In the past two decades there has been a major decline in use of the obstetric forceps internationally in association with an increased clinician preference for use of the vacuum extractor, or ventouse, to perform assisted vaginal deliveries.1 – 5 For assisted vaginal deliveries, when there is malposition of the fetal head, the use of rotational forceps has declined even more dramatically. Use of the Kielland's forceps, which is the most commonly used rotational forceps, has been largely discontinued, or was never initiated, by many obstetricians6 because of the controversy that has surrounded its use. This controversy has arisen because of reports, from relatively small case series, dated 20–30 years ago, from both European7 and U.S. groups,8 which described the adverse outcomes associated with its use and, particularly, its potential to cause fetal or maternal injury when used inappropriately. In addition, it is possible that the medicolegal environment that has pertained to this area of obstetrics has influenced medical practice and operator choice of instruments. Whether or not the Kielland's forceps should be abandoned completely remains an unsolved issue,8 and there are no real data from randomized controlled trials to outline best practice for midcavity rotational delivery, a relatively common clinical scenario in obstetric practice. Interestingly, the use of rotational forceps has been reviewed and supported in the American College of Obstetricians and Gynecologists Practice Bulletin Number 17,9 which includes a recommendation that it should be performed by only skilled health care practitioners. The aim of this study was to evaluate the use of the Kielland's forceps in current obstetric practice at a tertiary care obstetric department over a 15-year period from 1997 to 2011.
PATIENTS AND METHODS
The data for this study were obtained from an obstetric computerized database to which the information had been entered during the 15-year period from July 1997 to December 2011 inclusive at the Department of Obstetrics and Gynecology, Galway University Hospital, Ireland. This study was deemed exempt from the institutional review board at our institution. The database used was the Euroking System European Information Technology, Ottershaw, Surrey, United Kingdom. At the time of initial antenatal visit for the woman, a series of medical and demographic details pertaining to each woman was entered on the database by a midwife. Similarly, at the completion of delivery for every woman, and before discharge from the delivery suite, a further entry was made by the attending midwife, outlining the management of labor, mode of delivery, and neonatal details. In addition to the data entered on this database, the maternal and neonatal case records for each pregnancy for whom Kielland's forceps had been used were examined to validate all of the database findings. From these, the following outcome measures related to use of the Kielland's forceps were ascertained: 1) overall prevalence; 2) maternal age, parity, and body mass index (BMI, calculated as weight (kg)/[height (m)]2) at first antenatal visit for all mothers; 3) total duration of labor; 4) duration of second stage of labor; 5) presence of epidural analgesia; 6) position of the fetal head before attempted delivery; 7) station of the fetal head before attempted delivery; 8) indication for forceps delivery, ie, suspected fetal compromise or delayed second stage of labor (as defined below); 9) location of attempted Kielland's forceps assisted delivery (delivery suite room or operating room); 10) birth weight of neonates delivered; 11) degree of perineal trauma observed (perineum intact, episiotomy (mediolateral), first- or second-degree tear, third- or fourth-degree tear, or high vaginal laceration; 12) shoulder dystocia; 13) postpartum complications of hemorrhage or incontinence (urinary or fecal); 14) neonatal Apgar scores less than 7 at 1 minute and less than 7 at 5 minutes; 15) umbilical cord blood pH values (less than 7.20, 7.20–7.25, greater than 7.25); 16) the occurrence of forceps related neonatal trauma, neonatal encephalopathy, clavicular fracture, or brachial nerve palsy; 17) rate of admission to a neonatal special care baby unit; and 18) seniority of operator and presence or otherwise of a senior attending (consultant) obstetrician at delivery.
All senior obstetricians participating in this study had completed the Royal College of Obstetricians and Gynecologists, London, requirements for accreditation, which involved a period of 10 years of postgraduate clinical training. The clinical practice used was 1 hour of passive second stage after the diagnosis of full dilatation followed by 1 hour of maternal pushing. The second stage of labor was deemed delayed when delivery was not imminent after 1 hour of maternal pushing. Suspected fetal compromise consisted of evidence of a nonreassuring cardiotocogram. The technique used was that of direct application of the forceps blades when the fetal position was occipitoposterior, and for the occipitotransverse position, the anterior blade was applied by a technique of careful “wandering” of the blade over the fetal face into position with direct application of the posterior blade. After rotation, the delivery was completed using the Kielland's forceps. During the time period of the study (1997–2011), the overall operative delivery rates in the institution were as follows: the average forceps rate was 5.4% (1997, 10%; 2011, 5.5%); the average vacuum-assisted delivery rate was 13.2% (1997, 8.6%; 2011, 12%); and the average cesarean delivery rate was 24.8% (1997, 18.3%; 2011, 29.2%).
