Abstract
Objective: To evaluate the feasibility and efficacy of constraint-induced movement therapy (CIMT) for impaired upper extremity (UE) function in children with acquired brain injury (ABI).
Design: Multiple case studies.
Setting: Inpatient pediatric rehabilitation.
Participants: Seven consecutive ABI rehabilitation admissions with hemiparesis were recruited without regard to injury etiology, age, or cognitive capacities.
Main Outcome Measure: The actual amount of use test (AAUT) was used to evaluate change in UE function. AAUT amount of use (AOU) and quality of movement (QOM) scales were obtained at baseline and follow-up.
Results: AOU and QOM item improvements were significant, as were changes in activities of daily living. The effect sizes for these changes were large.
Conclusions: Stringent CIMT training, previously only implemented with adults, can be used effectively with children when everyday elements of a child's life are integrated into adult protocols. The use of child-friendly UE shaping exercises, "pushed into" activities by professional therapists as well as trained teachers, paraprofessionals, and parents, was supported. Effects of impairment, injury, and behavior on outcomes are discussed. Larger controlled studies with additional outcome measures are indicated.
CONSTRAINT-INDUCED Movement Therapy (CIMT) has received increasing scientific1-4 and media attention5 as the most promising rehabilitation technique for the affected upper extremity (UE) in stroke6 and traumatic brain injuries7 in adults. CIMT, characterized by short-term, high-intensity physical training accompanied by nearly constant restraint of the unaffected UE, has been demonstrated to provide superior immediate and long-term outcomes in comparison to standard therapies.4
CIMT was derived from a series of studies of "learned nonuse" in monkeys,8 with a single deafferented forelimb. Animals with the naturally occurring chronic nonuse of the deafferented extremity rapidly reacquired functional forelimb use after restriction of the unaffected forelimb. Subsequent work showed that behavioral shaping of the neglected extremity was also effective in reinstating use of the limb in life situations.9 Motor behavior shaping is the continuous reinforcement of each successive approximation of a skill. Typically, in CIMT, everyday tasks are used to train for accuracy, strength, and speed of movement in the affected UE. Initially, a checker might be placed in a child's hand to be inserted in columns of a "Connect Four" game board. Step by step the child would be given explicit feedback, in terms of the knowledge of results, for improved self-grasping, accuracy of release, and speed of repetitions. Massed practice and positive feedback are combined to maximize the learning effects on the performance of motor tasks.
The improved motor performance and increased use of the affected extremity in the free situation of deafferented monkeys had profound implications for extending these techniques to humans with neurologically based UE dysfunction.9 Although standard physical therapies often result in measured improvements in UE skills of hemiplegic individuals, those skills frequently fail to be employed after the individual leaves the therapeutic environment.10,11 CIMT, with both restraint and motor skill shaping, produced long-term changes in both motor skills and everyday use of the affected limb.1-7
Experts in the field of brain injury recovery and rehabilitation have explained the recovery of function by invoking "neural plasticity" as a general concept. Other than functional gains there was little direct evidence of what specific parameters of intervention produced CNS changes. Following the research of Merzernich and others,12-14 literature developed that described, in intact and lesioned animals and humans, the conditions under which cortical representation for perception and action could be changed in fully developed individuals. Those basic science papers suggested that such reorganization occurred only when an extraordinary number of trials occurred in a short period. For example, as a result of intense practice, expert string musicians show greater cortical representation of fingering hand's digits.13 Such changes frequently required some alteration in the "expected" way a stimulation or action took place, for example in the position the targeted item is placed or in the timing of presentation of items.12,14 In CIMT, the altered expectation is accomplished by applying a constraint to the unaffected extremity, thereby forcing the subject to use the affected extremity to accomplish tasks. CIMT clearly includes these attributes of intensity and altered "expectation" (i.e., the constraint). The links in adult stroke patients among constraint training, functional improvement, and brain reorganization have recently been further supported by direct brain-mapping techniques. Training following ischemic infarct resulted in expansion of the cortical representation of the digits to areas adjacent to the infarct, as measured by intracortical microstimulation in adult squirrel monkeys.15 This change correlated with improvements in hand function. Transcranial magnetic stimulation mapping demonstrated an increase in cortical representation of hand musculature in stroke patients following CIMT.1
Successful CIMT trials with adults inevitably evokes the question, will this work with children? Despite approximately 40,000 American children (5-14 years old) with hemiplegia, the CIMT literature is dominated by studies of adults. Children become hemiplegic as a result of stroke in sickle cell disease, perinatal vascular events, head traumas, or other acquired brain injuries (ABIs). Concerns about a minimal cognitive level suitable to training, the tolerance of youngsters for this restrictive and intensive therapy, and the ethics of using any experimental paradigm in youngsters, has resulted in only two published papers concerning children.16,17 Those studies only employed "average"-IQ children with hemiplegic cerebral palsy. More importantly, they did not attempt to replicate the most effective CIMT protocols, because they lacked both shaping techniques and the intensity of training applied in adult populations. Crocker et al.16 attempted to use constraint without explicit shaping. Only one of Crocker's two young subjects tolerated the 3-week treatment, showing a long-term increase in voluntary control and spontaneous use of the affected UE. Charles et al.17 employed constraint for only 6 hours per day and provided 2 hours of therapeutic activities without shaping. Two of three study children demonstrated improved hand function, all had an improved two-point discrimination threshold, but "real world" adaptive skills were not assessed.
In contrast to earlier work in children, we have adhered closely to the CIMT protocol developed by Taub for the research program at the University of Alabama. The preliminary nature of this study of children made the use of experimental controls somewhat premature. Three of seven subjects had reached a plateau after long periods of standard therapies aimed at improving upper extremity function. These seven well-documented cases could demonstrate the likelihood that children have the capacity for persevering and progressing during CIMT. We will use individual case material to show the roles age, cognitive level, time since injury, type of injury, and psychosocial factors play in training adherence and functional improvement.
Beyond the lack of background concerning criteria for applying CIMT in children with TBI or other ABIs, young people present unique challenges to the goal of constant constraint and 6 hours per day of shaping activities. Children attend school, children should play and socialize, and (especially) younger children cannot be expected to perform tedious tasks for long periods. These unique challenges lead to our creation of a child-centered program that:
* "pushed-in" through all of the everyday activities of the child
* trained nontherapist staff (e.g., teachers, recreation staff), friends, and families who were familiar to the children, from their school, nursing unit, and recreational activities to be "lay shapers"
* employed these individuals in their regular settings based upon the practical concerns raised in the CIMT and rehabilitation literature18
* minimized embarrassment caused by wearing a constraining device at all times; mitts were used and were often decorated with stickers, written messages, or self-selected designs
* provided a number of sessions involving parents and children in exposure to the constraining equipment and sample tasks to prepare them (especially cognitively compromised and younger children) for the training.
The main objectives of this investigation were to demonstrate that children could tolerate a standard CIMT protocol and to assess the effectiveness of CIMT in children with ABIs.