Stock, Sarah J. PhD; Josephs, Katherine BSc; Farquharson, Sarah BSc; Love, Corinne MD; Cooper, Sarah E. MD; Kissack, Chris MBChB; Akolekar, Ranjit PhD; Norman, Jane E. MD; Denison, Fiona C. MD
Kielland's rotational forceps are specifically designed for rotation and delivery of the fetal head, ie, in either a transverse or occipitoposterior position.1 There is regional variation in their use nationally and internationally,2–5 but generally there has been a decline in the use of rotational forceps over the past 30 years and there is an increasing preference among obstetricians to use cesarean delivery for this type of midcavity delivery.2–5 Although reasons for this decline are complex, the perception that rotational forceps are associated with increased complications and litigation is likely a strong influence.6,7 Concerns about litigation have also been cited as a reason for the increasing use of cesarean delivery as opposed to instrumental delivery for nonrotational midcavity deliveries.8,9
Modern obstetric guidelines10 reiterate the potential risks of rotational forceps, but the data that these recommendations are based on are more than 25 years old and originate from uncontrolled observational studies examining neonatal morbidity and mortality.11 Subsequent small studies have not supported the finding that rotational forceps are associated with increased neonatal morbidity and provide conflicting data on associations with maternal morbidity.3,12–21 There is therefore a need for contemporary data from large series to inform current obstetric practice.
The aim of this study was to estimate the rates of early neonatal and maternal complications in a series of consecutive Kielland's rotational forceps deliveries in a single center. Secondary aims were to compare the rates of complications of rotational forceps recorded in a maternity database with those of other types of delivery (spontaneous vertex delivery, nonrotational forceps delivery, ventouse delivery, and cesarean delivery) and to explore factors associated with complications of rotational delivery including the grade of operator.
MATERIALS AND METHODS
Ethical approval for the study was obtained from Lothian Research Ethics Committee (REC reference 09/S1102/28). The study was performed in the Simpson Centre for Reproductive Health. This is a tertiary referral obstetric center in the United Kingdom where Kielland's rotational forceps are regularly used for midcavity rotational operative deliveries.
To estimate the rates of early neonatal and maternal complications, we performed a retrospective cohort study of women with singleton pregnancies with cephalic presentation who were delivered successfully by Kielland's rotational forceps at 36 weeks of gestation or greater between January 2001 and December 2008. Data were not collected on deliveries subsequent to December 2008 because the hospital database system changed and there were difficulties with case ascertainment with the new system. There were 47,501 deliveries over the study period, 905 of which were Kielland's rotational forceps (1.90% deliveries). Records of all cases were requested for review. Five cases were misclassified and were actually nonrotational forceps deliveries (misclassification rate 0.55%), 10 Kielland's rotational forceps deliveries were performed at less than 36 weeks of gestation, and 17 were performed in one or more neonates of multiple pregnancies. These cases were excluded from analysis leaving a final cohort of 873 cases.
A secondary aim of the study was to compare complications of Kielland's forceps with other modes of delivery. Outcomes after Kielland's rotational deliveries were compared with a contemporaneous cohort of women with a singleton pregnancy at 36 weeks of gestation or more delivered in the same hospital between January 1, 2008 and December 31, 2008, by spontaneous vertex delivery (n=3,494), ventouse (n=159), nonrotational midcavity forceps (n=873), or emergency cesarean delivery (at any dilation; n=947). We restricted comparisons to this single year as a result of the large number of comparator deliveries over this time period. We examined the neonatal unit electronic patient records database (Badger) to estimate the proportion of neonates admitted after each mode of delivery during this period and the number diagnosed with neonatal encephalopathy. We also used data routinely recorded in the maternity database to compare the rates of postpartum hemorrhage and (vaginal deliveries only) obstetric anal sphincter injury associated with each mode of delivery.
Maternal outcomes after Kielland's rotational forceps delivery were collected by hand-searching maternity case records. Maternal outcomes were episiotomy, obstetric anal sphincter injury (third- or fourth-degree tear), cervical tears, bladder injury, postpartum hemorrhage of 1,000 mL or greater, postpartum pyrexia greater than 38°C, urinary retention (defined as recatheterization after removal of the catheter as a result of incomplete bladder emptying), urinary incontinence, fecal incontinence, and length of hospital stay. We also explored factors associated with complications. The following data were collected: birth weight, maternal age, body mass index (calculated as weight (kg)/[height (m)]2), presence of antenatal complication (preeclampsia or hypertensive disorder [defined as per International Society for Hypertension in Pregnancy Guidelines]),22 pregestational or gestational diabetes (diagnosed by the Scottish Intercollegiate Guideline Network, 200123), antepartum hemorrhage and other medical condition (including cardiac), onset of labor (spontaneous or induced), previous vaginal delivery, epidural anesthesia, fetal position (occipitotransverse or occipitoposterior), indication for Kielland's rotational forceps delivery (delay in second stage, nonreassuring cardiotocograph, or both, maternal medical indication for elective short second stage), and grade of operator (consultant, senior trainee [more than 5 years’ postgraduate experience and postgraduate qualification obtained], junior trainee).
