CHAPTER 27: Vaginal Delivery

As described by Williams, the natural culmination of second-stage labor is controlled vaginal delivery of a healthy neonate with minimal trauma to the mother. Vaginal delivery is the preferred route of delivery for most fetuses, although various clinical settings may favor cesarean delivery. Of delivery routes, spontaneous vaginal vertex delivery poses the lowest risk of most maternal comorbidity, and comparisons with cesarean delivery are found in Chapter 30 (Cesarean Delivery Risks). Delivery is usually spontaneous, although some maternal or fetal complications may warrant operative vaginal delivery, described in Chapter 29 (Indications). Last, a malpresenting fetus or multifetal gestation in many cases may be delivered vaginally but requires special techniques. These are described in Chapters 28 (Labor and Delivery Management) and 45 (Evaluation of Fetal Presentation).

The end of second-stage labor is heralded as the perineum begins to distend, the overlying skin becomes stretched, and the fetal scalp is seen through the separating labia. Increased perineal pressure from the fetal head creates reflexive bearing-down efforts, which are encouraged when appropriate. At this time, preparations are made for delivery. If the bladder is distended, catheterization may be necessary. Continued attention is also given to fetal heart rate monitoring. As one example, a nuchal cord often tightens with descent and may lead to deepening variable decelerations.

During second-stage labor, pushing positions may vary. But for delivery, the dorsal lithotomy position is most common and often the most satisfactory. For better exposure, leg holders or stirrups are used. Corton and associates (2012) found no increased rates of perineal lacerations with or without their use. With positioning, legs are not separated too widely or placed one higher than the other. Within the leg holder, the popliteal region should rest comfortably in the proximal portion and the heel in the distal portion. The legs are not strapped into the stirrups, thereby allowing quick flexion of the thighs backward onto the abdomen should shoulder dystocia develop. Legs may cramp during second-stage pushing, and cramping is relieved by repositioning the affected leg or by brief massage.

Preparation for delivery includes vulvar and perineal cleansing. If desired, sterile drapes may be placed in such a way that only the immediate area around the vulva is exposed. Scrubbing, gowning, gloving, and donning protective mask and eyewear protect both the laboring woman and accoucheur from infectious agents.

Delivery of the Head

By the time of perineal distention, the position of the occiput is usually known. In some cases, however, molding and caput formation may have precluded early accurate identification. At this time, careful assessment is again performed as described in Chapter 22 (Vaginal Examination). In most cases, position is directly occiput anterior (OA) or is rotated slightly oblique. But, in perhaps 5 percent, occiput posterior (OP) positioning persists.

With each contraction, the vulvovaginal opening is dilated by the fetal head to gradually form an ovoid and finally, an almost circular opening (Fig. 27-1). This encirclement of the largest head diameter by the vulvar ring is termed crowning. The perineum thins and may spontaneously lacerate. The anus becomes greatly stretched, and the anterior wall of the rectum can easily be seen through it.

Routine episiotomy is no longer recommended, and selective use aims to enlarge the vaginal opening for specific indications (Episiotomy). To limit spontaneous vaginal laceration, some perform antenatal massage of the perineal body to increase perineal distensibility or intrapartum perineal massage to widen the introitus for head passage. During massage with a lubricant, the perineum is grasped in the midline by both hands using the thumb and opposing fingers. Outward and lateral stretching to thin the perineum is repeatedly performed. But in randomized studies, this technique did not significantly prevent perineal laceration (Beckmann, 2013; Mei-dan, 2008; Stamp, 2001). Antepartum use of the Epi-No intravaginal pump balloon has a similar aim, but it also fails to prevent perineal trauma or levator injury (Brito, 2015; Kamisan Atan, 2016).

When the head distends the vulva and perineum enough to open the vaginal introitus to a diameter of 5 cm or more, a gloved hand may be used to support the perineum (Fig. 27-2). The other hand is used to guide and control the fetal head to deliver the smallest head diameter through the introitus and to avoid expulsive delivery. Slow delivery of the head may decrease lacerations (Laine, 2008). Overall, bracing the perineum lowers rates of anal sphincter injury compared with a “hands off” approach to delivery (Bulchandani, 2015; McCandlish, 1998).

Alternatively, if expulsive efforts are inadequate or expeditious delivery is needed, the modified Ritgen maneuver may be employed or an episiotomy cut. With the modified maneuver, gloved fingers beneath a draped towel exert forward pressure on the fetal chin through the perineum just in front of the coccyx. Concurrently, the other hand presses against the occiput (Fig. 27-3). Originally described in 1855, the maneuver allows controlled fetal head delivery (Cunningham, 2008). It also favors neck extension so that the head passes through the introitus and over the perineum with its smallest diameters. Comparing the Ritgen maneuver with simple perineal support in 1623 women, Jönsson and colleagues (2008) found a similar incidence of third- and fourth-degree tears, defined later (Immediate Postpartum Care).

Delivery of the Shoulders

Following delivery of the fetal head, a finger is passed across the fetal neck to determine whether it is encircled by one or more umbilical cord loops. The nuchal cord incidence increases with gestational age and is found in nearly 25 percent of deliveries at term (Larson, 1997; Ogueh, 2006). If an umbilical cord coil is felt, it is slipped over the head if loose enough. If applied too tightly, the loop is cut between two clamps. Tight nuchal cords complicate approximately 6 percent of all deliveries but are not associated with worse neonatal outcome than those without a cord loop (Henry, 2013).

Following its delivery, the fetal head falls posteriorly, bringing the face almost into contact with the maternal anus. The occiput promptly turns toward one of the maternal thighs, and the head assumes a transverse position. This external rotation indicates that the bisacromial diameter, which is the distance between the shoulders, has rotated into the anteroposterior diameter of the pelvis.

Most often, the shoulders appear at the vulva just after external rotation and are born spontaneously. If delayed, extraction aids controlled delivery. The sides of the head are grasped with two hands, and gentle downward traction is applied until the anterior shoulder appears under the pubic arch (Fig. 27-4). Next, by an upward movement, the posterior shoulder is delivered. During delivery, abrupt or powerful force is avoided to avert fetal brachial plexus injury.

The rest of the body almost always follows the shoulders without difficulty. With prolonged delay, however, its birth may be hastened by moderate outward traction on the head and moderate pressure on the uterine fundus. Hooking the fingers in the axillae is avoided. This can injure upper extremity nerves and produce a transient or possibly permanent paralysis. Immediately after delivery of the newborn, a gush of amnionic fluid that is often blood-tinged but not grossly bloody usually follows.

Previously, immediate nasopharyngeal bulb suctioning of the newborn was routine to remove secretions. It was found, however, that suctioning of the nasopharynx may lead to neonatal bradycardia (Gungor, 2006). The current American Heart Association neonatal resuscitation recommendations eschew most suctioning immediately following birth—even with meconium present (Chap. 33, Neonatal Encephalopathy and Cerebral Palsy). And with meconium-stained fluid, routine intubation for tracheal suction is not recommended for vigorous or for nonvigorous neonates. Suctioning is reserved for neonates who have obvious obstruction to spontaneous breathing or who require positive-pressure ventilation (Wyckoff, 2015). For suctioning, options are bulb syringe or suction catheter aspiration and may include intubation and suctioning if the airway is obstructed.

