Brian Hoare, BOccThy, PhD1,2, Christine Imms, BAppScOT, MSc, PhD2,3,4, Melinda Randall, BAppScOT, PhD2,3 and Leeanne Carey, BAppScOT, PhD1,5

From the 1School of Occupational Therapy, La Trobe University, Bundoora, 2Victorian Paediatric Rehabilitation Service Monash Medical Centre, Victoria, 3Royal Children’s Hospital, Melbourne, 4Murdoch Children’s Research Institute, Victoria and 5Head, Neurorehabilitation and Recovery, National Stroke Research Institute, Florey Neurosciences Institutes, Melbourne Brain Centre, Heidelberg, Australia

BACKGROUND: Intervention studies describe outcomes as measuring specific domains of the International Classification of Functioning, Disability and Health (ICF). However, the same measure may be described by different authors as assessing different domains, resulting in considerable confusion and inconsistent reporting of outcomes.

OBJECTIVE: To systematically link the scored items from the Melbourne Assessment of Unilateral Upper Limb Function, Quality of Upper Extremity Skills Test and Assisting Hand Assessment to domain(s) of the ICF.

METHODS: The meaningful concept for each scored item was defined. Using ICF linking rules, the concepts were assigned ICF codes to determine the outcome’s overall domain of measurement.

RESULTS: The Melbourne Assessment predominantly evaluates concepts in the body function domain. Coding of the Quality of Upper Extremity Skills Test indicated that dissociated movement, weight-bearing and protective extension predominantly measure concepts in the body function domain. Grasp was the only domain where concepts were coded in both the body function and activity domains. The Assisting Hand Assessment was the only measure where the majority of items assessed concepts in the activity domain.

CONCLUSION: Measures of upper limb function can be categorized according to ICF domains. These findings should resolve confusion surrounding the classification of these measures and provide a reference for reporting the impact of intervention.

Key words: cerebral palsy; assessment; upper extremity.

J Rehabil Med 2011; 00: 00–00

Correspondence address: Brian Hoare, School of Occupational Therapy, La Trobe University, Bundoora; and Occupational Therapist, Victorian Paediatric Rehabilitation Service, Monash Medical Centre, 246 Clayton Road, Clayton, Victoria, Australia. E-mail: brian.hoare@southernhealth.org.au

Submitted February 23, 2010; accepted August 17, 2011

INTRODUCTION

International Classification of Functioning, Disability and Health

The International Classification of Functioning, Disability and Health (ICF) was developed in 2001 by the World Health Organization (WHO) as a framework for measuring health and disability, based on a global consensus of multiple stakeholders (1). The WHO did not intend for the ICF to act as a static framework, rather it was to respond to researcher, clinician and consumer feedback with ongoing development and future revision. As such, recognition of the need for an ICF version that could be universally adopted for children and youth led to the publication of a specific version known as the ICF-Children and Youth (ICF-CY) (2). This adaptation was designed to record the unique characteristics of the developing child and their surrounding environment (2).

Over the past decade, the ICF conceptual framework and language has emerged as the international standard across health-related disciplines for understanding and communicating an individual’s health condition and functioning. Jette (3) acknowledges understanding of this framework as fundamental to advancing the science of disablement. The ICF views human functioning as a concept along a continuum that encompasses the domains of body functions and structures, activities and participation. Using this framework, the ability of an individual to function is seen as a dynamic interaction between elements of these domains and influenced by contextual factors including environmental and personal factors (1). The multi-dimensional framework and language of the ICF bear similarities with other disablement models such as Nagi’s (4) Disablement Model, which considers the health condition in association with personal and environmental factors as influences in functioning and disability. It is also consistent with the dynamic and interactive view of person and environment that underpin the core philosophies of occupational therapy practice (5) including the Model of Human Occupation (MOHO) (6) and the Canadian Model of Occupational Performance (CMOP) (7).

Across health-related disciplines, the ICF framework has acted to translate many discipline-specific concepts allowing the explicit identification and reporting of domains of practice and treatment effect in a commonly understood language (8). The development of linking rules by Cieza et al. (9) has also provided a standardized procedure to enable intervention and outcome measures to be linked to the ICF. Importantly, this provides “a connecting framework between interventions and outcome measures, facilitating the selection of the most appropriate outcome measure for the aim of the intervention” (10).

