Potential Predictors of Functional Outcomes After Home-Based Constraint-Induced Therapy for Children With Cerebral Palsy

Abstract

OBJECTIVE. Our objective was to identify predictors for treatment outcomes after home-based constraint-induced therapy (CIT) in children with cerebral palsy (CP).

METHOD. Forty-three children (aged 4–12 yr) with CP were treated with individualized CIT at home for 4 wk. Potential predictors were age, sex, affected hand, and upper-extremity motor capacity measured by the Peabody Developmental Motor Scale, 2nd edition (PDMS–2). Outcomes were the Pediatric Motor Activity Log (PMAL) Amount of Hand Use and Quality of Hand Use subscales and the Functional Independence Measure for Children (WeeFIM).

RESULTS. A higher PDMS–2 Visual–Motor Integration subscale score predicted a better WeeFIM score after home-based CIT (adjusted R ² = .35). Younger age predicted better performance on the PMAL Amount of Hand Use and Quality of Hand Use subscales (adjusted R ² = .06–.08) after home-based CIT.

CONCLUSION. The potential predictors may allow occupational therapy practitioners to target those children who will benefit most after home-based constraint-induced therapy.

Keywords

cerebral palsy forecasting home care services restraint,physical treatment outcome

Cerebral palsy (CP) is the most common childhood disability that usually requires services from occupational therapy practitioners to promote the child’s performance of activities of daily living (ADLs) and participation. Children with unilateral motor involvement or hemiparesis often neglect to use their more affected upper limb, a phenomenon termed developmental disregard (Deluca, Echols, Law, & Ramey, 2006). Constraint-induced therapy (CIT) is effective in improving developmental disregard, functional performance of the more affected arm (Deluca et al., 2006; Hoare, Imms, Carey, & Wasiak, 2007; Huang, Fetters, Hale, & McBride, 2009), ADLs, and quality of life (Hsin et al., 2012) for children with CP. CIT involves restraint of the less affected upper extremity, forced use of the more affected upper extremity, and massed task-related training of the affected upper extremity (Deluca et al., 2006).

Empirical evidence suggests that the generalization of clinic-based CIT effects to the context of daily living is limited; home-based constraint-induced therapy has been shown to provide more opportunity and experience for children with CP to learn the use of the more affected limb in a natural and familiar environment (Al-Oraibi & Eliasson, 2011; Chen et al., 2013; Hsin et al., 2012; Lin et al., 2011; Rostami & Malamiri, 2012). The improved performance after CIT in a home group compared with a clinic group suggests a natural context as the preferred method for treatment of children with CP (Rostami & Malamiri, 2012).

In several studies, home-based CIT protocols were administered by therapists with relatively moderate intensity and shortened constraint time (Chen et al., 2013; Hsin et al., 2012; Lin et al., 2011). Compared with the control group, the group that received home-based CIT exhibited significantly better motor control (Chen et al., 2013) and performance in grasping control and unilateral and bilateral upper-extremity motor efficacy immediately and at 3 and 6 mo after the treatment (Chen et al., 2013; Hsin et al., 2012; Lin et al., 2011). The moderate intensity and shortened constraint time might enhance the effectiveness and compliance of clients and caregivers.

The identification of potential predictors for CIT outcome helps occupational therapy practitioners target individuals who will benefit the most from the specific therapy. Very limited studies have investigated the possible predictors of clinic-based CIT in children with CP (Charles, Wolf, Schneider, & Gordon, 2006; DeLuca, Case-Smith, Stevenson, & Ramey, 2012; Gordon, 2011; Gordon, Charles, & Wolf, 2006; Rostami & Malamiri, 2012; Sakzewski, Ziviani, & Boyd, 2011), and no study, to our best knowledge, has looked at the predictors of home-based CIT outcome.