Cord Clamping

The umbilical cord is cut between two clamps placed 6 to 8 cm from the fetal abdomen, and later an umbilical cord clamp is applied 2 to 3 cm from its insertion into the fetal abdomen.

For term neonates, the timing of umbilical cord clamping remains debatable. Delayed umbilical cord clamping transfers a greater volume of blood to the newborn. A delay for up to 60 seconds may increase total body iron stores, expand blood volume, and decrease anemia incidence in the neonate (Andersson, 2011; Yao, 1974). This practice may be particularly valuable in populations in which iron deficiency is prevalent (Kc, 2017; World Health Organization, 2014).

Conversely, a higher hemoglobin concentration increases risks for hyperbilirubinemia and extended hospitalization for neonatal phototherapy (McDonald, 2013). Delayed cord clamping may also hinder timely and needed neonatal resuscitation. That said, early pilot studies are assessing the value of resuscitating newborns at the bedside to permit delayed clamping (Katheria, 2017; Winter 2017). Fortunately, in general, delayed umbilical cord clamping compared with early clamping does not worsen Apgar scores, umbilical cord pH, or respiratory distress caused by polycythemia. Regarding maternal outcomes, rates of postpartum hemorrhage are similar between early and delayed clamping groups (Andersson, 2013). Fewer data are available regarding cord “milking,” in which the operator pushes blood through the cord toward the newborn. This maneuver appears safe and may be advantageous if rapid cord clamping is clinically indicated (Upadhyay, 2013).

For the preterm neonate, delayed cord clamping has several benefits. These include higher red cell volume, decreased need for blood transfusion, and lower rates of intraventricular hemorrhage and of necrotizing enterocolitis (Backes, 2014; Rabe, 2012). For neonates who require expedited resuscitation, cord milking may have benefits to quickly transfer volume (Al-Wassia, 2015; Katheria, 2015; Patel, 2014). Still, because of rapid blood volume changes, the American Heart Association currently suggests against the routine use of cord milking for neonates born <29 weeks’ gestation (Wyckoff, 2015).

The American College of Obstetricians and Gynecologists (2017a) notes sufficient evidence to support delayed umbilical cord clamping for term and preterm neonates for at least 30 to 60 seconds after birth. This opinion is also endorsed by the American Academy of Pediatrics (2017a). The American Heart Association guidelines advise that the practice may have benefits for term or preterm neonates not needing immediate resuscitation at birth (Wyckoff, 2015).

Occiput Transverse Position

In the absence of a pelvic architecture abnormality or asynclitism, the occiput transverse (OT) position is usually transitory. Thus, unless contractions are hypotonic, the head usually spontaneously rotates to an OA position. If rotation ceases because of poor expulsive forces, vaginal delivery usually can be accomplished readily in several ways. The easiest is manual rotation of the occiput either anteriorly to OA or less commonly, posteriorly to occiput posterior. If either is successful, Le Ray and coworkers (2007) reported a 4-percent cesarean delivery rate compared with a 60-percent rate in women in whom manual rotation was not successful. Some recommend rotation with Kielland forceps for the persistent OT position as outlined in Chapter 29 (Occiput Transverse Positions). These forceps are used to rotate the occiput to the anterior position, and delivery is accomplished with the same forceps or by substitution with Simpson, Tucker–McLane, or similar forceps.

In some cases, pelvic shape leads to a persistent OT position that is not easily overcome. For example, a platypelloid pelvis is flattened anteroposteriorly and an android pelvis is heart shaped. With these, space may be inadequate for occipital rotation to either an anterior or posterior position (Fig. 2-17). Because of these concerns, undue force is avoided if forceps delivery is attempted.

Approximately 2 to 10 percent of singleton term cephalic fetuses deliver in an occiput posterior (OP) position (Cheng, 2010). Many fetuses delivering OP are OA in early labor and reflect malrotation during labor. Predisposing risks include epidural analgesia, nulliparity, greater fetal weight, and prior delivery with OP positioning (Cheng, 2006a; Gardberg, 2004; Lieberman, 2005). Regarding pelvic shape, an anthropoid pelvis and narrow subpubic angle can predispose (Barth, 2015; Ghi, 2016).

Morbidity

Women with a persistent OP position have higher associated rates of prolonged second-stage labor, cesarean delivery, and operative vaginal delivery. For women who deliver vaginally, rates of blood loss and of third- and fourth-degree lacerations are increased (Senécal, 2005).

Newborns delivered from an OP position have higher complication rates then those born positioned OA. Cheng and coworkers (2006b) compared outcomes of 2591 women undergoing delivery with a persistent OP position with those of 28,801 women whose newborns were delivered OA. Virtually every possible delivery complication was found more frequently with persistent OP position. Only 46 percent of these women delivered spontaneously, and the remainder accounted for 9 percent of cesarean deliveries performed. These investigators also found that an OP position at delivery was associated with more adverse short-term neonatal outcomes that included acidemic umbilical cord gases, birth trauma, Apgar scores <7, and intensive care nursery admission, among others. Similar results were reported by Ponkey (2003) and Fitzpatrick (2001) and their associates.

Methods to prevent persistent OP position and its associated morbidity have been investigated. First, digital examination for identification of fetal head position can be inaccurate, and transabdominal sonography can be used to increase accuracy (Dupuis, 2005; Zahalka, 2005). The transducer is placed transversely just cephalad to the maternal mons pubis. In the sonogram, fetal orbits and nasal bridge lie ventrally, whereas the occiput apposes the lower sacrum. Such information may provide an explanation for prolonged second-stage labor or may identify suitable candidates for rotation. Of other possible interventions, varying maternal position either antepartum or during labor does not appear to lower rates of persistent OP position (Desbriere, 2013; Kariminia, 2004; Le Ray, 2016).

Delivery

The fetus in an OP position may be delivered either spontaneously or by operative vaginal delivery. First, if the bony pelvic outlet is roomy and the perineum is somewhat relaxed from prior deliveries, rapid spontaneous OP delivery will often take place. Conversely, if the perineum is resistant to stretch, second- stage labor may be appreciably prolonged. During each expulsive effort, the head is driven against the perineum to a much greater degree than when the head position is OA. This leads to greater rates of third- and fourth-degree lacerations (Groutz, 2011; Melamed, 2013).

In some cases, spontaneous vaginal delivery from an OP position does not appear feasible or expedited delivery is needed. Here, manual rotation with spontaneous delivery from an OA position may be preferred. This technique is described fully in Chapter 29 (Occiput Posterior Positions). Successful rotation rates range from 47 to 90 percent. And, as would be expected, lower rates of cesarean delivery, vaginal laceration, and maternal blood loss follow rotation to OA position and vaginal delivery (Le Ray, 2005; Sen, 2013; Shaffer, 2006, 2011). Disadvantageously, manual rotation is linked with higher cervical laceration rates. Thus, careful inspection of the cervix following rotation is mandatory.

For exigent delivery, forceps or vacuum device can be applied to a persistent OP position. This is often performed in conjunction with an episiotomy. Also, if the head is engaged, the cervix fully dilated, and the pelvis adequate, forceps rotation may be attempted for those with suitable skills. These operative vaginal techniques are detailed in Chapter 29 (Occiput Transverse Positions).

