Australian Occupational Therapy Journal (2014)
doi: 10.1111/1440-1630.12150
Research Article
The Neurological Hand Deformity Classification for children with cerebral palsy Melissa Georgiades,1 Catherine Elliott,2 Judith Wilton,3,4 Eve Blair,5 Marie Blackmore4 and Simon Garbellini6 1 Department of Occupational Therapy, Edith Cowan University, Joondalup, 2Department of Paediatric Rehabilitation, Princess Margaret Hospital for Children, 3School of Paediatrics and Child Health, The University of Western Australia, Subiaco, 4The Centre for Cerebral Palsy, Mount Lawley, 5Telethon Kids Institute, The University of Western Australia, and 6Department of Paediatric Rehabilitation, Princess Margaret Hospital for Children, Subiaco, Western Australia, Australia
Background/aim: The purpose of this study was to evaluate the reliability of the Neurological Hand Deformity Classification and use it to describe changes in hand deformity over time in children with cerebral palsy. Methods: We identified 114 video clips of 26 children with cerebral palsy, aged 1–18 years (mean = 8.4, SD = 4.2), performing upper-limb tasks at multiple time points (n = 3–8) at least 6 months apart. Using the Neurological Hand Deformity Classification, three observers classified hand deformity in the video clips. Inter- and intra-observer reliabilities were estimated using Fleiss and Cohen’s kappa (j) and the temporal changes in classification of hand deformity were investigated. Results: Inter- and intra-observer reliability respectively were j = 0.87 and j = 0.91. Hand deformity was identified in all children at all time points, even before the age of 2 years. Ten children did not change hand classification, wrist flexion increased in eight, and eight showed changes from wrist flexion to extension or vice versa. Conclusions: The Neurological Hand Deformity Classification is a reliable tool to classify hand deformity in children with cerebral palsy. For more than one-third of children hand deformity classification did not change. For the remaining children, two patterns of change in hand Melissa Georgiades BSc(OT) Hons; Honours Student. Catherine Elliott PhD; Director of Research. Judith Wilton MSc, PGradDipHthSc, BAppSc(OT); Clinical Hand Specialist. Eve Blair PhD; Epidemiologist. Marie Blackmore PhD; Research Coordinator. Simon Garbellini MHlthServMgt, BSc(OT); Senior Occupational Therapist. Correspondence: Simon Garbellini, Department of Paediatric Rehabilitation, Princess Margaret Hospital for Children, Level 2, 37-39 Hay Street, Subiaco, WA 6008, Australia. Email: [email protected] Accepted for publication 14 July 2014. © 2014 Occupational Therapy Australia
deformity over time were identified. It is recommended that children with cerebral palsy involving their upper limbs be monitored regularly. Significance of the study: This is the first study to document longitudinal changes in hand deformity in children with cerebral palsy. KEY WORDS cerebral palsy, classification, hand deformity, occupational therapy, rater reliability.
Introduction Cerebral palsy (CP) is a multi-faceted condition resulting from a non-progressive cerebral lesion or abnormality, acquired prenatally, during labour and delivery, or postnatally and occurring in 2–2.5 individuals per 1000 live births (Reddihough & Collins, 2003). Heterogeneity in clinical presentation occurs due to the varied location and severity of the neurological lesion or abnormality, subsequent secondary impairments, response to treatments, and the effects of environment, occupation and growth (Brown, 1985; Damiano, Quinlivan, Owen, Shaffrey & Abel, 2001; Graham, 2004; Wilton, 2003). Clinical presentation is dynamic, owing to very rapid postnatal brain development (Brown, 1985) and the interaction of neurophysiological and biomechanical factors. These factors can alter bone and muscle properties, impair muscle growth, change biomechanical functioning, and cause contracture and deformity (Boyd & Graham, 1997; Flett, 2003; Graham & Selber, 2003; Priori, Cogiamanian & Mrakic-Sposta, 2006). Deformity is defined as a ‘structural deviation from the normal shape or size, resulting in disfigurement’ (Dirckx & Cadle, 2008, p. 414). The natural history of musculoskeletal deformity in children with CP tends towards progressive increase in deformity (Graham, 2004). Dynamic deformities, which can be modified by active or passive motion, may become fixed as children
2 become older (Graham, 2004). Hand deformity can occur in children with CP before 2 years of age (Park, Sim & Rha, 2011). Functional deficits of the hand have been shown to be strongly associated with the degree of deformity (Arner, Eliasson, Nicklasson, Sommerstein & Hagglund, 2008; Law et al., 2008; Park et al.). Impairments of hand function disrupt a child’s interaction with the environment and subsequent physical, cognitive, social and emotional development (Case-Smith, 2010; Pfeifer, Pacciulio, dos Santos, dos Santos & Stagnitti, 2011). Occupational therapists, who care for children with CP direct significant time and resources to the management of hand deformity and impairment of hand function (Wilton, 2003). Early intervention is considered essential to minimise future musculoskeletal complications (such as subluxation, contracture and hand deformity), promote and maintain hand function, and enhance independence and quality of life (Graham, 2004; Venkateswaran & Shevell, 2008). Strategies such as activity-based intervention, stretching, strengthening, upper limb orthoses, serial casting, pharmacological therapy (including botulinum toxin type A (BoNT-A)), and surgery are aimed at preserving the muscle-to-bone length ratio thereby preventing contracture and hand deformity (Boyd & Graham, 1997; Lin, 2004; Wilton, 2003; Zancolli, 2003). A recent review identified 10 published systems for classifying hand deformity (McConnell, Johnston & Kerr, 2011), 5 of which considered the thumb exclusively. All were designed by surgeons for pre- and post-operative classification, but only two (House, Gwathmey & Fidler, 1981; Zancolli, 2003) have any published evidence of reliability. The House classification is used to describe deformities of the thumb without consideration of wrist position. Zancolli describes a classification of voluntary grasp and release patterns in the spastic hand. However, it is limited to wrist flexion patterns of deformity and does not include extension deformities, nor does it encourage the observer to consider the contributions of extrinsic and intrinsic finger and thumb musculature when describing deformity patterns. Limitations in the use of surgical classifications to facilitate appropriate therapy treatment planning were identified by an occupational therapist who, using the basis of the Zancolli and House surgical classifications (House et al. 1981; Zancolli, 2003), developed the Classification of Deformities of the Hand in the Presence of Neurological Dysfunction (Wilton, 2003, 2004). This classification provides a framework for analysis of deformity and reasoning for therapy intervention. It also includes patterns where there is no voluntary grasp and release. This version of hand deformity classification consisted of five levels (Types I–V). Levels I–III presented patterns of increasing wrist flexion deformity with an increasing loss of active wrist extension. Level IV presented a pattern of wrist extension deformity. Level V presented a fisted hand pattern with either a wrist flexion or extension defor© 2014 Occupational Therapy Australia
M. GEORGIADES ET AL.
mity with minimal active movement evident (Wilton, 2003). To promote ease of use for therapists, the Classification of Deformities of the Hand in the Presence of Neurological Dysfunction was modified in 2012 by the developer and renamed the Neurological Hand Deformity Classification (NHDC) (see Table 1). To create a distinction between flexion and extension deformities the NHDC includes four separate categories for flexion deformities (F1 to F4) and two for extension deformities of the hand (E1 and E2). F1 refers to a hand position in which the wrist flexion is 20° or less. F2 and F3 refer to hand positions in which wrist flexion exceeds 20°. The difference is that in F2 the child is able to extend the wrist actively, whereas in F3 no active wrist extension is observed. In F4, both the wrist and fingers are flexed and there is no active wrist and finger movement. E1 and E2 both refer to hand positions in which the wrist is held in an extended position. In E1, the child is able to move the fingers, whereas in E2, no active finger movement is observed. The first row of Table 1 (entitled ‘Type of deformity’) is used to classify hand deformity. Once the classification is made, the remaining rows in Table 1 are designed to help the therapist identify which muscles and joints in the wrist and hand may be responsible for the deformity and functional deficits and thus plan individualised assessment and intervention. The NHDC was developed to facilitate analysis of the anatomical and biomechanical components of hand deformity in the presence of neurological dysfunction, thus providing a framework for therapists to consider intervention options (Wilton, 2003). The NHDC appears to be the only classification system for the extension deformities of the wrist that are observed in clinical practice and reported by occupational therapists. In this study, retrospective video analysis was used to investigate changes in the NHDC during the childhood years. Because all participants were receiving services for upper limb management from community or tertiary service providers, this study does not describe the natural progression of hand deformity unaffected by therapeutic intervention. Not all interventions that might have affected changes in hand deformity during the course of data collection were documented. However, the timing of BoNT-A and surgical interventions were recorded, and changes that occurred subsequent to these interventions are described. The first aim of the present study was to determine the inter-observer and intra-observer reliability of the NHDC. Musculoskeletal impairments such as contracture and deformity develop during childhood as children with CP age and grow (Boyd & Graham, 1997; DeLuca, 1996; Graham, 2004), but longitudinal changes in hand deformity in children with CP have not been systematically investigated. Therefore, the second aim of the present study was to describe the changes in hand deformity in children with CP over time as classified using the NHDC.
