THE CLINICAL UTILITY OF THE VIVOSONIC INTEGRITY CHILDREN WITH CEREBRAL PALSY

THE CLINICAL UTILITY OF THE VIVOSONIC INTEGRITY CHILDREN WITH CEREBRAL PALSY
THE CLINICAL UTILITY OF THE VIVOSONIC INTEGRITY
AUDITORY BRAINSTEM RESPONSE SYSTEM IN
CHILDREN WITH CEREBRAL PALSY
by
Christine van der Westhuizen
August 2010
Supervisor:
Co-supervisor:
Prof D Swanepoel
Prof J Hall
In partial fulfillment of the requirements for the degree
M.Communication Pathology in the Department of
Communication Pathology
Faculty of Humanities
University of Pretoria
© University of Pretoria
Abstract
Determining auditory functioning in difficult-to-test populations such as cerebral
palsy (CP) remains a challenge in paediatric audiology. The auditory brainstem
response (ABR) is favoured as the procedure to assess auditory functioning in
difficult-to-test populations such as CP.
The CP population, however, offers
unique challenges for the ABR procedure due to the presence of involuntary
muscular movements that may compromise the signal-to-noise ratio (SNR) of the
ABR.
Conventional ABR technology attempts to improve the SNR by the
modification of acquisition parameters e.g. adjusting the low cut filter or
implementing stricter artifact rejection criteria. However, such modifications may
compromise the waveform morphology of the ABR. Furthermore, sedation or
general anesthesia can also be used to improve the SNR by reducing excessive
muscular movements.
The CP population, however, displays a high risk for
developing upper airway obstruction when being sedated or anesthetized. Thus,
the feasibility and reliability of the conventional ABR may be compromised when
being employed in the CP population. In recent years a novel ABR system, the
Vivosonic Integrity (VS) ABR has become clinically available. The device
incorporates features such as pre-amplification of the ABR signal, Kalman
filtering and wireless recording.
These features promise to address the
limitations of conventional ABR technology to obtain a reliable recording in the
midst of excessive myogenic artifact. The aim of this study was therefore to
evaluate the clinical utility of the VS system when assessing a sample of children
with CP without the use of sedation. The clinical utility of the VS ABR system
was determined by comparing its success rates, the threshold correspondence to
behavioural pure tone (PT) thresholds and recording time to a conventional ABR
system when using click and 0.5 kHz TB stimuli.
A cross-sectional within-subject comparison research design was selected in
order
to
compare
thresholds
obtained
with
different
procedures.
The
experimental part of this study was represented by the within-subject control
condition where the VS ABR system and the conventional ABR system were
simultaneously conducted in each subject. This unique setup was important in
the research as equivalent test conditions in terms of EEG and environmental
conditions had to be ensured for both ABR systems. 15 CP subjects between
the ages of 12 and 18 years were included in the project.
A diagnostic
audiological test battery including immittance, distortion product otoacoustic
emissions and behavioural audiometry was conducted on each subject prior the
administration of the ABR procedures. The variability of the audiological test
battery results – between the subjects and when compared to previous research
– emphasized the heterogeneity of the CP population. Furthermore, more than
half of the research sample (53%; n=15) responded inconsistently to behavioural
pure tone (PT) stimuli. It was suggested that the severity of physical impairments
as well as additional impairments such as mental retardation might have
influenced the consistency of the subjects’ responses during behavioural PT
audiometry. The ABR results indicated that there were no significant differences
with regards to threshold correspondence and recording time between the two
ABR systems when using click and 0.5 kHz TB stimuli (p>0.05). With regards to
the success rates, the VS system was successful in more cases than the
conventional ABR system using click and 0.5 kHz TB stimuli. Although results
also showed no statistically significant value for click (p=.1121) and 0.5 kHz TB
stimuli (p=.1648), there was a tendency towards the 95% confidence level in both
cases suggesting that the VS ABR system may produce a statistically significant
success rate for click as well as for 0.5 kHz TB stimuli, provided a larger sample
is tested. The research indicated that, since the VS ABR system was more
successful across a wider range of subjects during click-evoked and 0.5 kHz TB
recordings, it may increase the clinical usefulness of the ABR especially in terms
of hearing screening in the CP population.
The research suggested that
excessive muscular movements during the recordings influenced not only the VS
ABR’s, but also the conventional ABR’s threshold correspondences to PT
thresholds as well as the recording time of the measurements. Therefore it may
still be necessary to use a light sedative in some CP patients to reduce excessive
myogenic interference despite the possible advantages of the VS ABR system.
Key words: Difficult-to-test, cerebral palsy, objective audiometry, auditory
brainstem response, signal-noise-ratio, feasibility, threshold correspondence,
recording time, Vivosonic Integrity ABR system, conventional ABR system
TABLE OF CONTENTS
TABLE OF CONTENTS……………………………………………………………..1
LIST OF FIGURES…………………………………………………………………...6
LIST OF TABLES…………………………………………………………………….9
LIST OF APPENDICES…………………………………………………………….12
1.
INTRODUCTION AND ORIENTATION……………………………………13
1.1
Introduction……………………………………………………………….13
1.2
Background……………………………………………………………….15
1.3
Rationale…………………………………………………………………..18
1.4
Problem statement………………………………………………………..21
1.5
Outline of the chapter contents………………………………………….24
1.6
Conclusion…………………………………………………………………24
1.7
Summary…………………………………………………………………...26
2.
A CRITICAL PERSPECIVE ON AUDITORY ASSESSMENT IN THE
CEREBRAL PALSY POPULATION………………………………………..27
2.1
Introduction………………………………………………………………..27
2.2
Background to cerebral palsy…………………………………………...27
2.2.1 Historical perspective on cerebral palsy…………………………….27
2.2.2 Prevalence and definition of cerebral palsy………………………...29
2.2.3 Risk factors associated with cerebral palsy………………………...30
2.2.4 Classification of cerebral palsy……………………………………….31
2.2.5 Cerebral palsy in the South African context…………………………35
2.3
Cerebral palsy and hearing loss…………………………………………37
2.4
Importance of early identification of hearing loss in children with
cerebral palsy……………………………………………………………...41
2.5
Auditory assessments and cerebral palsy……………………………...42
2.5.1 Behavioural audiometry and cerebral palsy………………………....42
2.5.2 Objective audiometry and cerebral palsy………………………….....45
2.6
The auditory brainstem response procedure in auditory
assessment…………………………………………………………………47
2.7
The auditory brainstem response as assessment method in the
cerebral palsy population…………………………………………………..51
2.7.1 Improving the signal-to-noise ratio in the conventional auditory
brainstem response system………………………............................55
2.7.2 Cerebral palsy and conventional auditory brainstem response
technology………………………………………………………………..59
2.7.3 Improving the signal-to-noise ratio by implementing patient
management techniques………………………………………………..61
2.7.4 Novel technology for improving ABR signal-to-noise ratio………….63
2.8
Conclusion…………………………………………………………………..66
2.9
Summary…………………………………………………………………….66
3.
METODOLOGY…………………………………………………………………68
3.1
Introduction…………………………………………………………………..68
3.2
Aims of the research………………………………………………………..68
3.2.1 Main aim…………………………………………………………………..68
3.2.2 Sub-aims………………………………………………………………….68
3.3.
Research design……………………………………………………………69
3.4
Ethical considerations…………………………………………………......70
3.4.1 Respect of privacy of research subjects ……………………………..71
3.4.2 Informed consent………………………………………………………..71
3.4.3 Beneficence and non-malfeasance……………………………….......72
3.5
Research sample………………………………………………………….72
3.5.1
Selection procedures…………………………………………………..72
3.5.2
Description of the research sample…………………………………..74
3.6 Material and apparatus………………………………………………………74
3.6.1
Material and apparatus for subject selection………………………..74
3.6.1.1 Otoscopic examination ………………………………………….74
3.6.1.2 Tympanometry…………………………………………………….77
3.6.1.3 Case history……………………………………………………….77
3.6.2
Material and apparatus used for data collection……………….. …77
3.6.2.1 Ipsilateral acoustic reflexes…………………………………….77
3.6.2.2 Distortion product otoacoustic emissions…………………….78
2
3.6.2.3 Behavioural pure tone audiometry…………………………….79
3.6.2.4 Auditory brainstem response…………………………………..79
3.7
Procedures………………………………………………………………..83
3.7.1 Data collection procedures: immittance, distortion product
otoacoustic emissions and behavioural pure tone
audiometry. …………………………………………………………….84
3.7.1.1 Immittance……………………………………………………....84
3.7.1.2 Distortion product otoacoustic emissions…………………….84
3.7.1.3 Behavioural pure tone audiometry…………………………....85
3.7.2
Data collection procedures: auditory brainstem response………...85
3.7.3
Procedures for analysis of auditory brainstem response……….....90
3.7.4
Procedures for analysis of the recording time of the
auditory brainstem response recordings………………………….....90
3.8
Data processing………………………………………………………......90
3.9
Data analysis………………………………………………………………91
3.9.1
Describing the feasibility and the characteristics of the auditory
procedures…………………………………………………………….....91
3.9.2
Comparison between the VS and BL ABR systems in terms
of the feasibility, correspondence to behavioural PT thresholds
and recording time…………………………………………………......92
3.9.2.1
Feasibility using click and 0.5 kHz tone burst (TB)
stimuli……………………………………………………………..92
3.9.2.2
Electrophysiological thresholds using click and 0.5 kHz
TB stimuli………………………………………………………..92
3.9.2.3
Threshold correspondence to behavioural PT
thresholds …………………………………………………….....92
3.9.2.4
Recording time using click and 0.5 kHz TB stimuli………....94
3.10
Reliability and validity……………………………………………………95
3.11
Conclusion………………………………………………………………....96
3.12
Summary…………………………………………………………………...96
4.
