Search tips
Search criteria 


Logo of spsJournal HomeThiemeInstructions for AuthorsSubscribeAboutEditorial Board
Semin Plast Surg. 2004 November; 18(4): 309–317.
PMCID: PMC2884801
Obstetrical Brachial Plexus Paralysis, Part 1
Guest Editors Julia K. Terzis M.D., Ph.D. Saleh M. Shenaq M.D.

Current Concepts in the Management of Obstetrical Brachial Plexus Injuries: The Taipei Experience


Excluding patients in the early period before 1992 because of immaturity and patients after 2000 because the follow-up period is not long enough, we collected and operated for 78 patients with an initial obstetrical brachial plexus operation during the period of infancy with an average age of 4.9 months (range, 3–9 months). Another 10 patients were operated beyond the period of infancy, with an average age of 19.2 months (range, 12–30 months). The results show that, when indicated, surgical treatment is still more valuable than conservation in both ruptured and avulsed cases. Some conclusions are drawn. An initial obstetrical brachial plexus palsy with no elbow flexion but good hand function is not an urgent indication for early exploration. However, an obstetrical brachial plexus palsy with total hand palsy is an urgent indication requiring early nerve surgery within 3 months. Nerve grafts for ruptured cases should be as short and numerous as possible. Direction of nerve grafts to the planned target nerves yielded better results than the cable nerve grafts randomly, achieving less aberrant reinnervation between shoulder and elbow. Nerve transfers should be applied often in avulsed cases, including intraplexus and extraplexus transfers. Nerve transfer has the further advantage of minimizing the aberrant reinnervation between shoulder and elbow.

Keywords: Nerve surgery, obstetrical brachial plexus palsy

Obstetrical brachial plexus palsy (OBPP) can be divided into two distinct categories: (1) initial OBPP (infant or early OBPP), in which some require early nerve surgery, and (2) sequelae OBPP (child or late OBPP with deformity), in which many request surgical correction for deformities of the shoulder, elbow, forearm, or hand.1,2,3 The senior author (DCCC) has been operating for adult brachial plexus injuries since 1985 and for OBPP reconstruction since 1990, after gaining more experience with adults. In a recent report, the senior author evaluated 78 OBPP patients who were operated at Chang Gung Memorial Hospital between 1992 and 1999 (Table 1). The inclusion criteria for this study were rigid. Fifteen cases operated before 1992 were excluded because of immaturity. Some had no neck immobilization after surgery. Another 18 cases were also excluded after 2000 because of inadequate follow-up time. At least 4 years follow-up is absolutely required for OBPP patients. Sixty-eight patients received the brachial plexus operation during the period of infancy with an average age of 4.9 months (range, 3–9 months), and 10 patients were operated beyond infancy, with an average of 19.2 months (range, 12–30 months). This article deals with our approaches regarding surgery versus conservation, timing of surgery, clinical examination, reconstructive strategy for different degrees of injury, postoperative management, functional assessment, and finally the results following primary nerve surgery.

Table 1
Intraoperative Findings


Many rehabilitation therapists,4,5 and some surgeons6 continue to argue against early nerve surgery. They cite a high rate of spontaneous recovery (70 to 92%) from what they reported, high risks for surgery and postoperative care in infancy, and no significant improvement after surgery. However, since the development of the microscope, loop magnification, improved techniques of microsurgery (such as nerve transfers), and more understanding of the anatomy of the brachial plexus and pathophysiology of nerve degeneration and regeneration, more and more microsurgeons7,8,9 prefer early nerve surgery for OBPP patients. They have found that the rate of spontaneous recovery is actually very low, from 7 to 50%. For example, Zancolli and Zancolli1 evaluated 368 cases presenting with late OBPP. Only 36 cases (10%) needed no surgery; 90% of the cases needed surgical correction for deformities of the shoulder, elbow, or hand; and 239 cases (65%) received operation for shoulder sequelae. If there was no primary nerve surgery, the reconstruction performed in this late group was found to be difficult and required complex procedures to achieve reasonable results.10,11,12

The results for our 78 patients prove that early nerve surgery, if indicated, is still the mainstream and recommended treatment of choice for the initial OBPP patients (Tables 2, ,3,3, and and44).