Not always present CMC adduction
Hyperextension of PIP joints
FCU AP
Wrist extensors Extrinsic and intrinsic finger flexors and extensors Nil wrist/fingers Thumb web space
Associated thumb deformity
Associated finger patterns
Spasticity primarily located in these muscles
Muscles not affected by spasticity
Contracture
F1. Wrist flexion 20°, active extension
Flexion type deformities
TABLE 1: Neurological Hand Deformity Classification (NHDC)
FCU, FCR, PL FDP and FDS, Thumb web space
Intrinsic finger musculature
FCU, FCR, PL FDP and FDS, AP
CMC adduction MCP and IP vary OR ? Hyperextension of PIP joints
F3. Wrist flexion >20°, extension absent
Potential severe deformity of the wrist fingers and thumb
Combined spasticity extrinsic and intrinsic musculature of the fingers and thumb Wrist musculature opposite to wrist position
Flexion
CMC adduction, MCP and IP flexion
F4. Wrist and finger flexion, no active movement
MCP flexion adduction of fingers and thumb. Palmar skin and fascial shortening. Hand hygiene critical
ECRL and ECRB, ECU contributes to ulnar deviation. Interossei, AP FPB FDP and FDS
CMC adduction MCP flexion IP neutral Skin and fascial shortening MCP flexion IP extension
E1. Wrist extension, finger movement powered by intrinsic muscle action
Extension type deformities
Wrist extensors and dorsal wrist capsule Palmar skin and fascial contracture Potential severe deformity of the wrist, fingers MCP joints and thumb
Flexion, adduction at MCP joints, flexion of IP joints associated with wrist extension position Combined spasticity extrinsic and intrinsic musculature of the fingers and thumb Wrist flexor musculature
CMC adduction MCP and IP flexion
E2. Wrist extension, no active movement
THE NHDC FOR CHILDREN WITH CP
3
© 2014 Occupational Therapy Australia
4
© 2014 Occupational Therapy Australia
Adapted by Wilton (2004). CMC, carpometacarpal; MCP, metacarpophalangeal; PIP, proximal interphalangeal; IP, interphalangeal; FCU, Flexor Carpi Ulnaris; AP, Adductor Pollicis; FDS, Flexor Digitorum Superficialis; FDP, Flexor Digitorum Profundus; 1st DI, First Dorsal Interosseous; PL, palmaris longus; FCR, Flexor Carpi Radialis; ECRL, Extensor Carpi Radialis Longus; ECRB, Extensor Carpi Radialis Brevis; ECU, Extensor Carpi Ulnaris; FPB, Flexor Pollicis Brevis.
Fingers and thumb disadvantaged by wrist extension – finger flexion/ extension possible if wrist in neutral and thumb abducted Palm orientation in grasp, wrist control during finger flexion Thumb disadvantaged effective opposition Nil ?thumb adduction Functional deficit
Flexion type deformities
TABLE 1: (Continued)
Approach and grasp compromised by wrist position. Grasp effective if object placed in hand. No manipulation
No function
Extension type deformities
No function
M. GEORGIADES ET AL.
Method Procedure We identified 450 video clips of children with CP performing upper limb gross and fine motor tasks during occupational therapy sessions. They had been recorded in Western Australia between 1982 and 2012 either at The Centre for Cerebral Palsy or at Princess Margaret Hospital for Children. Clips meeting the following criteria were selected for this study: (i) they recorded the active or passive approach of the hand to an object; (ii) movements of the wrist, fingers and thumb were clearly visible; (iii) at least three such clips taken at least 6 months apart were available for each participant; and (iv) the participant (or their parent/guardian, if they were under 18 years) consented to this use of their video clip. The type of activity recorded, camera position and format used to record the footage were not among the criteria for selection of video clips. Where more than one section of a video met selection criteria, the section in which the actions of wrist, finger and thumb were most clearly visible was selected. One hundred and fourteen of the 450 video clips met these selection criteria. They had been recorded between 1996 and 2012. Video clips ranged from 1 to 5 minutes (mean = 3 minutes 20 seconds) In these video clips, 61% of children were performing standard upper limb assessments (Melbourne Assessment of Unilateral Upper Limb Function (43%), Assisting Hand Assessment (8%), Quality of Upper Extremity Skills Test (10%)) and 39% were performing functional assessments, including clinical observation of free play. Whenever the clip was taken from the Melbourne Assessment of Unilateral Upper Limb Function, the same items from the assessment were always selected. Whenever clips were taken from the Assisting Hand Assessment, Quality of Upper Extremity Skills Test, and functional assessments (particularly those shot before the Melbourne Assessment of Unilateral Upper Limb Function and Assisting Hand Assessment were available), they were chosen to be as similar as possible with regard to size and shape of object and child’s direction of movement and position. All children were seated independently or with adult support. Clips were digitised using the software program ‘Cyberlink PowerDirector’ Version 8 and a video capture card. Audiovisual data were organised in chronological order for each participant, the date being recorded to the day for 42/114 video clips (36.8%), 2 weeks for 64 video clips (55.6%) and 6 months for 8 video clips (7.2%). Where the date was not recorded on the video itself, records of medical and/or therapy appointments were used to assign a date to the video. In cases where the exact dates were doubtful, successive videos were included only if there were clearly at least 6 months between them.
5
THE NHDC FOR CHILDREN WITH CP
To measure inter-observer reliability of the NHDC, two occupational therapists experienced in the management of neurological hand dysfunction (J. W. and S. G.) and, following instruction in applying the NHDC (twoone-to-one 30-minute practice sessions), a final year occupational therapy student (M. G.), independently viewed each video clip in the order in which they occurred in the folders and classified the children’s level of hand deformity using the NHDC. J. W. and S. G. had already been using the NHDC and its predecessor, the Classification of Deformities of the Hand in the Presence of Neurological Dysfunction, for at least 10 years. Blinded to the classification she had previously assigned, M. G. also reviewed and reclassified the video clips 3 weeks after the initial classification to determine intra-observer reliability. Observers generally classified the children’s hand deformities after a single viewing of each clip, but occasionally needed to re-watch certain sections of the clips once or twice in order to check or confirm their classifications. Information was obtained from each participant’s medical records concerning birth history, CP topographical classification, comorbidities, any upper limb BoNTA or surgical treatments, and levels of Gross Motor Function Classification System (GMFCS) and Manual Ability Classification System (MACS) recorded in the last entry in the medical record. Ethics approval was gained from the Human Research Ethics Committees of Princess Margaret Hospital for Children and Edith Cowan University. Work conformed to the provisions of the Declaration of Helsinki and informed consent was obtained from all participants or their parents/guardians.
Statistical analysis IBM SPSS Statistics Version 19.0 was employed for statistical analysis. Cohen’s kappa (j) (Cohen, 1960) was used to determine intra-observer reliability. Fleiss’s kappa (j) was used to determine inter-observer reliability. A macro for Fleiss’s j was developed from original syntax by David Nichols and modified by Geoffrey Hammond, to enable analysis to be run for three observers in SPSS. Reliability was reported in accordance with Landis and Koch values (Landis & Koch, 1977). For each participant, interventions and NHDC classifications were plotted against age and NHDC classifications at first and final time points were cross-tabulated.
Results Participants Twenty-six children (11 males and 15 females) were included. There were between three and eight video clips per child, covering 2–13 years. All children in the study had upper limb spasticity and seven had a mixed presentation, including dystonia. Twenty-two children
were classified as GMFCS levels I or II, two as GMFCS level III, and one each at GMFCS level IV and V. Twenty were classified as MACS level II and two were classified in each of the levels I, III and IV. Twenty children had CP with unilateral upper limb involvement and six had CP with bilateral upper limb involvement. Ten children had video clips taken at three time points, five at four time points, six at five time points, one at six time points, three at seven time points and one at eight time points, totalling 114 video clips taken between the ages of 1 and 18 years (mean = 8.4, SD = 4.2).
Inter-observer and intra-observer reliability The inter-observer reliability of the Neurological Hand Deformity Classification was j = 0.87 and the intraobserver reliability was j = 0.91. Most of the disagreements were between F1 and F2 (28%) and between F2 and F3 (39%) classifications. Disagreements were compared between ages 0–6, 7–12, and 13–18 years using chi square, and no significant effect for age was found (P = 0.60).
