Clinical Aspects of Normal and Abnormal Labor

Prolonged Latent Phase

On the basis of the 95th percentile limit of the distribution of latent-phase duration in the primiparous population, 20 hours is considered the definition of an abnormal latent phase. For multiparas, the corresponding definition of prolonged latent phase is 14 hours. Sometimes it is difficult to ascertain the difference between a prolonged latent phase and so-called false labor. Friedman found that prolongation of the latent phase was associated with excessive sedation, prematurely administered epidural anesthesia, unfavorable cervical status, or myometrial dysfunction.

Although early studies suggested that prolongation of the latent phase was not associated with increased perinatal mortality and was not the harbinger of other abnormalities of labor, subsequent research showed otherwise. In a study of 10,979 patients in San Francisco, Chelmow and colleagues found that prolonged latent phase of labor, defined as longer than 12 hours for nulliparous patients and longer than 6 hours for multiparous patients, was associated with an increased risk for subsequent labor abnormalities, cesarean delivery, low Apgar score, and need for neonatal resuscitation. These risks for adverse outcomes remained significantly elevated even when data were controlled for other labor abnormalities, prolonged rupture of membranes, meconium-stained amniotic fluid, parity, and epidural use. In addition to the increased risk of cesarean delivery, a prolonged latent phase of labor in patients who delivered vaginally was associated with an approximately twofold increased incidence of third-degree and fourth-degree lacerations, febrile morbidity, and intrapartum blood loss.

One of the major problems with evaluation and management of the latent phase of labor is knowing at what hour labor began. Some authorities have used the time of admission to the hospital as a convenient starting point for judging when to intervene in the progress of labor. ^{, ,} Hanley and colleagues systematically reviewed various definitions of the onset of labor and found little consensus or evidence of superiority of one definition over another. Friedman, however, regarded the onset of regular contractions as the beginning of labor and recommended intervention when the duration of the latent phase of labor reaches 20 hours in the primipara. He found that either adequate sedation (“therapeutic narcosis”) or oxytocin augmentation resulted in the resumption of normal cervical dilation. Because most patients are exhausted after 20 hours of labor, Friedman preferred therapeutic narcosis over oxytocin augmentation. For narcosis, he recommended morphine sulfate, 15 to 20 mg, with 10 to 15 mg more if the first dose has not made the patient somnolent and thereby inhibited uterine contractions. The obvious advantage of this therapy is that the patient awakens rested and refreshed and prepared for the active phase of labor.

Critics of this approach, especially O’Driscoll and Meagher, argued that waiting out 20 hours of latent phase before considering the labor to be abnormal only promotes exhaustion and discourages the patient. They advocated a protocol for active management of labor that has been practiced and evaluated at the National Maternity Hospital in Dublin. The protocol has several important features:

• 1.

  Patients are admitted to the labor unit only when they are experiencing painful uterine contractions as well as complete effacement of the cervix, ruptured membranes, or passage of blood-stained mucus.

• 2.

  Amniotomy is performed soon after admission for patients who have intact membranes.

• 3.

  Oxytocin augmentation of labor is performed if the progress of labor is less than 1 cm/hr over a 2-hour period. Oxytocin infusion is begun at 4 mU/min and is increased by 6 mU/min every 15 minutes until there are seven contractions per 15 minutes. The oxytocin infusion rate does not exceed 40 mU/min.

• 4.

  Continuous electronic fetal heart rate monitoring is used only if there is meconium-stained amniotic fluid and after the fetal scalp pH has been determined to rule out fetal acidosis.

• 5.

  A nurse-midwife is in constant attendance with the patient throughout labor.

• 6.

  The patient is assured that if her labor exceeds 12 hours, cesarean delivery will probably be performed.

• 7.

  The progress of labor is documented on a simple graphic form, and the senior obstetrician in charge of the unit reviews all cases daily. A partogram, as described earlier, is used to time interventions.

• 8.

  This approach to the management of labor is confined to nulliparas.

