Revisiting constraint-induced movement therapy: are we too smitten with the mitten? Is all nonuse "learned"? and other quandaries.

In the past decade, several critical reviews of poststroke rehabilitation (1-4) have indicated that task-oriented interventions can induce more substantial functional improvements than neuromuscular reeducation approaches for which weighting of supportive evidence is sparse. One such approach is called "constraint-induced movement therapy" (CIMT), also called "CI therapy."

A Brief Historical Primer

Constraint-induced movement therapy originates from seminal studies by Taub (5) of monkeys that had undergone deafferentation. Taub demonstrated that animals forced to use the insensate upper extremity through immobilization of the intact limb for short periods could soon learn to use the insensate limb even when the use of both limbs was possible. Training of the monkeys that had undergone deafferentation during the forced-use period was achieved through successive approximations as the basis for shaping intended movement. The animals would be rewarded as they progressively reached toward and subsequently grasped objects. Taub proposed that the animals had undergone "learned nonuse" of the affected limb and, given the appropriate behavioral training, could relearn to use it indefinitely. Although the duration of the effect of deafferentation was never assessed, Taub proposed that this approach be used for patients with hemiparesis and presumed that the diaschisis after stroke led to a learned suppression of movement comparable to the suppression of the spontaneous limb use in monkeys that had undergone cervical deafferentation.

At about the same time at which Taub was undertaking a series of elegant subhuman primate studies (6-8) from which the learned nonuse theory was formulated, Basmajian was initiating studies on electromyographic biofeedback applications to patients with stroke. (9-11) This approach consisted of monitoring individual limb muscles, usually with surface electrodes, and providing patients with visual and auditory cues about muscle activity. Electromyographic responses could be conditioned through the use of threshold detectors, and signal amplification settings could be controlled by clinicians to modify the presentation of auditory or visual cues to patients on the basis of whether responses were being "down-trained" (as in the case of hyperactive muscles) or "up-trained" (weakened antagonists).

These studies led to a series of subsequent investigations (12,13) that indicated that the primary predictor of the independent use of the hemiparetic upper extremity in patients with chronic stroke was the ability to initiate elbow, wrist, and finger extension. (14,15) This capability became the primary inclusion criterion for what was initially described as "forced use." (16,17) Forced use is defined as the process through which a patient is made to use the hemiparetic upper extremity through immobilization of the better limb in a sling or while wearing a mitt during most waking hours. During such time, the patient undertakes activities determined collaboratively with the clinician but performed in the home environment. Over a 2-week time interval, the patient is free to contact the therapist for alterations in tasks or questions regarding compliance. Many studies (18-21) continue to use this forced-use approach for patients after stroke.

Constraint-induced movement therapy includes forced use but also includes one-on-one training for as much as 6 hours per day over several weeks as well as repetitive task practice and adaptive task practice (also called "shaping"). Repetitive task practice refers to continuous efforts to execute movements that usually are repeated, for example, eating, grooming, or brushing teeth. During such efforts, the kinematics of the movements can be varied (made more challenging) on the basis of considerations of a patient's reacquisition of movement control. Therefore, the only interruptions that occur are those used to make the execution of the tasks easier or more difficult.

Adaptive task practice is a form of operant or instrumental conditioning (associating a reward with a correct response as a basis for reinforcing the correct response) characterized by repetitions of a defined movement, such as picking up blocks and moving them toward a pall, in a series of trials. Each trial has a defined duration, and often a participant is asked either to increase the successful numbers of repetitions or to reduce the time to complete the task demands successfully with one effort. During these efforts, the patient is coached or encouraged by the therapist. The patient then is shown a performance record over a number of trials and should be motivated to perform even more optimally on the basis of progressive improvement over trials. With respect to CIMT, this training procedure was developed by Taub and his group at the University of Alabama at Birmingham and contains the intense treatment approach and additional home work assigned along with a mutually agreed-on behavior contract as its signature piece (see Taub and Uswatte (22) for a comprehensive review of the basis for CIMT).

Modified CIMT, as developed by Page and colleagues, (23-25) represents a distributed practice pattern in which the mitt is worn for several hours each day over a 10-week period and this home-based practice is supplemented with outpatient therapy several times each week. It is interesting that 27 years after the original formulation of CIMT, the ability of a patient to initiate finger extension has been validated as a primary predictor of the successful application of CIMT. (26)

Exploring a Model for Studying CIMT

An important consideration for CIMT is the belief that the intervention is behavioral in nature and, in fact, a case has been made that the signature piece of CIMT can be defined by the predominant, if not exclusive, use of adaptive task practice, also called shaping. (27) This statement would imply that other factors, such as the use of principles governing motor learning (or relearning) or motor control, are not essential considerations in the formulation and execution of CIMT. Several superb perspectives (28-31) can easily be interpreted as challenging this implication.

Although the facts that CIMT has been shown to modify brain activity, especially in the affected motor and premotor cortexes, and that interconnections from undamaged hemispheric structures can be engaged, (32) there is a need to explore mechanisms through which CIMT can induce neuroplasticity. Further questions involve the possible presence of neural substrates that impede movement initiation and whether these substrates might be susceptible to modification with CIMT.

The flow diagram depicted in Figure 1 superimposes on the fundamental model of learned nonuse developed by Taub and coworkers (27) (bold type) additional components supplied by Sunderland and Tuke (30) (italic type). These components can influence the reacquisition of limb use through compensatory learning. The perspective of Taub and coworkers is that specific behavioral retraining will reduce basic impairments as more normal function is restored. Under the learned nonuse paradigm (Fig. 1), cortical or subcortical pathology affecting motor output (as well as reduced limb cortical representation; see below) would result in poor function, even if the potential for use existed. Frustration, fatigue, and teaching of compensatory strategies (defined as learning to use the better upper extremity in the interest of time, convenience, and demonstration of ability) inevitably would produce learned nonuse behavior and, consequently, little initiative to use the impaired hand. The additional factors supplied by Sunderland and Tuke are referred to as "compensatory learning." This form of compensation is different from the compensatory use of the better limb. Specifically, compensatory learning includes behavioral factors, such as attention, motivation, and perceived sense of effort, (33) that contribute to a patient's reacquisition of unique motor skills through attention, motivation, effort, and control over motor outflow from preserved or accessible pathways. This new skill capability thus may facilitate restoration of the cortical representation of movement through task practice.

[FIGURE 1 OMITTED]

Therefore, this model would suggest that overcoming learned nonuse (Fig. 1, bold type) and improved compensatory learning (Fig. 1, italic type) both contribute to limb use after CIMT. One could deduce that factors such as attention and sense of effort are emergent behaviors that are manifested during CIMT training.

