See All Evidence Sections
Pediatric-Onset Rehabilitation

Upper Extremity

The spinal cord is an integral aspect of the central nervous system because it is involved in the exchange of information between the body and brain; it carries motor information from the brain to the body and transports sensory information from the body back to the brain. Any injury to the spinal cord can disrupt and prevent the body’s ability to perceive sensation and create a motor response below the level of injury. As the spinal cord provides segmental innervation of the body, disruption or loss of motor function is determined by the level of injury. Because the upper extremities are controlled by nerve fibers that originate in the neck (C5-T1) damage to the cervical spine results in the loss of motor and/or sensory function to the upper extremities, trunk, and lower extremities, resulting in tetraplegia. The overall SCI incidence for children (aged 0-18 years) in the United States is 1.99 cases per 100,000 children (Vitale et al., 2006). US hospitals admit and provide SCI treatment to approximately 1455 children annually (Vitale et al., 2006). Children with SCIs experience paraplegia and complete injuries at higher rates than the adolescent and adult populations (DeVivo & Vogel, 2004; Massagli, 2000). However, it is essential to consider the prevalence of pediatrics with SCIs that experience tetraplegia. Although incidences of pediatric SCI, where upper extremity function is impaired, is not widely published, one study looking at traumatic injury (n=490) found that 6.1% of individuals 0-15 and 8.5% of individuals 16-21 experienced complete tetraplegia, while 33.8% of individuals 0-15 and 25.2% of individuals 16-21 experienced incomplete tetraplegia (Saunders et al., 2015). Therefore, roughly 40% of children with SCI will have some upper extremity dysfunction.

It is not surprising that individuals with tetraplegia find challenges in many aspects of life due to impairments in the function of their upper extremities. Two categories that negatively affect the quality of life of tetraplegics, more so than paraplegics, are physical function, specifically lack of hand function, and independence (Manns & Chad, 2001). The occupation of children is play. Play is a developmental activity that fosters necessary skills including sensory, motor, cognitive, communication and social domains (O’Brien J, 2019). Through use of their hands, children learn about cause and effect by interacting with their environment, and they gain a sense of control. Physical impairments, such as SCI, affect play behavior by limiting movement and access (O’Brien J, 2019). SCI significantly impacts children’s growth and development because an inability to engage in play could lead to decreased mastery over the environment and decreased social participation (Vogel et al., 1997). Through mobility, children explore and act on the environment, which leads to positive effects on emotional, social, and intellectual states (O’Brien J, 2019). The upper extremity function of a tetraplegic child facilitates mobility such as pushing their wheelchair, utilizing a joystick, opening a door, and performing transfers. Therefore, the use of upper extremities is required for increased independence. Engagement in social participation is also influenced by upper extremity function (Kim, 2016). Not only is social participation paramount to a child’s development of social skills, it is also related to emotional well-being and life satisfaction, which puts those with SCIs at risk for negative psychological outcomes (Law et al., 2007). Upper extremity function is positively correlated with increased independence. For example, performance of Activities of Daily Living is closely related to upper extremity function (Kim, 2016). A primary milestone of childhood is increased participation and independence in self-care activities, such as dressing, bathing, and feeding. Even if a child is able to help with Activities of Daily Living, tetraplegics often require increased time, effort, and adaptive equipment, as compared to their typically developing or less involved peers (Manns & Chad, 2001). The inability to gain independence as a child due to decreased upper extremity function could have a detrimental impact on a child’s self-concept and familial relationships. Upper extremity function is crucial to engagement and participation in daily living. This chapter will review the available literature on children with SCI with affected upper extremity function, discuss trends in interventions, and identify the gaps in the research, as compared to adult interventional studies.

Author, Year


Study Design

Sample Size



Outcome Measure


(Mulcahey et al., 1997)




Population: Age: 16.8±0.8 yr; Gender: males=4, females=1; Time since injury: 25.8±16.1 mo; Level of injury: C5=4, C6=1.

Intervention: Implantable Functional Electrical Stimulation (FES) and tendon transfers, lengthenings, and releases of the upper extremity unique to each patient.

Outcome Measures: Pinch and grasp, Grasp and Release Test (GRT), six activities of daily living (ADL): eating with a fork, drinking from a cup, placing a telephone call, writing with an ink pen, storing data on a diskette, and brushing teeth.

Muscle Strength

1.         Three of the four adolescents who underwent the deltoid to triceps transfer gained 4/5 muscle strength in elbow extension which, in all cases, was sufficient to stabilize the elbow and expand the horizontal and vertical work areas.

