SEARCH

SEARCH BY CITATION

Abstract

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

Objective To test whether a policy of active management of nulliparous labour would reduce the rate of caesarean section and prolonged labour without influencing maternal satisfaction.

Design Randomised controlled trial.

Setting Tertiary referral obstetric unit, Auckland, New Zealand.

Population Nulliparous women in spontaneous labour at term with singleton pregnancy and cephalic presentation and without fetal distress.

Methods After the onset of active labour, previously consented women were randomly assigned to active management (n= 320) or to routine care (n= 331). Active labour was defined as regular painful contractions, occurring at least once in five minutes, lasting at least 40 seconds, accompanied by either spontaneous rupture of the membranes, or full cervical effacement and dilatation of at least two centimetres. Active management included early amniotomy, two-hourly vaginal assessments, and early use of high dose oxytocin for slow progress in labour. Routine care was not prespecified. Prolonged labour was > 12 hours duration. Maternal satisfaction with labour care was assessed by postal questionnaire at six weeks postpartum.

Main outcome measures Mode of delivery, duration of labour, and maternal satisfaction.

Results Active management of labour did not reduce the rate of caesarean section 30/320 (9.4%), compared with 32/331 (9.7%) for routine care, but did shorten the length of first stage of labour (median 240 min vs 290 min; P= 0.02), and reduce the relative risk of prolonged labour (RR 0.39; 95% CI 0.22, 0.71). There were no differences between groups in the rates of newborn nursery admission, neonatal acidosis, low Apgar scores, or postpartum haemorrhage. Satisfaction with labour care was high (77%) and did not differ between groups.

Conclusions Active management of labour reduced the duration of the first stage of labour without affecting the rate of caesarean section, maternal satisfaction, or other maternal or newborn morbidity.


INTRODUCTION

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

Active management of labour was first described as a policy to reduce the rate of prolonged labour1. In studies using historical controls, implementation of active management of labour reduced the rate of nulliparous caesarean section due to dystocia2,3.

There are three large randomised controlled trials where a policy of active management of nulliparous labour has been compared with routine management, with the aim of reducing the rate of caesarean section4–6. In these studies active management consisted of an accurate diagnosis of labour, early amniotomy, frequent vaginal examinations, and high dose oxytocin augmentation for slow labour progress (cervical dilatation < 1 cm/hour). In one study actively managed women also had specific antenatal education and one-to-one midwifery care in labour5.

All three studies have reported a reduction in the duration of labour with a policy of active management and two found reduced maternal infectious morbidity4,5. One study found a reduction in caesarean section rate with active management of labour, but only after controlling for confounding factors4. The issue of whether a policy of active management of nulliparous labour can reduce caesarean section rates is therefore still unresolved and further studies are necessary.

All three previously published studies were conducted in the United States and no study to date has assessed maternal satisfaction with a policy of active management of labour. In the current study we tested the hypothesis that a policy of active management of nulliparous labour would: 1. reduce the rate of caesarean section; 2. reduce the rate of prolonged labour; 3. not influence maternal satisfaction with the birth experience.

METHODS

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

This randomised controlled trial was undertaken at National Women's Hospital, a large tertiary referral centre, in Auckland, New Zealand, from June 1993 to August 1997. The Health Funding Authority ethical committee approved the study. Nine thousand women gave birth at National Women's Hospital in 1996, with a caesarean section rate for nulliparous women of 22%. Private caregivers (e.g. obstetricians, family doctors, and independent midwives) provided antenatal care for 52% of women, whilst public obstetricians, midwives, and trainees provided care for 24%. Eleven percent of the women were booked for shared care, defined as antenatal care shared between the family doctor and the hospital clinic with delivery by the public obstetric team. The remaining women (13%) were transferred to the National Women's Hospital in pregnancy or labour.

