Development of a music therapy assessment tool for ... - IOS Press

319

NeuroRehabilitation 22 (2007) 319–324 IOS Press

Development of a music therapy assessment tool for patients in low awareness states Wendy L. Magee ∗ Institute of Neuropalliative Rehabilitation, London, UK Department of Palliative Care, Policy and Rehabilitation, Kings College, London, UK

Abstract. People in low awareness states following profound brain injury typically demonstrate subtle changes in functional behaviors which challenge the sensitivity of measurement tools. Failure to identify and measure changes in functioning can lead to misdiagnosis and withdrawal of treatment with this population. Thus, the development of tools which are sensitive to responsiveness is of central concern. As the auditory modality has been found to be particularly sensitive in identifying responses indicating awareness, a convincing case can be made for music therapy as a treatment medium. However, little has been recommended about protocols for intervention or tools for measuring patient responses within the music therapy setting. This paper presents the rationale for an assessment tool specifically designed to measure responses in the music therapy setting with patients who are diagnosed as minimally conscious or in a vegetative state. Developed over fourteen years as part of interdisciplinary assessment and treatment, the music therapy assessment tool for low awareness states (MATLAS) contains fourteen items which rate behavioral responses across a number of domains. The tool can provide important information for interdisciplinary assessment and treatment particularly in the auditory and communication domains. Recommendations are made for testing its reliability and validity through research. Keywords: Music therapy, minimally conscious, vegetative, assessment

1. Background: Assessment of patients in low awareness states Assessment of the person in a vegetative (VS) or minimally conscious state (MCS) is a complex task for all disciplines of the multidisciplinary team [1] even for those with considerable experience [2]. All the current tools for assessing VS and MCS patients rely on behavioral observation ratings using clear operational definitions of behaviors which assist with repeated measurements over time [22]. However, the sensitivity of assessment tools is a long standing issue in the assessment of patients in low awareness states [12]. In order to optimally assess and treat such patients, tools must be capable of measuring subtle changes in functional behaviors [11]. Those who show improvement usually make ∗ Address

for correspondence: Dr. Wendy Magee, Institute of Neuropalliative Rehabilitation, West Hill, London SW15 3SW, UK. Tel.: +44 20 8780 4500 ext. 5146; Fax: +44 20 8780 4569; E-mail: [email protected].

small and often subtle changes. If assessment fails to identify these changes then the treatment team may fail to plan effective treatment, or withdraw treatment altogether. Without treatment, optimal recovery may never be achieved. In fact Andrews et al. [2] found that failure to identify subtle changes in improvement contributed to 42% of patients being misdiagnosed. Shiel et al. [18] point out that failing to identify recovery in these patients not only leads to poor prognosis, but also can effect the motivation of the treatment team. Hence, optimal care is dependent upon measuring changes in the patient’s functioning which may be small, inconsistent and infrequent. In addition to scale sensitivity, the lack of comprehensive standardized assessments for this population contributes to misdiagnosis [11].

2. Assessment tools for low awareness states In response to this challenge, several assessment tools which are capable of measuring small changes in

ISSN 1053-8135/07/$17.00  2007 – IOS Press and the authors. All rights reserved

320

W.L. Magee / Development of a music therapy assessment tool for patients in low awareness states