Table 2
Results of Cases with Rupture Injury Alone
Table 3
Results of Cases with Rupture and Avulsion Injury
Table 4
Results of 10 OBPP Patients with Late Nerve Reconstruction


In the 78 cases of initial OBPP in this series, 56% of the patients had rupture injury with at least one root avulsion (44 of 78, 56%). Once root injury occurred, nearly 80% had at least two root avulsions (35 of 44, 79.5%), of which more than half (28 of 44, 64%) involved the lower plexus (C8,T1 ±C7) (Table 1). This means that once root injury occurred, the opportunity for lower plexus injury increased, resulting in more involvement of the hand functions.

Gilbert and Tassin7 recommended that at 3 months of age no elbow flexion is an indication for surgery. Total palsy with positive Horner's syndrome is an indication for earlier surgery, after 1 month of age. Clarke and Curtis8 recommended that at 9 months of age, failure to perform the “cookie test” (to put a cookie to the mouth) is an indication for exploration. From our investigation, the results for shoulder and elbow recovery following primary nerve surgery did not show any significant difference between 2 months and 11 months. This is also proved by the series of patients who had primary nerve surgery at age older than 1 year (Table 4). However, improvement in hand function was significantly poorer in the group with delayed primary nerve surgery (Table 4). Comparison between early nerve surgery for initial OBPP and late palliative reconstruction for late OBPP has been made (Table 5), showing that early nerve surgery in infant OBPP was not better than late reconstruction in late OBPP from the point of view of shoulder recovery . Moreover, hand reconstruction in patients with poor hand recovery is much more difficult than shoulder or elbow reconstruction in patients with poor shoulder and elbow recovery.10,11,12 The only opportunity to improve hand functions is provided by doing early nerve surgery, as was demonstrated in this study (Table 3).

Table 5
Comparison with Different Approaches (from the “Recovery of Shoulder” Point of View)

The results of this study demonstrate that Gilbert's “rule of 3 months” is an overestimate of the poor results for shoulder and elbow. On the other hand, Clarke's “rule of 9 months” is an underestimate of the poor results for forearm and hand. Early surgical treatment within 3 months is absolutely valuable in patients with associated lower plexus injuries (total palsy) but relatively valuable in patients with only upper plexus injuries (Erb's palsy). Poor shoulder and elbow function shown in infant OBPP is not an urgent indication for surgery. However, hand palsy is an urgent condition in infant OBPP that requires early exploration within 3 months. Therefore, our recommended timing of surgery is somewhere between those of Gilbert and Clarke. At 3 months of age, no biceps function with little or no hand function is an indication for exploration. At 3 months of age, if there is no biceps function but there is functional wrist or finger extension, we recommend “wait and see” for another 3 months. At 6 months of age, if there is still no elbow flexion, exploration is indicated. After 1 year of age, poor shoulder or elbow function is still an indication for exploration. But poor hand function at this late age is not an indication for exploration.


Physical examination of the newborn infant with OBPP, although challenging, is essential and more important than other diagnostic tests including electrodiagnosis and imaging studies. Observation of active movement of the infant while lying on the examination bed is more valuable than the strength detected. Usually the M2 (movement without weight resistance) strength score in a newborn infant is adequate to predict a good result. Even mild movements of elbow flexion, wrist extension, or finger flexion are given a score of M2. Horner's syndrome (ptosis is easily noted, but myosis, enophthalmos, and anhidrosis are difficult to detect) is not a reliable sign in OBPP. During the exploration, a patient with positive Horner's syndrome is often noted to have an intact C8 and T1. Disappearance of Horner's syndrome is also often noted when the patient grows up.