Cerebral palsy and the influence of the ICF

Cerebral palsy is a health condition that describes “a group of disorders of the development of movement and posture causing activity limitation, that are attributed to non-progressive disturbances that occurred in the developing fetal or infant brain” (11). The integration of the term activity limitation in the most recent description of cerebral palsy by Rosenbaum et al. (11) serves as evidence for the recognition and endorsement of the ICF as the framework for articulating and reporting outcomes related to children with cerebral palsy. Rosenbaum & Stewart (12) note the influence of the ICF upon cerebral palsy assessment and treatment as it has helped to expand thinking beyond fixing impairments to promoting functional activity and full participation of children in life activities. As a result, recent research has placed more emphasis on what children actually do rather than what they can do in a controlled environment or how normal their movements appear. This shift has significantly influenced the treatment of hand function in children with hemiplegic cerebral palsy and has lead to greater promotion, exploration and targeting of outcomes related to the activity and participation domain of the ICF.

ICF: Activity and Participation domain

The ICF defines activity as “the execution of a task or action by an individual” and participation as “involvement in a life situation” (1). The WHO reports that this domain can be used to denote activities or participation or both. The domain is further delineated by two qualifiers known as capacity and performance. Capacity describes an individual’s ability to execute actions or tasks in an optimal environment and performance describes what an individual does in his or her current environment or the real world (2). The ICF has led to a greater understanding that maximal capacity demonstrated during optimal conditions is not automatically an indicator of performance in everyday life (13). The gap between capacity and performance often reflects the impact of the ideal and current environment, providing a useful guide as to what can be altered or adapted in the individual’s environment to improve performance.

In 2003, Jette et al. (14) initially distinguished activity and participation as two separate concepts, however the authors later questioned the wisdom of adopting this view (3, 15). Unfortunately, the lack of operationally defined distinction between activity and participation currently remains the ICF’s greatest limitation (3, 16). It is felt that the ability to separate this domain as two distinct concepts remains essential if the ICF is to achieve longstanding acceptance as an international classification framework (3, 17). In addition, precise internal coherence within the ICF is necessary for the understanding of constructs within existing and newly developed assessment tools (3).

Despite the current lack of clarity in operational differentiation, occupational therapists, along with other health professionals, often view activity and participation as distinct concepts for both measurement and the articulation of health-related outcomes for children with cerebral palsy. This is particularly evident in recent cerebral palsy literature, where clinical measures have been categorized as distinct activity (18, 19) or participation measurement tools (20, 21). In addition, further differentiation within the activity domain of the ICF has also seen the capacity and performance qualifiers being used to categorize activity domain measures (18, 22, 23).

Inconsistency in reported ICF classification of upper limb measurement tools

There are a few commonly used outcome measures for children with cerebral palsy, such as the Quality of Upper Extremity Skills Test (QUEST) (24) and the Melbourne Assessment of Unilateral Upper Limb Function (Melbourne Assessment) (25), that have been used to evaluate change following upper limb intervention over the past two decades. Recently, the Assisting Hand Assessment (AHA) (26) has also emerged as a popular outcome measure. The lack of clear definition of, and distinction between, body function, activity and participation, and varying interpretation of the conceptual nature of these ICF domains has lead to inconsistent categorization of these commonly used outcome measures. This has led to inconsistent reporting of the type of outcome that can be expected following intervention. Inconsistencies are particularly evident for these 3 measures where items appear to assess change across more than one domain of the ICF. In addition, many items are administered within the context of functional activities; however the specific scoring criteria for these items measure components of the movement or body functions within the activity, rather than the outcome of the activity itself. Therefore, what might appear to be a change in activity level performance may actually reflect change in the body function domain. For example, the reach to brush from forehead to back of neck item in the Melbourne Assessment is scored from observation of the child performing the action of brushing the palm of their hand from their forehead to the back of their neck. Scoring this item involves rating two movement components observed as the child performs the action. These components are active range of movement (B7011, B7601) and fluency (B7651), both of which are items in the body function domain of the ICF.

This confusion has had detrimental effects on the interpretation of research outcomes. Without further clarification this confusion has the potential to hamper communication between researchers and clinicians and the advancement of knowledge on outcomes related to upper limb intervention in children with cerebral palsy. This is particularly relevant for emerging research that aims to explore relationships and interactions between domains of the ICF (23).

Current classifications of the Melbourne Assessment, QUEST and AHA according to ICF domains, as reported in cerebral palsy intervention trials and review papers are summarized in Table I. This summary highlights the current inconsistency in ICF classification and interpretation for these measures. The Melbourne Assessment has exclusively been classified as an activity domain measure, except by Wasiak et al. (27), and Hoare & Imms (28) who initially provided the classification of body function/body structure and, later, a combination of both body function/body structure and activity (29). A similar issue exists for the QUEST, where 5 out of 9 papers report the tool as a measure of activity. Hoare & Imms (28) and Olesch et al. (30), however, assign a classification of body function/body structure. Hoare et al. (29) later provided a classification of both body function/structure and activity, which was consistent with Klingels et al. (31). The only assessment to demonstrate consensus across all papers was the AHA, where all authors classified it as an activity level measure.

