SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. What this paper adds
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Aim  New tools that capture hand function in everyday activities and contexts are needed for assessing children with hemiplegic cerebral palsy. This study evaluates a wearable wrist monitor and tests the hypothesis that wrist extension frequency (FreqE) is an appropriate indicator of functional hand use.

Method  Fifteen children (four females, 11 males; age range 6–12y; mean age 10y [SD 2y]) with hemiplegia (seven at level I and eight at level II on the Manual Ability Classification System) participated in the Assisting Hand Assessment (AHA) while wearing the wrist monitor. FreqEs were captured via the wrist monitor and validated using video analysis. Correlations between FreqE and AHA scores were calculated and a multivariate linear regression was conducted to explore other measures of wrist activity.

Results  Wrist extensions observed in video analyses were reliably detected by the wrist monitor (intraclass correlation coefficient, r=0.88; p<0.001) and were strongly correlated with the AHA scores (r=0.93; p<0.001). AHA scores were significantly correlated with FreqE (r=0.80; p=0.001) and the range of wrist extensions/flexions (r=0.70; p=0.008). The multivariate linear regression combining the FreqE and range of wrist extensions/flexions yielded a strong correlation with AHA scores (r=0.84; p=0.0043).

Interpretation  The wearable wrist monitor may offer a convenient, valid alternative to observer reports for functional assessments of the hemiplegic hand in everyday contexts.


Abbreviations
AHA

Assisting Hand Assessment

FreqE

Extension frequency

RangeFE

Range of wrist flexion and extension

What this paper adds

  1. Top of page
  2. Abstract
  3. What this paper adds
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  •  This paper provides supporting evidence that the frequency of wrist extensions is an important indicator of hemiplegic hand function for children with CP.
  •  This article offers validation of a novel wearable wrist activity monitor for home, school, and play environments.
  •  Methods to establish functional assessments based on quantitative measures collected by the wrist activity monitor are assessed.

Developmental disregard refers to the tendency for children with hemiplegic cerebral palsy (CP) to neglect or underutilize their affected limb regardless of its skill or measured capacity in ideal or test conditions.1 A neurological basis for developmental disregard, similar to that of adult poststroke neglect syndromes, has been proposed for children born with a cerebral lesion.1 Thus, consistent practice and use of the hemiplegic limb is considered essential to function and skills advancement. Developmental disregard, therefore, poses a significant challenge in the rehabilitation of children with hemiplegia.1 As reviewed by Gilmore et al.,2 many well-established and important assessment tools (e.g. the Assisting Hand Assessment [AHA]; Quality of Upper Extremity Skills Test) are available to measure functional skill in children with hemiplegia. Few tools, however, quantify actual everyday use of the hemiplegic hand, which can be a critical factor to measure in children with disregard/neglect. At present, everyday hand use and function is largely assessed through self- or parent reports,2 which can be burdensome to complete and are prone to inconsistencies such as recall bias and interrater variations.

Wrist movement is an extremely important component of functional hand use. While trunk, shoulder, and elbow control are essential for positioning the hand to target an object, the wrist is largely responsible for the finer positioning of the hand and affects the power and efficiency of a grasp.3 Typically, individuals will move their wrist into extension to position their hand for grasp. Frequency of wrist extensions is, therefore, proposed as a marker to quantify how frequently the hand is being ‘used’ to grasp. This hypothesis is in line with previous literature which has suggested maximum wrist extension as one potential indicator of functional ability for children with hemiplegia.4 Although the effectiveness of wrist actigraphs (i.e. accelerometer-based devices designed to monitor rest and activity cycles through measures of gross motor activity) for quantifying wheelchair activity, sleep patterns, and/or non-sedentary behaviours in everyday life has been demonstrated,5 it is difficult to distinguish specific movements of the wrist such as extensions/flexions and degrees of bend using accelerometers. Wrist actigraphs are also sensitive to movements (e.g. arm swinging while walking) that are not necessarily related to functional hand use. Previous studies have explored the potential of other sensors, such as electrogoniometers, to measure specifically the magnitude of wrist flexions/extensions during hand activities.6,7 These devices, however, are sized for use with adults, are prone to cross-talk between wrist planes along with offset errors, and are usually not suitable as a portable, wearable device, with the exception of an instrument designed by Ugbolue et al.7 The use of electrogoniometers for home and/or paediatric applications is further limited by their high cost (i.e. over US$10 000 for the sensor and data logger) and size.