Infrequently, protrusion of fetal scalp through the introitus is the consequence of marked elongation of the fetal head from molding combined with formation of a large caput succedaneum. In some cases, the head may not even be engaged—that is, the biparietal diameter may not have passed through the pelvic inlet. In these, labor is characteristically long and descent of the head is slow. Careful palpation above the symphysis may disclose the fetal head to be above the pelvic inlet. Prompt cesarean delivery is appropriate.

At Parkland Hospital, spontaneous delivery or manual rotation is preferred for management of persistent OP position. When needed, either manual rotation to OA position followed by forceps delivery or forceps delivery from the OP position is used. If neither can be completed with ease and safety, cesarean delivery is performed.

Following complete emergence of the fetal head during vaginal delivery, the remainder of the body may not rapidly follow. The anterior fetal shoulder can become wedged behind the symphysis pubis and fail to deliver using normally exerted downward traction and maternal pushing. Because the umbilical cord is compressed within the birth canal, this dystocia is an emergency. Several maneuvers, in addition to downward traction on the fetal head and neck, may be performed to free the shoulder. This requires a team approach, in which effective communication and leadership are critical.

Consensus regarding a specific definition of shoulder dystocia is lacking. Some focus on whether maneuvers to free the shoulder are needed, whereas others use the head-to-body delivery time interval as defining (Beall, 1998). Spong and coworkers (1995) reported that the mean head-to-body delivery time in normal births was 24 seconds compared with 79 seconds in those with shoulder dystocia. These investigators proposed that a head-to-body delivery time >60 seconds be used to define shoulder dystocia. Currently, however, the diagnosis continues to rely on the clinical perception that the normal downward traction needed for fetal shoulder delivery is ineffective.

Because of these differing definitions, the incidence of shoulder dystocia varies. One recent review cites a clinically useful average of 1 percent of all deliveries (Ouzounian, 2016). The incidence has increased in recent decades, likely due to increasing fetal birthweight (MacKenzie, 2007; Øverland, 2014). Increased identification and documentation may also raise the incidence (Kim, 2016).

Maternal and Neonatal Consequences

In general, shoulder dystocia poses greater risk to the fetus than to the mother. The main maternal risks are serious perineal tears and postpartum hemorrhage, usually from uterine atony but also from lacerations (Gauthaman, 2016; Rahman, 2009). In contrast, significant neonatal neuromusculoskeletal injury and asphyxia are concerns. These specific injuries are described in Chapter 33 (Spinal Cord Injury). In one review of 1177 shoulder dystocia cases, brachial plexus injury was diagnosed in 11 percent and clavicular or humeral fracture in 2 percent (Chauhan, 2014). MacKenzie and associates (2007) reviewed 514 cases. Of the neonates, 7 percent showed evidence of acidosis at delivery, and 1.5 percent required cardiac resuscitation or developed hypoxic ischemic encephalopathy (HIE). In another review of 200 cases, rates of severe fetal acidosis and HIE were each 0.5 percent if delivery was completed within 5 minutes. These rates rose to 6 and 24 percent, respectively, with delivery delays ≥5 minutes (Leung, 2011a).

Prediction and Prevention

Fetal macrosomia, maternal obesity, prolonged second-stage labor, and a prior event raise risks for shoulder dystocia (Mehta, 2004; Overland, 2009; Schummers, 2015). Although these factors are clearly associated with this complication, identification of individual instances before the fact has proved to be impossible. The American College of Obstetricians and Gynecologists (2017c) reviewed studies and concluded that:

1. Most cases of shoulder dystocia cannot be accurately predicted or prevented.

2. Elective induction of labor or elective cesarean delivery for all women suspected of having a macrosomic fetus is not appropriate.

3. Planned cesarean delivery may be considered for the nondiabetic woman with a fetus whose estimated fetal weight is >5000 g or for the diabetic woman whose fetus is estimated to weigh >4500 g.

Birthweight

There is a corresponding rise in the incidence of shoulder dystocia with increasing birthweight (Acker, 1985; Øverland, 2012; Stotland, 2004). Commonly cited maternal characteristics associated with increased fetal birthweight are obesity, postterm pregnancy, multiparity, and diabetes (Jolly, 2003; Koyanagi, 2013). The combination of fetal macrosomia and maternal diabetes mellitus escalates the frequency of shoulder dystocia (Langer, 1991; Nesbitt, 1998). This predisposition may stem from the fact that fetuses of diabetic women have larger shoulder and extremity circumferences and greater shoulder-to-head and chest-to-head size differences relative to comparable-weight fetuses of nondiabetic mothers (McFarland, 1998; Modanlou, 1982). That said, translating these specific measurements into stand-alone sonographic clinical thresholds has shown poor predictive sensitivity (Burkhardt, 2014).

Preventively, early labor induction has yielded conflicting results. In one study, approximately 800 women with suspected macrosomic fetuses were randomized either to early induction between 37 and 39 weeks or to expectant care (Boulvain, 2015). Dystocia rates were lowered by two thirds in the intervention group, and neither group suffered brachial plexus injury. Although not measured, this practice is balanced against morbidity of early delivery. Moreover, the poor accuracy of antepartum fetal weight prediction should be considered as well (Hoopmann, 2010; Malin, 2016; Noumi, 2005). In contrast, an earlier randomized study of 284 women showed that rates of shoulder dystocia were not lowered by early induction at 38 weeks (Gonen, 1997).

As previously discussed, cesarean delivery may be considered to prevent shoulder dystocia. That said, Rouse and Owen (1999) concluded that a prophylactic cesarean delivery policy for macrosomic fetuses would require more than 1000 cesarean deliveries with attendant morbidity to avert a single permanent brachial plexus injury.

Prior Shoulder Dystocia

The risk of recurrent shoulder dystocia ranges from 1 to 13 percent (Bingham, 2010; Moore, 2008; Ouzounian, 2013). For many women with prior shoulder dystocia, a trial of labor may be reasonable. The American College of Obstetricians and Gynecologists (2017c) recommends that estimated fetal weight, gestational age, maternal glucose intolerance, and severity of prior neonatal injury be evaluated and risks and benefits of cesarean delivery discussed with any woman with a history of shoulder dystocia. After discussion, either mode of delivery may be appropriate.

Management

Because shoulder dystocia cannot be accurately predicted, clinicians should be well versed in its management principles. Because of ongoing cord compression with this dystocia, one goal is to reduce the head-to-body delivery time. This is balanced against the second goal, which is avoiding fetal and maternal injury from aggressive manipulations. Accordingly, an initial gentle attempt at traction, assisted by maternal expulsive efforts, is recommended. Adequate analgesia is certainly ideal. Some clinicians advocate performing a large episiotomy to provide room for manipulations. Episiotomy itself does not lower brachial plexus injury rates but raises third- and fourth-degree laceration rates (Gurewitsch, 2004; Paris, 2011; Sagi-Dain, 2015). Episiotomy may be elected to complete needed maneuvers.