This active management protocol, with minor modifications, has also been evaluated in several obstetric services in the United States and other countries. These studies have consistently demonstrated a small but significant shortening of labor associated with active management. Although most have also demonstrated a decrease in the incidence of cesarean delivery for dystocia, the largest prospective and best-designed controlled trial showed no difference in the incidence of cesarean delivery for dystocia, although it did show shortened labor and a decreased incidence of maternal infection with the active management protocol.

Holmes and associates clearly demonstrated that women who present to the hospital with less than 3 cm dilation are more likely to undergo cesarean or operative vaginal delivery than women presenting with more advanced dilation. Interestingly, they found that women presenting with less than 3 cm dilation had spent less time at home (2.0 versus 4.5 hours) since the onset of painful uterine contractions. Their results imply that women who present with reduced cervical dilation could have intrinsically different labors than those who present with more advanced dilation. Murphy and coworkers and Falzone and colleagues made similar observations, noting that nulliparous women presenting in labor with unengaged and particularly floating (above −3/3 station) fetal heads had higher risks for obstetric intervention.

Protraction Disorders

Protraction disorders are those in which the progress of cervical dilation and descent of the fetal head occur at a slower than normal rate during the active phase of labor. The work by Friedman originally demonstrated that the rate of cervical dilation for nulliparas should be at least 1.2 cm/hr; for multiparas, it should be 1.5 cm/hr or faster. ^, For descent of the fetal head, the rate for nulliparas should be 1.0 cm/hr or faster; for multiparas, it should be 2.0 cm/hr. These criteria for minimum rates of cervical change represent the 95th percentiles for each parity category. More recently, the Consortium on Safe Labor found that nulliparous women who presented in spontaneous labor with vaginal delivery and normal perinatal outcome had longer 95th percentiles labor durations depending on the cervical dilation ( Table 40.2 ). The 95th percentile for labor duration was 7 hours from 4 to 5 cm cervical dilation and over 3 hours from 5 to 6 cm cervical dilation for multiparous women. In 2012, a joint summary was released by the Eunice Kennedy Shriver National Institute of Child Health and Human Development, the Society for Maternal-Fetal Medicine, and the American College of Obstetricians and Gynecologists. After reviewing the available evidence, the workshop concluded, “These data suggest that the historical criteria defining normal labor progress—cervical change of 1.2 cm/hr for nulliparous women and 1.5 cm/hr for multiparous women—are no longer valid.”

Friedman found protraction disorders in primiparas to be associated frequently with cephalopelvic disproportion (CPD), use of conduction anesthesia, and fetal malposition. Whether these factors are related in a cause-and-effect manner is not known. Moreover, he found oxytocin augmentation and therapeutic narcosis to be of little value in these cases. Friedman also noted unusually high neonatal mortality and morbidity rates when this labor abnormality was terminated by mid-forceps delivery; however, the diagnosis of a primary dysfunctional labor (i.e., persisting at a cervical dilation rate of <1.2 cm/hr) is usually made in retrospect, after oxytocin augmentation has been used and found not to increase the dilation rate.

The experiences of Beazley and Kurjak and those of O’Driscoll and Meagher suggested that an early, more active use of oxytocin, as described in the active management of labor protocol, effectively corrects most protraction disorders, although these authors did not specifically separate protraction disorders from arrest disorders. Those who advocate the active management of labor explain that the use of x-ray pelvimetry is unnecessary in the nulliparous patient, because rupture of the uterus does not occur with the recommended oxytocin augmentation. Therefore, in nulliparous patients with suboptimal progress of labor from any cause, it is safe to use a trial of oxytocin to determine whether labor will progress to completion.

Ganström and associates demonstrated significant differences in collagen content and collagen remodeling in the cervix and lower uterine segment in patients with protracted labors compared with those having normal labors. This finding may explain why some patients with protracted labor do not respond to oxytocin augmentation.