Our goals were to design and validate a lightweight, portable, wearable, and inexpensive measurement device that can quantify wrist extension activity in children with hemiplegia in their home, play, and school environments, and to test the hypothesis that frequency of wrist extensions is strongly related to functional hand use. In pursuit of these goals, this paper describes (1) the design and initial validation of the wrist activity monitor, (2) clinical validation of a novel wrist activity monitor for detecting wrist extension frequency, and (3) verification that measures of wrist activity (i.e. extension frequency, range of extension/flexion) correlate well with clinical assessment scores (i.e. the AHA) and may therefore be a useful proxy measure of functional hemiplegic hand use in everyday activities.

Methods

  1. Top of page
  2. Abstract
  3. What this paper adds
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Wrist activity monitor

Device design

A lightweight, low-cost wrist monitor was designed to measure flexions and extensions during hand and arm activities. The device, depicted in Figure 1, consists of a bend sensor (Flexpoint Sensor Systems, Draper, UT, USA) that experiences a change in resistance when its radius of curvature is altered. The flex sensor is insensitive to externally applied pressures. The flex sensor weighs less than 1g and is 80.01 × 6.10 × 0.13mm. Its range of operation for temperature (−40 to 90°C) and humidity (0–100%) is compatible with both northern and southern climates. The flex sensor is connected to a Logomatic v2 Serial SD Datalogger board (Sparkfun Electronics, CO, USA), which records the voltage associated with changes in the flex sensor’s resistance as it bends. Data are stored on a 1-GB micro-secure digital card and are extracted from the device through a USB interface or by removing the secure digital card. The sampling rate of the datalogger is adjustable and was specified at 100Hz. This sampling frequency is sufficient to capture wrist flexions and extensions, which do not generally exceed 12.5Hz for activities of daily living and usually remain below 5Hz.8 The wrist monitor is limited to 12 hours of continuous use by the charge capacity of the rechargeable lithium polymer battery. The flex sensor is positioned on a one-size lightweight Lycra glove in a channel that allows the sensor to shift proximally and distally as the wrist extends and flexes in order to prevent it from buckling. The connecting wires are directed along the forearm to a water-resistant casing that is secured by an armband and houses the datalogger and battery, as depicted in Figure 2. The overall dimensions of the armband unit are 80 × 62 × 21mm, and its weight is 60g. The total cost of the wrist monitor was approximately US$200.

image

Figure 1.  Circuit and connection diagram of the device.

Download figure to PowerPoint

image

Figure 2.  Wrist monitor composed of (a) a wearable glove with (b) an embedded flex sensor that is controlled and powered by (c) a data logger and lithium battery contained in an armband.

Download figure to PowerPoint

Device validation protocol

The operation of the wrist monitor was initially validated with a group of 11 able-bodied adults (five males; six females), who performed a predefined series of wrist movements of varying magnitudes and speeds. The participants’ movements were captured by the wrist monitor and a seven-camera Vicon MX motion capture system (Vicon, Los Angeles, CA, USA), a well-established standard in biomechanics and movement analysis.9–11 Vicon has an accuracy of 63μm (SD 5) in determining marker position and high levels of repeatability (coefficient of multiple correlation between 0.87 and 0.96) for upper limb movements.12,13 The device accurately captured 98% (SD 2%) of wrist flexions and extensions. The minimum angle of flexion and/or extension that can be detected is 15° from neutral. Figure 3 depicts a sample comparison of angular deviations obtained via the Vicon and the wrist monitor device.

image

Figure 3.  Wrist monitor and Vicon-recorded angular deviations for a series of wrist flexion/extension cycles. Note: the biodirectional sensor demonstrated greater sensitivity in one direction of bend than for the other.

Download figure to PowerPoint

This study was approved by Holland Bloorview Kids Rehabilitation Hospital’s Science and Ethics Review Board. All parents and children provided written and informed consent.