During the third trimester of pregnancy, women attending hospital clinics and private obstetric providers were recruited to the study by a research midwife. Informed consent was obtained from each woman. Eligibility criteria were: nulliparity, singleton pregnancy, no severe cardiac disease, no uterine scar and no proven contracted pelvis. Neither antenatal care nor antenatal education nor labour caregiver was modified by the woman's agreement to participate in the study. At the onset of labour, some of the women who had given consent were excluded prior to randomisation for the following reasons: induced labour; noncephalic presentation; withdrawal of consent; gestation < 37 completed weeks; abnormal cardiotocograph or thick meconium suggesting fetal distress on admission to the labour ward; involvement in the study not recognised at admission in labour; elective caesarean section; intrauterine death; or multiparity.

Labour was defined as regular painful contractions, occurring at least once every five minutes, lasting at least 40 s, accompanied by either spontaneous rupture of the membranes or full effacement of the cervix and cervical dilatation of at least two centimetres. At the time of diagnosis of labour, randomisation to active or routine management was performed by opening a sealed envelope corresponding to the participant's preassigned study number. Randomisation was based on a computer-generated list of random numbers and was not stratified. A nonclinical person not involved in the study sealed the treatment assignments into opaque envelopes.

Women randomised to active management were encouraged to have an amniotomy at the diagnosis of labour. Cervical assessment was performed every two hours. Oxytocin augmentation was commenced if progress was delayed, defined as < 1 cm/hour in the first stage of labour and the absence of descent of the baby's head after 30 minutes of pushing or contractions less than one in five minutes and delivery not imminent in the second stage of labour. Oxytocin was commenced at 6 mU per minute and increased by 6 mU per minute every 15 minutes to a maximum dose of 36 mU per minute.

There was no written guideline describing routine management. The frequency of vaginal examinations, the use of amniotomy, and the commencement of oxytocin were determined by the caregiver. The labour ward protocol for oxytocin dosage was used in this arm of the study, consisting of an initial dose of 1 mU per minute, doubled every 10 minutes to 8 mU per minute, then increased by 2 mU per minute to a maximum dose of 40 mU per minute.

In both groups it was recommended that vaginal examinations were to be performed by the same observer; oxytocin dosage to be halved if hyperstimulation occurred; oxytocin to be stopped if there was prolonged bradycardia or if acidosis was found on scalp blood sampling (pH < 7.2); and oxytocin not to be stopped if fetal heart decelerations occurred, in which case the caregiver was contacted for advice about further management (e.g. scalp pH). One-to-one midwifery care in labour was the aim in both treatment arms. Continuous fetal monitoring was used for all women requiring oxytocin augmentation. It was requested that cord blood gas analysis be performed after delivery for study participants.

The main maternal outcome measures were rates of caesarean section, operative vaginal delivery, prolonged labour (> 12 hours), febrile complications, postpartum haemorrhage (> 500 mL), and satisfaction with labour care. Newborn complications were also compared between groups including rates of newborn nursery admission, depression at birth (Apgar score < 6 at 5 minutes), and acidosis.

Maternal satisfaction was assessed six weeks postpartum by a postal questionnaire. If necessary, the questionnaires were sent on two separate occasions and a letter sent on a third occasion in an attempt to maximise response rates. Women were asked two questions to assess overall satisfaction with labour and delivery care. The first was: ‘Thinking back now, how satisfied are you, overall, with the care you received during your labour and delivery?’ The responses were: ‘Very satisfied’, ‘Satisfied in some ways but not in others’, and ‘Very dissatisfied’. The response ‘Very satisfied’ was considered to indicate satisfaction with labour care7. The second question assessing satisfaction was as follows: ‘Would you choose the same type of management again?’ The following responses were possible: ‘Yes’, ‘No’, and ‘I'm not sure/don't know’.