Sakzewski et al. (2011) showed that older age was a significant predictor of favorable CIT outcomes with respect to perceived functional performance, whereas Gordon et al. (2006) suggested that age was not associated with CIT outcome on use of the involved upper extremity. Left hemiplegia and severity of hand impairment were suggested predictors of affected limb frequency of use and quality of movement after treatment with CIT (Charles et al., 2006). Charles et al. (2006) also reported that less severity of hand impairment was associated with favorable CIT outcomes; in contrast, Sakzewski et al. (2011) revealed the inverse relationship. The predictors for clinic-based CIT outcomes are not conclusive, and the predictors for clinic-based CIT might not be appropriate for the success of home-based CIT. Further investigation of potential predictors in the context of home-based CIT is warranted.

In this study we aimed to identify predictors influencing outcomes of functional ability of the more affected hand and ADLs in children with CP after home-based CIT treatment. The potential predictors selected in this study were age, sex, affected hand, and upper-extremity motor capacity. Upper-extremity motor capacity was evaluated using the Peabody Developmental Motor Scale—Second Edition (PDMS–2; Folio & Fewell, 2000), Visual–Motor Integration (PDMS–VMI) and Grasping (PDMS–G) subscales. The Pediatric Motor Activity Log (PMAL; Wallen, Bundy, Pont, & Ziviani, 2009) and the Functional Independence Measure for Children (WeeFIM; Ottenbacher et al., 1996) were used to assess functional performance after home-based CIT. We hypothesized that the four predictors would be associated with functional performance after home-based CIT.

Method

Research Design

This study used a pretest–posttest design. Children with CP received 4 wk of CIT at home and underwent outcome assessments before and immediately after the intervention. The institutional review board for human studies from the participating sites approved the study protocol, and caregivers of all participants gave written informed consent.

Participants

The participants in this study were a sample of convenience. All participants received independent examinations by a physiatrist, occupational therapist, or physical therapist to determine their eligibility for inclusion. The eligibility criteria were (1) diagnosed with congenital unilateral spastic CP, (2) considerable nonuse of the more affected upper extremity (amount-of-use score on the PMAL <2.5), (3) active extension movement of the wrist and metaphalangeal joint ≥10°, and (4) no excessive muscle tone (Modified Ashworth Scale; Bohannon & Smith, 1987; ≤2 for any muscle group on the upper extremity) before starting treatment. Exclusion criteria were a severe cognitive, visual, or auditory disorder; a severe concurrent illness or disease not typically associated with CP; active medical conditions such as pneumonia; any major surgery or nerve blockage (such as botulinum toxin or phenol injection) within 6 mo before intervention; and poor cooperation during assessments.

Instruments

Outcome Measures.

Outcome measures included the PMAL and WeeFIM, which are subjective (self-reported) measurements of the functional performance of participants. Parents completed the PMAL—a higher score indicates better performance—to rate their perceptions of how often and how well their child used the more affected upper extremity. The How Often and How Well subscales of the PMAL, documenting Amount of Hand Use and Quality of Hand Use, were included. Both scales have high test–retest reliability (.93–.94; Wallen et al., 2009), fair concurrent validity, and good response to change after CIT treatment (Lin et al., 2012).

The WeeFIM has 18 items for three functional subscales assessing self-care, mobility, and cognition. An ordinal rating system, ranging from 7 (complete independence) to 1 (total assistance), is used to rate performance. The WeeFIM has evidence of excellent test–retest reliability for subscales (.94–.99) and total items (.98) for children with disabilities.

Potential Predictors.

Five potential predictors, including age, sex, affected hand, and upper-extremity motor capacity, were selected on the basis of the empirical findings and theoretical grounds. Upper-extremity motor capacity was measured by the PDMS–2 Fine Motor domain (Grasping and Visual–Motor Integration). The Fine Motor domain of the PDMS–2 has evidence of good test–retest and interrater reliability (.84–.99; van Hartingsveldt, Cup, & Oostendorp, 2005). Construct validity of the PDMS–2 was established by showing its ability to reflect developing motor abilities and to differentiate between children known to be average and those expected to be low average or below average in motor abilities (Folio & Fewell, 2000). The Fine Motor domain, including the Grasping and Visual–Motor Integration subscales, assesses grasping skills and the ability to execute tasks requiring visual–perception–cognition and visual–motor coordination. To assess the motor change of the more affected hand, participants performed unimanual items such as drawing and cutting tasks with the more affected hand.

