Shoulder Dystocia





Is all brachial plexus injury caused by shoulder dystocia and/or “excessive” physician traction?

In his 2002 paper, Pecorari states the following:

Unfortunately for obstetricians and midwives, in court Erb palsy has been causally connected with shoulder dystocia and errors in management, although it is not always true. Perhaps the lack of an obvious explanation has contributed to the identification of the birth attendant as a handy scapegoat.

When there is a permanent brachial plexus injury following shoulder dystocia, responsibility for this injury is often presumed to lie with the obstetrician who supposedly did not foresee that a shoulder dystocia was likely to occur or mishandled it when it did. Yet a review of the literature does not substantiate such assumptions.

Gherman in his 1998 paper summarizes the refutation to these claims:

We feel that some cases of brachial plexus injury are unavoidable events. Recent reports have noted that brachial plexus palsies occur:

(1) In the absence of characteristic risk factors

(2) In the absence of shoulder dystocia

(3) In the posterior arm of infants whose anterior shoulder was impacted behind the symphysis pubis

(4) In vertex-presenting fetuses delivered by atraumatic Cesarean section

(5) Without apparent relationship to the type or number of maneuvers used to disimpact the fetal shoulder

(6) In association with other peripheral nerve injuries

(7) With electromyelographic evidence of muscular denervation during the immediate postpartum period [subsequently disproven].

Jennett commented in a similar vein in 1997:

Evidence continues to accumulate that renders a univariate theory of the causation of brachial plexus injury untenable . . . Intrauterine maladaption is responsible for some instances of brachial plexus injury.

He notes that in his series of deliveries from 1977 to 1990, 22 of 39 (56%) brachial plexus injuries were not associated with shoulder dystocia. In that same paper he quotes Pearl (1993) and Gimovsky (1995), both of whom reported brachial plexus injuries in babies delivered from the occiput posterior position without shoulder dystocias. He further cites Walle (1993) who observed in his patient population that 1/3rd of 175 shoulder dystocias involved the posterior shoulder.

There are many other similar reports:

Hardy (1981) reported 36 infants with brachial plexus injuries of whom only 10 had shoulder dystocia noted at birth.

Gilbert (1990) initially published a study of 1000 infants with brachial plexus injury in which 39% did not have shoulder dystocia at delivery. In a supplementary article in 1999, he reported that 47% of babies with brachial plexus injury in his now larger series experienced deliveries in which no shoulder dystocia was noted. Even among macrosomic fetuses in this series, 26% of brachial plexus injuries occurred in the absence of shoulder dystocia.

Gonik (1991) reported that 71% of all injured infants in his series were the product of deliveries without recognized shoulder dystocia.

Sandmire (1996) reported 17 babies in his series of 36 with brachial plexus injuries whose deliveries did not involve shoulder dystocia. This article included his personal review of the literature concerning brachial plexus injury (BPI) with and without shoulder dystocia:


# of deliveries greater than  4,500 grams

BPI with shoulder dystocia BPI without shoulder dystocia
Lipscomb (1995) 157 7/12 5/12
Mentocoglou (1992) 589 5/9 4/9
Sandmire (1996) 547 9/19 10/19
TOTALS 1,727 34/69 (49%) 35/69 (51%)

As can be seen, 51% of brachial plexus injuries in over 1727 deliveries of macrosomic babies did not involve shoulder dystocias.

Gram (1997) noted that only 8 babies had shoulder dystocia deliveries in a group of 15 who experienced brachial plexus injury. In the other seven cases, there had been no birth trauma.

Hillard's series of babies with Erb palsy (1997) described 15 of 51 babies who had not experienced shoulder dystocia during delivery.

Torki (2012) presented a series of eight cases from the University of Southern California where there was no shoulder dystocia yet there was severe brachial plexus palsy. While the authors do not state whether the injuries to these neonates were permanent, they were severe enough that the infants had to be admitted to the neonatal intensive care unit.

El-Sayed (2013) studied obstetrical brachial plexus palsy over the last 18 years in his clinic in Saudi Arabia. Of 751 cases 33% resulted from routine deliveries. They conclude that the normal forces of labor and delivery can lead to obstetric brachial plexus palsy.

Iskender (2014) reviewed 44,092 vaginal deliveries from his hospital in Ankara Turkey between 2009 and 2013. Among six neonates with permanent brachial plexus injury, five of them had not experienced shoulder dystocia.

Zuarez-Easton et. al. (2015) evaluated 83,806 deliveries from their hospital in Israel. There were 144 cases of brachial plexus injury for a rate of 1.7 per thousand. In 41 of 144 cases (28.9%) there was no documented shoulder dystocia—and this in an era and facility where shoulder dystocia simulation and documentation were both being practiced.