Arrest Disorders

Friedman originally defined an arrest disorder as the cessation of either cervical dilation or descent of the fetal head in the active phase of labor for longer than 2 hours. In their pure form, arrest disorders differ from protraction abnormalities in that, before the arrest of progress, the rate of cervical dilation or descent of the fetal head is normal. Arrest of progress can also complicate a protraction disorder. In either situation, Friedman found that 45% of the cases of arrest disorder were associated with CPD. Philpott and Castle also found that patients whose labor progress crossed the “action line” (i.e., those with protraction or arrest disorders) had smaller pelvic measurements and more often required cesarean delivery for CPD.

Because of the frequent association between arrest disorders and CPD, Friedman recommended radiographic cephalopelvimetry followed by cesarean delivery for those with CPD and oxytocin augmentation for the remainder. He found that 80% of women with arrest disorders who did not have CPD delivered after oxytocin augmentation. Philpott and Castle and O’Driscoll and Meagher found that radiographic studies are not required, especially in primiparas, and that a trial of oxytocin augmentation is indicated in all protraction and arrest disorders. With careful monitoring of mother and fetus and discontinuation of augmentation if there is no progress after 4 to 6 hours, patients are not in danger. This is the approach of most, if not all, obstetric services in the United States; radiographic cephalopelvimetry is seldom used in the management of abnormal labor in vertex presentations.

A notable exception is the use of the fetal-pelvic index by Morgan and Thurnau. This technique combines ultrasound measurement of the fetal head circumference (HC) and abdominal circumference (AC) with radiographic measurement of the maternal pelvic inlet circumference (IC) and midpelvic circumference (MC). The fetal-pelvic index is the sum of the two greatest positive circumference differences (i.e., HC−IC, HC−MC, AC−IC, or AC−MC). A positive fetal-pelvic index value indicates the presence of fetal-pelvic disproportion, and a negative fetal-pelvic index value indicates the absence of fetal-pelvic disproportion. This index had a 94% positive predictive value for cesarean delivery of patients with abnormal labor patterns. These authors also found the fetal-pelvic index to be useful in predicting the success of induction of labor and the success of attempted vaginal birth after previous cesarean delivery. Ferguson and colleagues were not able to confirm the efficacy of the fetal-pelvic index, and the method is not widely used.

Using labor progression guidelines based on the slower labor curves characteristic of modern parturients, Rouse and colleagues demonstrated the effectiveness of a new protocol to treat arrest disorders. Their protocol had three principal elements:

• 1.

  An intent to achieve a sustained uterine contraction pattern of greater than 200 Montevideo units as measured by an intrauterine pressure catheter

• 2.

  A minimum of 4 hours of oxytocin-augmented labor arrest with a contraction pattern of greater than 200 Montevideo units before proceeding to abdominal delivery for active-phase arrest (more liberal than the original Friedman cutoff of 2 hours)

• 3.

  For patients who cannot achieve a sustained uterine contraction pattern of greater than 200 Montevideo units, administration of a minimum of 6 hours of oxytocin augmentation before proceeding to cesarean delivery for active-phase labor arrest

The researchers demonstrated not only the effectiveness (92% vaginal delivery rate) but also the safety of this approach, with no serious adverse maternal or perinatal effects. The only cost of liberalization of the minimums was an increased risk of maternal infection, with the risk proportional to the time elapsed.

These recommendations of Rouse and colleagues raise the issue of whether intrauterine pressure catheters are useful tools for clinicians managing labor abnormalities. Lucidi and colleagues reviewed the literature in 2001 and noted that there were no clinical trials testing the utility of this device. The technique is not included in the routines of the active management protocols described earlier. It would seem that partograms have a much stronger evidence basis than measurement of uterine contractibility to guide the timing of obstetric interventions. Bakker and associates systematically reviewed internal versus external tacodynamometry during labor and found no differences in any maternal or fetal outcome. These same authors from the Netherlands reported a randomized trial in 1456 women of internal versus external modalities and showed no differences in operative delivery or newborn outcomes.