Although this behavioral linking of concepts is indeed intriguing, the model does call into question a fundamental concern about whether all nonuse is indeed learned. The model falls to account for several factors, including variations in neuronal synaptic behavior (neuromodulation), alterations in neurotransmitter regulation, and the impact of previous behaviors (movement experience) on skill reacquisition. (29) The model also inadequately addresses the notion that structural pathology may actually retard the genesis of movement and, consequently, nonmovement would not be learned but rather would be an acquired misfortune resulting from neurophysiological disarray.

Figure 2 explores this possibility. Note that in contrast to what is shown in Figure 1, although "contracted cortical limb representation" may result in impaired basic motor control, a more plausible explanation is that low spontaneous use is the cause of contracted cortical control. Such a contracted state may perpetuate impaired motor control but most likely is not the initial causative factor. The major modifications to the prevailing model can be best appreciated by orienting to the large shaded arrow and, particularly, to the dashed lines that surround many important structures whose output can profoundly influence descending motor systems. An accumulating body of evidence indicates that descending motor commands can be influenced by other outputs. Several examples help to illustrate this point.

[FIGURE 2 OMITTED]

Case study evidence suggests that the cingulate motor area, lying inferior to the presupplementary motor area, may govern the intention to move the contralateral upper extremity. (34) Artificial activation of this area resulted in repetitive hand movements that lasted throughout the duration of the stimulation, thus raising the intriguing notion that if this area of the cingulate gyrus drives movement initiation, then disruption of its output could impair volitional drive toward limb use. More recently, by combining positron emission tomography and transcranial magnetic stimulation (TMS) approaches, Chouinard and coworkers (35) demonstrated changes in (caudal) cingulate cerebral blood flow when applying TMS to the ipsilesional sensorimotor area after several weeks of CIMT. Although it is not known whether such cerebral blood flow changes are indicative of excitatory or inhibitory processes, they suggest that a change in cingulate motor outflow may represent a strengthening of connections to engage motor outputs, allowing improved function relative to prestroke activity.

Data derived from the temporal resolution of functional magnetic resonance imaging (fMRI) studies in which participants who were able-bodied were instructed either to initiate or to refrain from movement on the basis of the cueing stimulus suggested that engaging the right inferior prefrontal area was associated with inhibition to targets. (36) This observation raises the intriguing notion that exaggerated activation of this output system (disinhibition) could contribute to an inhibition of or a delay in movement activation. Such a delay could impair a patient's efforts at relevant movement, thereby reducing attempts at successful reinforcement while elevating perceived level of effort. Functional magnetic resonance imaging studies after sessions of repetitive task practice for patients with stroke also implicated other areas not previously used during initial rehabilitation, including bilateral superior cerebellar hemispheric areas, the premotor cortex, and the secondary somatosensory cortex. (37) Therefore, the use of these structures has been implicated in task practice paradigms after stroke, suggesting that improvement may be related to seeking appropriate training vehicles for engaging these structures rather than presuming that suppressed use is learned.

Some of the most compelling information regarding control over motor output has been derived from TMS studies that explored the role of the unaffected motor cortex. Substantial information from TMS (38-40) or fMRI (41) studies has shown that when activated, this structure can produce intracortical inhibition (ICI) of the contralateral (affected) motor cortex. Murase and colleagues (42) demonstrated that ICI is greater in patients with chronic stroke and is especially prevalent during attempts to initiate movement with the impaired hand. This observation supports the possibility that ICI plays a substantial role in impaired motor activation in patients with subcortical stroke. Transcranial magnetic stimulation experiments undertaken by Liepert and colleagues (43) showed that patients with pure motor strokes demonstrated a loss of ICI in the affected hemisphere and that patients with subcortical strokes had longer silent periods in the affected cortex, indicative of enhanced inhibition. Patients with pontine or internal capsule lesions had elevated motor thresholds to activation of the corticospinal system, suggesting a compromised ability to recruit corticospinal neurons.

More exacting detail can be extracted from TMS studies by Classen and coworkers, (44) who showed that severe impairment might result from hyperactive cortical inhibitory interneurons rather than direct disruption of descending motor systems. Modulation of behavior in these interneurons could result from exaggerated inhibition (hyperinhibition) if afferent pathways are damaged, thus implicating the importance of sensory inputs in determining the degree to which volitional motor output could be restricted. Moreover, when efforts were made to directly activate the impaired cortex with TMS over the primary motor cortex, response characteristics appeared to be more normal in less impaired patients but showed elevated thresholds in more impaired patients, thus confirming the notion that corticospinal output capabilities might be directly related to the ability to recruit descending systems. (45)

There are other factors contributing to this alternative mode. Success with limb movements certainly can serve to reinforce efforts, thus reducing fatigue and a sense of effort and resulting in enhanced endurance. This possibility is supported by observations from the recently completed Extremity Constraint-Induced Therapy Evaluation (EXCITE) Trial, (46,47) in which participants increased their training time 4-fold over the 2-week CIMT intervention. Although improved endurance does permit intensification of the CIMT intervention, the extent to which improvements lie within the exclusive domain of more concentrated work time, enhanced problem-solving capabilities, or both (see later discussion of uncertainties) requires further research. Participant awareness, concentration, and cooperation constitute some of the behaviors that can be influenced by the nature and location of the lesions and by premorbid behaviors. The extent to which these behaviors are influenced through CIMT rather than through more formal neuropsychological interventions is a topic worthy of intensive study. Psychosocial and cultural factors also can influence patient compliance. Any aspect of CIMT that motivates patients inevitably has the potential to affect limited use of the more impaired upper extremity.

Collectively, these observations suggest that movement initiation may be suppressed through several changes in pathway engagement and communication that may actually precede any learned nonuse phenomenon. However, these possibilities are not without their modifiers, many of which have not been thoroughly explored. The extent to which one lesion or multiple lesions affect connectivity and function in other neural substrates with which they interface is not known. This concern may be important in situations in which patients have sustained several strokes, with resultant multiple loci of injury. Nor does this model account for the extent to which task specificity interfaces with mechanism, especially because one cannot presume that within CIMT, the vast array of task selection implies the use of consistent problem-solving skills or a neuroanatomical substrate. Simply changing the spatial or temporal domains of a task could have profound implications for which pathways are interfaced. Interhemispheric influences on movement initiation or control are profound, but the exact nature of hemispheric interactions, although the subject of intense study, still warrants considerable exploration.