RESULTS…………………………………………………………………….. .98
3
4.1
Introduction…………………………………………………………………98
4.2 Results for sub-aim 1: Characteristics and feasibility of an
audiological test battery in children with cerebral palsy………………..98
4.2.1
Immittance measurements…………………………………………...99
4.2.2
Distortion product otoacoustic emissions………………………......103
4.2.3
Behavioural pure tone audiometry…………………………………..108
4.3 Results for sub-aim 2: Comparing the Vivosonic Integrity
ABR system with a conventional ABR system…………………………..111
4.3.1
Feasibility of the VS and BL ABR systems using click and
0.5 kHz TB stimuli………………………………………………….....111
4.3.2 Electrophysiological thresholds of the VS and BL ABR
systems using click and 0.5 kHz TB stimuli…………………………115
4.3.3 Threshold correspondence of the VS and BL ABR systems
to behavioural PT thresholds………………………………………..117
4.3.3.1
Threshold correspondence at 2 kHz, 4 kHz and the
average of 2 kHz and 4 kHz…………………………………119
4.3.3.2
Threshold correspondence at 0.5 kHz………………….....126
4.3.4 Recording time using click and 0.5 kHz TB stimuli……………....130
4.4
Conclusion………………………………………………………………..133
4.5
Summary………………………………………………………………….134
5.
DISCUSSION………………………………………………………………..135
5.1
Introduction………………………………………………………………...135
5.2 Sub-aim 1 discussion: Characteristics and feasibility of an
audiological test battery in children with cerebral palsy ………………137
5.2.1
Immittance measurements………………………………………….137
5.2.2
Distortion product otoacoustic emissions…………… …………..141
5.2.3
Behavioural pure tone audiometry………………………………...145
5.3
Sub-aim 2 discussion: Comparing the Vivosonic Integrity ABR
system with a conventional ABR system ……………………………...146
5.3.1
Feasibility of the VS and BL ABR systems using click and
0.5 kHz TB stimuli………............................................................146
4
5.3.2
Electrophysiological thresholds of the VS and BL ABR
systems using click and 0.5 kHz TB stimuli…………………………157
5.3.3
Threshold correspondence of the VS and BL ABR systems
to behavioural PT thresholds…………………………………………161
5.3.3.1
Threshold correspondence at 2 kHz, 4 kHz and the
average of 2 kHz and 4 kHz…………………………………161
5.3.3.2
5.3.4
Threshold correspondence at 0.5 kHz……………………..164
Recording time of the VS and BL ABR systems using click
and 0.5 kHz TB stimuli………………………………………………..165
5.4
Conclusion………………………………………………………………….167
5.5
Summary……………………………………………………………………161
6. CONCLUSIONS AND RECOMMENDATIONS……………………………..169
6.1 Introduction……………………………………………………………………169
6.2 Conclusions……………………………………………………………………169
6.2.1
Audiological tests in children with cerebral palsy……………..........170
6.2.2
ABR assessments in the CP population using the VS ABR
system……………………………………………………………………170
6.3 Implications of the findings…………………………………………………..172
6.4 Critical evaluation of study…………………………………………………...174
6.5 Recommendations for further research…………………………………….176
6.6 Final conclusion……………………………………………………………….177
REFERENCES……………………………………………………………………….179
5
LIST OF FIGURES
Figure 1.1: Principles of conventional ABR technology .................................19
Figure 1.2: Alternative setting arrangement of the VS ABR device…………..24
Figure 2.1: Egyptian carving of a person with right hemiplegia (Adapted
from Scherzer, 2001) ………………………………………………………...27
Figure 2.2: Distribution of the main CP types (Adapted from: Wilson-Jones,
2007; Cogher et al., 2002; Mechem, 2002)…………………….................34
Figure 2.3: Anatomic generators of the different components of the ABR
(Adapted
from: Hall, 2007; Arnold, 2000; Rowe, 1981)………………......48
Figure 3.1: Vanderbilt 65/55 95-5th percentile normative values .....................78
Figure 3.2: Positions of the ABR systems………………………………………..86
Figure 3.3: Position of the ABR electrodes as seen from behind……………...86
Figure 3.4:Position of the ABR electrodes as seen from the front……………..87
Figure 3.5: Position of the inverting and ground electrodes of the VS and
BL ABR systems as seen from the side………………………………….....88
Figure 3.6: Position of the Fz electrodes, as well as the ER- 3A insert
earphones each ABR system……………………………………………….89
Figure 3.7: Organization of threshold differences data for statistical
analysis ....................................................................................................93
Figure 4.1: Distribution of right ipsilateral acoustic reflexes ............................101
Figure 4.2: Distribution of left ipsilateral acoustic reflexes……………………...101
Figure 4.3: Noise floor levels attained during DPOAE measurements…………103
Figure 4.4: Distribution of distortion product emissions in the current
research as specified by distortion product (DP) - Noise Floor (NF)
criteria .......................................................................................................104
Figure 4.5: DPOAEs with elevated amplitudes in at least one ear of
a subject………………………………………………………………………...105
Figure 4.6: Subjects in Group A: Mean behavioural PT thresholds
for 0.5 kHz to 4 kHz……………………………………………………….......110
Figure 4.7: Subjects in Group B: Mean behavioural PT thresholds
6
for 0.5 kHz to 4 kHz.................................................................................110
Figure 4.8: Successful click-evoked and 0.5 kHz TB ABR recordings
using the VS and BL systems…………………………………………………113
Figure 4.9: Percentage of rejected sweeps in the BL ABR system
in subjects with unsuccessful ABR recordings using click and
0.5 kHz TB stimuli…………………………………………………………......115
Figure 4.10: Threshold differences between the click-evoked
thresholds (VS and BL ABR systems) and the 2 kHz behavioural
PT threshold…………………………………………………………………...124
Figure 4.11: Threshold differences between the click-evoked thresholds
(VS and BL ABR systems) and the 4 kHz behavioural PT
threshold………………………………………………………………………..124
Figure 4.12: Threshold differences between the click-evoked thresholds
(VS and BL ABR systems) and the average of the 2 kHz and
4 kHz behavioural PT thresholds.............................................................126
Figure 4.13: Threshold differences between the 0.5 kHz TB thresholds
(VS and BL ABR systems) and the 0.5 kHz behavioural PT
threshold……………………………………………………………………….129
Figure 4.14: Recording time per ear for VS and BL ABR systems using
click stimuli…………………………………………………………………….131
Figure 4.15: Recording time per ear for VS and BL ABR systems using
0.5 kHz TB stimuli…………………………………………………………….132
Figure 4.16: Mean recording time per ear for VS and BL ABR systems
using click and 0.5 kHz TB stimuli…………………………………………..133
Figure 5.1: An example of a feasible ABR recording using click
stimuli with the VS ABR system………………………………………….....146
Figure 5.2: An example of a feasible ABR recording using click stimuli
with the BL ABR system…………………………………………………….147
Figure 5.3: An example of an unsuccessful ABR recording using click
stimuli (a) and 0.5 kHz TB stimuli (b) with the BL ABR system…………155
Figure 5.4: An example of an unsuccessful ABR recording using click
7
stimuli (a) and 0.5 kHz TB stimuli (b) with the VS ABR system…………156
Figure 5.5: Illustration of the behavioural PT thresholds and ABR
thresholds obtained (VS and BL ABR systems) obtained
in Subject 5……………………………………………………………………160
Figure 5.6: Behavioural PT thresholds (2 kHz and 4 kHz) and
click-evoked thresholds (VS and BL ABR systems) obtained
in Subject 9……………………………………………………………………162
Figure 5.7: Behavioural PT thresholds (2 kHz and 4 kHz) and
click-evoked thresholds (VS and BL ABR systems) obtained
in Subject 6……………………………………………………………………163
Figure 5.8: Behavioural PT thresholds (2 kHz and 4 kHz) and
click-evoked thresholds (VS and BL ABR systems) obtained
in Subject 12………………………………………………………………….163
8
LIST OF TABLES
Table 1.1: Novel features of the VS ABR system………………………………..22
Table 1.2: Summary of dissertation contents by chapter…………………….....25
Table 2.1: Risk factors associated with CP………………………………………32
Table 2.2: Classification of CP (Adapted from: Wilson-Jones, 2007;
Cogher et al., 2002; Mechem, 2002; McDonald, 1987; Newton, 1977;
Minear, 1956)…………………………………………………………………...33
Table 2.3: Incidence of hearing loss in the CP and normal populations………38
Table 2.4: Outcomes of behavioural audiometric techniques in the
CP population…………………………………………………………………..44
Table 2.5: Acquisition parameters of click-evoked and TB ABR recordings
(Adapted from Hall, 2007)…………………………………...........................51
Table 2.6: Summary of previous studies in which the ABR procedure was
conducted in the population with multiple disabilities………………………53
Table 2.7: Principles and limitations of conventional ABR technology………...57
Table 2.8: Principles and supposed advantages of the VS ABR system……...65
