Correspondence: Dr I. Z. MacKenzie, Nuffield Department of Obstetrics and Gynaecology, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU, UK.
Objective To describe the time interval between decision for assisted vaginal delivery and the birth of the baby in different clinical circumstances.
Design A prospective analysis of 225 consecutive women with a singleton fetal cephalic presentation in the second stage of labour requiring an operative vaginal delivery for various reasons.
Setting A maternity unit in a district general hospital delivering more than 6000 women annually.
Main outcome measures The decision to delivery interval and the immediate and short term maternal and neonatal outcomes according to indication for operative vaginal delivery.
Results The mean (SD) decision to delivery interval was 34.4 minutes (28.3) with a range of 5 to 101 minutes. For those delivered because of suspected fetal distress, the interval of 26.5 minutes (14.0) was significantly shorter than for those performed without fetal distress 39.5 minutes (19.0) (P < 0.0001); for cases with fetal distress, forceps were significantly quicker at 23.3 minutes (14.3) than the ventouse 29.2 minutes (13.2) (P= 0.04). The longer the interval in cases of fetal distress the less favourable the condition of the neonate at birth, although this trend did not reach statistical significance and was not seen for deliveries expedited for other reasons. Perineal repair was required following 96% forceps deliveries compared with 87% ventouse (P= 0.015). Perineal trauma was not influenced by the interval between decision and delivery.
Conclusions If speed of delivery is important, use of forceps results in a quicker birth than use of the ventouse, without any compromise to the condition of the baby at delivery, and with similar rates of perineal trauma.
There has been a demand for standards to be set for the time taken to deliver once the decision has been made for a caesarean section to be performed1. While there is a paucity of objective data upon which to base these standards, obstetric units are striving to comply with these demands.
Logically, such standards should also be set for the time taken from decision to birth for assisted vaginal deliveries, but there is similarly a dearth of published objective data for obstetric units or advisory bodies to turn to for assistance. The present study was therefore undertaken to provide some idea of the intervals that occur in a large district teaching obstetric unit when an operative vaginal delivery was required in singleton pregnancies.
A prospective survey of all operative vaginal deliveries of live singleton pregnancies with the fetus presenting by the vertex was conducted over the period 1 November 1997 to 1 February 1998. The staff engaged in the clinical provision of the service were not advised of the study in advance or during the collection of data. Unit policy required the time of all decisions for operative delivery, whether vaginal or by caesarean section, to be documented in the patient record; the time of delivery has always been recorded.
As a general rule, deliveries were conducted by year 1–3 trainees or senior house officers under the supervision of a more senior obstetrician; rotational deliveries, other than those occurring with the ventouse, were supervised or conducted by year 4–5 trainees or consultants. More experienced senior house officers performed nonrotational forceps and ventouse deliveries without supervision if deemed appropriate following discussion with their senior colleague.
Demographic details were recorded as shown in Table 1. Additional information was collected and included length of first and second stages of labour, indication for assistance with delivery, analgesia used both intrapartum and for delivery, method of assistance, success at first and subsequent attempts, ultimate successful method of delivery, maternal perineal trauma, and neonatal birthweight and condition at and following delivery. The primary indication for assistance with delivery was recorded as the more major if more than one indication was noted; fetal distress was taken as the more major when identified. The diagnosis of suspected fetal distress was that used by the clinicians managing the case, and was almost invariably made by the interpretation of a cardiotocogram recording.
Table 1. Demographic details. Values are given as n or mean (SD).
Initial method for assisted delivery
No. of cases
Maternal age (years)
Maternal height (cm)
Maternal weight (kg)
Length of first stage of labour (min)
Length of second stage of labour (min)
Statistical analyses have been made using χ2 test, Student's t test and one-way analysis of variance.
Two hundred and twenty-five deliveries were studied. Table 1 shows the characteristics of the women according to the instrument initially selected to assist the delivery; there were no significant differences between the two groups. Figure 1 illustrates the success of achieving delivery with the first attempt made. Only one of the women who required more than one attempt at vaginal delivery was multiparous.