Aim

The aim of this paper is to systematically define and objectively link the meaningful concepts of scored items contained in 3 commonly administered standardized upper limb outcome measures used in cerebral palsy research, i.e. the Melbourne Assessment, QUEST and AHA, to specific codes within the ICF-CY. This process aims to: (i) resolve current confusion with identifying the conceptual understanding and domain classification of these important outcomes; (ii) serve as an important reference for clinicians and researchers for identifying and reporting the impact of upper limb intervention for children with hemiplegic cerebral palsy; and (iii) assist in the selection of appropriate outcome tools for future intervention trials. The paper does not aim comprehensively to describe or review the psychometric properties of each measure as these are been reported elsewhere in the literature (25, 42–44).

METHODS

The meaningful concept for each scored item on the Melbourne Assessment, QUEST and AHA were defined with careful consideration of the test situation, rationale and purpose of each measure. Using the 8 revised ICF linking rules outlined by Cieza et al. (10) (Appendix SI, available from http://www.medicaljournals.se/jrm/content/?doi=10.2340/16501977-0886), the meaningful concepts were assigned ICF-CY codes to determine the outcome’s overall ICF domain of measurement. Where possible, codes were assigned at the fourth level, indicated by 4 digits following the prefix. Items not specifically meeting fourth level ICF-CY descriptions were coded at the third level. Two raters (BH, MR) independently linked codes from the ICF-CY to the meaningful concept for individual items on each outcome measure. A third independent rater (CI) evaluated the assigned codes for agreement. Any disagreements were resolved by discussion between the 3 raters.

RESULTS

Tables II–IV provide a description of scored items, meaningful concept and consensus agreement for the ICF-CY codes assigned to each item for the Melbourne Assessment, QUEST and the AHA. Many of the activities within each measure included multiple scored items. For example, the release of crayon item in the Melbourne Assessment required scoring of range of movement, quality of movement and accuracy of release. Through discussion of this item consensus was reached and 3 distinct meaningful concepts were identified for each score. As a result, individual items for each measure could be assigned multiple ICF codes.

Assignment of ICF-CY codes to the meaningful concepts of score items on the Melbourne Assessment indicated that this assessment predominantly evaluates change at the body function domain of the ICF-CY (see Table II). Only 1/37 scored items relates to the activity domain alone, whilst 19/37 relate to body function and, 16/37 a combination of both body function and activity. One item, speed of upper limb movement on reach to mouth, was not definable. Consistent with the purpose of the Melbourne Assessment, to quantify the quality of upper limb motor function in children with unilateral upper limb impairment, the majority of the items score mobility of joints, control of simple or complex movement and tremor. The items in the Melbourne Assessment that measure activity level performance within the ICF-CY include concepts of hand skill development such as grasp, release, manipulation, pointing and reaching (see Table II).

The QUEST includes 34 items that evaluate both upper extremities separately in 4 domains including: dissociated movement, grasp, protective extension, and weight-bearing. Each meaningful concept for scored items from the 4 domains were coded separately. Dissociated movement items predominantly measure concepts in the body function domain with 15/19 scored items coded as the body function alone and the remaining 4/19 a combination of body function and activity. A similar outcome was obtained for weight-bearing, with 24/26 scored items coded as body function alone and 2/26 a combination of body function and activity. All meaningful concepts for protective extension were in the body function domain (18/18). Grasp was the only domain of the QUEST where concepts for scored items were coded as both the body function and activity (14/15). The remaining item was in the body function domain (1/15; Table III).

The AHA was the only measure where a majority of scored items were found to evaluate concepts in the activity domain or a combination of activity and body function (17/22). Only 5/22 scored items exclusively measured concepts in body function. Many aspects of hand function, including reach, grasp, release, putting down, picking up, and coordination are evaluated in the context of bimanual activities. The AHA is distinct from the Melbourne Assessment and QUEST as some concepts include the use of cognitive strategies required for hand function including solving simple problems, comprehending simple spoken messages and body gestures (Table IV).

DISCUSSION

Using the ICF, a universally acknowledged framework for measuring health and disability, this paper has defined the meaningful concept for each scored item on the Melbourne Assessment, the QUEST and the AHA. The Melbourne Assessment and the QUEST were found to predominantly include concepts within the body function domain, whilst the AHA predominantly includes concepts in the activity domain. All assessments however, possess items that include concepts within both the body function and activity domains.