patient responsiveness have been developed and standardized. A brief review of these follows, with particular attention to tool sensitivity and validation. Two sensory stimulation protocols in particular have been found to be sensitive to assessing small responses in slow-to-recover patients in low awareness states after the very acute stages of recovery. The Sensory Modality Assessment and Rehabilitation Technique (SMART) provides a graded assessment of a patient’s sensory, motor and communicative responses to stimuli across the five senses in addition to wakefulness/arousal [11]. The consistency of the patient’s response is linked to meaningful responses, suggesting the tool has diagnostic power which is pending future investigation. It involves formal components, carried out by trained assessors, and informal components implemented by caregivers respectively. Reliability and validity has been established for its use by trained assessors in discriminating awareness in VS and MCS patients [11] and the SMART has been found to have agreement with the Western Neuro Sensory Stimulation Profile [17] or ‘WNSSP’. The WNSSP measures responsiveness by evaluating arousal/attention, auditory comprehension, visual comprehension and tracking, object manipulation and expressive communication [3]. It has been critiqued for failing to allow comparison of responses across the sensory modalities, thus reducing its sensitivity to detect higher cognitive functions [10]. Unlike SMART, the Wessex Head Injury Matrix (WHIM) rates behavioral responses during everyday tasks and to a series of simple stimuli developed to elicit behaviors such as calling the patient’s name [20]. It is a scale which uses repeated observations across time and context to gauge consistency of behavioral response or establish treatment goals. It examines behaviors ordered hierarchically in items across motor ability, cognitive skills and social interaction. Its particular strengths are its focus on behaviors rather than diagnostic features and its potential to be used by all members of the multiprofessional treatment team. Reliability has been established for both inter-rater and test-retest reliability [20]. A further assessment format which has gained validation for use as an index of neurobehavioral function in VS and MCS patients in the revised version of the JFK Coma Recovery Scale or ‘CRS-R’ [13]. Reflecting the tools already discussed, it rates responses to stimuli hierarchically across the auditory, visual, motor, oro-motor, communication and arousal behavioral domains. It can predict functional outcome, has diagnostic power and has been shown to be reliable over repeated measures [13].

With a different emphasis, the Lowenstein Communication Scale (LCS) assesses basic, pre-linguistic skills and communicative awareness during recovery in patients in MCS after traumatic brain injury [4]. Its strength is the capacity to assess both oral and alternative communication, enabling the patient with an optimal means for communication at the earliest point. The scale measures communicative function in a hierarchical system across five domains; mobility, respiration, visual responsiveness, auditory comprehension and verbal or alternative communication. In an examination of its predictive power the LCS was found to be reliable with good inter-rater agreement, although limited operational definitions and guidelines for treatment protocol are provided [4]. In summary, all of the tools reviewed consistently include the behavioral domains of arousal, communication, motor skills, auditory and visual responsiveness. Additionally, all tools provide data for changes in patient response over time, which is important in terms of examining progress and demonstrating change [11]. However, the auditory modality has been found to be particularly sensitive in identifying responses indicating awareness in VS patients [10], suggesting greater thought is needed about assessment of auditory responsiveness.

3. Assessment of auditory responsiveness: Factors for consideration In a study with 60 VS subjects, 66.7% of patients demonstrated awareness only within the auditory modality, with 25% of patients demonstrating awareness in the visual and auditory modalities simultaneously [11]. Visual impairment is considered a contributory factor in misdiagnosis [2] and so may explain the importance of the auditory modality in the assessment procedures with this population. Given the significance which the auditory domain appears to play in assessing these patients, this paper argues for assessment methods with greater sensitivity within this particular modality. Strong interdisciplinary working from a range of experienced professionals is widely advocated in the assessment and rehabilitation of this patient group [1–3]. Whilst each discipline brings particular skills to assessing a behavioral function, most behavioral functions require complex interdisciplinary working from a wide range of specialists and no one discipline can take sole responsibility for assessing an isolated function [1]. Music therapy is


321

able to provide a protocol for assessment of the auditory domain in which components of the auditory stimuli can be subtly modified. This can produce a detailed report of a patient’s responsiveness to his or her auditory environment. Where responsive, the patient has the opportunity for active participation in tasks. It can also contribute to a more informed management of the auditory environment, given that music is often used by well-meaning carers and relatives to ‘stimulate’ the patient [1] or to provide a passive form of leisure for the person who can no longer access leisure pursuits independently.

behaviors might include tapping a finger to the pulse of music; vocal sounds at the pitch of music being played; vocal sounds only for the duration of a piece of music. Each of these behaviors can contribute to an understanding of the patient’s awareness and attempts to interact purposefully. Recognizing the need to record such information accurately, consistently and in detail to assist with treatment planning, the music therapy assessment tool for low awareness states or ‘MATLAS’, was developed through prolonged engagement with interdisciplinary assessment of people in low awareness states.

