INTRODUCTION AND PURPOSE
Measurement of gross motor function in children with cerebral palsy (CP) helps physical therapists (PTs) identify abilities and limitations, set individualized goals, and plan well-targeted interventions. The Gross Motor Function Measure (GMFM-66)1–3 is an internationally accepted approach to quantify gross motor abilities in children with CP. However, the measure's usefulness for elucidating advanced motor skill goals and documenting improvement with children in level I of the Gross Motor Function Classification System (GMFCS) is limited because those above age 5 years typically achieve scores of 85% to 100%.4 Indeed, the intended purpose of the GMFM-66 is to evaluate the foundations of gross motor function2 rather than high-level skills, speed, or quality of performance, all of which may limit participation in gross motor activities that children want to do.2 In thinking about how to measure these critical aspects of performance, motor skills assessments that are product-oriented (eg, focus on the number of successful repetitions, time required, or distance achieved) have been shown to be strongly linked to functional abilities and participation in children with CP, whereas process-oriented measures (eg, quality of movement) are best-suited to providing information on underlying impairments.5,6 Use of both approaches within a single measure might provide additive value in the process of strength or problem identification and goal setting.
Several well-known, validated pediatric measures were designed to evaluate gross motor abilities, but as articulated below, each has limitations when applied as an outcome tool for children with CP. For example, the Movement Assessment Battery for Children (M-ABC)7 is widely used to assess school-age children with mild to moderate motor delays. It focuses on 3 primary skill areas (ie, manipulative, ball, and balance skills7,8), and measures both quantitative (ie, time) and qualitative (ie, body posture) aspects of performance. The issue is that although the M-ABC has some potentially suitable advanced skills in its balance, aiming, and catching domains, it contains only 5 gross motor skills (2 of which are bilateral) and is diluted overall by the inclusion of manual dexterity skills. In addition, the M-ABC was not built to evaluate change. Although it has been used to measure change in children with developmental coordination disorder (DCD), issues have been documented with respect to its capacity to serve as an outcome tool,9 suggesting it perhaps serves best as a one-time performance test. Furthermore, even with children with DCD for whom it is most commonly used, reliability and validity limitations have been documented.10
The Bruininks-Oseretsky Test of Motor Proficiency (BOT-2) is often used by occupational therapists and PTs to assess fine and gross motor delay in children with neurodevelopmental disorders.11 The BOT-2 groups skills in areas of manual coordination, body coordination, strength, and agility. Although more frequently used than the M-ABC in published studies of children with CP, the BOT-2 also has a small group of gross motor items and the scaling is designed for comparison with age norms rather than change within the child. Test properties in children with CP are unknown and should not be assumed to be similar to the established properties of children with mild motor disorders such as DCD. The BOT-2 has tended to be used in upper limb intervention studies when applied to children with CP. Less frequently used in gross motor studies, the BOT-2 has shown mixed results as far as sensitivity to change.12–14
The Test of Gross Motor Development (TGMD) is divided into locomotor and object control skills.15 Although the skills tested are relevant to participation in school and extracurricular activities,16,17 the focus is on movement quality. Use of subcomponents of this test with children who have CP is described in 2 articles on product-oriented and process-oriented measures of fundamental movement skills among these children,5,6 but we did not find any published validation or outcome studies with the TGMD in children with CP.
The Physical and Neurological Examination for Subtle Signs (PANESS) Assessment18,19 and Zurich Neuromotor Assessment20 are measures of developmental status of motor control in children with suspected or identified neurodevelopmental disorders (eg, DCD, autism spectrum disorders, and CP). The focus of these norm-based measures is on movement speed (performance) and quality through measurement of timing in repetitive movement tasks, balance within gait tasks, and identification of associated movements during task performance (eg, dysrhythmia and overflow). However no published work was found on children with CP using the PANESS, and studies that have used the Zurich in CP have focused on prediction of outcomes in relation to the type of CP.21 In addition to these motor assessments, standardized fitness tests can be used in their original or adapted form to assess children with CP.22–24 However, they are not intended to examine a child's motor skill proficiency or developmental level.
