- Top of page
A recurrent obstetric problem is the inability to determine the most appropriate delivery method following a prolonged second stage of labor or an abnormal fetal heart rate at full dilatation. The delivery method is often decided in the operating room and, unfortunately, on some occasions after failed attempts at operative vaginal delivery, causing subsequent harm to both the mother and the fetus. The incidence of postpartum intracranial hemorrhage after a failed operative vaginal delivery has been reported as 1 in 334, a rate that is 5.7 times greater than that associated with spontaneous vaginal birth.
The instrument choice for an operative vaginal delivery is also a controversial decision. Forceps and vacuum are typically the instruments available. Recent reports have suggested the importance of training new obstetricians on how to use Kjelland's forceps. One study reported a lower rate of adverse fetal effects with Kjelland's forceps compared with vacuum extraction.
It is widely accepted that an instrumental delivery must only be performed when the clinical head station is at least +0 cm and that mid-forceps deliveries must only be performed by experienced obstetricians. Digitally palpated fetal head station is subjective and unreliable[3, 4]. In contrast, intrapartum transperineal ultrasound (ITU) measurements of the head station during labor correlate with birth progress and outcomes[5-11].
Although two previous studies have evaluated the usefulness of intrapartum ultrasound measurements in vacuum-assisted vaginal deliveries, to our knowledge, no report has evaluated the usefulness of these measurements in cases of instrumental forceps delivery[6, 12]. Our study aimed to test the hypothesis that intrapartum ultrasound measurements may predict complicated forceps delivery.
- Top of page
This observational prospective cohort study included 30 women. All participants provided signed informed consent. The study (PI-1392) was approved by the local Ethics and Research Committees. All ultrasound measurements were performed by the same person (M.J.C.). The managing obstetricians were blinded to the results of the sonographic measurements. Intrapartum ultrasound was not performed if the fetal scalp blood testing pH was < 7.21, or in cases of prolonged fetal bradycardia or late fetal heart-rate decelerations with absent fetal heart-rate variability.
Sample size was calculated using the results of a pilot study that considered the angle of progression (AOP) as the main outcome variable. In the pilot study, four women were included in each group. The pilot study subjects were not included in the final study. Using an alpha value of 0.05 and a power (1 – beta) of 80% for a mean difference of 10° between the complicated and uncomplicated forceps deliveries, and a common standard deviation of ± 7.1°, eight women were needed in each arm of the study. Consecutive cases indicated for operative forceps delivery were eligible for the study. None of the patients declined to participate in the study. After eight women with complicated forceps deliveries were included in the study, we continued the recruitment for a total of 30 patients to account for secondary variable analysis.
Inclusion criteria were indication for instrumental forceps delivery at term with an otherwise uncomplicated pregnancy, prolonged second stage of labor (i.e. at least 2 h) and epidural anesthesia. ITU was performed with the woman in a semirecumbent position, with an empty bladder and ruptured membranes. Fetuses in the occiput posterior position were excluded, as the occiput posterior position has been defined as an independent risk factor for perineal tears[13, 14].
The following parameters were assessed by ITU: AOP, progression distance (PD) and head direction (HD) between contractions (AOP1, PD1 and HD1, respectively); AOP, PD and HD with contraction and active pushing (AOP2, PD2 and HD2, respectively); and the difference between AOP1 and AOP2, PD1 and PD2 and HD1 and HD2 (dAOP, dPD and dHD, respectively).
ITU measurements were performed according to previously published methods[6, 8, 10, 15] with a curved array transducer (ProSound 6; Aloka, Tokyo, Japan). AOP is defined as the angle between a line through the midline of the pubic symphysis and another line from the anterior margin of the pubic symphysis to the leading edge of the fetal head. PD is defined as the distance between the infrapubic line (the line through the inferior margin of the pubic symphysis perpendicular to the long axis of the symphysis) and a parallel line through the deepest bony part of the fetal head. HD is defined as the angle between a line perpendicular to the widest diameter of the fetal head and a line parallel to the line through the midline of the pubic symphysis.