Many questions still remain to be answered. To what extent does training of the better limb actually inhibit output from the impaired limb? Is such an occurrence guided by cortical and subcortical substrates, and is it influenced by the context of the task for which the patient is being trained? Answers to these important questions can profoundly affect training strategies for CIMT or any repetitive task practice procedure.

Factors Influencing CIMT

Uncertainties prevail regarding the effects of other important variables associated with task practice in general and CIMT specifically. For example, claims have been made that key factors, such as sensory deficits, time since stroke, or provision of other therapies, do not affect outcomes from CIMT, (48) whereas other studies have suggested that hand dominance, (49-51) intensity of training, (52) or time since stroke (53) can affect outcomes. Moreover, extending the inclusion criteria to the treatment of patients with severe impairments has provided case history demonstrations of improved function, (54,55) but the clinical relevance or meaningfulness of such improvements is uncertain. Each of these issues can profoundly influence responses to any task training paradigm, and each also can be the topic of an individual treatise worthy of debate. In this context, the topic of dosing is important. Is more necessarily better (56) and, if so, does an intense, 2-week CIMT program yield better results than a distributed practice pattern? (57) Should all training be unimanual to effect functional changes and cortical reorganization, or does bimanual training facilitate such processes if undertaken judiciously? (58)

Although substantial data now exist to support the efficacy of CIMT in the treatment of patients with acute, (59) subacute, (47) and chronic (22) stroke, perhaps no single item associated with forced use or CIMT has received more attention than the mitt or sling restricting the use of the less affected upper extremity. Although this prop does draw attention to the patient and the technique, restraints are not imbued with special properties (60,61); there is no need to be "smitten with the mitten." Restraints do provide an invaluable reminder to engage the more affected upper extremity and to prevent grasp and manipulation of objects with the better hand. Such props confer unique powers on the user only to the extent that, combined with appropriate training, they allow a patient to recognize previously underestimated potential. An obvious concern is whether a patient would resort to using the better arm and hand if not prevented from doing so by wearing a mitt. It is clear that intense training for 2 or 3 weeks propels the patient to use the impaired limb for periods of time exceeding 1 year. (62) To date, there have been no studies tracking the progressive use of the impaired hand after restraint of the less affected hand, nor is there definitive proof that progressing from unimanual to bimanual task training will impede achievable levels of cortical reorganization.

Such concerns can be readily researched and are ripe for systematic exploration. The results obtained with distributed CIMT training (63) indicate that the mitt need not be worn continuously. Over the several weeks of distributed CIMT training, the patient presumably does undertake bimanual tasks when the better limb is not restrained.

Despite all these concerns, there is considerable evidence that CIMT can induce meaningful improvements in patients with stroke. The Table provides a partial list of strengths, limitations, and uncertainties regarding the utility of CIMT.

Relative Strengths of CIMT

Meaningful Changes

There is compelling evidence that CIMT can produce changes superior to those obtained with placebo treatments for patients with chronic stroke. (64) Even for patients who participated in the EXCITE Trial at 3 to 9 months after stroke, (46) evidence emerged that, after CIMT, kinetic behaviors in a key turning task took on force and torque generation qualities that paralleled those seen in able-bodied control participants undertaking the same biomechanical assessment. (65) Despite these outstanding achievements, there is a need to define what constitutes meaningful improvement in light of a progressively more austere reimbursement environment, especially because decisions regarding compensation for this treatment are predicated on degrees of patient independence and potential productivity that are greater than those seen from a patient's perspective. If a reasonable definition of meaningful improvement is the ability to use the impaired limb to manipulate and control the environment, then approximately 5% to 30% of patients with stroke have the potential to achieve this endpoint. This estimate is based on active extension range of motion at the wrist and digits. The relatively wide range, in turn, is determined by the variety of outcome measures used across studies. (66-69)

In the EXCITE Trial, "clinically meaningful" was defined on the basis of the primary outcome measures used: the Wolf Motor Function Test (WMFT) and the Motor Activity Log (MAL). Details regarding the psychometrics and definitions of these measures can be found elsewhere. (46,70-72) Clinically meaningful improvements were defined as a reduction in the numbers of the 15 timed WMFT tasks that could not be completed within 120 seconds (the maximum time limit) after the CIMT intervention compared with the numbers of tasks that could not be completed within 120 seconds before the intervention by participants randomized to treatment and control groups (numbers of tasks not completed: 2.20 to 0.94 and 3.30 to 3.00, respectively). The proportions of 30 MAL tasks for which participants could achieve scores of [greater than or equal to] 3 after the same time intervals, in comparison with the baseline, also were determined. This measure is an indication of independent use of the impaired upper extremity.

Participants who received CIMT showed significantly greater improvements after 2 weeks in the percentages of tasks reaching these scores (18 to 43 [MAL Amount of Movement] and 22 to 44 [MAL Quality of Movement]) than did those who received usual and customary care (control participants) (18 to 25 and 21 to 27, respectively), and this improvement persisted after 12 months. (47) These findings paralleled the overall improvements in change scores but targeted behaviors that most clinicians would believe are meaningful. The fact that recent critical reviews of upper-extremity treatment approaches for stroke rehabilitation have highlighted repetitive task practice and CIMT as evidence-based interventions (3,4) can serve as an impetus for further investigation, not only for the treatment of stroke but also for other neurological disorders, such as cerebral palsy (73-76) and traumatic brain injury. (77,78)

Exploration of Mechanisms of Action

Considerable effort has been directed toward exploring mechanisms implicating neuroplastic changes during CIMT or repetitive task practice with predominantly TMS (79-81) or fMRI. (37,82-84) The results of these studies demonstrated that the use of a paretic limb increases cortical representation for movements directed away from a hand and wrist flexion synergy and that metabolic activity changes can be observed in the primary motor cortex, premotor cortex, supplementary motor area, cerebellum, and other structures linked to motor outflow from either the ipsilesional or the contralesional brain.

Relative Limitations of CIMT

Practicality

One must question the clinical practicality of intense individualized training in a health care environment that continuously seeks to constrict time for treatment. Favorable results from clinical trials may result in greater acceptability of this approach, including reimbursable coding designations. At present, the provision of individualized CIMT is reimbursable in treatment segments usually designated as neuromuscular reeducation or therapeutic exercise. The development and implementation of group training sessions could be a reasonable alternative delivery mode worthy of exploration.

