Outcome of term breech births: 10-year experience at a district general hospital

Authors


Dr P. Pradhan, Heartlands and Solihull NHS Trust, Princess of Wales Maternity Unit, Bordesley Green east, Birmingham B9 5SS, UK.

Abstract

Objective  To review the short and long term outcomes among singleton infants with breech presentation at term delivered in a geographically defined population over a 10-year period.

Design  Retrospective, cohort study.

Setting  District General Hospital.

Population  1433 term breech infants alive at the onset of labour and born between January 1991 and December 2000.

Methods  Data abstracted from birth registers, neonatal discharge summaries and the child health database system were used to compare the short and long term outcomes of singleton term breech infants born by two different modes of delivery (prelabour caesarean section and vaginal or caesarean section in labour). Fisher's exact test was used to compare the categorical variables.

Main outcome measures  Short term outcomes: perinatal mortality, Apgar scores, admission to the neonatal unit, birth trauma and neonatal convulsions. Long term outcomes: deaths during infancy, cerebral palsy, long term morbidity (development of special needs and special educational needs).

Results  Of 1433 singleton term infants in breech presentation at onset of labour, 881 (61.5%) were delivered vaginally or by caesarean section in labour and 552 (38.5%) were born by prelabour caesarean section. There were three (0.3%) non-malformed perinatal deaths among infants born by vaginal delivery or caesarean section in labour compared with none in the prelabour caesarean section cohort. Compared with infants born by prelabour caesarean section, those delivered vaginally or by caesarean section in labour were significantly more likely to have low 5-minute Apgar scores (0.9%vs 5.9%, P < 0.0001) and require admission to the neonatal unit (1.6%vs 4%, P= 0.0119). However, there was no significant difference in the long term morbidity between the two groups (5.3% in the vaginal/caesarean section in labour group vs 3.8% in the prelabour caesarean group, P= 0.26); no difference in rates of cerebral palsy; and none of the eight infant deaths were related to the mode of delivery.

Conclusions  Vaginal breech delivery or caesarean section in labour was associated with a small but unequivocal increase in the short term mortality and morbidity. However, the long term outcome was not influenced by the mode of delivery.

INTRODUCTION

In the ‘Term breech trial’, planned caesarean section for singleton breech presentation at term resulted in significant reduction in the perinatal and neonatal deaths as well as serious short term neonatal morbidity compared with the planned vaginal births.1 The results of this trial were crucial for the recent RCOG recommendations on the management of singleton term breech birth.2 The trial has, however, been criticised for a number of methodological reasons.3,4 Subsequent retrospective cohort studies from various countries have provided conflicting results about the effects of mode of delivery on immediate perinatal outcomes.5–7 A recent Dutch study showed a sevenfold increase in low Apgar scores, a threefold increase in birth trauma and a twofold increase in perinatal mortality with vaginal delivery and emergency caesarean compared with planned caesarean in term singleton breech presentations.5 This contrasts with the results from the Irish study where safe vaginal delivery of term breech fetuses was achieved by adherence to strict selection criteria and intrapartum protocol.6 Although there has been an increasing amount of data available on the early neonatal mortality and morbidity, there is paucity of information on the long term outcome. The two-year outcomes of the term breech trial are yet to be reported.

Information on both short and long term outcomes is needed to help inform individualised decision making regarding management of breech presentation at term. We wished to review the short and long term outcomes among singleton infants with breech presentation at term delivered in a geographically defined population over a 10-year period.

METHODS

The Royal Shrewsbury Hospital is a large district general hospital serving a population of approximately 500,000. There are around 5000 births annually in its catchment area. The obstetric service is provided through a Consultant Unit at the Royal Shrewsbury hospital and six midwifery-led units in the county. The consultant unit is compliant with the RCOG standard of 40 hours of consultant cover on the labour ward providing a high standard of care to the labouring women. Shropshire has the lowest caesarean section rate in England and Wales.8 At our Hospital, a median (range) of 32% (22–42) of infants in breech presentation at term have been delivered vaginally over the last 10 years. Between 1991 and 1995, suitability of a pregnant woman with term breech presentation for vaginal delivery was assessed by the woman's consultant after X-ray pelvimetry and/or ultrasound scan examination for the type of breech presentation and estimated fetal weight measurement. The policy then included use of prostaglandins and syntocinon for augmentation.

