Maternal and neonatal morbidity in relation to the instrument used for mid-cavity rotational operative vaginal delivery: a prospective cohort study

Authors


Abstract

Objective

To compare the maternal and neonatal morbidity associated with alternative instruments used to perform a mid-cavity rotational delivery.

Design

A prospective cohort study.

Setting

Two university teaching hospitals in Scotland and England.

Population

Three hundred and eighty-one nulliparous women who had a mid-cavity rotational operative vaginal delivery.

Methods

A data collection sheet was completed by the research team following delivery.

Main outcome measures

Postpartum haemorrhage, third- and fourth-degree perineal tears, low cord pH, neonatal trauma, and failed or sequential operative vaginal delivery.

Results

One hundred and sixty-three women (42.8%) underwent manual rotation followed by non-rotational forceps delivery, 73 (19.1%) had a rotational vacuum delivery, and 145 (38.1%) delivered with the assistance of rotational (Kielland) forceps. The rates of postpartum haemorrhage were similar when comparing manual rotation with rotational vacuum (adjusted OR 1.42, 95% CI 0.66–3.98), and when comparing manual rotation with Kielland forceps (adjusted OR 1.22, 95% CI 0.71–2.88). The results were comparable for third- and fourth-degree perineal tears (adjusted OR 0.85, 95% CI 0.13–1.89; adjusted OR 0.94, 95% CI 0.39–1.82), low cord pH (adjusted OR 1.76, 95% CI 0.44–6.91; adjusted OR 1.12, 95% CI 0.44–2.83), neonatal trauma (adjusted OR 0.50, 95% CI 0.16–1.55; adjusted OR 3.25, 95% CI 0.65–16.17), and admission to the neonatal intensive care unit (adjusted OR 1.47, 95% CI 0.45–4.81; adjusted OR 1.04, 95% CI 0.49–2.19). The sequential use of instruments was less likely with manual rotation and forceps than with rotational vacuum delivery (0.6 versus 36.9%, OR 0.01, 95% CI 0.002–0.090).

Conclusions

Maternal and perinatal outcomes are comparable with Kielland forceps, vacuum extraction, and manual rotation, with few serious adverse outcomes. With appropriate training mid-cavity rotational delivery can be practiced safely, including the use of Kielland forceps.

Introduction

Mid-cavity rotational operative vaginal delivery (ROVD) is a complex procedure. The commonly used techniques are direct traction forceps or vacuum extraction after manual rotation of the fetal head, rotational vacuum extraction, and the use of rotational forceps such as Kielland forceps. However, there is concern regarding the potential for increased morbidity with these procedures, thereby leading to the increased use of caesarean section at full dilatation.[1, 2] There are historical case studies reporting increased neonatal morbidity with Kielland forceps.[3-5] This has led to a reduction in the use of Kielland forceps. Rotational vacuum extraction and manual rotation of the fetal head are now preferred, particularly by junior obstetricians.[6, 7] There are few studies investigating the morbidity related to rotational vacuum extraction or manual rotation.[8-10] In these case studies morbidity has been reported in comparison with the morbidity associated with a non-rotational operative vaginal delivery or a spontaneous vaginal birth.

When faced with fetal malposition in the second stage of labour, the available options are either an ROVD or a caesarean section at full dilatation. A cohort study comparing the maternal and neonatal morbidity between these two modes of delivery reported an increased incidence of the baby requiring admission to the neonatal intensive care unit (NICU) and also an increased incidence of postpartum haemorrhage when a caesarean section was performed in the second stage.[11] Although an increased incidence of trauma to the baby was reported in the early neonatal period, there was no difference in the long-term neurodevelopment of babies born by ROVD compared with babies born by caesarean section.[11, 12] There is a lack of robust evidence reflecting current obstetric practice when comparing maternal and neonatal morbidity associated with manual rotation followed by direct traction forceps delivery, rotational vacuum extraction, and the use of Kielland forceps for mid-cavity ROVD. The aim of this study was to compare the maternal and neonatal morbidity associated with the instrument used to assist a mid-cavity rotational delivery in current obstetric practice.