Table 3.1: Selection criteria for the research sample…………………………….73
Table 3.2: Description of research sample………………………………………..75
Table 3.3: Normative tympanometric values………………………………….. …77
Table 3.4: Criteria for DP/NF difference…………………………………………...79
Table 3.5: Summary of methodology followed to obtain reference values
for VS and BL ABR systems……………………………………………………80
Table 3.6: Acquisition parameters for the VS and BL ABR systems…………...83
Table 3.7: Stimuli parameters for the VS and BL ABR systems………………..84
Table 4.1: A summary of the feasibility of the various auditory procedures
employed in the current research...............................................................100
Table 4.2: Summary of ipsilateral acoustic reflexes obtained in different CP
sub-groups of the research sample…………………………………………..102
Table 4.3: The difference between the DPOAE amplitude (dB SPL) of the
left ear and the 95th percentile of the Vanderbilt criteria at various test
9
frequencies of subjects in whom elevated emissions were obtained…….106
Table 4.4: The difference between the DPOAE amplitudes (dB SPL) of the
right ear and the 95th percentile of the Vanderbilt criteria at various test
frequencies of subjects in whom elevated emissions were obtained……..107
Table 4.5: Behavioural PT thresholds (in dB HL) for each subject (n=15)…….108
Table 4.6: Consistency of behavioural PT responses (0.5 kHz to 4 kHz)
(n=15 subjects)…………………………………………………………………109
Table 4.7: Information related to the state of awareness, sound level
and inter-aural latency differences (n=15 subjects)…………………………112
Table 4.8: Subjects with unsuccessful ABR recordings………………………...114
Table 4.9: The distribution of the ABR thresholds (dB nHL) of the VS
ABR and BL ABR systems using click and 0.5 kHz TB stimuli……………116
Table 4.10: The mean ABR thresholds and SD (dB nHL) of the VS ABR
BL ABR systems using click stimuli (n=12) and 0.5 kHz TB
stimuli (n=11).............................................................................................117
Table 4.11: Distributions of threshold differences (click-evoked ABR
threshold – PT threshold) for the VS and BL ABR systems at 2 kHz,
4 kHz and the average of 2 kHz and 4 kHz………………………………...120
Table 4.12: Mean threshold differences at 2 kHz, 4 kHz and the average
for 2 kHz and 4 kHz for subjects in Group A (n=7)…………………………121
Table 4.13: Mean threshold differences at 2 kHz, 4 kHz and the average
of 2 kHz and 4 kHz for subjects in Group A and B sample (n=12)……….122
Table 4.14: Range of threshold differences of ABR systems utilizing click
stimuli for subjects in Group A and subjects in Group A and B…………..123
Table 4.15: Normal distributions of threshold differences (0.5 kHz TB
ABR threshold – behavioural PT threshold at 0.5 kHz) for the VS
and BL ABR systems at 0.5 kHz……………………………………………..127
Table 4.16: Mean threshold differences at 0.5 KHz for subjects in Group A
as well as subjects in Group A and B………………………………………..128
Table 4.17: Range of threshold differences of ABR systems using 0.5 kHz
TB stimuli for subjects in Group A and subjects in Group A and B……….129
10
Table 4.18: VS ABR recording time per ear for subjects in whom ABR
assessments using the BL ABR device were not successful………………130
Table 5.1: Summary of previous assessment procedures employed
in the population with multiple disabilities…………………………………....138
Table 5.2: Summary of OAE research involving the CP population…………...142
Table 5.3: A review of the percentage successful ABR recordings
conducted in CP populations………………………………………………....148
Table 5.4: Potential modifications of the acquisition parameters during
ABR assessments in the CP population……………………………………..150
Table 5.5: A summary of different ABR systems and parameters
implemented in previous studies in populations with multiple
disabilities………………………………………………………………………153
Table 5.6: Summary of ABR results in previous studies……………………......158
Table 5.7: The actual and suggested recording time for VS and BL ABR
systems using click and 0.5 kHz TB stimuli…………………………………166
11
LIST OF APPENDICES
APPENDIX A
APPROVAL BY ETHICS COMMITTEE OF THE
UNIVERSITY OF PRETORIA
APPENDIX B
APPROVAL BY THE GAUTENG DEPARTMENT OF
EDUCATION
APPENDIX C
CONSENT LETTER TO THE PRINCIPAL OF THE
PRETORIA SCHOOL FOR CHILDREN WITH CEREBRAL
PALSY
APPENDIX D
INFORMED CONSENT LETTER (PARENTS)
APPENDIX E
VERBAL ASSENT (SUBJECTS)
APPENDIX F
RATING SCALE FOR SUBJECTS’ AWARENESS DURING
ABR RECORDINGS
APPENDIX G
INFORMED CONSENT LETTER (NORMATIVE SAMPLE)
12
Chapter 1
INTRODUCTION AND ORIENTATION
The aim of this chapter is to introduce the research question, to provide the
rationale for the study, to explain the terminology used, and to present an
overview of the content and the organization of the study.
1.1
Introduction
‘A disabled child has the right to enjoy a full and decent life, in conditions which
ensure dignity, promote self-reliance, and facilitate the child’s active participation
in the community’ (UN Convention on the Rights of the Child, 1989).
This
statement endorses the right of each disabled child to embrace his/her life to the
fullest, implying that disabilities need to be identified and addressed
appropriately.
A disability is defined as ‘a physical or mental condition that limits a person’s
movements, senses or activities’ (South African Concise Oxford Dictionary,
2002). Cerebral palsy (CP) is viewed mainly as a physical disability which affects
an individual’s movements, senses and activities of daily living to varying
degrees depending on the type and severity of the condition (Armstrong, 2007;
Beckung & Hagberg, 2002; Hutton & Pharoah, 2002). These disabling conditions
have far reaching implications not only for the CP child, but also for his/her family
and the immediate community as self-reliance and active participation in the
community is reduced to a significant degree.
The term cerebral palsy refers to a group of disorders of the central nervous
system that result in abnormal control of movement or posture (Lawson &
Badawi, 2003; Stanton, 1992).
These disorders are the result of a lesion(s)
before birth (prenatally), during birth (perinatally) or after birth (postnatally), prior
to the developing brain reaching maturation (Mechem, 2002).
Although the
13
lesion to the brain is irreversible, CP is regarded as a non-progressive disorder
as the neuromuscular symptoms do not usually degenerate over a period of time
(Andersen, Irgens, Haagaas, Skranes, Meberg & Vik, 2008; Beckung & Hagberg,
2002; Hutton & Pharoah, 2002; Mechem, 2002; Stanton, 1992).
A variety of neuromuscular symptoms can occur within the population with CP
and a classification system was proposed by Minear (1956) in order to elucidate
these symptoms.
According to this classification, the diverse symptoms are
described in terms of the place of disability (e.g. quadriplegia); the type of
disability (e.g. spasticity) and the severity of the disability (e.g. mild) (Mechem,
2002; Minear, 1956). Recent literature simplifies this classification to a certain
extent and suggests that CP can be classified according to a description of the
motor characteristics and the limb involvement on the one hand, and the place of
lesion i.e. in pyramidal and extrapyramidal nerve pathways on the other (WilsonJones, Morgan & Shelton, 2007). As CP is a diverse medical condition the
utilization of a classification system provides a platform of knowledge from where
medical personnel including therapists such as occupational therapists and
speech therapists can commence intervention in order to address not only the
motor disabilities of the condition, but other accompanying disabilities as well.
Although CP is mainly a motor impairment, the condition is often accompanied by
associated musculoskeletal problems that are secondary to the brain lesion
(O’Shea, 2008; Beckung & Hagberg, 2002; Mechem, 2002). These impairments
include dysphagia, speech and language difficulties, auditory dysfunction, vision
and cognitive impairments, perceptual and behavioural problems as well as
epilepsy (O’Shea, 2008; Workinger, 2005; Mechem, 2002; Stanton, 1992).
Recent Norwegian data illustrated that only 28% of children with CP did not
display any associated impairments (Andersen et al., 2008), emphasizing the
importance of a holistic approach in the treatment of a child with CP. The fact
that the related impairments may vary in degree and nature over a period of time
(e.g. progressive degeneration is possible), stresses the importance of
considering and assessing each associated impairment to ensure appropriate
14
early intervention services (Kennes, Rosenbaum, Walter, Russel, Raina, Bartlett
& Galuppi, 2002; Workinger, 2005).