Microsoft Excel was used to tabulate and graphically summarize the data. The statistical package SPSS 18 was used to perform descriptive statistics. Finally, a logistic regression analysis for trend was performed to examine for any potential variation in the number of cases per year during the study.
During the time period of the study there were 45,335 neonates delivered, and Kielland's forceps was attempted in 144 cases (0.3%), leading to a successful assisted delivery for 129 neonates, 89.6% (95% confidence interval [CI] 83–94%), and a failed attempt in 15 cases (10.4%) (95% CI 6–17%). The results concerning the 15 cases of failed attempt at delivery are outlined separately at the end of this section.
For successfully assisted delivery with Kielland's forceps, the number of cases per year during the time period of the study varied from three to 21 with a mean of 8.6 cases (Fig. 1). There was no significant linear increase or decrease observed in the number of cases per year for the duration of the study (95% CI 0.93–1.02, P<.266). The vast majority of women for whom Kielland's forceps-assisted delivery took place were nulliparous (n=116; 89.9%), and the remaining were parous women (n=13; 10.1%). The mean age of the group was 31.9 years (±3.9, standard error of the mean [SEM]) with a range in age from 20–43 years. The rate of induction of labor for this group was 44.2% (n=57), and the remaining 55.8% (n=72) underwent spontaneous onset of labor. The indications for forceps-assisted delivery were as follows: delayed second stage of labor (n=94; 72.9%) and suspected fetal compromise (n=35; 27.1%).
The BMI details of the 129 women delivered using Kielland's forceps are demonstrated in Table 1. It is evident that 48% (n=59) were in the normal BMI category with 2% (n=2) in the low-weight category, and 50% of the women (n=61) were either overweight or obese (in seven cases the BMI was not recorded). The mean duration of labor was 10 hours 9 minutes (±32 minutes, SEM). The mean duration of the second stage of labor was 1 hour 48 minutes (±4 minutes, SEM). Epidural analgesia was in place in 96.1% of women (n=124), and five women (3.9%) received a pudendal nerve block with local anesthetic agent.
The details pertaining to both malposition and station of the fetal head, at the time of application of Kielland's forceps, are outlined in Table 1. A senior obstetrician (consultant) was present at all of the Kielland's forceps-assisted forceps deliveries, and the vast majority (n=124; 96.1%) were performed by this attending consultant obstetrician. In a small number of cases (n=5; 3.9%), the forceps-assisted delivery was performed by the junior obstetrician (resident–specialist registrar) under the supervision of the attending consultant. For 90.7% (n=117) of these cases. the delivery took place in the delivery suite room with a small proportion (n=12; 9.3%) performed in the operating room.
The mean birth weight of all neonates delivered was 3, 563 g (±44.2, SEM) ranging from 2,140 to 4,620 g. The details in relation to perineal trauma observed at delivery are presented in Table 1. There was one third-degree tear and no fourth-degree tear in the women successfully delivered. There were no cases of cervical laceration and seven cases of extension of high vaginal tears. For postpartum hemorrhage, there were 16 cases (12.4%). For one of these cases, there was massive hemorrhage, which was deemed to be the result of an atonic uterus. This ultimately was managed with pharmacologic uterotonic therapy, examination under anesthesia, and a laparotomy and the patient required 16 units of blood. In this case there was no noteworthy vaginal tear and a hysterectomy was not performed. For urinary incontinence in the immediate postpartum period, there were 10 documented cases (7.8%). There were no cases of fecal incontinence.
Finally, the neonatal details are presented in Table 1, including Apgar scores and cord vessel pH values for the neonates. An Apgar score of less than 7 at 1 minute was recorded in six cases (4.6%), and there were no cases of an Apgar score less than 7 at 5 minutes. Umbilical cord vessel blood gas pH values were recorded in 102 cases (79%) and were missing or unrecorded in 27 cases (21%). An acidotic pH (less than 7.20) was recorded in 13 of 102 cases (12.7%). The rate of admission to the neonatal intensive care unit was 8.5% (n=11). The reasons for admission to the neonatal unit were as follows: neonatal tachypnea, grunting, neonatal pyrexia, to rule out sepsis, or all of these, n=6; low Apgar scores, n=3; hypothermia, n=1; and social admission (to facilitate the mother in high-dependency care as a result of hemorrhage), n=1. There were no cases of neonatal trauma secondary to use of the forceps and no cases of hypoxic–ischemic encephalopathy. There was one case of Erb's palsy deemed to be related to shoulder dystocia.
