Once a decision is made to operate on a patient for persistent NBPP, a number of intraoperative challenges face the nerve surgeon. The main decision is what intervention to perform: neurolysis alone, nerve graft repair, or nerve transfer. For a number of reasons, this decision remains challenging. One main reason is the lack of comprehensive postoperative data that allow head-to-head comparison of interventions. This will be discussed in the next section. With the available data, how does the nerve surgeon make this decision?

In adults, recoding nerve action potentials (NAPs) across a lesion in continuity can be helpful. When nerve action potentials are recorded across a lesion, it is often best to perform neurolysis alone, as nerve action potentials traveling across a lesion in continuity suggest a recovering nerve [43]. However, in neonates, nerve action potentials are not similarly useful. Intraoperative nerve action potentials in neonates are thought to provide overly optimistic data. One study included ten lesions in continuity and found positive NAPs across the lesion in five patients. Neurolysis alone was performed in these patients and none had a good recovery [33]. In an additional study, Pondaag and colleagues found that the specificity for a severe lesion of absent NAPs and compound muscle action potentials (CMAPs) across a lesion in continuity was high (>90%). However, the sensitivity was very low (<30%) [41]. Taken together, the available data suggest that intraoperative NAPs and CMAPs in neonates are not useful in guiding decisions. Thus, the surgeon is challenged with relying on preoperative assessment to determine who should undergo nerve reconstruction and that is fraught with the challenges previously described.

Thus, once a decision for surgery is made, the real decision is whether to graft or to transfer. There are very little data and very few studies directly comparing nerve grafts to nerve transfers for NBPP. Thus, determining the optimal intervention remains challenging. There are currently disagreements about the role of nerve transfers in the treatment of NBPP. The International Federation of Societies for Surgery of the Hand suggests that the role of nerve transfers in NBPP is unclear but that nerve transfers are a viable option for Erb’s palsy but should not be first-line treatment for more severe injuries. The committee suggests that there should not be an overreliance on nerve transfers and there should remain an inclination toward brachial plexus exploration and nerve graft repair [52]. Further data, however, are needed to determine the optimal roles of both nerve transfer and nerve graft repair.

Erb’s palsy with C5 and C6 injury is the most common pattern of injury in NBPP. While nerve graft repair is the traditional intervention, nerve transfers have been shown to be a viable option. Recovery of elbow flexion has been shown to be good following ulnar or median nerve fascicle transfer to the biceps or brachialis branch of the musculocutaneous nerve. In one study, 87% of patients undergoing these transfers obtained functional elbow flexion recovery. Outcomes were worse for supination recovery with only 21% recovering functional supination [34]. While there was no direct comparison to nerve graft repair, these outcomes suggest nerve transfer is a viable option.

Reinnervation of the suprascapular nerve is important for restoration of external rotation of the shoulder following C5/C6 injury in NBPP. Early experience reinnervating the suprascapular nerve was poor regardless of whether nerve graft repair or nerve transfer was used [35]. More recently, however, outcomes have been better. There have been mixed data comparing spinal accessory nerve transfer with C5 nerve graft repair. Spinal accessory nerve transfer is at least equivalent to C5 nerve graft repair, but some data suggest it may have better outcomes [47, 53]. Seruya and colleagues found that C5 nerve graft repair led to poorer shoulder function and also increased secondary shoulder surgery compared to spinal accessory to suprascapular nerve transfer [47]. The major challenge remains making a decision to graft or to transfer in the setting of a lack of data comparing the two interventions. Future studies will need to focus on comparing outcomes. Additionally, as we discuss in the next section, it will be important to compare outcomes more in depth than simply motor outcome.