Gurewitsch and Allen (2006), while reporting that 90% of their persistent brachial plexus injury cases were related to documented shoulder dystocia, show that 10% of permanent injuries either had no documentation of or no occurrence of shoulder dystocia. The object of their paper was to ascertain whether brachial plexus palsy that occurs without shoulder dystocia represented a traction injury during unrecognized shoulder dystocia or a natural phenomenon with a different mechanism of injury. They compared risk factors and outcomes between shoulder dystocia-associated and non-shoulder dystocia-associated brachial plexus palsy cases using two datasets, one from Johns Hopkins Hospital from 1993 to 2004 and the other from a series of litigated cases.

In 11.5 years at Johns Hopkins Hospital there were 135 brachial plexus palsies among 23,273 deliveries. There were 206 cases in the litigated series. Of these, in 1 of 8 in the hospital group and in 6 of 206 of the litigated cases, no shoulder dystocia was documented. Gurewitsch and Allen state in the last paragraph of their paper that

The current investigation supports that intrauterine and intrapartum phenomena can contribute to the mechanism of birth -related brachial plexus palsy.

They conclude that non-shoulder dystocia brachial plexus palsy is real though uncommon and likely occurs by modes of injury mechanically distinct from shoulder dystocia brachial plexus palsy.

Finally, the definitive 2014 ACOG report on Neonatal Brachial Plexus Palsy states

“Neither high-quality nor consistent data exist to suggest that neonatal brachial plexus palsy can be caused only by a specific amount of applied force beyond that typically used by healthcare providers during any delivery. Instead, available data suggests that the occurrence of a neonatal brachial plexus palsy is a complex event, dependent not only on the forces applied at the moment of delivery, but also on the constellation of forces…. that have been acting on the fetus during the labor and delivery process as well as individual fetal tissue characteristics.

The report goes on to say that

“Recent multidisciplinary research now stresses that the existence of neonatal brachial plexus palsy following birth does not a priori indicate that exogenous forces are the sole cause of this (i.e. brachial plexus nerve) injury.”

What does cause brachial plexus injuries?

In court, the standard explanation for a brachial plexus injury is that it results from excessive downward traction by the obstetrician on the fetal head during an attempt to resolve a shoulder dystocia. This supposedly overstretches the brachial plexus thus injuring it.

Yet significant endogenous forces are generated through the natural physical events of labor. The forces of both uterine contractions and maternal pushing move a fetus from the uterus through the birth canal and out of the maternal pelvis. Stretching of the brachial plexus occurs during this process, as shown by both computer simulations and physical models. The stretch results from differential motion between the fetal head and shoulders after some element of the maternal anatomy—usually the symphysis pubis--retards the progression of the fetal shoulders.

The tractor-trailer theory

Sandmire (2000) was among the first to describe how the forces of labor and maternal pushing could result in brachial plexus injuries unrelated to physician traction at delivery. Sandmire studied what happens to the various parts of the fetus during uterine contractions and maternal pushing. He noted that the forces of contractions and maternal pushing act on the long axis of the fetus. If the fetus's anterior shoulder were to get stuck behind the maternal pubic bone and continued pressure was applied to the long axis of the fetus, the baby's brachial plexus would undergo considerable stretching.

This may be compared to what happens when a tractor-trailer truck gets stuck under a low overpass. While the cab may pass under the bridge, the trailer -- taller than the cab--may get caught on the overpass and will be unable to emerge no matter how hard and frequently the driver in the cab “guns” the engine. This process will, however, generate large stretching forces in the connectors between the tractor and the trailer. Sandmire suggests that an equivalent force acts upon a baby's brachial plexus during shoulder dystocia deliveries.

Evidence of such stretching is clinically observed with the “turtle sign”. This phenomenon is encountered at the moment of delivery when, after the head emerges, it is often pulled back against the perineum with significant, forceful recoil. This “recoil” is caused by the spring-like action of the markedly stretched neck skin, muscle, and brachial plexus tissue during the mother’s last push to deliver her baby’s head. During this last thrust, the shoulders remain unmoved--restrained by the mother’s symphysis pubis—while the head emerges several centimeters out from the vagina orifice, greatly stretching the brachial plexus nerves.

Stretching of the brachial plexus nerves does not only occur during the last push of labor. It likely takes place through much of the late second stage when the fetus’s head is low in the pelvis yet the fetal body is prevented from moving lower in the birth canal because of obstruction of its shoulders on the mother’s pubis.