Of the 15 women for whom the attempt at assisted delivery with Kielland's forceps was unsuccessful, they were all ultimately delivered by cesarean. The reasons for failure were as follows: failed application, n=3; failed rotation, n=6; failed descent with traction, n=5; and unclear documentation n=1. For 10 (66.6%) of these women, the attempt at forceps delivery was performed as a trial in the operating room, and for two of these latter 10 cases, there was an additional attempt at vacuum delivery after the unsuccessful attempt with Kielland's forceps, ie, failed sequential instruments. One of these neonates was admitted to the neonatal intensive care unit with Apgar scores of 3 at 1 minute and 8 at 5 minutes and an umbilical cord blood pH value of 7.04. All other neonates had Apgar scores greater than 7 at 1 and 5 minutes and umbilical cord blood gas pH readings greater than 7.10.
The findings from this study indicate that use of the Kielland's forceps for assisted rotational vaginal delivery, in selected circumstances, in current obstetric practice, with the fetal head mainly at the level of the ischial spines, or below, and in the presence of, or performed by, a senior attending consultant obstetrician, is associated with a high rate of successful vaginal delivery and apparently low rates of maternal and neonatal morbidity. The strengths of this study include the large number of rotational forceps procedures included, that the series is recent in timing, that the data were all entered prospectively at the time of delivery, that the findings were all validated with the medical case notes and, finally, that the maternal and neonatal follow up are complete. The weaknesses of this study are that it is essentially observational and retrospective in nature, is nonrandomized, and the results have all emanated from one tertiary care center.
The first notable observation from this study is that the successful vaginal delivery rate for rotational deliveries in this study, using Kielland's forceps, was 89.6% during the time period 1997–2011. This represents a failure rate of 10.3%, which compares favorably with previously reported failure rates for use of this forceps, which varied between 5.9% and 17.5%, in a tertiary care referral U.K. obstetric unit, during the years 1992–2001.1 For vacuum-assisted deliveries, in comparison, the success rates reported using soft cup instruments have been reported to be between 66% and 79%9,10 and for the more traditional rigid vacuum cup to be between 79% and 81%.9,11 These reported success rates for vacuum-assisted deliveries pertain to cohorts of patients in which the position of the fetal head was occipitoanterior in the majority of cases, ie, less than half of the deliveries were rotational. The data from our study clearly highlight the low failure rate associated with use of the Kielland's forceps for rotational deliveries and outline the established principle that use of the forceps is less likely to fail to achieve a vaginal birth than use of the vacuum extractor.12 However, when failure is encountered, it is recognized that persistent efforts to achieve a vaginal delivery using other instruments is not advisable and may be associated with a higher degree of fetal or maternal injury.9,13 In our case series, such use of sequential instruments was observed in two of the 15 cases of failed instrumental delivery. Finally, in view of the fact that five of the 15 failed delivery attempts in this study occurred in the delivery room, it is our view that a more liberal approach to forceps delivery in the operating room should be recommended.
The other interesting finding from this study is that the neonatal morbidity associated with use of the Kielland's forceps was low. In 1979, analysis of a case series of 86 Kielland's forceps deliveries reported that the neonatal mortality rate was 3.5%, associated birth trauma was 15.1%, and abnormal neonatal neurologic behavior was 23%.7 In 2001, a Scottish study, which included 93 women delivered by Kielland's rotational forceps, reported a 2% risk of physical trauma to the neonate.14 In this study of 129 women similarly delivered, there were no neonatal deaths, 12.7% of neonates demonstrated a cord pH value less than 7.20, and the rate of admission to the neonatal unit was 8.5% for all neonates delivered with use of the Kielland's forceps. There were no cases of forceps-related fetal trauma and no neonates were classified as having hypoxic–ischemic encephalopathy.
It is difficult to speculate, or understand, why the neonatal morbidity, and mortality, associated with the use of this forceps are lower than reported from earlier series7,8 on consideration of the findings presented in this study. There are many possible explanations. Case selection may have improved in current practice. There may be a lower threshold for performing a cesarean delivery in current practice, and hence the forceps deliveries described in this study may have been easier or more amenable for assisted vaginal delivery than earlier reported cases. The presence of a senior and experienced obstetrician at all of these deliveries was likely to have had a positive influence on outcome.
In summary, the data from this study support a continuing use of rotational forceps in current obstetric practice. It is apparent that in well-selected cases, and using the expertise of a senior obstetrician with experience in using the instrument, that it achieves a relatively high rate of successful delivery with low neonatal and maternal morbidity. Although use of this instrument remains controversial, the challenge for the future is that of training junior obstetricians in the skill of rotational forceps after many decades of decreasing use.