Methods

The study took place at Ninewells Hospital and Medical School, Dundee, and St Michael's Hospital, Bristol. Both units are consultant-led maternity units in university teaching hospitals with approximately 3800 and 5500 deliveries per year, respectively. The maternity unit in Scotland recruited women from October 2004 to September 2006, and the unit in South-West England recruited women from June 2005 to August 2006. The annual rates of operative vaginal delivery (OVD) were 14.5 and 11.0%, respectively, which are representative of UK national rates. The forceps used for rotational delivery included direct traction forceps (Anderson, Neville Barnes, and Milne Murray forceps) following manual rotation of the fetal head from occipito-transverse (OT) or occcipito-posterior (OP) to occipito-anterior (OA) position, and Kielland forceps. In each centre soft vacuum devices, Kiwi omnicup disposable devices, and metal cup vacuums were used for rotational delivery with or without manual rotation of the fetal head to OA. The choice of instrument was left to the individual obstetrician or supervising obstetrician according to experience and personal preference for the clinical circumstances.

Cohort

All women delivered by OVD or caesarean section following an attempted OVD were eligible for inclusion in the study if they were nulliparous (no previous delivery after 24 weeks of gestation), with a live singleton pregnancy, and cephalic presentation at term (37 weeks of gestation). The analyses were based on an intention-to-treat principle. Therefore, women who had a caesarean section following a failed OVD were also included in the cohort. Participants were identified from labour ward records and electronic maternity databases. Women were not included in the cohort study if they were participating in the randomised controlled trial (RCT) of episiotomy use that was recruiting at that time (n = 200).[13]

Women who gave birth with the aid of forceps and/or vacuum extraction following rotation of the fetal head at the pelvic mid-cavity were identified. The clinical findings recorded at the outset of the procedure were used to identify all cases where mid-cavity rotation was per-formed. These were cases where the station of the fetal head was at the ischial spines, or 1 cm below the ischial spines, and there was a malposition of the fetal head, OT, or OP. Cases in which the delivery was completed in persistent OP position were excluded from the analysis.

The explanatory variables are detailed in a previous publication from this cohort.[14, 15] The data were obtained from the maternity hand-held records and the computerised databases for maternity and neonatology. The maternal and intrapartum factors of interest were recorded.

Outcome measures

The outcome measures for the mother included intrapartum events such as postpartum haemorrhage (a global estimate), extensive perineal tearing, anal sphincter involvement (third- and fourth-degree tears, as classified in the Royal College of Obstetricians and Gynaecologists' Green-top Guideline),[16] and complications with the healing of the perineal tears. Length of postnatal hospital stay, and urinary or bowel symptoms were reviewed. Prolonged postnatal stay was defined as a postnatal in-patient stay of greater than 3 days. If the urinary catheter was in situ beyond 24 hours, it was regarded as prolonged.

The outcome measures for the baby included neonatal trauma (cephalhaematoma, retinal haemorrhage, facial nerve palsy, brachial plexus injury, and fractures), low Apgar scores, and low arterial blood gases of pH 7.10 and base excess greater than −12.0. Admission to the NICU was considered an independent outcome measure.

The secondary outcome measures included the sequential use of instruments to complete the delivery and caesarean section after a failed OVD.

Statistical analysis

Descriptive statistics of the maternal, neonatal, labour, and delivery factors were used to characterise the cohort in relation to risk factors for mid-cavity rotational delivery. The morbidity analyses were in three groups comparing manual rotation followed by direct traction forceps (MR), Kielland rotational forceps (RF), and rotational vacuum delivery with or without manual rotation (RV). Results are presented as odds ratios (ORs) and 95% confidence intervals (95% CIs), or with chi-square tests for differences in proportions and Student's t-tests for differences in means. Multivariable logistic regression models were performed adjusting for important confounding factors. The most commonly used procedure (MR) was used as the index group and the other two groups were compared with the index group. Factors were tested in the multivariable models based on a statistically significant difference between the two groups in the univariate analyses (< 0.05) or if biologically plausible. Results are reported as adjusted odds ratios (ORs) with 95% confidence intervals (95% CIs). The statistical package spss 16.0 was used for analysis.