Forces involved in shoulder dystocia deliveries

In 1991 Robert Allen published an article wherein it was suggested that 100lbs of force was the amount necessary to injure a brachial plexus, a claim which he has subsequently made in other forums (Allen, 1999). Yet this data was based on the measurement of a single baby who experienced a temporary brachial plexus at birth. Moreover Gonik—who was the co-author with Allen on the 1991 paper—subsequently showed (2000) that maternal endogenous forces—those of maternal contractions and pushing—resulted in several fold the force level applied by a delivering physician.

Deering (2011) also published an article in which he expressed skepticism about the “rather arbitrary cutoff of 100 N [that] is generally accepted as the maximum force that should be applied during a delivery”. He reports how studies by Croft in the United Kingdom show that forces exceeded 100 Newton in more than two thirds of shoulder dystocia delivery simulations. He concludes that either simulations do not correctly represent what happens at regular deliveries or the 100 Newton number is incorrect. If not, than a very large percentage of babies would suffer brachial plexus injury at birth.

Another point—made by Allen (2005) —is that some fetuses may be more predisposed than others to brachial plexus injury merely because of “biologic variation” That is, some fetuses may have structurally weaker nerve tissue than others, be born in a position that more significantly stretches this tissue, or at the time of delivery--due to hypoxia or other causes—have less protection from surrounding intact muscles than do other neonates.

There is another force at birth that operates on the fetal brachial plexus: compression. Gonik (2003) has shown that during a shoulder dystocia the fetus’s shoulder and neck are pushed against the maternal symphysis pubis with such large pressures that the neonate’s brachial plexus may be injured. This methodology of injury is contested by some who claim that compression cannot cause brachial plexus injuries. Yet we know from many other examples in obstetrics and gynecology that compression injuries can cause permanent nerve damage: forceps damage to facial nerves, self-retaining retractor damage to obturator and femoral nerves, etc.

The 2014 ACOG report “Neonatal Brachial Plexus Policy” discusses the forces potentially involved in brachial plexus injuries in detail. They conclude:

Neonatal brachial plexus palsy can occur even when axial traction is properly applied. The occurrence of this injury does not automatically indicate that the practitioner applied forces or maneuvers that caused the nerve injury. The forces of uterine contractions and maternal pushing alone are probably sufficient to cause excessive traction on the brachial plexus. Many cases of brachial plexus injury occur independently of shoulder dystocia or excessive force by the provider. A substantial portion of neonatal brachial plexus palsy cases are not associated with antecedent shoulder dystocia. These injuries occur during cesarean section deliveries.

Posterior shoulder

Most brachial plexus injuries occur to a baby's right arm (60%). This is because babies most commonly "present" into the mother's pelvis in the left occiput anterior position (LOA). The LOA position is when the back of the baby's head -- the occiput -- points towards the mother's left arm while the fetal face is oriented towards the mother's right buttock. In this fetal position the baby's right arm will be anterior -- and thus more likely to get caught under the mother's pubic bone. But many brachial plexus injuries have also been reported in the posterior shoulder. It is thought that in these cases the posterior shoulder gets caught on and restrained by the sacral promontory while the remainder of the baby is being pushed forward by the mother's expulsive efforts and uterine contractions.

Brachial plexus injuries following Cesarean section

Reports of brachial plexus injury in the absence of shoulder dystocia are subject to the criticism that perhaps shoulder dystocias were under-reported or that "excess' traction might have been placed on the baby's head during the course of a routine delivery. But reports of brachial plexus injury following Cesarean section are less subject to criticism. There are many such reports in the literature:

Ecker (1997): Two infants born by Cesarean section who sustained brachial plexus injuries, one of a nondiabetic mother, the other of a diabetic mother.

Hardy (1981): Two infants born in vertex position at Cesarean section who sustained brachial plexus injuries.

McFarland (1986): Four patients delivered by Cesarean section who experienced brachial plexus injuries.

Graham (1997): Report of an Erb palsy from cesarean section.

Gilbert (1999): Evaluated data on all brachial plexus injuries from California in the years 1994 to 1995. Of the 1,094,298 babies born in those two years there were 1,611 brachial plexus injuries reported (0.15%). Of these, 60 of the injured babies were born via cesarean section.

The above is only a brief sample of many such papers.

Arguments by claimants that brachial plexus injury is always or almost always caused by physician traction

A study by the Swedish obstetrician Mollberg (2008) is often quoted as demonstrating that brachial plexus injury is correlated with an increase in physician traction during delivery. However a closer look at her data reveals an interesting feature: Of the 18 cases of permanent neonatal brachial plexus palsy she describes, fundal pressure was used in the delivery of 17 of them. Thus, as the 2014 ACOG publication on Neonatal Brachial Plexus Palsy states,

Despite the conclusions by Mollberg and colleagues that forceful downward traction was causative, it is not possible to separate out the effect of increased traction from the effect of increased expulsive force induced by the application of fundal pressure.