Associated sensory impairments, in particular hearing loss, occur with relative
high incidence in the CP population (Mechem, 2002; Northern & Downs, 2002;
Zafeiriou, Andreou & Karasavidou, 2000).
Although there are discrepancies
regarding the exact incidence of additional hearing loss in this population, data
suggest that auditory disorders may occur in 1% to 25% of this population
(Fawke, 2007; Kolker, 2004).
Auditory dysfunction disorders refer to pathology in any part of the auditory
system.
This would include a conductive hearing loss due to middle ear
pathology, a sensory hearing loss due to cochlear damage, a NVIII hearing loss
due to a lesion of the auditory nerve or to auditory neuropathy/dys-synchrony, or
auditory processing problems (Romero, Mendez, Tello & Torner, 2008; Ngo, Tan,
Balakrishnan, Lim & Lazaroo, 2006; Sano, Kaga, Kitazumi & Kodama, 2005;
Kolker, 2004; Mechem, 2002; Sheykholeslami & Kaga, 2000; Stanton, 1992). It
is important for the audiologist to identify each type of auditory disorder as soon
as possible since undetected hearing loss can have detrimental effects on the
communication development of a child, especially in terms of language and
speech acquisition (Yoshinaga-Itano, 1998).
1.2
Background
A hearing loss is a sensory disability that, if not identified timeously and
intervened appropriately, can negatively influence speech and language
development and, ultimately, may prevent the child from leading a full and
integrated life (Sininger, Doyle & Moore, 1999; Yoshinaga-Itano, 1998).
Jamieson (1994:596) stated that 'the essence of a hearing loss is its effect on
communication and the resulting impact on cognitive, speech, language and
psychosocial development and functioning'. If a hearing loss, in isolation, can
have such an impact on the developing child, it is expected that the combined
15
effect of a hearing loss in combination with another disabling condition such as
CP could be significantly more detrimental to development.
The CP population is generally considered as 'multi-handicapped', since the
condition is often characterized by the presence of more than one disability
(Workinger, 2005; Mecher, 2002). Considering the fact that CP in itself is an
established risk factor for a communication delay, the presence and/or neglect of
an additional hearing loss could disable the child’s development to an even
greater extent (Rossetti, 1996). Hence, the early detection and intervention of a
hearing loss in the multi-handicapped population is of the utmost importance, not
only to minimize the adverse effects of the sensory deficit (the hearing loss), but
also of the overall handicap (Zafeiriou et al., 2000).
Following identification and diagnosis of a hearing loss, an appropriate
intervention plan can be implemented and the type of amplification (hearing aids
or cochlear implant) and the communication approach (aural communication or
augmentative/alternative communication) can be decided on. This decision relies
on precise audiometric information obtained by the audiologist. It is, however,
often problematic to obtain reliable audiometric results in the CP population as
various physical, perceptual and intellectual impairments may hinder the
execution of certain auditory test procedures.
Within the field of Audiology, any population with special needs is referred to as a
difficult-to-test population (Northern & Downs, 2002). This would include children
with CP as this population presents with impairments in various developmental
areas (Northern & Downs, 2002; Mechem, 2002; Newton, 1977). This population
runs a higher risk for an associated hearing loss (Kolker, 2004). Therefore, the
administration of a sensitive audiometric test-battery that is not influenced by
various developmental impairments is needed.
16
Within the audiometric test-battery various auditory assessment procedures are
employed to obtain specific auditory information (Roeser, 2000; Gans & Gans,
1993). Immittance measurements (tympanometry and acoustic reflexes) and
otoacoustic emissions (OAEs) provide important diagnostic value to the
audiological test-battery as these procedures can identify the place of lesion in
the auditory pathway (Danhauer, 1997; Hall & Mueller, 1997). However, the
hearing sensitivity level needs to be established in order to determine the
presence of a hearing loss, and, if a hearing loss is present, the type and degree
of the hearing loss.
The hearing sensitivity level is determined by obtaining hearing thresholds across
the frequency spectrum of 0.25 kHz to 8 kHz.
Behavioural pure tone (PT)
audiometry is the first choice for hearing assessments as it is the only true test of
hearing sensitivity (Folsom & Diefendorf, 1999). Subsequently, conventional PT
audiometry is modified to suit the chronological and the developmental age of the
child in order to elicit the best responses (Northern & Downs, 2002; Hodgson,
1994). The behavioural audiometric procedure which suits the developmental
and chronological age of the child will therefore be the method of choice
(Hodgson, 1994: 472).
However, the administration of behavioural audiometric procedures are
influenced by the voluntarily participation of the child (Yantis, 1994). Voluntary
participation during behavioural PT audiometry is often compromised by factors
such as poor motivation to participate, limited intelligence level (e.g. mental
retardation), short attention span (e.g. hyper-attention or hypo-attention) as well
sensory disabilities such as cortical blindness (Yantis, 1994).
In addition,
physical constraints such as involuntarily reflexes and poor head control may
further constrain the voluntary participation of the child during behavioural
audiometry (Mechem, 2002).
Since children with CP may typically present with a spectrum of disabilities
including sensory and motor impairments as mentioned in the previous
17
paragraph, their ability to engage appropriately in any activity during subjective
procedures may be compromised to a great extent. This inability may lead to
inaccurate behavioural responses during subjective audiometric procedures.
Hence, an accurate assessment of the child’s hearing sensitivity may still remain
difficult to obtain.
1.3
Rationale
When behavioural audiometry is not possible or the validity and reliability of the
results may be questioned, the audiologist needs to administer objective
procedures to determine the hearing sensitivity. Objective audiometry refers to
audiological procedures that are not dependent on voluntary responses from the
individual being assessed, making it especially relevant for difficult-to-test
populations such as children with CP (Hall & Mueller, 1997).
Auditory evoked responses (AER) are objective audiometric procedures that can
be employed to determine the integrity of the auditory system (Hall, 2007; Arnold,
2000). The auditory brainstem response (ABR) is a short latency AER that is the
preferred choice for auditory assessment of infants and other difficult-to-test
populations (Hall, 2007; JCIH, 2007; Jiang, Andrew & Wilkinson 2006; Folsom &
Diefendorf, 1999). The objectivity of the ABR, its sensitivity for the type (cochlear
versus retro-cochlear pathologies) and degree of hearing loss as well as its long
history and significant research database probably favour it as the current
procedure of choice for difficult-to-test populations (Hall, 2007; JCIH, 2007;
Jiang, Andrew & Wilkinson 2006; Arnold, 2000; Folsom & Diefendorf, 1999;
Galambos, Hicks & Jo Wilson, 1984).
Being a difficult-to-test population, the CP population may, however, challenge
the signal-to-noise ratio (SNR) during ABR recording.
ABR recording is
dependent on an adequate SNR; this implies limited interference of background
noise within the frequency spectrum of the ABR (30 Hz - 3000 Hz) during the
recording (Hall, 2007).
However, the CP population frequently displays
18
involuntarily or uncontrollable muscular movements which generate myogenic
potentials during the ABR recording (Hall, 2007; Workinger, 2005; Mechem,
2002).
These myogenic potentials are regarded as background noise which
negatively affects the SNR in two ways: firstly, the frequency spectrum of
myogenic potentials (50Hz - 500Hz) overlaps with the frequency spectrum of the
ABR (30Hz - 3000Hz). Secondly, the amplitude of myogenic potentials may
exceed the amplitude of the vulnerable ABR which is generally between 0.1 to 1
microvolt (Hall, 2007).
These two factors – interference with the frequency
spectrum and exceeding the amplitude of the ABR – may ultimately lead to a
poor SNR during the recording.
The improvement of a poor SNR has traditionally been addressed by
modification or implementation of specific techniques on the conventional ABR
system (Hall, 2007; Sanchez & Gans, 2006; Kurtz & Steinman, 2005).
Conventional ABR technology incorporates certain techniques such as the
amplification of the signal, inclusion of band-pass filters, signal averaging as well
as artifact rejection (viewed in Figure 1.1) (Hall, 2007; Sanchez & Gans, 2006).
Generation of the ABR and additional muscle activity
Amplification of responses (ABR as well as muscle activity)
Filtering process (to reduce the amplitude of
unwanted electrical noise)
Signal averaging and artifact rejection processes
Visualization of the ABR recording
Figure 1.1:
Principles of conventional ABR technology
In the case of assessing a CP child, an attempt to obtain a ‘purer’ ABR signal
(i.e. improved SNR) will typically involve modifying acquisition parameters of the
19
conventional ABR system. Some of these modifications include increasing the
amplification scale (e.g. x150 000 in stead of x100 000), increasing the low cut
filter (i.e. 150Hz instead of 30Hz), using more sweeps (i.e. 2000 in stead of 1000)
or employing a more conservative artifact rejection value such as 10 microvolt
rather than 20 microvolt. Theoretically, these modifications will improve the SNR,
provided the child is relatively quiet during testing. However, regular and
extensive muscular movements, typical of the CP population, may strain the
recording of the ABR to such an extent, that even maximum modification of
various settings may not improve the SNR.