Taken collectively, these measures provide information on global movement properties including quality and speed of performance, but they were not designed as outcome measures for children with CP, and their ability to serve in this role has either been called into question or remains untested. Furthermore, it was noted by the clinical experts in the Challenge Module development study25 that the tasks in these measures do not sufficiently represent skills or components of performance that children require for participation in school or outside activities. Because children in GMFCS level I have strong potential to benefit from physical therapy and other rehabilitation therapies that are targeted toward performance of skills such as running, jumping, skipping, and ball skills that facilitate participation with peers,26,27 the lack of a tool that comprehensively measures these advanced skills is a serious limitation both for preintervention evaluation (the lead up to goal setting) and measurement of outcomes.
For all of these reasons, the initial version of the Challenge Module25 was created as a proposed adjunct to the GMFM-66.2 Despite its strength overall as outcome measures for children with CP,3,28,29 the GMFM-66 has limited sensitivity when used with children in GMFCS level I. The GMFM-66 was a strong foundation on which to build the Challenge as far as focus, style, and format, and from the outset, thoughts were that the Challenge might ultimately be able to be structurally linked the GMFM-66 through Rasch scaling procedures. Indeed, reliability evaluation and Rasch scaling30 of the Challenge with a large sample of children with CP (GMFCS level I) were in progress at the time this article was written.
Items for the Challenge were generated from the gross motor measures listed above, and refined during a focus group with 3 expert pediatric PTs.25 This focus group identified 35 nonoverlapping priority items for further review. Item reduction was then conducted using pediatric PTs’ ratings of item importance and safety in 2 online surveys. The first item reduction survey (n = 86 PT respondents) resulted in 20 highly endorsed items. The second survey (n = 49 PTs) yielded 2 more items.
The original response options for the initial version of the Challenge were created on the basis of the GMFM's 4-point “extent of task completion” scale (rated as “0” to “3”), with skill performance wording individualized to each Challenge item. The Challenge was tested with 7 children with CP (ages 6-14 years) in GMFCS level I who had scored at least 90% on the GMFM.25 The 22-item test revealed performance challenges beyond those of the GMFM and was found to be feasible to administer. The Challenge total mean score with the pilot sample was 74.5% (standard deviation [SD] = 19.4), demonstrating lack of a ceiling effect overall. Three items were found, however, in which all 7 children scored a perfect “3” for performance. The children confirmed that these 3 tasks were too easy, and hence all 3 were removed. Two other items, catches ball bounced outside base of support and bounces/catches a volleyball were deemed similar and thus combined to reduce test burden. Seven new items arose from the children and parents as a result of conversations with the assessors at the end of their test about functional skill priorities at school and play. Two of these 7 new items were endorsed for inclusion in the Challenge as per the results of a third online survey sent to pediatric PTs (n = 26). The result was a 20-item Challenge25 that was ready for the next development stage.
During the above process, Wilson et al25 observed that although most of the children with CP in the pilot sample could score the full “3” points on what were thought to be the difficult items, they did so with considerable effort and slow performance (ie, not in a functionally meaningful period). Hence, the scores of this first version of the Challenge did not capture important issues in performance speed (movement efficiency). To better reflect the functional demands of everyday physical activities (eg, play and sports), we decided to expand the Challenge's response options to have extent of task completion as the first requirement, and then include speed as a higher level performance requirement (ie, introduce a product-oriented assessment component).5 This type of dual-criterion scoring has been built into measures such as the Berg Balance Scale31–33 and the Community Balance and Mobility Scale34 to evaluate upper levels of performance. However, the approach taken by developers of those tests to determine the time cut-points was not articulated in their test manuals. For example, in the Berg Balance Scale times of 3 and 10 seconds were the selected cut-points for the standing unsupported with eyes closed item. These times may have been derived from observation of patients or able-bodied adults, or as a best guess estimate of a functional performance time.
Given the recommendations that stemmed from the work to date on the Challenge, the purpose of this study was to refine the Challenge to increase its sensitivity to change as well as its functional and interpretational relevance. The specific objectives were 3-fold: (1) to expand the item set slightly to increase the skill breadth of the test, (2) to revise the response scales to include empirically based performance (time/distance) cut-points from children with typical development (TD) aged 4 to 10 years, and then (3) to evaluate the revised Challenge as far as score patterns and test feasibility when used with children with TD.