Change in hand deformity with age All children were classified as having some level of hand deformity in each of the video clips, even the four youngest children who were below 2 years of age (three subsequently classified at MACS level II and one at level III). Table 2 shows the numbers of children with each combination of initial and final NHDC classification. Fourteen children (54%) had the same classification at the final assessment as at the first. For four of these children (15%), the classification changed in the intervening time points but reverted to their original classification by the final time point. The number classified as F3 increased from three children to seven children during the course of the study period. No child was classified as F4 at any time point.
TABLE 2: Cross-tabulation of NHDC classifications at first and final time points Final classification Initial Classification
F1
F2
F3
E1
E2
Total
F1 F2 F3 E1 E2 Total
2 1 0 1 0 4
4 6 0 0 0 10
1 2 3 1 0 7
0 0 0 3 0 3
1 1 0 0 0 2
8 10 3 5 0 26
Shading indicates no change between first and final classifications. © 2014 Occupational Therapy Australia
6 Table 3 shows the NHDC at each time point for each study participant, arranged according to the types of interventions: (i) BoNT-A and surgery, (ii) BoNT-A only, (iii) surgery only, and (iv) no BoNT-A or surgery. The changes in hand deformity in this observational study were grouped into three categories. Firstly, 8 children, one with an extension classification and seven with a flexion classification at first video recording, changed from an extended or less flexed wrist posture to a more flexed wrist posture over time. These children are designated as ‘↑ flexion’ in Table 3. Secondly, 10 children did not change classification at any time point, 9 of whom had a flexion classification. These children are designated as ‘no change’ in Table 3. Thirdly, 8 children did not follow a consistent sequence in change of classification level, and either changed from a flexion classification to an extension classification or vice versa (e.g. F1 to E2, E1 to F3). These children, 4 of whom had dystonia, are designated as ‘variable’ in Table 3. Classification of hand deformity of children at older ages generally showed more flexed (F2 and F3) classifications. Out of 14 children with initial video clips at 6 years of age or younger (within an Early Intervention Program), 11 were classified as F1 or E1, 2 were classified as F2 and 1 as F3. Out of 12 children with initial video clips over 6 years of age (within a School Aged Intervention Program), 11 were classified as F2/F3 or E2 classification. Seven children presented with dystonia as indicated in Table 3. Three children were in the ‘no change’ group and four in the ‘variable’ group following classification. None were in the ‘↑ flexion’ group. The Occupational Therapists classifying hand deformity for the children with dystonia did not view any fluctuation in level of hand deformity from the clips to conflict with the classification assigned to the individual child at that point in time. The variability in classification, particularly fluctuations between flexion and extension deformities for those with dystonia in the ‘variable’ group, was evident only between one time point and the next. Nineteen participants received upper limb BoNT-A and/or hand surgery. Their ages at intervention and NHDC classifications are shown in Table 3. Seven children did not receive hand surgery or upper limb BoNTA during the period of the video recordings. For 5 of these 7 NHDC classification did not change. However, none of these had a video clip taken before the age of 7 years. Age at first video for the remaining two participants was 1 year and 10 years.
Conclusion The first aim of the study was to determine the interobserver and intra-observer reliability of the NHDC. The NHDC demonstrated ‘almost perfect’ (Landis & Koch, 1977) inter-observer and intra-observer reliability in classifying hand deformity in children with CP. © 2014 Occupational Therapy Australia
M. GEORGIADES ET AL.
However, two or three viewings of the video clips were sometimes required when classifying children with fluctuating tone, dystonic posturing, and well-established compensatory movement patterns. The almost perfect inter-observer reliability between novice and experienced clinicians under the conditions of this retrospective study suggests that the NHDC is easy to use and does not require special training. Whether the NHDC would be equally reliable for the in vivo classification of hand deformity depends on how much within-subject variability exists in the way children in this population approach an object. Investigation of this question is recommended for future research. Given the potential for fluctuation in classification of hand deformity in children with dystonia, caution must be taken when comparing hand deformity classification over time. The NHDC is capable of classifying hand deformity in children with dystonia at a particular point in time. In this situation the NHDC may be used by the observer as a guide for clinical consideration of the factors contributing to the deforming forces and the structures involved. The second aim of the study was to describe observed changes in hand deformity over time in children with CP. These changes do not represent natural history because the children in this study were receiving various interventions from community and tertiary service providers for upper limb spasticity. Our results demonstrate that hand deformity may occur before the age of 2 years, although it must noted that children attending a specialist hand clinic before the age of 2 years are unlikely to be representative of all children with CP involving upper limb impairment. Although all children had hand deformity at all ages, children aged 6 years or younger had less wrist flexion deformity than older children. The study did not provide evidence for a consistent pattern of changes in hand deformity for all children: 10 children did not change hand classification, wrist flexion increased in 8, and 8 showed changes from wrist flexion to extension or vice versa. Children who moved from an extension classification to a flexion classification or from an F1 to F2 or F2 to F3 classification on the NHDC lost active wrist extension range of motion. Due to the retrospective nature of the study and inconsistent recording of passive range of motion in the children’s medical and therapy records, it is impossible to report the loss of passive range of motion for the children in the study. The study illustrates the difficulty in predicting future hand deformity from a child’s current classification. For example, an F1 classification may be taken to indicate that no immediate intervention is required. However, in this study, three quarters of the children either moved from less than 20 degrees (F1) to greater than 20 degrees of wrist flexion (F2/F3) or changed to a wrist extension classification over time. It is important,
7
THE NHDC FOR CHILDREN WITH CP
TABLE 3: NHDC classifications and upper limb interventions for individual participants
ID
GMFCS/ MACS
Ages at classification
NHDC levels at each age
Category
Ages at BoNT-A
F2,F2,F2,F2 F2,F2,F3,F3,F3,F3 F2,F2,F1,F2,F2
No change ↑ flexion Variable
11,12,17 12,12,12,14,14 11,11,13,13
9,10,12,13,15,16,17 6,10,11
F2,F2,F3,F2,F2,F2,F1 E1,E1,E1
Variable No change
5,12,13,14,15,16,16,17,17 3,3,5,8,8,9,10,10,11,11
3,7,9,11,12,14,15,16 1,2,4,4,5,7,8
E1,F2,F2,F3,F2,F2,E1,E1 E1,E1,E1,E1,F2,F2,F1
Variable ↑ flexion
10,11,14 4,4,5,5,6
3,4,8,9 3,4,5,7,8 1,3,4,5,5,9,10 3,4,4,8,9 1,2,4,4,6 1,3,4,6,8 7,10,13 5,9,11 3,4,5 7,9,10,13
F1,F1,F1,F1 F1,F2,F2,F2,F2 F1,F1,F1,F2,F2,F3,F2 F1,F1,F2,F2,F2 F1,F1,E1,F2,F2 F1,F1,F2,F3,F3 F2,F2,F2 F2,F3,F2 F3,F3,F3 E1,E1,F3,E1
No change ↑ flexion ↑ flexion ↑ flexion ↑ flexion ↑ flexion No change Variable No change Variable
E1,E2,F3
Variable
4,7,7,8 5,5,6,7,8,8 4,5,6,6,7,7,8,8 4,5,5,6,6,7,8,8 3,3,5,5,6 3,4 12 11 2,3,3,4,4,5 5,6,8,8,9,10,10,11,11,11, 12,12,13 4,4,5,5,6,6
F2,F2,F2,F2,E2
Variable
F1,F1,F1 F1,E1,E2 F2,F2,F2 F2,F2,F2,F2 F2,F3,F3 F3,F3,F3 F3,F3,F3
No change Variable No change No change ↑ flexion No change No change
BoNT-A and surgery 1 I/II 9,10,14,17 2 III/III 9,11,11,14,14,15 3 I/II 7,9,10,12,15 4 5
II/II II/II
6 I/II 7 I/II BoNT-A only 8‡ V/IV 9 II/II 10 II/II 11 II/II 12 II/II 13 II/III 14 I/II 15 I/II 16 I/II 17‡ I/II
18‡ I/II 2,4,6 Surgery only 19‡ IV/IV 5,9,11,12,17 No BoNT-A or surgery 20‡ I/I 12,13,14 21‡ I/II 1,2,3 22 I/I 7,10,10 23 III/II 8,11,13,13 24 I/II 10,12,13 25‡ I/II 7,10,10 26 I/II 10,12,13
Ages at surgery
Types of surgery†
14 14 11 14 16 6 10 14 7
1,3,4 2,4 4 2 1,5 5,10 7 1,4,5 4,6,9
13
1,6,8
†Surgery type: 1 – tendon transfer – Flexor Carpi Ulnaris (FCU) to Extensor Carpi Radialis Brevis or Longus; 2 – tendon transfer – FCU to supination transfer (transfer to dorsal radius insertion to periosteum); 3 – FDSR to Abductor Pollicis Brevis; 4 – musculotendinous release of forearm pronator; 5 – musculotendinous release of extrinsic flexor muscles; 6 – musculotendinous release of thumb intrinsic (Adductor Pollicis); 7 – intrinsic releases of fingers; 8 – Ulnar Neurectomy; 9 – thumb MCP joint stabilisation; 10 – Abductor Pollicis Longus Tenodesis surgery. ‡Dystonia. GMFCS – Gross Motor Function Classification Scale; MACS – Manual Ability Classification Scale.
therefore, to monitor changes in the hands of children with CP over time, regardless of classification. Consistent methods of recording and maintaining the records of change are essential, particularly when children’s occupations change from early intervention to schoolaged therapy programmes or if there is a change in primary therapist or service provider.