A similar dilemma exists in the adult population of brachial plexus injury patients. What is the optimal repair strategy to maximize outcomes? For upper trunk injuries with loss of shoulder abduction, external rotation, and elbow flexion, there is little in the way of direct comparisons between nerve graft repair and nerve transfer. However, two recent meta-analyses help compare the two strategies, and both concluded that nerve transfer strategies are superior to nerve graft repair. These studies utilized the Medical Research Council (MRC) grading scale where M5 is normal strength, M4 is movement against active resistance, M3 is movement against gravity but no active resistance, M2 is movement with gravity eliminated, and M1 is flicker movement or contraction only. Garg and colleagues found that 83% of patients with nerve transfers achieved M4 or greater elbow flexion strength and 96% achieved M3 or greater. Comparatively, only 56% of patients with nerve graft repair achieved M4 or greater strength and 82% achieved M3 or greater. Shoulder outcomes were similarly better with nerve transfers. Seventy-four percent of dual nerve transfer patients achieved M4 or greater shoulder abduction strength versus 46% with nerve graft repair. Both shoulder abduction and external rotation were better in the nerve transfer group [26]. Ali and colleagues recently supported these findings. They found that nerve transfer techniques were superior to nerve graft repair for the restoration of elbow flexion and shoulder abduction. Specifically, with regard to elbow flexion, the Oberlin procedure (transfer of an ulnar fascicle to the biceps branch of the musculocutaneous nerve) was superior to all other strategies [4]. Thus, for upper trunk brachial plexus injuries, nerve transfer seems to be superior to nerve graft repair, but no direct comparative data are available. This data is not conclusive, however, and there certainly remains controversy. In fact, in a systematic review, we previously found that the data did not support the sole use of nerve transfers for upper brachial plexus injury. We recommended at that time that the standard should still include brachial plexus exploration with nerve graft repair when feasible [55]. Additional comparative studies are needed to better elucidate the optimal strategy.

Restoration of hand function following lower trunk injuries is similarly challenging. In addition to nerve graft and nerve transfer techniques, an additional consideration is the Doi procedure (double free muscle transfer) [20]. Ray and colleagues initially described a series of four patients with isolated lower trunk injuries in whom they performed transfer of the nerve to the brachialis to the anterior interosseous nerve, with good clinical outcomes [42]. Isolated lower trunk injuries, however, are relatively uncommon. With concomitant involvement of the upper brachial plexus, nerve transfer options become more limited. Dodakundi and colleagues initially reported success of the double free muscle transfer in total brachial plexus injury [19]. As an adjunctive intervention, wrist arthrodesis has been shown to improve both finger range of motion and overall hand function in patients with double free muscle transfer for pan-plexus injury [2]. Recently, Satbhai and colleagues reported an improvement in overall functional outcome and quality of life using the double free muscle transfer versus single free muscle transfer or nerve transfer for patients with pan-plexus injury [46]. However, it is not clear that hand function was significantly better. In addition, this study pertains to patients with pan-plexus injury and focuses on the overall function of the limb. In cases of isolated lower trunk injury, it is not clear what strategy, whether nerve graft, nerve transfer, free muscle transfer, or tendon transfer, yields the best results. Thus, determining the optimal reconstructive strategy remains challenging.

Postoperatively or, in the case of those neonates who are managed nonoperatively, throughout the natural history of the condition, we are tasked with evaluating these children in some way. This is particularly important in order to collect data to determine if operative intervention is helpful and in order to compare different types of intervention head to head. To this point, most evaluations have focused on motor outcomes and grading individual motor movements on scales such as the Medical Research Council (MRC), Active Movement Scale (AMS), and Louisiana State University motor grading scales. While a variety of outcome measures have been used, the five most common in the published literature include range of motion of the shoulder, range of motion of the elbow, the Mallet scale, MR imaging findings, and the MRC grading scale [45]. Very few evaluation instruments/metrics are specifically validated for use in the NBPP population. Validated evaluation instruments/metrics include the Active Movement Scale, Toronto Scale Score, Mallet Score, Assisting Hand Assessment, and Pediatric Outcomes Data Collection Instrument [16]. While gross motor function and evaluation of body structure and function are important, this may not capture the complete picture, as simply grading motor strength ignores other important factors such as sensation, arm preference, proprioception, functional use of the extremity, cognitive development, pain, quality of life, and language development [22]. Thus, it remains a specific challenge to determine how best to evaluate patients with NBPP. While a number of these domains of evaluation are specifically to the NBPP population, a similar problem exists when evaluating adults with brachial plexus injury following intervention. In this population, it also remains a specific challenge to go beyond purely evaluating motor recovery and rather to also evaluate quality of life, functional use of the affected limb, and pain [22].