Data were collected by the midwife who had admitted the participant at the time of randomisation and by the research midwife after delivery using the hospital notes and the hospital obstetric database. All data from the hospital database were checked manually against the hospital clinical record for accuracy. Compliance with the active management protocol was assessed in each case by the research midwife who reviewed all labour and delivery notes. She determined whether: 1. the definition of labour was met at randomisation; 2. vaginal assessments were performed as frequently as dictated by the active management protocol (less than three hourly); 3. oxytocin was commenced within one hour when indicated as dictated by the active protocol; and 4. the dosage of oxytocin followed the active management protocol.

Duration of labour was defined as the time from randomisation (diagnosis of labour) until delivery. Prolonged labour was defined as > 12 hours. Maternal infectious morbidity included any infection during the postnatal stay. Postpartum haemorrhage was defined as blood loss > 500 mL, estimated by delivery staff. Hyper-stimulation was defined as a single contraction lasting more than two minutes or more than six contractions in 10 minutes. Acidosis at birth was empirically defined as cord arterial pH <7.15 and base excess >−8.

Statistical methods

The caesarean section rate for low risk nulliparous women in our hospital was estimated at 18%. Three hundred and twenty women were needed in each arm of the study to show a reduction in caesarean section rate from 18% to 10%, which was the rate in nulliparous women in a neighbouring hospital with a policy of active management (power = 80%, α= 0.05).

Statistical analysis was performed using SAS Version 6.12 (Cary, North Carolina, USA). Normally distributed continuous data were compared using Student's t test. Nonparametric continuous data were compared using the Wilcoxon rank sum test. Frequency data were compared using the χ2 or Fisher's exact test, and are presented where possible as relative risks with 95% confidence interval for active management, compared with routine care. A P value < 0.05 was required for statistical significance. Logistic regression was used to adjust for any imbalance in baseline characteristics in determining the adjusted odds ratio for caesarean section for active compared with routine management.

A meta-analysis was performed using Revman 3.1 Software (Update Software, Oxford, UK) to produce a summary relative risk for caesarean section for active compared with routine management. The studies included were those, identified by Medline in the English literature, where active management was compared with routine management. Active management was defined as including at least all of the following: a strict definition of labour, two hourly vaginal assessments, early amniotomy, and early augmentation with high dose oxytocin. The summary statistic presented is a relative risk, using a fixed effects model and the Mantel-Haenszel method. A test for homogeneity was performed.

RESULTS

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

Twelve hundred and twenty-seven women were recruited to the study over a four year period. Of these, 22 were delivered at other hospitals and are excluded from further analysis. Five hundred and fifty-four (46%) became ineligible before randomisation at the onset of labour for pre-defined reasons (Fig. 1). The most common reason for ineligibility was induction of labour, affecting 370 women (67%). Randomised women were significantly younger and less likely to be European than ineligible women, but did not differ by smoking, education, or marital status (data not presented).

image

Figure 1. Flow chart of trial recruitment.

Download figure to PowerPoint

Three hundred and twenty women were randomised to active management and 331 to routine management. There were no differences in baseline data according to treatment assignment (Table 1).

Table 1.  Demographic characteristics of randomised women by treatment group. Values are given as n (%) or mean [SD].
 Active management (n= 320)Routine management (n= 331)
Age (years)25.7 [5.1]25.7 [5.0]
  Ethnicity  
  European121 (38)110 (33)
  Maori41 (13)51 (16)
  Pacific Island92 (29)100 (30)
  Asian/Other66 (20)70 (21)
Smoker (n= 615)60 (19)67 (22)
Primary care giver  
  Public135 (42)137 (41)
  Private43 (14)50(15)
  Shared care142 (44)144 (44)
Height (cm) (n= 433)164 [7]164 [7]
Booking weight (kg) (n= 562)71 [16]71 [17]

Dilatation at randomisation (4.5 cm) was the same in both groups (Table 2). More women in the actively managed group had an amniotomy (72%vs 63%), and it was performed at a lower cervical dilatation (5.1 cm vs 5.7 cm). Oxytocin augmentation was used more commonly (53%vs 39%) and at higher dosages in the actively managed group. Oxytocin augmentation was begun in second stage in 10.9% (35/320) of the actively managed women and 6.6% (22/331) of the routinely managed women. Epidural rates did not differ between groups. Fifty percent of women received one-to-one midwifery care, and the rates did not differ between groups.