A dedicated external cephalic version clinic was set up in 1995. Pregnant women with breech presentation at term are referred to this clinic, where they are assessed by a specialist. The assessment includes a detailed ultrasound scan to determine the type of breech, the suitability for external cephalic version and also the suitability of vaginal delivery in case of failed or unattempted external cephalic version. The ultrasound scan encompasses various measurements including the abdominal and head circumferences. In cases of failed or unattempted external cephalic version, the decision regarding the mode of delivery is made jointly by the mother and the specialist after careful counselling. If vaginal delivery is contemplated, then spontaneous onset of labour is awaited. On admission to the delivery suite, the on-call consultant is informed of the mother's condition and progress. The decision to use syntocinon for augmentation is made at a consultant level and, if possible, avoided. An experienced team consisting of the on-call consultant and/or senior registrar along with the obstetric and anaesthetic registrars is present on the delivery suite when a vaginal breech delivery is anticipated.

We identified all instances of breech presentations at or after 37 completed weeks of gestation delivered in the consultant unit at the Royal Shrewsbury hospital between January 1991 and December 2000 from the Labour ward registers. The exclusion criteria were multiple pregnancy and antenatal fetal death. Short term outcomes of interest were the perinatal mortality, Apgar scores, admission to the neonatal unit, birth trauma and neonatal convulsions. The long term outcomes evaluated included deaths during infancy and childhood, cerebral palsy and development of special needs and special educational needs. A child with special needs is defined as one who has a higher than average risk of needing paediatric input. Special educational needs include children with learning difficulties who need special educational provision.

The details of the labour and delivery were obtained from the birth registers, and the necessary neonatal outcomes were abstracted from the neonatal discharge summaries. The long term outcome was identified from an electronic Child Health System (CHS) database of the county. The CHS database maintains records of all children residing in Shropshire from 1965. The records are transferred with the child in case of change of residence. The National Informatics Council examined the database in 1998 and found it to have an accuracy rate of 95% in the new NHS tracing scheme.9 To ensure a minimum follow up of at least three years, only infants born between January 1991 and December 1998 were included in the analysis of long term outcomes. The maximum follow up period was 10 years for children born in 1991. Obstetric case notes of severely handicapped children were examined where possible.

We used Fisher's exact test for comparison of categorical variables, and calculated the odds ratios (OR) with 95% confidence intervals for comparison of the short and long term morbidity between the infants born by prelabour caesarean section and those delivered vaginally or by caesarean section in labour.

RESULTS

Between January 1991 and December 2000, there were 1433 singleton term breech infants alive at the onset of labour. Five hundred and fifty-two (38.5%) infants were born by prelabour caesarean section. Of the remaining 881 infants, 416 (29.1%) were delivered vaginally and 465 (32.4%) were born by caesarean section in labour.

Data on short term outcomes were available for all 1433 infants. There were four (0.2%) intrapartum deaths (Table 1). One death was due to a lethal congenital malformation. In the second case, abnormal fetal heart rate pattern was noted in labour but no attempts were made at fetal blood sampling. The case showed evidence of suboptimal care in labour, however, no difficulty was experienced at the vaginal delivery. Massive abruption with resultant fetal death occurred in the third case. Labour was allowed to progress and the baby was delivered vaginally with no birth difficulties. Birth trauma was directly attributed as the cause of death in the fourth case where difficulty was experienced at the delivery of fetal arms due to nuchal extension.