Valid and Relevant Outcome Measures

As is so often the case with neuro-rehabilitation interventions, inconsistency in the use of outcome measures as well as inadequacy in relating them to patient attributes or even to the underlying purpose for upper-extremity treatment has led to inconsistency in quantification or interpretation of the data. Therefore, outcome measures such as the Functional Independence Measurement (85) or the Barthel Index (86) are appropriate for evaluating interventions that highlight teaching upper-extremity compensatory contralateral limb behaviors but are not appropriate for assessing changes at the impairment level. Similarly, impairment-based measures must be matched against the patient attributes for which they were intended. The WMFT, originally designed as the Emory Motor Test, is explicitly geared toward participants who have sustained mild to moderate strokes and whose impairments match inclusion criteria for most patients receiving CIMT. Outcome measures such as the WMFT should include extensive clinometric assessments to maximize reliability and validity. (70,72,87,88)

Costs

The relative costs for providing the signature CIMT approach are high. Modified CIMT, which represents a distributed practice and treatment pattern, and forced use, in which a patient works primarily in the home environment and has far fewer treatments, are less expensive. However, there still is a need to subject this intervention to a cost-effectiveness analysis. To date, this task has not been undertaken formally. However, the costs necessary to produce a clinically meaningful difference or the number of patients who need to be treated to produce a change acknowledged to be of economic or personal benefit should be determined. (89)

Multisite Validation

Amassing evidence to support the value of an intervention can be strengthened through the implementation of a multisite randomized clinical trial. Constraint-induced movement therapy has benefited from the successful execution of the EXCITE Trial. (46,47) That study constituted a form of validation, especially because the differences in outcomes between participating sites were minimal. Attempting similar trials with variations of the signature CIMT approach would strengthen the recognition of this form of training as an important treatment approach for patients with specific upper-extremity movement capabilities.

Effect Sizes and Relevance of Findings

The notoriety gleaned from CIMT must be expressed by specifying the magnitude of the treatment effect and how that improvement is manifested in terms of functional relevance. In this regard, the demonstration of an effect size holds no more value than a significant P calculation if the outcome lacks clear indexes of functional improvement. In this context, CIMT, like any therapeutic intervention, requires that clinicians assess changes with their minds and not with their hearts, as improvements must be sufficiently compelling to ensure payers and families alike that the outcome warrants the investment. Therefore, the expression of the outcome must reach an equilibrium somewhere between statistical significance and clinical significance. The time when the generation of a significant P value was proof positive of the efficacy of a treatment approach has passed. Particularly in clinical trial studies, substantial P values are relegated to one step beyond proof of principle. Funding agencies and insurance carriers now seek evidence, and "evidence" is rapidly becoming code for "minimal clinically important difference." (90,91)

Given these issues, several questions arise. To what extent is individualized CIMT practical? Should other modes of delivery of CIMT beyond the signature treatment approach and modified CIMT be explored? Can or should outcome measures be standardized? What measures will define CIMT cost-effectiveness? How critical should consumers of CIMT information be in their interpretation of the literature? Given the emerging published data for this intervention, can it serve as a model with which clinicians can better discern functional significance from statistical significance?

Uncertainties Associated With CIMT

Persistence of Training Effect

Just how persistent is CIMT treatment? Findings from studies of patients with subacute (47) and chronic (62) stroke have suggested that at follow-up at 1 or 2 years, the magnitude of improvement remains. With the length of hospitalization after stroke becoming shorter and resources for continued therapy being substantially reduced, attention has been drawn to very early CIMT. (59) However, the persistence of some documented improvements is unclear, as is the extent to which intense CIMT training during the acute poststroke interval can be beneficial or detrimental to long-term rehabilitation goals.

Distribution of Training

As with many interventions, the relationship between the dosing of CIMT and outcome remains uncertain. The signature form of CIMT provides for 6 hours of training per day for 2 weeks (10 days), for a total dose of approximately 60 hours, whereas a distributed delivery pattern, such as that advocated by Page and colleagues, (92,93) would involve less than half that amount of training but longer periods of restraint. Therefore, the signature form of CIMT developed by Taub (94) would yield greater daffy intensity as well a greater total dose than the modified form of CIMT advocated by Page and colleagues. However, there is often a presumption that "more is better," but "more" can be defined as concentrated or distributed sessions of training. As correctly ascertained for patients with chronic stroke and receiving CIMT, (52) the notion that greater intensity will yield more favorable outcomes over other time intervals after stroke is still questionable. (63,95) Consequently, how the dosing of CIMT compares with the dosing involved in usual and customary care still requires elaboration. This issue raises the perpetual question that clinicians have pondered for decades: "Are successful outcomes linked to quality or quantity of care?"

Elements of Training

If long-term studies on CIMT fail to demonstrate a correlation between dosing and treatment outcome, then a logical explanation may be ascertained after precise determination of the elements included in the treatment sessions. To what extent is improvement contingent on enhancement of the problem-solving skills of a patient (engaging movement control in both spatial and temporal domains) rather than on repetition of the task trials that serve as the cornerstone of shaping successful behavior (engaging movement exclusively in the temporal domain without regard to movement control or calibration of the space through which it must be demonstrated)? This question raises the issue of the primary modus operandi for CIMT. The signature approach is based predominantly on operant conditioning as a vehicle to overcome learned nonuse. This behavioral paradigm places less emphasis on the continuity of continuous task performance (repetitive task practice, such as grooming or eating) and more emphasis on segmented trials, in which performance time is used to shape subsequent responses. Accordingly, there is a need to better delineate the extent to which each of these treatment components contributes to successful CIMT.

Unimanual Versus Bimanual Training

The previously described concern, then, brings into focus the importance of exploring the extent to which CIMT must be directed toward unimanual training of the hemiparetic limb. To date, there have been no studies demonstrating that such signature training leads to cortical reorganization that is superior to that achieved with bimanual training or a combination of the 2 types of training. Understandably, there is concern that allowing a patient to engage in bimanual training (96-98) will diminish the effectiveness of reacquired use of the impaired limb, whether such retraining be due to overcoming learned nonuse, other mechanisms, or both (see "Exploring a Model for Studying CIMT" section and Fig. 2). Therefore, studies designed to achieve a successful titration of unimanual and bimanual upper-extremity training in the context of CIMT so that a patient does not resort to the exclusive use of the better limb appear to be warranted. The relevance of such a consideration is embedded in acknowledging the multitude of bimanual tasks that define successful manipulation of the environment.

