Assisted (operative) vaginal birth

INTRODUCTION — Assisted vaginal birth (also called operative vaginal birth) refers to a birth in which the operator uses forceps, a vacuum device, or another instrument to extract the fetus from the vagina, with or without concurrent maternal pushing. The decision to use an instrument to assist in delivering the fetus balances the maternal, fetal, and neonatal impact of the procedure against the alternative options of cesarean birth or expectant management.

This topic will provide an overview of issues related to assisted vaginal birth. The technique for vacuum-assisted birth is reviewed separately. (See "Procedure for vacuum-assisted vaginal birth".)

PREVALENCE — In the United States, 3.1 percent of all births in 2017 were accomplished via an assisted vaginal approach (the most recent data available) [1]. Forceps births accounted for 0.5 percent of vaginal births, and vacuum births accounted for 2.6 percent of vaginal births. However, there is a wide range in the prevalence of assisted vaginal birth both across and within geographic regions in the United States, which suggests that evidence-based guidelines for assisted vaginal birth are either inadequate or randomly applied, or familiarity and expertise with the technique is declining in some areas [2]. In the United States, the Midwest has the highest rates for both forceps and vacuum-assisted births, the Northeast has the lowest forceps rate, and the South has the lowest vacuum rate [2]. Overall, the rates of assisted vaginal birth have been decreasing both nationally and regionally in the United States [3].

Prevalence rates also vary worldwide depending on local practice patterns and availability of trained clinicians and other necessary resources [4]. Forceps- and/or vacuum-assisted birth was not taught or performed in some regions [5]. In a large prospective study of low- and middle-income countries, the assisted vaginal birth rate declined from 1.6 to 0.3 percent, while the cesarean rate more than doubled to reach 14.4 percent [6].

INDICATIONS

Overview — Use of either forceps or vacuum is reasonable when an instrumental intervention to complete labor is indicated and assisted vaginal birth can be safely and readily accomplished; otherwise, cesarean birth is the better option.

We agree with an American College of Obstetricians and Gynecologists (ACOG) practice bulletin that considers the following scenarios potentially appropriate reasons for assisted vaginal birth [7]:

●Maternal exhaustion and an inability to push effectively.

●Maternal medical indications, such as maternal cardiac disease and a need to avoid pushing in the second stage of labor.

●Prolonged second stage of labor.

●Suspicion of immediate or potential fetal compromise.

However, no indication is absolute, and cesarean birth is also an option in these clinical settings.

Although one can never be certain of a successful outcome, we attempt an assisted vaginal birth when we believe success is likely since the rate of birth trauma may be higher after failed attempts at assisted birth [8-10]. The decision to proceed with assisted vaginal birth is an ongoing process with constant reconsideration based on assessment of the success of sequential steps in the procedure. Preprocedure risk factors do not accurately predict whether an assisted vaginal birth attempt will succeed or fail [11].

Prolonged second stage of labor — To reduce the rate of cesarean birth for failure to progress in the second stage, ACOG and the Society for Maternal-Fetal Medicine recommend allowing three hours of pushing for nulliparous patients and two hours of pushing for multiparous patients before diagnosing arrest of labor, when maternal and fetal conditions permit [12]. They also opined that longer durations may be appropriate on an individual basis (eg, epidural anesthesia, fetal malposition) as long as progress is being documented, but they did not provide specific criteria for the upper limit of the second stage. Many obstetric providers allow an extra hour of pushing for patients with epidural anesthesia when the fetal heart rate pattern is reassuring of fetal well-being. These criteria are also useful for deciding when to perform an assisted vaginal birth.

For patients whose second stage is prolonged by these criteria and who have a normal fetal heart tracing and no other indication for expediting birth, we evaluate the relative value of an assisted birth versus expectant management. We favor expectant management when we believe a spontaneous vaginal birth is likely because fetal descent is progressing, albeit slowly, or because there has been a recent favorable change in the clinical situation, such as rotation from occiput posterior to occiput anterior, oxytocin augmentation, or more effective pushing. We favor assisted vaginal birth when further progress seems unlikely, and we believe assisted vaginal birth is the least morbid operative strategy, given the fetal station, position, and estimated size. Many of these cases appear to be related to ineffective pushing due to maternal exhaustion or, less commonly, to a maternal neurologic or muscular disease. Patients with a prolonged second stage who are not good candidates for assisted vaginal birth are delivered by cesarean. (See 'Prerequisites' below.)

Fetal compromise — Use of forceps or vacuum is appropriate when expeditious birth is indicated because of fetal compromise or probable imminent fetal compromise (eg, in the setting of acute abruption) and assisted vaginal birth can be safely and readily accomplished; otherwise, cesarean birth is the better option. (See "Intrapartum category I, II, and III fetal heart rate tracings: Management".)

Maternal medical disorder — Forceps or vacuum can be used to shorten the second stage of labor if the Valsalva maneuver is contraindicated or exertion should be minimized because of maternal medical disorders (typically cardiac or neurologic disease, also cystic lung disease). As discussed above, maternal neurologic or muscular disease can impair effective pushing and can be a reason to shorten the second stage. Instrumental intervention is performed when uterine contractions descend the fetus to a station where the clinician believes forceps or vacuum extraction can be performed safely and effectively.

CONTRAINDICATIONS — Assisted vaginal birth is contraindicated if the clinician or patient believes that the risk to mother or fetus is unacceptable. Examples include, but are not limited to [7,13]:

●Extreme fetal prematurity. (See 'Minimum and maximum estimated fetal weight' below.)

●Fetal demineralizing disease (eg, osteogenesis imperfecta). The safety of forceps or vacuum birth has not been established in disorders that result in demineralization of the skull. There is a theoretic risk for intracranial bleeding, extracranial bleeding, and other brain injuries due to cranial deformation or fracture from these instruments.

●Fetal bleeding diathesis (eg, fetal hemophilia, neonatal alloimmune thrombocytopenia [14]).

●Unengaged head. (The head is engaged when the widest diameter [the biparietal diameter] has reached or passed through the pelvic inlet. This typically occurs when the leading bony part has reached or passed through the ischial spines and when no more than one-fifth of the fetal head remains palpable from the maternal abdomen (in most cases, the head is not palpable abdominally once engaged) [15].

●Unknown fetal position.

●Brow or face presentation.

●Suspected fetal-pelvic disproportion. (See "Labor: Overview of normal and abnormal progression", section on 'Cephalopelvic disproportion'.)

Relative contraindications to vacuum extraction — Relative contraindications to use of vacuum devices, but not forceps, include gestational age <34 weeks or prior scalp sampling (which is rarely performed in contemporary United States practice). (See "Procedure for vacuum-assisted vaginal birth", section on 'Contraindications'.)