Clinical assessment

Participants

Fifteen children (four females, 11 males) aged six to 12 years (mean age 10y [SD 2y]) with a diagnosis of hemiplegic CP and who were receiving occupational therapy at Holland Bloorview were recruited to participate in this study. Children with a range of wrist extension abilities were recruited (seven at level I and eight at level II on the Manual Ability Classification System).14 All 15 participants were assessed at level I on the Gross Motor Function Classification System.15

Clinical validation protocol

This study was designed as a blinded analytical observational study to evaluate the correlation between wrist activity (quantified by a novel wrist monitor) and functional hand use (measured via an established standard, the AHA). Participants attended a single session at Holland Bloorview. Maximum active wrist extension and flexion were measured using a goniometer. While wearing the measurement glove, participants were video recorded as they engaged in the AHA. These videos were used (1) to complete the AHA scoring and (2) to count the number of wrist extensions observed. Previous studies have successfully used video analyses to quantify and qualitatively assess upper limb movements in children with CP.16,17

The AHA is a well-established and validated assessment tool for measuring bilateral hand function in children with hemiplegia during bimanual play that separates and spreads individuals along an ability continuum to a high degree.18 It has excellent intraclass correlation coefficients (ICCs) for interrater, intrarater, and test–retest reliability at 0.98, 0.99, and 0.99 respectively.18–20 The appropriate version of the AHA for this age group was the School Kids AHA, which engages children in a game with a series of bimanual tasks (e.g. cutting paper, opening a box).20 The AHA was conducted by a trained and certified occupational therapist (LF). The AHA consists of six subsections (i.e. grasp–release, fine motor adjustment, arm use, coordination, pace, overall usage). The scores for each section are combined and then scaled to give the overall score for the child. The raw score for the AHA ranges from 22 (limited hand function) to 88 (high level of hand function) points, which is then scaled to a 100-point scale.18

Data analysis

SPSS version 17 (SPSS Inc., Chicago, IL, USA) was used for all statistical analyses. The normality of all data was assessed with the Shapiro–Wilk test. Wrist extensions past the participant’s relaxed or neutral wrist position were counted for all AHA game play activities by two independent and blinded observers (JH, EB). Absolute observer agreement was calculated via the ICC in using the two-way mixed model and single measure reliability, wherein the unit of analysis is the individual rating in counts per minute. Averaged frequency counts were then correlated with the AHA scores for each child via Pearson’s correlation coefficient with a significance level of p=0.05.

To validate the accuracy of the wrist activity monitor for use with children with hemiplegic CP, frequency counts for wrist extensions observed in the video analysis were compared with those detected by the wrist monitor. To determine the wrist extension frequency (FreqE, counts/min) detected by the wrist monitor, the raw data were first passed through a fifth-order Butterworth low-pass filter with a cut-off frequency of 5Hz, as appropriate for typical wrist activity.8 A peak detection algorithm was constructed and applied to the filtered data. Specifically, the algorithm assessed the data in 3-minute windows and detected peaks (i.e. extensions) that exceeded 90% of the standard deviation from the mean for more than 0.4 seconds. These thresholds along with the time interval of 3 minutes were determined through an optimization process. The total number of peaks for all windows was summed to determine the FreqE. Signal processing was conducted in Matlab 2009a (Mathworks, Natick, MA, USA). Consistency between the video analysis and the wrist monitor was quantified via the ICC (two-way mixed model, single measure reliability). The ICC is frequently used to assess the degree of consistency between two different methods for measuring the same quantitative value when both methods use the same unit system.21

Lastly, a quantitative model was developed to investigate the relationship between features extracted from the wrist monitor data and the AHA assessment scores. In addition to FreqE, a second characteristic of the wrist activity was determined from the device data: the change in voltage associated with the range of wrist flexions and extensions (RangeFE, mV). RangeFE was calculated and averaged for 35-second windows, which proved to be an appropriate interval to separate different activities from each other and from rest during AHA game play. Pearson’s correlation coefficients were calculated to examine the bivariate correlations between RangeFE and the AHA score, and FreqE with the AHA score with a significance level of p=0.05. A multivariate linear regression model was then constructed to predict the dependent variable (i.e. the AHA score) from features of the wrist monitor data. Features with a significant univariate correlation coefficient were included in the model. Possible interactions between input variables were also explored. Based on the Pearson’s correlation coefficient, the model correlation was defined as strong (r>0.6), moderate (0.4<r≤0.6), or weak (r≤0.4).22

Results

  1. Top of page
  2. Abstract
  3. What this paper adds
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Of the 15 children, wrist monitor data collected from two participants with particularly small hands were not viable owing to poor glove fit and unacceptable sensor positioning (i.e. the length of the sensor extended from the wrist to the metacarpophalangeal joints, making the device sensitive to finger as well as wrist activity). The following section presents clinical validation of the wrist activity monitor (based on data collected from the remaining 13 children) and video analyses showing the relationship between the frequency of wrist extensions and the AHA scores (based on all 15 children). The mean AHA score for the participants was 55 (SD 23) with a range from 12 to 100. All data outputs (i.e. video-derived data, device-derived data, and AHA scores) were normally distributed (p>0.05).

