British Journal of Obstetrics and Gynaecology Novcmber 1992, Vol. 90, pp. 903-906
FETAL AND NEONATAL M E D I C I N E
The prognostic value of fetal ultrasonography before induction of labour NIGEL SAUNDERS Senior Lecturer SHAMIM AMIS Senior House O f i c e r MICHAEL MARSH Registrar Academic Department of Obstetrics and Gynaecology, Imperial College of Science, Technology and Medicine, St Mary’s Hospital Medical School, London W2 1PG, UK
ABSTRACT Objective To investigate whether ultrasound examination ot the fetus and amniotic fluid before induction of labour can provide useful prognostic information about the course of labour and mode of delivery. Design A prospective observational study. Setting The delivery suite of St Mary’s Hospital, London. Suhjects 101 women undergoing induction of labour. Measured variables Biparietal diameter, head circumference, abdominal circumference, occipital position, amniotic fluid volume and umbilical artery Doppler velocimetry waveforms. Main outcome measures The outcome of labour was assessed in terms of the induction delivery interval, the mode of delivery and the incidence of abnormal cardiotocograms in labour. Results Seven women were delivered by emergency caesarean section early in the first stage of labour because of a significant abnormality of the fetal heart trace and these pregnancies were characterized by small fetal abdominal circumference measurements. Marked oligohydramnios was also noted in five of these seven subjects. Small or average size babies who were surrounded by an adequate volume of amniotic fluid tended to be born spontaneously, whereas the majority of operative deliveries for failure to progress in labour were associated with fetal abdominal circumference measurements 3340 mm. The mean Bishop score of women before labour who delivered spontaneously was not significantly different from the scores of women who had a forceps delivery or caesarean section. Abnormal cord Doppler waveforms were observed in only one instance. Conclusions In women undergoing induction of labour. the measurement by ultrasonography of two variables (fetal abdominal Circumference and amniotic fluid volume) may allow the prospective identification of pregnancies at increased risk of fetal distress o r dystocia. If these findings can be replicated in carly spontaneous labour than more rational utilization of resources may be possible in hospital based obstetric practice.
Diagnostic ultrasonography has had a major influence on obstetric practice over the past two decades. The technique has been applied primarily to the diagnosis of fetal malformation in early pregnancy and the assessment of fetal growth and well being during the third trimester. Surprisingly little research has been carried out to determine whether an ultrasound examination of the fetus during early labour may be of value in the prediction of fetal asphyxia o r difficult birth. To explore this possibility. a pilot study was carried out in which an ultrasound examination was undertaken immediately before induction of labour in 101 pregnancies.
Subjects and methods The study was approved by the hospital ethics committee and Corrcspondence: Nigel Saunders, Consultant Obstetrician, The Princess Anne Hospital, Coxford Road, SouthamptonSO9 4HA, UK.
the women’s consent was obtained before the examination. All 101 women had a singleton fetus with cephalic presentation. Ultrasound measurements were obtained using an A L O K A SSD-650 scanner (Aloka Co., Japan) equipped with a 3.5 MHz convex sector probe with pulsed wave Doppler facility. Fetal size was assessed by measuring abdominal circumference, head circumference and biparietal diameter using standard methodology (Neilson 1990). Occipital position was assessed by noting the position of fetal intracranial structures in relation to the maternal pelvis (Rayburn et al. 1989). Amniotic fluid volume was measured using thc 4-quadrant technique (Phelan et al. 1987). Doppler waveform patterns in the umbilical vessels were recorded with the wall filter set at 100 Hz. This variable was recorded simply as the presence or absence of end-diastolic flow in the umbilical artery (Pearce & McParland 1991). Most of the scans were performed by a single observer who was not responsible for
N . S A U N D E R S ET A L .
Table 1. Indications for induction of labour ( n ) .
Gestation 42 weeks Hypertension Prelabour rupture of membranes Maternal diabetes Suspected fetal growth retardation Abnormal fetal heart tracing Rhesus iso-immunisation Other
36 26 13 4 3 2 2 14
labour ward management and the information derived from the examination was not conveyed to the labour ward staff. O n e exception was a case of unsuspected placenta praevia which has been excluded from the analysis. Before induction of labour the condition of the cervix was assessed using the Bishop score (Bishop 1964). The outcome of labour was recorded with respect to the induction-delivery interval, the length of labour, the incidence of abnormal cardiotocograms during labour and the mode of delivery. If caesarean delivery was required a note was made as to whether the procedure had been performed primarily because of concern about fetal well being or primarily because of failure to progress in labour. The 95% confidence intervals (CI) for differences in mean values and proportions were calculated using appropriate computer software (Gardner et al. 1989).