In 26 cases, maternity case records were missing or incomplete. For these records, missing maternal data were imputed from the hospital maternity database, which holds more limited data. The following data only were therefore available for these women: obstetric anal sphincter injury, postpartum hemorrhage of 1,000 mL or greater, maternal age, maternal body mass index, onset of labor, previous vaginal delivery, and indication for Kielland's rotational forceps delivery.
Neonatal outcomes after Kielland's rotational forceps were collected from the maternity records and neonatal case records if a neonate was admitted to the neonatal unit. To ensure full case ascertainment of neonatal complications, the neonatal unit admission database was interrogated to identify admitted neonates and case notes were reviewed. This was to ensure that no neonatal admissions were missed. Similarly, local and national perinatal mortality databases were searched for every case of Kielland's rotational forceps delivery to identify whether any of the 873 neonates delivered by Kielland's rotational forceps delivery had died. No additional cases were found. The neonatal outcomes collected were perinatal mortality (stillbirth or death within first 28 days), admission to a neonatal unit for special or intensive care, duration of neonatal unit admission, indication for neonatal unit admission, arterial cord pH of less than 7.00, 5-minute Apgar score of less than 7, seizures, nerve palsy (including facial nerve and Erb's palsy), cephalohematoma, and any other significant trauma. Although transient bruising and forceps marks did occur and were sometimes described in the medical notes, these were not recorded as a result of variation in reporting.
No formal sample size calculation was calculated. We aimed to collect data on all cases through the specified time period to maximize power for detection of outcomes. Data were analyzed using SPSS 19. Fisher’s exact test was used to compare rates of neonatal unit admission, maternal trauma, and rate of postpartum hemorrhage between different modes of delivery. Logistic regression was performed to examine the influences of factors on complication rates. The following factors were included as dichotomous events in the model: age older than 35 years, previous vaginal delivery, presence of any antenatal complication, occipitotransverse position of the head, position, birth weight greater than 4 kg, and grade of operator performing the procedure. Factors found to be significantly associated with risk of maternal complications after univariable analysis were entered as independent variables in multivariable analysis using multiple logistic regression. Results were expressed as adjusted odds ratio (OR, 95% confidence intervals [CIs]). Missing fields were excluded from analysis. The number of missing fields for each outcome stated in the results were relevant. Significance was defined as P<.05.
Details of cases of Kielland's rotational forceps deliveries are provided in Table 1. The cohort had a mean age of 29.4 years (standard deviation 6.01). Primiparas accounted for 81.8% of the group, 62.5% had epidural anesthesia, and 85.3% were free from medical or obstetric complications. The indication for Kielland's rotational forceps delivery was delay in second stage in 62.3%.
Neonatal outcomes are summarized in Table 2. There was one intrapartum stillbirth. This was diagnosed before Kielland's forceps were applied and was associated with a congenital anomaly. This case was excluded from subsequent analysis of neonatal outcomes. There were no other perinatal deaths in the cohort.
In the 872 live-born neonates delivered by Kielland's rotational forceps, 17 neonates (1.9%) had Apgar scores less than 7 at 5 minutes. Arterial cord pH results were available in 845 neonates (96.9%) and were below 7.00 in 17 neonates (1.9%).
Congenital injury was recorded in 27 neonates (3.1%). Twelve neonates had cephalohematoma, 13 had nerve palsies (seven Erb's, four facial nerve, one with both Erb's and facial nerve palsies, one vocal cord paralysis). One neonate had a corneal abrasion, one neonate had a fractured clavicle, and one neonate had a subdural hematoma. One neonate had both cephalohematoma and facial nerve palsy. All Erb's and facial palsies were transient.
In total 58 neonates (6.7%) were admitted for neonatal special or intensive care. Fifteen of these were otherwise well neonates admitted from the postnatal ward or readmitted from home with transitional feeding difficulties (with or without jaundice or dehydration); one neonate was admitted for social reasons and one neonate for assessment of dysmorphism. If these cases, which are unlikely to relate to delivery, are excluded, the neonatal admission rate was 41 of 872 (4.7%). The most common primary reason for neonatal unit admission was respiratory problems (20 neonates), low cord pH (five neonates), and suspected sepsis (three neonates). In seven neonates (0.8%), a diagnosis of neonatal encephalopathy was recorded.