Numerous clinical, ultrasound and forceps-extraction characteristics were recorded, including newborn weight, sex and gestational age and maternal age, height, body mass index (BMI), obstetric history, hemoglobin level, need for episiotomy, vaginal or anal sphincter tear (using Sultan's classification of perineal tears), digital assessment of head station and position between contractions, number of tractions performed, utilization of fundal pressure, time between ITU and delivery, time spent performing ITU and measurements, time between application of forceps and delivery, subjective impression of the operator regarding use of forceps (i.e. easy, medium, difficult or failed), cord arterial pH, Apgar scores at the first and fifth minutes, presence of marks produced by the forceps on the head of fetus and the presence of caput succedaneum.
The use of forceps for each case was classified as complicated or uncomplicated, depending on the performance characteristics and the maternal consequences. A forceps delivery was classified as complicated when one or more of the following situations occurred: three or more tractions were performed, difficult or failed forceps delivery (according to the subjective impression of the operator), a third-degree or higher tear according to Sultan's classification of perineal tears, significant bleeding during the episiotomy repair, major tear reported by the obstetrician and confirmed by a drop in the hemoglobin level of ≥ 2.5 g/dL following the delivery or a significant traumatic neonatal lesion.
The distribution of variables was verified using the Kolmogorov–Smirnov test and histograms. As all numeric variables were normally distributed, they were displayed as the mean and SD. Between-group comparisons were performed using the Student's t-test. Categorical variables were shown as number and percentage. Individual predictive capabilities were evaluated using the receiver–operating characteristics (ROC) curve and the area under the curve (AUC). A conditional binary logistic regression was used to find the best predictive models for complicated forceps delivery. The level of significance was set at 95% (P < 0.05).
- Top of page
Kjelland's forceps were used in all cases. The majority of women (28/30, 93.3%) were primiparous. The fetal head position was left occiput transverse in three (10%) cases, right occiput transverse in one (3.3%), left occiput anterior in 14 (46.7%), right occiput anterior in nine (30%) and occiput anterior in three (10%). All fetuses were delivered in the occiput anterior position.
An episiotomy was performed in 90% (27/30) of cases and a third degree or greater perineal tear occurred in 10% (three of 30); there were no fourth-degree tears. In five (16.7%) cases, significant bleeding occurred during the episiotomy repair and a tear was reported and confirmed by a drop of ≥ 2.5 g/dL in the hemoglobin level. Three or more tractions were performed in five (16.7%) cases. Fundal pressure was utilized in 19 (63.3%) deliveries. Only two (6.7%) of the 30 operative deliveries were considered difficult by the operator; in one (3.3%) case, forceps delivery failed and a Cesarean section delivery was performed. The time between the forceps application and delivery was 1.50 ± 1.56 min. In total, nine forceps deliveries were considered to be complicated (according to our classification) and none was classified as complicated solely because of the subjective impression of the obstetrician or significant neonatal lesions.
The mean birth weight was 3099 ± 473 g. In two (6.7%) cases, the Apgar score was < 7 at 1 min, and no Apgar scores of < 7 were obtained at 5 min. The mean umbilical arterial blood pH was 7.24 ± 0.08; this measure was < 7.20 in five (16.7%) cases and < 7.10 in two (6.7%). Forceps marks were apparent in four (13.3%) newborns. A mild-to-moderate caput succedaneum was present in all neonates and no significant traumatic lesions were observed.
AOP1, PD1, HD1, AOP2, PD2 and HD2 assessments were possible in all cases. The digitally assessed head station was at least +2 cm in all cases. The mean time from ITU to delivery was 5.6 ± 3.7 min. For the case in which the forceps failed and a Cesarean section delivery was performed, the time from ITU to delivery was 20 min. The mean time for performing the ITU and measurements was 1.6 ± 0.6 min.
There were no significant differences between the uncomplicated forceps delivery group and the complicated forceps delivery group regarding the maternal or neonatal demographic data, except for the gender of the newborn (Table 1). There were also no significant between-group differences with regard to presence of caput succedaneum, head position or timing, except for time between forceps application and delivery time (P = 0.01).