Results The indications for induction of labour are listed in Table 1. Prostaglandin gel was used as the initial method of induction in 73% of cases and amniotomy with or without an intravenous’infusion of oxytocin in the remainder. Forty nine per cent of the women were primiparae and the mean gestational age at induction was 40 weeks (range 33-43). The median interval between induction and delivery was 13 h (range 1 4 3 ) . The median length of the first stage of labour was 4 h (range 1-27) and the median length of the second stage was 38 min (range 5-300). Mean birth weight was 3 373 g (range 1 6 3 0 4 700). Sixty women achieved a spontaneous vaginal delivery, 18 required forceps o r ventouse delivery and 22 were delivered by lower segment caesarean section (the overall caesarean section rate in this unit is currently 15Y0). The majority of women delivered by forceps were primiparae but the parity distribution of women delivered by caesarean section was similar to that of women who delivered spontaneously. The majority of vaginal operative deliveries (11/18) were undertaken because of lack of progress in the second stage of labour.
Table 3. Occipital position on scan before induction compared with occipital position at the time of operative delivery.
Position on scan Position at delivery*
Occipito-posterior Occipito-transverse Occipito-anterior
*In another 5 cases the position of the occiput at the time of operative delivery was not recorded in the case notes.
Of the caesarean sections one was performed for cord prolapse and six were performed primarily because of cardiotocographic abnormalities early in the course of labour before the cervix was sufficiently dilated to allow fetal scalp blood sampling. The other caesarean deliveries were performed for failure to progress in labour and the mean interval between induction and delivery in these cases was 26 h. In this group significant cardiotocographic abnormalities occurred during labour in two women, but fetal scalp blood sampling did not reveal acidosis. There were no significant differences in the mean Bishop scores between women who delivered spodtaneously (4.8, 95% CI 4.2-5.3), and those who required forceps (4.6,9.5% CI 34-58) or caesarean section (4.3, 95% CI 3.8-4.8). The relation between occipital position before induction and mode of delivery is shown in Table 2. There was a nonsignificant trend towards more caesarean deliveries when the occiput was posterior. Although the number of women with occipito-posterior positions is too small to allow statistical analysis, the mean abdominal circumference was less in infants born spontaneously (340 mm) than in those born by caesarean section for failure to progress (364 mm). In women who required operative delivery there was a good correlation between the position of the occiput before induction and the position at delivery (Table 3). There were several cases where an occipito-transverse position on scan was noted to be occipito-anterior at delivery (and vice versa) but there were no anomalous outcomes (such as right occipito-posterior on scan delivered as left occipito-transverse). The abdominal circumference and biparietal diameter measurements of babies who had to be born by caesarean section for fetal distress during the first stage of labour were significantly smaller than the corresponding measurements of babies who were born spontaneously or who required caesarean section for failure to progress (Figs. 1 and 2). In only one of these cases had labour been induced primarily because of suspected intrauterine growth retardation o r an abnormal fetal heart tracing. In contrast, four of the five women who had labour induced primarily for these reasons achieved a spontaneous vaginal delivery.
Table 2. Occipital position before induction and mode of delivery.
Mode of delivery
( n = 46)
( n = 40)
( n = 14)
Spontaneous Forceps Caesarean section
26 (57%) 12 (26%) 8 (17%)
25 (63%) 6 (15%) 9 (22%)
P R O G N O S T I C V A L U E O F FETAL ULTRASONOGRAPHY
2 390 c
LSCS (failUre to progress)
96 a, a, 94
- 0.6 a
,z -0 m
LSOCS (fetal distress)
The mean abdominal circumference measurement was significantly smaller in babies born spontaneously (341 mm, 95% CI 334-348) than in those born by forceps (366 mm, 95% CI 358-375) but not significantly smaller than in those born by caesarean section for failure to progress (359 mm, 95% CI 34.5-372). Head measurements were, in general, less discriminating than abdominal measurements (Fig. 2) and in six cases the station of the vertex precluded the accurate measurement of the biparietal diameter. Although smaller babies were associated with a shorter second stage of labour, fetal measurements were not useful for predicting the induction-delivery interval or the overall length of labour.
.-c .-> .-Ill
I W > 'i:0.4 .-
Fig. 1. Fetal abdominal circumference measurements (mm) in relation to method of delivery. The thick bars indicate the mean values for each group and the light bars represent the 95% confidence intervals (CI) for mean values. The open circles refer to cases with a 4-quadrant liquor volume