Abstract
The study findings support the clinical efficacy and feasibility of constraint-induced movement therapy and bimanual intensive training protocols for children with unilateral cerebral palsy.
Children with unilateral cerebral palsy (UCP) demonstrate unimanual motor impairment that is due to a disturbance in the developing brain. Constraint-induced movement therapy (CIMT) and bimanual intensive training (BIT) have been developed for this population to ameliorate their upper limb functions and to further improve participation in daily activities (Sakzewski et al., 2014). Over the past decade, cumulative studies on CIMT and BIT have indicated that these well-defined and intensive training models are more promising than the usual care (Hoare et al., 2019; Ouyang et al., 2020). Current research has continuously focused on comparing the effectiveness of these neurorehabilitative programs and beneficial components to establish optimal intervention plans for children with UCP (Hoare et al., 2019; Simon-Martinez et al., 2020; Walker et al., 2022).
Previous studies have tried to identify whether CIMT or BIT is more effective than the other. The results demonstrated similar improvements from these two neurorehabilitative programs in various outcomes, such as motor performance, daily participation, and quality of life (Hoare et al., 2019; Sakzewski et al., 2012). However, two knowledge gaps that warrant further investigation were identified. First, those previous findings were generally based on comparisons of conventional high-dose neurorehabilitation programs with intensive designs of at least 60 to 90 training hours. Because the feasibility and practicality of those high-dose programs presented significant challenges, several studies have made modifications to improve their clinical accessibility by reorganizing or decreasing the intervention hours (Amjad et al., 2016 ; Chen et al., 2019; Wu et al., 2020). Nevertheless, the differences in efficacy between these modified CIMT and BIT programs have not been investigated sufficiently.
To the best of our knowledge, only one study has compared the effectiveness of CIMT and BIT with fewer training hours (2 hr/day, 6 days/wk, for 2 wk, 24 hr in total; Amjad et al., 2016). The results of that study showed that CIMT yielded more improvement than BIT did in unilateral performance, such as in grasp and dissociated movements. However, the study used nonequivalent dosages for the two interventional programs. The CIMT group received an extra 6 hr of constraint of the less affected hand to force use of the more affected hand. Thus, the more beneficial outcomes of CIMT might have been a result of the extra hours of forced use. Also, bimanual outcomes and motor improvement progress were not included in Amjad et al.’s (2016) study. To fill the first knowledge gap, a longitudinal randomized controlled trial with an equal dosage design and comprehensive measurements covering unimanual and bimanual outcomes is needed.
Second, psychosocial outcomes, including the children’s engagement and parental stress levels, have not been compared between the CIMT and BIT programs. It is known that programs that are based on neuroplasticity and motor learning theories generally consist of intensive and repetitive practice; maintaining the child’s engagement in the training activities presents another challenge (Mancini et al., 2013). Because engagement in an intervention is rated as one of the influential characteristics determining treatment outcomes (Cramer et al., 2011), it is particularly important for clinicians and researchers to investigate a child’s engagement in the neurorehabilitative clinic. On the basis of literature review, the comprehensive measure of engagement is suggested to include concepts of extrinsic motivation, intrinsic motivation, and basic psychological needs (Tatla et al., 2015), and it is recommended that engagement should be measured from different perspectives (Wright & Majnemer, 2014). However, because of the lack of sound standardized assessments for investigating engagement, this issue has rarely been addressed; furthermore, this issue has been addressed only with the use of qualitative or anecdotal documents in a previous neurorehabilitative study (Mancini et al., 2013). In addition, unlike adult clients, who are self-motivated to regain lost functions, children with cerebral palsy (CP) often participate in training programs to meet their parents’ expectations. The potential conflicts between a parent’s and a child’s expectations may lead to parenting stress. Lin et al. (2011) demonstrated that the intensive and intrusive constraint principles led to deterioration in the interactions between parents and their children. Thus, understanding and comparing the parental stress level for these modified protocols is also important.
The aim of this study was to investigate and compare the efficacy of 36-hr dosage–matched CIMT and BIT programs on motor and psychosocial outcomes. To demonstrate the progress of motor improvement precisely and the parental stress status continuously, unilateral and bilateral motor performance and parental stress levels were evaluated four times: at preintervention, midterm of intervention (after 18 hr of training), postintervention (after 36 hr of training), and follow-up (6 mo after training). In addition, we documented the children’s engagement weekly from both the children’s and the parents’ perspectives to investigate their engagement status throughout the whole intervention period.