2.        One subject achieved 215 elbow extension strength.

3.        Three subjects who had brachioradialis transfer gained at least 4+/5 in wrist extension strength; however, they did not have sufficient strength to stabilize the wrist during stimulated finger and thumb flexion so this movement had to be limited to preserve each of the subject’s ability to control their wrists; FPL split tendon transfer provided good positioning of the thumb during lateral pinch without compromising stimulated force of FPL.

4.        In three of the four subjects, intrinsic tenodesis transfer prevented MCP joint hyperextension during stimulated finger extension.

5.        The intrinsic transfer of subject 1 had minimal effect on the intrinsic minus posturing of her hand.

6.        One subject, who underwent a capsulodesis procedure, had poor stimulated finger extension because of MCP flexion deformities.

Grasp and Release Test

7.        FES forces were significantly greater than tenodesis forces for lateral and palmer grasps (p=0.043).

8.        The primary difference in performance was with the heavier objects; of the four heavier objects (can, weight, tape, fork) no subject could manipulate them with tenodesis, but with FES all subjects could manipulate the weight and fork, 3 subjects could manipulate the can, and 1 could move the tape.

Activities of Daily Living

9.        Using the FES hand system, independence scores increased in 25 out of 30 cases as compared to baseline testing (six activities, five subjects).

10.      All baseline activities performed with PA before surgery could be achieved using the FES hand system without attendant assistance.

11.       In 11 out of 12 cases, the FES hand system eliminated the need to don and use AE.

12.      After system training, FES was preferred in 21 out of 30 cases; every subject preferred FES for eating, and except for one, preferred FES for writing.

13.      The one subject who preferred writing with a splint was unable to maintain his wrist in extension against the stimulated force of the lateral pinch.

14.      Satisfaction with the FES hand system came from no longer needing adaptive equipment, citing “I can press harder” (writing, brushing teeth), “it’s easier” (writing, phoning, eating) and “it makes me look more normal” (writing, eating, phoning, storing data).

15.      For the times FES was not preferred, the most frequently cited reason was “it’s too hard” (phoning, storing data).

16.      Although they were more independent as defined by the ADL test scoring, it was easier for several of the adolescents to place a phone call and manipulate a diskette with multiple pieces of adaptive equipment or physical assist.

17.      For the drinking activity, three subjects had difficulty stabilizing their wrists against stimulated flexion while holding the cup and felt more confident that they would not spill the water when using alternative strategies.

(Smith et al., 1996)




*Same study sample from (Mulcahey et al., 1994)

Population: Age: 15.8±2.6 yr; Gender: males=3, females=2; Time since injury: 29.8±33.8 mo (<1 yr=3, >4 yr=2); Level of injury: C5=2, C6=3.

Intervention: Functional neuromuscular stimulation (FNS) neuroprosthesis for the upper limb; site of stimulation included fingers extensors, thumb abductors, thumb extenders, finger flexors, and thumb flexors.

Outcome Measures: Grasp and Release Test (GRT).

FNS versus Tenodesis

1.         With FNS, subjects were able to manipulate each test object in at least 1 test session with the exception of subject 4 who could never complete the tape task.

2.        With a tenodesis, all subjects were able to complete the peg task, 1 subject could not manipulate the block, 2 subjects could never complete the can task and no subject was able to pass the pretest with the weight, fork or tape.

3.        For 23 of the 30 (77%) task comparisons, performance was significantly improved with FNS.

4.        In 14 of the 15 cases involving the heaviest test objects (weight, fork, tape), tasks could only be completed with FNS.

5.        For the lighter test objects (peg, block, can), FNS was more effective in 9 of 15 cases (60%):

·          In 3 cases (2 can, 1 block) FNS was needed to complete the task; 2 of these situations involved subject 1, the only individual who lacked wrist extension.

·          In 4 cases (2 can, 1 block, 1 peg) there was no difference in completions but significantly more trials where there were fewer failures using FNS.

·          In 2 cases (1 block, 1 peg) there were more completions with FNS in a greater number of trials.

·          Of the 6 remaining cases with the lighter objects, there was 1 case (can) in which there were no differences in completions or failures and 5 situations (3 peg, 2 block) in which more completions, but also more failures, occurred with a tenodesis.

6.        Lateral pinch forces ranged from 8.9 N to 22.5 N and palmar grasp forces from 2.1 N to 11. 1 N; tenodesis grasp force was not measurable.

7.        Of 29 testable cases with FNS, completions were consistent across sessions in 8 instances (28 %); 6 of which involved the peg or block.

8.        The number of failures was consistent in 10 instances (34%).

9.        Tenodesis performance was consistent in 3 of the 12 (25%) testable cases for completions and 7 of 12 (58%) instances for failures.