Maternal outcomes are also summarized in Table 2. Fifty-three women (6.1%) had obstetric anal sphincter injury, 121 (13.9%) a vaginal laceration, and two women (0.2%) had a cervical tear. All women had episiotomies performed. There were no cases of bladder injury recorded. Fifty-seven women (6.5%) had a postpartum hemorrhage greater than 1,000 mL. Overall 15 women (2.0%) had postpartum pyrexia, and 12 women (1.4%) had perineal infection or wound breakdown. Twenty-nine (3.3%) women had urinary retention, seven women (0.8%) had urinary incontinence, and four women (0.5%) had bowel symptoms (one woman with fecal urgency, one woman with pain on defecation, and two women with fecal incontinence). The median length of stay postnatally was 3 days (range 1–9 days) with 64 women (7.3%) staying more than 4 days. Overall, 276 women (31.6%) had one or more complication (defined as obstetric anal sphincter injury, genital tract trauma, postpartum hemorrhage, other early postnatal complication, or postnatal hospitalization of more than 4 days).
Table 3 compares outcomes for women delivered by Kielland's rotational forceps compared with other modes of delivery in the Simpson Centre for Reproductive Health during 2008. The maternal demographics of women delivered by Kielland's rotational forceps delivery in 2008 were not significantly different from those of overall cohort (data not shown).
The rate of admission to the neonatal unit after Kielland's rotational forceps was not significantly different from rates after other modes of vaginal birth (3.3% [5 of 150] compared with 6.1% [53 of 873] for nonrotational midcavity forceps delivery, 3.8% [6 of 159] for ventouse, and 3.7% [128 of 3,494] for spontaneous vertex delivery; P>.05). The rate of neonatal admission after emergency cesarean delivery was significantly higher than for other modes of delivery (11.2% [106 of 947], P=.002). There was no significant difference in rates of neonatal unit admission with neonatal encephalopathy after Kielland's rotational forceps delivery compared with other modes of delivery, but the overall number of cases of neonatal encephalopathy was small.
The incidence of postpartum hemorrhage was not significantly different after Kielland's rotational forceps delivery than after nonrotational forceps delivery (5.3% compared with 7.2%; P=.49) or ventouse delivery (2.5%; P=.25) and was significantly lower than after emergency cesarean delivery (15.0%; P=.008). It was however higher than that associated with spontaneous vertex delivery (2.3%, P=.027). The incidence of obstetric anal sphincter injury after Kielland's rotational forceps was also similar to that of nonrotational forceps (9.3% compared with 8.5%; P=.64) but higher than after ventouse delivery or spontaneous vertex delivery (1.9% and 2.9%, respectively; P=.005 and<.001).
Because numbers of cases with neonatal complications was low, we did not undertake further analysis of associated factors.
The contribution of factors that might be associated with maternal complications after Kielland's rotational forceps delivery was assessed using univariable analysis. Results are presented in Table 4 as ORs with 95% CI. The primary outcome of maternal complication was defined as a dichotomous event. Induction of labor and epidural were associated with increased odds of maternal complications (OR 1.42 [95% CI 1.03–1.96] and 1.67 [95% CI 1.23–2.28], respectively), whereas previous vaginal delivery was associated with a reduction in odds of maternal complication (OR 0.38 [95% CI 0.22–0.66]). No other factor was associated with a significant change in risk of maternal complication. The three significant factors were entered as independent variables in multivariable analysis with multiple logistic regression. Adjusted ORs with 95% CI were calculated and shown in Table 5. After adjustment, induction of labor was no longer associated with increased odds of maternal complications after Kielland's rotational forceps delivery, whereas previous vaginal delivery remained protective (adjusted OR 0.39 [0.23–0.67]) and epidural remained associated with increased odds of maternal complications (adjusted OR 1.60 [1.17–2.19]).
The findings of this large consecutive series of Kielland's forceps deliveries suggest a low neonatal complication rate after Kielland's forceps delivery. This supports smaller earlier studies12,17,24 and provides evidence that the perception that rotational forceps are dangerous may be unfounded. This study was primarily designed as a descriptive study of outcomes of Kielland's rotational forceps. To put these findings in context, we compared complication rates associated with Kielland's rotational forceps with those occurring after other types of deliveries. In this secondary analysis we made no attempt to match the different delivery groups or adjust for potential confounding factors between the groups. The mode of delivery reflects, among other things, maternal and fetal well-being, the anticipated difficulty of delivery, and maternal and operator preference for mode of delivery. This means that different delivery groups are fundamentally different with regard to case complexity. Furthermore, it is impossible to accurately determine such factors retrospectively from maternity records, yet they are likely to influence both the mode of delivery and outcomes and thus confound attempts to demonstrate superiority of one method over another retrospectively. We therefore decided to simply present descriptive outcomes, believing that this approach will be useful in informing the operator of the likely outcomes for women and their neonates delivered by Kielland's rotational forceps. The limitations of this approach to draw conclusions regarding the superiority of one mode of delivery over another in individual women should be recognized.