Social and Cultural Factors

Because realization of the importance of health-related quality of life dictates many clinical decisions, there is a heightened sensitivity to the roles of both cultural factors and social influences in therapeutic outcomes. Compliance with and adherence to CIMT instruction as well as honoring of a mutually agreed-on behavior contract can be influenced by cultural factors, (99) unmet needs, (100,101) support between patient and caregiver, (102-104) caregiver stress, (105) caregiver perception of the patient's memory or behavioral changes, (106) and the number of patient comorbidities. (99) The complexity of these factors is supplemented by the effect that premorbid or acquired poststroke depression may have on patient and caregiver compliance with a behavior contract. Given that the relative risk ratio for depression among adults with stroke is 2.18 compared with that for adults without disability, (107) patients with higher levels of depression may make less efficient use of rehabilitation services. (108) In addition, if present, cognitive impairments are associated with poststroke depression, (109) potentially further complicating compliance with a home-based activity program during CIMT.

Adherence to home-based CIMT instruction may be further complicated by the possibilities that both caregiver and patient perceptions of stroke severity, perspectives regarding full recovery, and involvement with the rehabilitation process may vary with time and experience. (99,105,106) As would be the case for many catastrophic injuries, the extent to which patient attitudes about recovery change over time can influence the effectiveness of CIMT. These possibilities warrant further investigation.

Conclusion

Constraint-induced movement therapy continues to hold great promise as a neurorehabilitation approach that can be classified as a functional retraining procedure. This procedure makes use of behavioral principles of operant conditioning supplemented by a judicious use of findings derived from studies of motor learning and motor control and fundamental neurophysiology.

This article has provided a brief historical overview. A model to explain the modus operandi that linked repetitive use to favorable outcomes, including evidence for cortical reorganization, was presented. This model emphasizes the importance of CIMT in assisting patients to overcome learned nonuse of the hemiparetic upper extremity and, in this context, has been associated exclusively with a behavioral approach to functional restoration. However, such an explanation may be an oversimplification of an exceedingly complex series of interactions that cannot be ignored in explaining factors contributing to improvement. Accordingly, an expansion of the model that considers structural and functional relationships was provided. These relationships may be compromised after cortical or subcortical lesions, because controlled motor outputs may be altered by inhibitory processes. Thus, the failure to initiate movement may transcend an inherent learning phenomenon to include structural pathology. Such a perspective calls into question whether activities teaching compensatory behaviors with the less affected upper extremity reinforce a suppression of movement that is exclusively based on behavioral considerations, that amplifies existing inhibitory processes from contralesional or surrounding structures, or both.

The emerging popularity of this treatment approach requires a critical review of its relative strengths and weaknesses as well as definition of the relevant clinical criteria on which meaningful improvement must be based. A perspective on these strengths, weaknesses, and limitations can foster further examination of this exciting avenue of clinical treatment.

The author thanks his EXCITE colleagues, both team members and principal investigators, for a remarkably engaging experience. Their input and frequent dialogue served as an impetus for the formulation of ideas contained in this perspective. Special acknowledgment is given to Steve Cramer, MD, Associate Professor, Department of Neurology, University of California at Irvine; Leonardo G Cohen, MD, Human Cortical Physiology Section, National Institute of Neurological Disorders and Stroke, National Institutes of Health; Patricia C Clark, PhD, RN, FAHA, Associate Professor, Byrdine F Lewis School of Nursing, Georgia State University; and Sarah Blanton, PT, DPT, NCS, Division of Physical Therapy, Department of Rehabilitation Medicine, Emory University, for invaluable discussions and insightful references.

This article is modified from a presentation at the III STEP Symposium on Translating Evidence Into Practice: Linking Movement Science and Intervention; July 15-21, 2005; Salt Lake City, Utah.

This presentation was supported, in part, by National Institutes of Health grant R01-HD37606.

This article was submitted November 28, 2006, and was accepted April 18, 2007.

DOI: 10.2522/ptj.20060355

References

(1) Gresham GE, Duncan PW, Stason WB, et al. Post-Stroke Rehabilitation. Clinical Practice Guideline, No. 16. Rockville, Md: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research; 1995. AHCPR Publication No. 95-0662.

(2) Management of the Post Stroke Arm and Hand: Treatment Recommendations of the 2001 Consensus Panel. Hamilton, Ontario, Canada: McMaster University; 2001.

(3) Brosseau L, Wells GA, Finestone HM, et al. Ottawa Panel Evidence-Based Clinical Practice Guidelines for Post-Stroke Rehabilitation. Topics in Stroke Rehabilitation. 2006;13:1-279.

(4) Duncan PW, Zorowitz R, Bates B, et al. Management of adult stroke rehabilitation care: a clinical practice guideline. Stroke. 2005;36:e100-e143.

(5) Taub E. Somatosensory deafferentation research with monkeys: implications for rehabilitation medicine. In: Ince L, ed. Behavioral Psychology in Rehabilitation Medicine: Clinical Applications. Baltimore, Md: Williams & Wilkins; 1980:371-401.

(6) Taub E, Bacon RC, Berman AJ. Acquisition of a trace-conditioned avoidance response after deafferentation of the responding limb. J Comp Physiol Psychol. 1965;59:275-279.

(7) Taub E, Goldberg IA. Use of sensory recombination and somatosensory deafferentation techniques in the investigation of sensory-motor integration. Perception. 1974;3:393-405.

(8) Taub E, Goldberg IA, Taub P. Deafferentation in monkeys: pointing at a target without visual feedback. Exp Neurol. 1975;46:178-186.

(9) Basmajian JV. Control of individual motor units. Am J Phys Med. 1967;46:480-486.

(10) Basmajian JV. Electromyographic investigation of spasticity and muscle spasm. Physiotherapy. 1976;62:319-323.

(11) Basmajian JV. Biofeedback in medical practice. Can Med Assoc J. 1978;119: 8-10.

(12) Wolf SL, Baker MP, Kelly JL. EMG biofeedback in stroke: effect of patient characteristics. Arch Phys Med Rehabil. 1979;60:96-102.

(13) Wolf SL, Baker MP, Kelly JL. EMG biofeedback in stroke: a 1-year follow-up on the effect of patient characteristics. Arch Phys Med Rehabil. 1980;61: 351-355.

(14) Wolf SL. Electromyographic biofeedback applications to stroke patients: a critical review. Phys Ther. 1983;63:1448-1459.

(15) Wolf SL, Binder-MacLeod SA. Electromyographic biofeedback applications to the hemiplegic patient: changes in upper-extremity neuromuscular and functional status. Phys Ther. 1983;63:1393-1403.

(16) Ostendorf CG, Wolf SL. Effect of forced use of the upper extremity of a hemiplegic patient on changes in function: a single-case design. Phys Ther. 1981; 61:1022-1028.

