Schlüsselwörter
Vakuumentbindung - Vakuumextraktion - vakuumbedingte unerwünschte Outcomes - vakuumbedingte
mütterliche Komplikationen - vakuumbedingte neonatale Komplikationen
Keywords
vacuum-assisted delivery - vacuum extraction - adverse outcomes of vacuum delivery
- maternal complications of vacuum delivery - neonatal complications of vacuum delivery
Introduction
Vacuum extraction (VE) is a common modality for delivery worldwide; used in 5.9% of
the deliveries in Israel [1], and in up to 13% of the
deliveries in the United Kingdom [2]. It is an important component of modern obstetric care and can be used to avoid
cesarean delivery
(CD) during the second stage of labor in times of fetal distress, arrest of descent
or other maternal indications that require shortening the second stage [3].
Under the appropriate circumstances, VE is the preferred modality compared to CD,
because it can often be accomplished more quickly. and CD is also associated with
short- and long-term
maternal morbidities, such as extensive hemorrhage, infection, prolonged healing,
repeat CD, uterine rupture and risk of placental abnormalities such as placenta accreta
[3]
[4]
[5].
However, VE is also a risk factor for several maternal morbidities compared to second
stage CD, such as third- or fourth-degree perineal lacerations [6], with an adjusted odds ratio of up to 13.9–14.9 [7]. Perineal lacerations can cause permanent damage to the anal
sphincter and result in lifetime disability [8]
[9]. At times, VE can also lead to
excessive maternal blood loss during delivery, just as CD does [10].
As for neonatal complications, VE compared to intrapartum CD, has been linked to several
neonatal injuries related to birth trauma, such as clavicular fracture [11], humerus fracture and Erb’s palsy [12]
[13]. It has also been specifically associated with neonatal head injuries, which have
long-term implications such as subgaleal hematoma (SGH), intracranial hemorrhage and
skull fracture [14]
[15].
VE is also known as a risk factor for shoulder dystocia [16], a life-threatening situation in which, even if a successful extraction is
achieved, some neonates will incur permanent disability due to brachial plexus nerve
injury.
Many considerations affect a physician’s decision to perform VE or CD during the second
stage of labor. At times, looking at the severe complications sustained by the mother
or the newborn
due to VE, many physicians wish they would have chosen CD, even when conditions for
performing VE met the American College of Obstetricians and Gynecologists (ACOG) guidelines
[3] and the procedure was not contraindicated.
Previous studies have evaluated risk factors for maternal complications during VE,
others have found risk factors for neonatal complications associated with VE. None
of the studies to date
have examined potential risk factors for the combined outcome of serious maternal
and neonatal complications.
In addition, since non-metal vacuum cups are not used worldwide, maternal and fetal
adverse outcomes have not been fully explored in large cohorts.
The aim of the current study, was to explore VE performed using a non-metal cup and
to determine potential risk factors for maternal or neonatal adverse outcomes that in retrospect
would have led the physician to avoid the procedure. We aimed to recognize unfavorable
conditions for VE in which a thorough discussion regarding mode of delivery is needed.
Materials and Methods
This retrospective cohort study included women delivered by means of a non-metal cup
VE, January 2014 to August 2019, in a tertiary care medical center. All VE were performed
at ≥ 34 weeks of
gestation. Additional inclusion criteria were singleton pregnancies, without known
genetic or structural anomalies.
We divided our cohort into two groups. The first group included deliveries with VE-related
feto-maternal adverse outcomes, defined as one or more of the following: third- or
fourth-degree
perineal laceration, SGH, intracranial hemorrhage, shoulder dystocia, clavicular fracture,
Erb’s palsy, fractures of humerus (VE adverse outcomes group). The second group included
deliveries
without any VE-related feto-maternal adverse outcomes (control group). We compared
the two groups in terms of basic maternal, labor and delivery characteristics to determine
potential risk
factors for VE-related adverse outcomes.
We excluded all cases of failed vacuum, since some of the outcomes included in the
composite adverse outcome defined were necessarily related to the extraction of the
body of the fetus
through the birth canal, and also to assure that all adverse outcomes described were
associated only with VE, and not with a subsequent forceps or CD.