C7 is the most commonly injured spinal nerve (Tables 2 and and3).3). When injury of C7 is found intraoperatively, either rupture or root avulsion injury with a preoperative wrist drop, the nearby C6 and C8 are expected to be injured too, partially or completely. When M2 wrist extension is found on clinical examination, it can be taken as a good sign that C8 and T1 are intact. Therefore, the presence of wrist extension is a good sign to wait for another 3 months. Good finger flexion and extension even with drop wrist is also a valuable sign to wait another 3 months for observation.


The operated baby is always put in a supine position with endotracheal intubation and general anesthesia. An artery line (radial artery on the contralateral side or femoral artery) and a central venous pressure (CVP) line through the femoral vein or the internal jugular vein on the contralateral neck are critically required for postoperative care and should be well set and protected during and after the surgery.

A C-curved incision along the posterior border of the sternocleidomastoid muscle, curved laterally parallel to and above the clavicle, stopped with a little curve laterally at the junction of the deltopectoral groove and clavicle is a routine incision for the supraclavicular approach. The platysma in infants is very thin and invisible; therefore the incision can be deep to the surface of the sternocleidomastoid muscle. The posteriorly based skin flap is elevated. The omohyoid muscle is usually preserved. The underlying adipofascial tissue contains rich lymphatic ducts, lymph nodes, and transverse cervical vessels. A similar C-shaped dissection along the internal jugular vein medially and subclavicular vein inferiorly is made and the posteriorly based adipofascial flap is elevated; the scalene anterior muscle, phrenic nerve, and beneath the brachial plexus are then seem. All the lymphatic tissues including ducts coming from the deep plane of the internal jugular vein should be coagulated first and then divided to avoid postoperative lymph leak. The transverse cervical vessels are routinely preserved as much as possible.

The phrenic nerve should be isolated and carefully protected without traction or long-time compression, which may cause postoperative transient phrenic nerve palsy and prolong the endotracheal intubation period. The scalene anterior muscle, if injured and mixed with the underlying neuroma tissue and fibrotic change, should be resected, sometime nearly totally, to expose the underlying injured neuroma, subclavian artery, and more proximal spinal nerves. The subclavian artery should be carefully dissected and protected. The nerve beneath the subclavian artery is C8 or the lower trunk. The whole spinal nerves, C5 to T1, should be identified and their health assessed. Neuromas of the upper and middle trunk due to rupture are often seen. Microneurolysis is required for each neuroma to judge the severity of scar tissue or degree of nerve injury. If the scar is bad, the neuroma is resected and nerve grafts are placed for reconstruction. If the scar is acceptable with some fascicles still in continuity, only microneurolysis is performed. If the lesions involve the division more distally, an extended incision down to the deltopectoral groove is made. The triangle rounded by the main branch of the cephalic vein on one side and the clavicle on the other side is approached. From here under the subclavius muscle, the supra- and infraclavicular fossae are connected. The clavicle can be elevated easily and the reconstructive procedure continued.


Nerve grafts and nerve transfers are the most common operative procedures for the initial OBPP. In the cases with rupture injury alone (34 patients, Table Table6),6), most ruptured lesions were located at the trunk level. Nerve grafts were the main reconstructive procedure. Microneurolysis was used as an adjuvant procedure. In most cases, the suprascapular nerve and anterior and posterior divisions of the trunks were found in the supraclavicular fossa through the supraclavicular approach. Short nerve grafts (2–3 cm) with cable (4–6 nerve grafts as a cable) without any specific arrangement, from proximal trunk or spinal nerve stumps to the distal trunk stump (group 1, Table Table2),2), seemed to provide good results in shoulder and elbow function (Fig. 1A–D). But the group series was small (only five patients). However, they all had significant combined movement (synkinesis) between shoulder and elbow. Long nerve grafts (4–6 cm) interposed between C5 and C6 spinal nerves to the infraclavicular terminal nerves (group 2, Table Table2)2) seemed to provide the worst results. The elbow flexion strength was M2–3 with a severe trumpet angle, which was due to weak elbow and not to aberrant reinnervation. The result for shoulder elevation was fair with an average of 90 degrees (range, 80–100 degrees), and shoulder external rotation was poor with an average of 50 degrees (range, 40–60 degrees). The results obtained with cable short nerve grafts (3–4 cm) directed to the posterior division of the upper trunk from C5 and the anterior division of the upper trunk from C6 (group 3, Table Table2)2) seem good and constant. Shoulder abduction was 132 degrees on average (range, 80–180 degrees), and shoulder external rotation was good to fair, 67 degrees on average (range, 50–90 degrees). Elbow flexion was M3 in most cases.