Table 1.  Perinatal mortality in term breech infants delivered over a 10-year period (1991–2000).
Year of birthMode of deliveryWeeks of gestationCause of death
1994Caesarean section in labour38Lethal congenital malformations
1996Vaginal delivery39Suboptimal care in labour and delivery
1998Vaginal delivery41Massive abruption during labour
2000Vaginal delivery40Birth trauma

Compared with the infants born by prelabour caesarean section, those delivered vaginally or by caesarean section in labour were more likely to have low 5-minute Apgar score and were also more likely to need admission to the neonatal unit (Table 2). Two infants in the entire cohort had neonatal convulsions and both were delivered vaginally. In one, there was evidence of asphyxia at birth (abnormal fetal heart rate pattern during labour and cord blood acidosis) with neonatal encephalopathy in the neonatal period. The labour records did not suggest any difficulty with the delivery. The second case was complicated with prolonged prelabour rupture of membranes and the delivery was uncomplicated. The seizures were thought to be related to neonatal infection. There was no evidence of birth trauma in any of the surviving infants.

Table 2.  Short and long term morbidity in relation to the mode of delivery among term breech infants.
MorbidityVaginal breech/caesarean section in labour (n= 881)Prelabour caesarean section (n= 552)OR (95% CI)P
Short term
5-minute Apgar score <752 (5.9)5 (0.9)6.5 (2.6–16.2)<0.0001
Admission to the Neonatal Unit36 (4.0)9 (1.6)2.5 (1.2–5.2)0.0119
Neonatal convulsions2 (0.2)00.526
Birth trauma00
 
Long term(n= 669)(n= 462)  
Special needs and special educational needs36 (5.3)18 (3.8)1.4 (0.8–2.4)0.26
Cerebral palsy1 (0.1)01.00

Eight infants from the entire cohort died during infancy and childhood. Four of these deaths were secondary to congenital malformations (trisomies 18 and 21, spina bifida and vein of Galen malformation), two were due to accidental causes (house fire and road traffic accident) while infection and childhood cancer accounted for the deaths of the remaining infants. None of these deaths were related to the mode of delivery.

Information on the long term outcomes was available for 1131 infants (94.3% of survivors). Fifty-four (4.7%) infants from this cohort developed long term needs. There were no statistically significant differences in the development of long term needs (which include both special needs and special educational needs) between the prelabour caesarean group and the vaginal delivery/caesarean section in labour group (Table 2). One infant among the survivors developed cerebral palsy. This infant was born vaginally and had suffered perinatal asphyxia and showed signs of moderately severe neonatal encephalopathy. This infant was one of the two babies that suffered with neonatal convulsions.

Table 3 illustrates the relationship between short and long term morbidities among the infants for whom long term data were available. The majority of babies that developed long term needs were born with normal Apgar scores, did not require admission to the neonatal unit and did not demonstrate convulsions during the neonatal period. Three of 36 babies in the vaginal/caesarean section in labour group with long term needs were born with Apgar score of less than 7 at 5 minutes as was 1 of the 18 babies in the prelabour caesarean group (P= 1.0). Similar findings were obtained when the need for neonatal unit admission was assessed in these babies—8 of the 36 babies needed admission in the vaginal group as 2 of the 18 in the prelabour caesarean group (P= 0.47).

Table 3.  Relationship between short and long term outcomes among term breech infants (n= 1131).
Short term morbiditynLong term morbidityP
5-minute Apgar score
<7374 (10.8)0.095
≥7109450 (4.6) 
 
Admission to the Neonatal Unit
Yes3310 (30.3)<0.0001
No109844 (4.0) 
 
Neonatal convulsions
Yes110.093
No113053 (4.7) 

DISCUSSION

This study provides incremental evidence about the lack of significant differences in long term morbidity between babies with breech presentation at term delivered by prelabour caesarean section and those born vaginally or by caesarean section in labour.