4. Music therapy in the assessment of patients in low awareness states

5. A description of the MATLAS

Music therapy (MT) for patients in low awareness states offers planned sensory stimulation which involves the subtle manipulation of auditory components within the environment. A review of published literature indicates it is a useful clinical tool in stimulating a range of behavioral, physiological and expressive responses in patients in low awareness states [14,15]. As a treatment medium music provides a non-verbal medium which is inherently emotional in nature [21], using visually appealing objects such as guitars, keyboards and drums. Music can be a motivating medium for patients and can therefore be useful for assessing a range of functions [9]. However, similar to many other disciplines, it lacks a body of rigorous research to support its role in rehabilitation and also lacks valid and reliable tools for measuring its effects in brain injury rehabilitation. Furthermore, as yet there is no data to guide frequency and duration of intervention, nor clear guidance on indicators for cessation of treatment. Within music therapy sessions, live and recorded music is presented to assess a patient’s recovery of function in a less formalized environment. When using live music, parameters such as volume, pitch, melodic contour, rhythm and articulation can be modified to assess the patient’s behavioral responses to specific auditory information. In line with standard rehabilitation of this patient group, measurement of the patient’s behavioral responses within the music therapy context is an issue needing further refinement and rigorous measurement. Whilst existing tools such as the WHIM might be useful for recording behavioral responses in music therapy, they do not allow registration of small changes in the patient’s musical behaviors which can indicate other non-musical functioning. Examples of such musical

Developed over many years, the MATLAS has been found to have clinical value in interdisciplinary assessments and treatment planning. Understanding the contribution music therapy can make in assessment and treatment programs with this complex population depends on its integration with interdisciplinary assessment and its ability to report on behavioral functions which are relevant to the broader assessment of the patient [14]. The MATLAS met this objective by providing a standard assessment format which rates small changes of behavior in a hierarchical manner providing clinically relevant information in the management of complex cases. Music therapy offers a particular environment which is valuable for behavioral assessment over a range of behavioral domains and senses. In line with the existing tools for low awareness states, assessment is undertaken on repeated occasions. The MATLAS contains 14 items covering the five behavioral domains consistently included in other assessment formats; motor responses, communication, arousal, auditory and visual responsiveness. Each item is categorized hierarchically into levels of observed behavioral responses with a numerical grading. Each level specifies an observable behavior with definitions of behaviors provided in an accompanying instruction manual [16]. Within all items level zero rates ‘no response’, with the numerical rating of highest level responses varying between three and seven. As the assessment format allows differential rating of small increments of behaviors there is a greater number of levels of response within some items where behavioral responses can be categorized more sensitively. Within several items level one rates non-meaningful behaviors which are not contingent to the stimuli presented.

322


6. Items to assess the auditory domain Given music’s primary quality as an auditory medium, responses within the auditory domain are given particular attention within the MATLAS. The item ‘Responses to auditory stimuli’ records localizing and tracking behaviors, primarily providing information to the team about the patient’s hearing status. The auditory stimuli presented vary from simple single musical sounds to more complex musical stimuli which can be subtly manipulated in parameters such as pitch and volume. Another item, ‘Behavioral responses to music’, rates immediate spontaneous responses to musical stimuli such as changes in respiration and physical movements. In line with the reported effects of music therapy [14], observable behavioral changes in respiration rate, physical and facial gesture, arousal and eye opening should be monitored, particularly at the initial presentation of music. Throughout the published literature, changes in respiration, localizing and changes in emotional behavior are frequently noted in response to the presentation of live music [14]. If the patient is fully aroused but shows no behavioral changes at the initial outset of musical stimuli, experience has indicated that questions should be asked about the patient’s ability to hear or about awareness of the environment. A further item breaks down the auditory stimulus into musical parameters of ‘pulse/rhythm’, ‘melody/pitch’, ‘timbre’, ‘dynamics/intensity’, and ‘tempo’ with the assessor required to rate whether observed behaviors were related to a particular component. This item is able to provide detailed information about responses to specific parameters of the auditory environment. This has been instrumental for both treatment planning and making recommendations for the use of sound in the patient’s environment. For example, examining responses to pitch and volume can assist staff in considering the non-verbal musical aspects of their speaking voice when communicating with the patient whose comprehension of language is unknown.