Interventions

Three certified therapists, including one physical therapist and two occupational therapists, were trained to administer the home-based CIT protocol by the investigator and completed a written competency test before treatment began. Each participant received therapy at home from the same therapist throughout the 4-wk intervention period. The home-based CIT protocol involved practice of functional tasks with the more affected upper extremity for 3.5–4 hr daily, twice per week, and restraint of the less affected upper extremity for 3.5–4 hr daily for 4 wk.

This protocol of home-based CIT has been shown to be feasible for implementation at home (Chen et al., 2013; Hsin et al., 2012; Lin et al., 2011) and has been proved to improve motor control (Chen et al., 2013), motor and functional outcomes (Chen et al., 2013; Hsin et al., 2012; Lin et al., 2011), and quality of life (Hsin et al., 2012) in children with CP.

The functional training of the more affected upper extremity is focused on massed practice of functional activities using the more affected arm, with the principles of shaping and repetitive task practice applied during training. Through shaping training, a motor or behavioral objective is approached in small steps by successive approximations. Repetitive task practice involves functional tasks that are performed continuously during a specific period of time (Wolf et al., 2006).

Between training sessions, children with CP were encouraged to exercise or perform daily activities (e.g., reaching, grasping, manipulating, and self-care activities) with their affected upper extremity at home under parental supervision. In addition, participants were required to wear an elastic bandage and restraint glove on the less affected upper extremity that limited their wrist and individual finger movement for 3.5–4 hr daily for 4 wk (including therapy sessions). Parents were asked to document in daily logs the number of restraint hours outside therapy. No adverse effects or harmful events related to either intervention occurred during the study period.

Data Collection

Tests were administered before and immediately after the 4-wk intervention by two occupational therapists who did not provide the therapy in this study. The therapists were trained to properly administer the PDMS–2. This training included careful review of written instructions and repeated practice. Rater competence was assessed by a senior certified occupational therapist. The interrater reliability for the subscales of the PDMS–2 was high (intraclass correlation coefficients >.9). The PMAL and WeeFIM were completed by parents.

Data Reduction

The relative change in the child’s pretreatment and posttreatment assessments served as the dependent variables. To consider the effect of different starting points of pretreatment scores for each child, the relative change score was used rather than simply a raw change score (Marx & Cummings, 2007). Performance improvement after home-based CIT was indexed using the following formula: [improved score (posttest − pretest)/max possible improvement score (total scores − pretest scores)] × 100%. If the child’s performance deteriorated after home-based CIT, the following formula was applied: [deteriorated score (posttest − pretest)/pretest scores] × 100% (Marx & Cummings, 2007).

Statistical Analysis

Skewness (±1) and visual inspection of histograms were used to verify the normality of the data, including all outcome measures and predictors. The scores for only 1 outcome measure (WeeFIM) at pretest and posttest required transformation using the natural log [ln]. Two steps were used to identify predictors of home-based CIT outcome. First, we examined the associations between the predictors at the pretreatment assessment and the relative change scores on the three outcome measures using the Pearson correlation coefficient (r). To avoid the exclusion of important factors in the model development, the criterion for the predictors included in the regression analysis was set to a value of p of up to .25 (Hosmer & Lemeshow, 1989).

In the second step, the included predictors were used in a backward stepwise procedure to generate a linear regression model for each outcome measure at posttreatment. Adjusted R ², p values, and regression coefficients (β) were examined to assess the goodness of fit of the fitted regression models. We also performed regression diagnostics to examine the presence of multicollinearity, indicated by the variance inflation factor (VIF), among predictors in the models.