The decision to base the cut-points on time scores from 4- to 10-year-old children with TD arose from the idea that they would display a range of performance abilities, and their maximum proficiency would be limited to a level that might be feasible for some older children with CP (GMFCS level I) to attain. Specifically, we hypothesized that the youngest children with TD (ages 4-7.5 years) would exhibit various stages of development in the extent of skill completion, whereas the older children (7.5-10 years) would complete the skills but show varying degrees of proficiency and efficiency. It is important to stress that the goal was to gain a deeper understanding of how young children with TD perform the Challenge skills so as to develop timing expectations for level I children/youth. We did not intend to determine age-specific norms for the items. Rather, this study's cut-point estimates serve as a starting point for fine-tuning the cut-points in the subsequent Rasch scaling of the Challenge with children in GMFCS level I (ie, with the understanding of item difficulty that the Rasch analysis provides, cut-times then can be adjusted as needed to achieve the desired gradient of item difficulty in the final version of the Challenge).
Preparation of the Challenge for the Next Phase of Development
The rescaling and testing of items that was done in this study began with the 20-item Challenge.25 We proceeded to systematically expand the Challenge to include additional items that required a higher degree of gross motor ability. The goal of adding these items was to measure children's abilities across a greater diversity of items related to balance, coordination, and speed that might be important foundations for participation in school recess/gym activities with friends, and also to enhance the Challenge's ability to discriminate among children. To maintain a link to the previous item selection process, these new items were those that had just missed the 75% “importance rating” endorsement cut by the expert PTs in the first phase of the Challenge's item reduction work (ie, items in the 58%-74% “importance rating” endorsement range).25
In preparation for time cut-point integration into the response sets, the general structure of the original Challenge's 4-level (“0” to “3”) response sets (Figure 1, version A) was re-designed because Wilson et al25 determined that the score of “0” (“does not initiate”) rarely applied to the children with CP in the pilot sample. This score, therefore, was modified to indicate up to partial performance (ie, <50% of the skill) and a score of “1” was altered to reflect considerable, but not full, completion of the task without error. We planned that a score of “2” would indicate 100% task completion with possible movement style variations, and “3” would be reworked by adding a time or distance component to capture movement efficiency (Figure 1, version A). Placeholders (eg, “x” seconds as shown in Figure 1, version A) were introduced into the response sets for the time and distance criteria to prepare for insertion of the cut-point values determined from testing of the children with TD in this study.
The next step was a review of the scores of the children with TD's performance on the Challenge using these new cut-points. This review allowed consideration of whether a second revision was needed to enhance the Challenge's discriminative abilities. If so, this revised version then would be rescored from the Challenge test videos to determine the effect of the response option alterations on scores.
Participant Enrollment for the Cut-Point Determination and Revised Challenge Scoring
The children were recruited through e-mails to staff at Holland Bloorview Kids Rehabilitation Hospital and through e-mail distribution to staff in the Department of Physical Therapy at the University of Toronto. Children were eligible to participate if they were 4 to 10 years of age inclusive, with TD, and judged by their parents to be able to follow instructions to perform the motor skills in the test. Examples of skills, such as “throwing and catching a ball,” “walking backward on a line,” “jumping rope,” and “running as fast as possible for 10 m and then stopping on a finish line,” were given in the information letter to provide concrete examples for parents. Children were not eligible if they had a known neurological condition; orthopedic injury requiring limitation of weight-bearing or other physical activity restriction within the last 12 months; or respiratory, cardiovascular or rheumatologic disease or visual impairment that limited the child's full participation in the physical education program in school. Informed consent was obtained from a parent/guardian before beginning the session. Children provided verbal, and when appropriate, written assent.
Children attended a single 45- to 60-minute Challenge test session that took place in a quiet hallway on the Challenge track at Holland Bloorview Kids Rehabilitation Hospital. Both authors were present for testing. The first author (CG) was responsible for conducting the test, and the second author (VW) observed and made notes on the child's performance for later review. CG had been trained to administer the Challenge first by practice scoring a set of Challenge videos of children with CP, and then the 2 authors completed a practice session together before data collection to ensure accurate administration according to the Challenge guidelines.