The number of children with an F3 classification more than doubled from three at first classification to seven at final classification. These children have no active wrist extension. Intervention to maintain passive wrist and finger flexor muscle length is paramount for these children to prevent flexion contractures. © 2014 Occupational Therapy Australia
8 The study had several limitations owing to its retrospective design. The study population was a sample of convenience, because only children with appropriate video clips were included. Not all treatment records could be collected and video clips were not at regular intervals. Given the heterogeneity of the children in this study, subgroup analyses would have been desirable, but the small sample size of 26 precluded this. It was not possible to blind observers to the children’s identity nor to the chronological order of their clips because the observers could sometimes see part of the child on the screen and could infer the chronological order from the size of the child’s hand. However, the observers were blind to the timing of any interventions. They viewed and classified each video independently, without reference to how the child had been classified in other video clips. In all video clips, the child was actively or passively reaching for an object on a horizontal surface, but the task and type of object and camera position were not standardised. It is recommended that the effects of these factors on hand position be examined in future research. However, the NHDC is designed to be much less affected by task or object than are comprehensive assessments of hand function, provided the task given to the child involves approach of the hand to an object. This is because the biomechanical limitations in the child’s hand constrain the range of possible movements of the wrist and fingers during approach to an object, and the same constraints will operate whatever the task. Typically, in clinical practice, the NHDC would guide a comprehensive functional assessment with a range of tasks. As this is the first study to be published on the NHDC, there is great scope for further investigation of this tool. It is recommended that a prospective longitudinal study be undertaken to examine changes in hand deformity systematically over time, the reliability of NHDC in vivo, and whether the NHDC classification correlates with functional impact of hand deformity. The NHDC should be used in the context of comprehensive hand assessments. Videos should be taken at regular intervals to identify if a child’s hand deformity classification changes over time. In clinical settings, the NHDC, together with a comprehensive upper limb assessment and accurate documentation, would enable clinicians to identify changes in the level of hand deformity in children with CP over time. Ideally, video data should be routinely collected during the assessment process to support effective documentation and communication, and provide evidence to guide clinical practice. In conclusion, this is the first study to document longitudinal changes in hand deformity in children with CP. It shows that the NHDC is reliable in classifying hand deformity by video analysis, is easy to use, and does not require special training. The study also demonstrates that children with CP can present with hand © 2014 Occupational Therapy Australia
M. GEORGIADES ET AL.
deformities from an early age, and therefore need to be monitored regularly.
Acknowledgements The authors declare no conflicts of interest. Financial support was received from Edith Cowan University by Melissa Georgiades as part of her honours study. We wish to convey our gratitude to the children and families involved in this study and to The Centre for Cerebral Palsy and Princess Margaret Hospital for Children for access to the children’s video data and treatment histories. Thanks also to Mr Geoffrey Hammond from the Telethon Institute for Child Research for the syntax modification, to Dr Sonya Girdler from Curtin University for her academic support, to Dr Siobhan Reid from University of Western Australia for access to video data and to Mrs Stephanie Mann from The Centre for Cerebral Palsy for her assistance in data collection.
References Arner, M., Eliasson, A., Nicklasson, S., Sommerstein, G. & Hagglund, G. (2008). Hand function in cerebral palsy. Report of 367 children in a population-based longitudinal health care program. Journal of Hand Surgery, 33A, 1337– 1347. Boyd, R. N. & Graham, H. K. (1997). Botulinum toxin A in the management of children with cerebral palsy: Indications and outcomes. European Journal of Neurology, 4, S15– S21. Brown, K. (1985). Positional deformity in children with cerebral palsy. Physiotherapy Theory and Practice, 1, 37–41. Case-Smith, J. (2010). Development of childhood occupations. In: J. Case-Smith & J. C. O’Brien (Eds.), Occupational therapy for children (6th ed., pp. 56–83). Missouri: Mosby Elsevier. Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37–46. Damiano, D. L., Quinlivan, J., Owen, B. F., Shaffrey, M. & Abel, M. F. (2001). Spasticity versus strength in cerebral palsy: relationships among involuntary resistance, voluntary torque, and motor function. European Journal of Neurology, 8, 40–49. DeLuca, P. A. (1996). The musculoskeletal management of children with cerebral palsy. Pediatric Clinics of North America, 43, 1135–1150. Dirckx, J. H. & Cadle, K. J. (Eds.). (2008). Stedman’s medical dictionary for the health professions and nursing (6th ed.). London: Wolters Kluwer Health/Lippincott Williams & Wilkins. Flett, P. J. (2003). Rehabilitation of spasticity and related problems in childhood cerebral palsy. Journal of Pediatrics and Child Health, 39, 6–14. Graham, H. K. (2004). Mechanisms of deformity. In: D. Scrutton, D. L. Damiano & M. Mayston (Eds.), Management of the motor disorders of children with cerebral palsy (2nd ed., pp. 105–129). London: Mac Keith Press.
THE NHDC FOR CHILDREN WITH CP
Graham, H. K. & Selber, P. (2003). Musculoskeletal aspects of cerebral palsy. Journal of Bone and Joint Surgery, British, 85B, 157–166. House, J. H., Gwathmey, F. W. & Fidler, M. O. (1981). A dynamic approach to the thumb-in palm deformity in cerebral palsy. The Journal of Bone and Joint Surgery. American Volume, 63, 216–225. Landis, J. R. & Koch, G. G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33, 159–174. Law, K., Lee, E. Y., Fung, B. K., Yan, L. S., Gudushauri, P., Wang, K. W. et al. (2008). Evaluation of deformity and hand function in cerebral palsy patients. Journal of Orthopaedic Surgery and Research, 3, doi:10.1186/1749-799X-3-52. Lin, J.-P. (2004). The assessment and management of hypertonus in cerebral palsy: A physiological atlas (‘road map’). In: D. Scrutton, D. L. Damiano & M. Mayston (Eds.), Management of the motor disorders of children with cerebral palsy (2nd ed., pp. 85–104). London: Mac Keith Press. McConnell, K., Johnston, L. & Kerr, C. (2011). Upper limb function and deformity in cerebral palsy: A review of classification systems. Developmental Medicine and Child Neurology, 53, 799–805. Park, E. S., Sim, E. G. & Rha, D. W. (2011). Effect of upper limb deformities on gross motor and upper limb func-
9 tions in children with spastic cerebral palsy. Research in Developmental Disabilities, 32, 2389–2397. Pfeifer, L. I., Pacciulio, A. M., dos Santos, A. C., dos Santos, J. L. & Stagnitti, K. E. (2011). Pretend play of children with cerebral palsy. Physical and Occupational Therapy in Pediatrics, 31, 390–402. Priori, A., Cogiamanian, F. & Mrakic-Sposta, S. (2006). Pathophysiology of spasticity. Neurological Sciences, 27, s307–s309. Reddihough, D. S. & Collins, K. J. (2003). The epidemiology and causes of cerebral palsy. Australian Journal of Physiotherapy, 49, 7–12. Venkateswaran, S. & Shevell, M. I. (2008). Comorbidities and clinical determinants of outcome in children with spastic quadriplegic cerebral palsy. Developmental Medicine and Child Neurology, 50, 216–222. Wilton, J. C. (2003). Casting, splinting, and physical and occupational therapy of hand deformity and dysfunction in cerebral palsy. Hand Clinics, 19, 573–584. Wilton, J. C. (2004). Splinting and casting in the presence of neurological dysfunction. Hand splinting: Principles of design and fabrication (pp. 168–197). Perth, WA: Success Print. Zancolli, E. A. (2003). Surgical management of the hand in infantile spastic hemiplegia. Hand Clinics, 19, 609–629.