Each VE was carried out by a senior physician who performed a full evaluation before
the procedure and assured the conditions met ACOG guidelines [17]. Either a Ventouse-Mityvac (VM) or Kiwi Omnicup vacuum cup were used. The type of
vacuum cup was selected by the physician. A pediatrician was present at every VE.
After delivery, the performing physician completed a detailed electronic report regarding
assessment of the labor pattern before and during the procedure.
Data collection
Data were retrieved using the electronic maternal database of the delivery room, then
crossed-tabulated with data from the Neonatal Unit and the Neonatal Intensive Care
Unit (NICU). All
medical records were reviewed manually to complete missing data. Data collected included:
-
Maternal demographics: age, body mass index (BMI), gravidity, parity, gestational
age at delivery, diabetes (pre-gestational or gestational) [18] and history of CD.
-
Maternal outcomes: rates of third- or fourth-degree perineal lacerations [19] and maternal blood loss during delivery.
-
Labor and delivery characteristics: use of epidural anesthesia, intrapartum fever,
duration of first, second and third stages of labor, indication for VE, fetal head
station and
position at time of VE, vacuum cup type, vacuum duration and the presence of cup detachments.
-
Neonatal characteristics and outcomes: fetal weight, NICU hospitalization and VE-related
neonatal adverse outcomes (SGH, intracranial hemorrhage, shoulder dystocia, Erb’s
palsy,
fracture of humerus or clavicle).
Indications for VE were categorized as:
-
Non-reassuring fetal heart rate [20];
-
Prolonged second stage [21] and
-
Maternal indications, including medical background requiring shortening second stage
or maternal exhaustion.
For historical reasons, fetal head station was defined by thirds from ischial spines
−3 to +3, and was divided into Mid-pelvis: S+1, Low: S+2, and Outlet: S+3 and below.
VE was performed
according to ACOG guidelines.
Neonatal diagnoses were determined by the senior pediatrician present at VE and during
neonatal hospitalization, according to international standards and relevant imaging.
Statistical analysis
Comparison of continuous variables between groups was performed using t-test. Categorical
data were compared using Chi-square or Fisher’s exact test, each when appropriate.
Multivariate
logistic regression and adjusted odds ratios were calculated to examine variables
that had an independent effect on severe VE complications. A probability value of
< 0.05 was considered
significant. All analyses were performed using SPSS-25 software (IBM, Armonk, NY,
USA).
Results
During the study period, 33889 women delivered vaginally in our institution, of which
3410 had a successful VE (9.8%). A total of 3331 women met the inclusion criteria
and were included in
the study ([Fig. 1]).
Fig. 1
Flowchart describing the study population.
Overall, 263 (7.9%) deliveries had a maternal or neonatal adverse outcome related
to VE, defined as at least one of the following: third- or fourth-degree perineal
laceration, SGH,
intracranial hemorrhage, shoulder dystocia, clavicular fracture, Erb’s palsy and fracture
of humerus (VE adverse outcome group), whereas 3068 (92.1%) did not have VE-related
adverse outcomes
(control group). The characteristics of the VE adverse outcome group are described
in [Table 1].
Table 1
Characteristics of vacuum extraction (VE) adverse outcome group, N = 263.
|
Injury
|
Incidence among VE adverse outcome group
|
|
Third- or fourth-degree perineal laceration, N (%)
|
66 (25.1%)
|
|
Subgaleal hematoma, N (%)
|
154 (58.6%)
|
|
Intracranial hemorrhage, N (%)
|
1 (0.4%)
|
|
Shoulder dystocia, N (%)
|
22 (8.4%)
|
|
Clavicular fracture, N (%)
|
23 (8.7%)
|
|
Erb’s palsy, N (%)
|
14 (5.3%)
|
|
Fracture of humerus, N (%)
|
1 (0.4%)
|
|
Multiple injuries (more than one), N (%)
|
19 (7.2%)
|
We compared the two groups in terms of basic maternal, and labor and delivery characteristics
to determine potential risk factors for VE-related feto-maternal adverse outcomes.
Maternal characteristics
Women in the VE adverse outcome group were statistically younger than those without
adverse outcomes (29.3 ± 4.5 vs. 30.3 ± 5.2 years, respectively; p = 0.004), but this
difference is not
clinically significant ([Table 2]). More were nulliparous (80.6% vs. 71.8%, p = 0.002).