Table 1. Maternal and neonatal demographic data in 30 pregnancies requiring forceps delivery
|Variable||Uncomplicated forceps delivery(n = 21)||Complicatedforceps delivery (n = 9)||P|
|Age (years)||33.5 ± 4.0||30.1 ± 6.9||0.103|
|Height (cm)||161 ± 6 ||158 ± 3 ||0.190|
|BMI||27.4 ± 3.8||26.9 ± 2.9||0.172|
|GA (weeks)||39.5 ± 1.4||39.4 ± 1.3||0.952|
|Birth weight (g)||3050 ± 408||3212 ± 611||0.399|
|Gender|| || ||0.025|
| || || || |
|1 min|| 8.5 ± 0.8|| 7.9 ± 1.2||0.097|
|5 min|| 9.4 ± 0.7|| 9.2 ± 0.7||0.556|
|UCA pH|| 7.26 ± 0.07|| 7.21 ± 0.08||0.149|
Significant differences were observed between the uncomplicated and complicated cases with regard to AOP1, PD1, AOP2, PD2, HD2, dPD, dHD and digitally assessed head station. There were no significant differences regarding HD1 or dAOP. The average AOP1 measurement was significantly higher in the uncomplicated group (145 ± 6°) compared with the complicated group (125 ± 8°). The mean PD1 was 5.1 ± 0.1 cm in the uncomplicated group and 3.5 ± 0.2 cm in the complicated group. The AOP2 (156 ± 7° vs 142 ± 12°), PD2 (5.7 ± 0.6 cm vs 4.6 ± 0.8 cm), HD2 (35 ± 12° vs 21 ± 9°) and dHD (8 ± 10° vs −2 ± 13°) measurements were also significantly higher in the uncomplicated group compared with the complicated group. The digitally assessed head station was 2.7 ± 0.6 cm in the uncomplicated group and 2.1 ± 0.3 cm in the complicated group (Table 2) (Figure 1).
Table 2. Intrapartum ultrasound (ITU) data in 30 pregnancies requiring forceps delivery
|Variable||Uncomplicatedforceps delivery(n = 21)||Complicatedforceps delivery(n = 9)||P|
|AOP (°)|| || || |
|AOP1||145 ± 6||125 ± 8||0.0000001|
|AOP2||156 ± 7||142 ± 12 ||0.0004|
|dAOP||11 ± 7||17 ± 12||0.089|
|PD (cm)|| || || |
|PD1||5.1 ± 0.1||3.5 ± 0.2||0.00000004|
|PD2||5.7 ± 0.6||4.6 ± 0.8||0.0002|
|dPD||0.6 ± 0.6||1.2 ± 0.8||0.029|
|HD (°)|| || || |
|HD1||28 ± 12||23 ± 16||0.394|
|HD2||35 ± 12||21 ± 9||0.005|
|dHD||8 ± 10||−2 ± 13||0.035|
|Dig. station||2.7 ± 0.6||2.1 ± 0.3||0.0057|
Figure 1. Angle of progression (AOP) measurements in uncomplicated (Cases 1–21) and complicated (Cases 22–30) forceps deliveries. AOP1, angle of progression between contractions; AOP2, angle of progression during contraction and active pushing.
Download figure to PowerPoint
The greatest AUCs for the different measurements were observed in AOP1 (98.9%) and PD1 (98.4%). AUCs for HD1 (61.4%) and digital head station (79.9%) were lower (Figure 2a). AUCs were 84.7% for AOP2, 82% for HD2 and 89.2% for PD2 (Figure 2b). For dAOP, dPD and dHD, AUC was > 50 for dHD only (73%). The best cut-offs to predict a complicated forceps delivery were an AOP of 138° (sensitivity = 85.7%, specificity = 100%) and a PD1 of 4.8 cm (sensitivity = 85.7%, specificity = 100%).
Both AOP1 and PD1 showed a strong linear correlation (R2 = 0.930) (Figure 3), whereas a weaker linear correlation was observed for PD2 and AOP2 (R2 = 0.703).
Figure 3. Linear regression of intrapartum ultrasound measurements of angle of progression between contractions (AOP1) and progression distance between contractions (PD1) (R2 = 0.930).
Download figure to PowerPoint
Several logistic regression models were built to predict the likelihood of a complicated forceps delivery. The best model used AOP1 and HD2 (R2 = 0.705 (Cox and Snell)):
The best model using only one variable was for AOP1 (R2 = 0.632 (Cox and Snell)):
- Top of page
Our findings demonstrate that ITU is helpful for predicting complicated operative forceps delivery of fetuses in the occiput anterior position. Measurements of AOP, HD and PD are easily obtained by ITU, and a high concordance between observers has been demonstrated for both two-dimensional (2D) and three-dimensional (3D) ultrasound systems[5, 10, 11]. In our study, we used a 2D ultrasound system and the cases were studied prospectively. It took under 5 min (1–4 min) to perform the ITU examination and measurements. The AOP1, AOP2, PD1, PD2 and HD2 measurements were statistically associated with complicated operative forceps delivery.