Each child's Challenge test was videotaped for subsequent review and scoring. The child completed as many Challenge items as possible. Each item was tested just twice to avoid overtaxing the child, although a third attempt was allowed, as is done with the GMFM, if the child appeared to have difficulty understanding the requirements of the skill or wanted to better their previous performance. Children who were able to complete an item in full without error (ie, scored at least “2” on the redesigned response scale template) had their best time/distance scores entered into the cut-point calculations for that item. Conversely, if a child was unable to complete an item, then his/her data were not included in the calculation of the median and lower quartile times. If the time measured during live testing was missing or marked as being in question, times were calculated from the video.
Use of the Children's Videos to Test the Cut-Points
Time/distance cut-points for the revised response scales were determined as described below in the analysis, and these were inserted into each item's predefined response scale. The new response scales were then used to score the children's videos to gain a sense of the functionality of the new response scale with children with TD. Small adjustments were made to the initial time cut-points to adjust for any issues that were evident on review of the distribution of each item's scores (as described in the analysis). The same rater then rescored the videos to determine the effect of these cut-point changes and any other response set revisions on Challenge scores.
For each item, the median and lower quartile times for the time/distance data from the test of the children with TD were calculated for the whole sample as well as for the older (7.5 to 10 years) and younger children (4.0 to <7.5 years). Details of performance on each item, including the number of children who completed each item in full, the median, interquartile range, and range of scores of these children, as well as any comments on the style/quality of performance were determined. For the more advanced motor skills, this meant that the time scores obtained typically represented the older children in the sample.
Several arbitrary rules were made a priori to keep the cut-point determination process consistent. Specifically, when 1 cut-point was required, the median score for the total group was selected. When 2 cut-points were required, the lower quartile for the younger group was selected as the first point, and the lower quartile of the older group was selected as the second point as long as this resulted in reasonable separation in the performance times. Otherwise, the total group scores from quartile and median scores were used as first and second cut-points.
Evaluation of the Challenge Scores
The empirically derived benchmark values were inserted into the time/distance placeholders in the response sets as the new cut-points for the revised response scales. As noted in the Methods section, the response scales for each item were reviewed again including examination of histograms of item score distributions. They were then refined on the basis of our observations from the videos of children's performance as well as from consideration of the distribution of scores across the response scale. This refinement typically involved shifting a cut-point slightly (ie, by 0.2-0.3 seconds depending on the spread of time scores) if this helped the item to attain a more normal distribution. Finally, the videos were rescored using the revised response scales to assess the effect of these changes on the Challenge Total and item scores.
Expansion of the Challenge's Item Set
The 7 new items that were added from the original item reduction survey were jumping jacks, single-leg stance (right and left), tandem stance, dribbles soccer ball within pathway, step-ups (right and left), sideways jumping over line, and ins and outs (Tables 1 and 2). These items had all just missed the 75% importance rating endorsement cut (ie, were in the 58%-74% endorsement score range) during the Challenge's item reduction work by Wilson et al.25 One item (hopscotch) that was in the original Challenge was removed before testing in this study because it was noted by Wilson et al25 to be problematic as most boys had never tried the skill before (ie, a gender-based skill). Finally, testing of the item ascends and descends 2 flights of stairs resulted in concerns about potential safety issues given its new speed component (Table 1). Because the new step-up item covers part of this skill, stair ascent/descent was removed. Thus, the revised Challenge consisted of 25 items that became the focus of the subsequent response scale revisions (Tables 1 and 2).
Response Scale Preparation for the Cut-Points
On review of the focus of the 25 items in the revised Challenge, use of a time component was observed to be appropriate for the response scales of 23 items as a way to capture movement efficiency, whereas a distance element was required for 1 item (long jump). One last item (throw ball to target on wall) already had considerable performance demands in the original Challenge as far as style of throw and target areas, and thus remained as it was without time or distance qualifiers (Table 2).