Zhang and colleagues in the Consortium on Safe Labor compared oxytocin augmentation protocols in 15,054 women. Oxytocin regimen had no effect on cesarean delivery or other perinatal outcomes. Compared to regimens starting at 1 mU/min of oxytocin, those starting at 2 mU/min and 4 mU/min shortened duration of labor by 0.8 and 1.3 hours, respectively, in nulliparous women. Similar results were seen in parity 1+ women. Although the authors endorsed starting oxytocin augmentation protocols at higher doses, the reader should be mindful that these data were obtained within large, tertiary units. The findings may not be generalizable to smaller units with fewer resources to monitor mothers and their unborn babies. The small decreases in labor length must be weighed against any potential harms from higher-dose regimens.

Several authors have evaluated the effects of ambulation on the progress of labor. Flynn and coworkers found that patients who ambulated had more rapid labor with fewer instances of fetal distress than a similar number of patients who labored in bed. Williams and associates, studying 48 patients who ambulated, could find no differences in duration of labor or frequency of fetal distress compared with control patients. Read and colleagues studied 14 patients whose labors were regarded as requiring augmentation because of lack of progress attributed to inadequate contractions. Progress of labor was as rapid in the 8 patients who were randomized to an ambulation study protocol as in 6 control patients whose labors were augmented with oxytocin.

These studies suggest that ambulation is not detrimental to the progress of labor or the well-being of the fetus. However, these studies had an insufficient sample size and it has not been established whether ambulation is clearly beneficial or whether it is a substitute for pharmacologic augmentation of labor in cases of abnormal progress.

Abnormalities of Rotation and Descent

Textbooks of obstetrics traditionally have discussed the first and second stages of labor as if they were separate clinical and biologic entities, which they are not. Descent and rotation of the fetal head frequently occur before complete dilation of the cervix, a phenomenon that is clear to most clinicians and that was confirmed by the studies of Friedman. In addition to showing slower rates of cervical dilation than did Friedman, the contemporary data of Zhang et al. showed a slower rate of fetal head descent. As demonstrated in Table 40.6 and Fig. 40.5 , it can take up to 3 hours to descend from +1/3 to +3/3 station and an additional 30 minutes for delivery. Again, there is a clear need for practice guidelines to incorporate these new data.

Arrest of descent and rotation, whether it occurs before or after complete dilation of the cervix, is a matter of concern and requires evaluation. Arbitrary limits on the duration of the second stage of labor probably resulted from misinterpretation of the data presented by Hellman and Prystowsky. In their study, patients whose second stage of labor was longer than 2 hours were at increased risk for perinatal and maternal morbidity. This observation was interpreted by many clinicians to mean that delivery of the fetus should be accomplished, by whatever means, before 2 hours of the second stage had elapsed. This interpretation occasionally resulted in traumatic mid-forceps operations or unnecessary cesarean deliveries, not to mention overzealous use of the vacuum extractor. The reader should note that their recommendation of a 2-hour second stage limit antedates electronic fetal monitoring.