Clinical validation of the wrist activity monitor

Interrater reliability for video analyses of wrist extension frequency counts had an ICC of 0.93 (95% confidence interval [CI] 0.88–0.95; p<0.001). Wrist extension frequencies as determined from (1) video analysis or (2) data extracted from the wrist monitor had an ICC of 0.88 (95% CI 0.66–0.96; p<0.001). This indicates an agreement between the wrist monitor and observer reports suggesting the concurrent validity of the device. The monitor demonstrated some sensitivity to radial and ulnar deviations identified by real-time and video qualitative observations. This sensitivity was responsible for disparities between wrist extension frequencies observed via video analyses and those extracted from the wrist monitor.

Relationship between wrist activity and functional hand use

A significant strong correlation between the frequency of wrist extensions (observed via video analysis) and functional hand use, as indicated by the AHA score, was observed (r=0.93; p<0.001). This analysis was based on data collected from all 15 children. There was no change in this association when the two children whose wrist monitor data were not viable were excluded from this analysis (r=0.93; p<0.001).

The FreqE, as measured by the wrist monitor, was also significantly and strongly correlated with the AHA score (r=0.80; p=0.001), as was the RangeFE (r=0.70; p=0.008). The combination of RangeFE and FreqE in a multivariate linear regression model yielded a strong correlation with the AHA score (r=0.84; p=0.0043). The equation describing this relationship was:

  • image

This model had no outliers, leverage effects, or strong indications of collinearity (variance inflation factor=1.636).

A significant interaction effect was observed between RangeFE and FreqE, as depicted in Figure 4. This suggests that when RangeFE is high (i.e. greater wrist mobility), functional scores on the AHA depend less on how often the wrist is actively extended (FreqE). Conversely, for children with a limited range of movement (low RangeFE), the FreqE is a far more important predictor of the AHA functional scores. The regression equation incorporating the interaction term was

  • image
image

Figure 4.  Predicted Assisting Hand Assessment (AHA) scores for a range of extension frequencies (FreqE; extensions per minute) for selected range of wrist flexions and extensions (RangeFE; mV). Noted is the significant interaction effect between FreqE and RangeFE.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. What this paper adds
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Key findings

This study examined the clinical value of a wearable wrist monitor to capture functional use of a child’s hemiplegic hand during play activities. Three key findings were established: (1) FreqE can be measured using an inexpensive, wearable, and portable wrist activity monitor that yields results consistent with observer reports; (2) FreqE is a suitable proxy measure to assess functional use of the hemiplegic hand as indicated by its strong association with the AHA; (3) a combination of two wrist activity measures (i.e. frequency of wrist extensions, range of flexions/extensions) more closely reflects functional hand use than either individual measure on its own.

To our knowledge, this is the first study to measure and demonstrate definitively the strong relationship between the frequency of wrist extensions and functional hand use in children with hemiplegia. Our findings support clinical acumen in addition to a number of studies suggesting the importance of maximizing wrist extension in rehabilitation strategies.23–25 This study is well aligned with recent findings by Murgia et al.,26 who observed that the active range of movement of the wrist exhibited by adults who had suffered a distal radius fracture was smaller than that of adults with no hand impairments during a typical daily activity (i.e. page turning).

Study implications

This is the first clinical demonstration of a portable and wearable wrist activity monitor suitable for use by children in everyday contexts and activities. This addition to the assessment tool kit for children may augment the ability to assess functional hand use, much in the way that Holter monitors supplement in-clinic cardiac assessments. This study substantiates the high potential of the novel wrist activity monitor, which offers the following advantages: (1) it appears to yield clinically valid measures in line with occupational therapists’ observations of functional hand use; (2) it provides quantitative measures that can be tracked continuously over extended periods of time and in different contexts; and (3) it is portable, low-cost, and wearable, and may be appropriate for home or school use, as needed, to measure the activity and participation components of the World Health Organization’s International Classification of Functioning, Disability and Health.27 Future investigations are needed to assess the ability of the wrist monitor to detect wrist movements in the home or school environment.