(17) Wolf SL, Lecraw DE, Barton LA, Jann BB. Forced use of hemiplegic upper extremities to reverse the effect of learned nonuse among chronic stroke and head-injured patients. Exp Neurol. 1989; 104:125-132.

(18) Liepert J, Uhde I, Graf S, et al. Motor cortex plasticity during forced-use therapy in stroke patients: a preliminary study. J Neurol. 2001;248:315-321.

(19) Pierce SR, Gallagher KG, Schaumburg SW, et al. Home forced use in an outpatient rehabilitation program for adults with hemiplegia: a pilot study. Neurorehabil Neural Repair. 2003;17:214-219.

(20) Schabitz WR, Berger C, Kollmar R, et al. Effect of brain-derived neurotrophic factor treatment and forced arm use on functional motor recovery after small cortical ischemia. Stroke. 2004;35: 992-997.

(21) Ploughman M, Corbett D. Can forced-use therapy be clinically applied after stroke? An exploratory randomized controlled trial. Arch Phys Med Rehabil. 2004; 85:1417-1423.

(22) Taub E, Uswatte G. Constraint-induced movement therapy: answers and questions after two decades of research. NeuroRehabilitation. 2006;21:93-95.

(23) Page SJ, Sisto SA, Levine P, et al. Modified constraint induced therapy: a randomized feasibility and efficacy study. J Rehabil Res Dev. 2001;38:583-590.

(24) Page SJ, Sisto SA, Levine P. Modified constraint-induced therapy in chronic stroke. Am J Phys Med Rehabil. 2002;81:870-875.

(25) Page SJ, Sisto S, Johnston MV, Levine P. Modified constraint-induced therapy after subacute stroke: a preliminary study. Neurorehabil Neural Repair. 2002; 16:290-295.

(26) Fritz SL, Light KE, Patterson TS, et al. Active finger extension predicts outcomes after constraint-induced movement therapy for individuals with hemiparesis after stroke. Stroke. 2005;36:1172-1177.

(27) Taub E, Uswatte G, Elbert T. New treatments in neurorehabilitation founded on basic research. Nat Rev Neurosci. 2002; 3:228-236.

(28) Johansson BB. Brain plasticity in health and disease. Keio J Med. 2004;53: 231-246.

(29) Winstein C, Prettyman MG. Constraint-induced therapy for functional recovery after brain injury: unraveling the key ingredients and mechanisms. In: Baudry M, Bi X, Schreiber SS, eds. Synaptic Plasticity: Basic Mechanisms to Clinical Applications. Boca Raton, Fla: Taylor & Francis; 2005:281-228.

(30) Sunderland A, Tuke A. Neuroplasticity, learning and recovery after stroke: a critical evaluation of constraint-induced therapy. Neuropsychol Rehabil. 2005;15: 81-98.

(31) Calautti C, Baron JC. Functional neuroimaging studies of motor recovery after stroke in adults: a review. Stroke. 2003;34:1553-1566.

(32) Luft AR, Hanley DF. Stroke recovery: moving in an EXCITE-ing direction. JAMA. 2006;296:2141-2143.

(33) Gandevia SC. The perception of motor commands or effort during muscular paralysis. Brain. 1982;105:151-159.

(34) Kremer S, Chassagnon S, Hoffmann D, et al. The cingulate hidden hand. J Neurol Neurosurg Psychiatry. 2001;70: 264-265.

(35) Chouinard PA, Leonard G, Paus T. Changes in effective connectivity of the primary motor cortex in stroke patients after rehabilitative therapy. Exp Neurol. 2006;201:375-387.

(36) Konishi S, Nakajima K, Uchida I, et al. Common inhibitory mechanism in human inferior prefrontal cortex revealed by event-related functional MRI. Brain. 1999;122:981-991.

(37) Johansen-Berg H, Dawes H, Guy C, et al. Correlation between motor improvements and altered fMRI activity after rehabilitative therapy. Brain. 2002;125: 2731-2742.

(38) Ward NS, Cohen LG. Mechanisms underlying recovery of motor function after stroke. Arch Neurol. 2004;61: 1844-1848.

(39) Floeter MK, Rothwell JC. Releasing the brakes before pressing the gas pedal. Neurology. 1999;53:664-665.

(40) Schambra HM, Sawaki L, Cohen LG. Modulation of excitability of human motor cortex (M1) by 1 Hz transcranial magnetic stimulation of the contralateral M1. Clin Neurophysiol. 2003;114:130-133.

(41) Allison JD, Meador KJ, Loring DW, et al. Functional MRI cerebral activation and deactivation during finger movement. Neurology. 2000;54:135-142.

(42) Murase N, Duque J, Mazzocchio R, Cohen LG. Influence of interhemispheric interactions on motor function in chronic stroke. Ann Neurol. 2004;55:400-409.

(43) Liepert J, Restemeyer C, Kucinski T, et al. Motor strokes: the lesion location determines motor excitability changes. Stroke. 2005;36:2648-2653.

(44) Classen J, Schnitzler A, Binkofski F, et al. The motor syndrome associated with exaggerated inhibition within the primary motor cortex of patients with hemiparetic stroke. Brain. 1997;120:605-619.

(45) Butefisch CM, Netz J, Wessling M, et al. Remote changes in cortical excitability after stroke. Brain. 2003;126:470-481.

(46) Winstein CJ, Miller JP, Blanton S, et al. Methods for a multisite randomized trial to investigate the effect of constraint-induced movement therapy in improving upper extremity function among adults recovering from a cerebrovascular stroke. Neurorehabil Neural Repair. 2003;17:137-152.

(47) Wolf SL, Winstein CJ, Miller JP, et al. Effect of constraint-induced movement therapy on upper extremity function 3 to 9 months after stroke: the EXCITE randomized clinical trial. JAMA. 2006;296: 2095-2104.

(48) Rijntjes M, Hobbeling V, Hamzei F, et al. Individual factors in constraint-induced movement therapy after stroke. Neurorehabil Neural Repair. 2005;19:238-249.

(49) Harris JE, Eng JJ. Individuals with the dominant hand affected following stroke demonstrate less impairment than those with the nondominant hand affected. Neurorehabil Neural Repair. 2006; 20:380-389.

(50) Lutz K, Koeneke S, Wustenberg T, Jancke L. Asymmetry of cortical activation during maximum and convenient tapping speed. Neurosci Lett. 2005;373:61-66.

(51) McCombe Waller S, Whitall J. Hand dominance and side of stroke affect rehabilitation in chronic stroke. Clin Rehabil. 2005;19:544-551.