After gentle traction, various techniques can be used to free the anterior shoulder from its impacted position behind the symphysis pubis. A more detailed discussion of these and the topic is found in Cunningham and Gilstrap’s Operative Obstetrics, 3rd edition (Cunningham, 2017). Of these, moderate suprapubic pressure can be applied by an assistant, while downward traction is applied to the fetal head. Pressure is applied with the heel of the hand to the anterior shoulder wedged above and behind the symphysis. The anterior shoulder is thus either depressed or rotated, or both, so the shoulders occupy the oblique plane of the pelvis. Here, the anterior shoulder can be freed.

The McRoberts maneuver is often selected next if additional steps are needed. The maneuver consists of removing the legs from the stirrups and sharply flexing them up toward the abdomen. Suprapubic pressure is often concurrently applied (Fig. 27-5). Gherman and associates (2000) analyzed the McRoberts maneuver using x-ray pelvimetry. They found that the procedure caused straightening of the sacrum relative to the lumbar vertebrae, rotation of the symphysis pubis toward the maternal head, and a decrease in the angle of pelvic inclination. Although this does not increase pelvic dimensions, pelvic rotation cephalad tends to free the impacted anterior shoulder. Gonik and coworkers (1989) tested the McRoberts position objectively with laboratory models and found that the maneuver reduced the forces needed to free the fetal shoulder. If unsuccessful, most move next either to free the posterior shoulder or to rotate the bisacromial diameter into one of the oblique diameters of the maternal pelvis.

With delivery of the posterior shoulder, the accoucheur carefully sweeps the posterior arm of the fetus across its chest, followed by delivery of the arm (Fig. 27-6). If possible, the operator’s fingers are aligned parallel to the long axis of the fetal humerus to lower bone fracture risks. The shoulder girdle is then rotated into one of the oblique diameters of the pelvis with subsequent delivery of the anterior shoulder.

Of rotational maneuvers, Woods (1943) reported that by progressively rotating the posterior shoulder 180 degrees in a corkscrew fashion, the impacted anterior shoulder could be released. This is frequently referred to as the Woods corkscrew maneuver (Fig. 27-7). Rubin (1964) recommended two maneuvers. First, the fetal shoulders are rocked from side to side by applying force to the maternal abdomen. If this is not successful, the pelvic hand reaches the most easily accessible fetal shoulder, which is then pushed toward the anterior surface of the chest. This maneuver most often abducts both shoulders, which in turn produces a smaller bisacromial diameter. This permits displacement of the anterior shoulder from behind the symphysis (Fig. 27-8).

If the above are initially unsuccessful, they may be repeated, and finally other methods may be elected. With an all-fours maneuver, also called the Gaskin maneuver, the parturient rolls onto her knees and hands. Here, downward traction against the head and neck attempts to free the posterior shoulder (Bruner, 1998). Challenges with this include immobility from regional analgesia and time lost in patient repositioning.

In some, the posterior arm is inaccessible for delivery. Cluver and Hofmeyr (2009) described posterior axilla sling traction to deliver the posterior arm. With this alternative method, a suction catheter is threaded under the axilla and both ends are brought together above the shoulder. Upward and outward traction on the catheter loop delivers the shoulder. In a small series of 19 cases, this maneuver was successful in 18 cases. However, neonatal injury included three cases of humeral fracture and one permanent and four transient cases of Erb palsy (Cluver, 2015).

Deliberate fracture of the anterior clavicle using the thumb to press it toward and against the pubic ramus can be attempted to free the shoulder impaction. In practice, however, deliberate fracture of a large neonate’s clavicle is difficult. If successful, the fracture will heal rapidly and is usually trivial compared with brachial nerve injury, asphyxia, or death.

The Zavanelli maneuver involves replacement of the fetal head into the pelvis followed by cesarean delivery (Sandberg, 1985). Terbutaline, 0.25 mg, is given subcutaneously to produce uterine relaxation. The first part of the maneuver consists of returning the head to an OA or OP position. The operator flexes the head and slowly pushes it back into the vagina. Cesarean delivery is then performed. Sandberg (1999) reviewed 103 reported cases. It was successful in 91 percent of cephalic cases and in all cases of breech head entrapments. Despite successful replacement, fetal injuries were common but may have resulted from the multiple manipulations used before the Zavanelli maneuver (Sandberg, 2007).

Symphysiotomy, in which the intervening symphyseal cartilage and much of its ligamentous support is cut to widen the symphysis pubis, is described in Chapter 28 (Total Breech Extraction). It has been used successfully for shoulder dystocia (Goodwin, 1997; Hartfield, 1986). Maternal morbidity can be significant due to urinary tract injury. Cleidotomy consists of cutting the clavicle with scissors or other sharp instruments and is usually done for a dead fetus (Schramm, 1983).

Shoulder Dystocia Drill

Hernandez and Wendel (1990) suggest use of a shoulder dystocia drill to better organize emergency management:

1. Call for help—mobilize assistants and anesthesia and pediatric personnel. Initially, a gentle attempt at traction is made. Drain the bladder if it is distended.

2. A generous episiotomy may be desired at this time to afford room posteriorly.

3. Suprapubic pressure is used initially by most practitioners because it has the advantage of simplicity. Only one assistant is needed to provide suprapubic pressure, while normal downward traction is applied to the fetal head.

4. The McRoberts maneuver requires two assistants. Each assistant grasps a leg and sharply flexes the maternal thigh toward the abdomen.

   These maneuvers will resolve most cases of shoulder dystocia.

   If the above listed steps fail, the following steps may be attempted, and any of the maneuvers may be repeated:

5. Delivery of the posterior arm is attempted. With a fully extended arm, however, this is usually difficult to accomplish.

6. Woods screw maneuver is applied.

7. Rubin maneuver is attempted.

The American College of Obstetricians and Gynecologists (2017c) has concluded that no one maneuver is superior to another in releasing an impacted shoulder or reducing the chance of injury. Performance of the McRoberts maneuver, however, is deemed a reasonable initial approach. In one review of more than 2000 cases, Hoffman and colleagues (2011) noted an 84-percent success rate with posterior shoulder delivery and comparable rates of neonatal injury compared with other standard methods. This contrasts with a review of 205 cases, in which posterior shoulder delivery yielded greater neonatal injury rates than rotational methods (Leung, 2011b). Spain and associates (2015) found that duration rather than a specific maneuver increased neonatal injury.

Importantly, progression from one maneuver to the next should be organized and methodical. As noted, the urgency to relieve the dystocia should be balanced against potentially injurious traction forces and manipulations. Lerner and coworkers (2011) in their evaluation of 127 shoulder dystocia cases reported that all neonates without sequelae from shoulder dystocia were born by 4 minutes. Conversely, most depressed neonates–57 percent–had head-to-body delivery intervals >4 minutes. The percentage of depressed neonates rose sharply after 3 minutes.

Shoulder dystocia training and protocols using simulation-based education and drills has evidence-based support. These tools improve performance and retention of drill steps (Buerkle, 2012; Crofts, 2008; Grobman, 2011). Their use has translated into improved neonatal outcome in some, but not all, investigations (Crofts, 2016; Fransen, 2017; Kim, 2016; Walsh, 2011). The American College of Obstetricians and Gynecologists (2012) also has created a Patient Safety Checklist to guide the documentation process with shoulder dystocia.