Like with any retrospective study, this study was dependent on the quality of the data recorded. We performed careful and systematic review of medical records collecting predefined data in an attempt to maximize accuracy. We used multiple sources of data to collect neonatal outcome data so as to increase confidence that no cases of perinatal mortality or significant morbidity were missed. In 26 cases, some or all of the maternal outcome data were obtained from the maternity database, which may be less accurate than data obtained from review of the maternity case records because the maternity case records are the primary source of outcome data. Nevertheless, we found that the database fields included were concordant with maternity case records in more than 98% of cases in which both sources of data were available.
A limitation of our study is the inclusion of all emergency cesarean deliveries rather than only those performed at full dilatation. Many emergency cesarean deliveries are performed for suspected fetal compromise, and the higher neonatal unit admission rate seen in association with cesarean delivery may reflect this. A prospective study of operative delivery in the second stage of labor found that both maternal and neonatal morbidity was lower after operative vaginal delivery than after full dilation cesarean delivery, supporting our findings.25 Another limitation is that we were unable to identify patients undergoing manual rotation before instrumental delivery and compare outcomes of neonates who underwent manual rotation with those of Kielland's rotational forceps. Finally, our maternity database only records the final mode of delivery. Thus, for women whose final mode of delivery was by cesarean delivery, we were unable to identify those women in whom delivery was initially attempted by Kielland's rotational forceps, ventouse, or nonrotational forceps. We were therefore unable to report the failure rate after Kielland's rotational forceps delivery for our cohort and compare maternal and neonatal outcomes between successful and unsuccessful Kielland's rotational forceps deliveries. The incomplete ascertainment of unsuccessful attempts of operative vaginal delivery and reporting of outcomes of only successful deliveries means that our study is representative of outcomes only in women who were ultimately delivered by Kielland's forceps. However, it is possible that outcomes in all women in whom Kielland's forceps were attempted are better or worse than those reported here.
We have only examined immediate complication rates associated with delivery. Qualitative work has shown that women frequently feel inadequately prepared for forceps delivery and that it affects their views about future deliveries.26 Nevertheless, a prospective cohort study found that women were more likely to prefer a future vaginal delivery after a successful forceps delivery than after a cesarean delivery.27 Further research is needed on longer-term outcomes of Kielland's rotational forceps, maternal satisfaction, and preference for future deliveries.
This study is not powered to estimate the incidence of rare events such as neonatal mortality. In 2009, a Kielland's rotational forceps delivery was associated with a neonatal death in the Simpson Centre for Reproductive Health.28 However, this neonatal death fell outside our prespecified study period and as a result of changes in our hospital database system, we were unable to collect data on deliveries subsequent to December 2008. We were therefore unable to determine whether there were any neonatal deaths associated with other modes of delivery as a comparator for the neonatal death associated with the Kielland's rotational forceps delivery. Despite this, the low incidence of neonatal complications after Kielland's rotational forceps delivery in the present study is reassuring.
This study was performed in a single center where the majority of obstetric consultants have expertise in rotational forceps delivery and trainees are taught and supported in performing rotational forceps deliveries. This is an increasingly unusual situation,2 and in many centers rates of operative vaginal delivery are low and skills for performing rotational deliveries have been lost. Widespread reintroduction of rotational forceps delivery would be challenging and would depend on a program of training and support by obstetricians experienced in rotational delivery. This may be particularly difficult in the United States where those with recent experience in Kielland's forceps deliveries are in short supply. However, we believe that there is a place for Kielland's rotational forceps delivery of neonates with malposition of the head. Increasing the availability of this form of delivery may have potential to reduce maternal complications associated with cesarean delivery, both in the short term and in subsequent pregnancies.
1. Dunn PM. Dr Christian Kielland of Oslo (1871–1941) and his straight forceps. Arch Dis Child Fetal Neonatal Ed 2004;89:F465–7.
2. Chinnock M, Robson S. An anonymous survey of registrar training in the use of Kjelland's forceps in Australia. Aust N Z J Obstet Gynaecol 2009;49:515–6.