Table 2
Baseline characteristics of women with or without feto-maternal adverse outcomes related
to vacuum extraction (VE).
|
Variable
|
VE adverse outcome group (n = 263)
|
Control group (n = 3068)
|
P value
|
|
SD = standard deviation
|
|
Maternal age, years (mean ± SD; median
|
29.3 ± 4.5; 29
|
30.3 ± 5.2; 30
|
0.004
|
|
Maternal BMI ± SD
|
23.6 ± 4.6
|
23.3 ± 4.9
|
0.643
|
|
Parity ± SD
|
0.3 ± 0.7
|
0.4 ± 0.9
|
0.023
|
|
Nulliparous (N, %)
|
212 (80.6%)
|
2202 (71.8%)
|
0.002
|
|
Previous cesarean delivery, N (%)
|
9 (3.4%)
|
244 (8.0%)
|
0.008
|
|
Gestational age [days, weeks]
Mean ± SD; median
|
278 (39.7 w) ± 8.8 (1.3 w); 279
|
277.3 (39.6 w) ± 9.4 (1.3 w); 279
|
0.193
|
|
Gestational or pregestational DM, N (%)
|
23 (8.7%)
|
216 (7.0%)
|
0.304
|
Previous CD was less common in the VE adverse outcome group (3.4% vs. 8.0%, p = 0.008).
No differences were found between the groups regarding maternal BMI, gestational age
at delivery, and pre-gestational or gestational diabetes mellitus rates.
Labor and delivery characteristics
The use of epidural anesthesia was more common in the VE adverse outcome group (29.3%
vs. 18.1%, p < 0.001). Rates of intrapartum fever were also higher (12.9% vs. 8.3%,
p = 0.011; [Table 3]).
Table 3
Labor and delivery characteristics of women with or without feto-maternal adverse
outcomes related to vacuum extraction (VE).
|
VE adverse outcome group (N = 263)
|
Control group (N = 3068)
|
P value
|
|
SD = standard deviation; VM = Ventouse-Mityvac; NRFHR = Non-reassuring fetal heart
rate
|
|
Epidural (N %)
|
77 (29.3%)
|
555 (18.1%)
|
< 0.001
|
|
Intrapartum fever (N, %)
|
34 (12.9%)
|
255 (8.3%)
|
0.011
|
|
Vacuum type
|
Kiwi
|
116 (45.7%)
|
1907 (65.4%)
|
< 0.001
|
|
VM
|
138 (54.3%)
|
1007 (34.6%)
|
|
|
Procedure duration (min)
|
Mean ± SD; median
|
6.9 ± 5.8; 5
|
5.0 ± 3.3; 5
|
< 0.001
|
|
Vacuum detachment (N, %)
|
78 (30.7%)
|
571 (19.6%)
|
< 0.001
|
|
Fetal position OA (N, %)
|
180 (73.5%)
|
2307 (81.0%)
|
0.004
|
|
Head station at VE (N, %)
|
Mid-pelvis
|
154 (63.1%)
|
1585 (56.2%)
|
0.038
|
|
Low
|
89 (36.5%)
|
1159 (41.1%)
|
0.160
|
|
Outlet
|
1 (0.4%)
|
74 (2.6%)
|
0.031
|
|
Vacuum indication (N, %)
|
NRFHR
|
151 (61.4%)
|
2194 (76.7%)
|
< 0.001
|
|
Prolonged second stage
|
89 (36.2%)
|
622 (21.7%)
|
|
|
Maternal indication
|
6 (2.4%)
|
46 (1.6%)
|
|
|
Labor duration, minutes, mean, SD
|
First stage
|
600.6 ± 249.5
|
645.1 ± 429.5
|
0.631
|
|
Second stage
|
173.7 ± 79.6
|
135.7 ± 81.8
|
< 0.001
|
|
Third stage
|
9.2 ± 6.9
|
9.5 ± 6.9
|
0.451
|
|
Neonatal birth weight, g
|
Mean ± SD
|
3419.1 ± 460.3
|
3169.6 ± 433.6
|
< 0.001
|
|
Delivery blood loss, ml
|
Mean ± SD
|
221 ± 300
|
333.6 ± 200.6
|
0.021
|
|
NICU admission (N, %)
|
11
|
4.2%
|
85
|
2.8%
|
0.189
|
The use of VM cup type was more common in deliveries with VE adverse outcomes (54.3%
vs. 34.6%, p < 0.001). Procedure duration in these deliveries was longer (6.9 min
± 5.8 vs.