Cohen demonstrated that when patients with fetal distress or traumatic delivery are excluded, the duration of the second stage bears no relationship to perinatal outcome. If there are no serious fetal heart rate abnormalities, the mother is well hydrated and reasonably comfortable, and there is some progress in descent or rotation of the fetal head (regardless of how slow), there is no need for operative delivery. Similarly, Menticoglou and associates confirmed that the duration of the second stage of labor is in itself not related to untoward outcomes. Hansen and coauthors, in a trial of active versus passive pushing in the second stage, found that second-stage lengths as long as 4 to 9 hours had no harmful effects. Moreover, delayed pushing was better tolerated by patients and was associated with fewer fetal heart rate decelerations. In a meta-analysis of 20 studies, Lemos and associates found that delayed pushing leads to a shortening of actual time pushing and an increase in spontaneous delivery at the cost of an overall longer second stage of labor. They concluded that there is not evidence to endorse any style of pushing with or without epidural anesthesia. Rouse and colleagues performed a secondary analysis on 5341 participants in the fetal pulse oximetry trial and demonstrated that 55% of women with a second-stage labor lasting 3 hours or longer had a successful vaginal delivery. There was an increased risk of maternal morbidity, including chorioamnionitis, severe perineal lacerations, uterine atony, and blood transfusions. Admission to a neonatal intensive care unit (NICU) and composite serious neonatal morbidity were also increased with a second stage lasting 3 hours or longer. Laughon and colleagues in the Consortium on Safe Labor assessed maternal and neonatal outcomes for deliveries ≥36 weeks with a prolonged second stage, defined as nulliparous women with epidural >3 hours and without >2 hours and multiparous women with epidural >2 hours and without >1 hour. Prolonged second stage was associated with increased chorioamnionitis and third-degree or fourth-degree perineal lacerations, regardless of parity or epidural status. Nulliparous women with an epidural who delivered after a prolonged second stage had additional increased rates of endometritis, wound separation, and postpartum hemorrhage; multiparous women who delivered with prolonged second stage regardless of epidural status had higher rates of postpartum hemorrhage. Neonatal morbidity with prolonged second stage included small increases in sepsis in nulliparous women regardless of epidural status, asphyxia in nulliparous women with epidural, and perinatal mortality in deliveries without an epidural. However, among the offspring of women with epidurals whose second stage was prolonged (3533 nulliparous and 1348 multiparous women), there were no cases of hypoxic ischemic encephalopathy or perinatal death. More data are needed to help determine the optimal cutoff for second stage labor.

After the cervix is dilated more than 7 cm, descent or rotation of the fetal head can be expected. If this does not occur, uterine contractions, if they are not adequate, should be augmented with oxytocin. Manual examination to determine the position of the fetal head and the dimensions and shape of the pelvis often helps at this point. Posterior presentation, brow presentation, marked degrees of asynclitism, and very large infants are associated with longer labors, even with adequate contractions.

The Mueller-Hillis maneuver also may help. The obstetrician applies pressure to the uterine fundus with one hand and detects descent of the fetal head with the examining finger in the vagina. If the fetal head descends 1 cm or more with fundal pressure, the prognosis for vaginal delivery is good; if no descent occurs, the prognosis for delivery is poor. This maneuver is not predictive of outcome if it is performed early in labor, but it is helpful late in the first stage or in the second stage of labor.

Kominiarek and colleagues from the Consortium on Safe Labor showed that although obesity lengthened the first stage of labor in a linear fashion (i.e., with increases in BMI, the first stage was longer), there was no relationship between maternal weight and the length of the second stage. Robinson and coworkers studied this point further in 5341 nulliparous women from the Fetal Pulse Oximetry Trial. Obesity did not have any relationship with second-stage length and moreover was not related to need for abdominal delivery performed in the second stage. Therefore obesity seems to affect labor by lengthening the first stage and to have no impact on the length of the second stage.

Cahill and associates completed a randomized controlled trial in 2404 nulliparous women receiving epidural anesthesia to compare immediate versus delayed voluntary propulsive efforts (“pushing”) in the second stage. Both groups had a high proportion of spontaneous delivery (around 80%), with the immediate group having a shorter mean second-stage length by 32 minutes, lower occurrence of chorioamnionitis, and fewer postpartum hemorrhages. One can conclude that delayed pushing has no benefit on spontaneous vaginal delivery and increases the risk of complications and argue for immediate pushing.

Shoulder Dystocia

Shoulder dystocia occurs in 0.24% to 2.00% of vaginal deliveries. This wide range of prevalence estimates highlights the lack of a standard definition. The cause is impingement of the biacromial diameter of the fetus against the symphysis pubis anteriorly and the sacral promontory posteriorly. Why the shoulders do not descend in the oblique diameters of the pelvis is unclear, although sometimes the fetus is simply too large. Although the risk of shoulder dystocia rises with increasing birth weight, 40% to 50% of cases occur in infants whose birth weight is less than 4000 g.

