytocin/NaCl) with regard to the number (n) of cases of postpartm haemorrhage (PPH) > 800 mL (also given in relation to parity and fetal size), PPH > 500 mL and PPH > 1000 mL. Total PPH, and PPH divided into bleeding before and after placenta! delivery (PD), is also given (mean [±1 SD])
 NaCl(n= 487)Oxytocin(n= 513)f test or OR (95% CI)
Total bleeding > 800 mL74450-63 (0.36–0.79)
Para   
I29/15719/1730-55 (0.29–1.01)
II-III41/28924/2990-52 (0.31–0.89)
>IV4/412/410-47 (0.08–1.31)
Fetal weight   
>4000g26/11215/1080-53 (0.27–1.07)
>4500g8/192/200-15 (0.03–0.76)
Total bleeding   
> 500 mL1751040-45 (0.34–0.60)
>1000mL43320-69 (0.43–1.10)
Total bleeding (mL)527 [412]409 [345]P < 0-001
Bleeding   
Before PD (mL)359 [315]270 [260]P < 0-001
After PD (mL)162 [192]138 [212]p = 0-061

In analysing the importance of a history of previous postpartum haemorrhage with regard to the occurrence of postpartum haemorrhage in a subsequent pregnancy, we used different cutoff levels: > 600 mL (the value recommended in the International Classification of Diseases (ICD 9), and generally adopted in Sweden) and > 1000 mL. Of the 670 multiparae in the present series, information regarding postpartum haemorrhage in a previous pregnancy was available in 44% of the cases (n= 296). Twelve percent (19/158) of the oxytocin group and 6% (8/138) of the saline group had had postpartum haemorrhage > 600 mL in a previous pregnancy; of these, 26% (5/19) and 38% (3/8), respectively, had postpartum haemorrhage > 800 mL in the present pregnancy. Eight (5%) of the oxytocin group and four (3%) of the saline group had had postpartum haemorrhage > 1000 mL in a previous pregnancy; of these, 25% (2/8 and 1/4, respectively) had postpartum haemorrhage > 800 mL in the present pregnancy.

Group requirements in terms of additional treatment with oxytocic drugs on clinical indications are presented in Table 3. Manual removal of the placenta was more common in the oxytocin group, although the difference did not reach statistical significance. The mean maternal haemoglobin level on day two after delivery was slightly higher in the oxytocin group (11.7 g/dL versus 11.4 g/dL, P < 0.001), though the mean duration of postpartum hospitalisation did not differ between the two groups (4.6 versus 4.5 days, P= 0.24). The small number of parturients given puerperal blood transfusion were also evenly distributed between the groups (seven and five cases, respectively).

Table 3.  Numbers of cases requiring additional oxytocic treatment on clinical indications or manual removal of the placenta in the two groups. Values are shown as n (%). IV = intravenous
 NaCl(n= 487)Oxytocin (n= 513)OR(95C1)
Methylergometrine IV67 (13-8)40 (7-8)0-53 (0.35–0.80)
Oxytocin 30 IU/500 mL glucose IV infusion18 (3-7)16(3-1)0-84 (0.42–1.66)
Prostaglandin F2a (IV infusion)3 (0-6)1 (0-2)0-32 (0.04–2.70)
Manual removal of placenta11 (2-3)18 (3-5)1-57 (0.74–3.33)

Table 4 shows the distribution of low postpartum Hb values in the two groups. Subgrouping according to maternal puerperal haemoglobin levels (< 10 g/dL versus > 10 g/dL) showed anaemia to be associated with a significantly longer duration of maternal hospitalisation (5.3 versus 4.5 days, P < 0.0001).

Table 4.  Distribution of low haemoglobin values on the second post-partum day in the two groups. Values are given asn.
Hb (g/dL)NaCl (n= 458)Oxytocin (n= 485)OR(95%CI)
<10.070470-60 (0.40–0.88)
<9-030200-61 (0.34–1.09)
<8-01270-55 (0.21–1.38)
<7-0430-71 (0.16–3.12)

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

In the present study we found a 43% reduction in the diagnosis of postpartum haemorrhage (> 500 mL) in the oxytocin group, a figure consistent with that of 40% elicited by meta-analysis3. Applying traditional criterion for postpartum haemorrhage (> 500 mL) to the present series, 20.3% of the oxytocin group and 35.9% of the saline group were classifiable as cases of postpartum haemorrhage, which are higher frequencies than those reported by others. The discrepancy is difficult to explain, except by the notorious inaccuracy of blood loss estimation. As the present study was blind, the difference observed between the two groups is likely to be true. However, we believe this cutoff value for blood loss to be too low to be of clinical significance, and wanted to test an arbitrary cutoff level for postpartum haemorrhage of > 800 mL. Using this cutoff level, the frequency of postpartum haemorrhage was 8.8% in the oxytocin group but almost double in the saline group (15.2%). To the best of our knowledge, in only one previous study of appropriate size have oxytocin and saline solution been compared, though this trial was actually carried out in the fourth stage of labour4. In the quoted study 3 IU of oxytocin was given intravenously, in comparison with 10 IU in the present series. Their findings regarding the proportion of reduction in postpartum haemorrhage > 500 mL (3.2%versus 5.3%) were similar to those of the present study, although no analysis was made of cases of more severe postpartum haemorrhage (i.e. with a higher cutoff level).