7. Items to assess communicative intent The MATLAS differs from other assessment tools used with this population in relation to how it assesses communicative intent. There are five items relating to communication in all, encompassing pragmatic, musical and language-based communication. One of these items is an expanded item on vocal responses, important within the music therapy session where

a high incidence of spontaneous vocal sounds occurs within music-based social exchanges. If movement is impaired, vocal responses may also be the most available means for responding. An item for vocal responses documents all observed vocal responses across eight hierarchically organized levels ranging from ‘no vocalization to musical stimuli’ up to ‘sang all the words to familiar songs’. For this item, a level nine has been added to rate ‘Unable to vocalize’ when fibreoptic endoscopic evaluation has indicated impairment at an organic level which is unlikely to change. Other items falling under the domain of communication include ‘non-verbal communication’, ‘choice-making’ and ‘responses to verbal commands’. Using a non-language based medium which allows for mutual participation is useful for assessing features of social communication such as turn-taking [5]. Similarly, because of its motivational nature, the music therapy session can assist with the assessment of the early signs of communicative intent, such as following one-step commands rated under ‘Response to verbal commands’, which includes commands such as ‘Look at the guitar’ or ‘Touch the drum’. Strategies for communicating preference by means of eye gaze can be reinforced using a musical reward. Similarly, reinforcing the developing use of ‘yes/no’ in line with interdisciplinary communication strategies can be assessed and reinforced using musical stimuli known to have personal salience, such as familiar songs or particular instrumental sounds.

8. A music therapy assessment scale in the context of existing scales The MATLAS has produced useful clinical information at a local level over sixteen years of use. A pilot study to examine its validity found strong concurrent validity with the SMART and the WHIM scales, both validated and widely accepted for use with this population [7]. A larger validity and reliability study is now in progress which will test both intra and inter-rater reliability as well as measuring item discrimination in order to refine the tool [6]. Development of a standard protocol which can be repeated across assessors remains a challenge and at current time is in the final stages of refinement [6]. Shiel et al. [18] criticize existing tools on their poor relationship with everyday life, poor predictive power, lack of usefulness in designing and guiding treatment programs, and their lack of ability to detect subtle changes in functioning. The MATLAS does not


attempt to address all of these factors, however, it has been found a useful tool for the assessment of such a complex patient group and in guiding interdisciplinary treatment. It is not designed to be used in everyday life as the information it generates about auditory awareness is particularly finely tuned. It has evolved to assist with detection of subtle changes over repeated measures. It is a tool for use during specific stimuli by a qualified practitioner and requires training to use in addition to specialist knowledge of the patient population. It is anticipated that the MATLAS may assist in measuring the effects of music therapy intervention with this complex patient group in future investigations. Optimizing the patient’s potential performance is significant in the assessment of the patient in low awareness states. It has been observed that impaired motor function following profound damage may limit the patient’s performance within a task, risking a lower rating overall of cognitive awareness [10], with recommendations to offer a wider range of alternative methods of stimuli to enable the patient to show their potential for functional ability. In music therapy assessment, participation is possible and measurable through vocalization, bypassing physical functioning which has been found to confound accurate scoring in other assessment methods. Wilson et al. [23] suggest that providing ‘alternative means for VS patients to express their abilities’ through a variety of treatments may lead to more refined perceptions of the condition. Multimodal stimulation has been found to produce greater behavioral changes than unimodal stimuli, with personally salient stimuli in multimodal stimulation producing the greatest changes of all [23]. The use of meaningful stimuli in the attempt to elicit behaviors is endorsed by others [15, 19]. Using music which is known to be meaningful to the individual provides a non-verbal, emotionally significant stimulus in the modality which has been found to enable optimal demonstration of awareness [5,9,14, 15]. The call for multimodal stimulation supports a music therapy assessment which uses visual and tactile sensory stimuli rather than merely auditory stimuli alone. The MATLAS places a heavy prominence on communication function, with five items within this domain. It should be emphasized that music and language are related within the human condition at the earliest stages of life, and continue to have a complex relationship throughout the life span [8]. However, regaining communication ability has been identified as significant in providing information about the regaining of