Risk factors for shoulder dystocia ^, include fetal macrosomia, diabetes, a history of shoulder dystocia in a previous birth, and prolonged second stage of labor. Other factors that have been inconsistently reported as increasing the risk ^{, ,} include a history of macrosomia or postterm pregnancy, multiparity, obesity, and operative vaginal delivery from the midpelvis. In 50% of cases of shoulder dystocia, no risk factors are identified.

Maternal morbidity from shoulder dystocia includes postpartum hemorrhage and rectal injuries. The morbidity for the infant is attributable to asphyxia from delay in delivery or to trauma from the maneuvers used to deliver the fetus. Infant morbidity related to trauma includes brachial plexus and phrenic nerve injuries and fractures of the humerus and clavicle. The most serious traumatic morbidity is brachial plexus injury (Erb palsy), which occurs in 10% to 20% of infants born after shoulder dystocia. ^, If they are recognized early, 80% to 90% of brachial plexus injuries recover completely with proper physical therapy and, in some situations, neurosurgical management. For this reason, permanent neurologic injury is rare, occurring in 1 or 2 of every 10,000 births. This low prevalence greatly limits the ability to conduct prospective prevention studies.

Most brachial plexus injuries resulting from shoulder dystocia involve the arm and shoulder that are in the anterior pelvis at the time of delivery. The brachial plexus is believed to be injured when excessive downward traction and lateral extension of the fetal head and neck occur during the attempt to deliver the anterior shoulder ; however, there are exceptions to this cause of brachial plexus injury. Some infants with brachial plexus injury were born by vaginal delivery in which there was no evidence of shoulder dystocia. Also, brachial plexus palsy can involve the arm that was in the posterior pelvis at the time of delivery. Furthermore, Erb palsy has occurred in infants born by cesarean delivery. ^, Finally, several reports have described brachial plexus injuries in newborns that were confirmed by physical findings and electromyographic tests to have occurred before the onset of labor. ^, It is postulated that these injuries resulted from chronic nerve compression due to malposition in utero. ^, The presence of a permanent injury does not imply that the delivering clinician applied excessive force, despite the res ipsa loquitur argument seen in so many torts arising from these births.

Management of Shoulder Dystocia

Conventional wisdom dictates that the most effective treatment includes prompt recognition that delivery of the shoulders will be difficult and avoidance of excessive downward traction on the fetal head when attempting to deliver the anterior shoulder. Retraction of the fetal head immediately on its delivery (turtle sign) is an early warning that delivery of the shoulders may be difficult. Studies using simulated models of the fetus and pelvis have demonstrated that obstetricians frequently underestimate the amount of traction they apply to the fetal head. ^,

Several maneuvers have been useful in resolving shoulder dystocia. Hyperflexion of the mother’s thighs, known as the McRoberts maneuver, flattens the lumbosacral curve, thereby removing the sacral promontory as an obstruction to the inlet. ^, The knee-chest position tends to accomplish the same end. Suprapubic pressure can be applied in conjunction with the McRoberts maneuver.

Rubin described rotating the fetal shoulders into the oblique position by inserting the fingers of one hand vaginally behind the most accessible shoulder (usually the posterior) and pushing the shoulder toward the fetal chest. This is a substantial improvement on the commonly described Woods maneuver, which involves pushing the shoulder toward the fetal back.

If these maneuvers are not successful, the posterior arm of the fetus can be delivered if the obstetrician inserts one hand posteriorly, grasps the elbow, and draws the arm across the chest of the fetus. This maneuver may result in fracture of the humerus or the clavicle, which is a consistently remediable injury and preferable to a brachial plexus injury of the opposite arm. Finally, replacement of the fetal head in the uterus followed by cesarean delivery—the Zavanelli maneuver—may be necessary in rare instances.

Although the soft tissue of the perineum does not contribute to shoulder dystocia, many protocols recommend a wide episiotomy to facilitate one or more of the described maneuvers.