Figure 2 shows the cumulative deliveries by time according to the indication for delivery assistance, whether for fetal distress or not and Fig. 3 illustrates the mean decision to delivery interval according to the time of day. The range of decision to delivery interval was 5–101 minutes. For the 134 cases with no fetal distress, the primary indication was delay in the second stage of labour in 125, maternal distress in seven, and for prophylaxis with a history of intracranial haemorrhage in two women.
Table 2 illustrates the mean (SD) decision to delivery intervals according to the initial instrument selected and primary indication for assisted delivery. The mean (SD) decision to delivery interval for women delivered by a consultant (8 deliveries) was 36.8 minutes (27.2), for year 4–5 trainees (52 deliveries) was 28.4 minutes (18.0), for year 1–3 trainees (93 deliveries) was 38.4 minutes (19.7), and for senior house officers (72 deliveries) was 33.2 minutes (14.1); in three instances the consultant was called out-of-hours to attend. Consultants and year 4–5 trainees selected forceps on 63% and 54% occasions respectively, and year 1–3 trainees and senior house officers used forceps on 30% and 40% occasions respectively. Forceps or manual rotations were performed for 21 of the forceps cases. Delivery was achieved more rapidly with both ventouse and forceps if fetal distress was suspected compared with cases without fetal distress; forceps resulted in a quicker delivery than the ventouse in this situation.
Table 2. Decision to delivery intervals according to indication for assisted delivery and selected initial mode of delivery. Values are given as mean (SD), unless otherwise indicated.
No fetal distress
The anaesthetic used for the delivery influenced the time interval. For 196 women delivered with a regional block (182 epidural, 14 spinal) the mean (SD) was 35.9 minutes (18.5) compared with 24.1 minutes (12.8) for 29 women delivered with local perineal infiltration (P= 0.001). If an epidural was in place for the first stage of labour, the decision to delivery interval was 33.3 minutes (16.3) in those 177 cases, which was significantly quicker than the 59.8 (21.4) for the 19 cases in whom the block was given for the delivery (P < 0.0001). For the 91 women having an assisted delivery because of suspected fetal distress, the mean (SD) interval was 27.8 minutes (14.3) using a regional block in 73 cases, and 19.0 minutes (8.9) using local perineal infiltration in 13 cases (P= 0.3).
Trauma to perineum occurred in 100 (87%) women delivered by the ventouse and 104 (96%) ultimately delivered with forceps (P= 0.015). The grade of the accoucheur did not influence the trauma rates, nor did the delivery interval. Six women suffered a third degree tear, five of whom were nulliparae; three followed delivery by ventouse (3%) and three following forceps (3%).
Table 3 illustrates the effect of the delivery interval on neonatal condition at birth depending upon the indication for assistance. The longer the interval over the first hour for fetal distress, the more acidaemic the cord arterial blood values became, but this did not reach statistical significance (P= 0.4). There were too few deliveries born after 60 minutes to assess the impact on acidaemia. Apgar scores did not show a similar trend. Increasing acidaemia was not observed following deliveries without fetal distress until the interval was greater than 60 minutes, although one baby was born after 67 minutes with a 5-minute Apgar score of < 9, cord arterial pH < 7.15 and base excess of > −12. The total length of second stage of labour did not affect the cord arterial pH value over the range 5–319 minutes.
Table 3. Neonatal condition according to indication for assisted delivery and decision to delivery interval. BE = base excess.
1'Apgar score (n*)
Cord artery pH
Cord artery BE (nmol/L)
*Number of cases in each time period.
There were 25 cases where the cord blood values were not obtained, 8 were in the distress group and 17 in the nondistress group.
Twenty-one babies were initially observed in the special care baby unit. Nine deliveries were for fetal distress and 12 for other reasons. Seventeen of these babies were delivered using the initially selected instrument and one after a second attempt at forceps delivery; three were delivered after an initial failed ventouse attempt, one successfully achieved with forceps, and two by caesarean section. Eleven babies had been delivered within 30 minutes of the decision and four were delivered after 60 or more minutes. Six babies developed a cephalhaematoma; four of these followed a ventouse delivery one of which was completed with forceps, and two followed a forceps delivery. All were discharged from the nursery having made good progress.