10.      With FNS, 5 of the 21 (24%) inconsistent cases were due to increases in completions in early sessions; in 4 of those cases, the median number of completions plateaued by the second or third session whereas for the last subject, they were only able to complete the tape task in the eighth session, after surgery to facilitate stimulated finger extension.

11.       With tenodesis, 7 of the 9 (78%) inconsistent cases were related to improved performance, all on peg or block tasks; plateaus in performance occurred between the second and fourth session.

12.      With FNS and tenodesis, each case of improved performance in later sessions was significantly better as compared to the initial session (p<0.05).

(Mulcahey et al., 1994)




*Same study sample from (Smith et al., 1996)

Population: Age: 15.8±2.6 yr; Gender: males=3, females=2; Time since injury: 29.8±33.8 mo (<1 yr=3, >4 yr=2); Level of injury: C5=2, C6=3.

Intervention: Functional neuromuscular stimulation (FNS) neuroprosthesis for the upper limb; site of stimulation included fingers extensors, thumb abductors, thumb extenders, finger flexors, and thumb flexors.

Outcome Measures: Common Object Test (COT) involving performance and satisfaction of five activities: eating with a fork, drinking from a cup, writing, applying toothpaste and brushing

teeth; device usage survey (activity patterns in home, work, and school setting) with open-ended questions.

Acquire Phase

1.         Without FNS, two hands (self-assist) were required in almost all activities to acquire the objects.

2.        Two subjects scored physical assist for eating and writing since they required wrist splints specifically for those two activities and were unable to don them without help.

3.        With FNS, independence increased for at least one subject in each activity.

4.        Three subjects were able to use stimulation to acquire toothpaste with one hand (independent) which freed the non-FNS extremity to hold or stabilize the toothbrush; all three could independently acquire the toothbrush.

5.        The remaining two subjects acquired a pen and fork with two hands (self-assist), eliminating the need for attendant care (physical assist).

6.        For the drinking activity, one subject was able to acquire the cup independently.

7.        One subject was unable to grasp the cup with his fingers because of insufficient finger extension (physical assist).


Performance Phase (repetitive activity or performing for extended period of time)

8.        During the hold phase in the majority of the activities without FNS, adaptive equipment or two hands (self-assist) were required to maintain the objects in the hand; for example, to hold a toothbrush and a pen, most subjects used a universal cuff, and two subjects relied on Wanchik splints to hold their pens.

9.        Four subjects used a universal cuff to hold a fork (thereby not requiring stimulation) and one was able to weave his utensil through his tight fingers independently.

10.      Without FNS four subjects required modifications to the handle of the fork (adaptive equipment) to stab food; with FNS, no subject required any modifications to the fork to stab food (independent).

11.       Each subject was able to write and grasp a cup independently and, for each activity, lift and lower the arm without assistance (independent).

12.      Without FES, all subject typically required two hands (self-assist) for squeezing and applying toothpaste, brushing teeth and drinking the first and last sip; with FES two subjects used lateral pinch (independent) to squeeze the toothpaste and four subjects were able to use one hand (independent) to brush both sides of their mouths; one subject used two hands (self-assist) to brush the contra-lateral side.


Release Phase

13.      Without FNS, release of objects in each activity usually required two hands (self-assist) during tenodesis flexion, or the mouth (self-assist) to doff adaptive equipment.

14.      One subject required a physical assistance to remove the wrist splint used specifically in the eating and writing tasks.

15.      With FNS, all subjects scored higher on the independence; for most, stimulated lateral and palmar extension was sufficient to release the objects (independent).

16.      One subject no longer needed to insert a fork in the cuff and was able to release the toothbrush and fork with one hand (independent).

17.      For most subjects their quality of performing activities improved and they preferred using the FES system.

18.      Reasons for not using the system included mood (4/5), no time (3/5), no attendant (3/5), skin irritation (1/5), system complications (1/5) and illness (1/5).

Author, Year


Study Design


(Smith et al., 2001)


Case Report


Population: 10 yr, female, C5 SCI, 10 mo post injury.

Intervention: Freehand System (functional electrical stimulation neuroprosthesis).

Outcome Measures: Manual Muscle Test, stimulated pinch force, Grasp and Release Test, Functional Independence Measure, bone growth, lead unwinding.

(Mulcahey et al., 1999b)


Case Report


Population: Case I: 9 yr, C7 SCI; Case II: 6 yr, C7 SCI; Case III 9 yr, C6 SCI.