Home Birth

In 2014, 0.7 percent of deliveries in the United States were planned home births and 0.2 percent were unplanned (MacDorman, 2016). Of unplanned births in a 15-year epoch in Norway, 69 of 6027 or 1.1 percent resulted in fetal or neonatal death. This high rate was attributable to infection, prematurity, and placental abruption (Gunnarsson, 2017). Multiparity and distance from the hospital were ascribed risks (Gunnarsson, 2014). In the United States, youth, lack of prenatal care, minority race, and lower educational attainment were associated risks for unplanned home birth (Declercq, 2010).

In contrast, the demographics of women choosing planned home birth in the United States favor those who are white, nonsmoking, self-pay, college-educated, and multiparous (MacDorman, 2016). As perceived benefits, planned delivery at home for those with low-risk pregnancies results in fewer medical interventions that include labor augmentation, episiotomy, operative vaginal delivery, and cesarean delivery (Bolten, 2016; Cheyney, 2014). Regarding the safety of planned home birth, data from randomized trial are lacking, and large observational studies derive from heterogeneous care systems, whose results may not be generalizable. For example, several developed countries deliver at home a large volume of carefully screened women, delivered by midwives with substantial training and in a setting closely integrated with the local health-care system (Birthplace in England Collaborative Group, 2011; de Jonge, 2015; Hutton, 2016). The level of such coordination in the United States is less uniform.

Overall, risks of home births in the United States are small but greater than those of hospital delivery. Midwife-attended home births carry a neonatal mortality risk of 1.3 per 1000 births. This is a nearly fourfold greater rate compared with midwife-attended hospital births. The most common underlying causes of death are those attributed to labor and delivery events, to congenital anomalies, and to infection. Of neonatal injuries, rates of neonatal seizure and serious neurological dysfunction are similarly elevated in home-birth groups (Grünebaum, 2013, 2014, 2017; Snowden, 2015; Wasden, 2016). Importantly, substantial risks attend home birth for those with prior cesarean delivery, with breech presentation, and with multifetal gestation (Cheyney, 2014; Cox, 2015). The American College of Obstetricians and Gynecologists (2017b) considers these to be absolute contraindications. Further, the College considers accredited hospitals and birthing centers to be the safest site for birth but recognizes the autonomy of the well-counseled patient.

Water Birth

As one option for pain relief, some women choose to spend part of first-stage labor in a large water tub. With this practice, one Cochrane review found lower rates of anesthesia block use and no greater adverse neonatal or maternal effects compared with traditional labor (Cluett, 2009).

For delivery, however, water birth carries greater concern for neonatal harm and without proven benefits. Case reports describe aspiration leading to fresh-water drowning (Pinette, 2004). The risk of cord avulsion during water birth approximates 3 per 1000 births, and stems primary from abruptly bringing the newborn out of the water (Schafer, 2014). Last, case reports also enumerate serious infections, which emphasize the need for rigorous sanitizing protocols. That said, in most large studies comparing land and water births, overall maternal or neonatal infection rates are not increased (Bovbjerg, 2016; Burns, 2012; Thoeni, 2005). In sum, several reviews comment on study shortcomings and isolated complications but do not identify definitive evidence for overall greater rates of neonatal harm from water birth in low-risk populations (Davies, 2015; Taylor, 2016). However, given the paucity of robust data and potential for serious complications, the American College of Obstetricians and Gynecologists (2016a) currently recommend that “birth occur on land, not in water.”

Female Genital Mutilation

This practice refers to medically unnecessary vulvar and perineal modification. In the United States, it is a federal crime to perform unnecessary genital surgery on a girl younger than 18 years. That said, forms of female genital mutilation are practiced in countries throughout Africa, the Middle East, and Asia. As many as 200 million women worldwide have undergone one of these procedures, and approximately 513,000 girls in the United States were at risk for this practice in 2012 (Goldberg, 2016; UNICEF, 2016). Cultural sensitivity is imperative, because many women may be offended by the suggestion that they have been assaulted or mutilated (American College of Obstetricians and Gynecologists, 2014).

The World Health Organization (2008) classifies genital mutilations into four types (Table 27-1). Long-term complications from surgery and its associated scarring include infertility, genital pain, diminished sexual quality of life, and propensity for urogenital infection (Almroth, 2005; Andersson, 2012; Nour, 2015). In general, women with significant symptoms following type III procedures are candidates for corrective surgery. Specifically, division of midline scar tissue to reopen the vulva is termed deinfibulation.

Female genital mutilation has been associated with some adverse maternal and neonatal complications. The World Health Organization (2006, 2008) estimated that these procedures increased perinatal morbidity rates by 10 to 20 per 1000. Small increased risks for prolonged labor, cesarean delivery, and postpartum hemorrhage are also found (Berg, 2014; Chibber, 2011; Wuest, 2009). Importantly, the psychiatric consequences can be profound.

To prevent obstetrical complications, deinfibulation can be performed either antepartum or intrapartum (Fig. 27-9) (Esu, 2017). In women not undergoing deinfibulation, anal sphincter tear rates with vaginal delivery may be increased (Berggren, 2013; Rodriguez, 2016). In our experiences, intrapartum deinfibulation in many cases allows successful vaginal delivery without major complications.

Prior Pelvic Reconstructive Surgery

These surgeries are performed with increasing frequency in reproductive-aged women, and thus pregnancy following these procedures is not uncommon. Logically, there are concerns for symptom recurrence following vaginal delivery, and high-quality data to aid evidenced-based decisions are limited. For women with prior stress urinary incontinence surgery, slightly greater protection against postpartum incontinence is gained by elective cesarean delivery (Pollard, 2012; Pradhan, 2013). Stated another way, most women with prior corrective surgery for incontinence can be delivered vaginally without symptom recurrence. Also, cesarean delivery is not always protective. Obviously, symptom recurrence and the need for additional vaginal surgery should be weighed against the surgical risk of cesarean delivery (Groenen, 2008). In those with prior surgeries for anal incontinence or pelvic organ prolapse, only scant information regarding outcomes is available. Such cases require individualization.

Anomalous Fetuses

Rarely, delivery can be obstructed by extreme hydrocephaly, by body stalk anomaly, or by massive fetal abdominal enlargement from a greatly distended bladder, ascites, or organomegaly (Costa, 2012; Sikka, 2011). With milder forms of hydrocephaly, if the biparietal diameter is <10 cm or if the head circumference is <36 cm, then vaginal delivery may be permitted (Anteby, 2003).

In rare cases in which neonatal death has occurred or is certain due to associated anomalies, vaginal delivery may be reasonable, but the head or abdomen must be reduced in size for delivery. Removal of fluid by cephalocentesis or paracentesis with sonographic guidance can be performed intrapartum. As described in Shoulder Dystocia Drill, cleidotomy can shorten the bisacromial diameter. For hydrocephalic fetuses that are breech, cephalocentesis can be accomplished suprapubically when the aftercoming head enters the pelvis. Currently, these practices are more germane in developing countries.

Delivery of the Placenta

Third-stage labor begins immediately after fetal birth and ends with placental delivery. Goals include delivery of an intact placenta and avoidance of uterine inversion or postpartum hemorrhage. The latter two are grave intrapartum complications and constitute emergencies, as described in Chapter 41 (General Considerations).