CLASSIFICATION — The American College of Obstetricians and Gynecologists' classification system for forceps births is based on station and extent of rotation, as these factors correlate with the level of difficulty and procedure-related risk (eg, lower fetal station and smaller degree of head rotation are associated with less risk of maternal and fetal injury [16]) [7]. Fetal station is measured using the -5 to +5 centimeter classification system (figure 1).

●Outlet forceps

•The leading point of the fetal skull has reached the pelvic floor, and at or on the perineum, the scalp is visible at the introitus without separating the labia.

•The sagittal suture is in anteroposterior diameter or a right or left occiput anterior or posterior position.

•Rotation does not exceed 45 degrees.

●Low forceps

•The leading point of the fetal skull is ≥2 cm beyond the ischial spines but not on the pelvic floor (ie, station is at least +2/5 cm).

•Low forceps have two categories that are based on whether rotation of the head is more or less than 45 degrees from the median sagittal plane:

-Without rotation – Rotation is 45 degrees or less (right or left occiput anterior to occiput anterior, or right or left occiput posterior to occiput posterior).

-With rotation – Rotation is greater than 45 degrees.

●Midforceps

•The head is engaged (ie, at least 0 station), but the leading point of the skull is not ≥2 cm beyond the ischial spines (ie, station is 0 to +1/5 cm).

Vacuum-assisted births do not have a separate classification system. The clinician should document the station at which the vacuum was applied. Rotational maneuvers should not be performed with vacuum because of the risk of severe scalp lacerations and concern for causing subgaleal hemorrhage. (See "Procedure for vacuum-assisted vaginal birth".)

PREREQUISITES

Overview — The operator should be experienced in assisted vaginal birth and responsible for determining that the following prerequisites are met prior to application of instruments:

●Cervix is fully dilated.

●Membranes are ruptured.

●Head is engaged (at least 0/5 cm station). Forceps and vacuum devices should never be used when the head is not engaged.

Large infants, extreme caput or molding, extension of the fetal head, pelvic deformities, and asynclitism may falsely suggest engagement, which can be recognized by assessing descent abdominally as well as vaginally (when maternal habitus allows). In these cases, the leading bony part may be at the ischial spines, although the biparietal diameter has not passed through the pelvic inlet. If the vertex is engaged, then no more than one-fifth of the fetal head should be palpable abdominally above the symphysis pubis [17].

●Fetal presentation, position, station, and any asynclitism are known, and extent of molding is estimated. The fetus must be in a cephalic presentation (unless the purpose is to use Piper forceps to assist in birth of an after-coming head in a breech presentation). It is important to reassess the fetal position in the second stage given that 68 percent of occiput posterior positions at delivery occur due to malrotation from occiput anterior or occiput transverse positions during labor [18], and 50 to 80 percent of occiput posterior positions will spontaneously rotate to occiput anterior during the second stage of labor [19].

●Fetal size is neither too large nor too small. (See 'Minimum and maximum estimated fetal weight' below.)

●Clinical pelvimetry suggests an adequate pelvis relative to estimated fetal size. A flat sacrum, a narrow subpubic angle, or a narrow distance between the ischial spines may prevent the fetal head from descending to fill the sacral hollow. Significant molding and caput may suggest some degree of pelvic obstruction.

●The patient consents to the procedure. The medical record should document the indication for the procedure, relevant clinical assessment of mother and fetus, and a summary of the informed consent discussion (specific risks, benefits, alternatives) [20].

●The option of performing an immediate cesarean birth is available if complications arise. Personnel for neonatal resuscitation are available, if needed.

●The patient has adequate anesthesia for the planned procedure.

●The maternal bladder is empty, as this may provide more room for fetal descent and possibly reduce injury to the bladder.

●Willingness to abandon the procedure if indicated.

Minimum and maximum estimated fetal weight — The minimum and maximum estimated fetal weights for assisted vaginal birth depend on the choice of instrument (vacuum versus forceps), size of available instruments, and patient-specific factors (eg, progress of labor, previous pregnancy history).

●Upper threshold – We believe that estimated fetal weight is one of several factors to assess when considering assisted birth of a suspected macrosomic infant. Multiple maternal factors (eg, diabetes, body mass index [BMI], prior newborn size in successful vaginal births, clinical pelvimetry, progress in the second stage) and fetal factors (eg, head position and station, caput and molding, estimated abdominal circumference compared with head circumference) can influence the decision to attempt an assisted vaginal birth.

In general, patients with severe obesity (defined as a BMI >40 kg/m²), diabetes, slow progress in the second stage of labor with significant caput/molding, or an estimated newborn birth weight over 4000 grams are not ideal candidates for assisted vaginal birth, but these cases must be individualized. An example of a case with estimated fetal weight 4000 grams where we would consider assisted vaginal birth might be a patient with fetal compromise at +3 station who has a normal BMI, no gestational or pregestational diabetes mellitus, and a history of a prior spontaneous vaginal birth of a newborn of similar size. (See "Shoulder dystocia: Risk factors and planning birth of high-risk pregnancies".)

Assisted vaginal birth of the macrosomic infant may be associated with an increased risk of injury. As an example, in a study including almost 3000 newborns ≥4000 grams, the risk of persistent significant injury at six months of age was 1.5 percent (4/261) after forceps birth, 0.24 percent (4/1666) after spontaneous birth, and 0 (0/862) after cesarean or vacuum birth (0/135) [21]. The authors estimated that a policy of planned cesarean birth for macrosomia would necessitate 148 to 258 cesarean births to prevent a single persistent injury; avoidance of assisted vaginal birth would require 50 to 99 cesarean births per injury prevented. However, these estimates are imprecise because of the small number of observed injuries and the poor performance of intrapartum identification of macrosomic fetuses. (See "Fetal macrosomia".)

The American College of Obstetricians and Gynecologists practice bulletin on assisted vaginal birth suggests that judicious use of forceps or vacuum extraction is not contraindicated for most fetuses suspected to be macrosomic, if the maternal pelvis and progress of labor are adequate [7]. However, the obstetrician should be aware of the risk of shoulder dystocia, especially when the second stage of labor is prolonged.

●Lower threshold – Use of vacuum devices is limited to births ≥34 weeks of gestation because the risk of intraventricular hemorrhage appears to be increased above baseline when these devices are employed at earlier gestational ages.