Method
Participants
Children were recruited from the Cerebral Palsy Association of R.O.C., medical centers, and special education systems in Taipei and New Taipei City in Taiwan. Those who met the following inclusion criteria were included: (1) congenital UCP; (2) ages 6 to 12 yr; (3) active extension at the wrist and metacarpophalangeal joint of the more affected hand ≥10°; (4) no excessive muscle tone (Modified Ashworth Scale score ≤2 at any joints of the upper limb) before treatment; (5) the ability to follow instructions according to medical documents, parental reports, and clinical observation; and (6) no injections of botulinum toxin type A or operations on the hand within 6 mo.
Design and Procedure
Eligible participants were randomly allocated to either the CIMT group or the BIT group using a web-based randomization tool. The allocation sequence was generated by the corresponding author (Tien-Ni Wang), and information was delivered to the attending therapist directly. A sample size of at least 21 children for each group was suggested, given a large effect size (d = .80), a power of .80, and a one-sided Type 1 error of .05. The effect size was determined by the findings of primary outcomes in a previous study that compared the efficacy of CIMT with that of BIT (Amjad et al., 2016). Considering the possibility of a 10% to 20% dropout rate, we initially recruited 25 patients for each intervention group. Motor outcomes were performed at pretreatment, midterm (18 hr), posttreatment (36 hr), and 6 mo after the intervention. All motor measures were conducted by three trained occupational therapists masked to the study design. Each therapist was trained to properly administer the outcome measures on the basis of the careful review of written instructions and repeated practice. Rater competence, defined as the ability to implement and score the assessments accurately and independently, was evaluated by a senior certified occupational therapist (Tien-Ni Wang). Psychosocial outcomes included the weekly assessments of the child’s engagement as well as parental stress level at pretreatment, midterm, posttreatment, and 6 mo after the intervention. The study was approved by the National Taiwan University Ethics Committee and was registered with ClinicalTrials.gov (NCT02808156).
Interventions
To address existing challenges regarding the highly intensive protocol, a child- and family-friendly intervention approach was implemented in this study. The approach was child-friendly in that training activities were provided according to each child’s hand function to fit the child’s preferences. A playful atmosphere was emphasized. The CIMT and BIT programs were also conducted in the child’s natural environment (home or school) to facilitate the generalizability of treatment effects to real-world function. The approach was also family-friendly in that the interventions were arranged and distributed with appropriate training dosages that were easy to merge into the participants’ daily schedules. The optimal training hours were decided on the basis of a literature review and a pilot survey to estimate the feasible intervention intensity for the families with school-age children (Chen et al., 2019; Jackman et al., 2020). The final intervention dosage that fit their family schedules and daily routines was 2.25 hr/day, twice a week, for 8 wk (a total of 36 hr of intervention). The two intervention sessions were scheduled, one on a weekday with only a half day of school and one on the weekend. Participants could continue their usual rehabilitation care during the study period.
The intervention was provided by certified occupational therapists who had completed a 2-day training program on both the CIMT and BIT protocols. Interventional dialogue for each training section was documented and supervised by a senior certified occupational therapist to ensure the fidelity of the protocols. The principles of shaping and repetitive task practice were applied in both groups. Shaping is a training method in which a motor or behavioral objective is approached in small steps by successive approximations (e.g., a task is gradually made more difficult for a person’s motor capabilities), and repetitive task practice involves functional tasks that are performed continuously over a specific period. The therapists graded the intervention tasks according to each child’s hand function and gave appropriate feedback to enhance motor learning. Types of target movements, including grasp, release, carry, stabilization, and in-hand manipulation, were identified according to a comprehensive literature review (Charles & Gordon, 2006). Finally, the tasks of each intervention protocol were chosen with consideration of the child’s specific upper limb impairments and the appropriate level of difficulty, as well as the child’s preferences. The 2.25 hr of training were divided into four to five sessions (30 min per session) to fit the children’s attention span. A short break (5–10 min) was allowed between sessions. Each training session usually included one treatment activity that integrated developmentally appropriate games for each participant, such as board games, card games, manipulation activities, crafts, and ball games. During the therapeutic sessions, parents were welcome to observe and learn about the training activities. There was no extra assignment outside the therapy sessions, but the children in both groups were encouraged to use their more affected hands in daily activities.