Table 6
Functional Assessment for OBPP Postoperatively
Figure 1
(A) An infant with left OBPP who was operated at the age of 4 months. Rupture of upper and middle trunks was noted intraoperatively. Interposition of nerve grafts from C5 to the suprascapular nerve and posterior division, C6 to the anterior ...

For the cases of rupture injury with at least one root avulsion (Table 3), intraplexus neurotization for the avulsed C8 or lower trunk or median nerve from C5, C6 did result in improvement in hand functions (Table 3). Extraplexus neurotization by intercostals nerve transfer to the musculocutaneous nerve could resolve the problem of elbow flexion, and C6 was utilized for the hand functions(Fig. 2A–F). Spinal accessory nerve transfer to the suprascapular nerve could resolve shoulder problems, and C5 was utilized for hand functions. In three patients, the contralateral C7 was transferred to the median nerve for hand function: two using free nerve grafts (8 cm × four nerve grafts), one using a segment of vascularized ulnar nerve graft. All achieved good to fair (M2–3) finger flexion.

Figure 2
(A) An infant with left OBPP who was operated at the age of 4 months. Rupture of upper trunk and C7-T1 three-root avulsions were noted intraoperatively. C5 nerve grafting to the suprascapular and radial nerve, C6 nerve grafting to the ...

We used contralateral C7 transfer in the adult brachial plexus injuries very often13 but used very few in the OBPP patients (only three patients). The reasons include the fact that the proximal remaining stumps and nearby nerves for transfer are usually adequate for shoulder, elbow, and hand reconstructions. In addition, ipsilateral spinal nerve (C5, C6, or C7) transfer is always better than contralateral C7 transfer. Contralateral C7 transfer is good only for finger flexion (small resistance of the finger weight) and not for shoulder or elbow function (more resistance from the whole or half upper limb weight).

Oberlin et al14 used part of the ulnar nerve transfer to the musculocutaneous nerve and was able to achieve good results for elbow flexion. However, this was performed only in adult patients. Al-Qattan15 used Oberlin's method to treat Erb's birth palsy. In this series, we did not perform this procedure in OBPP patients with Erb's palsy. In cases with combined C5 and C6 two-root avulsions, Oberlin's method may be considered instead of using intercostal nerve transfer, which is more demanding in terms of technique and time. But intercostal nerve transfer to the musculocutaneous nerve is actually very effective and yields high success rates (M > 3) in this group of OBPP patients.

Microneurolysis is one of the operative techniques used when a lesion in continuity is found. The results of this procedure are quite unpredictable. Intraoperatively, under the microscope, the amount of perineural or intraneural scarring is judged. If the scar is found to be moderate or severe, resection of neuroma and nerve grafting are performed. If there are only mild or acceptable scars, microneurolysis alone may be adequate. Nevertheless, if there is any doubt, resection and nerve grafts are preferable.

In four-root avulsions, the proximal stump of ruptured C5 is usually not healthy enough for intraplexus transfer. This is quite different from the adult pattern. The reason is that the whole five roots of the spinal nerve are so much closer in a small fossa in OBPP than in the adult. Using C5 stump transfer to the C8 or median nerve is therefore not recommended. The proximal C5 stump in cases of C6-T1 four-root avulsions is better transferred to the anterior division of the upper trunk for elbow instead of for hand function.