"Baby" Elliot and "baby" Simpson forceps have the same blade dimensions as standard forceps, but the handles and shafts are shorter and lighter than standard forceps; they have been used to assist the birth of fetuses as small as 1000 grams [22]. We were unable to identify any studies or manufacturer guidelines regarding prerequisites for estimated fetal weight or gestational age for use of these instruments. They are well suited to outlet deliveries and fetal head extraction at cesarean birth. (See "Delivery of the low birth weight singleton cephalic fetus", section on 'Do forceps protect the head during vaginal birth?' and "Delivery of the low birth weight singleton cephalic fetus", section on 'Use of episiotomy, vacuum, and forceps'.)

PATIENT PREPARATION

Anesthesia — Before beginning an assisted vaginal birth, maternal anesthesia should be satisfactory. Neuraxial anesthesia provides more effective analgesia than pudendal block and should always be offered for forceps-assisted birth, especially mid-cavity or rotational procedures [23]. Pudendal block may be adequate for low forceps, or for vacuum extraction (because, unlike forceps blades, the vacuum cup does not significantly displace the walls of the birth canal or increase the cephalic diameter).

Ancillary procedures

Ultrasound — We often use ultrasound to determine fetal position and station before assisted vaginal birth to confirm our diagnosis from physical examination and to assess its chances of success and its risks. We always perform an ultrasound examination when we are uncertain of the head position.

Intrapartum sonographic visualization of fetal intracranial structures, including the cerebellum, orbits, and midline falx, can be used to determine fetal head position and station and thus could reduce morbidity of improperly placed instruments at the time of an assisted vaginal birth. Multiple studies comparing ultrasound with digital vaginal examination of head position have shown digital examination is incorrect in approximately 20 to 40 percent of cases, regardless of the experience of the person performing the examination, whereas ultrasound is incorrect in only 1 to 2 percent of cases [24-26]. Despite more accurate diagnosis of head position, a meta-analysis of randomized trials of ultrasound before assisted vaginal birth found no clear improvement in composite maternal or neonatal outcomes [27]. However, the trials were underpowered to allow a definitive conclusion.

Ultrasound can diagnose lack of engagement and predict a difficult assisted vaginal birth. In a meta-analysis of 16 studies with >2800 patients), angle of progression, progression distance, and midline angle measured during pushing had high prognostic accuracy for predicting a complicated or failed assisted vaginal birth [28].

Use of ultrasound to determine fetal position during late labor when cervical dilation is ≥8 cm may have unanticipated consequences. In a randomized trial comparing digital examination versus both ultrasound and digital examination at ≥8 cm dilation, occiput posterior and occiput transverse positions were under-detected on digital examination alone and knowing correct fetal position increased the likelihood of cesarean birth [29].

Antibiotics — Prophylactic antibiotics are not routinely administered before assisted vaginal birth because unnecessarily exposing mothers and fetuses to antibiotics has potential adverse effects (eg, emergence of antimicrobial resistance, selection of pathogenic organisms such as Clostridioides [formerly Clostridium] difficile, and drug toxicity). However, a trial (ANODE) of 3427 patients randomized to a single dose of intravenous co-amoxiclav versus placebo as soon as possible after any assisted vaginal birth reported that the intervention resulted in a 42 percent reduction in suspected or confirmed infection in the six-week postpartum period (11 versus 19 percent, relative risk [RR] 0.58, 95% CI 0.49-0.69) [30]. The primary outcome was defined by a new prescription of antibiotics for presumed perineal wound-related infection, endometritis or uterine infection, urinary tract infection with systemic features or other systemic infection, or confirmed systemic infection on culture. Importantly, the secondary outcomes of superficial and deep incisional infections were both reduced in the antibiotic group (4 versus 8 percent [RR 0.53, 95% CI 0.37-0.75] and 2 versus 5 percent [RR 0.46, 95% CI 0.28-0.77]). This appeared to translate into less perineal discomfort and less need for additional at-home or in-office perineal evaluation and care.

Some factors that limit the generalizability of the ANODE trial to United States and other populations include higher rate of assisted vaginal birth in the United Kingdom than in the United States (13 versus 3 percent [1]), higher episiotomy rate (89 versus 8 percent [31]), higher proportion of forceps use (2/3 versus 1/5 assisted vaginal births performed [1]), and the different approach to group B streptococcus (GBS) intrapartum antibiotic prophylaxis at the time of the trial (risk-based rather than culture-based). Other limitations of the trial are that only 78 patients (2.4 percent) had no episiotomy or tear [32], the higher observed than expected rate of infection in the placebo group (19 versus 10 percent) [30], and a change in an outcome metric during the trial from "any antibiotics" to "antibiotics only for the indication of perineal infection" [30]. The finding that antibiotics administered at a median time of 3.2 hours after birth were effective is also surprising since tissue concentrations would not be optimal at the time of the procedure and repair. However, a secondary analysis of data from the ANODE trial noted that each 15-minute increment between birth and antibiotic administration was associated with a 3 percent higher risk of infection (adjusted risk ratio [aRR] 1.03, 95% CI 1.01-1.06) [32].

Additional trials are needed to confirm a benefit in other populations before recommending prophylactic antibiotics for all patients who undergo an assisted vaginal birth. A 2020 systematic review [33] found only one additional trial [34], which was small (394 participants) and only evaluated the incidence of postpartum endometritis.

●Our approach – When an assisted vaginal birth is performed with an episiotomy, we believe it is reasonable to administer a single dose of intravenous co-amoxiclav immediately after the birth, based on the ANODE trial and until better data are available. The vast majority of patients in ANODE had episiotomies. A World Health Organization guideline endorsed administration of a single dose of intravenous amoxicillin 1 g and clavulanic acid 200 mg as soon as possible and no more than 6 hours after assisted vaginal birth [35].

Although the secondary analysis of ANODE also indicated benefit in the 8.6 percent of patients with tears alone (no episiotomy), the degree of tear was not reported; thus, it is unknown whether the group with first- and second-degree tears would benefit from antibiotics [32]. We routinely administer prophylactic antibiotics before repairing a third- or fourth-degree laceration. The American College of Obstetricians and Gynecologists recommends a single dose of prophylactic antibiotics at the time of repair of obstetric anal sphincter injuries (ie, extending into or through the anal sphincter complex), regardless of mode of birth [36].

Episiotomy — We do not routinely perform an episiotomy. The only randomized trial comparing routine versus restrictive episiotomy at assisted vaginal birth found no significant differences between groups in the rate of anal sphincter tear, postpartum hemorrhage, neonatal trauma, or pelvic floor symptoms until 10 days postpartum; however, this was a pilot study with only 200 participants (the type of episiotomy was not described, but mediolateral episiotomy is preferred in Europe, where the trial was performed) [37].