Despite recent findings demonstrating a high correlation between the Melbourne Assessment and the QUEST (indicating concurrent validity) (31), the Melbourne Assessment’s emphasis on evaluation of quality of movement provides distinctly different information when compared with the QUEST. The Melbourne Assessment includes multiple test items that measure control of simple or complex movement and tremor, making it ideally suited for measurement of children with movement-based disorders such as dystonic or athetoid cerebral palsy. Recently, further investigation of the construct validity of the tool established that the 37 score items on the Melbourne Assessment do not comprise a uni-dimensional scale (45). A series of Rasch analyses established evidence to support the Melbourne Assessment as consisting of 4 separate uni-dimensional sub-scales. The sub-scales identified separately measure elements of movement quality including: range movement, accuracy (of reach and release), fluency of upper limb movement and dexterity (of grasp). These sub-scales, developed for the updated modified Melbourne Assessment (45) will continue to provide measurement at the body function or a combination of both body function and activity domains of the ICF-CY.

The QUEST was designed in 1993 to capture patterns of movement that are part of normal development and considered to be the basis for upper limb performance (43). At a time where a popular emphasis was on the use of neuro-developmental therapy techniques for children with cerebral palsy, the QUEST provided an evaluation tool that was able to measure a child’s ability to move out of pathological patterns against gravity and their protective reflex responses (43). Following analysis of longitudinal data obtained using the QUEST, it has been suggested that the impairments that underlie many of the items in the QUEST are unlikely to improve through movement or task-related practice (46). Improved clarity from assignment of ICF-CY codes to the meaningful concepts of the QUEST now provides additional support and evidence for this suggestion. Movement-based interventions predominantly target change in the activity domain. Except for the grasp domain, the QUEST overwhelmingly evaluates concepts in the body-function domain, making it more appropriate for evaluation of interventions that target improvements in body functions. The changes on the QUEST seen in previous clinical intervention trials of movement-based therapies may predominantly be related to change in the grasp domain. This warrants analysis and reporting of the separate domains of the QUEST. Future research evaluating upper limb practice-based or movement-based interventions (i.e. constraint-induced movement therapy, bimanual occupational therapy) in children with cerebral palsy should question the use of total QUEST scores in these trials. Improvements in the body structure and function are unlikely to be associated with similar levels of improvement in activities and participation (47). Expectations and hopes for additional influence across domains are common features of many recent intervention trials targeting change in the body function domain; however, we cannot expect change to be seen in any other domain other than the one on which treatment is focused (48).

Aside from scored items, the administration procedures for the 3 assessments demonstrate important differences. It appears that these differences have contributed to the confusion in the literature surrounding the ICF classification of each measure, particularly the Melbourne Assessment and QUEST. These assessments have often been referred to as measures of a child’s capacity (18, 31), suggesting activity domain measurement. In the context of functional activities, administration of the QUEST and in particular, the Melbourne Assessment often requires children to perform test items at their best capacity. The administration process itself however, is not the scored construct. The specific scoring criteria for many of the items measure components of the movement or body functions within these functional activities not the outcome of the activity itself. As demonstrated by the identification of meaningful concepts and application of the ICF coding procedure, it is inappropriate to continue to refer to the Melbourne Assessment or QUEST as measures of activity domain capacity. Only the grasp domain of the QUEST could be used for this purpose as the meaningful concepts for grasp relate to the activity domain or a combination of body function and activity domains.

The central aim of all upper limb motor-based interventions in children with hemiplegic cerebral palsy is to improve the actual use (performance) of their affected upper limb in a range of daily tasks, particularly those requiring bimanual performance (49). As stated by Gordon (41), and now supported by results of this ICF code assignment, at this time the only commonly used upper limb specific activity-level measure of performance for children with hemiplegic cerebral palsy is the AHA, which “sets the criterion standard in quantifying upper limb activity limitation” (41). The AHA has been constructed with the underlying principle that skilled hand use is influenced by a number of complex components including motor, perceptual, cognitive and environmental aspects. These components often represent the areas targeted by therapists using practice or movement-based interventions such as constraint-induced movement therapy or bimanual occupational therapy. Unlike other assessments, the AHA does not set out to capture these specific aspects individually or instruct a child to perform at their best. It attempts to synthesize all these components by observing the actual behaviour or functional use of the assisting hand when performing bimanual tasks (44). Changes on the AHA are therefore more likely to reflect what a child may do in their usual environment or assess the transfer of intervention effects into daily life.

CONCLUSION

The outcome of this identification and linking process has provided improved clarity and contributed evidence to support the validity of the measured concepts for the Melbourne Assessment, QUEST and AHA. The results can be used to guide clinicians and researchers in the interpretation of upper limb clinical intervention trials for children with cerebral palsy and in the selection of appropriate outcome measures for future intervention trials.

REFERENCES