Intervention: Surgical transfers of the brachioradialis to the flexor pollicis longus and extensor carpi radialis longus to flexor digitorum profundus for thumb and finger flexion, respectively.

Outcome Measures: Pinch and finger flexion force, Grasp and Release Test, Functional Independence Measure.

(Davis et al., 1997)


Case Report


Population: 17 yr, male, C5 complete SCI.

Intervention: Bilateral tendon transfers and unilateral implementation of the Freehand System (functional electrical stimulation neuroprosthesis).

Outcome Measures: Range of motion, Manual Muscle Test, Pinch force.

(Mulcahey et al., 1995)


Case Report


Population: 11 yr, male, C7 complete SCI.

Intervention: Surgical transfers of the brachioradialis to the flexor pollicis longus and the extensor carpi radialis longus to the flexor digitorum profundus.

Outcome Measures: Pinch force, Jebsen Test of Hand Function for Children, Grasp and Release, Functional Independence Measure, Common Object Test.

(Smith et al., 1992)


Case Report


Population: 8 yr, male, C7-8 Frankel C SCI.

Intervention: Functional neuromuscular stimulation neuroprosthetic hand system.

Outcome Measures: Manual muscle test, muscle strength, Grasp and Release Test.


Pediatric patients with cervical SCI face challenges when performing daily tasks and resuming developmentally appropriate roles due to their lack of upper extremity function (Mulcahey et al., 1994). Although research indicates that family members may find it easier to complete tasks for their children (Mulcahey et al., 1994), encouraging children’s independence is crucial to their growth and development. There is a paucity of research on upper extremity interventions and outcomes for children with tetraplegia. Nearly all articles are low level research, including observational cohorts, limited case reports, and expert reviews, and focus primarily on surgical reconstruction and functional electrical stimulation (FES).

Surgical reconstruction of the upper extremity aims to improve positioning of the upper extremity and augment overall hand function (Mulcahey et al., 1997). Surgical reconstruction occurs through various methods such as transferring tendons or nerves, synchronizing, releasing or lengthening tendons, and fusing adjacent bones to immobilize a joint (Mulcahey MJ, 1997). Different surgical approaches are described in the literature, most notably tendon transfers (Davis et al., 1997; Mulcahey et al., 1999a; Mulcahey et al., 1995; Vova & Davidson, 2020) and more recently nerve transfers (Vova & Davidson, 2020). Tendon transfers involve unipolar transfer of a strong redundant muscle to replace a function lost to SCI (Davis et al., 1997; Mulcahey et al., 1995).  Nerve transfers involve unipolar transfer of a nerve, branch, or fascicle to reinnervate a target muscle (Vova & Davidson, 2020). In theory, nerve transfers could innervate more than one muscle and restore multiple functions, depending on the location of coaptation (Tung & Mackinnon, 2010). Nerve transfers have gained recent favor in peripheral nerve conditions, like brachial plexus injury, but the SCI community has been slower to adopt these practices. The most common upper extremity transfers aim to restore elbow extension, wrist extension, finger flexion, and thumb flexion and opposition (Bryden et al., 2012; Mulcahey et al., 1999a; Vova & Davidson, 2020).

Articles that review surgical reconstruction as a method to increase upper extremity function report various approaches and functional outcomes related to stability, strength, and cosmesis. Tendon transfers from the posterior deltoid to the triceps are found to be successful in stabilizing the elbow (Davis et al., 1997; Mulcahey et al., 1999a), and brachioradialis transfers to wrist extensors successfully stabilize the wrist (Mulcahey et al., 1999a). Surgical reconstruction in the form of tendon transfers also positively influences upper extremity positioning. Split tendon transfers of the Flexor Pollicis Longus muscle are used to promote thumb flexion and opposition, which is required for grasping, and intrinsic transfers are used to reduce intrinsic minus positioning to improve the hand’s usability (Mulcahey et al., 1999a). Following tendon transfers, parents report that their children look more normal (Mulcahey et al., 1999a) which most likely contributes to the child’s self-esteem and acceptance. Tendon transfers positively influence individuals’ range of motion, allowing them greater environmental access by expanding the horizontal and vertical workspace (Davis et al., 1997; Mulcahey et al., 1999a). Tendon transfers also result in improvements in upper extremity strength, even for nonexistent movement before surgery (Davis et al., 1997; Mulcahey et al., 1999a). Tendon transfers were also reported to improve grasp and pinch function, functional independence and mobility, bilateral coordination and unilateral control, and eliminate the need for orthoses (Davis et al., 1997; Mulcahey et al., 1999a; Mulcahey et al., 1995).