We suggest performing episiotomy rarely and selectively in patients at high risk for sphincter laceration and with shared decision-making. If performed, we suggest a mediolateral or lateral episiotomy rather than a median episiotomy to reduce the number of apparent anal sphincter injuries [38-40], although there is more initial postpartum discomfort compared with a median incision and early complaints of flatal incontinence are common (9 percent) [41-45]. (See "Approach to episiotomy", section on 'Mediolateral versus median (midline) episiotomy'.)

Our approach is based on observational studies that suggest that a median (ie, midline) episiotomy increases, rather than decreases, the risk of perineal trauma in assisted vaginal births compared with no episiotomy [46-51] and mediolateral or lateral episiotomies performed in certain situations, such as in nulliparous patients undergoing forceps or vacuum extraction or in a persistent occiput posterior presentation, may reduce anal sphincter lacerations compared with no episiotomy [38,40,51]. However, long-term prospective data are lacking, thus it is unclear whether long-term fecal continence and anal sphincter function are better in patients who routinely undergo mediolateral episiotomy at the time of assisted vaginal birth than those who undergo selective mediolateral episiotomy or no episiotomy. A comprehensive review of community prevalence and associated factors with fecal incontinence found that obstetric history or previous obstetric trauma was not associated with reported fecal incontinence [52].

CHOICE OF INSTRUMENT — Both forceps and vacuum are acceptable instruments for assisted vaginal birth. Our approach depends on patient-specific factors, as described below.

When to choose vacuum versus forceps — We choose vacuum extraction when a relatively easy extraction is anticipated (eg, occipito-anterior position at low stations with no signs of relative cephalopelvic disproportion). Lower station vacuum procedures are associated with a higher chance of success compared with attempts at higher stations (93 to 99 versus 76 percent) [53]. Because success is likely, the primary consideration in these cases is to minimize the risk of maternal injury. If a difficult extraction is anticipated, we choose forceps despite a slightly higher risk of maternal injury because vacuum extraction is likely to fail [54].

The choice of instrument is determined by the clinician's expertise with the various forceps and vacuum devices, availability of the instrument, level of maternal anesthesia, and knowledge of the risks and benefits associated with each instrument in various clinical settings. Vacuum-assisted birth is generally less traumatic for the mother than forceps-assisted birth. Vacuum devices are easier to apply and require less maternal anesthesia than forceps. Fetal head rotation may occur passively during fetal extraction. The advantages of forceps are that they have a significantly higher success rate, are unlikely to detach from the head during a difficult extraction, can be used on preterm fetuses or to actively rotate the fetal head, and do not aggravate bleeding from scalp lacerations. It is not clear which procedure is safer for the fetus; the complication profiles for the two procedures are different. (See 'Complications' below.)

Choice of vacuum cup — All vacuum extraction devices consist of a soft or rigid plastic cup, a vacuum pump to provide suction between the cup and fetal scalp, and a traction system. A soft vacuum cup (which is associated with less scalp trauma) is appropriate for most births but is especially suited to outlet deliveries with an occiput anterior position. Rigid cups may be preferable for occiput posterior, occiput transverse, and difficult occiput anterior deliveries because they are less likely to detach. A more detailed discussion regarding the choice of an extractor cup can be found separately. (See "Procedure for vacuum-assisted vaginal birth", section on 'Cup'.)

Choice of forceps — The type of forceps selected for a particular procedure depends on several factors, including:

●The size and shape of the fetal head and maternal pelvis, which should match the size, cephalic curve, and pelvic curve of the forceps. A good head application (bimalar, biparietal) is a key goal in choice of forceps.

•Simpson type forceps, which have long tapered blades, tend to be the best fit for a molded head (picture 1) because of the less concave cephalic curve.

•Elliott type forceps (picture 2) or Tucker-McLane type forceps (picture 3) are better suited to a round, unmolded head as the cephalic curve of the forceps is more concave.

Fenestrated blades (picture 4) allow for a better grip and therefore are less likely to slip, but the fenestrations increase the risk for tissue laceration when greater forces are applied. Solid blades (picture 3) are less likely to lacerate the fetal head but may be more likely to slip with increased traction. Pseudo fenestrated blades have a shallow indentation rather than a true fenestration, which may reduce slippage while also reducing risks of laceration.

●Fetal head position and whether rotation is planned. Choosing the right forceps for the direction of traction and type of rotation is another key goal.

•Kielland forceps are useful for rotations because of their minimal (reverse) pelvic curve and sliding lock (picture 4). A sliding lock is helpful when there is asynclitism.

•Piper forceps are used to assist the birth of the aftercoming head in vaginal breech births (picture 5).

●Station.

•Midpelvic procedures may be facilitated by an instrument that can be used with a traction handle (eg, Bill's axis traction handle or Irving forceps). Traction is applied in the axis of the pelvis, which is curved in most patients. If the fetal head is at a station that requires an axis of traction that is not feasible with a standard manual method of axis traction (the Pajot-Saxtorph maneuver), then an instrument with axis traction is helpful.

●Operator experience and preference.

Application of forceps is more difficult, requires more manipulation for a good application, and is more likely to result in maternal or fetal trauma with higher stations, head asynclitism, non-anterior positions, rotations beyond 45 degrees, and unusual pelvic types; therefore, choosing the correct instrument is particularly important in these settings. A detailed discussion of the hundreds of types of instruments available for forceps-assisted birth and their application is beyond the scope of this review. The classic resource is Dennen's Forceps Deliveries, but books on operative obstetrics are also helpful.

Novel devices

Thierry or Teissier spatula — The Thierry and Teissier spatulas consist of two independent and symmetric branches which include a shank, handle, and wide solid blade [55]. The shanks do not articulate; thus, each branch acts as an independent lever and the head is not compressed between the blades. In a meta-analysis (nine observational studies; nearly 17,000 patients), spatulas had a similar failure rate as forceps but a lower incidence of maternal trauma/lacerations and neonatal soft tissue injuries (low-certainty evidence) [56].

Odon device — The Odon device was developed by the World Health Organization for use in areas that have limited or no access to cesarean birth. It is a low-cost device made of film-like polyethylene material that creates a sac filled with air that surrounds the entire head and enables extraction when traction is applied. It has the potential to be safer and easier to apply than forceps or a vacuum extractor. Videos showing application of the device and fetal extraction are available online (eg, online).

No randomized trials have published data regarding its safety and efficacy. In a pilot observational study, the Odon Device alone assisted in the birth of 19 of the 40 newborns, with no serious maternal or neonatal adverse outcomes related to its use [57].

PROCEDURE — Use of a checklist for preparation, performance, and documentation of assisted vaginal birth can help to ensure that the important elements of the procedure have been addressed and documented (table 1) [58].

Forceps

●Application – Appropriately applied forceps grasp the occiput anterior (OA) fetal head such that:

•The long axis of the blades corresponds to the occipitomental diameter (figure 2).