Of the 10 cases who had primary nerve surgery at an age older than 1 year (Table 4), all had at least one root avulsion. The most significant benefits from these late primary nerve reconstruction were nearly 100% improved shoulder and elbow functions (Table 4); however, little or no improvement was achieved for the hand functions. One patient had contralateral C7 transfer to the median nerve, using a vascularized ulnar nerve graft (14 cm), but without improved hand function after 3 years of rehabilitation. Further reconstruction for the hand is required.


Immediate postoperative neck splinting is applied for every case. The patient is transferred to the intensive care unit without extubation. The endotracheal tube is always left for 3 days or longer (if the diaphragm is involved in palsy) with the help of a ventilator machine. Once the ventilation tube is removed, the patient is observed for one more day and transferred to a regular ward bed for another couple of days. Usually hospitalization is about 1 week. The neck splint is kept for 3 weeks. Regular follow-up in the rehabilitation center and outpatient clinics for at least 4 years is very important and should be explained to the patient's parents before and emphasized again after the discharge. Electric muscle stimulation twice a day at home may start in the fourth postoperative week and continue for at least 1 year or until the muscle has function at the M2 level.


There is a lack of consensus about the method of assessment7,8,16 because of the complexity of the lesions (e.g., aberrant reinnervation), complexity of the repairs (e.g., different surgical strategies), involvement of different parts of the limb (shoulder, elbow, forearm, and hand), and age of the patients, who are too young to cooperate well. The reported methods of functional assessment include the Medical Research Council Muscle Grading System, Gilbert and Tassin Muscle Grading System,7 Clark and Curtis Active Movement Scale,8 Narakas Grading System,17 and others.13 Although the Narakas17 and Raimondi13 grading systems include evaluation of sensation, sensory disturbance of the hand in OBPP patients is rarely a major factor.

All patients in this study were observed for at least 4 years. The results were evaluated on the basis of shoulder abduction (elevation), shoulder external rotation, elbow flexion and extension, and finger flexion (Table 6) and were divided simply into “good,” “fair” and “poor.” For shoulder abduction, the modified Gilbert six-stage system4 is used: stage 0, no abduction; stage 1, abduction angle less than 45 degrees; stage 2, between 45 and 90 degrees; stage 3, between 90 and 120 degrees; stage IV, between 120 and 160 degrees; stage 5, normal. If the elevation angle is greater than 120 degrees, it is accepted as a good result; between 120 and 90 degrees is considered a fair result; and less than 90 degrees is defined as a poor result. For shoulder external rotation, the patient is asked to place the hand on the occipital region. Good means that the hand can touch the occipital region (shoulder external rotation ≥ 90 degrees); fair means that the hand can reach the ear (shoulder external rotation 60–90 degrees); poor means that the hand is at a line of prechest or pre-face (shoulder external rotation is < 60 degrees). For elbow flexion, the Medical Research Council Scale system (M0–5)14 is accepted for evaluation. M3 or greater is considered good. The patient can reach the mouth easily against one finger resistance without tilting the head anteriorly. Fair means a score of M2; the patient can flex the elbow to touch the mouth only when the patient abducts the shoulder more than 80 degrees (trumpet sign). If the score is less than M2, it is defined as a poor result: it is impossible to have the hand touch the mouth.

For evaluation of hand function, the Gilbert Muscle Grading System16 is more accepted but is modified. Good (M =3) means finger movement against the weight of one finger resistance. Fair (M =2) means finger movement with weight eliminated. Poor (M < 2) means a finger with little or no movement. M3 is considered functional and successful.


The results for the group of patients with rupture injury alone, patients with rupture with at least one root avulsion injury, and patients with delayed nerve surgery are summarized in Table Table2,2, ,3,3, and and44.