Although surgical reconstruction has various benefits, it does have drawbacks. An individual must have voluntary control and sufficient strength of at least 4/5 in two or more muscles that perform a similar function in order to be a candidate for tendon transfer surgery (Davis et al., 1997; Mulcahey et al., 1999a). Following surgery, the upper extremity needs to be immobilized, and surgical management such as the prevention of contractures and edema control need to be employed (Mulcahey et al., 1995). The child also needs to receive tendon transfer education (Mulcahey et al., 1995). Clinicians should keep in mind how contractures or severe spasticity would affect surgical reconstruction outcomes (Mulcahey et al., 1999a). Furthermore, there is not a specific universal measure that is used to evaluate upper extremity function following surgical reconstruction in pediatrics (Mulcahey et al., 1999a).

The other significant area of intervention discussed in the literature is FES. Many articles describe the use of implanted percutaneous electrodes, as part of the now defunct Freehand System. The specific technology notwithstanding, there are lessons to be gleaned on the use of both implanted and surface stimulation.  FES is demonstrated to improve both physical and social aspects of upper extremity function for pediatrics with SCI. When utilizing grasp and release abilities to manipulate objects, FES is found to be significantly more effective than tenodesis, owing to the improvements in palmar and lateral power (Davis et al., 1997; Smith et al., 1996). FES improves grip efficiency and consistency, particularly as the weight of objects increases (Mulcahey et al., 1997; Smith et al., 1996).  FES is useful in increasing children’s ability to engage with the environment, be independent with Activities of Daily Living, reduce reliance on adaptive equipment, and improve overall self-concept and autonomy (Mulcahey et al., 1997; Mulcahey et al., 1994; Smith et al., 1996). Overall independence scores for Activities of Daily Living performance increased in 83% of cases when using the FES hand system compared to tenodesis (Mulcahey et al., 1997). Worth noting, compared to children’s and adolescents’ baseline performance that required physical assistance to complete daily tasks, the need for physical assistance was eliminated with the use of FES (Mulcahey et al., 1997; Mulcahey et al., 1994)

Although there are many advantages, the limitations of FES must be considered. Due to the increased pinch force created by FES, it may be harder for children to maintain wrist extension while pinching (Mulcahey et al., 1997), which reduces the natural tenodesis effects. Because it requires learning and practice, children may feel as though utilizing FES is too hard, and may prefer tenodesis (Mulcahey et al., 1997). Furthermore, children who have had their injuries for over one year felt more comfortable using FES in public places such as school, stores, and restaurants than those with recent SCI (Mulcahey et al., 1994). FES may be most beneficial to pediatrics with C5 level injuries, as they typically experience less independence than those with C6 injuries who are able to use tenodesis to their advantage (Mulcahey et al., 1997). It is also worth noting that, despite all this evidence, FES use in general clinics is still limited and the evidence for children has not progressed since the mid-1990s. While adult literature is available, it should be applied with caution, as parameter selection, torque production, and growth considerations would be different in children.

Both surgical reconstruction and FES identify increased independence as a positive outcome; this reduces the burden on parents and family members and promotes a feeling of accomplishment not only for the child but also for the entire family unit (Mulcahey et al., 1999a; Mulcahey et al., 1994). Increased upper extremity function is positively correlated with emotional well-being and life satisfaction, and interventions such as surgical reconstruction and FES should be considered to promote these positive outcomes that give children with cervical SCI the best opportunity to lead a happy and productive life (Law et al., 2007).

Notably missing from the current body of literature is any interventional studies for children post SCI and predictive studies regarding the likelihood of recovering function based on injury level, type, and early indicators. While these outcomes have been reported in the context of case studies and anecdotal evidence, there has been no systematic examination of interventions improving upper extremity function. And so, we are left to extrapolate from adult studies of robotic devices, gravity compensation training, massed practice of component skills, telerehabilitation, virtual reality, and transcranial direct current stimulation/neuromodulatory inputs. Interventional studies in children are challenging for multiple reasons. First, there is an issue of development on top of recovery and, in some cases, cultivating skills which the child never had, if the injury was sustained in infancy or early childhood. Secondly, there are issues related to measurement. Many adult studies use biomechanic or kinematic evaluation to measure range of motion and tissue extensibility or actigraph data for information on frequency of use. These metrics are more limited in children and are largely dependent on normal value comparisons not available for children. Finally, there is the challenge of recruitment to sufficiently power an interventional study, as the population of children with SCI, is relatively small. For these reasons, the literature on upper extremity function in children with tetraplegia lacks the breadth of interventions and careful measurement of function present in the adult literature.

Related Downloads
    Outcome Measures
      Related Toolkits
        Related Videos