The major difference between the CIMT and BIT groups was that, in the CIMT group, the focus was on training the more affected hand with constraint of the less affected hand, whereas in the BIT group, both hands were involved in the activities. In the CIMT group, to reduce the negative emotional response and relieve parental stress, the therapist used verbal instructions and gentle physical guidance (holding the child’s hand) instead of a glove or splinting to constrain the less affected hand (Chen et al., 2019). Because of the different characteristics of the CIMT and BIT interventions, the more affected hand in the BIT group was more often the assisting hand, whereas in the CIMT group, it was more often the task hand.
Outcome Measures
Motor Outcomes
Motor outcomes included two assessments for unilateral performance and two for bilateral performance. Each contained one test of observation-based motor capacity and one self-report of daily motor functions. We used the Melbourne Assessment 2 (MA–2) and Pediatric Motor Activity Log–Revised (PMAL–R) to measure the children’s unilateral movement skills and daily function, and we used the Bruininks–Oseretsky Test of Motor Proficiency, Second Edition (BOT–2), and the ABILHAND–Kids measure to examine the children’s bimanual coordination and bimanual daily activities.
The MA–2 was developed to evaluate the unilateral upper limb movement quality in children ages 2.5 to 15 yr with neurological impairment (Randall et al., 2014). Versions in several languages are available on the official website. Each child was asked to perform 14 functional tasks, which were videotaped for subsequent scoring of four subscales. Items were scored according to each element of movement measured and categorized into four corresponding subscales measuring (1) range of movement (ROM); (2) accuracy of reach and placement; (3) dexterity of grasping, releasing, and manipulation; and (4) fluency of movement. Each subscale was scored on a 3-, 4-, or 5-point ordinal scale using specifically developed criteria. Final scores assigned to each subscale were converted into percentage scores by dividing them by the maximum possible score. The MA–2 has sound psychometric properties with high test–retest reliability (rs = .92–.98), good concurrent validity, and large responsiveness (Wang et al., 2017). We used the PMAL–R, a questionnaire-based assessment, to evaluate the spontaneous use of the affected upper limb in 22 daily activities (Uswatte et al., 2012). Each activity was rated by parents or caregivers on 6-point ordinal scales (0 = not used, 5 = normal) with 0.5-point increments possible. The ratings were provided for two subtests: how often (HO), which measured the amount of use of the affected hand, and how well (HW), which measured the quality of use of the affected hand. Higher scores indicate that a child used the affected hand more frequently than, or similarly to, an age-matched child. The PMAL–R was developed for children with CP and has high internal consistency (Cronbach’s α = .93), sound test–retest reliability (r = .89), and fair concurrent and predictive validities (rs = .31–.48; Lin et al., 2012; Uswatte et al., 2012). The BOT–2 is a standardized assessment that is frequently used in upper limb neurorehabilitation effectiveness studies for children with CP (Bruininks & Bruininks, 2005). It has been reported to have excellent test–retest reliability (r = .99), good internal consistency (Cronbach’s α = .92), and sound construct validity (Deitz et al., 2007; Wuang & Su, 2009). Subtest 3 of the BOT–2, manual dexterity, uses goal-directed activities that include reaching, grasping, and bimanual coordination with small objects to investigate a child’s upper limb manual function. The point scores of manual dexterity (range = 0–45) were used for analysis. The ABILHAND–Kids questionnaire is a Rasch-based assessment completed by the caregivers. Versions in several languages are available on the official website. Caregivers were instructed to rate their child’s difficulty in performing 21 daily activities on a 3-point response scale (0 = impossible, 1 = difficult, 2 = easy). The ABILHAND–Kids demonstrated good construct validity, internal consistency reliability (r = .94), and test–retest reliability (r = .91; Arnould et al., 2004). The logit score (i.e., the conversion of the ordinal score into a linear measure of ability) was used.
Psychosocial Outcomes
Psychosocial outcomes included the child’s weekly engagement and parental stress level at pretreatment, midterm, posttreatment, and 6 mo after the intervention. The Engagement Questionnaire (EQ) was used to investigate the participants’ engagement in the therapeutic programs. The items in the EQ were modified from existing questionnaires (Green & Wilson, 2012) to contain the concepts of basic psychological needs, extrinsic motivation, and intrinsic motivation. In this study, the EQ was modified into two versions to represent the participants’ engagement from both the children’s and parents’ perspectives (e.g., Do you look forward to the next session? vs. Do you think your child looks forward to the next session?). Items on the EQ are rated from 1 (strongly disagree) to 5 (strongly agree), and the total score ranges from 0 to 30, with higher scores indicating better engagement. Finally, the Chinese version of Parenting Stress Index–Short Form (PSI–SF) was used to measure the stress level of the parents. It includes 36 items rated from 1 (strongly agree) to 5 (strongly disagree). High scores on the PSI–SF indicate lower stress in the parent. The PSI–SF demonstrated sound internal consistency (Cronbach’s α = .86–.95) and construct validity (Weng, 2011).