•The tips of the blades lie over the cheeks (figure 3).

•The blades are equidistant from the sagittal suture, which should bisect a horizontal plane through the shanks.

•The posterior fontanelle should be one finger breadth anterior to this plane.

•Fenestrated blades should admit no more than one finger breadth between the heel of the fenestration and the fetal head.

•No maternal tissue has been grasped.

●Midforceps – Midforceps births are generally avoided because they are associated with a higher rate of severe perinatal morbidity/mortality compared with cesarean births performed in the second stage, especially when the indication for intervention is dystocia. They are also associated with higher rates of maternal trauma. (See 'Complications of midpelvic procedures' below.)

However, clinicians highly experienced with forceps-assisted births may choose to attempt a midforceps birth in select circumstances, such as sudden severe fetal or maternal compromise, if the clinician believes that they can safely expedite and effect a safe assisted vaginal birth. Simultaneous preparation for cesarean birth should be underway. We suggest performing midforceps births in the operating room so that cesarean birth can be performed promptly, if necessary.

When attempting a midforceps birth, the leading point of the skull should be at or just beyond the ischial spines (0 to +1/5 station) to ensure the head is engaged. It would be rare to have an engaged head when the leading point is at -1/5 cm station and extraordinarily unlikely at -4 or -5 station.

●Rotation – A rotational delivery is an appropriate option in select clinical circumstances [7], as neonatal morbidity is not increased compared with appropriate controls when intervention is indicated and when performed by experienced clinicians [59-62]. Rotation, when needed, is performed between contractions. Rotation followed by extraction is more difficult and associated with a higher risk of maternal and fetal complications than simple traction applied to the non- or minimally-rotated head. Forceps application and rotation when the fetal head is not directly OA is beyond the scope of this topic review.

●Traction – Traction should be steady (not rocking) and in the line of the birth canal, rotating under the symphysis pubis, along the curve of Carus (ie, pelvic axis). It should be applied with each contraction and in conjunction with maternal expulsive efforts. In most cases, progress is noted with the first or second pull, and birth occurs by the third or fourth pull [63]. The procedure should be abandoned if descent does not occur with appropriate application and traction.

The forceps pressure on the fetal head can be relaxed between contractions to reduce fetal cranial compression.

●Removal – To reduce the risk of laceration, forceps are disarticulated and removed when expulsion is certain, but before the widest diameter of the fetal head passes through the introitus. The head can then be delivered with no or minimal maternal assistance.

Vacuum — The procedure for vacuum extraction is reviewed separately. (See "Procedure for vacuum-assisted vaginal birth".)

When to abandon the procedure — Assisted vaginal birth should be abandoned if it is difficult to apply the instrument, descent does not easily proceed with traction, or the fetus has not been delivered within a reasonable time [17]. Some experts suggest abandoning the procedure if birth has not occurred within 15 to 20 minutes or after three pulls. A cohort study found that 82 percent of completed assisted vaginal births (vacuum or forceps) occurred with one to three pulls and that pulling more than three times was associated with infant trauma in 45 percent of such births [64]. A secondary analysis of a multicenter observational cohort study found that increasing duration of assisted vaginal birth time was more strongly associated with adverse neonatal outcomes than the number of forceps pulls or vacuum cup pop-offs, but the optimum threshold for abandoning the procedure was not identified [65]. Durations greater than 12 minutes had the strongest association with both adverse neonatal outcomes and failed assisted vaginal births. Based on these and other data, we believe it is prudent to abandon the procedure if good instrument placement and adequate traction are followed by no descent with three attempts. The duration of the procedure should be monitored and used as an additional factor for consideration when deciding whether to continue with an assisted vaginal birth. If descent has occurred and birth is clearly imminent, then proceeding with assisted-vaginal birth after three pulls or a longer duration may be appropriate and less morbid than a cesarean birth of a newborn with its head on the perineum.

The operator should not be fixated on achieving a vaginal birth. It is essential that the operator be willing to abandon a planned or attempted assisted vaginal birth and have the ability to perform a cesarean birth if evaluation or reevaluation of the clinical status shows that an instrument-assisted birth is contraindicated (eg, the fetal head is not engaged, the position is uncertain, the procedure is not succeeding).

The most common and highest risk clinical factors associated with failed assisted vaginal birth are occiput posterior position, macrosomia, prolonged second stage, primiparity, and maternal body mass index (BMI) ≥30 kg/m² at delivery [8,66-74]. Other characteristics that have been associated with failure include higher station and excessive molding of the fetal head. (See "Occiput posterior position", section on 'Management' and "Shoulder dystocia: Risk factors and planning birth of high-risk pregnancies", section on 'Planning birth in high-risk pregnancies'.)

Higher rates of neonatal morbidity have been observed when cesarean birth was performed after a failed assisted vaginal birth than when performed during labor without such attempts (table 2) [75,76]; there are many limitations to these observational data, including confounding by indication for intervention and lack of appropriate controls.

Second attempt with a different instrument — Sequential attempts with different instruments should not be performed routinely. For a procedure to be considered sequential, traction is applied sequentially by two different instruments. Situations in which an instrument is placed, but no traction applied, should not be considered a sequential attempt: for instance, when proper placement of forceps cannot be achieved, or a vacuum device fails to achieve suction and no traction has been applied. Sequential instrumentation might occur in the rare situation of potential or immediate fetal compromise when a vacuum attempt by a less experienced provider has failed and a more experienced provider believes that vaginal birth can be safely achieved with forceps more quickly than with cesarean. We believe that sequential use may be avoided if less experienced providers consult more experienced providers before attempting a challenging assisted vaginal birth. This facilitates both a reduction in sequential instrument use and ongoing learning for less experienced providers.

The American College of Obstetricians and Gynecologists recommends against routinely performing sequential attempts at assisted vaginal birth using different instruments due to the greater potential for maternal and/or fetal injury [7]. Population-based data have reported increased maternal and neonatal morbidity from sequential application of vacuum and forceps [75,77,78]. Although a few studies have demonstrated no adverse effects from sequential use of vacuum and forceps, even when vaginal birth was not achieved, a type II error may have resulted from the small number of participants and events in these studies [79,80].