The successful management of shoulder dystocia is a matter of considerable obstetric judgment and skill. There is an inverse relationship between the incidence of brachial plexus injuries from shoulder dystocia and the experience of the obstetrician. Shoulder dystocia culminating in an injury that is permanent (1 in 10,000 births) is so rare that all of the recommendations on prevention are based on accumulated wisdom and opinion rather than evidence-based medicine. Athukorala and coauthors demonstrated the paucity of evidence for interventions in their systematic review and pointed to the flimsy foundation underlying any obstetrician’s criticism of another’s management of shoulder dystocia or suggestion that a particular approach in his or her own hands would have prevented an injury.

Hoffman and colleagues, using data from the Consortium on Safe Labor, studied 132,098 deliveries in which a total of 2018 cases of shoulder dystocia were reported, for an incidence of 1.5% (range among 12 centers, 0.2% to 3.0%). Among these cases, 101 babies (5.2%) had a birth injury, and 64 of those involved a nerve palsy. There were no neonatal deaths. The injury rates were the same whether the births were attended by residents, attending physicians, or midwives. Maneuvers performed are listed in Table 40.7 . As the number of maneuvers needed to resolve the shoulder dystocia increased, so did the likelihood of temporary birth injury ( Fig. 40.6 ). The McRoberts maneuver and suprapubic pressure were most often used as initial responses, and delivery of the posterior shoulder had the highest rate of success. However, because it is not known how many of the palsies were permanent, the conclusions of this study are applicable only to transient injuries.

TABLE 40.7

Outcomes of Obstetric Maneuvers ^a

From Hoffman MK, Bailit JL, Branch DW, et al. Consortium on Safe Labor. A comparison of obstetric maneuvers for the acute management of shoulder dystocia. Obstet Gynecol. 2011;117:1272.

Maneuver	Total N	N With Order Documented	First	Second	Third	Fourth	Fifth+	P Value	Overall	P Value
Rate of Success with a Particular Maneuver
McRoberts	1679	1123 (66.9%)	213/918 (23.2%)	49/186 (26.3%)	11/19 (57.9%)	—	—	.0067	273/1123 (24.3%)	<.001
Suprapubic pressure	1386	875 (63.1%)	58/116 (50.0%)	406/635 (63.9%)	74/116 (63.8%)	6/8 (75.0%)	—	.0002	544/875 (62.2%)	<.001
Delivery of posterior shoulder	262	179 (68.3%)	7/8 (87.5%)	28/32 (87.5%)	55/73 (75.3%)	40/45 (88.9%)	21/21 (100%)	.4642	151/179 (84.4%)	Referent
Rubin maneuver	86	50 (58.1%)	4/6 (66.7%)	4/5 (80.0%)	16/27 (59.3%)	9/12 (75.0%)	—	.7760	33/50 (66.0%)	<.005
Woods corkscrew	315	221 (70.2%)	14/19 (73.7%)	27/34 (79.4%)	78/114 (68.4%)	35/49 (71.4%)	5/5 (100%)	.7031	159/221 (72.0%)	<.005
Rate of Injury With Attempted Maneuver
McRoberts	1679	1123 (66.9%)	51/918 (5.6%)	15/186 (8.1%)	2/19 (10.5%)	—	—	.15	68/1123 (6.1%)	.25
Suprapubic pressure	1386	875 (63.1%)	6/116 (5.17%)	39/635 (6.1%)	10/116 (8.6%)	1/8 (12.5%)	—	.26	56/875 (6.4%)	.34
Delivery of posterior shoulder	262	179 (68.3%)	0/8 (0.0%)	0/32 (0.0%)	2/73 (2.7%)	8/45 (17.78%)	5/21 (23.8%)	<.0001	15/179 (8.4%)	Referent
Rubin maneuver	86	50 (58.1%)	0/6 (0.0%)	0/5 (0.0%)	5/27 (18.5%)	2/12 (16.7%)	—	.27	7/50 (14.0%)	.23
Woods corkscrew	315	221 (70.2%)	0/19 (0.0%)	0/34 (0.0%)	13/114 (11.4%)	7/49 (14.3%)	1/5 (20.0%)	.01	21/221 (9.5%)	.7

a The Cochran-Armitage test for trend was performed with delivery of posterior shoulder as the referent.