While working on creating the templates for the revised response scales, it became evident that the use of 2 cut-points within an item would generally enhance the Challenge's ability to capture gains in speed (Table 2). For some of the single-task component items (eg, jumping jacks, running and kicking a ball, and single-leg stance), these 2 cut-points were easily built into the score level “2” and “3” responses. However, for 14 items that were more complex skills, or already had multiple performance components that needed to be considered in the response sets (eg, bouncing a ball down the path, crossovers on a line, and running and picking up a bowling pin and returning), it made more sense to split the top score of “3” into 2 levels (ie, a “3” and a “4”) (Figure 1, version B). The resulting mix of 4- and 5-point level response scales across items creates a slightly uneven item weighting in the measure's total score. However, keeping the number of response options constant across items will no longer be necessary once the Rasch scaling that will align the Challenge with the GMFM-66 is completed.
Participants and Determination of Cut-Points
Thirty-four Children with TD (19 female, 15 male) performed the revised Challenge test. They were 4 to 10 years of age, mean age 7.0 years (SD = 1.8). Their median/quartile values (Table 3) were inserted into the predesigned response sets according to the system described in the analysis. An example of a response set with a cut-point determined from the children with TD's data is shown in Figure 1 (version C).
Scoring of the Challenge With the Benchmarked Response Scales
The Challenge's total mean score for the children in the younger group (n = 16) was 48.9% (46/94 points; SD = 20.4), whereas the mean total score for the older group (n = 14) was 87.2% (82/94 points; SD = 8.2). Bar graphs of each child's total score (plotted according to ascending age) are in shown in Figures 2A and 2B, and median scores for each item are reported in Table 4. The sample size for these calculations was 30 because the Challenge could not be rescored for 2 participants because of damaged video files, and scores for 2 children from the initial test were missing.
Upon review of the item scores, the plan to use the lower quartile scores from the younger and older child groups for the 2 required cut-points was not suitable for 2 items. Specifically, only 3 of the younger children were able to skip rope with enough skill to record completion times for use in benchmarking. Furthermore, for walking with a lunch tray, the lower quartile scores for the 2 groups were separated by less than 1 second. Thus, in both cases, we selected the lower quartile and median scores from the total sample for the cut-points.
Although developmental progression was very evident between the younger and older child groups (ie, higher item scores in the older child group as shown in Table 4 and Figure 2B, as well an indication of age-linked progression of scores in the younger group as shown in Figure 2A), the first revision of the response scales did not seem to capture the differences in advanced motor control that many older children exhibited. In other words, it was clear that some older children completed the items with greater ease, fewer added movements, and overall greater precision than others who demonstrated features such as extraneous arm movements, lack of controlled finish, and erratic cadence. Thus, a decision was made to build movement quality into item response sets as appropriate. It was possible to consider expanding response sets that were still 4-point into 5-point scales to incorporate quality into the upper levels, or for the 5-point scales, to rework the wording to add quality considerations to the upper level of scores. We used the children's Challenge test videos to do a second scoring of the Challenge using these revised response sets to determine the effect on the scores. We hypothesized that both younger and older children's scores would tend to be lower than with the original response scales.
Redevelopment and Testing of the New Response Scales
After rescoring the videos using the revised response scales with the quality compensations considered, the total mean performance scores decreased from 48.9% to 40.6% (43/106) (SD = 22.9) (P < .01) in the younger group and from 87.2% to 81.1% (86/106) (SD = 10.4) (P < .01) in the older group (also illustrated in Figures 2A and 2B on a child–by-child basis). Individual item scores are reported in Table 5. Only 2 of the 11 items with 4-point response scales were not expanded to 5-point scales (ie, jumps forward with 2 feet and skips forward along pathway). This was because on review of children's item scores, it seemed that the existing 4 levels differentiated the performance of children with TD on these items.
The revised Challenge captured a range of gross motor abilities in a group of 4- to 10-year-old children with TD. The 7 new items were feasible to test and successfully increased the range of task difficulty in the test. Indeed, 3 of the 4 most challenging items in the present version were new (ie, skipping with a rope, step-ups, and ins and outs), and the fourth was adapted for the revised Challenge test on the basis of suggestions of Wilson et al25 (ie, for the walking with the tray item, water was added to the cup to add a no-spill challenge). These more difficult skills clearly identify the more skilled and activity-experienced older children as shown by the comparative numbers of children who could complete these items in the younger and older groups, as well as the differences in the group median scores (see Table 4).