Results

The original cohort study including all women who delivered by OVD had 1424 women. The mode of delivery of the participating women is shown in Figure 1. Of these, 1360 achieved an OVD, and 64 had a caesarean section following an unsuccessful attempt at OVD. In total, 381 mid-cavity ROVDs were attempted, and 355 were successfully completed. The most commonly used method of ROVD was MR, used for 163 deliveries (42.8%), followed by RF for 145 deliveries (38.1%), and 73 deliveries (19.1%) were conducted using RV. When maternal and fetal factors likely to affect the mode of delivery were compared, women with body mass indices (BMIs) of greater than 30 were found to be more likely to receive MR and RF, compared with RV (OR 4.7, 95% CI 1.06–20.71; Table 1). A pudendal block was less likely to be used that an epidural analgesia for MR, when compared with RV (OR 0.14, 95% CI 0.03–0.54; Table 2). A comparison of labour and operator characteristics (Table 2) showed that specialist trainees (STs) in the third to fifth years of training were less likely to use RF for ROVD, when compared with consultants (OR 0.36, 95% CI 0.18–0.72) and with specialist trainees in their sixth and seventh years (OR 0.34, 95% CI 0.20–0.59). If the fetal head position was OT the operator was more likely to use MR (OR 2.57, 95% CI 1.62–4.08), and was less likely to opt for RF. The above confounding factors were taken into account in multivariable logistic regression models, as were birthweight greater than 4.0 kg and head circumference greater than 37 cm.

Table 1. Demographic details of women who had a mid-cavity rotational operative vaginal delivery, in relation to the choice of instrument
 MR n = 163 (%)RV n = 73 (%)RF n = 145 (%)OR (95% CI) MR versus RVOR (95% CI) MR versus RF
  1. a

    Statistically significant odds ratio.

  2. b

    IUGR, intrauterine growth restriction.

Age >35 years11 (6.7) 8 (10.9)  8 (5.5%)0.58 (0.22–1.53)1.24 (0.48–3.17)
Body mass index >3019 (11.7)2 (2.7)  16 (11.1)4.7 (1.06–20.71)a1.06 (0.52–2.15)
Pre-eclampsia11 (6.7)3 (4.1)  10 (6.9)1.7 (0.45–6.20)0.97 (0.40–2.37)
Suspected IUGRb4 (2.5)2 (2.7)   4 (2.8)0.89 (0.16–4.99)0.88 (0.22–3.61)
Birthweight >4.0 kg26 (15.9)7 (9.5)  14 (10.3)1.78 (0.74–4.33)1.34 (0.70–2.56)
Birthweight mean (g)35193467 3494
SD 439 477  434
Range2120–46601870–43302430–4640
Low birthweight (<10th centile)16 (9.8)7 (9.5)  15 (10.6)1.03 (0.41–2.64)0.96 (0.46–2.04)
Male gender85 (52.1)39 (53.4)  90 (62.1)1.03 (0.59–1.79)0.66 (0.43–1.05)
Head circumference ≥37.0 cm8 (5.4)7 (9.5)  3 (2.1)1.27 (0.71–2.26)2.44 (0.63–9.39)
Head circumference: mean (cm) 35.0 34.8  34.9
SD1.21.4   1.2
Range 30.5–38.231–37.5  31.7–37
Table 2. Details of labour and operator grade according to the instrument of choice for mid-cavity rotational delivery
 MR n = 163 (%)RV n = 73 (%)RF n = 145 (%)OR (95% CI) MR versus RVOR (95% CI) MR versus KF
  1. a

    Epidural analgesia used as the reference category for analgesia.

  2. b

    Statistically significant odds ratio.

  3. c

    Years of specialist training curriculum in obstetrics and gynaecology.

  4. d

    ST 3–5 used as the reference category for grade of operator.