Immediately after newborn birth, uterine fundal size and consistency are examined. If the uterus remains firm and there is no unusual bleeding, watchful waiting until the placenta separates is the usual practice. Neither massage nor downward fundal pressure is employed, but the fundus is frequently palpated to ensure that it does not become atonic and filled with blood from placental separation. To prevent uterine inversion, umbilical cord traction must not be used to pull the placenta from the uterus. Signs of separation include a sudden gush of blood into the vagina, a globular and firmer fundus, a lengthening of the umbilical cord as the placenta descends into the vagina, and elevation of the uterus into the abdomen. With the last, the placenta, having separated, passes down into the lower uterine segment and vagina. Here, its bulk pushes the uterine body upward.

These signs appear within minutes after newborn delivery, and the median time ranges from 4 to 12 minutes (Combs, 1991; Frolova, 2016; Shinar, 2016b). Once the placenta has detached from the uterine wall, the mother may be asked to bear down, and the intraabdominal pressure often expels the placenta into the vagina. These efforts may fail or may not be possible because of analgesia. After ensuring that the uterus is contracted firmly, the umbilical cord is kept slightly taut but is not pulled. Pressure is exerted by a hand wrapped around the fundus to propel the detached placenta into the vagina (Fig. 27-10). Concurrently, the heel of the hand exerts downward pressure between the symphysis pubis and the uterine fundus. This also aids inversion prevention. Once the placenta passes through the introitus, pressure on the uterus is relieved. The placenta is then gently lifted away. Care is taken to prevent placental membranes from being torn off and left behind. If the membranes begin to tear, they are grasped with a clamp and removed by gentle teasing (Fig. 27-11).

Management of the Third Stage

Practices within the third stage of labor may be broadly considered as either expectant or active management. Expectant management involves waiting for placental separation signs and allowing the placenta to deliver either spontaneously or aided by nipple stimulation or gravity (World Health Organization, 2012). In contrast, active management of third-stage labor consists of early cord clamping, controlled cord traction during placental delivery, and immediate administration of prophylactic oxytocin. The goal of this triad is to limit postpartum hemorrhage (Begley, 2015; Jangsten, 2011; Westhoff, 2013).

As noted earlier, delayed cord clamping does not increase postpartum hemorrhage rates, and thus early clamping is a less important component of this trio. Similarly, cord traction may also be less critical (Deneux-Tharaux, 2013; Du, 2014; Gülmezoglu, 2012). Uterine massage following placental delivery is recommended by many, but not all, to prevent postpartum hemorrhage. We support this with the caveat that evidence for this practice is not strong (Abdel-Aleem, 2010).

Therefore, uterotonics play an essential role to decrease postpartum blood loss. Choices include oxytocin (Pitocin), misoprostol (Cytotec), carboprost (Hemabate), and the ergots, namely ergonovine (Ergotrate) and methylergonovine (Methergine). In addition, a combination agent of oxytocin and ergonovine (Syntometrine) is used outside the United States. Also in other countries, carbetocin (Duratocin), a long-acting oxytocin analogue, is available and effective for hemorrhage prevention during cesarean delivery (Attilakos, 2010; Su, 2012). Of these, the World Health Organization (2012) recommends oxytocin as a first-line agent. Ergot-based drugs and misoprostol are alternatives in settings that lack oxytocin.

Uterotonics may be given before or after placental expulsion without affecting rates of postpartum hemorrhage, placental retention, or third-stage labor length (Soltani, 2010). If they are given before delivery of the placenta, however, they may entrap an undiagnosed, undelivered second twin. Thus, abdominal palpation should confirm no additional fetuses. Notably, this concern is less relevant with current widespread sonography use.

High-Dose Oxytocin

Synthetic oxytocin is identical to that produced by the posterior pituitary. Its action is noted at approximately 1 minute, and it has a mean half-life of 3 to 5 minutes. When given as a bolus, oxytocin can cause profound hypotension. Secher and coworkers (1978) reported that an intravenous bolus of 10 units of oxytocin caused a marked transient fall in blood pressure with an abrupt increase in cardiac output. Svanström and associates (2008) confirmed those findings. These hemodynamic changes could be dangerous for women hypovolemic from hemorrhage or those with certain types of cardiac disease. Thus, oxytocin should be given as a dilute solution by continuous intravenous infusion or as an intramuscular injection.

Water intoxication can result from the antidiuretic action of high-dose oxytocin if administered in a large volume of electrolyte-free dextrose solution (Whalley, 1963). Thus, if oxytocin is to be administered in high doses for a considerable period of time, its concentration should be increased rather than increasing the infusion flow rate.

Despite the routine use of oxytocin, no standard prophylactic dose has been established for its use following either vaginal or cesarean delivery. Our practice is to add 20 units (2 mL) of oxytocin per liter of infusate. This solution is administered after delivery of the placenta at a rate of 10 to 20 mL/min—200 to 400 mU/min—for a few minutes until the uterus remains firmly contracted and bleeding is controlled. The infusion rate then is reduced to 1 to 2 mL/min until the mother is ready for transfer from the recovery suite to the postpartum unit. The infusion is usually then discontinued. For women without intravenous access, 10 units of intramuscular oxytocin are injected.

Other Uterotonics

Ergonovine and methylergonovine have similar activity levels in myometrium, and only methylergonovine is currently manufactured in the United States. These ergot alkaloid agents do not provide superior protection against postpartum hemorrhage compared with oxytocin. Moreover, safety and tolerability are greater with oxytocin (Liabsuetrakul, 2011). For these reasons, ergot alkaloid agents are considered second-line for prevention of postpartum hemorrhage. If selected, a 0.2-mg dose of methylergonovine is slowly given intravenously in a period not less than 60 seconds to avoid sudden hypertension (Novartis, 2012). Methylergonovine is relatively contraindicated in the hypertensive woman.

Misoprostol is a prostaglandin E₁ analogue, which has proved inferior to oxytocin for postpartum hemorrhage prevention (Tunçalp, 2012). However, in resource-poor settings that lack oxytocin, misoprostol is suitable for hemorrhage prophylaxis and is given as a single oral 600-μg dose (Mobeen, 2011; World Health Organization, 2012). Notably, although oxytocin is preferred for prevention of hemorrhage, ergot alkaloids and prostaglandins play a greater role in postpartum hemorrhage treatment, discussed in Chapter 41 (Risk Factors).

Manual Removal of Placenta

In approximately 2 percent of singleton births, the placenta may not deliver promptly (Cheung, 2011). Three possibilities include placenta adherens, in which uterine contractions are insufficient to detach the placenta; lower uterine segment constriction and a detached but trapped placenta; or a morbidly adherent placenta. Consistent risks for retained placenta include stillbirth, prior cesarean delivery, prior retention, and preterm delivery (Belachew, 2014; Coviello, 2015; Endler, 2014; Nikolajsen, 2013). For the last, in one study with nearly 46,000 deliveries, analysis predicted that 90 percent of placentas would spontaneously deliver by 180 minutes for gestations at 20 weeks; 21 minutes at 30 weeks; and 14 minutes at 40 weeks (Dombrowski, 1995).