To compensate for the effects of the muscular movements, external patientrelated methods to reduce these movements can be employed, including natural
sleep, sleep deprivation or melatonin (Schmidt, Knief, Deuster, Matulat &
Zehnhoff-Dinnesen, 2006; Surya, Harkera, Begentb, & Chongc, 2005). Although
there are minimum risks involved when utilizing these techniques, they may not
be effective in the CP population. Muscle artifacts may be present during natural
sleep as children with CP may display involuntarily movements even when
sleeping (Surya et al., 2005). Additionally, sleep deprivation may be impractical
for parents and children; whilst the natural sleep agent, melatonin, is more
effective in infants and young children (Schmidt et al., 2006).
In cases where natural sleep, sleep deprivation or the use of melatonin are
inappropriate or impractical for ABR recordings in the CP population, sedation or
general anaesthesia may be resorted to (Hall, 2007).
Sedation or general
anaesthesia will reduce body movements by manipulating the child’s sleeping
pattern (Surya et al., 2005). A reduction in body movements will result in minimal
myogenic potentials; thus enabling the audiologist to obtain an ABR recording
with an adequate SNR.
Although it seems that sedation or general anaesthesia is a relatively
straightforward solution in objective audiometry, especially for a difficult-to-test
20
population, it is not without problems (Elwood, Hansen & Seely, 2001). Sedation,
or any form of general anaesthesia, increases the risk for apnoea or airway
obstruction especially in multi-handicapped and/or developmentally delayed
children such as the population with CP (Schmidt, Knief, Deuster, Matulat &
Zehnhoff-Dinnesen, 2006; Surya et al., 2005; Elwood et al., 2001). Hence, great
care must be taken when implementing sedation and general anaesthesia in the
clinical facility.
Coté and Wilson (2006) formulated recommendations in order to reduce and
alleviate
the
risks
associated
with
sedation
or
anaesthesia.
The
recommendations include the presence of medical supervision as well as the
application of appropriate airway management equipment (Coté & Wilson, 2006).
These recommendations are imperative, especially for difficult-to-test subjects or
any high risk population in the clinical setting. However, considering the costs
involved, it might not be practical and cost-effective in a public health care
system of a developing country such as South Africa.
1.4
Problem statement
The South African public health care system is burdened by poverty and
infectious diseases, e.g. HIV/AIDS.
Priorities within this system include the
prevention, management and cure of infectious diseases which imply that any
disease that is less life threatening is regarded as secondary (Theunissen &
Swanepoel, 2008). The lack of resources and expensive procedures such as
anaesthesia for an ABR may be considered an inappropriate and excessive
expenditure for a health system burdened by coping with acute life-threatening
diseases. Alternative sedation such as chloral hydrate may be offered, though it
too has an increased risk of airway obstruction, especially for the multihandicapped CP population (Surya et al., 2005).
Considering that anaesthesia may be too expensive in the public sector and
sedation still poses a risk, the applicability of the conventional ABR may be
21
limited in auditory assessments of difficult-to-test populations. It is obvious that
the audiologist is in need of an ABR system that estimates hearing thresholds
reliably – even in difficult-to-test cases where children are awake or illustrate
uncontrollable and involuntary movements.
The Vivosonic Integrity™ ABR system (VS) may possibly realize this ideal as it
aims to address some of the challenges faced by audiologists using the
conventional ABR system (Hall, 2007; Sokolov et al., 2007).
Whereas the
outcome of reliable wave components of the conventional ABR systems is
seriously challenged by the presence of excessive myogenic potentials, the VS
ABR system proposes to be less affected by the incorporation of three novel
features presented in Table 1.1.
These features of the VS system were
purposively designed to improve the SNR for optimal ABR recording.
Table 1.1:
Novel features of the VS ABR system
Feature
Objective
Alternative filtering (Pre-
To improve the SNR by differentiating (filtering) between the ABR
amplification)
signal and myogenic potentials prior the amplification process (Hall,
2007).
Kalman filtering/averaging
To improve the SNR by adding value to the evoked responses
(sweeps).
Each sweep is individually considered during the
averaging process and more value is awarded to signals with less
noise and less value is awarded to a 'noisier' signal (Steinman &
Kurtz, 2005).
Wireless recording
To eliminate electrical noises conducted from the computer or
power line (Hall, 2007).
As seen in Figure 1.2, the arrangement of the filters differs from that of the
conventional system (Figure 1.1).
The significance of this arrangement, as
explained in Table 1.1, is that evoked responses (the ABR signal as well as
22
myogenic potentials) are digitally filtered prior to the amplification process (Hall,
2007).
The amplification process may benefit from this arrangement as
contamination due to myogenic potentials and electrically conducted noises are
eliminated.
The elimination of the unwanted potentials reduces the risk of
saturation in the first stage of the amplifier which enables the audiologist to
optimize the gain of the amplifier (Hall, 2007). As the amplifier is positioned at
the electrode site (rather than at a distance like the arrangement of conventional
ABR systems), the quality of the ABR response may further be optimized.
In addition to the alternative arrangement of the filters and the location of the
amplifier, Kalman filtering may offer another advantage for improving the quality
of the ABR recording, especially in the presence of sporadic myogenic potentials
(Kurtz & Steinman, 2005). Kalman filtering is an alternative weighting averaging
technique (algorithm) which promises to minimize the effects of muscular activity
during the ABR recording (Hall, 2007; Kurtz & Steinman, 2005). Kalman filtering
has been designed to evaluate each sweep (stimulus repetition) individually
during the averaging process and to add a certain value accordingly – more
value is awarded to signals with less noise and less value is awarded to a
'noisier' signal (Kurtz & Steinamn, 2005). By using this information an estimate
of the ABR is produced. In this estimated ABR the likelihood of error in the
amplitude estimate at each latency point is minimized (Hall, 2007). Thus, with
the inclusion of the Kalman filtering technique, the adverse effects of sporadic
myogenic potentials are reduced (Hall, 2007; Steinman & Kurtz, 2005). This
suggests that ABR recordings might be more feasible in the CP population when
using the techniques provided by the VS system as compared to techniques
used in conventional ABR systems.
The research question this study therefore proposes is: What is the clinical
utility of the Vivosonic Integrity ABR system in children with Cerebral
Palsy?
23
Generation of the ABR and additional muscle activity
Filtering
Amplification of the response
Kalman filtering
Visualization of the ABR recording
Figure 1.2:
1.5
Alternative setting arrangement of the VS ABR system
Outline of the chapter contents
The current dissertation provides an in-depth description of the procedures
followed to address the research question as described in this chapter. Table 1.2
provides a concise summary of the content of each of the chapters of the
dissertation.
1.6
Conclusion
Auditory assessments in the CP population might be challenging due to the
occurrence of various disabilities including physical and cognitive disabilities.
The ABR is favoured as the current procedure of choice to assess auditory
functioning in difficult-to-test populations.
However, the applicability of the
conventional ABR system is often limited with populations who present with
involuntarily reflexes or spasms such as the CP population. The inclusion of the
novel features of the VS system may alleviate the effects of involuntarily reflexes
or spasms which produce large myogenic potentials (Hall, 2007; Steinman &
Kurtz, 2005).
If proven to be successful, the novel features of the VS ABR
system could be a valuable way to improve the in which difficult-to-test
populations are assessed. Reducing the need for anaesthesia or sedation may
save expenditures and, more importantly, the risks associated with these
procedures are avoided.
24
Table 1.2:
Summary of dissertation contents by chapter
Chapter 1
Introduction and Orientation: This chapter provides an overview of the CP
condition and the need to identify a hearing loss in the CP population. The
challenges to identify and diagnose a hearing loss in this population by
means of various audiological procedures, including the ABR, are also
briefly discussed. The novel VS ABR system is contrasted to the
conventional ABR system to assess auditory functioning in the CP
population. This delineates the purpose of the study, to determine the
clinical utility of the VS ABR system when assessing auditory functioning in
children with CP.
Chapter 2
A Critical Perspective on Auditory Assessment in the Cerebral Palsy
Population: This chapter discusses the CP condition extensively and
emphasizes the importance of early identification of a hearing loss in this
population. The challenges to reliably identify auditory functioning in the CP
population are discussed extensively. The ABR procedure is discussed as
the most widely used auditory evoked response for determining auditory
functioning and estimating hearing thresholds in difficult-to-test populations.
A critical discussion of the conventional ABR serves as an introduction to
the discussion of the VS ABR system with novel technology that has
become available. Theoretical and clinical advantages of how the VS ABR
system might address the limitations of the conventional ABR system are
provided.
Chapter 3
Methodology:
This
chapter
describes
the
operational
framework
implemented to conduct the empirical research. This chapter includes the
aims of the study, the research design, ethical considerations, research
sample, material, apparatus, procedures as well as the validity and
reliability of the research.
Chapter 4
Results: The results of the empirical research are presented in this chapter.
Chapter 5
Discussion: This chapter provides an interpretation of the results obtained in
Chapter 4. The meaning and the significance of the results obtained is
discussed extensively and compared against previous research studies.