We conclude that an initial OBPP with no elbow flexion but good hand function is not an urgent indication for early exploration. However, an OBPP with total hand palsy is an urgent indication requiring early nerve surgery within 3 months. Short nerve grafts and as many nerve grafts as possible for the rupture injuries, direction of nerve grafts to the target nerves instead of a randomized arrangement, and nerve transfer (intraplexus or extraplexus) are preferable and recommended.


  • Zancolli E A, Zancolli E R. Palliative surgical procedures in sequelae of obstetrical palsy. Hand Clin. 1988;4:643. [PubMed]
  • Leffert R D, Pess G M. Tendon transfers for brachial plexus injury. Hand Clin. 1988;4:273. [PubMed]
  • Chuang D CC, Ma H S, Wei F C. A new evaluation system to predict the sequelae of late obstetrical brachial plexus palsy. Plast Reconstr Surg. 1998;101:673. [PubMed]
  • Brown K L. Review of obstetrical palsies, nonoperative treatment. Clin Plast Surg. 1984;11:181. [PubMed]
  • Greenwald A G, Schute P C, Shiveley J L. Brachial plexus birth palsy, a 10-year report on the incidence and prognosis. J Pediatr Orthop. 1984;4:689. [PubMed]
  • Kline D G, Hudson C G. In: Kline DG, editor. Nerve Injuries. Philadelphia: WB Saunders; 1995. Birth palsies. pp. 461–471.
  • Gilbert A, Tassin J L. In: Terzis JK, editor. Microreconstruction of Nerve Injuries. Philadelphia: WB Saunders; 1987. Obstetrical palsy: a clinical, pathologic, and surgical review. pp. 529–553.
  • Clarke H M, Curtis C G. In: Gilbert A, editor. Brachial Plexus Injuries. London: Martin Dunitz; 2001. Examination and prognosis. pp. 159–172.
  • Terzis J K, Liberson W T, Levine R. In: Terzis JK, editor. Microreconstruction of Nerve Injuries. Philadelphia: WB Saunders; 1987. Our experience in obstetrical brachial plexus palsy. pp. 513–528.
  • Chuang D CC, Ma H S, Wei F C. A new strategy of muscle transposition for treatment of shoulder deformity caused by obstetric brachial plexus palsy. Plast Reconstr Surg. 1998;101:686. [PubMed]
  • Chunag D CC, Hattori Y, Ma H S, Chen H C. The reconstructive strategy for improving elbow function in late obstetric brachial plexus palsy. Plast Reconstr Surg. 2002;109:116. [PubMed]
  • Chuang D CC, Ma H S, Borud L J, Chen H C. Surgical strategy for improving forearm and hand function in late obstetric brachial plexus palsy. Plast Reconstr Surg. 2002;109:1934. [PubMed]
  • Chuang D CC. Contralateral C7 transfer (CC7T) for avulsion injury of the brachial plexus. Techniques in Hand and Upper Extremity Surgery. 1999;3:185–192. [PubMed]
  • Oberlin C, Beal D, Leechavengvongs S, et al. Nerve transfer to biceps muscle using a part of ulnar nerve for C5–6 avulsion of the brachial plexus: anatomical study and report of four cases. J Hand Surg [Am] 1994;19:232. [PubMed]
  • Al-Qattan Oberlin's ulnar nerve transfer to the biceps nerve in Erb's birth palsy. Plast Reconstr Surg. 2002;109:405. [PubMed]
  • Clarke H M, Curtis C G. An approach to obstetrical brachial plexus injuries. Hand Clin. 1995;11:563. [PubMed]
  • Narakas A O. In: Lamb DW, editor. The Paralysed Hand, Volume 2. Edinburgh: Churchill Livingstone; 1987. Obstetrical brachial plexus injuries. pp. 116–135.

Articles from Seminars in Plastic Surgery are provided here courtesy of Thieme Medical Publishers