Statistical Analysis
We evaluated baseline differences between the two groups by using an independent t test and χ2 test. Normality was verified using the Shapiro–Wilk test and inspection of the histograms for symmetry. Internal reliabilities were examined for each measure by calculating Cronbach’s α. To conduct parametric statistics, we applied a Box-Cox transformation to the parameters of range of motion, the accuracy of the MA–2, and the PMAL–R. Outcome measures were compared between groups by fitting a linear mixed model including random intercepts, group, time, and Time × Group interactions as independent factors, as well as potential covariates. Covariates were retained if the model fit showed a lower corrected Akaike information criterion value. The level of significance was set at p < .05. For significant main effects and interactions, we conducted post hoc analyses using t tests applying Bonferroni corrections for multiple comparisons in case of significant trends (<.10), as this allowed us to capture tendencies immediately after the intervention (Simon-Martinez et al., 2020). We calculated the effect sizes of the Group × Time interaction and the main effect of time from the F values, according the Cohen’s η2 p formula: η2 p = (F × df between-subjects)/ ((F × df between-subjects) + df within-subjects) (Cohen, 1973). To explore children’s engagement across treatment sessions, we used a linear mixed model, as well as a simple line graph, to present each week’s scores on the EQ.
Results
Fifty children were randomly assigned to the CIMT group or the BIT group, and 48 children completed the study. One child in the CIMT group dropped out because of injections of botulinum toxin during the intervention, and one child in the BIT group dropped out because of issues with transportation. All 48 participants completed the 36-hr intervention within 8 to 10 wk. Each participant was assigned one specific certified occupational therapist; a total of 17 certified occupational therapists provided interventions for this study. No adverse events were reported throughout the study. The recruitment process is described in Figure 1. There were no significant group differences in baseline scores for any measures (Table 1). The internal reliabilities of all measures were excellent, with Cronbach’s α values of .87–.94 for the MA–2, .88–.92 for the PMAL–R, .81 for the BOT–2, .93 for the ABILHAND–Kids, and .95 for the PSI–SF.
Description of recruitment following CONSORT guidelines.
Demographic Characteristics of the Study Participants
Note. BIT = bimanual intensive training; CIMT = constraint-induced movement therapy; MACS = Manual Ability Classification System.
Table 2 shows the means for each group at each time point for the unilateral and bilateral motor outcomes. For the unilateral motor outcome, the results of the mixed-effect model showed a significant Group × Time interaction on the HW and HO subtests of the PMAL–R at the 6-mo follow-up, favoring CIMT. There were no significant Group × Time interactions on the MA–2 at any time point. We found main effects for time on all subscales of the MA–2. Post hoc analysis revealed initial improvements on three subscales (range of movement and fluency) of the MA–2 at midterm assessment but not the accuracy and dexterity subscales of the MA–2, which reached significant improvement at posttreatment. All improvements on the MA–2 were retained at the 6-mo follow-up. For the bilateral motor outcome, there were no significant Group × Time interactions on the BOT–2 and ABILHAND–Kids. Both groups showed initial significant improvements on the BOT–2 at midterm assessment and on the ABILHAND–Kids at posttreatment. These improvements were retained at the 6-mo follow-up.
Estimated Marginal Means of Outcome Measures at Each Time Point, and Statistical Comparison
Note. BIT = bimanual intensive training; BOT–2 = Bruininks–Oseretsky Test of Motor Proficiency, Second Edition; CIMT = constraint-induced movement therapy; MA–2 = Melbourne Assessment 2; PMAL–R = Pediatric Motor Activity Log–Revised; PSI–SF = Parenting Stress Index–Short Form; ROM = range of movement.
p < .05.
For the psychosocial outcomes, the PSI–SF showed no significant interaction or time effects, reflecting that the parents’ stress did not differ between groups and was stable during the interventions. The weekly scores on the EQ, completed by the children and parents, are presented in Figure 2. The CIMT and BIT groups showed similar trends on the child and parent versions of the EQ. In addition, the parent version of the EQ yielded consistently high weekly scores throughout the intervention. The child version of the EQ yielded relatively lower EQ scores in the beginning but continuous increases during the 8 wk of intervention.