Results

We recruited 45 children with CP from four medical centers in Taiwan. Two children failed to adhere to the requirement of restraint of the less affected hand and withdrew from the study. Consequently, 43 children (22 boys, 21 girls) with a mean age of 7.7 yr (range, 4–12 yr) completed the 4-wk therapy and posttreatment assessment (Table 1). In Table 2 we report the mean score of the relative change in each outcome measure and the grouping of participants according to status of change (improvement, no change, or deterioration). Most participants improved after home-based CIT, especially on the WeeFIM. A very small number of participants (about 2–3 children) deteriorated after therapy.

Table 1.

Demographic and Clinical Characteristics of the Participants (N = 43)

Characteristic	Value
Age, yr	7.7 ± 2.2
Gender
Male	22 (48.8%)
Female	21 (51.2%)
Affected hand
Right	19 (44.2%)
Left	24 (55.8%)
Upper-extremity function
PDMS–2
Grasp	42.6 ± 5.1
Visual–Motor Integration	128.3 ± 8.9
PMAL
AOU	1.9 ± 0.4
QOU	2.1 ± 0.4
WeeFIM	92.5 ± 19.5

Note. Values are expressed as mean ± standard deviation for continuous variables and number (%) for categorical variables. AOU = Amount of Hand Use; PDMS–2 = Peabody Developmental Motor Scale—Second Edition; PMAL = Pediatric Motor Activity Log; QOU = Quality of Hand Use; WeeFIM = Functional Independence Measure for Children.

Table 2.

Mean of the Relative Change Scores and the Grouping of Participants According to Status of Change

		Participants’ Status of Change After Therapy
Outcome Measure	Score of Relative Change, M (SD)	Improved (n)	No Change (n)	Deteriorated (n)
PMAL
AOU	21.8 (19.4)	40	0	3
QOU	16.9 (16.6)	38	2	3
WeeFIM	24.0 (20.4)	41	0	2

Note. AOU = Amount of Hand Use; M = mean; PMAL = Pediatric Motor Activity Log; QOU = Quality of Hand Use; SD = standard deviation; WeeFIM = Functional Independence Measure for Children.

In Table 3 we list Pearson correlation coefficients between the five predictors and the relative change scores on the three outcome measures at posttreatment. Two predictors—age and the PDMS–VMI subscale—were entered into the PMAL–Amount of Hand Use and PMAL–Quality of Hand Use models; one predictor, the PDMS–VMI subscale, was entered into the [ln] WeeFIM model.

Table 3.

Relationships Between the Five Predictors and the Relative Change Scores of the Four Outcome Measures

	Pearson’s r
Candidate Predictor	PMAL–AOU	PMAL–QOU+	[ln]WeeFIM
Age	−.30*	−.30*	−.05
Gender	−.16	.02	.12
Affected hand	.01	.10	.09
PDMS–2
PDMS–G	.03	.08	.10
PDMS–VMI	.19*	.28*	.61*

Note. AOU = Amount of Hand Use; [ln] = natural log; PDMS–2 = Peabody Developmental Motor Scale—Second Edition; PDMS–G = Grasping subscale; PDMS–VMI = Visual–Motor Integration subscale; PMAL = Pediatric Motor Activity Log; PMAL–QOU+ = removed 2 outliers; QOU = Quality of Hand Use; WeeFIM = Functional Independence Measure for Children.

p < .25.

The values of all variables (VIF ≤ 1.02) were well within normal ranges (VIF ≤ 10), indicating that the assumptions regarding multicollinearity were not violated. The results on the normal probability plot and the histogram of the three final regression models revealed the residuals were normally distributed. The presence of outliers was examined using jackknife residual analyses, and three outliers were found (residuals < mean ± 3 SD).

In Table 4 we present the results of backward stepwise multiple regression analyses after the outliers were removed. Age was the significant predictor for the PMAL–Amount of Hand Use and PMAL–Quality of Hand Use models, which accounted for 6%–8% of the variance in change scores. The PDMS–VMI subscale score was the significant predictor for the [ln] WeeFIM model, and the model explained 35% of the variance in change scores. The three final regression equations are as follows:

Table 4.