Postpartum hemorrhage can complicate a retained placenta, and bleeding risk accrues with third-stage length. Thus, in the absence of bleeding, some recommend expectant management for 30 minutes, whereas others use a 15-minute threshold (Cummings, 2016; Deneux-Tharaux, 2009; Shinar, 2016a). The World Health Organization (2012) cites a 60-minute threshold. Notably, if brisk bleeding ensues and the placenta cannot be delivered by standard technique, manual removal of the placenta is indicated (Fig. 27-12). When performed, some administer a single dose of intravenous antibiotics, however, one systematic review of observational studies found no benefits (Chibueze, 2015). Although the American College of Obstetricians and Gynecologists (2016c) concludes that data neither support nor refute this practice, the World Health Organization (2012) recommends prophylaxis. At our institution, we administer a single dose to women not already receiving antibiotics.

The hour immediately following delivery of the placenta is critical. During this time, lacerations are repaired. Although uterotonics are administered, postpartum hemorrhage as the result of uterine atony is most likely at this time. Hematomas may expand. Consequently, uterine tone and the perineum are frequently evaluated. The American Academy of Pediatrics and the American College of Obstetricians and Gynecologists (2017b) recommend that maternal blood pressure and pulse be recorded immediately after delivery and every 15 minutes for the first 2 hours. The placenta, membranes, and umbilical cord are examined for completeness and for anomalies, as described in Chapter 6 (Normal Placenta).

Birth Canal Lacerations

Lower genital tract lacerations may involve the cervix, vagina, or perineum. Those of the cervix and vagina are described in Chapter 41 (Injuries to the Birth Canal). Those of the perineum often follow vaginal delivery, and most are first- and second-degree lacerations. Lacerations are classified by their depth, and complete definitions and visual examples are given in Figure 27-13. Of these, third-degree lacerations reflect anal sphincter injury and are now subcategorized as:

• (3a) <50 percent external anal sphincter (EAS) tear;

• (3b) >50 percent EAS tear; and

• (3c) EAS plus internal anal sphincter (IAS) tears.

Third- and fourth-degree lacerations are considered obstetrical anal sphincter injuries (OASIS), and their combined incidence varies from 0.5 to 5 percent (Blondel, 2016; Friedman, 2015). Risk factors for these more complex lacerations include nulliparity, midline episiotomy, persistent OP position, operative vaginal delivery, Asian race, short perineal length, and increasing fetal birthweight (Ampt, 2013; Dua, 2009; Gurol-Urganci, 2013; Landy, 2011). Mediolateral episiotomy is protective in most, but not all, studies (Jangö, 2014; Räisänen, 2011; Shmueli, 2016).

Morbidity rates rise as laceration severity increases. Compared with simpler lacerations, anal sphincter injuries are associated with greater blood loss and puerperal pain. Wound disruption and infection rates are other risks (Goldaber, 1993; Lewicky-Gaupp, 2015). Stock and coworkers (2013) reported that approximately 7 percent of 909 OASIS lacerations had complications. Long term, anal sphincter injuries are linked with approximately doubled rates of fecal incontinence compared with vaginal delivery without OASIS (Evers, 2012; Gyhagen, 2014). Data on long-term dyspareunia are limited, and rates are increased in some but not all studies (Mous, 2008; Otero, 2006; Salim, 2014; Sundquist, 2012).

To ensure appropriate repair, identification and correct categorization is essential. Diagnosis rates of OASIS improve with clinical experience (Andrews, 2006). Intrapartum endoanal ultrasound, performed in research studies, also boosts detection, and rates of clinically occult tears in primiparas range from 6 to 12 percent (Corton, 2013; Faltin, 2005; Ozyurt, 2015). That said, few data currently support routine intrapartum endoanal sonography, and the American College of Obstetricians and Gynecologists (2016b) does not recommended it (Walsh, 2015).

Women with a prior OASIS have a higher recurrence rate compared with multiparas without prior OASIS (Baghestan, 2012; Edozien, 2014; Elfaghi, 2004). That said, the risk mirrors that of primiparas in the general population and is low (Basham, 2013; Boggs, 2014; Priddis, 2013). Fetal macrosomia and operative vaginal delivery are notable risks in this cohort of parturients and can influence counseling in future pregnancies. Specifically, patients may choose to deliver by cesarean to avoid repeat OASIS. This consideration may be most pertinent for those with prior postpartum anal incontinence, with OASIS complications requiring corrective surgery, or with psychological trauma (American College of Obstetricians and Gynecologists, 2016b). However, planned cesarean delivery is balanced against its associated operative risks discussed in Chapter 30 (Cesarean Delivery Risks).

Episiotomy

Types

In contrast to spontaneous lacerations, perineotomy is intended incision of the perineum. Episiotomy is incision of the pudendum–the external genital organs. In common parlance, however, the term episiotomy often is used synonymously with perineotomy, a practice that we follow here. Obstetrical textbooks and organizational guidelines differ considerably in their description of episiotomy techniques. Kalis and associates (2012) have presented a classification, and we agree with the need for terminology standardization.

Midline and mediolateral episiotomies are the two main types and vary by the angle of perineal incision. Involved structures mirror those found with second-degree laceration, and their repairs are analogous. The midline episiotomy begins at the fourchette, incises the perineal body in the midline, and ends well before the external anal sphincter is reached. The incision length varies from 2 to 3 cm depending on perineal length and degree of tissue thinning. The mediolateral episiotomy begins at the midline of the fourchette and is directed to the right or left at an angle 60 degrees off the midline (Fig. 27-14). This angle accounts for perineal anatomy distortion during crowning and ultimately yields an incision 45 degrees off the midline for suturing (El-Din, 2014; Kalis, 2011). The lateral episiotomy begins at point 1 to 2 cm lateral from the midline. It too is angled toward either the right or the left ischial tuberosity.

Before episiotomy, analgesia may be provided by existing labor regional analgesia, by bilateral pudendal nerve blockade, or by local infiltration of 1-percent lidocaine. Some instead advocate 2.5-percent lidocaine-prilocaine cream (EMLA cream), but this requires application an hour before expected delivery, which may be logistically difficult (Franchi, 2009; Kargar, 2016).

If episiotomy is performed unnecessarily early, incisional bleeding may be considerable before delivery. If it is performed too late, lacerations will not be prevented. Typically, episiotomy is completed when the head is visible during a contraction to a diameter of approximately 4 cm, that is, crowning. When used in conjunction with forceps delivery, most perform an episiotomy after application of the blades.

Few data directly compare midline and mediolateral types. As noted, midline episiotomy has a greater likelihood of associated anal sphincter lacerations (Coats, 1980; de Leeuw, 2001). Short-term rates of self-perceived pain and dyspareunia are similar or increased with mediolateral episiotomy (Fodstad, 2013, 2014; Sartore, 2004).

Even fewer studies compare lateral episiotomy to either mediolateral or midline. One randomized trial compared lateral and mediolateral types in nulliparas. Groups did not differ in pain scores, in sexual quality of life, or in vaginal or perineal trauma, including OASIS (Karbanova, 2014a,b; Necesalova, 2016). The authors also reported that mediolateral episiotomies required less time and suture for the repair. Thus, among the three, mediolateral episiotomy may be the preferred incision to reduce OASIS rates.