© 2014 Occupational Therapy Australia
doi: 10.1111/1440-1630.12150
Research Article
The Neurological Hand Deformity Classification for children with cerebral palsy Melissa Georgiades,1 Catherine Elliott,2 Judith Wilton,3,4 Eve Blair,5 Marie Blackmore4 and Simon Garbellini6 1 Department of Occupational Therapy, Edith Cowan University, Joondalup, 2Department of Paediatric Rehabilitation, Princess Margaret Hospital for Children, 3School of Paediatrics and Child Health, The University of Western Australia, Subiaco, 4The Centre for Cerebral Palsy, Mount Lawley, 5Telethon Kids Institute, The University of Western Australia, and 6Department of Paediatric Rehabilitation, Princess Margaret Hospital for Children, Subiaco, Western Australia, Australia
Background/aim: The purpose of this study was to evaluate the reliability of the Neurological Hand Deformity Classification and use it to describe changes in hand deformity over time in children with cerebral palsy. Methods: We identified 114 video clips of 26 children with cerebral palsy, aged 1–18 years (mean = 8.4, SD = 4.2), performing upper-limb tasks at multiple time points (n = 3–8) at least 6 months apart. Using the Neurological Hand Deformity Classification, three observers classified hand deformity in the video clips. Inter- and intra-observer reliabilities were estimated using Fleiss and Cohen’s kappa (j) and the temporal changes in classification of hand deformity were investigated. Results: Inter- and intra-observer reliability respectively were j = 0.87 and j = 0.91. Hand deformity was identified in all children at all time points, even before the age of 2 years. Ten children did not change hand classification, wrist flexion increased in eight, and eight showed changes from wrist flexion to extension or vice versa. Conclusions: The Neurological Hand Deformity Classification is a reliable tool to classify hand deformity in children with cerebral palsy. For more than one-third of children hand deformity classification did not change. For the remaining children, two patterns of change in hand Melissa Georgiades BSc(OT) Hons; Honours Student. Catherine Elliott PhD; Director of Research. Judith Wilton MSc, PGradDipHthSc, BAppSc(OT); Clinical Hand Specialist. Eve Blair PhD; Epidemiologist. Marie Blackmore PhD; Research Coordinator. Simon Garbellini MHlthServMgt, BSc(OT); Senior Occupational Therapist. Correspondence: Simon Garbellini, Department of Paediatric Rehabilitation, Princess Margaret Hospital for Children, Level 2, 37-39 Hay Street, Subiaco, WA 6008, Australia. Email: [email protected] Accepted for publication 14 July 2014. © 2014 Occupational Therapy Australia
deformity over time were identified. It is recommended that children with cerebral palsy involving their upper limbs be monitored regularly. Significance of the study: This is the first study to document longitudinal changes in hand deformity in children with cerebral palsy. KEY WORDS cerebral palsy, classification, hand deformity, occupational therapy, rater reliability.
Introduction Cerebral palsy (CP) is a multi-faceted condition resulting from a non-progressive cerebral lesion or abnormality, acquired prenatally, during labour and delivery, or postnatally and occurring in 2–2.5 individuals per 1000 live births (Reddihough & Collins, 2003). Heterogeneity in clinical presentation occurs due to the varied location and severity of the neurological lesion or abnormality, subsequent secondary impairments, response to treatments, and the effects of environment, occupation and growth (Brown, 1985; Damiano, Quinlivan, Owen, Shaffrey & Abel, 2001; Graham, 2004; Wilton, 2003). Clinical presentation is dynamic, owing to very rapid postnatal brain development (Brown, 1985) and the interaction of neurophysiological and biomechanical factors. These factors can alter bone and muscle properties, impair muscle growth, change biomechanical functioning, and cause contracture and deformity (Boyd & Graham, 1997; Flett, 2003; Graham & Selber, 2003; Priori, Cogiamanian & Mrakic-Sposta, 2006). Deformity is defined as a ‘structural deviation from the normal shape or size, resulting in disfigurement’ (Dirckx & Cadle, 2008, p. 414). The natural history of musculoskeletal deformity in children with CP tends towards progressive increase in deformity (Graham, 2004). Dynamic deformities, which can be modified by active or passive motion, may become fixed as children
2 become older (Graham, 2004). Hand deformity can occur in children with CP before 2 years of age (Park, Sim & Rha, 2011). Functional deficits of the hand have been shown to be strongly associated with the degree of deformity (Arner, Eliasson, Nicklasson, Sommerstein & Hagglund, 2008; Law et al., 2008; Park et al.). Impairments of hand function disrupt a child’s interaction with the environment and subsequent physical, cognitive, social and emotional development (Case-Smith, 2010; Pfeifer, Pacciulio, dos Santos, dos Santos & Stagnitti, 2011). Occupational therapists, who care for children with CP direct significant time and resources to the management of hand deformity and impairment of hand function (Wilton, 2003). Early intervention is considered essential to minimise future musculoskeletal complications (such as subluxation, contracture and hand deformity), promote and maintain hand function, and enhance independence and quality of life (Graham, 2004; Venkateswaran & Shevell, 2008). Strategies such as activity-based intervention, stretching, strengthening, upper limb orthoses, serial casting, pharmacological therapy (including botulinum toxin type A (BoNT-A)), and surgery are aimed at preserving the muscle-to-bone length ratio thereby preventing contracture and hand deformity (Boyd & Graham, 1997; Lin, 2004; Wilton, 2003; Zancolli, 2003). A recent review identified 10 published systems for classifying hand deformity (McConnell, Johnston & Kerr, 2011), 5 of which considered the thumb exclusively. All were designed by surgeons for pre- and post-operative classification, but only two (House, Gwathmey & Fidler, 1981; Zancolli, 2003) have any published evidence of reliability. The House classification is used to describe deformities of the thumb without consideration of wrist position. Zancolli describes a classification of voluntary grasp and release patterns in the spastic hand. However, it is limited to wrist flexion patterns of deformity and does not include extension deformities, nor does it encourage the observer to consider the contributions of extrinsic and intrinsic finger and thumb musculature when describing deformity patterns. Limitations in the use of surgical classifications to facilitate appropriate therapy treatment planning were identified by an occupational therapist who, using the basis of the Zancolli and House surgical classifications (House et al. 1981; Zancolli, 2003), developed the Classification of Deformities of the Hand in the Presence of Neurological Dysfunction (Wilton, 2003, 2004). This classification provides a framework for analysis of deformity and reasoning for therapy intervention. It also includes patterns where there is no voluntary grasp and release. This version of hand deformity classification consisted of five levels (Types I–V). Levels I–III presented patterns of increasing wrist flexion deformity with an increasing loss of active wrist extension. Level IV presented a pattern of wrist extension deformity. Level V presented a fisted hand pattern with either a wrist flexion or extension defor© 2014 Occupational Therapy Australia
M. GEORGIADES ET AL.
mity with minimal active movement evident (Wilton, 2003). To promote ease of use for therapists, the Classification of Deformities of the Hand in the Presence of Neurological Dysfunction was modified in 2012 by the developer and renamed the Neurological Hand Deformity Classification (NHDC) (see Table 1). To create a distinction between flexion and extension deformities the NHDC includes four separate categories for flexion deformities (F1 to F4) and two for extension deformities of the hand (E1 and E2). F1 refers to a hand position in which the wrist flexion is 20° or less. F2 and F3 refer to hand positions in which wrist flexion exceeds 20°. The difference is that in F2 the child is able to extend the wrist actively, whereas in F3 no active wrist extension is observed. In F4, both the wrist and fingers are flexed and there is no active wrist and finger movement. E1 and E2 both refer to hand positions in which the wrist is held in an extended position. In E1, the child is able to move the fingers, whereas in E2, no active finger movement is observed. The first row of Table 1 (entitled ‘Type of deformity’) is used to classify hand deformity. Once the classification is made, the remaining rows in Table 1 are designed to help the therapist identify which muscles and joints in the wrist and hand may be responsible for the deformity and functional deficits and thus plan individualised assessment and intervention. The NHDC was developed to facilitate analysis of the anatomical and biomechanical components of hand deformity in the presence of neurological dysfunction, thus providing a framework for therapists to consider intervention options (Wilton, 2003). The NHDC appears to be the only classification system for the extension deformities of the wrist that are observed in clinical practice and reported by occupational therapists. In this study, retrospective video analysis was used to investigate changes in the NHDC during the childhood years. Because all participants were receiving services for upper limb management from community or tertiary service providers, this study does not describe the natural progression of hand deformity unaffected by therapeutic intervention. Not all interventions that might have affected changes in hand deformity during the course of data collection were documented. However, the timing of BoNT-A and surgical interventions were recorded, and changes that occurred subsequent to these interventions are described. The first aim of the present study was to determine the inter-observer and intra-observer reliability of the NHDC. Musculoskeletal impairments such as contracture and deformity develop during childhood as children with CP age and grow (Boyd & Graham, 1997; DeLuca, 1996; Graham, 2004), but longitudinal changes in hand deformity in children with CP have not been systematically investigated. Therefore, the second aim of the present study was to describe the changes in hand deformity in children with CP over time as classified using the NHDC.