Grobman and colleagues studied the impact of a shoulder dystocia protocol focused on team response in a large urban tertiary care hospital. Informal training was concentrated on response to this event and not on specific maneuvers. Key elements were (1) an overt and unequivocal announcement of the diagnosis, (2) a simplified approach to summoning the team through overhead notification and a paging tree, (3) declaration of the duration of shoulder dystocia to team members on their arrival, (4) role clarity for various team members, and (5) structured documentation. The team consisted of nurses, available attending obstetricians, anesthesia, pediatricians, and obstetric residents. During the three 6-month periods before, during, and after introduction of the protocol, documentation improved and the incidence of brachial plexus palsy diagnosed at delivery declined from 10.1% to 4.0% and then to 2.6%. The authors could not comment on the rate of permanent injury given the low prevalence of that phenomenon. Although these results may be generalizable to other tertiary care settings, implementation would not be possible in smaller community hospitals that lack the resources of academic health centers. This report does add evidence for the arguments that a coordinated team response and prearranged implementation process can reduce transient injury from shoulder dystocia.

Placental Separation and Control of Uterine Bleeding

The third stage of labor is defined as the time from delivery of the infant to delivery of the placenta. For all practical purposes, it should also include the hour after the delivery of the placenta, because it is during this time that the patient is at greatest risk for postpartum hemorrhage.

After the infant is born, the uterus contracts and placental separation occurs by cleavage along the plane of the decidua basalis. Placental separation usually is complete by the time two contractions have occurred, although several additional contractions may be necessary to accomplish expulsion of the placenta from the uterus. Large venous sinuses are exposed after separation of the placenta, and control of bleeding from these sinuses depends primarily on contraction of uterine muscle and only secondarily on coagulation and thrombus formation in the placental site. The average blood loss during a normal vaginal delivery is about 600 mL. In the young, healthy parturient, acute blood loss is well tolerated because of the increased blood volume of pregnancy and the decrease in vascular volume that occurs with the reduction of the uteroplacental circulation at the time of birth.

Management of the placenta in the third stage is a matter of debate among qualified obstetricians. Elective manual removal of the placenta, if performed promptly, has been associated with no increase in puerperal morbidity. Advantages include the immediate identification of retained placental fragments and intrauterine extensions of cervical lacerations and the shortened time of placental removal. ^, However, manual removal is not a painless procedure in the unanesthetized patient, and it is unnecessarily invasive in most cases. Gentle massage of the uterine fundus encourages uterine contractions and helps one to detect changes in the shape of the uterus that signal placental separation. Vigorous fundal massage accomplishes nothing and is painful; if combined with excessive traction on the umbilical cord of a placenta implanted in the fundus of the uterus, it could promote uterine inversion.

The prophylactic administration of a uterotonic medication to reduce blood loss, either immediately after delivery of the infant or after delivery of the placenta, is a generally accepted practice, and prospective trials show that it decreases blood loss and reduces the need for therapeutic oxytocics. These trials also found no difference in effectiveness of ergot preparations compared with oxytocin, although there were nonsignificant trends toward increased need for manual removal of the placenta, increased need for blood transfusions, and increased blood pressure associated with the use of ergot alkaloids. Intravenous oxytocin is the drug of choice on most obstetric services. Prophylactic administration of the thermostabile prostaglandin E ₁ analogue misoprostol has been used to reduce bleeding in the third stage of labor. Though inferior to parenteral oxytocin in both efficacy and safety, orally administered misoprostol would be a reasonable choice as a uterotonic agent in low-resource settings when oxytocin is not available.