Upon review of the first round of scoring revisions, it was clear that the 4-level scoring with its new time cut-points did not accurately separate younger children who could do the skill quickly but lacked movement refinement (eg, had extraneous movements) from those who could perform it both quickly and precisely. For example, many children sacrificed control for speed when performing sideways cross-over steps on a line by excessively using their arms for balance and lacking control at the end of the movement sequence. To capture the quality aspect of skill development, the second round of response option revisions focused on embedding consideration of movement quality as appropriate within the response sets.
Given the recommendations from the Challenge pilot work,25 our initial focus was to build in scoring of movement speed. Upon further reflection, we realized that true skill mastery is evident when an individual can perform a motor skill with both high movement quality and efficiency (process- and product-oriented).5,35 The concept of capturing quality and efficiency has been used successfully in a variety of measures designed to quantify movement performance. For example, both the PANESS and Zurich Neuromotor Assessment report the presence of “associated movements” and time to complete each task. As noted earlier, this dual-criterion scoring is also used in the Berg Balance Scale and Community Balance and Mobility Scale.31–34 From an evaluation and goal-setting perspective, clinicians can note whether movement quality, efficiency, or both are the primary goal(s) for the client.
Conducting this work on Challenge refinement with a group of children with TD provided reality-based estimates of time to complete the items, as well as a wealth of insights on the wide variety of movement capabilities and styles that children with TD exhibit. The performance of children with TD helped us to confirm and refine the lower response levels of the Challenge (scores of “0” and “1” that reflect acquisition of the skill) that had been created by Wilson et al25 without the advantage of viewing children performing the tasks. The younger children in particular gave us insights into developmental precursors or movement quality issues when asked to perform a task that was at the edge of their current abilities. For example, many younger children opted to drop the tennis ball with 1 hand but catch it with 2 hands when they were asked to bounce and catch with 1 hand. This 2-handed catch became a “1” response level for this item.
A number of tasks in the Challenge require separate or leading use of the nondominant arm or leg (eg, bounce and catch tennis ball, standing on one leg, or step-ups). Observations about nondominant side performance of children with TD were readily transferable to movement quality considerations in the Challenge. For example, while young children with TD might be able to hold the standing on one leg position for the full 30 seconds, they likely use compensating balance strategies to do so. These compensations are captured within the “3” level of the item's revised response scale. A similar quality issue may be something that is experienced when a child with CP performs this item using their involved side despite the ability to hold for 30 seconds; however, for an older child with CP, this limitation in performance is likely impairment-related.
A potential limitation of this study is that the cut-points selected were based on a small sample of children with TD. However, we did not intend to determine age-related norms, so the sample size is less of an issue; nor was it assumed that these cut-points from children with TD are the best fit for the scoring of children with CP. Rather, the performance time estimates were intended to be an empirically based starting point for a Rasch scaling study with the Challenge that is underway by the senior author (VW) at the time of writing this article. That study will provide an evaluation of the relative difficulty of each item and permit an opportunity for some cut-point fine-tuning on the basis of Challenge testing with more than 300 children with CP in level I of the GMFCS.
Assessing the child's ability first to complete the functional activity (as is done with the GMFM) and then see what happens when the child's performance is pushed to what might be required to keep up/use the skill in real life has potential to open the discussion with the child and parent about how the child does these skills, and the meaning of being able to do these things in one's life. This could inform goal setting and therapeutic planning if advanced motor skill performance is an area of interest/priority for the child and parent.36
Implications for Future Research
Through this work, we created a novel process to establish empirically based benchmarks for the time cut-points as starting estimates for subsequent Rasch scaling work with the Challenge. When movement quality was introduced into the scoring criteria in the second revision, the performance of the children was further differentiated (both within and between age groups). In the future, the possibility of capturing effort could also be considered because this component of motor performance may influence how functional the skill is when integrated into the demanding context of daily life.