(52) Sterr A, Elbert T, Berthold I, et al. Longer versus shorter daily constraint-induced movement therapy of chronic hemiparesis: an exploratory study. Arch Phys Med Rehabil. 2002;83:1374-1377.

(53) Kwakkel G, Kollen B, Twisk J. Impact of time on improvement of outcome after stroke. Stroke. 2006;37:2348-2353.

(54) Bonifer N, Anderson KM. Application of constraint-induced movement therapy for an individual with severe chronic upper-extremity hemiplegia. Phys Ther. 2003;83:384-398.

(55) Bowman MH, Taub E, Uswatte G, et al. A treatment for a chronic stroke patient with a plegic hand combining CI therapy with conventional rehabilitation procedures: case report. NeuroRehabilitation. 2006;21:167-176.

(56) Sterr A. Training-based interventions in motor rehabilitation after stroke: theoretical and clinical considerations. Behav Neurol. 2004;15:55-63.

(57) Dettmers C, Teske U, Hamzei F, et al. Distributed form of constraint-induced movement therapy improves functional outcome and quality of life after stroke. Arch Phys Med Rehabil. 2005;86: 204-209.

(58) Luft AR, McCombe-Waller S, Whitall J, et al. Repetitive bilateral arm training and motor cortex activation in chronic stroke: a randomized controlled trial. JAMA. 2004;292:1853-1861.

(59) Dromerick AW, Edwards DF, Hahn M. Does the application of constraint-induced movement therapy during acute rehabilitation reduce arm impairment after ischemic stroke? Stroke. 2000; 31:2984-2988.

(60) Taub E, Wolf SL. Constraint induction techniques to facilitate upper extremity use in stroke patients. Topics in Stroke Rehabilitation. 1997;4:38-61.

(61) Taub E, Crago JE, Uswatte G. Constraint-induced movement therapy: a new approach to treatment in physical medicine. Rehabil Psychol. 1998;43:152-170.

(62) Taub E, Miller NE, Novack TA, et al. Technique to improve chronic motor deficit after stroke. Arch Phys Med Rehabil. 1993;74:347-354.

(63) Page SJ. Intensity versus task-specificity after stroke: how important is intensity? Am J Phys Med Rehabil. 2003;82: 730-732.

(64) Taub E, Uswatte G, King DK, et al. A placebo-controlled trial of constraint-induced movement therapy for upper extremity after stroke. Stroke. 2006; 37:1045-1049.

(65) Alberts JL, Butler AJ, Wolf SL. The effects of constraint-induced therapy on precision grip: a preliminary study. Neurorehabil Neural Repair. 2004;18:250-258.

(66) Bard G, Hirschberg GG. Recovery of voluntary motion in upper extremity following hemiplegia. Arch Phys Med Rehabil. 1965;46:567-572.

(67) Wade DT, Langton-Hewer R, Wood VA, et al. The hemiplegic arm after stroke: measurement and recovery. J Neurol Neurosurg Psychiatry. 1983;46: 521-524.

(68) Parker VM, Wade DT, Hewer RL. Loss of arm function after stroke: measurement, frequency and recovery. Int Rehabil Med. 1987;8:69-73.

(69) Binkofski F, Seitz RJ, Hacklander T, et al. Recovery of motor functions following hemiparetic stroke: a clinical and magnetic resonance-morphometric study. Cerebrovasc Dis. 2001;11:273-281.

(70) Morris DM, Uswatte G, Crago JE, et al. The reliability of the Wolf Motor Function Test for assessing upper extremity function after stroke. Arch Phys Med Rehabil. 2001;82:750-755.

(71) Wolf SL, Catlin PA, Ellis M, et al. Assessing Wolf Motor Function Test as outcome measure for research in patients after stroke. Stroke. 2001;32:1635-1639.

(72) Wolf SL, Thompson PA, Morris DM, et al. The EXCITE trial: attributes of the Wolf Motor Function Test in patients with subacute stroke. Neurorehabil Neural Repair. 2005;19:194-205.

(73) Charles JR, Wolf SL, Schneider JA, Gordon AM. Efficacy of a child-friendly form of constraint-induced movement therapy in hemiplegic cerebral palsy: a randomized control trial. Dev Med Child Neurol. 2006;48:635-642.

(74) Gordon AM, Charles J, Wolf SL. Methods of constraint-induced movement therapy for children with hemiplegic cerebral palsy: development of a child-friendly intervention for improving upper-extremity function. Arch Phys Med Rehabil. 2005;86:837-844.

(75) Gordon AM, Charles J, Wolf SL. Efficacy of constraint-induced movement therapy on involved upper-extremity use in children with hemiplegic cerebral palsy is not age-dependent. Pediatrics. 2006;117: e363-e373.

(76) Charles J, Gordon AM. A critical review of constraint-induced movement therapy and forced use in children with hemiplegia. Neural Plast. 2005;12:245-261.

(77) Sterr A, Freivogel S, Schmalohr D. Neurobehavioral aspects of recovery: assessment of the learned nonuse phenomenon in hemiparetic adolescents. Arch Phys Med Rehabil. 2002;83:1726-1731.

(78) Shaw SE, Morris DM, Uswatte G, et al. Constraint-induced movement therapy for recovery of upper-limb function following traumatic brain injury. J Rehabil Res Dev. 2005;42:769-778.

(79) Liepert J, Miltner WH, Bauder H, et al. Motor cortex plasticity during constraint-induced movement therapy in stroke patients. Neurosci Lett. 1998;250:5-8.

(80) Liepert J, Bauder H, Wolfgang HR, et al. Treatment-induced cortical reorganization after stroke in humans. Stroke. 2000;31:1210-1216.

(81) Ro T, Noser E, Boake C, et al. Functional reorganization and recovery after constraint-induced movement therapy in subacute stroke: case reports. Neurocase. 2006;12:50-60.

(82) Schaechter JD, Kraft E, Hilliard TS, et al. Motor recovery and cortical reorganization after constraint-induced movement therapy in stroke patients: a preliminary study. Neurorehabil Neural Repair. 2002;16:326-338.

(83) Kim YH, Park JW, Ko MH, et al. Plastic changes of motor network after constraint-induced movement therapy. Yonsei MedJ. 2004;45:241-246.

(84) Dong Y, Dobkin BH, Cen SY, et al. Motor cortex activation during treatment may predict therapeutic gains in paretic hand function after stroke. Stroke. 2006; 37:1552-1555.

(85) Keith RA, Granger CV, Hamilton BB, Sherwin FS. The functional independence measure: a new tool for rehabilitation. Adv Clin Rehabil. 1987;1:6-18.