5.0 ± 3.3, p < 0.001), and vacuum cup detachment occurred more often (30.7% vs. 19.6%,
p < 0.001).
Occiput anterior (OA) position was less common in the VE adverse outcome group (73.5%
vs. 81.0%, p = 0.004).
Fetal head stations at VE were higher in the VE adverse outcome group. In both groups,
most of the fetuses were at S+1 (midpelvis) when VE was initiated, fewer (37.5%) were
at S+2 (low) and
the minority (2.3%) were at S+3 and below (outlet). Yet, we found a relatively higher
portion of midpelvis station in the VE adverse outcome group (63.1% vs. 56.2%, p = 0.038)
and fewer at
outlet station (0.4% vs. 2.6%, p = 0.031).
Non-reassuring fetal heart rate was the most common indication for VE, but its rate
was relatively low in the VE adverse outcome group (61.4% vs. 76.7%, p < 0.001). Prolonged
second
stage was the second most common indication, and its rate was relatively high in the
VE adverse outcome group (36.2% vs. 21.7%, p < 0.001).
Second stage of labor was longer in the VE adverse outcome group (173.7 min ± 79.6
vs. 135.7 min ± 81.8, p < 0.001).
Neonatal birth weight was significantly higher in the VE adverse outcome group (3419.1 g
± 460.3 vs. 3169.6 g ± 433.6, p < 0.001).
Interestingly, blood loss during delivery was lower in the VE adverse outcome group
(221 ml ± 300 vs. 333.6 ml ± 200.6, p = 0.021).
No differences were noted between the groups regarding the duration of first and third
stages of labor and the rates of NICU admission.
Logistic regression
Logistic regression indicated seven independent risk factors for VE-related feto-maternal
adverse outcomes ([Table 4]). These included
nulliparity (adjusted odds ratio [OR] 1.82 (95% CI 1.11–2.98, p = 0.018), use of epidural
anesthesia (OR 1.99, 95% CI 1.42–2.80, p < 0.001), use of VM cup (OR 1.86, 95% CI
1.35–2.54,
p < 0.001), prolonged second stage as indication for VE (OR 1.54, 95% CI 1.11–2.15,
p = 0.010), vacuum cup detachment (OR 1.66, 95% CI 1.18–2.34, p = 0.004), increasing
procedure duration
(OR 1.07 for every additional minute 95% CI 1.03–1.11, p < 0.001) and increasing neonatal
birthweight (OR 3.42 for every additional kg, 95% CI 2.33–5.02, p < 0.001). OA position
was
found to be a protective factor against severe VE complications (OR 0.62, 95% CI 0.43–0.89,
p = 0.010).
Table 4
Multivariable logistic regression analysis of factors associated with VE-related feto-maternal
adverse outcomes.
|
Variable**
|
Adjusted Odds Ratio (OR)
|
95% CI
|
P value
|
|
* R square for procedure characteristics (duration and cup detachment): 0.041
** R square for all variables: 0.152
|
|
Nulliparity
|
1.82
|
1.11–2.98
|
0.018
|
|
Epidural anesthesia
|
1.99
|
1.42–2.80
|
< 0.001
|
|
Vacuum cup type – Ventouse-Mityvac
|
1.86
|
1.35–2.54
|
< 0.001
|
|
Prolonged 2nd stage as VE indication
|
1.54
|
1.11–2.15
|
0.010
|
|
OA position
|
0.62
|
0.43–0.89
|
0.010
|
|
Vacuum cup detachment*
|
1.66
|
1.18–2.34
|
0.004
|
|
Procedure duration (each additional minute)*
|
1.07
|
1.03–1.11
|
< 0.001
|
|
Increasing neonatal BW (each additional kg)
|
3.42
|
2.33–5.02
|
< 0.001
|
Maternal age, history of CD, intrapartum fever and fetal head station at VE were not
significant in terms of VE adverse outcomes ([Table 4],
[Table 5]).