The present study was primarily conducted to test the hypothesis that exclusion of routine oxytocin administration in the third stage of labour would not significantly increase the frequency of postpartum haemorrhage > 800 mL. The randomised groups (oxytocin versus saline treatment) were comparable regarding other factors that might affect postpartum haemorrhage. A 22% reduction in mean total blood loss was demonstrated in the oxytocin group, a finding in accord with those of previous studies regarding the use of oxytocic drugs3,4, but in contrast to those in a study by Thilaganathan et al.9. However, as the latter study was not blind, blood loss estimation may have been biased. Moreover, patients at high risk of postpartum haemorrhage were excluded (e.g. multiparas, and patients with a previous history of postpartum haemorrhage or caesarean section). The small size of their series, 193 patients, may also have impeded the detection of group differences.

Regarding risk groups for postpartum haemorrhage, we found the frequency of postpartum haemorrhage to be higher in the saline group than in the oxytocin group, irrespective of parity, although the difference reached significance only in the para II and para III subgroups. The presence of fetal macrosomia (fetal weight > 4500 g), was also a factor of significance for post partum haemorrhage, where oxytocin treatment reduced the frequency from 42% to 10%.

The frequency of a history of postpartum haemorrhage (> 600 mL, the criterion usually adopted in Sweden) seems to be underestimated in clinical practice (e.g. the figure of 9% obtained in the present study, as compared with that of 18% obtained in our pilot study in 1992 (L. Nordstrom, unpublished data). Not only was the frequency of repeat postpartum haemorrhage high in the present series as a whole, but it was also high in the oxytocin group, despite the prophylactic treatment. Owing to the small proportion of gravidae so identified, a history of previous postpartum haemorrhage would appear to be a poor criterion for selecting a risk group who might benefit from prophylactic treatment.

Additional treatment with methylergometrine was given in the event of excessive bleeding in 7.8% (1/13) of cases in the oxytocin group and in 13.8% (1/7) of cases in the saline group. Corresponding figures in the study by Howard et al.4 were 4.4% and 12.2%, respectively. We also found a postpartum haemoglobin level < 10 g/dL to be more frequent in the saline group, and to be a factor correlated to longer duration of postpartum hospitalisation. The tendency in Sweden during recent years to reduce the duration of puerperal care has not penetrated the county of Jämtland where the present study was carried out. Owing to the large size of the county, and thus of the county hospital catchment area (with a maximum radius of about 300 km), postpartum care has traditionally been of rather long duration (mean 4-5 days). Accordingly, the need for prolonged postpartum maternal care due to excessive postpartum haemorrhage may have been obscured, if recovery occurred during the usually long period of hospitalisation traditional at our hospital.

The overall number of blood transfusions was too small for meaningful group differences to be elicited.

The frequency of manual placenta removal was higher in the oxytocin group than in the saline group, though the difference did not reach statistical significance. There has been discussion as to the choice of oxytocic drug for use in the third stage of labour7,8,12, and oxytocin with ergometrine has previously been advocated7,12. Yuen et al. found this drug to be associated with a 3.7-fold increase in the incidence of retained placenta, as compared with oxytocin7. Had oxytocin with ergomertine been used instead of oxytocin alone in the present series, a sixfold increase in the incidence of this complication would have been expected.

The incidence of heavy postpartum haemorrhage (> 1000 mL), a clinically more serious condition, did not differ between the two groups studied by Yuen et al.7 or by McDonald et al.8

The choice of oxytocin for use in the present series was based on its documented immediate effect, with® intermittent uterine contractions, and especially the lower frequency of retained placenta (as compared with Syntometrine®). As we consider manual removal of the placenta to be a more severe complication (usually involving gastric emptying and intravenous anaesthesia) than moderate postpartum haemorrhage, we favour the use of oxytocin. Not only is oxytocin a cheap drug, but it has neither any important side effects nor any significant effect on blood pressure7.

Oxytocin can be given either intramuscularly or intravenously, the former route delaying its pharmacological effect by a few minutes13, which might be of some clinical importance as the main purpose is to reduce blood loss before placental delivery. The midwifery staff at our hospital also find the intravenous route easier, and less painful to the patient, than the intramuscular route. However, the parenteral route of administration could be difficult to carry out in many developing countries, where it might be needed the most. It would therefore be useful to assess other active procedures like manual fundal pressure, controlled cord traction and draining of blood from the placenta in a similar trial.

In all, we found intravenous oxytoxin (Partocon®10 IU) routinely given after delivery to be associated with substantial effects in terms of a 22% reduction in mean blood loss volume, a 43% reduction in the frequency of postpartum haemorrhage > 500 mL, a 42% reduction in postpartum haemorrhage > 800 mL, a 43% reduction in additional use of methylergometrine, and a 36% reduction in the frequency of postpartum haemoglobin values of < 10 g/dL.

When routine treatment with oxytocin is given, one case of postpartum haemorrhage of more than 800 mL is avoided for every 16 injections. Were treatment to have been restricted to high risk patients, defined as multiparas and/or parturients delivering fetuses of more than 4500 g, 68% of the parturients in this study would have been treated. Thus, 28 patients would have to be treated to avoid one case of postpartum haemorrhage of more than 800 mL. Can we afford not to give routine oxytocin treatment?

Acknowledgements

We would like to express our sincere gratitude to Ing-Mari Hjalmars at the Pharmacy, County Hospital, Östersund, for excellent preparation of random numbered ampullas. This study was supported by grants from the County Council and County Health Authority Research and Development Foundation in the County of Jamtland, Sweden.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References