Table 2.  Labour data by treatment group. Values are given as n (%), mean [SD] or median {interquartile range}, unless otherwise indicated.
 Management style  
 Active (n= 320)Routine (n= 331)PRR (95% CI)
  1. *Dose maintained for longest interval.

Dilatation at randomisation (cm)4.5 [1.8]4.5 [2.1]0.8 
Amniotomy231 (72)209 (63) 1.14 (1.03, 1.27)
  Dilatation at amniotomy (cm)5.1 [2.1]5.7 [2.3]0.006 
Oxytocin augmentation168 (53)129 (39) 1.35 (1.14, 1.60)
  Dilatation at augmentation (cm)5.9 [2.7]6.1 [2.7]0.5 
  Main syntocinon dose (mU/min)*12.6 [8.2]9.3 [6.1]0.0001 
  Maximum syntocinon dose (mU/min)16.2 [8.9]12.0 [7.2]0.0001 
Epidural156 (49) 149 (45)1.08(0.92, 1.28)
Number of vaginal examinations3.8 [1.7]3.5 [1.6]0.02 
Received 1:1 midwifery care (n= 595)156 (52)142 (48) 1.09 (0.93, 1.28)
     
Vaginal deliveries only(n= 290)(n= 299)  
Total duration of labour (min)326 (185–485)376 (212–543)0.05 
  First stage240 (120–390)290 (145–460)0.02 
Second stage59 (33–106)60 (35–100)0.9 
Labour > 12 hours duration14 (5)37 (12) 0.39 (0.22, 0.71)

The median length of labour for women who delivered vaginally was reduced by 50 minutes in the active management group. This was due to a 50 minute reduction in the duration of the first stage of labour with no effect on second stage duration (Table 2). Five percent of actively managed women had prolonged labour, compared with 12% of routinely managed women.

Mode of delivery did not differ by study group (Table 3). Thirty women (9%) in the actively managed group were delivered by caesarean section, compared with 32 (10%) in the routine group. Neither total rates nor indications for caesarean section differed between groups. Logistic regression, controlling for maternal age, ethnicity (European versus other), gestational age, birthweight, epidural use, and dilatation at randomisation, did not alter the odds of caesarean section. Unadjusted odds ratio (95% CI) for caesarean section for the actively managed group was 0.97 (0.57, 1.63) and adjusted odds ratio 0.97 (0.54, 1.73).

Table 3.  Mode of delivery and indication for caesarean section by treatment group. Values are given as n (%), unless otherwise indicated.
Mode of deliveryActive management (n= 320)Routine management (n= 331)P
Vaginal delivery   
  Spontaneous227 (71)245 (74) 
  Operative63 (20)54 (16) 
Caesarean section30 (9)32 (10)0.5
First stage17 (5)22 (7) 
Second stage13 (4)10 (3) 
    
Indication for caesarean section(n= 30)(n= 32)P
Failure to progress26 (87)25 (78) 
  First stage15 (50)18 (56) 
  Second stage11 (37)7 (22) 
Fetal distress4 (13)4 (13) 
Other03 (9)0.3

There were no differences in newborn outcomes by treatment group (Table 4). Cord blood gas results were obtained in 61 % (n= 196) of actively managed and 48% (n= 158) of routinely managed women. Six routinely managed women (1.8%) and seven actively managed women (2.2%) had babies who were acidotic at birth. There were no differences in maternal infectious morbidity or postpartum haemorrhage by treatment group (Table 4).