323

awareness in this group [4]. Rehabilitation must provide the patient with the optimal means for communication at the earliest stages. As music therapy can offer such a forum, it is essential to develop means by which to measure this objectively. The presentation of live music in music therapy often prompts spontaneous vocal responses which can indicate the intent to communicate immediate feeling states or participate within a non-verbal musical interaction. Using the voice for spontaneous communication of this nature is innate in the human condition [8]. Indeed, the WHIM scale has ‘volitional vocalization, to express feelings’ as item six, with ‘Vocalizes to express mood or needs’ appearing later as item 20 [20]. Within the music therapy assessment, vocalization is an independent item which has been considerably extended to maximize sensitivity. This was developed as, for the patient who has no active movement, vocalizing may be the only active intentional behavior which can be demonstrated. Whereas the WHIM does not require any assumptions to be made about the purpose of behaviors, the MATLAS does make analyses using operational definitions of behavior being ‘goal-directed’. Thus, differential comparisons are made of physical responses to contrasting stimuli. For example, reaching a hand and upper limb out in a forward flexion movement towards a guitar to strum can be differentiated from flexing and extending at the wrist to beat a drum. Similarly, vocalizations can be measured in terms of their musical qualities such as pitch, melodic contour, volume and duration and compared to the musical stimuli being used at the time the response was elicited. Similar to WHIM, the music therapy scale records change over repeated measures but does not relate this to treatment or to natural recovery. The important factor is to rate the patient’s responsiveness at that moment to defined stimuli in order to contribute to interdisciplinary observations of behavior and assist with goal-planning. Determining the reliability and validity of the assessment is now a priority.

9. Conclusion Assessment of the patient in low awareness states consistently includes five relevant behavioral domains. Of these, the auditory domain has shown greatest sensitivity in revealing responses indicating awareness [10]. Sensitivity of assessment of this population pertains not only to the tools which are available for use, but the treatment medium and protocols as well [9,5,10,19].

324


Enabling means for communicative ability should be prioritized in assessment of these patients at the earliest stages [4]. Music therapy as a treatment can provide a forum for non-formalized assessment of a range of functions, and in particular of communicative ability. The MATLAS was developed over many years in interdisciplinary assessment and has been found useful in rating responses across the five behavioral domains to produce information relevant to interdisciplinary assessment. A planned reliability and validity study of this scale is now a priority to enhance the range of reliable assessment tools available for use with this complex patient group [6].

[8]

[9] [10]

[11]

[12]

[13]

Acknowledgements The author would like to acknowledge financial support from the Neuro-disability Research Trust and the Department of Health NHS Research and Development funding for the preparation of this paper. The views expressed in this publication are those of the author and not necessarily those of the NHS Executive. The author would like to thank Dr. Barb Daveson for her helpful comments on an earlier draft of this paper.