Backward Stepwise Multiple Regression Analyses of the Predictors for the Outcome Measures

	PMAL–AOU		PMAL–QOU		[ln]WeeFIM
Adjusted R ²	.06		.08		.35
F (significance)	3.58 (.07)		4.47 (.04)		22.70 (<.01)
Predictors	β	p	β	p	β	p
Constant	.40	<.01	.27	<.01	−9.24	<.01
Age, yr	−.02	.07	−.02	.04
PDMS–VMI					.06	<.01

Discussion

This pioneer study demonstrates demographic and clinical characteristics predicting outcomes in functional performance in children with CP immediately after home-based CIT. Five demographic and clinical characteristics were used to predict functional outcomes evaluated by the PMAL and WeeFIM. The best predictor for WeeFIM outcome changes after home-based CIT was the PDMS–VMI subscale scores. Age also could predict changes in the amount and quality of the use of the affected hand for daily activities after home-based CIT.

These findings may support conclusions that young children or children with high capacity in visual–motor integration skills before CIT can achieve greater functional outcomes than can older children or those with low capacity. The baseline score of the PDMS–VMI subscale, which can be easily and quickly examined, might be used to stratify patients for participation in home-based CIT at this dose. This knowledge could allow occupational therapy practitioners to identify the inclusion criteria for treatment success when implementing home-based CIT. For example, high visual–motor skill or younger age may lead to great improvements in the amount and quality of the affected hand use for daily activities after home-based CIT.

Visual–motor integration skills (measured by the PDMS–VMI subscale) rather than grasping skills (measured by the PDMS–G subscale) could predict changes in the WeeFIM after home-based CIT, because the PDMS–VMI subscale involves visual–perception–cognition and visual–motor coordination tasks such as tracking, drawing, cutting, and tool use. The PDMS–G subscale involves grasping, releasing, and manipulating skills. The WeeFIM measures daily activities in various domains, including self-care, mobility, and cognition domains. The visual–motor integration skills are more basic and important for performing ADLs than the grasping skills.

Children with a high level of visual–motor integration respond better to home-based CIT in daily activities than those with a low level. Children who had a high PDMS–VMI subscale score preserved a high level of visual–motor control for visual–perception, visual–cognition, and visual–motor integration and coordination before the intervention. Children with a high level of visual–motor integration might be able to complete daily tasks more efficiently and experience a larger variety of activities and problem-solving opportunities to achieve better functional performance during an intervention than those with a lower level. These results were compatible with the study by Charles et al. (2006) but differed from other studies (Gordon et al., 2006; Sakzewski et al., 2011) in which children who initially used their involved hand less effectively during bimanual tasks improved more after intervention than children who initially used it more effectively.

The contradictory findings might come from the nature of movement predictor and outcome measures. We used the predictor reflecting visual–motor efficacy primarily during unimanual tasks and outcomes representing functional performance, whereas previous studies (Gordon et al., 2006; Sakzewski et al., 2011) used predictors reflecting movement performance during bimanual tasks and outcomes relevant to movement performance. Accordingly, the ideal candidates for obtaining subjective improvements in ADLs after home-based CIT might be those children who preserve high visual–motor skill.

Age significantly predicted changes in the amount and quality of the affected hand use for daily activities, with 6% of variance. Our findings suggest that among children with CP aged 4–12 yr, younger children might demonstrate a greater amount and higher quality of affected hand use after home-based CIT than older children. A possible reason may be that home-based CIT forces younger children to adapt earlier to use the more affected hand correctly for exploring the natural environment than older children. Home-based CIT at an early age may also help reduce or prevent developmental disregard and promote upper-extremity skills more effectively than at a later age (Charles & Gordon, 2005; Willis, Morello, Davie, Rice, & Bennett, 2002).