Induction of labour51 (31.3)17 (23.2)48 (33.1)1.51 (0.79–2.93)0.92 (0.57–1.48)
Meconium-stained liquor52 (31.9)21 (28.7)36 (24.8)1.16 (0.63–2.12)1.06 (0.52–2.15))
Epidural anaesthesia124 (76.1)53 (72.6)89 (61.4)1.00a1.00a
Pudendal anaesthesia4 (2.5)12 (16.3)1 (0.7)0.14 (0.03–0.54)b2.15 (0.22–21.04)
Spinal anaesthesia50 (30.7)10 (13.6)60 (41.4)2.30 (1.05–5.07)b0.57 (0.35–1.02)
CTG abnormality in second stage89 (54.6)43 (58.9)62 (42.8)0.84 (0.48–1.47)1.61 (0.98–2.52)
First stage of labour >12 hours73 (44.8)17 (23.2)67 (46.2)2.67 (1.43–4.98)0.96 (0.46–2.04)
Second stage of labour >2 hours65 (39.8)21 (28.7)63 (43.2)1.64 (0.90–2.98)0.66 (0.43–1.05)
Primary indication: suspected fetal compromise54 (33.1)37 (50.6)41 (28.3)0.48 (0.29–1.02)1.25 (0.39–1.01)
Operator grade: ST 1–2c13 (7.9)9 (12.2)11 (7.6)0.52 (0.19–1.36)0.65 (0.27–1.54)
Operator grade: ST 3–5c (or equivalent)100 (61.3)32 (43.8)55 (37.9)1.00d1.00d
Operator grade: ST 6–7c (or equivalent)34 (20.9)25 (34.2)54 (37.2)0.40 (0.21–0.77)b0.34 (0.20–0.59)b
Operator grade: Consultant16 (9.8)7 (9.5)25 (17.2)0.72 (0.26–2.10)0.36 (0.18–0.72)b
ST6 or above present for deliveryc122 (74.8)59 (80.8)144 (99.3)0.70 (0.36–1.39)0.02 (0.001–0.15)b
Position at outset: OT103 (63.1)38 (52.1)58 (40.0)1.58 (0.90–2.76)2.57 (1.62–4.08)b
>3 contractions during which pulled6 (3.7)15 (20.5)1 (0.7)0.14 (0.05–0.39)b5.37 (0.64–45.19)
Severe caput9 (5.5)6 (8.2)4 (2.8)0.65 (0.22–1.90)2.06 (0.62–6.38)
Figure 1.

Mode of delivery for the women participating in the cohort study.

When maternal morbidities between the three groups were compared, there were no significant differences in vaginal wall lacerations, third- or fourth-degree perineal tears, postpartum haemorrhage, or urinary or faecal incontinence (Table 3). There was greater use of moderate to strong analgesia at the time of discharge from the hospital by women who had MR in comparison with those who had an RV (OR 2.38, 95% CI 1.23–4.59; Table 3). When compared with RF, women who underwent MR required less in-patient and less out-patient analgesia (OR 0.05, 95% CI 0.01–0.44; OR 0.56, 95% CI 0.28–0.98, respectively; Table 3). A urinary catheter was less likely to remain in situ for more than 24 hours in women who had MR compared with RF (OR 0.15, 95% CI 0.08–0.27). There were no significant differences between the three groups when comparing neonatal morbidities in terms of low Apgar scores, low umbilical cord pH, neonatal trauma, or admission to the neonatal intensive care unit (Table 4). In comparison with RV, the use of MR was less likely to result in the sequential use of instruments (OR 0.01, 95% CI 0.002–0.090). It was routine practice for some practitioners to use RF for rotation of the fetal head, remove the forceps, and then use non-rotational direct traction forceps to complete the delivery of the fetal head. This technique was reported in ten cases. However, as this was an intentional use of a second instrument, and traction was not applied with the RF, it was not classified as sequential instrument use. For the purpose of this study the sequential use of instruments was defined as the need to use a second instrument if the first instrument failed to effect rotation or traction as intended. In this cohort, a second instrument was not used following a failed attempt at RF delivery. There were no significant differences between instruments in the number of failed ROVDs and caesarean sections (Table 5).

Table 3. Maternal morbidity in relation to the first instrument of choice for mid-cavity rotational delivery
 MR n = 163 (%)RV n = 73 (%)RF n = 145 (%)Adjusted OR (95% CI)a MR versus RVAdjusted OR (95% CI)a MR versus RF
  1. a

    Adjusted for body mass index (BMI) >30, birthweight >4 kg, head circumference >37 cm, operator experience, number of contractions >3, and position of fetal head at the start of OVD.

  2. b

    Statistically significant odds ratio.