Others who claim that physician expulsive forces are the only cause of permanent brachial plexus often cite cadaveric studies in an attempt to show that the amount of force necessary to sever the nerves of the brachial plexus are sometimes exceeded by physicians during deliveries. But most of these quoted studies are between 50 and 100 years old: Sever 1916, Adson 1922, Morris 1955. The author of a more recent study (Metaizeau (1979) says that his data is not a good means of assessing the forces in a live baby during a delivery complicated by shoulder dystocia. As the ACOG 2014 publication suggests, such anatomic studies are “quite crude” by today’s standard of biomechanics. Furthermore, “they do not provide a complete picture of how and why neonatal brachial plexus palsy may occur during delivery”. The ACOG publication concludes this section of the report by stating:

It is inappropriate to conclude that lateral bending is the only cause of the injury on the basis of these early studies when similar research has not examined other mechanisms.”

The ACOG 2014 publication from the Neonatal Brachial Plexus Palsy Task Force

This task force, convened by James Breeden MD, then President of the American College of Obstetricians and Gynecologist (2012-2013), and Albert Strunk JD, MD, Deputy Executive Vice President of the American College, was established to provide “a well-researched, informative, objective, and dispassionate presentation of the existing state of knowledge” concerning neonatal brachial plexus palsy. The members who made up the panel are among the most knowledgeable experts in the world on shoulder dystocia and brachial plexus injury, having between them published scores of articles on this topic. The following are some conclusions from this ACOG report:

The existence of neonatal brachial plexus palsy following birth does not a priori indicate that exogenous forces are the sole cause of this injury p. ix

Neonatal brachial plexus palsy also has been shown to occur entirely unrelated to traction, with studies demonstrating cases of both transient and persistent neonatal brachial plexus palsy in fetuses delivered vaginally without clinically evident shoulder dystocia or fetuses delivered by cesarean section without shoulder dystocia. P. 17

Traction applied in the plane of the fetal cervical-thoracic spine is typically along a vector estimated to be 25 – 45° below the horizontal plane when the woman in labor is in lithotomy position. P. 24

Uterine contractions result in a compression force to the fetus that acts to move the entire fetus down the birth canal. If, during this movement, a structure obstructs a body part while another body segment continues moving forward, the difference in motion can result in either a pulling force on the tissues that connect the two regions or a bending force on a rigid, bony body part. P. 24.

Maternal forces, the combination of uterine contractions and maternal pushing, are likely to be at least 140 – 160 N during the second stage of labor when intrauterine pressure of 120 mmHg is common….. Thus it can be anticipated that maternal forces during the second stage of labor will reach at least 140 N with an average size fetus. P. 25

If a shoulder is restrained, maternal forces will continue to move the head and neck forward, widening the angle between the neck and the shoulder and causing traction on the brachial plexus. P. 27

Gherman and colleagues hypothesized that several mechanisms of injury may occur, depending on the characteristics of the fetus and the delivery, for example:

Continued movement of the head following impaction of the anterior shoulder behind the symphysis pubis or impaction of the posterior shoulder on the sacral promontory.

Normal downward traction applied by the physician in the presence of observed shoulder dystocia.

Compression of the brachial plexus against the symphysis pubis.

Abnormal intrauterine pressure arising from uterine anomalies or uterine hypertonicity. P. 28

No published clinical or experimental data exist to support the contention that the presence of persistent (as compared to transient) neonatal brachial plexus palsy implies the application of excessive force by the birth attendant. P. 28

In summarizing the pathophysiology and causation of neonatal brachial plexus injury, the ACOG report states:

Neither high-quality nor consistent data exist to suggest that neonatal brachial plexus palsy can be caused only by a specific amount of applied force beyond that typically used by healthcare providers and experienced during a delivery without neonatal brachial plexus palsy. Instead, much of the data suggest that the occurrence of neonatal brachial plexus palsy is a complex event, dependent not only on the forces applied at the moment of delivery, but also on the constellation of forces (e.g., vector and rate of application) that have been acting on the fetus during the labor and delivery process, as well as individual fetal tissue characteristics (e.g., in situ strain and acid-base balance). P. 37

As has been shown, there is much evidence that not all instances of brachial plexus injury are due to the actions of a physician during a shoulder dystocia delivery. Thus the automatic assignment of responsibility to an obstetrician or midwife for a brachial plexus injury is inappropriate and is not supported by the obstetrical literature.

next >>

Copyright © 2017 Henry Lerner  

View printer friendly version of Shoulder Dystocia or PDF

Home | Text | Conclusions | Bibliography 
Shoulder Dystocia Registry | Contact


Shoulder Dystocia

Designed and developed by FSB Associates
Maintained by Danielle Campisi