Our results were not compatible with previous studies (Eliasson, Krumlinde-Sundholm, Shaw, & Wang, 2005; Gordon et al., 2006; Sakzewski et al., 2011). Eliasson et al. (2005) and Sakzewski et al. (2011) showed that older age was a significant predictor of favorable individualized outcome. The differences among studies may be the result of differences in participant age, treatment setting (home based vs. clinic based), treatment dosage, and treatment approaches (individualized vs. group or camp therapy). For example, the Eliasson et al. (2005) study used a restricted range of ages (1.5–4 yr) for the participants. The age range of 4–8 yr that is critical for rapid motor functional development of the upper limb (Hart, 2005; Holmefur, Krumlinde-Sundholm, Bergström, & Eliasson, 2010) was not addressed in the Eliasson et al. study. However, the relatively small variance for the regression model did not allow us to draw a definite conclusion of age effects of home-based CIT on functional outcome. Further studies with a larger sample size are needed to validate the age effects of home-based CIT on functional outcome. In addition, other potential predictors, such as participant’s motivation and family participation, may be included in future studies to increase the predicted variance for home-based CIT.

This study used several potential predictors to explore the factors related to home-based CIT changes in functional outcomes, whereas most previous studies (Charles et al., 2006; DeLuca et al., 2012; Gordon, 2011; Gordon et al., 2006; Rostami & Malamiri, 2012; Sakzewski et al., 2011) investigated the association of few predictors and final outcomes after clinic-based CIT. The combination of different relevant factors, as was done in this study, may consider the possible interaction or effect among those factors. Our findings might suggest that the potential predictors varied depending on the forms of CIT used (e.g., clinic based and home based).

Age and, especially, visual–motor integration skill are the best predictors for changes in functional outcomes after home-based CIT. The child’s sex and affected hand were not predictive of any outcome measures after home-based CIT. The result may be explained by the individualization and home-based therapy provided in this study. Although boys and girls might prefer different activities, the therapist could slightly adjust activities during the functional training of the more affected hand according to the child’s individual preferences and home environment to increase compliance. Similarly, for children with either side of affected hand, the therapist could also facilitate the use of the more affected hand by adjusting therapeutic activities or modifying the home environment. In this study, both boys and girls and children with either side of hemiplegia might have responded equally to individualized CIT implemented at home. The predictive value of sex was not investigated in previous studies of children with CP. In a previous study, however, Charles et al. (2006) suggested that left-sided hemiplegia and the severity of hand impairment were associated with clinic-based CIT outcome. The difference may result from different treatment settings (clinic vs. home) in the two studies.

The clinician should select children for treatment on the basis of the best predictor identified from different forms of CIT. For example, younger children and those with high visual–motor skills may respond better to home-based CIT. Children with left hemiplegia and less severe hand impairment may respond better to clinic-based CIT. To identify the best predictors for different forms of CIT, future studies may include multidimensional potential predictors, such as treatment dosages (DeLuca et al., 2012; Gordon, 2011), treatment settings, constraint time, participant motivation, and family participation. In addition, identifying the factors that may be predictive of clinically meaningful change after home-based CIT is important for selecting children who could make clinically meaningful improvement in upper-extremity function relevant for daily life situations.

This study should be interpreted in light of several limitations, including the sample size, participant characteristics, and limited predictors. We used a small sample and mild to moderate impairment of hand function as an inclusion criterion. Future researchers may recruit a larger sample size and children with severely impaired hand function to investigate the possible predictors of home-based CIT outcomes.

Implications for Occupational Therapy Practice

The results of this study have the following implications for occupational therapy practice:

Potential predictors identified in this study allow occupational therapy practitioners to target the children who will benefit most from home-based CIT.

Better visual–motor integration skills predicted immediate improvement in ADLs after home-based CIT in children with CP.

Younger age predicted immediate improvements in the amount and quality of use of the affected hand after home-based CIT.

Footnotes

Acknowledgments

This research was supported in part by the National Health Research Institutes (grant nos. NHRI-EX102-9920PI and NHRI-EX102-10010PI), the National Science Council (grant nos. NSC 95-2314-B-182-045, NSC 98-2314-B-182-006-MY3, NSC 99-2314-B-182-014-MY3, and NSC 100-2314-B-002 -008 -MY3), and Chang Gung University (grant no. CMRPG 391671-3). We thank the children and parents who participated in this study for their time and effort.