Table 4.  Maternal and newborn outcomes by treatment group. Values are given as n (%) or mean [SD], unless otherwise indicated.
 Management style  
 Active (n= 320)Routine (n= 331)PRR (95% CI)
Birthweight3444 [432]3433 [436]0.7 
Gestational age at delivery39.7 [1.1]39.6 [1.1]0.6 
Apgar at 5 minutes < 61 (0.3)00.5 
Baby admitted to newborn nursery17 (5)16 (5) 1.10 (0.57, 2.14)
Fully breastfed at 6 weeks postpartum (n= 472)123 (50)117 (52) 0.96 (0.80, 1.14)
Maternal infectious morbidity37 (12)34 (10) 1.12 (0.72, 1.74)
Postpartum haemorrhage48 (15)47 (14) 1.04 (0.72, 1.51)
Overall very satisfied with labour and delivery care (n= 468)190 (78)169 (75) 1.04 (0.94, 1.15)
Would choose the same management again (n= 466)179 (73)155 (70) 1.05 (0.94, 1.18)

The overall response rate to the maternal satisfaction questionnaire was 72%, although significantly more actively managed women [243/320 (76%)] responded than routinely managed women [225/331 (68%)], P= 0.02. Satisfaction with labour care was high and did not differ between groups.

One or more aspects of the active management protocol were not adhered to in 127 actively managed women (40%) (Table 5). The majority of these related to failure to initiate or follow the high dose oxytocin augmentation protocol. Only one caesarean section was performed in the 24 women for whom oxytocin augmentation was considered to have been commenced late or not at all, when indicated, in the second stage of labour.

Table 5.  Compliance by treatment group. Values are given as n (%). N/A = not applicable.
 Active management (n= 320)Routine management (n= 331)
  1. *Time lapsed between examinations < 3 hours.

  2. Less than one hour after indicated.

Compliance193 (60)N/A
Correct definition of labour308 (96)328 (99)
Correct frequency of vaginal examinations*288 (90)N/A
Oxytocin started when indicated  
  First stage275 (86)N/A
  Second stage296 (92)N/A
Dose of oxytocin followed protocol245 (77)N/A

Two meta-analyses are presented. The first includes the intention-to-treat data from all trials4–6 (Fig. 2) showing a summary relative risk (95% CI) of caesarean section for active management of 0.93 (0.80–1.08). The second includes intention-to-treat data from the trials of Lopez-Zeno et al.4, Rogers et al.6 and the current paper (i.e. randomisation at the onset of labour), along with data from women protocol-eligible at the onset of labour from the trial of Frigoletto et al.5 (i.e. randomisation before 30 weeks) (Fig. 3). The summary relative risk (95% CI) of caesarean section for active management in this second scenario is 0.86 (0.70–1.05). Neither meta-analysis shows a significant reduction in caesarean section rate with active management of labour. There was no evidence for heterogeneity between the studies.

image

Figure 2. Meta-analysis of trials of active management of labour (including intention to treat data from Frigoletto et al5).

Download figure to PowerPoint

image

Figure 3. Meta-analysis of trials of active management of labour (including data from women protocol-eligible at onset of labour from Frigoletto et al.5).

Download figure to PowerPoint

DISCUSSION

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

In this randomised controlled trial of active management of nulliparous labour conducted in New Zealand, incorporating a strict definition of labour, two-hourly vaginal assessments, early amniotomy, and early augmentation with high dose oxytocin, active management of labour did not reduce the rate of caesarean section. There was a 50 minute reduction in the length of labour and a reduction in the prevalence of prolonged labour but this was not associated with reduced maternal infectious morbidity.

Only one of the previous large randomised trials reported a significant reduction in caesarean section rate with active management of labour4. Meta-analyses of these three earlier trials combined with data from the current study demonstrate that active management is not associated with a significant reduction in caesarean section. Measures of perinatal morbidity were not increased in any trial by active management of labour. The reduced duration of labour found in all trials, ranging from 50 minutes in the current study to 102 minutes in the study by Rogers et al.6, may be important to some women who as a result may wish to make an informed choice to have their labour actively managed. Furthermore, data from the current study demonstrate that a policy of active management of labour did not reduce satisfaction with care.