Not always present CMC adduction
Hyperextension of PIP joints
FCU AP
Wrist extensors Extrinsic and intrinsic finger flexors and extensors Nil wrist/fingers Thumb web space
Associated thumb deformity
Associated finger patterns
Spasticity primarily located in these muscles
Muscles not affected by spasticity
Contracture
F1. Wrist flexion 20°, active extension
Flexion type deformities
TABLE 1: Neurological Hand Deformity Classification (NHDC)
FCU, FCR, PL FDP and FDS, Thumb web space
Intrinsic finger musculature
FCU, FCR, PL FDP and FDS, AP
CMC adduction MCP and IP vary OR ? Hyperextension of PIP joints
F3. Wrist flexion >20°, extension absent
Potential severe deformity of the wrist fingers and thumb
Combined spasticity extrinsic and intrinsic musculature of the fingers and thumb Wrist musculature opposite to wrist position
Flexion
CMC adduction, MCP and IP flexion
F4. Wrist and finger flexion, no active movement
MCP flexion adduction of fingers and thumb. Palmar skin and fascial shortening. Hand hygiene critical
ECRL and ECRB, ECU contributes to ulnar deviation. Interossei, AP FPB FDP and FDS
CMC adduction MCP flexion IP neutral Skin and fascial shortening MCP flexion IP extension
E1. Wrist extension, finger movement powered by intrinsic muscle action
Extension type deformities
Wrist extensors and dorsal wrist capsule Palmar skin and fascial contracture Potential severe deformity of the wrist, fingers MCP joints and thumb
Flexion, adduction at MCP joints, flexion of IP joints associated with wrist extension position Combined spasticity extrinsic and intrinsic musculature of the fingers and thumb Wrist flexor musculature
CMC adduction MCP and IP flexion
E2. Wrist extension, no active movement
THE NHDC FOR CHILDREN WITH CP
3
© 2014 Occupational Therapy Australia
4
© 2014 Occupational Therapy Australia
Adapted by Wilton (2004). CMC, carpometacarpal; MCP, metacarpophalangeal; PIP, proximal interphalangeal; IP, interphalangeal; FCU, Flexor Carpi Ulnaris; AP, Adductor Pollicis; FDS, Flexor Digitorum Superficialis; FDP, Flexor Digitorum Profundus; 1st DI, First Dorsal Interosseous; PL, palmaris longus; FCR, Flexor Carpi Radialis; ECRL, Extensor Carpi Radialis Longus; ECRB, Extensor Carpi Radialis Brevis; ECU, Extensor Carpi Ulnaris; FPB, Flexor Pollicis Brevis.
Fingers and thumb disadvantaged by wrist extension – finger flexion/ extension possible if wrist in neutral and thumb abducted Palm orientation in grasp, wrist control during finger flexion Thumb disadvantaged effective opposition Nil ?thumb adduction Functional deficit
Flexion type deformities
TABLE 1: (Continued)
Approach and grasp compromised by wrist position. Grasp effective if object placed in hand. No manipulation
No function
Extension type deformities
No function
M. GEORGIADES ET AL.
Method Procedure We identified 450 video clips of children with CP performing upper limb gross and fine motor tasks during occupational therapy sessions. They had been recorded in Western Australia between 1982 and 2012 either at The Centre for Cerebral Palsy or at Princess Margaret Hospital for Children. Clips meeting the following criteria were selected for this study: (i) they recorded the active or passive approach of the hand to an object; (ii) movements of the wrist, fingers and thumb were clearly visible; (iii) at least three such clips taken at least 6 months apart were available for each participant; and (iv) the participant (or their parent/guardian, if they were under 18 years) consented to this use of their video clip. The type of activity recorded, camera position and format used to record the footage were not among the criteria for selection of video clips. Where more than one section of a video met selection criteria, the section in which the actions of wrist, finger and thumb were most clearly visible was selected. One hundred and fourteen of the 450 video clips met these selection criteria. They had been recorded between 1996 and 2012. Video clips ranged from 1 to 5 minutes (mean = 3 minutes 20 seconds) In these video clips, 61% of children were performing standard upper limb assessments (Melbourne Assessment of Unilateral Upper Limb Function (43%), Assisting Hand Assessment (8%), Quality of Upper Extremity Skills Test (10%)) and 39% were performing functional assessments, including clinical observation of free play. Whenever the clip was taken from the Melbourne Assessment of Unilateral Upper Limb Function, the same items from the assessment were always selected. Whenever clips were taken from the Assisting Hand Assessment, Quality of Upper Extremity Skills Test, and functional assessments (particularly those shot before the Melbourne Assessment of Unilateral Upper Limb Function and Assisting Hand Assessment were available), they were chosen to be as similar as possible with regard to size and shape of object and child’s direction of movement and position. All children were seated independently or with adult support. Clips were digitised using the software program ‘Cyberlink PowerDirector’ Version 8 and a video capture card. Audiovisual data were organised in chronological order for each participant, the date being recorded to the day for 42/114 video clips (36.8%), 2 weeks for 64 video clips (55.6%) and 6 months for 8 video clips (7.2%). Where the date was not recorded on the video itself, records of medical and/or therapy appointments were used to assign a date to the video. In cases where the exact dates were doubtful, successive videos were included only if there were clearly at least 6 months between them.
5
THE NHDC FOR CHILDREN WITH CP
To measure inter-observer reliability of the NHDC, two occupational therapists experienced in the management of neurological hand dysfunction (J. W. and S. G.) and, following instruction in applying the NHDC (twoone-to-one 30-minute practice sessions), a final year occupational therapy student (M. G.), independently viewed each video clip in the order in which they occurred in the folders and classified the children’s level of hand deformity using the NHDC. J. W. and S. G. had already been using the NHDC and its predecessor, the Classification of Deformities of the Hand in the Presence of Neurological Dysfunction, for at least 10 years. Blinded to the classification she had previously assigned, M. G. also reviewed and reclassified the video clips 3 weeks after the initial classification to determine intra-observer reliability. Observers generally classified the children’s hand deformities after a single viewing of each clip, but occasionally needed to re-watch certain sections of the clips once or twice in order to check or confirm their classifications. Information was obtained from each participant’s medical records concerning birth history, CP topographical classification, comorbidities, any upper limb BoNTA or surgical treatments, and levels of Gross Motor Function Classification System (GMFCS) and Manual Ability Classification System (MACS) recorded in the last entry in the medical record. Ethics approval was gained from the Human Research Ethics Committees of Princess Margaret Hospital for Children and Edith Cowan University. Work conformed to the provisions of the Declaration of Helsinki and informed consent was obtained from all participants or their parents/guardians.
Statistical analysis IBM SPSS Statistics Version 19.0 was employed for statistical analysis. Cohen’s kappa (j) (Cohen, 1960) was used to determine intra-observer reliability. Fleiss’s kappa (j) was used to determine inter-observer reliability. A macro for Fleiss’s j was developed from original syntax by David Nichols and modified by Geoffrey Hammond, to enable analysis to be run for three observers in SPSS. Reliability was reported in accordance with Landis and Koch values (Landis & Koch, 1977). For each participant, interventions and NHDC classifications were plotted against age and NHDC classifications at first and final time points were cross-tabulated.
Results Participants Twenty-six children (11 males and 15 females) were included. There were between three and eight video clips per child, covering 2–13 years. All children in the study had upper limb spasticity and seven had a mixed presentation, including dystonia. Twenty-two children
were classified as GMFCS levels I or II, two as GMFCS level III, and one each at GMFCS level IV and V. Twenty were classified as MACS level II and two were classified in each of the levels I, III and IV. Twenty children had CP with unilateral upper limb involvement and six had CP with bilateral upper limb involvement. Ten children had video clips taken at three time points, five at four time points, six at five time points, one at six time points, three at seven time points and one at eight time points, totalling 114 video clips taken between the ages of 1 and 18 years (mean = 8.4, SD = 4.2).
Inter-observer and intra-observer reliability The inter-observer reliability of the Neurological Hand Deformity Classification was j = 0.87 and the intraobserver reliability was j = 0.91. Most of the disagreements were between F1 and F2 (28%) and between F2 and F3 (39%) classifications. Disagreements were compared between ages 0–6, 7–12, and 13–18 years using chi square, and no significant effect for age was found (P = 0.60).