3. Tan KH, Sim R, Yam KL. Kielland's forceps delivery: is it a dying art? Singapore Med J 1992;33:380–2.
4. Jain V, Guleria K, Gopalan S, Narang A. Mode of delivery in deep transverse arrest. Int J Gynaecol Obstet 1993;43:129–35.
5. Olah KS. In praise of Kielland's forceps. BJOG 2002;109:492–4.
6. Patel RR, Murphy DJ. Forceps delivery in modern obstetric practice. BMJ 2004;328:1302–5.
7. Park JS, Robinson JN, Norwitz ER. Rotational forceps: should these procedures be abandoned? Semin Perinatol 2003;27:112–20.
8. Localio AR, Lawthers AG, Bengtson JM, Hebert LE, Weaver SL, Brennan TA, et al.. Relationship between malpractice claims and cesarean delivery. JAMA 1993;269:366–73.
9. Werner EF, Janevic TM, Illuzzi J, Funai EF, Savitz DA, Lipkind HS. Mode of delivery in nulliparous women and neonatal intracranial injury. Obstet Gynecol 2011;118:1239–46.
10. RCOG. Operative vaginal delivery. Green-top guidelines. London (UK): Royal College of Obstetricians and Gynaecologists; 2005.
11. Chiswick ML, James DK. Kielland's forceps: association with neonatal morbidity and mortality. Br Med J 1979;1:7–9.
12. Chow SL, Johnson CM, Anderson TD, Hughes JH. Rotational delivery with Kielland's forceps. Med J Aust 1987;146:616–9.
13. Hinton L, Ong S, Danielian PJ. Kiellands forceps delivery—quantification of neonatal and maternal morbidity. Int J Gynaecol Obstet 2001;74:289–91.
14. Herabutya Y, O-Prasertsawat P, Boonrangsimant P. Kielland's forceps or ventouse—a comparison. Br J Obstet Gynaecol 1988;95:483–7.
15. Feldman DM, Borgida AF, Sauer F, Rodis JF. Rotational versus nonrotational forceps: maternal and neonatal outcomes. Am J Obstet Gynecol 1999;181:1185–7.
16. Hankins GD, Leicht T, Van Hook J, Uckan EM. The role of forceps rotation in maternal and neonatal injury. Am J Obstet Gynecol 1999;180:231–4.
17. Healy DL, Quinn MA, Pepperell RJ. Rotational delivery of the fetus: Kielland's forceps and two other methods compared. Br J Obstet Gynaecol 1982;89:501–6.
18. Schiff E, Friedman SA, Zolti M, Avraham A, Kayam Z, Mashiach S, et al.. A matched controlled study of Kielland's forceps for transverse arrest of the fetal vertex. J Obstet Gynaecol 2001;21:576–9.
19. Leo MV, Odibo A, Ling PY, Rodis J, Borgida A, Campbell W. Transverse arrest: a review of outcomes of rotational forceps and cesarean delivery at a single center. Prim Care Update Ob Gyns 1998;5:186.
20. Traub AI, Morrow RJ, Ritchie JW, Dornan KJ. A continuing use for Kielland's forceps? Br J Obstet Gynaecol 1984;91:894–8.
21. Cardozo LD, Gibb DM, Studd JW, Cooper DJ. Should we abandon Kielland's forceps? Br Med J (Clin Res Ed) 1983;287:315–7.
22. Brown MA, Lindheimer MD, de Swiet M, Van Assche A, Moutquin JM. The classification and diagnosis of the hypertensive disorders of pregnancy: statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens Pregnancy 2001;20:IX–XIV.
23. SIGN. Management of diabetes. SIGN Publication No. 55. Edinburgh (Scotland): SIGN; 2001.
24. Al-Suhel R, Gill S, Robson S, Shadbolt B. Kjelland's forceps in the new millennium. Maternal and neonatal outcomes of attempted rotational forceps delivery. Aust N Z J Obstet Gynaecol 2009;49:510–4.
25. Murphy DJ, Liebling RE, Verity L, Swingler R, Patel R. Early maternal and neonatal morbidity associated with operative delivery in second stage of labour: a cohort study. Lancet 2001;358:1203–7.
26. Murphy DJ, Pope C, Frost J, Liebling RE. Women's views on the impact of operative delivery in the second stage of labour: qualitative interview study. BMJ 2003;327:1132.
27. Murphy DJ, Liebling RE. Cohort study of maternal views on future mode of delivery after operative delivery in the second stage of labor. Am J Obstet Gynecol 2003;188:542–8.
28. Feinmann J. Why do doctors still use forceps when they killed our baby? Mail Online 2010, February 22.