Thus, the wrist activity monitor promises to become a useful tool for assessment of the hemiplegic hand. It provides a unique opportunity to obtain insight into functional hand use in natural contexts/activities that is not typically captured through clinical assessments. Additionally, the monitor could be incorporated into virtual reality therapy tools and/or games to encourage active extension of the wrist and to discourage developmental disregard.

Future work

Despite these promising results, future development and research is required to address a number of current limitations. As with electrogoniometers, cross-talk between the wrist planes reduced the wrist monitor’s accuracy and explained discrepancies between observed and measured wrist extension counts. The wrist monitor did not perform as well in the clinical setting as it did during initial validation testing conducted with able-bodied participants (as documented in the Method section). This is probably because of (1) the less structured activities performed in the clinical study; (2) the tendency for children with CP to have a neutral hand position that involves some degree of ulnar deviation – this caused the sensor to be bent in the coronal plane, which made measurements of wrist flexion and extension less accurate (sagittal plane); and (3) motion capture was used as the criterion standard metric in validation testing, whereas observer reports were relied upon in the clinical study – this was to minimize participant discomfort and deviations from standard care.

A number of design revisions are planned based on experiences gained through this study. First, multiple and/or adjustable glove sizes to accommodate a larger range of hand sizes are needed. Alternatively, preliminary tests suggest that it may be possible to incorporate a smaller sensor into a wrist band, which would probably increase aesthetic appeal and comfort. Future research will evaluate different lengths of flex sensors to determine the optimal trade-off between angular sensitivity and user comfort. A wrist band may also enable the sensor to be more suitably positioned on the wrist and better accommodate neutral hand positions with varying degrees of radial deviation. Associated electronics (e.g. the datalogger, battery) may also be miniaturized and incorporated directly into the wrist band. Lastly, alternative sensor designs may be explored to optimize sensitivity and accuracy. The current flex sensor was more sensitive to bending in one direction than in the other (as depicted in Fig. 3). In this study, the high sensitivity direction was consistently aligned to correspond with wrist extension, which was the primary measure of interest. An improved sensor with a high level of sensitivity in both directions would be preferable for subsequent iterations. Future studies will include a larger sample size and will be performed outside of clinical settings to evaluate the device in context and to assess user compliance and satisfaction.

Conclusion

  1. Top of page
  2. Abstract
  3. What this paper adds
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

A low-cost, wearable, and portable device was successfully designed and validated for detection of wrist extensions in children with hemiplegic CP. Quantitative characteristics of these captured wrist extensions were linked to and can be used to assess functional use of the hemiplegic hand. The wrist monitor may have future applications for home-based and contextually relevant assessments of hemiplegic hand function that are difficult to observe in the clinic environment.

Acknowledgements

  1. Top of page
  2. Abstract
  3. What this paper adds
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The authors acknowledge the contributions of Dr John Runciman, Adam Smith, Amanda Fleury, Cameron Farrow, and Joanna Weber in the initial design and testing of the wrist monitor, and Jomy Varghese for his assistance in data collection. This study was funded by the Natural Sciences and Engineering Research Council of Canada and Bloorview Research Institute.