Chapter 6
Conclusions and Recommendations: This chapter infers conclusions on the
findings of the study. A critical evaluation of the current study is provided.
Recommendations for further research are made in light of the findings of
the current study.
25
1.7
Summary
This chapter provided an overview of the CP condition and the importance of
identifying and diagnosing a co-occurring hearing loss in this population. The
ABR was briefly discussed as the preferred procedure for assessing auditory
functioning in this population. Novel ABR technology in the VS ABR system was
briefly compared to conventional ABR technology. Theoretical and clinical
advantages of how the VS ABR system might address the limitations of
conventional ABR technology have been indicated. An outline of the chapter
contents was also provided.
26
Chapter 2
A CRITICAL PERSPECTIVE ON AUDITORY
ASSESSMENT IN
THE CEREBRAL PALSY POPULATION
The aim of this chapter is to provide a theoretical foundation for the
empirical research and to provide a critical evaluation and interpretation of
the relevant literature.
2.1
Introduction
The population with cerebral palsy (CP) not only displays a cluster of motor
disorders, but often present with associated disorders such as hearing loss as
well (Mechem, 2002). Within the field of Audiology, this population is regarded
as a difficult-to-test population (Northen & Downs, 2002). The high incidence of
hearing loss in the CP population stresses the importance of accurate, efficient
and risk-free auditory procedures.
2.2
Background to cerebral palsy
The following section is dedicated to the background of CP and its classification,
as well as its general prevalence and its specific prevalence in the South African
context.
2.2.1 Historical perspective on cerebral palsy
The recorded history of CP dates back to eras before Christ when physical
impairments, some of which might be referred to as CP, were depicted and
described through Egyptian carvings and were recorded in both Greek and
Hebrew scriptures (Scherzer, 2001; Newton, 1977).
Figure 2.1 illustrates an
Egyptian carving of an individual with right hemiplegia dating back to 5 BC.
Throughout history CP has been communicated, described and illustrated in
various forms of art and literature: during the medieval and Renaissance periods
artists such as Raphael and Nicolas Poussin illustrated the crippled and the
27
palsied through paintings (Newton, 1977) and in Elizabethan times William
Shakespeare mentioned this condition in his play Richard III (Scherzer, 2001).
Figure 2.1:
Egyptian carving of a person with right hemiplegia
Adapted from: Scherzer (2001:2)
It was, however, only in the 19th century that the modern day understanding of
CP was established when French orthopaedist, Delpech, expressed his interest
in this condition after which the English surgeon, William John Little, described its
symptoms in 1843 (Scherzer, 2001). Subsequently, CP has often been referred
to as Little’s Disease (Lawson & Badawi, 2003; Newton, 1977).
Few connected this condition to a lack of oxygen during birth and suggested that
these children were affected during the first year of life (Wilson-Jones, Morgan,
Shelton & Thorogood, 2007). In 1897 Sigmund Freud suggested that CP might
be due to insufficient brain development prior to birth and related the child’s
abnormal development to factors influencing the developing foetus (WilsonJones et al., 2007). Both Little and Freud made valuable contributions to our
understanding of the nature of this disorder as it is known today; yet it was only in
the 1930-50’s that Winthrop Phelps, an orthopaedic surgeon, described the
cluster of symptoms (i.e. a combination of motor and sensory disturbances)
28
which laid the foundation for more comprehensive definitions of and perspectives
on CP (Mechem, 2002; Newton, 1977).
2.2.2 Prevalence and definition of cerebral palsy
CP has been acknowledged as a universal phenomenon with a prevalence of
between 2 to 4 per 1000 live births (Andersen et al., 2008; Jeseja, 2008; Donnely
et al., 2007; Fawke, 2007; Beckung & Hakung, 2002; Hutton & Pharoa, 2002;
Winter, Autry, Boyle & Yeargin-Allsopp, 2002). Some authors have reported that
the prevalence of CP has remained constant for approximately 40 years despite
technological advances that decrease mortality in term as well as preterm infants
(Wilson-Jones et al., 2007; Lawson & Badawi, 2003). Others have argued that
the prevalence of CP in especially low birth weight infants has increased over a
period of 10 to 20 years (Winter et al., 2002).
This might be attributed to
improvements in neonatal care and obstetric services which may lead to
increased survival of high risk infants, even where CP might be present (Fawke,
2007).
The awareness of a possible increase in the prevalence of this condition
highlights the need for further research in the development of appropriate
assessment and intervention procedures. However, prior to the implementation
of assessment and habilitation services, the term cerebral palsy needs to be
clarified and appreciated by medical personnel, family other individuals involved.
CP has been defined by various researchers at different stages over past
decades and subsequently many definitions of this condition have been
proposed (Andersen et al., 2008; Donnely et al., 2007; Lawson & Badawi, 2003;
Hutton & Pharoah, 2002; Kennes et al., 2002; Mechem, 2000; Zafeiriou, Andreou
& Karasavidou, 2000; Stanton, 1992; MacDonald, 1987; Newton, 1977; Minear,
1956). Definitions found in literature focus mainly on three characteristics of CP:
the static nature of the condition, the fact that it manifests as a motor impairment
as well as the fact that this condition is due to an insult to the immature brain
29
(Donnely et al., 2007; Wilson-Jones et al., 2007; Kennes et al., 2002; Stanton,
1992; MacDonald, 1987; Newton, 1977; Minear, 1956). Considering all three
characteristics, this condition was defined by the International Working Group on
Definition and Classification of Cerebral Palsy in 2004 as “…a group of
permanent 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” (O’Shea, 2008:36).
The above definition clearly states that any insult to the immature brain may
cause CP. Any congenital (i.e. prior, during or directly after birth) or acquired
(e.g. during childhood) insult that disrupts the anatomical and physiological
maturation of the brain can be referred to as a risk factor for CP (Workinger,
2005).
2.2.3 Risk factors associated with cerebral palsy
The aetiology of CP may include a number of risk factors even though 40% to
50% of cases currently diagnosed with CP have no known causes (WilsonJones, 2007; Mechem, 2002; Stanton, 1992; Newton, 1977).
Generally, the
aetiology of CP is categorized into three groups that reflect the risk factors during
the prenatal period (i.e. prior labour), the perinatal period (i.e. from birth to the
first week of life and the postnatal period (i.e. after the first week of life until the
developing brain has matured) (Wilson-Jones et al., 2007; Lawson & Badawi,
2003).
It seems that the risk factors responsible for the majority of CP cases include
problems during intrauterine development, intrauterine infection such as rubella,
congenital
disorders,
asphyxia
(occurring
in
any
gestational
period),
hyperbilirubinemia and prematurity (Wilson-Jones, 2007; Willoughby & Nelson
2002).
Problems during intrauterine development, intrauterine infection and
asphyxia have been accepted by various researchers as established and
consistent risk factors for CP (Wilson-Jones, 2007; Workinger, 2005; Lawson &
30
Badawi, 2003; Cogher et al, 2002; Mechem, 2002; Willoughby & Nelson, 2002;
McDonald, 1987). However, it appears that inconsistent data exists regarding
the incidence of prematurity causing CP. According to Andersen et al. (2008)
prematurity accounts for 12% of the CP population, whereas Wilson-Jones,
Morgon, Shelton & Thorogood (2007) and Mechem (2002) considered
prematurity (birth prior to 37 weeks gestational age) to be the causative factor in
25 to 40% of the CP population in the United States (US) and Sweden. This high
incidence of prematurity resulting in CP could be accurate considering the higher
susceptibility of premature infants for developing high risk conditions such as
asphyxia and hyperbilirubinemia (Mechem, 2002). Thus, the presence of high
risk conditions may be seen as contributing factors for acquiring CP during
prematurity.
Understanding the contributing risk factors for CP remains complex. A single
factor may be insufficient to cause cerebral damage, but if the same factor is
present to an overwhelming degree, it may cause CP (Lawson & Badawi, 2003).
In addition, multiple causes may be responsible for the irreversible brain injury in
CP (Lawson & Badawi, 2003).
Nevertheless, knowledge of aetiologies is
imperative for diagnostic and rehabilitative purposes because the type of CP
often correlates with a specific aetiology and can be classified according to
specific symptoms. Different risk factors play a role in the various periods and are
summarized in Table 2.1.
2.2.4 Classification of cerebral palsy
Similar to the greatly varying aetiology, the presentation of the various disabilities
is unique to each child (Stanton, 1992). Throughout the decades a number of
attempts have been made to construct/develop a method to classify the various
presentations of disabilities (Workinger, 2005; Stanton, 1992; McDonald, 1987,
Newton, 1977; Minear, 1956).
31
Table 2.1:
Risk factors associated with CP
Prenatal risk factors
Perinatal risk factors
Postnatal risk factors
™ Maternal neurological
™ Asphyxia
™ Asphyxia
™ Premature birth (< 28
™ Seizures within 48
disorders/diseases
Infertility treatment
Thyroid disease
weeks)
hours of birth
™ Low birth weight
™ Cerebral infarction
(< 2500 g)
™ Hyperbilirubinemia
™ Hypoxia
™ Metabolic disorders
™ Blood incompatibility
™ Sepsis
™ Multiple gestation
™ Infection e.g.