Finally, at the time of writing this article, the authors were conducting funded work with the revised Challenge with children with CP to evaluate inter-/intra-assessor and test-retest reliability with a large multicenter sample of children with CP and perform Rasch scaling. This research also includes a qualitative phase that will provide a look at the value of the Challenge and associated work on advanced motor skill performance to children with CP and their parents.
In summary, the Challenge captures a range of advanced motor skills that may collectively be an important foundation for participation in physical activities in the child's daily life. The skills and response sets are relevant to children with CP functioning at GMFCS level I and address a range of skill development.
1. Hanna SE, Bartlett DJ, Rivard LM, Russell DJ. Reference curves for the gross motor function measure: Percentiles for clinical description and tracking over time among children with cerebral palsy. Phys Ther. 2008;88:596–607.
2. Russell DJ, Avery LM, Rosenbaum PL, Raina PS, Walter SD, Palisano RJ. Improved scaling of the gross motor function measure for children with cerebral palsy: evidence of reliability and validity. Phys Ther. 2000;80:873–885.
3. Alotaibi M, Long T, Kennedy E, Bavishi S. The efficacy of GMFM-88 and GMFM-66 to detect changes in gross motor function in children with cerebral palsy (CP): a literature review. Disabil Rehabil. 2013 June 26. [Epub ahead of print]
4. Beckung E, Carlsson G, Carlsdotter S, Uvebrant P. The natural history of gross motor development in children with cerebral palsy aged 1 to 15 years. Dev Med Child Neurol. 2007;49:751–756.
5. Capio C, Sit C, Abernethy B. Fundamental movement skills testing in children with cerebral palsy. Disabil Rehabil. 2011;33:2519–2528.
6. Capio C, Sit C, Abernethy B, Masters R. Fundamental movement skills and physical activity among children with and without cerebral palsy. Res Dev Disabil. 2012;33:1235–1241.
7. Smits-Engelsman BC, Fiers MJ, Henderson SE, Henderson L. Interrater reliability of the Movement Assessment Battery for Children. Phys Ther. 2008;88:286–294.
8. Brown T, Lalor A. The Movement Assessment Battery for Children–Second Edition (MABC-2): a review and critique. Phys Occup Ther Pediatr. 2009;29:86–103.
9. Holm I, Tveter A, Aulie V, Stuge B. High intra- and inter-rater chance variation of the movement assessment battery for children 2, ageband 2. Res Dev Disabil. 2013;34:795–800.
10. Venetsanou F, Kambas A, Ellinoudis T, Faouras I, Giannakidou D, Kourtessis T. Can the Movement Assessment Battery for Children-Test be the “gold standard” for the motor assessment of children with Developmental Coordination Disorder? Res Dev Disabil. 2011;32:1–10.
11. Deitz JC, Kartin D, Kopp K. Review of the Bruininks-Oseretsky test of motor proficiency (BOT-2). Phys Occup Ther Pediatr. 2007;27:87–102.
12. Jelsma J, Pronk M, Ferguson G, Jelsma-Smit D. The effect of the Nintendo Wii Fit on balance control and gross motor function of children with spastic hemiplegic cerebral palsy. Dev Neurorehabil. 2013;16:27–37.
13. Chen C-L, Hong W-H, Cheng H-YK, Liaw M-Y, Chung C-Y, Chen C-Y. Muscle strength enhancement following home-based virtual cycling training in ambulatory children with cerebral palsy. Res Dev Disabil. 2012;33:1087–1094.
14. Sandlund M, Waterworth ECH. Using motion interactive games to promote physical activity and enhance motor performance in children with cerebral palsy. Dev Neurorehabil. 2011;14:15–21.
15. Ulrich DA. Test of Gross Motor Development. Examiners Manual. 2nd ed. Austin, TX: Pro-Ed; 2000.
16. Evaggelinou C, Tsigilis N, Papa A. Construct validity of the test of gross motor development: a cross-validation approach. Adapt Phys Act Q. 2002;19:483–495.
17. Pang AW, Fong DT. Fundamental motor skill proficiency of Hong Kong children aged 6-9 years. Res Sport Med. 2009:125–144.