Scores on the EQ During Weekly Interventions: (A) Parents and (B) Children.
Discussion
The study findings showed that, whether they received CIMT or BIT, children with UCP achieved similar motor improvements in both unilateral and bilateral motor performance. Overall, there were small differences in motor effectiveness between the two programs, and such differences only appeared at the 6-mo follow-up: At this evaluation, the CIMT group demonstrated greater improvements than the BIT group on the HO and HW of the PMAL–R, indicating that CIMT led to a higher frequency and quality of use of the more affected hand in daily activities. Children in both groups steadily maintained high engagement throughout the whole intervention period, and the parental stress levels were consistently static. These positive psychosocial outcomes provided evidence to support our design of clinically friendly features for the two proposed intervention programs.
The between-groups differences were found only for the scores on the HO and HW of the PMAL–R at the 6-mo follow-up, with the CIMT group showing greater improvement. The findings, consistent with previous high-dose effectiveness studies (Gelkop et al., 2015; Hoare et al., 2019), suggested that both CIMT and BIT yielded similar improvements in motor performance within the interventional process (from baseline to midterm and from baseline to immediately after the intervention). However, at the 6-mo follow-up, CIMT demonstrated additional benefits for daily use of the more affected hand as measured by the PMAL–R. This result indicated that children in the CIMT group might have had more potential for transferring learned abilities to daily motor function than the children in the BIT group. In addition, the characteristics of CIMT might provide another explanation for this result. The core therapeutic principles of CIMT, including intensive use of the more affected hand and forbidden use of the less affected hand, were easy to duplicate for the caregivers in their daily routines. Thus, in the long run, we found additional benefits of using the more affected hand in daily routines.
The unimanual CIMT approach had more favorable results for unilateral daily use at the 6-mo follow-up; however, the same result was not found for the BIT group for bimanual motor outcomes. This might be explained by the fact that, for the BIT group, even if parents were exposed to the bimanual protocol, they would intuitively still focus on the child’s more affected hand (Ferre et al., 2017). By observing the therapists’ interventional strategies, parents learned the instructions that directed the child to use both hands together. However, they might not have been sensitive enough to notice the role of the more affected hand in different bimanual tasks, or they may have lacked the motor knowledge to enhance bilateral coordination. Thus, parents in the CIMT group seemed to be able to extend and expand improvements in daily functional use of the more affected hand, whereas parents in the BIT group did not demonstrate this influence.
To our knowledge, this study is the first to compare and track children’s weekly engagement in CIMT and BIT from both the children’s and parents’ points of view. No significant differences were found between the two protocols; thus, the results of both groups were merged together for analysis. For engagement from the children’s perspective, it is interesting that, despite the high average score, both groups demonstrated relatively lower engagement in the beginning, although it gradually increased over the following weeks. These findings echoed those of previous qualitative studies, which stated that children might show marked frustration and complain in the early stages of the intensive neurorehabilitative program (Mancini et al., 2013). This study further provides quantitative evidence to confirm and support the existing qualitative documents. Moreover, on the basis of our clinical observation, there might be a potential psychosocial difference between the two groups in the child’s self-efficacy. In the CIMT group, therapists provided more assistance or grading because the children were required to use the more affected hand as their task hand during entire sessions—a requirement they might not have been familiar with. Unlike the BIT program, in which children used the more affected hand as their assisting hand, the CIMT program’s challenges might, as stated earlier, negatively affect children’s self-efficacy. Measurements reflecting self-efficacy should be considered in future work.
In addition, the consistently high engagement scores over the whole intervention period supported our child-friendly design in terms of gentle constraint and providing training activities that fit the children’s preferences. Regarding engagement from the parents’ perspective, it is noted that the scores on the parent version of the EQ remained steady and high for the whole intervention period. The relatively higher scores of parents on the weekly EQ may suggest that they overestimated their child’s engagement during the intervention. These findings should remind clinicians and researchers that children have different perceptions of interventional engagement (Kennedy et al., 2012). Thus, the opinions of the children themselves should be considered simultaneously in future work.
Regarding the parental stress level, similar scores on the PSI–SF were found for both groups at each evaluation visit, indicating that the parents demonstrated stable emotional statuses. These findings supported our modification of existing neurorehabilitative protocols to make them family-friendly. The protocols in this study were arranged as intervention delivery schedules of 2.25 hr/day, twice a week, for 8 wk, to fit the ecological features of local families’ routines, which did not increase parental stress. Previous studies have also demonstrated that, for children with CP as well as their caregivers, training implementation at home for 2 hr/day is more preferable to, and more comfortable than, training for 3 to 4 hr/day (Eliasson et al., 2005; Gelkop et al., 2015; Wu et al., 2020). Because parental stress levels influence parents’ well-being and children’s development, delivering intervention protocols that do not increase parents’ burden is important.