One challenge of the postoperative evaluation is determining the optimal duration of time to follow these patients. From age 5 onward, these patients generally have stable to improved hand and shoulder function. However, over the same time course, elbow function tends to slightly deteriorate. This is true whether or not nerve reconstruction was performed. Children who have poor shoulder external rotation benefit from shoulder surgery with significant improvement postoperatively [50]. Because of the continued decrease in elbow function and the significant benefit to shoulder external rotation following surgery for those patients in whom external rotation limitation is recognized, it is important to follow these patients throughout childhood and adolescence and into adulthood.

In the general population, approximately 90% of people have a right arm preference/dominance. In children with left upper extremity brachial plexus palsy, that percentage remains roughly the same, 93% in our previous study. However, when the right upper extremity is the affected limb, only 17% preferred the right limb. This is a significant deviation away from the population average [54]. This suggests neural plasticity is at work early in the development of these children. However, what is not clear is how dominant the unaffected extremity becomes. Is the affected extremity essentially a useless limb, or is there only a slight preference for the unaffected extremity? More importantly, do surgical interventions improve the functional use of the extremity and reduce the preference for the unaffected extremity? Finally, do nerve transfers that offer earlier, though some would argue less complete, recovery offer advantages over nerve graft repair due to the fact that recovery occurs when motor patterns are being established? These are the challenges in evaluation that remain to be answered.

It may not simply be weakness that leads to altered limb preference and reduced functionality. Proprioception plays a large role in the functional use of extremities. However, to this point, little focus has been given to evaluating proprioception following brachial plexus injury. We have previously assessed elbow position sense in adolescents with a history of NBPP. We found that position sense is impaired in the affected limb following NBPP [14]. Similarly, tactile spatial perception is reduced in the hand of the affected limb following NBPP [15]. It is unclear how much this affects daily use of the limb and overall limb preference. However, it may be an important component not assessed by purely focusing on gross motor function. Further assessments of proprioception and advanced sensory modalities are needed in future studies to determine their importance in daily activities and which interventions improve these modalities that contribute to complex functional use.

Delayed or altered use of the affected limb may also affect development in a more global fashion. Motor impairments in children have previously been reported to delay language [31]. The nature of the relationship between motor function and language is unclear. Decreased motor function may impair the ability of the child to explore the world around them, thus delaying language. We have previously shown a high rate of language delay in toddlers with a history of NBPP [17]. This finding has several important implications. First, it suggests that treating children with NBPP is more complex than simply focusing on motor rehab. Recognizing the association of language delay and NBPP means that rehabilitation focused on language development should be part of the overall rehabilitation program. Furthermore, it suggests that assessment of language is an important component of the global assessment of these patients. A further understanding of exactly how language development and motor deficits, and more specifically NBPP, are linked may lead to a better understanding of interventions that may address this issue. For example, if delays in language development result from a decreased ability to explore the surroundings at a very young age, those interventions that favor early recovery, i.e., nerve transfers as opposed to nerve graft repair, may favor improved language development. This remains hypothetical, however, but points to the challenge of needing more complex evaluations to determine optimal interventions.

With language development being affected, one might hypothesize that behavioral issues may arise in children with a history of NBPP. This hypothesis turns out to be correct. Children with a history of NBPP show global developmental delays, difficulty with hand-eye coordination, and a higher incidence of emotional and behavioral problems. This was closely associated with the severity of initial injury [6]. One might assume that earlier or more complete recovery may be associated with a reduction in behavioral problems, but this has never been demonstrated. Thus, it remains a challenge to evaluate behavioral outcomes and to determine what factors are associated with reduced behavioral issues, including which interventions may help reduce these issues.

All of these challenges point to need for more global and comprehensive evaluation of patients with NBPP, both managed operatively and nonoperatively. Ultimately, what is important to these patients is having the highest quality of life possible. A number of factors have been identified as affecting the quality of life in these patients including social impact and peer acceptance, emotional adjustment, aesthetics and body image, functional limitations, finances, pain, and family dynamics [49]. The diversity of these factors points to the fact that assessment necessarily involves more than simply assessing motor function. It remains the challenge of the nerve surgeon taking care of patients with NBPP to develop the optimal assessment metrics and intervals and to compare interventions head to head using optimized global metrics, ultimately moving beyond simply the World Health Organization International Classification of Functioning, Disability, and Health Body Function and Structure domain and moving into evaluations in the Activity and Participation domain (http://​www.​who.​int/​classifications/​icf/​en/​).