Vaginal wall lacerations31 (18.7)9 (12.3)29 (20.0)1.82 (0.71–4.62)0.84 (0.64–1.53)
Third- or fourth-degree tears17 (10.4)8 (10.9)15 (10.3)0.85 (0.13–1.89)0.94 (0.39–1.82)
Postpartum haemorrhage ≥ 500 ml51 (31.3)16 (21.9)36 (24.8)1.42 (0.66–3.98)1.22 (0.71–2.88)
Urinary retention1 (0.6)3 (4.1)0 (0)0.28 (0.02–3.73)
Urinary incontinence8 (4.9)2 (2.7)5 (3.4)1.29 (0.24–6.76)1.41 (0.44–4.52)
Faecal incontinence3 (1.8)2 (2.7)2 (1.4)0.74 (0.12–4.69)1.47 (0.24–8.97)
In-patient moderate/strong analgesia140 (85.9)55 (75.3)142 (97.9)2.11 (0.90–4.69)0.05 (0.01–0.44)b
Out-patient moderate/strong analgesia104 (63.8)29 (39.7)125 (86.2)2.38 (1.23–4.59)b0.56 (0.28–0.98)b
Postnatal admission >3 days30 (18.4)10 (13.6)43 (29.6)1.82 (0.47–3.88)0.66 (0.28–1.14)
Perineal haematoma1 (0.6)1 (1.3)1 (0.7)0.29 (0.02–4.96)0.74 (0.04–13.35)
Perineal infection2 (1.2)3 (4.1)4 (2.8)0.14 (0.01–1.41)0.67 (0.11–3.96)
Catheter >24 hours18 (11.1)4 (5.4)67 (46.2)2.27 (0.66–7.78)0.15 (0.08–0.27)b
Table 4. Neonatal morbidity in relation to the first instrument of choice for mid-cavity rotational delivery
 MR n = 163 (%)RV n = 73 (%)RF n = 145 (%)Adjusted OR (95% CI)a MR versus RVAdjusted OR (95% CI)a MR versus KF
  1. a

    Adjusted for body mass index (BMI) >30, birthweight >4 kg, head circumference >37 cm, operator experience, number of contractions >3, and position of fetal head at the start of OVD.

  2. b

    All neonatal trauma, excluding superficial bruising or swelling.

Apgar score at 1 minutes ≤ 35 (3.1)2 (2.7)5 (3.4)1.15 (0.20–6.65)0.85 (0.19–3.76)
Apgar score at 5 minutes <71 (0.6)1 (1.4)3 (2.1)0.49 (0.02–7.81)0.36 (0.04–3.54)
pH umbilical artery <7.1013 (7.9)4 (5.4)10 (6.9)1.76 (0.44–6.91)1.12 (0.44–2.83)
Base excess umbilical artery <−12.05 (3.1)1 (1.4)6 (4.1)2.30 (0.26–20.19)0.84 (0.25–2.86)
Neonatal traumab8 (4.9)7 (9.5)3 (2.1)0.50 (0.16–1.55)3.25 (0.65–16.17)
Admission to NICU20 (12.3)4 (5.4)15 (10.3)1.47 (0.45–4.81)1.04 (0.49–2.19)
Shoulder dystocia8 (4.9)1 (1.4)9 (6.2)3.08 (0.35–26.80)0.66 (0.21–2.00)
Table 5. Incidence of sequential use of instruments or failed operative vaginal delivery in relation to first instrument of choice
 MR n = 163 (%)RV n = 73 (%)RF n = 145 (%)Adjusted OR (95% CI)a MR versus RVAdjusted OR (95% CI)a MR versus KF
  1. a

    Adjusted for body mass index (BMI) >30, birthweight >4 kg, head circumference >37 cm, operator experience, number of contractions >3, and position of fetal head at the start of OVD.

  2. b

    When Kielland forceps were removed following the successful rotation of fetal head, and the direct traction forceps were then used for delivery of fetal head, the practice was not classified as sequential use.

  3. c

    Statistically significant odds ratio.

Second instrument used1 (0.6)27 (36.9) 0b0.01 (0.002–0.09)c
Caesarean section7 (4.2)5 (6.8)14 (9.6)0.51 (0.10–1.83)0.39 (0.14–1.06)

Discussion

Summary of main findings

This study shows that mid-cavity ROVDs are associated with few adverse maternal and perinatal outcomes overall, and that the morbidities with RF are comparable with RV and MR. The sequential use of instruments was less common with MR and RF. This is one of the largest cohort studies addressing the technique of MR for mid-cavity ROVD, and reports a high success rate and morbidity that is comparable with other techniques.