Change in hand deformity with age All children were classified as having some level of hand deformity in each of the video clips, even the four youngest children who were below 2 years of age (three subsequently classified at MACS level II and one at level III). Table 2 shows the numbers of children with each combination of initial and final NHDC classification. Fourteen children (54%) had the same classification at the final assessment as at the first. For four of these children (15%), the classification changed in the intervening time points but reverted to their original classification by the final time point. The number classified as F3 increased from three children to seven children during the course of the study period. No child was classified as F4 at any time point.
TABLE 2: Cross-tabulation of NHDC classifications at first and final time points Final classification Initial Classification
F1
F2
F3
E1
E2
Total
F1 F2 F3 E1 E2 Total
2 1 0 1 0 4
4 6 0 0 0 10
1 2 3 1 0 7
0 0 0 3 0 3
1 1 0 0 0 2
8 10 3 5 0 26
Shading indicates no change between first and final classifications. © 2014 Occupational Therapy Australia
6 Table 3 shows the NHDC at each time point for each study participant, arranged according to the types of interventions: (i) BoNT-A and surgery, (ii) BoNT-A only, (iii) surgery only, and (iv) no BoNT-A or surgery. The changes in hand deformity in this observational study were grouped into three categories. Firstly, 8 children, one with an extension classification and seven with a flexion classification at first video recording, changed from an extended or less flexed wrist posture to a more flexed wrist posture over time. These children are designated as ‘↑ flexion’ in Table 3. Secondly, 10 children did not change classification at any time point, 9 of whom had a flexion classification. These children are designated as ‘no change’ in Table 3. Thirdly, 8 children did not follow a consistent sequence in change of classification level, and either changed from a flexion classification to an extension classification or vice versa (e.g. F1 to E2, E1 to F3). These children, 4 of whom had dystonia, are designated as ‘variable’ in Table 3. Classification of hand deformity of children at older ages generally showed more flexed (F2 and F3) classifications. Out of 14 children with initial video clips at 6 years of age or younger (within an Early Intervention Program), 11 were classified as F1 or E1, 2 were classified as F2 and 1 as F3. Out of 12 children with initial video clips over 6 years of age (within a School Aged Intervention Program), 11 were classified as F2/F3 or E2 classification. Seven children presented with dystonia as indicated in Table 3. Three children were in the ‘no change’ group and four in the ‘variable’ group following classification. None were in the ‘↑ flexion’ group. The Occupational Therapists classifying hand deformity for the children with dystonia did not view any fluctuation in level of hand deformity from the clips to conflict with the classification assigned to the individual child at that point in time. The variability in classification, particularly fluctuations between flexion and extension deformities for those with dystonia in the ‘variable’ group, was evident only between one time point and the next. Nineteen participants received upper limb BoNT-A and/or hand surgery. Their ages at intervention and NHDC classifications are shown in Table 3. Seven children did not receive hand surgery or upper limb BoNTA during the period of the video recordings. For 5 of these 7 NHDC classification did not change. However, none of these had a video clip taken before the age of 7 years. Age at first video for the remaining two participants was 1 year and 10 years.
Conclusion The first aim of the study was to determine the interobserver and intra-observer reliability of the NHDC. The NHDC demonstrated ‘almost perfect’ (Landis & Koch, 1977) inter-observer and intra-observer reliability in classifying hand deformity in children with CP. © 2014 Occupational Therapy Australia
M. GEORGIADES ET AL.
However, two or three viewings of the video clips were sometimes required when classifying children with fluctuating tone, dystonic posturing, and well-established compensatory movement patterns. The almost perfect inter-observer reliability between novice and experienced clinicians under the conditions of this retrospective study suggests that the NHDC is easy to use and does not require special training. Whether the NHDC would be equally reliable for the in vivo classification of hand deformity depends on how much within-subject variability exists in the way children in this population approach an object. Investigation of this question is recommended for future research. Given the potential for fluctuation in classification of hand deformity in children with dystonia, caution must be taken when comparing hand deformity classification over time. The NHDC is capable of classifying hand deformity in children with dystonia at a particular point in time. In this situation the NHDC may be used by the observer as a guide for clinical consideration of the factors contributing to the deforming forces and the structures involved. The second aim of the study was to describe observed changes in hand deformity over time in children with CP. These changes do not represent natural history because the children in this study were receiving various interventions from community and tertiary service providers for upper limb spasticity. Our results demonstrate that hand deformity may occur before the age of 2 years, although it must noted that children attending a specialist hand clinic before the age of 2 years are unlikely to be representative of all children with CP involving upper limb impairment. Although all children had hand deformity at all ages, children aged 6 years or younger had less wrist flexion deformity than older children. The study did not provide evidence for a consistent pattern of changes in hand deformity for all children: 10 children did not change hand classification, wrist flexion increased in 8, and 8 showed changes from wrist flexion to extension or vice versa. Children who moved from an extension classification to a flexion classification or from an F1 to F2 or F2 to F3 classification on the NHDC lost active wrist extension range of motion. Due to the retrospective nature of the study and inconsistent recording of passive range of motion in the children’s medical and therapy records, it is impossible to report the loss of passive range of motion for the children in the study. The study illustrates the difficulty in predicting future hand deformity from a child’s current classification. For example, an F1 classification may be taken to indicate that no immediate intervention is required. However, in this study, three quarters of the children either moved from less than 20 degrees (F1) to greater than 20 degrees of wrist flexion (F2/F3) or changed to a wrist extension classification over time. It is important,
7
THE NHDC FOR CHILDREN WITH CP
TABLE 3: NHDC classifications and upper limb interventions for individual participants
ID
GMFCS/ MACS
Ages at classification
NHDC levels at each age
Category
Ages at BoNT-A
F2,F2,F2,F2 F2,F2,F3,F3,F3,F3 F2,F2,F1,F2,F2
No change ↑ flexion Variable
11,12,17 12,12,12,14,14 11,11,13,13
9,10,12,13,15,16,17 6,10,11
F2,F2,F3,F2,F2,F2,F1 E1,E1,E1
Variable No change
5,12,13,14,15,16,16,17,17 3,3,5,8,8,9,10,10,11,11
3,7,9,11,12,14,15,16 1,2,4,4,5,7,8
E1,F2,F2,F3,F2,F2,E1,E1 E1,E1,E1,E1,F2,F2,F1
Variable ↑ flexion
10,11,14 4,4,5,5,6
3,4,8,9 3,4,5,7,8 1,3,4,5,5,9,10 3,4,4,8,9 1,2,4,4,6 1,3,4,6,8 7,10,13 5,9,11 3,4,5 7,9,10,13
F1,F1,F1,F1 F1,F2,F2,F2,F2 F1,F1,F1,F2,F2,F3,F2 F1,F1,F2,F2,F2 F1,F1,E1,F2,F2 F1,F1,F2,F3,F3 F2,F2,F2 F2,F3,F2 F3,F3,F3 E1,E1,F3,E1
No change ↑ flexion ↑ flexion ↑ flexion ↑ flexion ↑ flexion No change Variable No change Variable
E1,E2,F3
Variable
4,7,7,8 5,5,6,7,8,8 4,5,6,6,7,7,8,8 4,5,5,6,6,7,8,8 3,3,5,5,6 3,4 12 11 2,3,3,4,4,5 5,6,8,8,9,10,10,11,11,11, 12,12,13 4,4,5,5,6,6
F2,F2,F2,F2,E2
Variable
F1,F1,F1 F1,E1,E2 F2,F2,F2 F2,F2,F2,F2 F2,F3,F3 F3,F3,F3 F3,F3,F3
No change Variable No change No change ↑ flexion No change No change
BoNT-A and surgery 1 I/II 9,10,14,17 2 III/III 9,11,11,14,14,15 3 I/II 7,9,10,12,15 4 5
II/II II/II
6 I/II 7 I/II BoNT-A only 8‡ V/IV 9 II/II 10 II/II 11 II/II 12 II/II 13 II/III 14 I/II 15 I/II 16 I/II 17‡ I/II
18‡ I/II 2,4,6 Surgery only 19‡ IV/IV 5,9,11,12,17 No BoNT-A or surgery 20‡ I/I 12,13,14 21‡ I/II 1,2,3 22 I/I 7,10,10 23 III/II 8,11,13,13 24 I/II 10,12,13 25‡ I/II 7,10,10 26 I/II 10,12,13
Ages at surgery
Types of surgery†
14 14 11 14 16 6 10 14 7
1,3,4 2,4 4 2 1,5 5,10 7 1,4,5 4,6,9
13
1,6,8
†Surgery type: 1 – tendon transfer – Flexor Carpi Ulnaris (FCU) to Extensor Carpi Radialis Brevis or Longus; 2 – tendon transfer – FCU to supination transfer (transfer to dorsal radius insertion to periosteum); 3 – FDSR to Abductor Pollicis Brevis; 4 – musculotendinous release of forearm pronator; 5 – musculotendinous release of extrinsic flexor muscles; 6 – musculotendinous release of thumb intrinsic (Adductor Pollicis); 7 – intrinsic releases of fingers; 8 – Ulnar Neurectomy; 9 – thumb MCP joint stabilisation; 10 – Abductor Pollicis Longus Tenodesis surgery. ‡Dystonia. GMFCS – Gross Motor Function Classification Scale; MACS – Manual Ability Classification Scale.
therefore, to monitor changes in the hands of children with CP over time, regardless of classification. Consistent methods of recording and maintaining the records of change are essential, particularly when children’s occupations change from early intervention to schoolaged therapy programmes or if there is a change in primary therapist or service provider.