™ Respiratory distress
™ Intrauterine infections
e.g. cytomegalovirus
meningitis
™ Abnormal foetal
(CMV), rubella
presentation
syndrome/chronic
lung disease
™ Infection e.g.
™ Thrombophilic
™ Placental abruption
meningitis
disorders
™ Instrument delivery
™ Intraventricular
™ Teratogenic exposure
™ Toxoplasmosis
™ Chorioamnionitis
™ Hyperbilirubinemia
™ Maternal fever
™ Exposure to toxins
™ Malformation of brain
structures
hemorrhage
™ Periventricular
leukomalacia
™ Shaken baby
syndrome
™ Head trauma
™ Intrauterine growth
restriction
™ Abdominal trauma
™ Vascular insults
Adapted from: Wilson-Jones (2007); Lawson & Badawi (2003); Cogher et al. (2002); Mechem (2002); Sheykholeslami &
Kaga (2000); Stanton (1992); Newton (1977)
In the early 19th century, Sigmund Freud proposed a broad spectrum
classification which primarily described the visible neuro-muscular symptoms
which included hemiplegia, general cerebral spasticity, paraplegic spasticity,
centralized chorea, bilateral athetosis and bilateral spastic hemiplegia (Shoup &
Roeser, 2000; Stanton, 1992).
Although this classification manner of
classification merely described the orthopaedic aspect of the condition, it created
a platform for researchers such as Minear (1956) to produce a more
32
comprehensive classification that involved neurological as well as orthopaedic
perspectives on CP.
Recent literature refers to two different approaches that are employed to
categorize CP (Wilson-Jones, et al., 2007). The first approach divides CP into
two categories according to the predominant motor impairment (e.g. spasticity,
athetotic or hypotonic) and the topographical pattern of limb movement (e.g.
monoplegia, diplegia or quadriplegia) which are involved (Wilson-Jones, et al.,
2007). The second approach to categorize CP focuses not only on predominant
motor impairments, but also on the area of brain lesion (Wilson et al., 2007).
Two main physiological categories, i.e. pyramidal (spastic) and extra-pyramidal
(non-spastic) serve as the foundation from which various subtypes such as
athetosis and hemiplegia are identified. This approach is summarized in Table
2.2.
Table 2.2:
Classification of CP
Main types of CP
Spastic
Brain lesion
Pyramidal lesion
Cortico-spinal lesion
(upper motor
neurons)
Non-spastic
Subtype
Monoplegia
Diplegia
Triplegia
Quadriplegia
2) Damage to
neurons in
cerebellum
Lower (pathological)
threshold of
stretch reflex.
Increased muscle tone
which can
lead to contractures
Extra-pyramidal
lesion
1) Basal ganglia
lesion/thalamus
(deep motor neurons)
Characteristics
Incontrollable and/or
involuntarily movements
1)Dyskinetic:
Athetosis and
dystonia
Possible difficulty in initiating
voluntarily movement
Hearing and visual
impairments common
2) Ataxic
Disturbance in sense of
balance and equilibrium
Adapted from: Wilson-Jones (2007); Cogher et al. (2002); Mechem (2002); McDonald (1987); Newton (1977)
33
The spastic type of CP accounts for the majority of the CP population (70% to
80%) as demonstrated in Figure 2.2 (Wilson-Jones et al., 2007; Cogher et al.,
2002; Mechem, 2002). This type of CP is occasionally referred to as upper
motor neuron damage and is caused by damage to the cortex and/or the corticospinal pathways (Wilson-Jones, 2007; Cogher et al., 2002; Mechem, 2002). Any
part of the body could be involved in spastic CP, resulting in paraplegia,
hemiplegia, triplegia or quadriplegia could result (Cogehr et al., 2002). The main
characteristic of this group is the presence of an exaggerated stretch reflex as
well as increased muscle tone which can ultimately lead to contractures (O’Shea,
2008; Workinger, 2005; Mechem, 2002).
Non spastic CP
20%
Spastic CP
80%
Mixed CP
25%
Figure 2.2:
Distribution of the main CP types
Adapted from: Wilson-Jones (2007); Cogher et al. (2002); Mechem (2002)
Contrary to the increased muscle tone of the spastic group, non-spastic CP is
characterized by incontrollable and involuntarily (athetiod and dystonic CP) or
disturbances related to kinesis and/or balance and the in-coordination of the
movements (ataxic CP) (Workinger, 2005; Cogher et al., 2002; Mechem, 2002;
Stanton, 1992). The non-spastic group comprises of the athetoid, dystonic and
ataxic types of CP and accounts for approximately 5% to 15% of CP cases as
presented in Figure 2.2 (Wilson-Jones, 2007). Athetiod, dystonic and ataxic CP
are caused by extrapyramidal lesions due to damage to the basal ganglia, the
34
deep motor neurons of the thalamus or the cerebellum (Cogher et al., 2002;
Mechem, 2002; Stanton, 1992). The functions of these brain structures focus on
the regulation of movement. Consequently, damage to these structures may
cause difficulties in terms of this regulation.
Although spastic and non-spastic CP are the main types, a combination between
these two CP types may occur which is referred to as mixed CP (Mechem,
2002). In fact, it appears that this group accounts for approximately 25% of the
total population as illustrated in Figure 2.2 (Mechem, 2002) and displays
characteristics of both main types (Mechem, 2002). The precise classification of
children with a mixed form of CP remains a challenge as different brain lesions
bring forth diverse disabilities (Andersen et al., 2007).
In such cases, the
predominant motor symptoms may determine the classification, though
accompanied symptoms need to be addressed in the intervention process
(Minear, 1956).
By categorizing the symptoms of this complex entity an appropriate intervention
plan can be compiled accordingly.
Thus, the classification of CP remains
imperative for each clinical setting worldwide, also in the South African context
where the prevalence of CP seems significant.
2.2.5 Cerebral palsy in the South African context
CP is a universal phenomenon that does not discriminate between races or
cultures (Andersen, 2008; Winter, 2002; Arens & Molteno, 1989). Although the
prevalence of CP has not been established in developing countries such as
South Africa, it appears to be significantly higher, up to five times, compared to
the estimated prevalence of 2 to 4 per 1000 live births in developed countries,
e.g. Ireland, US and Sweden (Donnely et al., 2007; Fawke, 2007; Couper, 2005;
Winter, Autry, Boyle & Yeargin-Allsopp, 2002).
35
A number of recent studies provided valuable insight into the occurrence of CP in
different regions of South Africa (Couper, 2002; Christianson et al., 2002).
Research conducted in a rural in the Limpopo province indicated that 8.4% of
children assessed at primary health care institutions presented with CP
(Christianson et al., 2002). Couper (2002) reported a prevalence of 10 per 1000
children presenting with CP in a rural area in the Kwazulu-Natal province.
Although these findings are representative of only two rural areas in South Africa,
it seems that the prevalence of CP is rather substantial.
Explanations for the higher CP prevalence rates in developing countries are not
entirely clear (Couper, 2002), though it has been argued that socio-economic
factors could play a significant role (Arens & Molteno, 1989). In South Africa, the
socio-economic situation is adversely affected by challenges such as poverty and
HIV/AIDS (Theunissen & Swanepoel, 2008). Poverty is the cause of undesirable
living conditions such as overcrowding, malnutrition, poor hygiene and
tuberculosis which are contributory factors that could lead to widespread
diseases such as meningitis (Cooper, 2002; Stanley, Blair & Alberman, 2002;
Arens & Molteno, 1989). If not treated timeously and effectively, these diseases
can be a potential causative factor for CP acquired in childhood years prior to
brain maturation (Arens & Molteno, 1989).
Whilst poverty can indirectly be a potential causative factor for postnatally
acquired CP, the presence of HIV/AIDS affects the child directly and can result in
pre-, peri- or postnatally acquired CP (Mitchell, 2001).
The HI-virus has
extensive medical consequences in the paediatric population, including the
presence of CP as a secondary sequel. UK-based research illustrated that 29%
of children with paediatric HIV/AIDS presented with CP (Cooper, Lyall, Walters,
Tudor-Williams, Habibi, De Munter, Britto & Nadel, 2003). This is not surprising
as in-utero infections such as cytomegalovirus (CMV) or toxoplasmosis,
premature birth or low birth weight that are all closely associated with HIV
(Newell, 1998) are also confirmed risk factors for CP (Lawson & Badawi, 2003;
36
Cogher et al., 2002; Mechem, 2002; Willoughby & Nelson, 2002; Sheykholeslami
& Kaga, 2000; Stanton, 1992; Newton, 1977).
In the South African context, HIV/AIDS and poor socio-economic conditions are
but a few factors contributing to the high prevalence of CP. The reality that
confronts medical professionals is not only to identify the children presenting with
CP (despite their HIV-status), but also to manage the condition inclusive of its
widespread effects on development including speech, language and hearing.