Table 5
Multivariable logistic regression analysis of factors unassociated with VE-related
feto-maternal adverse outcomes.
|
Variable
|
Adjusted Odds Ratio (OR)
|
95% CI
|
P value
|
|
Maternal age
|
0.99
|
0.96–1.02
|
0.422
|
|
History of cesarean delivery
|
1.23
|
0.51–2.99
|
0.643
|
|
Intrapartum fever
|
1.14
|
0.71–1.82
|
0.586
|
|
Fetal head station before VE
|
0.91
|
0.71–1.82
|
0.583
|
Failed VE
A total of 47 women had a failed VE and were excluded from the study. Their basic
maternal characteristics of age, parity, nulliparity, previous CD, gestational age
at delivery, gestational
diabetes mellitus and maternal BMI were similar to those included in the study, but
their labor and delivery characteristics differed significantly. Women who experienced
failed VE had a
lower percentage of OA position (47.6% vs. 74.7%, p < 0.001), higher head station
at the beginning of VE (82.6% vs. 56.7% at midpelvis, p = 0.004) and a higher rate
of prolonged second
stage as the indication for VE (58.7% vs. 22.9%, p < 0.001). Mean neonatal BW in the
failed VE group was higher than in the successful VE group (3429 ± 422 g vs. 3189 ± 441 g,
respectively; p < 0.001). Rates of use of each vacuum cup type were similar in both
groups.
Discussion
The purpose of this study was to determine potential risk factors for adverse maternal
or neonatal outcomes related to VE, including third- or fourth-degree perineal laceration,
SGH,
intracranial hemorrhage, shoulder dystocia, clavicular fracture, Erb’s palsy and fracture
of humerus.
While previous studies explored neonatal complications and maternal complications
separately, we chose to combine these complications under the same composite outcome,
as they all have
long-term implications and should be considered when choosing to perform VE. By doing
so, we aimed to help physicians avoid performing VE under unfavorable conditions that
could result in
permanent injury to the mother or the newborn, and consider CD instead. This study
is also the first large cohort focusing on adverse outcomes related to non-metal vacuum
cups.
We found seven independent risk factors for VE-related feto-maternal adverse outcomes:
nulliparity, epidural anesthesia, use of VM cup, prolonged second stage as indication
for VE, vacuum cup
detachment, increasing procedure duration and greater neonatal weight. OA position
was a protective factor against VE adverse outcomes.
Our findings agree with those of previous studies regarding risk factors for SGH and
intracranial hemorrhage [22]
[23]
[24], and risk factors for shoulder dystocia or neonatal birth trauma [16]
[25]
[26]. Third- or
fourth-degree perineal lacerations during VE were reported to be associated with nulliparity
and increasing neonatal weight [24]
[27]. Our study results agree with those of previous reports.
Our findings may be explained by the logic assumption that more difficult VEs increase
the amount of force exerted on the mother’s pelvis and on the fetus, and thus, cause
greater trauma to
both.
The use of the VM mushroom-shaped cup was a risk factor for severe VE complications
in our cohort, compared to the Kiwi-OmniCup. These findings were not reported in a
previous study conducted
on a smaller cohort [28], but it might have been underpowered for these rare outcomes.
Of note, the overall rate of SGH in our cohort was 4.6% (154/3331). This prevalence
is similar to previously published studies [29]
[30].
The worldwide consensus is that VE is the preferred modality of delivery for the mother
compared to CD during the second stage of labor, as it is associated with lower rates
of maternal
morbidity and mortality [7]
[31]. CD is also associated with increased risks for
fertility and future pregnancy [32]. In 2014, the ACOG recommended encouraging operative vaginal delivery as a strategy
to reduce the
rates of CD [21]. However, a recent study raised the concern that encouraging higher rates of VE
could result in increases in severe
perinatal and maternal morbidity [32].
When exploring the published data, we found that severe maternal complications during
CD are relatively rare, with an incidence of about 0.6–2.7% [33]
[34]. As for long-term outcomes of CD, morbidity is also rare, and there are even some
beneficial maternal effects
to CD compared to vaginal deliveries, such as reduced rates of pelvic organ prolapse
[32]. On the other hand, the risk for third- or
fourth-degree perineal lacerations during VE is about 4.5–6.5% [35]. Of these, 61% of women will remain with symptoms that impair their
quality of life [9]. Thus, in the presence of unfavorable conditions, we believe that one should not
try to perform a VE at any cost.