The number of children with an F3 classification more than doubled from three at first classification to seven at final classification. These children have no active wrist extension. Intervention to maintain passive wrist and finger flexor muscle length is paramount for these children to prevent flexion contractures. © 2014 Occupational Therapy Australia
8 The study had several limitations owing to its retrospective design. The study population was a sample of convenience, because only children with appropriate video clips were included. Not all treatment records could be collected and video clips were not at regular intervals. Given the heterogeneity of the children in this study, subgroup analyses would have been desirable, but the small sample size of 26 precluded this. It was not possible to blind observers to the children’s identity nor to the chronological order of their clips because the observers could sometimes see part of the child on the screen and could infer the chronological order from the size of the child’s hand. However, the observers were blind to the timing of any interventions. They viewed and classified each video independently, without reference to how the child had been classified in other video clips. In all video clips, the child was actively or passively reaching for an object on a horizontal surface, but the task and type of object and camera position were not standardised. It is recommended that the effects of these factors on hand position be examined in future research. However, the NHDC is designed to be much less affected by task or object than are comprehensive assessments of hand function, provided the task given to the child involves approach of the hand to an object. This is because the biomechanical limitations in the child’s hand constrain the range of possible movements of the wrist and fingers during approach to an object, and the same constraints will operate whatever the task. Typically, in clinical practice, the NHDC would guide a comprehensive functional assessment with a range of tasks. As this is the first study to be published on the NHDC, there is great scope for further investigation of this tool. It is recommended that a prospective longitudinal study be undertaken to examine changes in hand deformity systematically over time, the reliability of NHDC in vivo, and whether the NHDC classification correlates with functional impact of hand deformity. The NHDC should be used in the context of comprehensive hand assessments. Videos should be taken at regular intervals to identify if a child’s hand deformity classification changes over time. In clinical settings, the NHDC, together with a comprehensive upper limb assessment and accurate documentation, would enable clinicians to identify changes in the level of hand deformity in children with CP over time. Ideally, video data should be routinely collected during the assessment process to support effective documentation and communication, and provide evidence to guide clinical practice. In conclusion, this is the first study to document longitudinal changes in hand deformity in children with CP. It shows that the NHDC is reliable in classifying hand deformity by video analysis, is easy to use, and does not require special training. The study also demonstrates that children with CP can present with hand © 2014 Occupational Therapy Australia
M. GEORGIADES ET AL.
deformities from an early age, and therefore need to be monitored regularly.
Acknowledgements The authors declare no conflicts of interest. Financial support was received from Edith Cowan University by Melissa Georgiades as part of her honours study. We wish to convey our gratitude to the children and families involved in this study and to The Centre for Cerebral Palsy and Princess Margaret Hospital for Children for access to the children’s video data and treatment histories. Thanks also to Mr Geoffrey Hammond from the Telethon Institute for Child Research for the syntax modification, to Dr Sonya Girdler from Curtin University for her academic support, to Dr Siobhan Reid from University of Western Australia for access to video data and to Mrs Stephanie Mann from The Centre for Cerebral Palsy for her assistance in data collection.
References Arner, M., Eliasson, A., Nicklasson, S., Sommerstein, G. & Hagglund, G. (2008). Hand function in cerebral palsy. Report of 367 children in a population-based longitudinal health care program. Journal of Hand Surgery, 33A, 1337– 1347. Boyd, R. N. & Graham, H. K. (1997). Botulinum toxin A in the management of children with cerebral palsy: Indications and outcomes. European Journal of Neurology, 4, S15– S21. Brown, K. (1985). Positional deformity in children with cerebral palsy. Physiotherapy Theory and Practice, 1, 37–41. Case-Smith, J. (2010). Development of childhood occupations. In: J. Case-Smith & J. C. O’Brien (Eds.), Occupational therapy for children (6th ed., pp. 56–83). Missouri: Mosby Elsevier. Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37–46. Damiano, D. L., Quinlivan, J., Owen, B. F., Shaffrey, M. & Abel, M. F. (2001). Spasticity versus strength in cerebral palsy: relationships among involuntary resistance, voluntary torque, and motor function. European Journal of Neurology, 8, 40–49. DeLuca, P. A. (1996). The musculoskeletal management of children with cerebral palsy. Pediatric Clinics of North America, 43, 1135–1150. Dirckx, J. H. & Cadle, K. J. (Eds.). (2008). Stedman’s medical dictionary for the health professions and nursing (6th ed.). London: Wolters Kluwer Health/Lippincott Williams & Wilkins. Flett, P. J. (2003). Rehabilitation of spasticity and related problems in childhood cerebral palsy. Journal of Pediatrics and Child Health, 39, 6–14. Graham, H. K. (2004). Mechanisms of deformity. In: D. Scrutton, D. L. Damiano & M. Mayston (Eds.), Management of the motor disorders of children with cerebral palsy (2nd ed., pp. 105–129). London: Mac Keith Press.
THE NHDC FOR CHILDREN WITH CP
Graham, H. K. & Selber, P. (2003). Musculoskeletal aspects of cerebral palsy. Journal of Bone and Joint Surgery, British, 85B, 157–166. House, J. H., Gwathmey, F. W. & Fidler, M. O. (1981). A dynamic approach to the thumb-in palm deformity in cerebral palsy. The Journal of Bone and Joint Surgery. American Volume, 63, 216–225. Landis, J. R. & Koch, G. G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33, 159–174. Law, K., Lee, E. Y., Fung, B. K., Yan, L. S., Gudushauri, P., Wang, K. W. et al. (2008). Evaluation of deformity and hand function in cerebral palsy patients. Journal of Orthopaedic Surgery and Research, 3, doi:10.1186/1749-799X-3-52. Lin, J.-P. (2004). The assessment and management of hypertonus in cerebral palsy: A physiological atlas (‘road map’). In: D. Scrutton, D. L. Damiano & M. Mayston (Eds.), Management of the motor disorders of children with cerebral palsy (2nd ed., pp. 85–104). London: Mac Keith Press. McConnell, K., Johnston, L. & Kerr, C. (2011). Upper limb function and deformity in cerebral palsy: A review of classification systems. Developmental Medicine and Child Neurology, 53, 799–805. Park, E. S., Sim, E. G. & Rha, D. W. (2011). Effect of upper limb deformities on gross motor and upper limb func-
9 tions in children with spastic cerebral palsy. Research in Developmental Disabilities, 32, 2389–2397. Pfeifer, L. I., Pacciulio, A. M., dos Santos, A. C., dos Santos, J. L. & Stagnitti, K. E. (2011). Pretend play of children with cerebral palsy. Physical and Occupational Therapy in Pediatrics, 31, 390–402. Priori, A., Cogiamanian, F. & Mrakic-Sposta, S. (2006). Pathophysiology of spasticity. Neurological Sciences, 27, s307–s309. Reddihough, D. S. & Collins, K. J. (2003). The epidemiology and causes of cerebral palsy. Australian Journal of Physiotherapy, 49, 7–12. Venkateswaran, S. & Shevell, M. I. (2008). Comorbidities and clinical determinants of outcome in children with spastic quadriplegic cerebral palsy. Developmental Medicine and Child Neurology, 50, 216–222. Wilton, J. C. (2003). Casting, splinting, and physical and occupational therapy of hand deformity and dysfunction in cerebral palsy. Hand Clinics, 19, 573–584. Wilton, J. C. (2004). Splinting and casting in the presence of neurological dysfunction. Hand splinting: Principles of design and fabrication (pp. 168–197). Perth, WA: Success Print. Zancolli, E. A. (2003). Surgical management of the hand in infantile spastic hemiplegia. Hand Clinics, 19, 609–629.
© 2014 Occupational Therapy Australia