References

Al-Oraibi

, & Eliasson

A. C.

(2011). Implementation of constraint-induced movement therapy for young children with unilateral cerebral palsy in Jordan: A home-based model. Disability and Rehabilitation, 33, 2006–2012. http://dx.doi.org/10.3109/09638288.2011.555594

Bohannon

R. W.

, & Smith

M. B.

(1987). Interrater reliability of a modified Ashworth scale of muscle spasticity. Physical Therapy, 67, 206–207.

Charles

, & Gordon

A. M.

(2005). A critical review of constraint-induced movement therapy and forced use in children with hemiplegia. Neural Plasticity, 12, 245–261. http://dx.doi.org/10.1155/NP.2005.245

Charles

J. R.

Wolf

S. L.

Schneider

J. A.

, & Gordon

A. M.

(2006). Efficacy of a child-friendly form of constraint-induced movement therapy in hemiplegic cerebral palsy: A randomized control trial. Developmental Medicine and Child Neurology, 48, 635–642. http://dx.doi.org/10.1017/S0012162206001356

Chen

C. L.

Kang

L. J.

Hong

W. H.

Chen

F. C.

Chen

H. C.

, & Wu

C. Y.

(2013). Effect of therapist-based constraint-induced therapy at home on motor control, motor performance and daily function in children with cerebral palsy: A randomized controlled study. Clinical Rehabilitation, 27, 236–245. http://dx.doi.org/10.1177/0269215512455652

DeLuca

S. C.

Case-Smith

Stevenson

, & Ramey

S. L.

(2012). Constraint-induced movement therapy (CIMT) for young children with cerebral palsy: Effects of therapeutic dosage. Journal of Pediatric Rehabilitation Medicine, 5, 133–142.

Deluca

S. C.

Echols

Law

C. R.

, & Ramey

S. L.

(2006). Intensive pediatric constraint-induced therapy for children with cerebral palsy: Randomized, controlled, crossover trial. Journal of Child Neurology, 21, 931–938. http://dx.doi.org/10.1177/08830738060210110401

Eliasson

A. C.

Krumlinde-Sundholm

Shaw

, & Wang

(2005). Effects of constraint-induced movement therapy in young children with hemiplegic cerebral palsy: An adapted model. Developmental Medicine and Child Neurology, 47, 266–275. http://dx.doi.org/10.1017/S0012162205000502

Folio

M. R.

, & Fewell

R. R.

(2000). Peabody Developmental Motor Scales—Second edition. Austin, TX: ProEd.

10.

Gordon

A. M.

(2011). To constrain or not to constrain, and other stories of intensive upper extremity training for children with unilateral cerebral palsy. Developmental Medicine and Child Neurology, 53(Suppl 4), 56–61. http://dx.doi.org/10.1111/j.1469-8749.2011.04066.x

11.

Gordon

A. M.

Charles

, & Wolf

S. L.

(2006). Efficacy of constraint-induced movement therapy on involved upper-extremity use in children with hemiplegic cerebral palsy is not age-dependent. Pediatrics, 117, e363–e373. http://dx.doi.org/10.1542/peds.2005-1009

12.

Hart

(2005). Can constraint therapy be developmentally appropriate and child-friendly? Developmental Medicine and Child Neurology, 47, 363. http://dx.doi.org/10.1017/S0012162205000708

13.

Hoare

Imms

Carey

, & Wasiak

(2007). Constraint-induced movement therapy in the treatment of the upper limb in children with hemiplegic cerebral palsy: A Cochrane systematic review. Clinical Rehabilitation, 21, 675–685. http://dx.doi.org/10.1177/0269215507080783

14.

Holmefur

Krumlinde-Sundholm

Bergström

, & Eliasson

A. C.

(2010). Longitudinal development of hand function in children with unilateral cerebral palsy. Developmental Medicine and Child Neurology, 52, 352–357. http://dx.doi.org/10.1111/j.1469-8749.2009.03364.x

15.

Hosmer

D. W.

, & Lemeshow

(1989). Applied logistic regression. New York: Wiley.