This study encompassed all term singleton breech presentations in a geographically defined area with a stable population. The results of this study are more suited to this population than the term breech trial which was carried out in 126 centres worldwide with more than 65 centres recruiting 10 patients or less. We chose to use special needs and special educational needs as markers of long term outcome as these provide functional information about the neurodevelopmental problems. We were able to procure the long term outcomes in 94% of the surviving infants with a minimum follow up age of at least three years and a maximum follow up of 10 years.

The limitations of this study include its retrospective design, possibility of secular changes affecting the outcomes, and the difficulty in defining the selection criteria for vaginal breech delivery. The retrospective design of the study may have introduced a selection bias making it difficult to state the other indications for the prelabour (planned) caesarean section. Similarly, there may have been a coding error for women in the planned caesarean section group that may have presented earlier in labour. We therefore chose to compare the group of women that underwent a prelabour caesarean section with the group consisting of vaginal breech births and caesarean section in labour. Equally, it was difficult to exclude the undiagnosed breech presentations at term.

Of the three perinatal deaths among infants without lethal malformations, there was evidence of suboptimal care in one death. The seventh annual report of the CESDI confirmed that the single most avoidable factor in causing breech stillbirths and death was suboptimal care in labour.10 In the term breech trial, too, there were three perinatal deaths in the countries with a low perinatal mortality.1 In two cases, there were fetal heart abnormalities recorded during labour. In our study, there was no reported case of trauma in the surviving infants. This may be due to the selective use of prostaglandins and syntocinon in our study compared with the term breech study. In the term breech study, ultrasound assessment of the fetuses prior to labour was performed in 59.9% of babies in the planned caesarean section group and 59.0% in the planned vaginal delivery group. The practice in our unit has involved either an ultrasound or pelvimetry assessment of all fetuses diagnosed in breech presentation at term. This might have allowed for better assessment and selection of babies suitable for vaginal delivery and therefore resulted in no real cases of birth trauma in the surviving infants. In keeping with previous studies, there were a significantly greater number of babies born with a low Apgar score following vaginal delivery than after prelabour caesarean section,1,5,11 but this difference was not reflected in the long term neurodevelopmental morbidity. Although Apgar scores provide a quick and a subjective method of assessing the newborn infant, by themselves, they are known to be poor predictors of long term neurological outcome. As in the term breech trial, babies born by vaginal breech delivery were more likely to need admission to the neonatal unit than those born by prelabour caesarean section.

The causes of late infant and childhood deaths in our cohort were not related to the mode of delivery or even to the presentation. Most other studies that have looked at long term outcome have not reported the long term deaths.12–14 Information on the long term development was obtained in 94.3% of the study population. This compares favourably with a similar study from Aberdeen where long term outcome was analysed in 1387 out of 1645 infants included in the study (84.3%).13 There was no statistical difference in the development of long term needs in the infants delivered by the two routes. Apart from the need for admission to the neonatal unit, short term morbidities were not predictive of adverse long term outcome in our cohort. In their study, Danielian et al.13 reported similar rates of severe handicap in the vaginal delivery group and the elective caesarean section group. Surprisingly, a higher number of infants with severe handicap were born to primigravidas following elective caesarean section in the Danielian study. Similarly, a recent Swedish study reported no differences in the rates of neurodevelopmental handicaps in term breech infants born by planned caesarean section or those intended to deliver vaginally.7 Mortimer et al.,15 in their case–control study, examined the long term neurological abnormality in 70 breech infants born by vaginal delivery and elective caesarean section matched to 70 vertex infants born by the two routes. The results document the absence of significant differences with respect to major brain damage affected by birth route changes for breech fetuses and between the entire breech and vertex populations.

CONCLUSION

Similar to the conclusion of the term breech trial, our data show that planned vaginal delivery of term breech infants is associated with a small but unequivocally increased risk of perinatal death and short term morbidity. However, the long term outcome of survivors is not affected by the method of delivery.

The information from this study is made routinely available to pregnant women with term breech presentation at our hospital for an informed choice about the mode of delivery.

Conflict of Interest

None declared.

Accepted 27 May 2004

Ancillary