Before this study was started the caesarean section rate for nulliparous women in our hospital was 21 %. We estimated the background caesarean section rate in the lower risk study population at 18%, but at that time did not have data to confirm this estimate. An audit of groups of nulliparous women who would have been eligible for the study, but who were not recruited, showed a caesarean section rate during the study period of 12%. This is comparable to the caesarean section rate of 10% in our control group and, along with the continued rise in the overall nulliparous caesarean section rate in our unit (28% by 1998), perhaps argues against existence of the Hawthorne effect, where management in the control group comes to resemble that in the study group. Further evidence for lack of a Hawthorne effect is the significant difference between study groups in some of the measured aspects of active management (i.e. more frequent and earlier use of amniotomy and more frequent and higher dose augmentation). The lower than expected caesarean section rate in our control group reduces the power of our study. Given the caesarean section rate of 10% in the control group, a retrospective power calculation shows that we had sufficient power to detect a reduction in the caesarean section rate to 4% (power = 80%, α= 0.05).

The rate of caesarean section in the second stage of labour in our study [3.5% (23/651)] is similar to that found in the three North American studies4–6, compared with 0.2% of nulliparous women at the National Maternity Hospital in Dublin, Ireland8. The rate of first stage caesarean section in control and intervention groups in all four trials is similar to that reported from the National Maternity Hospital. It is the rate of caesarean section in the second stage that is so vastly different, yet none of these studies have found any effect of active management of labour on caesarean section in the second stage of labour. This suggests that either some aspect of the Dublin active management protocol has not been implemented in any of these trials or that there is some other, as yet unknown, factor that is responsible for the very low rate of second stage caesarean in Dublin. Impey and Boylan8, in their commentary on active management of labour, attribute this to their use of oxytocin for the first time in the second stage of labour in 14.7% of nulliparous women. In the current trial the rate of commencement of oxytocin in second stage was 10.9% in the actively managed group and 6.6% in the routinely managed group. The review of compliance undertaken in this study suggests that failure to use oxytocin when or as indicated in second stage may have resulted in one caesarean section in an actively managed woman. Second stage management therefore needs further consideration in our hospital and in other centres.

This is the first study of active management of labour to assess women's satisfaction with the care they received in labour. Most women (77%) were very satisfied and rates of satisfaction were not different between study groups. These results are consistent with those from a meta-analysis9 of three trials of early amniotomy where early amniotomy was not associated with increased dissatisfaction with care. The current trial had sufficient power to detect a reduction in satisfaction from 75% in the control group to 64% in the active management group (power = 80%, α= 0.05).

In this trial the active management protocol was not strictly followed in 40% of women randomised to active management. In one-quarter of women the regimen of high dose oxytocin augmentation was not followed, and lesser doses or less frequent increments in dosage occurred. This may have been because our labour ward midwives, who are mainly trained in New Zealand or England, were not comfortable with high dose augmentation. Information received by the research midwife concerning midwives' compliance indicated that a number of midwives considered the active management protocol to be medical interference and thus contrary to the philosophy of midwifery. As the other studies4–6 have not reported compliance with the active management protocol, it is impossible to know if the failure to find any effect of active management on caesarean section in the current study was related to compliance.

In conclusion, our randomised controlled trial of active management of nulliparous labour showed this protocol to be associated with shorter duration of labour, fewer prolonged labours, and high maternal satisfaction. Meta-analyses of this and three other similar trials, demonstrate no significant reduction in caesarean section rates with active management.

Acknowledgements

The authors would like to thank the other research midwives, Ms V. Masson, Ms A. Matthews, and Ms D. Lawton, who were involved with this project. Financial support was provided by Auckland Health Care, the Health Research Council of New Zealand, and the Evelyn Bond Obstetric Research Fund.

References

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