16.

Hsin

Y. J.

Chen

F. C.

Lin

K. C.

Kang

L. J.

Chen

C. L.

, & Chen

C. Y.

(2012). Efficacy of constraint-induced therapy on functional performance and health-related quality of life for children with cerebral palsy: A randomized controlled trial. Journal of Child Neurology, 27, 992–999. http://dx.doi.org/10.1177/0883073811431011

17.

Huang

H. H.

Fetters

Hale

, & McBride

(2009). Bound for success: A systematic review of constraint-induced movement therapy in children with cerebral palsy supports improved arm and hand use. Physical Therapy, 89, 1126–1141. http://dx.doi.org/10.2522/ptj.20080111

18.

Lin

K. C.

Chen

H. F.

Chen

C. L.

Wang

T. N.

C. Y.

Hsieh

Y. W.

, & Wu

L. L.

(2012). Validity, responsiveness, minimal detectable change, and minimal clinically important change of the Pediatric Motor Activity Log in children with cerebral palsy. Research in Developmental Disabilities, 33, 570–577. http://dx.doi.org/10.1016/j.ridd.2011.10.003

19.

Lin

K. C.

Wang

T. N.

C. Y.

Chen

C. L.

Chang

K. C.

Lin

Y. C.

, & Chen

Y. J.

(2011). Effects of home-based constraint-induced therapy versus dose-matched control intervention on functional outcomes and caregiver well-being in children with cerebral palsy. Research in Developmental Disabilities, 32, 1483–1491. http://dx.doi.org/10.1016/j.ridd.2011.01.023

20.

Marx

, & Cummings

(2007). Normalized change. American Journal of Physics, 75, 87–91. http://dx.doi.org/10.1119/1.2372468

21.

Ottenbacher

K. J.

Taylor

E. T.

Msall

M. E.

Braun

Lane

S. J.

Granger

C. V.

, & Duffy

L. C.

(1996). The stability and equivalence reliability of the Functional Independence Measure for Children (WeeFIM). Developmental Medicine and Child Neurology, 38, 907–916. http://dx.doi.org/10.1111/j.1469-8749.1996.tb15047.x

22.

Rostami

H. R.

, & Malamiri

R. A.

(2012). Effect of treatment environment on modified constraint-induced movement therapy results in children with spastic hemiplegic cerebral palsy: A randomized controlled trial. Disability and Rehabilitation, 34, 40–44. http://dx.doi.org/10.3109/09638288.2011.585214

23.

Sakzewski

Ziviani

, & Boyd

R. N.

(2011). Best responders after intensive upper-limb training for children with unilateral cerebral palsy. Archives of Physical Medicine and Rehabilitation, 92, 578–584. http://dx.doi.org/10.1016/j.apmr.2010.12.003

24.

van Hartingsveldt

M. J.

Cup

E. H.

, & Oostendorp

R. A.

(2005). Reliability and validity of the fine motor scale of the Peabody Developmental Motor Scales–2. Occupational Therapy International, 12, 1–13. http://dx.doi.org/10.1002/oti.11

25.

Wallen

Bundy

Pont

, & Ziviani

(2009). Psychometric properties of the Pediatric Motor Activity Log used for children with cerebral palsy. Developmental Medicine and Child Neurology, 51, 200–208. http://dx.doi.org/10.1111/j.1469-8749.2008.03157.x

26.

Willis

J. K.

Morello

Davie

Rice

J. C.

, & Bennett

J. T.

(2002). Forced use treatment of childhood hemiparesis. Pediatrics, 110, 94–96. http://dx.doi.org/10.1542/peds.110.1.94

27.

Wolf

S. L.

Winstein

C. J.

Miller

J. P.

Taub

Uswatte

Morris

Nichols-Larsen

; EXCITE Investigators. (2006). Effect of constraint-induced movement therapy on upper extremity function 3 to 9 months after stroke: The EXCITE randomized clinical trial. JAMA, 296, 2095–2104. http://dx.doi.org/10.1001/jama.296.17.2095