One study has suggested the HD2 measurement as a good predictor of a successful operative vacuum delivery, dividing this measurement into categories of more than 30° (head-up sign), 0–30° and < 0°. In that study, the head-up sign was associated with vacuum extractions rated as simple or moderate based on the number of tractions and on the subjective assessment of the operator who performed the ITU. Another study suggested that HD2, dAOP and AOP2 measurements were predictive of a vacuum delivery. The ITU measurements were included in the clinical decision and were studied retrospectively, comparing a group of 15 vacuum deliveries with five Cesarean section deliveries. Although all 15 vacuum deliveries presented with the head-up sign, none of the Cesarean deliveries presented in such a manner (in four of the five Cesarean deliveries, the fetuses were in occiput posterior position). The AOP2 and dAOP measurements were also higher for the vacuum deliveries.
In our study of non-occiput posterior deliveries, HD1 was not correlated with complicated forceps delivery, but HD2 and dHD were. There is evidence that dHD is greatest at head stations that may correlate with a simple operative vaginal delivery. In our cohort, AOP1 and the combination of AOP1 and HD2 correlated best with complicated forceps delivery.
It is widely accepted that digital palpation of fetal head station is unreliable[3, 4]. According to Tutschek et al., an AOP of 138° represents a head station of +2 cm5. In our study, digital assessment of head station was at least +2 cm in all cases; however, AOP1 for complicated forceps deliveries was 125 ± 8°, corresponding to a head station of +0.8 cm.
Forceps deliveries were classified as complicated and uncomplicated, giving the highest relevance to the possible maternal damage. The only neonatal variable included was the presence of significant neonatal lesions. The number of tractions performed is related to the difficulty and strength required to perform the extraction. Although there is no consensus regarding the appropriate number of tractions, classically, a limit of two tractions has been established. In our opinion, no more than three tractions should be performed and a forceps delivery can be considered as complicated when more than two tractions are required. Following bleeding deemed significant by the obstetrician, a drop in hemoglobin of ≥ 2.5 g/dL is a cut-off that has been used to define profuse bleeding. A perineal tear of the third degree or higher is considered to be significant damage and is associated with a longer, more complicated, postpartum recovery. The only non-objective measure used was the operator's subjective assessment; however, no cases were classified as complicated solely based on this measure.
In our study, nearly one-third of the operative forceps deliveries were classified as complicated, which was a higher proportion than expected.
The main goal of our study was to examine which of the sonographic markers of head descent and birth progress could predict complicated forceps delivery. An AOP of 138° had sensitivity of 85.7% and specificity of 100% for predicting a complicated forceps delivery. The alternative options to operative forceps delivery are Cesarean section at full cervical dilatation, vacuum extraction or prolonged second stage of labor. Cesarean section at full cervical dilatation is associated with a high rate of both intraoperative complications and significant blood loss requiring blood transfusion. Vacuum extraction is not free of complications and prolonging the second stage of labor has been associated with postpartum hemorrhage and pelvic floor damage. As every alternative has a significant rate of maternal complications, it is reasonable to set a low threshold for performing an operative forceps delivery, corresponding to the low rate of maternal complications. In our opinion, the threshold can be set at an expected level of 10% for forceps deliveries classified as complicated. According to our model, an AOP1 of 137.4° predicts a 10% rate of complicated forceps delivery, and we suggest using this value as a cut-off for performing operative forceps delivery.
Fetuses in occiput posterior position were excluded because the occiput posterior position has been described as an independent risk factor for perineal tears[13, 14]. Further study of this patient group is warranted to evaluate the capability of ITU measurements for predicting deliveries in occiput posterior position.
In conclusion, sonographic intrapartum measurements may be helpful in predicting a safe instrumental forceps delivery and in clinical decision-making. Higher head stations, as determined by lower AOP or PD measurements, and an unfavorable HD during a contraction, are more likely to result in complicated forceps delivery. In our cohort, a combination of AOP1 and HD2 yielded the best predictive value, and as a single parameter AOP1 was the best predictor. We believe that a single measurement of head station may be sufficient, and according to the existing data, this measure should be the AOP[3, 5, 8, 11, 13].