For the mother, sequential use of vacuum and forceps has been associated with increased rates of third-/fourth-degree lacerations and postpartum hemorrhage [78]. For the neonate, sequential use of these instruments has been associated with increased rates of subdural hematomas and intracranial hemorrhage. In one large study, the incidence of subdural or cerebral hemorrhage in newborns delivered by vacuum and forceps, vacuum alone, or forceps alone was approximately 21, 10, and 8 per 10,000 births, respectively [75]. These findings were corroborated by another analysis based upon statewide birth certificate data [77] and a small study of asymptomatic term newborns who underwent routine magnetic resonance imaging within 48 hours of birth [81]. In the latter, 9 of 111 asymptomatic newborns had a subdural hematoma, and the highest proportion was in the group exposed to failed vacuum followed by successful forceps-assisted birth (five subdural hematomas among 18 newborns [28 percent]) [81]. In comparison, 3 subdural hematomas occurred in 49 newborns spontaneously delivered from vertex presentation (6 percent), 1 subdural hemorrhage occurred among 13 successful vacuum-assisted births (8 percent), and 0 subdurals occurred among 4 successful forceps-assisted births. All hematomas had resolved without clinical sequelae when reevaluated four weeks later.

Maternal and newborn examination — The lower genital tract, peritoneum, and anus/rectum should be examined after the birth for lacerations. It is important to remember to perform this examination in patients who undergo cesarean birth after a failed attempt at assisted vaginal birth.

The neonatal care provider should be informed that vacuum or forceps were used to assist the birth. Since most serious complications, such as a subgaleal hematoma, occur within hours of birth [82], it is important to inform these providers by either a reliable charting method, direct notification, or both.

SUCCESS RATE — In a systematic review of randomized trials, failed to achieve a vaginal birth occurred in approximately 7.9 percent of forceps- and 13.7 percent of vacuum-assisted births (relative risk 0.58, 95% CI 0.39-0.88; 11 trials, 3080 births) [54]. Birth certificate data including over two million births in the United States suggest that attempted assisted vaginal birth (forceps, vacuum) failed in approximately 18 percent of cases in which an instrument was applied [83]. When unsuccessful, 60 percent of patients went on to have a vaginal birth, and 40 percent had a cesarean.

Midforceps birth is more likely to fail than low forceps birth; failure rates were 8.9 and 0.3 percent, respectively, in one large prospective study [84]. Other risk factors for failure include higher body mass index (especially ≥35), neonatal birth weight more than 4000 g, lower operator procedure volume, and advanced maternal age [85]. 

Historically, failed forceps was more likely to lead to cesarean birth than failed vacuum since failed vacuum extraction was sometimes followed by a successful trial of forceps, but the converse rarely occurs. As sequential use of instruments carries much higher morbidities, it is no longer considered acceptable to perform sequential instrument use. (See 'Second attempt with a different instrument' above.)

COMPLICATIONS

Overview — Maternal and fetal/neonatal complication rates vary widely in published series and depend on a number of factors, which are not independent. These factors include type of instrument, head position at application, station, indication for intervention, and operator experience. Rotation, higher station, longer active second stage of labor, and operator inexperience variably increase the risk of complications. However, complications can occur even when instruments are correctly applied and used. Virtually all complications associated with assisted vaginal birth can also occur in the course of a spontaneous vaginal birth, but the incidence is lower in the latter.

Newborn birth trauma is the major potential complication of assisted vaginal birth. A population-based analysis of over 11 million singleton births in the United States provided crude comparative neonatal morbidity/mortality data for unassisted (spontaneous), forceps-assisted, and vacuum-assisted births (table 3) [86]. Although based on far fewer births, state-based data for California and New Jersey provide a more detailed description of the specific types of neonatal injuries associated with different instruments and modes of birth (table 4 and table 5 and table 6) [75,86]. These data should be interpreted with respect to appropriate control groups and reasonable alternative procedures. For example, second stage cesarean birth is an alternative to assisted vaginal birth, but prelabor cesarean and spontaneous vaginal birth are not realistic alternatives in the setting of second stage labor complications. Bias in patient selection is also an important factor; the vacuum approach is often favored over forceps in patients most likely to give birth readily. When considering the morbidity of assisted vaginal versus cesarean birth, it should be noted that, when dystocia is the indication, forceps- and vacuum-assisted births are associated with higher rates of perinatal morbidity and mortality than cesarean births performed in the second stage [87].

Maternal birth trauma also needs to be considered. One in four attempted forceps-assisted births and one in eight attempted vacuum-assisted births in Canada resulted in maternal trauma, most commonly obstetric anal sphincter injury [88].

Neonatal complications

Vacuum-assisted births — Torsion and traction by the vacuum cup can cause life-threatening complications following use of vacuum-assisted devices [89]. These complications include intracranial hemorrhage (epidural, subdural, intraparenchymal, subarachnoid), intraventricular hemorrhage, and subgaleal hemorrhage (figure 4). Other potential complications include fetal scalp abrasions and lacerations, cephalohematoma, retinal hemorrhage, and brachial plexus injury [75,90-93]. (See "Neonatal birth injuries".)

●In a prospective study of retinal hemorrhage in healthy newborns, the incidence was higher for vacuum-assisted than for spontaneous vaginal or cesarean births (75, 33, and 7 percent, respectively) [94]. The hemorrhages typically resolved without sequelae within four weeks of birth.

●In a systematic review, cephalohematoma was more than twice as common after vacuum-assisted extraction than forceps-assisted birth (9.5 versus 3.7 percent) [54]. The majority of cephalohematomas resolve spontaneously over the course of a few weeks without any intervention.

●In two retrospective studies, shoulder dystocia was twice as common with vacuum-assisted than forceps-assisted births (3.5 versus 1.5 percent [95], 1.1 versus 0.54 percent [86]). For this reason, vacuum-assisted births are at higher risk of brachial plexus injury than forceps-assisted or cesarean births [93].

Additionally, a prospective study of vacuum-assisted birth found that an increasing number of pulls and prolonged traction were associated with subgaleal hemorrhage [96]. Suboptimal cup placement was associated with failed procedures but not with subgaleal hemorrhage or other vacuum-related birth trauma; however, the small number of events precludes making a clear conclusion about the risk of these latter outcomes.

Forceps-assisted births — Complications of forceps-assisted birth include skin markings and lacerations, external ocular trauma, intracranial hemorrhage, subgaleal hemorrhage, retinal hemorrhage, lipoid necrosis, facial nerve injury, skull fracture, and, rarely, death [13,75,97,98]. In a meta-analysis of randomized trials comparing forceps- with vacuum assisted-births, forceps trended toward a higher risk of scalp, facial, and intracranial injury (scalp injury: RR 1.29, 95% CI 0.89-1.87; facial injury: RR 7.17, 95% CI 0.92-55.71; intracranial injury: RR 1.37, 95% CI 0.60-3.11), but data were limited [54]. Forceps resulted in less cephalohematoma (3.7 versus 9.7 percent, RR 0.41, 95% CI 0.30-0.56) and retinal hemorrhage (16.8 versus 24.2 percent, RR 0.66, 95% CI 0.46-0.94).