Strengths and weaknesses

This study included a cohort of all nulliparous women who gave birth by ROVD at the two teaching hospitals during the study period. The data were collected prospectively from the maternity notes and the maternity databases. The findings reflect current obstetric practice in the two units. Both units have senior obstetric support on the delivery suite in the form of a consultant obstetrician or senior trainee (ST6 or above or equivalent). Therefore, the majority of deliveries were supervised by a senior obstetrician with adequate experience and skill to conduct these complex deliveries.

The main limitation of this study is the cohort design, and although we have attempted to reduce the effect of confounding by taking the likely confounding factors into consideration when performing multivariate logistic regression analyses, there is likely to be some residual confounding in the findings of the study. Also, we have used the documented prescription of moderate or strong analgesia as a measure of pain after birth. The prescriptions may reflect an obstetrician's preferences in response to the existing evidence that a forceps delivery is likely to need stronger analgesia. Another limitation of the study is that the number of cases reviewed is limited in terms of statistical power to detect differences in less frequent but more severe neonatal complications, such as skull fractures or intracranial haemorrhages, or indeed perinatal deaths.

Comparison with published literature

This study focused on the comparative morbidities associated with mid-cavity rotational deliveries. However, the participants are part of a larger cohort of OVDs carried out in the participating centres during the study period. We have previously published the morbidity related to all vacuum extractions and forceps deliveries (low and mid cavity, rotational and non-rotational) as well as morbidity associated with the sequential use of instruments.[14, 15] The maternal and neonatal morbidity rates were similar when rotational forceps were compared with all forceps deliveries. However, the morbidity was greater when rotational vacuum extractions were compared with all vacuum extractions. This is likely to reflect an increased use of a second instrument with a rotational vacuum delivery, and we have previously reported a greater morbidity with the sequential use of instruments.[15] The sequential use of instruments is greatest when using vacuum extraction, and the increased morbidity associated with the sequential use of instruments should be considered when selecting the primary instrument for a mid-cavity rotational delivery.[15, 16]

In this study the rate of complications was low in all groups, including the RF group. There were no significant differences in the key maternal outcomes, such as third- and fourth-degree perineal tears and postpartum haemorrhage. The incidence of third- and fourth-degree perineal tears is reported to be greater with forceps when compared with vacuum extraction. In this cohort the difference was not significant. This may be because more than a third of the vacuum extraction deliveries were completed with obstetric forceps. The overall rate of third- and fourth-degree perineal tears was comparable with the existing literature.[17-20]

When comparing the neonatal outcomes, there were no significant differences in the number of babies born with low cord pHs, low Apgar scores, and needing admission to the NICU. The incidence of severe neonatal trauma in cases of RF is less than the incidence of trauma reported in previous studies.[3-5] The likely explanation for the lower morbidity in this study is that the majority (99%) of RFs were either performed or supervised by a senior obstetrician. It has been suggested that the morbidity associated with RFs can be lower if performed by a competent or senior obstetrician.[21-24] Our study supports RF as a safe instrument when performed by a skilled obstetrician. There was an increased need for analgesia when forceps were used to conduct the delivery. These findings are similar to the Cochrane systematic review comparing vacuum extraction and forceps delivery.[25] In this study the need for a second instrument to complete the delivery was greater when a vacuum extractor was used. Increased sequential instrument use has been reported with the vacuum extraction when comparing all OVDs, and our study shows that the trend continues with mid-cavity rotational deliveries.[25-27]

Clinical implications

In this study, mid-cavity rotational deliveries accounted for 26% of all OVDs, with a low failed ROVD rate of 6.8%. This shows that mid-cavity rotational deliveries constitute a substantial proportion of all attempted OVDs, and with appropriate expertise the vast majority can be delivered safely vaginally. Therefore, training programmes for obstetricians should include structured training in performing mid-cavity rotational deliveries. Appropriate training in decision making and the technique of conducting mid-cavity rotational delivery may also contribute to a reduction in the caesarean section rate at the second stage of labour by giving obstetricians confidence and competence in performing these deliveries safely.