References

  1. Top of page
  2. Abstract
  3. What this paper adds
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • 1
    Sutcliffe TL, Logan WJ, Fehlings DL. Paediatric constraint-induced movement therapy is associated with increased contralateral cortical activity on functional magnetic resonance imaging. J Child Neurol 2009; 24: 12305.
  • 2
    Gilmore R, Sakzewski L, Boyd R. Upper limb activity measures for 5- to 16-year-old children with congenital hemiplegia: a systematic review. Dev Med Child Neurol 2010; 52: 1421.
  • 3
    Utley A, Sugden D. Interlimb coupling children with hemiplegic cerebral palsy during reaching and grasping at speed. Dev Med Child Neurol 1998; 40: 396404.
  • 4
    Vaz DV, Mancini MC, Fonseca ST, Rocha DS, Pertence AE. Muscle stiffness and strength and their relation to hand function in children with hemiplegic cerebral palsy. Dev Med Child Neurol 2006; 48: 72833.
  • 5
    Warms C. Physical activity measurement in persons with chronic and disabling conditions methods, strategies, and issues. Fam Community Health 2006; 29 (1 Suppl.): 78S88S.
  • 6
    Johnson PW, Jonsson P, Hagberg M. Comparison of measurement accuracy between two wrist goniometer systems during pronation and supination. J Electromyogr Kinesiol 2002; 12: 41320.
  • 7
    Ugbolue UC, Nicol AC, Maclean J. Development and Validation of an Electrogoniometric Wrist Alignment Device. Proc Inst Mech Eng 2008; 222: 63744.
  • 8
    Mann KA, Werner FW, Palmer AK. Frequency spectrum analysis of wrist motion for activities of daily living. J Orthop Res 1989; 7: 3046.
  • 9
    Fitoussi F, Diop A, Maurel N, Lassel EM, Pennecot GF. Kinematic analysis of the upper limb: a useful tool in children with cerebral palsy. J Pediatr Orthop B 2006; 15: 24756.
  • 10
    Henmi S, Yonenobu K, Masatomi T, Oda K. A biomechanical study of activities of daily living using neck and upper limbs with an optical three-dimensional motion analysis system. Mod Rheumatol 2006; 16: 28993.
  • 11
    Metcalf CD, Notley SV, Chappell PH, Burridge JH, Yule VT. Validation and application of a computational model for wrist and hand movements using surface markers. IEEE Trans Biomed Eng 2008; 55: 1199209.
  • 12
    Windolf M, Gotzen N, Morlock M. Systematic accuracy and precision analysis of video motion capturing systems-exemplified on the Vicon-460 system. J Biomech 2008; 41: 277680.
  • 13
    Mackey AH, Walt SE, Lobb GA, Stott NS. Reliability of upper and lower limb three-dimensional kinematics in children with hemiplegia. Gait Posture 2005; 22: 19.
  • 14
    Eliasson AC, Rösblad B, Beckung E, Arner M, Öhrwall AM, Rosenbaum P. The manual ability classification system (MACS) for children with cerebral palsy: scale development and evidence of validity and reliability. Dev Med Child Neurol 2006; 48: 54954.
  • 15
    Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol 1997; 39: 21423.
  • 16
    Bensmail D, Robertson J, Fermanian C, Roby-Brami A. Botulinum toxin to treat upper-limb spasticity in hemiparetic patients: grasp strategies and kinematics of reach-to-grasp movements. Neurorehabil Neural Repair 2010; 24: 14151.
  • 17
    Sorsdahl AB, Moe-Nilssen R, Strand LI. Observer reliability of the Gross Motor Performance Measure and the Quality of Upper Extremity Skills Test, based on video recordings. Dev Med Child Neurol 2008; 50: 14651.
  • 18
    Krumlinde-Sundholm L, Holmefur M, Kottorp A, Eliasson A-C. The assisting hand assessment: current evidence of validity, reliability, and responsiveness to change. Dev Med Child Neurol 2007; 29: 25964.
  • 19
    Holmefur M, Krumlinde-Sundholm L, Eliasson A-C. Interrater and Intrarater Reliability of the Assisting Hand Assessment. Am J Occup Ther 2007; 61: 7984.
  • 20
    Holmefur M, Aarts P, Hoare B, Krumlinde-Sundholm L. Test-retest and alternated forms reliability of the Assisting Hand Assessment. J Rehabil Med 2009; 41: 88691.
  • 21
    Daniel WW. Biostatistics: A Foundation for Analysis in Health Science, 9th edn. New Jersey: John Wiley and Sons, Inc., 2009.
  • 22
    Shin JC, Yoo JH, Jung TH, Goo HR. Comparison of lower extremity motor score parameters for patients with motor incomplete spinal cord injury using gait parameters. Spinal Cord 2011; 49: 52933.
  • 23
    Wright PA, Granat MH. Improvement in hand function and wrist range of motion following electrical stimulation of wrist extensor muscles in an adult with cerebral palsy. Clin Rehabil 2000; 14: 2446.
  • 24
    Rösblad B, Andersson G, Pettersson K. Effects of botulinum toxin type A and a programme of functional activity to improve manual ability in children and adolescents with cerebral palsy. Scand J Plast Reconstr Surg Hand Surg 2007; 41: 2508.
  • 25
    Barbosa AP, Vaz DV, Gontijo AP, Fonseca ST, Mancini MC. Therapeutic effects of electrical stimulation on manual function of children with cerebral palsy: evaluation of two cases. Disabil Rehabil 2008; 30: 7238.
  • 26
    Murgia A, Kyberd P, Barnhill T. The use of kinematic and parametric information to highlight lack of movement and compensation in the upper extremities during activities of daily living. Gait Posture 2010; 31: 3006.
  • 27
    World Health Organization. International Classification of Functioning, Disability and Health (ICF). Geneva: World Health Organization, 2001.