To understand motor improvement progress during the interventions and whether these changes were retained after the interventions, this study implemented evaluations at four time points (pretreatment, 18-hr midterm, 36-hr posttreatment, and 6-month follow-up). For the unilateral outcome, both groups showed initial improvements at the 18-hr midterm on two subtests (range of movement and fluency) of the MA–2 and the HO of the PMAL–R but not on the dexterity and accuracy subtests of the MA–2 or HW of the PMAL–R. These initial improvements indicated that 18 hr of intensive training could lead to significant motor improvements on the basic components of motor abilities, as well as the frequency of daily use. Nonetheless, improvements in advanced motor skills and quality of daily use did not appear until the end of the 36-hr training program. For bilateral outcomes, Subtest 3 of the BOT–2 reached significant improvement after 18 hr of training, whereas the ABILHAND–Kids measure needed 36 hr for improvements to appear. The initial improvements after 18 hr of intensive training indicated that the child’s more affected hand could play a better role as the assisting hand in performing bilateral activities. Bilateral daily motor function improved later, after completion of the 36-hr training program. These promising results suggest that children who received either CIMT or BIT would demonstrate motor improvements after 36 hr of training, and some even after 18 hr. Given the intensive protocols that demand large amounts of resources and high costs, understanding the possible minimum dosage to yield changes for different motor prospects is vital (Ilieva & Ilieva, 2020).
Despite the promising results, several limitations of this study need to be considered. First, we did not stratify participants by age or Manual Ability Classification Scale level to investigate the differences between CIMT and BIT because of the various sample sizes in each stratified group. Future research with larger sample sizes to examine treatment effects within stratified groups is recommended to further clarify the effects of age and motor status. Second, subjective questionnaires (e.g., PMAL–R and ABILHAND–Kids) were used to investigate daily use of the upper limbs in this study. Because it is difficult to blind parents to the content of interventions, the parents may have had a bias that was based on the intervention their children received. Therefore, future studies should include an objective measure such as an accelerometer for exploring real-world hand use to validate the study findings.
Implications for Occupational Therapy Practice
The study results have the following clinical implications for occupational therapy practice: ▪ Motor improvements in basic skills appeared after 18 hr of intensive training, whereas improvements in advanced motor skills were achieved after 36 hr of training. Occupational therapists could provide at least 36 hr of CIMT or BIT to achieve improvements in both basic and advanced motor skills. ▪ CIMT yielded greater improvements than BIT on frequency and quality of daily use of the more affected hand at the 6-mo follow-up, indicating that, for parents, the core principle of CIMT (i.e., remind the child to use the more affected hand) might be easier to duplicate than those of BIT. Thus, occupational therapists should provide more guidance and education for parents when implementing BIT. ▪ In interventions, occupational therapists should consider the opinions of not only the parents but also the children themselves.
Conclusion
To our knowledge, this study is the first to compare the efficacy of the motor and psychosocial outcomes of CIMT and BIT, with a relatively smaller 36-hr dosage. We found small differences in motor and psychosocial performance between the CIMT and BIT groups. The only major difference was that CIMT yielded better frequency and quality of daily use of the more affected hand at the 6-mo follow-up, which might imply that the therapeutic principle of CIMT (i.e., remind the child to use the more affected hand) was easier for the parents to imitate than those of BIT. The possible minimum dosage to yield changes has been established for different motor prospects. Eighteen hours of intensive training can achieve motor improvements in basic skills, whereas improvements in advanced motor skills require 36 hr of training. For psychosocial outcomes, children in both groups exhibited similarly high levels of engagement, and their parents showed stable stress levels during the whole intervention period. However, the children and their parents reported different perceptions of treatment engagement, indicating that the opinions of the children themselves should be considered simultaneously in future work. In summary, promising results for motor and psychosocial outcomes were found in this study for our child- and family-friendly CIMT and BIT protocols.
Footnotes
Acknowledgments
We thank the children and their families for participating in this study. This project was supported in part by the Ministry of Science and Technology (MOST 110-2314-B-002-061 to Tien-Ni Wang and MOST 110-2628-E-002-004 to Hao-Ling Chen).