[14]

[15]

[16]

[17]

References [1]

[2]

[3]

[4]

[5] [6]

[7]

K. Andrews, Rehabilitation practice following profound brain damage, Neuropsychological Rehabilitation 15(3–4) (2005), 461–472. K. Andrews, L. Murphy, R. Munday and C. Littlewood, Misdiagnosis of the vegetative state: retrospective study in a rehabilitation unit, British Medical Journal 313 (1996), 13–16. B.J. Ansell and M.A. Keenan, The Western Neuro Sensory Stimulation Profile: A tool for assessing slow to recover head injured patients, Archives Physical Medical Rehabilitation 70 (1989), 104–108. N. Borer-Alafi, M. Gil, L. Sazbon and C. Korn, Loewenstein communication scale for the minimally responsive patient, Brain Injury 16(7) (2002), 593–609. L. Bunt, Music Therapy: An Art beyond Words, Routledge, London, 1994. B. Daveson and W.L. Magee, A validation and reliability study of a musical assessment tool for low awareness states (MATLAS): Research Proposal, Royal Hospital for Neuro-disability, London, 2005. B. Daveson, W.L. Magee, L. Crewe, G.J. Beaumont and P. Kenealy, A pilot study to investigate the validity of a music assessment tool for low awareness states, (MATLAS) in Patients with Complex Neuro-disability, Under review.

[18]

[19]

[20]

[21] [22]

[23]

I. Deliege and J. Sloboda, eds, Musical Beginnings: Origins and Development of Musical Competence, Oxford University Press, Oxford, 1996. E.T. Gaston, ed., Music in Therapy, Macmillan Publishing Co. Inc., New York, 1968. H. Gill-Thwaites, The Sensory Modality Assessment Rehabilitation Technique – A tool for assessment and treatment of patients with severe brain injury in a vegetative state, Brain Injury 11(10) (1997), 723–734. H. Gill-Thwaites and R. Munday, The sensory modality assessment and rehabilitation technique (SMART): a valid and reliable assessment for vegetative state and minimally conscious state patients, Brain Injury 18(2) (2004), 1255–1269. S. Horn, A. Shiel, D.L. McLellan, M. Campbell, M. Watson and B.A. Wilson, A review of behavioural assessment scales for monitoring recovery in and after coma with pilot data on a new scale of visual awareness, Neuropsychological Rehabilitation 3 (1993), 121–137. K. Kalmar and J.T. Giacino, The JFK Coma Recovery Scale – Revised, Neuropsychological Rehabilitation 15(3–4) (2005), 454–460. W.L. Magee, Music therapy with patients in low awareness states: assessment and treatment approaches in multidisciplinary care, Neuropsychological Rehabilitation 15(3–4) (2005), 522–536. W.L. Magee, Music as a diagnostic tool in low awareness states: Considering limbic responses, Brain Injury 21(6) (2007), 593–599. W.L. Magee and B. Daveson, Music Therapy Assessment Tool for Low Awareness States (MATLAS): Assessment Manual. Instructions for Use, Royal Hospital for Neuro-disability, London, 2005. D. Malkmus, B. Booth and C. Kodimer, Rehabilitation of the head injured adult. Comprehensive cognitive management. Professional staff. Association of Rancho Los Amigos Hospital, Inc., Downey CA, 1980. A. Shiel, S.A. Horn, B.A. Wilson, M.J. Watson, M.J. Campbell and D.L. McLellan, The Wessex Head Injury Matrix (WHIM) main scale: a preliminary report on a scale to assess and monitor patient recovery after severe head injury, Clinical Rehabilitation 14 (2000), 408–416. A. Shiel and B.A. Wilson, Can behaviours observed in the early stages of recovery after traumatic brain injury predict poor outcome? Neuropsychological Rehabilitation 15(3–4) 2005, 494–502. A. Shiel, B.A. Wilson, L. McLellan, S. Horn and M. Watson, WHIM The Wessex Head Injury Matrix – Manual, Thames Valley Text Company Limited: Bury St Edmunds, England, 2000. J. Sloboda, Music Structure and Emotional Response: Some Empirical Findings, Psychology of Music 19 (1991), 110–120. B.A. Wilson, Behavioural assessment and rehabilitation techniques: Foreword, Neuropsychological Rehabilitation 15(3– 4) (2005), 428–430. S. Wilson, G. Powell, D. Brock and H. Thwaites, Vegetative state and responses to sensory stimulation: an analysis of 24 cases, Brain Injury 10(11) (1996), 807–818.