18. Denckla M. Revised neurological examination for subtle signs. Psychopharmacol Bull. 1985;21:773–800.
19. Barber A, Srinivasan P, Joel S, Caffo B, Pekar J, Mostofsky S. Motor “dexterity”?: evidence that left hemisphere lateralization of motor circuit connectivity is associated with better motor performance in children. Cereb Cortex. 2012;22:51–59.
20. Rousson V, Gasser T, Caflisch J, Largo R. Reliability of the Zurich Neuromotor Assessment. Clin Neuropsychol. 2008;22:60–72.
21. O'Reilly M, Vollmer B, Vargha-Khadem F, et al. Ophthalmological, cognitive, electrophysiological and MRI assessment of visual processing in preterm children without major neuromotor impairment. Dev Sci. 2010;13(5):692–705.
22. Freedson P, Cureton K, Heath G. Status of field-based fitness testing in children and youth. Prev Med. 2000;31:77.
23. Zhu W, Mahar MT, Welk GJ, Going SB, Cureton KJ. Approaches for development of criterion-referenced standards in health-related youth fitness tests. Am J Prev Med. 2011;41:S68–S76.
24. Safrit MJ. The validity and reliability of fitness tests for children: a review. Pediatr Exerc Sci. 1990;2:9–28.
25. Wilson A, Kavanaugh A, Moher R, et al. Development and pilot testing of the challenge module: a proposed adjunct to the Gross Motor Function Measure for high-functioning children with cerebral palsy. Phys Occup Ther Pediatr. 2011;31(2):135–149.
26. Morris C, Bartlett D. Gross Motor Function Classification System: impact and utility. Dev Med Child Neurol. 2004;46:60–65.
27. Shevell MI, Dagenais L, Hall N, Repacq C. The relationship of cerebral palsy subtype and functional motor impairment: a population-based study. Dev Med Child Neurol. 2009;51:872–877.
28. Scholtes V, Becher J, Comuth A, Dekkers H, van Dijk L, Dallmeijer A. Effectiveness of functional progressive resistance exercise strength training on muscle strength and mobility in children with cerebral palsy: a randomized controlled trial. Dev Med Child Neurol. 2010;52:e107–e113.
29. Taylor N, Dodd K, Baker R, Willoughby K, Thomason P, Graham H. Progressive resistance training and mobility‐related function in young people with cerebral palsy: a randomized controlled trial. Dev Med Child Neurol. 2013;55:806–812.
30. Belvedere SL, de Morton NA. Application of Rasch analysis in health care is increasing and is applied for variable reasons in mobility instruments. J Clin Epidemiol. 2010;63:1287–1297.
31. Conradsson M, Lundin-Olsson L, Lindelöf N, et al. Berg balance scale: intrarater test-retest reliability among older people dependent in activities of daily living and living in residential care facilities. Phys Ther. 2007;87:1155–1163.
32. Donoghue D, Stokes EK, Murphy A, et al. How much change is true change? The minimum detectable change of the Berg Balance Scale in elderly people. J Rehabil Med. 2009;41:343–346.
33. Berg K. Measuring balance in the elderly: preliminary development of an instrument. Physiother Can. 1989;41:304–311.
34. Howe JA, Inness EL, Venturini A, Williams JI, Verrier MC. The Community Balance and Mobility Scale—a balance measure for individuals with traumatic brain injury. Clin Rehabil. 2006;20:885–895.
35. Ericsson KA, Lehmann AC. Expert and exceptional performance: evidence of maximal adaptation to task constraints. Ann Rev Psychol. 1996;47:273–305.
36. Ziebell M, Imms C, Froude E, McCoy A, Galea M. Justification for working on gross motor skills in kids with CP. The relationship between physical performance and self-perception in children with and without cerebral palsy. Aust Occup Ther J. 2009;56:24–32.
activities of daily living; adolescent; cerebral palsy/physiopathology; child; disability evaluation; female; focus groups; humans; male; motor skills; movement disorders/diagnosis; pilot projects; task performance© 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins and the Section on Pediatrics of the American Physical Therapy Association.