Conclusion

Mid-cavity ROVD has comparable morbidities for RF, RV, and MR, with few serious adverse outcomes. Larger studies are needed to explore the rarer but clinically significant morbidities associated with OVDs, such as cerebral palsy and perinatal death.

Disclosure of interests

Nothing to declare.

Contribution to authorship

DJM had the idea for the original cohort study and, with BS, RB, and MV, carried out the design of the study. DJM, MM, and BS obtained funding. MM, MV, and RB were responsible for data collection. RB and MV carried out the analysis. RB drafted the article, which was revised by all authors.

Details of ethics approval

The study was approved by the Multi-centre Research Ethics Committee for Scotland (04/MRE00/48).

Funding

Tenovus Scotland. RB was supported by a Wellbeing of Women (WoW) Research Training Fellowship.

Acknowledgements

We are very grateful to the women in the study and the NHS and university staff who recorded and entered the data, and to Karen Goyder and Louise Howarth for assistance with data collection.

Commentary on ‘Maternal and neonatal morbidity in relation to the instrument used for mid-cavity rotational operative vaginal delivery: a prospective cohort study’

Much of clinical obstetrics is becoming a lost art. As evidence accumulates regarding clinical outcomes from appropriately designed studies, obstetric judgment is often replaced by fear of adverse outcomes. For example, in many developed countries vaginal breech deliveries have been relegated to history because of reports of poorer immediate outcomes, even though long-term outcomes indicate no harm (Hannah et al. Am J Obstet Gynecol 2004;191:864–871). Regrettably, going back is difficult, as a generation of clinical teachers has been lost and there are few who can teach such operative obstetric skills.

Similarly, the art of the operative vaginal delivery is becoming increasing subverted to fears of poor obstetric outcomes and litigation. In the USA, operative vaginal delivery has largely been replaced by caesarean delivery. Skill and judgment have been lost to expedient surgery in the name of ‘safety’. However, there is no compelling evidence that caesarean delivery is necessarily the safest route of delivery, particularly when examining long-term costs.

The work by Bahl et al., in this issue, highlights clinical outcomes that indicate that when instructed correctly, using sound obstetric judgment, trainees can successfully perform mid-pelvic rotational deliveries, and that the instrument chosen is not as relevant as is the skill of the accoucheur. However, the safety of their outcomes cannot be addressed, as an appropriate control group is not provided.

The rates of fetal trauma (4.7%), admission to NICU (10.2%), and shoulder dystocia (4.7%) seem high. Similarly, maternal complications seem excessive. Vaginal wall lacerations (18.1%), postpartum haemorrhage (27%), urinary catheterisation for longer than 24 hours (23.4%), and postnatal admission exceeding 3 days (21.8%) indicates significant maternal morbidity in their cohort. Although long-term outcomes, and costs, are not reported, these data support equivalent outcomes, regardless of the instrument used. A recent report incorporated an appropriate control group of women, who had a second-stage caesarean delivery after achieving complete dilation (Walsh et al. Obstet Gynecol 2013;121:122–128). Reassuringly, the rates of fetal and neonatal injury were the same in both groups.

In the USA, several convergent forces have conspired to dampen enthusiasm for mid-pelvic and rotational deliveries. Data reported from California birth certificates indicated that the sequential use of operative approaches (i.e. vacuum followed by forceps) resulted in significantly higher adverse perinatal outcomes (Towner et al. New Eng J Med 1999;341:1709–1714). In 1998 the US Food and Drug Administration (FDA) issued an alert regarding the use of the vacuum, warning providers to pay close attention to product inserts after several case reports of lethal intracranial haemorrhage in infants following the inappropriate use of the vacuum. The old adage that ‘one never gets sued for performing caesarean delivery’ certainly applies here.

So, reports such as the one by Bahl et al. are welcome, as they highlight the continued need to practice these valuable obstetric techniques and skills. Critical to this teaching, however, is imparting the obstetric judgment that only comes with experience and wisdom. Until more complete outcome data are available, obstetric prudence must be exercised in the use of the operative mid-pelvic delivery.

Disclosure of interests

The author has no relevant disclosures.

  • DJ Dudley

  • University of Texas, Health Science Center, San Antonio, TX, USA

Ancillary