When delivery was expedited for suspected fetal distress, it was achieved more quickly at 26.5 minutes compared with 39.5 minutes when fetal distress was not suspected. These results are in broad agreement with those quoted by Vacca et al.2, but less impressive than those described by Johanson et al.3 in their analyses of morbidity outcomes using the ventouse and forceps. In both these studies, fewer than 50% of the women had a regional block for the delivery: the mean times for delivery using local anaesthesia were similar in both these reports and in the present study. However unlike Vacca et al.2, the time intervals were shorter when delivery was planned using forceps than ventouse. Of note, therefore, are the observations of others that the mean time to conduct the delivery is of the order of three to four minutes with similar values for one standard deviation4.
The status of the accoucheur did not significantly influence the time interval but as expected, senior house officers generally took less time than their senior colleagues who would manage the less straightforward cases. It was also reassuring to discover that the time of day when the delivery occurred did not influence the outcome.
The method of operative anaesthesia significantly influenced the decision to delivery interval. Local pelvic anaesthesia was quicker than a regional block. This was most marked for those women who were not using an epidural pain relief during the first stage of labour.
There was evidence that the longer the interval between decision and delivery, the less favourable was the condition at birth of those babies born with evidence of second stage fetal distress. However, this deterioration did not persist for those babies born after a delay of an hour or more but the number of such babies was small. A similar trend over the first 60 minutes delay was not seen for those babies born following an assisted delivery when fetal distress was not suspected.
The success with which the selected instrument was used to complete the delivery inevitably influenced the time taken from the decision to the delivery. While the success rate with the ventouse was relatively high in some studies2,4, with not more than 6%–15% being delivered ultimately with forceps or caesarean section, other studies have reported rates around 25% needing delivery at a second attempt3,5,6. By contrast, forceps deliveries were unsuccessful in 2%–10% of the studies comparing the two approaches2–6. The results from the more recent studies5,6 question the recommendation that the ventouse should be the instrument of first choice7 when a vaginal delivery needs to be expedited. In our analysis, which we acknowledge is observational, forceps resulted in delivery on average six minutes more quickly than with the ventouse (P < 0.05).
Perineal trauma, including episiotomy and/or laceration, was more commonly associated with delivery using forceps (96%) than with the ventouse (87%). These rates are similar to those reported by other studies2–4. The rate of third degree tears in our study, which was identical for both ventouse and forceps at 3%, was lower than in the other studies. These studies found a two to three-fold increase with the use of forceps (10%–21%) compared with the ventouse (5%–9%). The higher rates in the earlier studies could be due to a more assiduous definition of a third degree tear. The trend towards more cephalohaematomata following the ven-touse is in keeping with the experiences of others2,4.
The mean interval from decision to delivery of 26.5 minutes for fetal distress is approximately seven minutes shorter than the median time of 34 minutes quoted by Dunphy et al.8 for emergency caesarean section for fetal distress in 1989. In that study, as in the present one, the anaesthesia used for delivery influenced the interval, with regional blockade resulting in a longer delay. They found no direct correlation between the interval and neonatal condition, although there was a direct correlation between longer intervals and admission to the special care baby unit. In view of the results reported here for assisted vaginal for fetal distress, it is perhaps unreasonable to expect a caesarean section to result in delivery within 30 minutes of the decision as is presently recommended1.
While it is clear that every effort should be made to improve upon our results, it must be remembered that the decision to delivery intervals reported were obtained in a large well staffed delivery unit with resident senior trainee obstetricians and anaesthetists present at all times. Less well staffed units without resident staff and dedicated obstetric theatres could have difficulties achieving similar or better times.
The assistance of the District Audit Unit of the Oxford Radcliffe Hospital NHS Trust in keying-in data is gratefully acknowledged.