Maternal complications — Maternal complications associated with assisted vaginal birth include lower genital tract laceration, vulvar and vaginal hematomas, urinary tract injury, and anal sphincter injury [99-108]. Occiput posterior position is a risk factor for maternal trauma during assisted vaginal birth [109-111], particularly third-/fourth-degree perineal lacerations (spontaneous birth 2 percent, vacuum extraction 10 to 11 percent, forceps-assisted birth 17 to 20 percent [106,112]).

Increasingly complex instrumental intervention is associated with increasing maternal morbidity such that spontaneous vaginal birth is least morbid, followed by vacuum extraction, then forceps-assisted birth, and lastly cesarean birth, where venous thromboembolism, endometritis, and wound infection are particular concerns. Although cesarean birth protects the genital tract from forceps/vacuum-related trauma, the long-term risks of urinary incontinence, anal incontinence, and prolapse symptoms were not lower with second stage cesarean birth compared with assisted vaginal birth in longitudinal cohort studies [113,114].

Vacuum-assisted births — Randomized trials generally report less maternal genital trauma with vacuum versus forceps extraction (refer to data below), which is not unexpected given that a correctly applied vacuum cup does not take up additional space between the fetal head and the birth canal and does not make contact with maternal soft tissue [54].

This difference has not been proven to impact long-term maternal outcomes, such as urinary and anal dysfunction and pelvic organ prolapse. In a trial that randomly assigned 75 patients to forceps- or vacuum-assisted birth and then surveyed them five years postpartum, long-term morbidity rates were similar for both instruments; 47 percent had some degree of urinary incontinence and 20 percent had loss of bowel control "sometimes" or "frequently" [115]. This trial was limited by the small number of subjects. In a prospective study including over 2100 subjects that used multivariable logistic regression and propensity score methods to control for indication bias, the risk of urinary or anal incontinence at six months postpartum was similar for attempted forceps-/spatula- and vacuum-assisted births [116].

Vacuum-assisted births have been associated with lower rates of maternal morbidity and mortality than cesarean births in the second stage [87]. (See "Fecal and anal incontinence associated with pregnancy and childbirth: Counseling, evaluation, and management" and "Effect of pregnancy and childbirth on urinary incontinence and pelvic organ prolapse".)

Forceps-assisted births — Randomized trials generally report a higher risk of maternal genital trauma with forceps versus vacuum extraction (any perineal, vulval, or vaginal trauma: 95 versus 92.5 percent, OR 1.53, 95% CI 0.98-2.40; third-/fourth-degree laceration: 15 versus 8.2 percent, RR 1.83, 95% CI 1.32-2.55) [54].

Rotational and midforceps procedures have been considered major risk factors for serious maternal trauma during assisted vaginal birth. In case reports and small series, direct bladder injury, ureteral lacerations/transections, and uterine rupture have been reported [117]. A direct comparison meta-analysis of 13 observational studies published after 2000 found no statistically significant difference in the risk of obstetric anal sphincter injury between rotational vaginal birth with Kielland forceps versus rotational vacuum (RR 1.09, 95% CI 0.70-1.71) or between rotational vaginal birth with Kielland forceps versus nonrotational forceps delivery (RR 0.81, 95% CI 0.60-1.09) [118]. Rates of shoulder dystocia were higher with Kielland's delivery compared with vacuum delivery (RR 1.79, 95% CI 1.08–2.98), but rates of neonatal birth trauma were lower (RR 0.49, 95% CI 0.26–0.91). Interpretation of these findings is limited by bias in selection of cases for each method of birth.

Complications of midpelvic procedures — A study of perinatal and maternal morbidity and mortality associated with attempted midpelvic assisted procedures compared with cesarean birth reported the following observations [119]:

●Among patients with dystocia and a prolonged second stage, midpelvic procedures (n = 7521) were associated with higher risks of severe perinatal morbidity/mortality than cesarean birth (n = 9300):

•Forceps adjusted odds ratio (aOR) 1.81, 95% CI 1.24-2.64 (1.1 versus 0.7 percent)

•Vacuum aOR 1.81, 95% CI 1.17-2.80 (1.2 versus 0.7 percent)

•Sequential instruments aOR 3.19, 95% CI 1.73-5.88 (2.0 versus 0.7 percent)

In particular, midpelvic assisted vaginal birth was strongly associated with a high risk of severe newborn birth trauma compared with cesarean birth:

•Forceps aOR 5.01, 95% CI 2.75-9.15 (0.7 versus 0.2 percent)

•Vacuum aOR 4.47, 95% CI 2.27-8.80 (0.7 versus 0.2 percent)

•Sequential instruments aOR 9.46, 95% CI 4.11-21.8 (1.4 versus 0.2 percent)

The overall risk of severe maternal morbidity and mortality was similar for midpelvic procedures and cesarean birth, but midpelvic procedures were associated with a high rate of maternal obstetric trauma (eg, third- and fourth-degree perineal lacerations, cervical laceration, high vaginal laceration: 22.9 percent with forceps, 15.4 percent with vacuum), which significantly exceeded the 6.3 percent rate of obstetric trauma (injury to pelvic organ/joint, pelvic hematoma and extension of uterine incision) with cesarean birth.

●Among patients with fetal distress and a prolonged second stage, the difference in severe perinatal morbidity/mortality between routes of birth was not statistically significant, but midpelvic procedures were still associated with significantly higher rates of severe newborn birth trauma and maternal obstetric trauma.

These data underscore the importance of case-by-case assessment of the relative risks and benefits of midpelvic vaginal procedures versus cesarean birth.

NEURODEVELOPMENTAL OUTCOME — Assisted vaginal birth, whether by forceps or vacuum extraction, does not appear to have an adverse effect on neurodevelopment when evaluated in school-aged children.

In two randomized trials, forceps- and vacuum-assisted births were associated with similar neurodevelopmental outcomes [115,120]. Both of these trials lacked a comparison with children born spontaneously or by cesarean. A third trial compared neurodevelopmental outcome at age five for children born by successful assisted vaginal birth, failed assisted vaginal birth, and cesarean birth in the second stage of labor [121]. Neurodevelopmental morbidity was low with no significant differences among the three groups, but the study was underpowered.