Indications

In the past, routine episiotomy was practiced to avoid a ragged laceration and to limit postoperative pain and anal sphincter injury rates. But, a Cochrane review of randomized trials showed lower rates of severe perineal/vaginal trauma in women managed with a restrictive, that is, selective use of episiotomy for spontaneous delivery rather than with routine episiotomy (Jiang, 2017). Importantly, this review did not discern between midline and mediolateral episiotomies.

The American College of Obstetricians and Gynecologists (2016b) has concluded that restricted use of episiotomy is preferred to routine use. We are of the view that the procedure should be applied selectively for appropriate indications. Thus, episiotomy can be considered for indications such as shoulder dystocia, breech delivery, fetal macrosomia, operative vaginal deliveries, persistent OP positions, markedly short perineal length, and other instances in which failure to perform an episiotomy will result in significant perineal rupture. The final rule is that there is no substitute for surgical judgment and common sense.

With this new approach, episiotomy rates have dropped. Oliphant and coworkers (2010) used the National Hospital Discharge Survey to analyze episiotomy rates between 1979 and 2006 in the United States. They noted a 75-percent decline in the age-adjusted episiotomy rate. In the United States in 2012 episiotomy was performed in approximately 12 percent of vaginal births (Friedman, 2015).

Laceration and Episiotomy Repairs

Typically, perineal repairs are deferred until the placenta has been delivered. This policy permits undivided attention to the signs of placental separation and delivery. A further advantage is that the repair is not interrupted or disrupted by placenta delivery. This is especially true if manual removal must be performed. The major disadvantage is continuing blood loss until the repair is completed. Direct pressure from an applied gauze sponge will help to limit this volume.

For suitable repair, an understanding of perineal support and anatomy is necessary and is discussed in Chapter 2 (Perineum). Adequate analgesia is imperative, and women without regional analgesia can experience high levels of pain during perineal suturing. Again, locally injected lidocaine can be used solely or as a supplement to bilateral pudendal nerve blockade. In those with epidural analgesia, additional dosing may be necessary.

First-degree lacerations do not always require repair, and sutures are placed to control bleeding or restore anatomy. Here, few data guide suture selection, and fine-gauge absorbable or delayed-absorbable suture or adhesive glue is suitable.

Second-degree laceration correction as well as midline and mediolateral episiotomy repairs include similar steps. Namely, these close the vaginal epithelium and reapproximate the bulbospongiosus and superficial transverse perineal muscles during restoration of the perineal body (Figs. 27-15 and 27-16). For this, most studies support a continuous suturing method, which is faster than placing interrupted sutures and, with few exceptions, yields less pain (Grant, 2001; Kettle, 2012; Kindberg, 2008; Valenzuela, 2009). Blunt needles are suitable and likely decrease the incidence of needle-stick injuries (El-Refaie, 2012; Mornar, 2008). Commonly used suture materials are 2–0 polyglactin 910 (Vicryl) or chromic catgut. With the former, a decrease in postsurgical pain and lower risk of wound dehiscence are cited as major advantages (Jallad, 2016; Kettle, 2010). Closures with traditional polyglactin 910, however, occasionally require removal of residual suture from the repair site because of pain or dyspareunia. This disadvantage may be reduced by using a rapidly absorbed polyglactin 910 (Vicryl Rapide) (Bharathi, 2013; Kettle, 2002; Leroux, 2006).

For third-degree laceration repair, two methods are available to repair the external anal sphincter. The first is an end-to-end technique, which we prefer, and is shown in Figure 27-17. Initially, the cut ends of the external anal sphincter, which often retract, are isolated and brought to the midline. Importantly, the strength of this closure is derived from the connective tissue surrounding the sphincter—often called the capsule—and not the striated muscle. Thus, serial interrupted sutures incorporate sphincter fibers and perisphincter connective tissue, to bring sphincter ends together. There are few evidence-based data to guide suture selection for sphincter repair, but delayed-absorbable material can provide sustained tensile strength during healing. This theory is supported by the above study by Jallad and coworkers (2016), which showed higher perineal breakdown rate following OASIS repair with chromic gut.

With the overlapping technique, the ends of the external anal sphincter are brought to the midline and lie atop one another. This method is only suitable for type 3c lacerations—those involving the external and internal anal sphincter. Two rows of mattress sutures travel through both sphincter ends to recreate the anal ring. In comparing the two methods, neither yields superior long-term anatomical or functional results (Farrell, 2012; Fernando, 2013; Fitzpatrick, 2000). Also with type 3c lacerations, the IAS is repaired before the EAS and is described next.

With fourth-degree laceration repairs, the torn edges of the rectal mucosa are reapproximated (Fig. 27-18). At a point 1 cm proximal to the wound apex, sutures are placed approximately 0.5 cm apart in the rectal muscularis and do not enter the anorectal lumen. Clinicians often use 4–0 polyglactin 910 or chromic gut for this running suture line. Some recommend a second reinforcing layer above this (Hale, 2007). If this is not done, then the next layer to cover the anorectal mucosa is formed by reapproximation of the internal anal sphincter. This running, nonlocking closure is completed with 3–0 or 4–0 suture (see Fig. 27-18B). Following any repair, needle and sponge counts are reconciled and recorded in the delivery note.

For reduction of infectious morbidity associated with anal sphincter lacerations, a single dose of antibiotic at the time of repair is recommended by the American College of Obstetricians and Gynecologists (2016c). This practice is supported by evidence (Buppasiri, 2014; Duggal, 2008; Lewicky-Gaupp, 2015; Stock, 2013). A single dose of a second-generation cephalosporin is suitable, or clindamycin for penicillin-allergic women. With OASIS, postoperatively, stool softeners are prescribed for a week, and enemas and suppositories are avoided.

Unfortunately, normal function is not always ensured even with correct and complete surgical repair. Some women may experience continuing fecal incontinence caused by injury to the innervation of the pelvic floor musculature (Roberts, 1990).

Perineal Laceration Care

Initially, locally applied ice packs help reduce swelling and allay discomfort (de Souza Bosco Paiva, 2016). In subsequent days, warm sitz baths aid comfort and hygiene. Additionally, a small squirt bottle of warm water can cleanse the site after voiding or stooling. For pain, topical application of 5-percent lidocaine ointment was not effective in relieving episiotomy or perineal laceration discomfort in one randomized trial (Minassian, 2002). Oral analgesics containing codeine provide considerable relief. For lesser degree of discomfort, NSAID tablets can be given.

Because pain may signal a large vulvar, paravaginal, or ischiorectal fossa hematoma or perineal cellulitis, these sites should be examined carefully if pain is severe or persistent. Management of these complications is discussed in Chapters 37 and 41 (Perineal Infections and Puerperal Hematomas). In addition to pain, urinary retention may complicate episiotomy recovery (Mulder, 2012, 2016). Its management is described in Chapter 36 (Perineal Care).

For those with second-degree lacerations or anal sphincter tears, intercourse is usually proscribed until after the first puerperal visit at 6 weeks. Compared with women with intact perineum, those with perineal trauma show higher rates of delayed intercourse at 3 and 6 months, but not at 1 year (McDonald, 2015; Rådestad, 2008; Signorello, 2001).