VE is considered beneficial to the newborn compared to second stage CD because it
is quicker in terms of fetal extraction; thus lowering the risk for hypoxic damage
when fetal distress is
suspected during delivery [36]. Yet, data comparing these two modalities are conflicting, with one study reporting
higher rates of
neonatal adverse outcomes with CD [37], and another showing higher rates with VE [38].
Although an attempt for quick extraction to avoid hypoxic ischemic damage is understood,
fetal distress is not the only indication for VE. Many VEs are performed for other
indications, such
as prolonged second stage or maternal conditions requiring shortening the second stage
[39]. In these situations, time from decision to
extraction has minimal importance; thus, VE loses its potential beneficial effect
for the fetus compared to CD. Moreover, it seems that VEs performed specifically for
prolonged second stage
are associated with higher rates of severe maternal and perinatal morbidity/mortality
compared to CD [7]
[38]. This makes sense because the indication itself implies feto-pelvic dystocia. Our
study also found prolonged second stage as indication
for VE to be a risk factor for severe complications. Thus, when considering VE due
to prolonged second stage or due to maternal indications, an informed decision should
be made based on the
risks versus benefits. Intrapartum trans-perineal ultrasound or digital feedback might
also help predict an upcoming complicated VE and assist in the decision-making process
[40]
[41].
Interestingly, although previous reports have suggested that midpelvic VE places the
neonate at risk for adverse outcomes compared to CD [31]
[42], our study found that head station was not a significant risk factor for VE-related
adverse outcomes.
Another aspect to consider are the similarities found between the characteristics
of deliveries that resulted in VE-related adverse outcomes and the failed VE population.
Both had relatively
higher neonatal birthweights, lower rates of OA position, higher fetal head stations
and higher rates of prolonged second stage as the indication for VE. Previous studies
also found these
characteristics as risk factors for failed VE [43]
[44]. As is known, failed VE can
result in high morbidity for the mother and the newborn [44]
[45], which should be
considered when deciding whether to perform VE under unfavorable conditions or CD.
Strengths and limitations
The strengths of our study are that it included a large, homogenous cohort. Data were
retrieved from a single institution with a strict protocol for VE. Only successful
procedures using a
non-metal cup were included.
The main limitation of our study is its methodological design as a retrospective cohort
study. Relevant information, such as the severity of maternal and neonatal injuries,
and their
long-term outcomes, such as fecal incontinence and neurological deficits are lacking.
However, we assume that the rates of permanent damage due to VE-related adverse outcomes
are similar to
those reported in the literature; thus, enabling an informed decision based on our
study findings. Data regarding whether the onset of labor was induced or spontaneous
are also missing, but
we believe this information is relatively unimportant. Diagnosis of fetal head station
by vaginal examination alone might have been subjective. Factors such as operator
technique and
experience could have affected the outcomes, and data regarding location of the cup
in relation to the neonatal head sutures were also missing and might have influenced
neonatal head
injuries.
Conclusions
VE is an important modality in modern obstetrics; yet, occasionally it results in
maternal and neonatal adverse outcomes, which can cause lifetime disability. Our study
found seven
independent risk factors for VE-related feto-maternal adverse outcomes: increasing
neonatal weight, epidural anesthesia, use of VM cup, vacuum cup detachment, prolonged
second stage as
indication for VE and increasing procedure duration. OA position was a significant
protective factor against VE adverse outcomes. This information may help medical staff
performing VEs make an
informed decision regarding VE or CD during the second stage of labor. Additional
large-scale, prospective studies are needed to establish the risk factors for VE-related
adverse outcomes.
Declarations
Funding: This study was not funded.
Availability of data and material: Data can be made available upon reasonable request from the corresponding author.
Author contributions: Gal Cohen: Project development, Data Collection, Manuscript writing and editing.
Hanoch Schreiber, Michal Ovadia: Data collection. Gil Shechter-Maor, T
Biron-Shental: Revised manuscript critically.
Ethics approval: The study was approved by the Meir Medical Center Ethics Committee in August 2019,
approval number 0246–19-MMC. Since the study was based on patient records, informed
consent was not required.