One of the few follow-up evaluations comparing outcome at school age after assisted or spontaneous birth in over 3000 five-year-olds found no differences in cognitive testing [122]. This series included 1192 forceps-assisted births, of which 114 were midforceps. Another study compared the neurologic outcome of 295 10-year-old children born by vacuum extraction with that of 302 children born spontaneously in the same hospital by the same doctors and matched for maternal age, gestational age, and birth weight [123]. Both groups had similar results for tests of fine and gross motor control, perceptual integration, behavioral maturity, scholastic performance, speech ability, and self-care. In addition, a national cohort study of 126,032 16-year-olds born as nonanomalous vertex singletons ≥34 weeks of gestation to Swedish-born parents noted that those born by vacuum extraction had slightly lower mean mathematics test scores than children born vaginally without instruments after adjustment for major confounders, but similar scores to those born by intrapartum cesarean birth [124]. A subsequent large cohort study using data from nationwide Swedish registries reported that children born after successful vacuum extraction at 37 to 41.6 weeks had no increased risk of neonatal death, cerebral palsy or epilepsy compared with those delivered by emergency cesarean, but there was an increased risk of neonatal death among those born after failed vacuum extraction [125].

RATE OF RECURRENT ASSISTED VAGINAL BIRTH — Approximately 5 percent of patients who have an assisted vaginal birth will have a second assisted vaginal birth [126,127].

USE OF VACUUM OR FORCEPS AT CESAREAN BIRTH — (See "Cesarean birth: Management of the deeply impacted head and the floating head".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Delivery".)

SUMMARY AND RECOMMENDATIONS

●Indications – The risks and benefits of assisted (also called operative) vaginal birth for each patient need to be balanced against those for cesarean birth and less invasive interventions. Use of forceps or vacuum is reasonable when an intervention to complete labor is indicated and assisted vaginal birth can be safely and readily accomplished; otherwise, cesarean birth is the better option. Situations where assisted vaginal birth may be preferable to cesarean birth or less invasive interventions include prolonged second stage when maternal exhaustion impedes further progress, fetal compromise where expeditious birth is desirable, and maternal medical disorders where pushing (Valsalva) needs to be avoided or minimized. Nonmedically indicated shortening of the second stage is not an indication for assisted vaginal birth. (See 'Indications' above.)

●Prerequisites – Before resorting to an assisted vaginal birth, the clinician should ensure that prerequisites are met (eg, head is engaged, membranes ruptured, presentation and position known, anesthesia is satisfactory, the fetus is of appropriate gestational age and size, maternal bladder is empty) and there are no contraindications. (See 'Prerequisites' above and 'Contraindications' above.)

●Classification

•Forceps deliveries are classified as outlet, low, or mid, depending on the fetal station (figure 1) and degree of head rotation. (See 'Classification' above.)

•Midforceps deliveries are generally avoided because they are associated with a higher rate of severe perinatal morbidity/mortality compared with cesarean births performed in the second stage, especially when the indication for delivery is dystocia. They are also associated with higher rates of maternal trauma. However, clinicians highly experienced with forceps assisted birth may choose to attempt a midforceps delivery in select circumstances, such as sudden severe fetal or maternal compromise, if the clinician believes that they can safely expedite and effect a safe assisted vaginal birth. Simultaneous preparation for cesarean birth should be underway. (See 'Forceps' above.)

●Choice of instrument – In general, vacuum extraction is generally less traumatic for the mother than forceps assisted birth and the devices are easier to apply and require less maternal anesthesia than forceps. The advantages of forceps are that they are associated with a higher success rate, are unlikely to detach from the head, can be sized to a premature cranium, may be used for a rotation, and do not aggravate bleeding from scalp lacerations. (See 'When to choose vacuum versus forceps' above.)

For patients who are to undergo an assisted vaginal birth at >34 weeks and have a high likelihood of success (eg, outlet procedure), we suggest use of vacuum over forceps (Grade 2C). Maternal morbidity is lower with vacuum than forceps, and neonatal morbidity is likely to be low in this setting with either approach. When success is uncertain, primary use of forceps may reduce the likelihood of failure, as use of sequential instruments should be avoided. (See 'Choice of instrument' above and 'Second attempt with a different instrument' above.)

●Procedure

•Checklist – We suggest using a checklist before performing the procedure and documenting the procedure after it has been completed (table 1)

•Personnel and location – Staff and providers needed in the event of an emergency cesarean should be notified and nearby. If the procedure is anticipated to be more difficult, such as from a midstation, then an attempt at assisted vaginal birth in the operating room may allow for a more expedited cesarean if the attempt fails. A pediatric provider should also be notified and readily available. (See 'Overview' above.)

•Ultrasound – If fetal presentation or position is uncertain, intrapartum sonographic visualization of fetal intracranial structures can be used to determine fetal head position and is more accurate than digital examination. (See 'Patient preparation' above.)

•Episiotomy – We recommend not performing an episiotomy routinely (Grade 1B). If episiotomy is performed in selected patients (eg, forceps-assisted vaginal birth in a patient with no prior vaginal births, assisted birth of persistent occiput posterior), we suggest a mediolateral or lateral episiotomy (Grade 2B).

•Forceps position – The tips of the blades should lie over the cheeks (figure 3).

•When to abandon the procedure – The decision to proceed with assisted vaginal birth is ongoing and decided moment by moment based on assessment of the success of the various steps in the procedure. Assisted vaginal birth should be abandoned if it is difficult to apply the instrument, descent does not easily proceed with traction, or the newborn has not been delivered within a reasonable time (eg, 15 to 20 minutes) or after three pulls with no progress. (See 'When to abandon the procedure' above.)

•Antibiotics – For patients who are going to have an assisted vaginal birth, we suggest not administering prophylactic antibiotics before the birth (Grade 2C).

After an assisted vaginal birth in which an episiotomy was performed, we suggest administering a broad-spectrum antibiotic (Grade 2C). Antibiotic administration in this population appears to reduce the risk for perineal infection and wound breakdown. We use a single dose of intravenous co-amoxiclav. However, use of a different broad spectrum antibiotic or no antibiotic prophylaxis is also reasonable, given limited data of efficacy, no comparative data from use of other antibiotics, and no long-term data regarding harms. (See 'Antibiotics' above.)

●Complications

•Maternal and fetal/neonatal complication rates vary widely and depend on a number of factors, which are not independent. Virtually all complications associated with assisted vaginal birth can also occur in the course of a spontaneous vaginal birth, but the incidence is lower in the latter. (See 'Complications' above.)

•Vacuum-assisted births are associated with an increased risk of neonatal cephalohematomata and retinal hemorrhage (figure 4) compared with forceps or spontaneous births. These complications generally resolve without sequelae. Forceps-assisted births cause more acute maternal injury and fetal facial nerve injury than vacuum-assisted or spontaneous births. (See 'Vacuum-assisted births' above and 'Forceps-assisted births' above.)

•Although short-term neonatal morbidity varies between procedures, developmental outcome appears to be equivalent for both forceps- and vacuum-assisted births. (See 'Neurodevelopmental outcome' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Elisabeth K Wegner, MD, who contributed to earlier versions of this topic review.