(86) Granger CV, Dewis LS, Peters NC, et al. Stroke rehabilitation: analysis of repeated Barthel Index measures. Arch Phys Med Rehabil. 1979;60:14-17.

(87) Wolf SL, Catlin PA, Ellis M, et al. Assessing Wolf Motor Function Test as outcome measure for research in patients after stroke. Stroke. 2001;32:1635-1639.

(88) Whitall J, Savin DN Jr, Harris-Love M, Waller SM. Psychometric properties of a modified Wolf Motor Function Test for people with mild and moderate upper-extremity hemiparesis. Arch Phys Med Rehabil. 2006;87:656-660.

(89) Weeks DL, Noteboom JT. Using the number needed to treat in clinical practice. Arch Phys Med Rehabil. 2004;85: 1729-1731.

(90) Wyrwich KW, Nienaber NA, Tierney WM, Wolinsky FD. Linking clinical relevance and statistical significance in evaluating intra-individual changes in health-related quality of life. Med Care. 1999;37:469-478.

(91) Jaeschke R, Singer J, Guyatt GH. Measurement of health status: ascertaining the minimal clinically important difference. Control Clin Trials. 1989;10:407-415.

(92) Page SJ, Sisto S, Levine P, McGrath RE. Efficacy of modified constraint-induced movement therapy in chronic stroke: a single-blinded randomized controlled trial. Arch Phys Med Rehabil. 2004;85: 14-18.

(93) Page SJ, Levine P, Leonard AC. Modified constraint-induced therapy in acute stroke: a randomized controlled pilot study. Neurorehabil Neural Repair. 2005;19:27-32.

(94) Taub E. Harnessing brain plasticity through behavioral techniques to produce new treatments in neurorehabilitation. Am Psychol. 2004;59:692-704.

(95) Sterr A, Szameitat A, Shen S, Freivogel S. Application of the CIT concept in the clinical environment: hurdles, practicalities, and clinical benefits. Cogn Behav Neurol. 2006;19:48-54.

(96) Whitall J, McCombe Waller S, Silver KH, Macko RF. Repetitive bilateral arm training with rhythmic auditory cueing improves motor function in chronic hemiparetic stroke. Stroke. 2000;31: 2390-2395.

(97) Mudie MH, Matyas TA. Can simultaneous bilateral movement involve the undamaged hemisphere in reconstruction of neural networks damaged by stroke? Disabil Rehabil. 2000;22:23-37.

(98) Tijs E, Matyas TA. Bilateral training does not facilitate performance of copying tasks in poststroke hemiplegia. Neurorehabil Neural Repair. 2006;20:473-483.

(99) Nichols-Larsen DS, Clark PC, Zeringue A, et al. Factors influencing stroke survivors' quality of life during subacute recovery. Stroke. 2005;36:1480-1484.

(100) Bakas T, Austin JK, Okonkwo KF, et al. Needs, concerns, strategies, and advice of stroke caregivers the first 6 months after discharge. J Neurosci Nurs. 2002;34:242-251.

(101) King RB, Semik PE. Stroke caregiving: difficult times, resource use, and needs during the first 2 years. J Gerontol Nurs. 2006;32:37-44.

(102) Glass TA, Matchar DB, Belyea M, Feussner JR. Impact of social support on outcome in first stroke. Stroke. 1993; 24:64-70.

(103) Maeshima S, Ueyoshi A, Osawa A, et al. Mobility and muscle strength contralateral to hemiplegia from stroke: benefit from self-training with family support. Am J Phys Med Rehabil. 2003;82: 456-462.

(104) Tsouna-Hadjis E, Vemmos KN, Zakopoulos N, Stamatelopoulos S. First-stroke recovery process: the role of family social support. Arch Phys Med Rehabil. 2000;81:881-887.

(105) Clark PC, Dunbar SB, Shields CG, et al. Influence of stroke survivor characteristics and family conflict surrounding recovery on caregivers' mental and physical health. Nurs Res. 2004;53:406-413.

(106) Clark PC, Dunbar SB, Aycock DM, et al. Caregiver perspectives of memory and behavior changes in stroke survivors. Rehabil Nurs. 2006;31:26-32.

(107) McDermott S, Moran R, Platt T, et al. Depression in adults with disabilities, in primary care. Disabil Rehabil. 2005;27: 117-123.

(108) Gillen R, Tennen H, McKee TE, et al. Depressive symptoms and history of depression predict rehabilitation efficiency in stroke patients. Arch Phys Med Rehabil. 2001;82:1645-1649.

(109) Kauhanen M, Korpelainen JT, Hiltunen P, et al. Poststroke depression correlates with cognitive impairment and neurological deficits. Stroke. 1999;30:1875-1880.

SL Wolf, PT, PhD, FAPTA, FAHA, is Professor, Department of Rehabilitation Medicine and Department of Medicine and Cell Biology, Emory University School of Medicine, 1441 Clifton Rd NE, Atlanta, GA 30322 (USA); and Professor of Health and Elder Care, Nell Hodgson Woodruff School of Nursing, Rehabilitation Research and Development Center, Atlanta VA Medical Center. Address all correspondence to Dr Wolf at: [e-mail].

Table.

Strengths, Limitations, and Uncertainties Regarding the Utility of
Constraint-Induced Movement Therapy

Category        Features

Strengths       Preliminary evidence that this intervention yields
                  meaningful functional gains in patients with chronic
                  stroke and having specific movement criteria (23%-30%
                  of the population)

                Reenergizing of research and clinical approaches
                  targeting repetitive task practice in a functional
                  context

                Fostering of further investigation of
                  constraint-induced movement therapy for patients with
                  acute and subacute stroke

                Promotion of research activities in other diagnostic
                  categories

                Promotion of research into mechanisms (transcranial
                  magnetic stimulation-functional magnetic resonance
                  imaging coregistration)

Limitations     Practicality of individualized training

                Need for valid outcome measures

                Cost-effectiveness

                Validation across clinics and research centers

                Potential misrepresentation (magnitude of effect)

                Mismatch between statistical significance and functional
                  significance

Uncertainties   Persistence of effect

                Distribution of training (intensity and dosing)

                Best training methods (shaping vs repetitive task
                  practice)

                Rigorous comparisons with alternative interventions
                  (eg, bimanual training and functional activity
                  training)

                Cultural factors contributing to adherence

                Behavioral and social influences over time after stroke
                  (ie, does constraint-induced movement therapy for
                  patients with acute subacute, and chronic stroke
                  present different problem sets influencing outcome?)