2.3
Cerebral palsy and hearing loss
CP is one of many childhood disabilities that not only compromises general
motor development, health status and general behaviour of a child; but is also
associated with secondary impairments (Donnely, Parks, McDowell & Duffy,
2007; Lawson & Badawi, 2003; Kennes, Rosenbaum, Hanna, Walter, Russell,
Raina & Galuppi, 2002; Mechem, 2002; Cogher, Savage & Smith, 1992; Newton,
1977). Norwegian data illustrated that only 28% of children presenting with CP
have been diagnosed without any associated impairments (Andersen, Irgens,
Haagaas, Skranes, Meberg & Vik, 2008).
Secondary impairments include
epilepsy, intellectual impairments or mental retardation, perceptual impairments
and sensory impairments (vision and hearing disabilities) (Andersen et al., 2008;
O’Shea, 2007; Lawson & Badawi, 2003; Kennes et al. 2002; Mechem, 2002;
Stanton, 1992; Newton, 1977).
The population with CP seems to present a higher incidence or occurrence of
hearing loss than the normal population as is illustrated in Table 2.3 (Sano et al.,
2005; Kolker, 2004; Russman & Ashwal, 2004; Shapiro, 2003; Mechem, 2002;
Sheykholeslami & Kaga, 2000; Northern & Downs, 1991; McDonald, 1987;
Newton, 1977).
37
Table 2.3: Incidence of hearing loss in the CP and normal populations
Estimated % of hearing
Estimated % hearing loss in
Estimated % hearing loss in
loss in CP population
infants in Sub-Saharan
infants in industrialized areas
Africa (no CP cases)
(no CP cases)
11% Newton (1977)
0.006% Swanepoel &
0.002% -0.004% (Olusanya, 2008)
0.3% - 3.7% Kennes et al.
Storbeck (2008)
(2002)
4% -25% Kolker (2004)
0.8%-6% Fawke (2007)
The discrepancies that exist between the incidence percentages shown in Table
2.3 could be attributed to a number of reasons (Newton, 1977).
Firstly, the
criteria for judging the presence of hearing loss were not the same for all studies.
If specific criteria for a hearing loss was 0.5 kHz to 4 kHz, and some children
presented with a high frequency loss that fell above the speech frequency range
(that would imply a hearing loss between 6 kHz to 8 kHz), those children would
have been excluded from the incidence numbers that were measured. Secondly,
throughout the decades, different types of hearing assessment procedures were
used. Initially, behavioural audiometry by means of behavioural observations
and a psychogalvanometer which both depended on conditioned responses
(Byers et al., 1955) were used to detect a hearing loss. Finally, Newton (1977)
commented on the variation that exists regarding the skills of different examiners
(the audiologists) in choosing the most appropriate audiological procedures and
at the same time considering the age of the child, the severity of the condition as
well as associated impairments. A child with CP may wrongly be diagnosed with
a hearing loss or, alternatively, may be misdiagnosed as presenting no hearing
loss due to the difficulty in obtaining accurate behavioural test results (Shoup &
Roeser, 2000).
38
Though the specific cause for the presence of a hearing loss in the CP
population is not entirely clear, many of the risk factors are shared, for example
hyperbilirubinemia due to erythroblastosis fetalis (often associated with Rhincompatibility), rubella, prematurity, low birth weight, asphyxia, meningitis,
toxoplasmosis and/or CMV (Sano et al., 2005; Shapiro, 2003; Parker & Parker,
2002; Nakamura, Takada, Shimabuku, Matsuo, Matsuo & Negishi, 1985;
Newton, 1977; Byers et al., 1956). Toxoplasmosis, CMV, bacterial meningitis
and rubella have generally been accepted as established risks for the presence
of a congenital or an acquired sensorineural hearing loss in the population
without CP (Stein & Boyer, 1994), although a relationship between these risk
factors and a hearing loss in the population with CP has not yet been
established. However, it seems that these risk factors are the most apparent
cause of a hearing loss in the CP population. Hearing loss therefore appears to
be subjected to the same risk factors that predispose CP and is not directly
associated with the CP as such.
The majority of the risk factors mentioned above may be responsible for damage
to cochlear structures which lead to the assumption that a sensory hearing loss is
the main type of hearing loss in the CP population (Sano et al., 2005;
Sheykholeslami et al., 2000).
However, middle ear pathologies, in particular
otitis media, are among the most common childhood diseases (Orlin, Effgen,
Handler, 1997) This holds true for the CP population as well (Newton, 1977).
Thus, the presence of otitis media needs to be constantly monitored for
especially since it may be difficult for the CP child to communicate common
symptoms (e.g. otalgia) associated with the condition.
Additionally, the presence of another type of hearing loss, i.e. auditory
neuropathy spectrum disorder (ANSD), must also be kept in mind. ANSD may
be the result of damage to the inner hair cells of the cochlea or dysynchrony of
the synapse of the cochlea and the auditory nerve (Romero, Mendez, Tello &
39
Torner, 2008; Ngo, Tan, Balakrishnan, Lim & Lazaroo, 2006; Sano et al., 2005;
Shapiro, 2003; Sheykholeslami et al., 2000; Nakamura et al., 1985).
Various factors including prematurity and neonatal hyperbilirubinemia may
contribute to the presence of ANSD (Hall, 2007; Shapiro, 2003). Prematurity and
neonatal hyperbilirubinemia are also risk factors for a sensory hearing loss
(Shapiro, 2003) which stresses the importance of a careful and accurate
differential diagnosis by the audiologist.
Differentiating between the types of hearing loss is extremely important as the
selection of an intervention mode will be based on the type of hearing loss. The
child with a middle ear pathology will be referred for an appropriate medical
examination whilst the child with a sensory hearing loss or ASND will be a
candidate for specific amplification (hearing aids or cochlear implants) and
speech therapy (aural rehabilitation and/or augmentative or alternative
communication) depending on the severity of the hearing loss and other
impairments.
In the general population differential diagnosis usually is a straightforward and
uncomplicated process. However, the complex, multi-facetted condition of CP
presents challenges for differential diagnosis.
Differential diagnosis may be
demanding since the audiologist is not confronted with the presence of a hearing
loss only, but with a spectrum of disabilities including motor, cognitive,
perceptual, speech, and language disabilities (Wilson-Jones et al., 2007;
Workinger, 2005; Cogher et al., 2005; Mechem, 2002; Stanton, 1992).
The audiologist needs to overcome these challenges when assessing auditory
functioning of a child with CP. Early identification of a hearing loss and the
implementation of appropriate intervention could make a significant difference in
the CP child’s life and could result in the child living life to his/her optimum
potential, given the physical and/or cognitive constraints.
40
2.4
Importance of early identification of hearing loss in children with
cerebral palsy
The CP population is a vulnerable group with a high risk for global developmental
delay as well as additional hearing loss (Workinger, 2005; Newton, 1977). If the
additional hearing loss is not timeously intervened, it may compromise the global
development to an even greater extent.
Research has demonstrated the adverse effects of a hearing loss on the global
development (including communication, cognitive, motor and emotional abilities)
in populations without any secondary disabilities such as CP (Olusanya, 2008;
Sininger et al, 1999; Yoshinaga-Itano, 1998).
To avoid or minimize these
negative effects, early identification of a hearing loss followed by early
intervention is recommended (Downs & Yoshinaga-Itano, 1999; Singerer et al,
1999).
The efficacy of early intervention following early identification and
diagnosis of a hearing loss has been convincingly demonstrated (Driscoll et al.
2002; Zafeiriou, 2000; Singer, Doyle & Moore, 1999; Downs & Yoshinaga-Itano,
1999; Yoshinaga-Itano, 1998) advantageous to children with and without
disabilities.
By reducing the contributing speech and language impairments, the disabled
child may participate in a community more actively and independently, resulting
in improved quality of life outcomes. Quality of life is the ultimate goal for a
disabled child as stated by the UN Convention on the Rights of the Child (1989).
Since research has proven that the CP population survive well into adulthood
(Hemming et al., 2005; Beckung & Hagberg, 2002; Bottos, Feliciangeli, Sciuto,
Gericke & Vianello, 2001), it is essential that the child’s communication abilities
(language and hearing) are maximized to ensure optimal functioning in their
environment. In order for this goal to be realised, a communication mode needs
to be introduced and implemented.
The foundation of these interventions,
however, remains the accurate and reliable assessment of hearing abilities.
41
2.5
Auditory assessment and cerebral palsy
In the CP population the entire auditory pathway is at risk for a congenital or
acquired hearing loss (Fawke, 2007; Sano, 2005; Kolker, 2004; Northern &
Downs, 2002). The type, degree and configuration of the hearing loss need to be
determined in order for appropriate habilitation to commence.
Thus, a
comprehensive auditory assessment which comprises of behavioural and
objective assessments, and which reveals reliable results, is of great importance.
2.5.1 Behavioural audiometry and cerebral palsy
Behavioural audiometry is the preferred choice in hearing assessment as it is the
only true test of hearing (Folsom & Diefendorf, 1999).
The philosophy of
behavioural audiometry is based on specific responses, unconditioned or
conditioned, obtained from the individual being assessed (Hodgson, 1994). The
behavioural audiometry procedure will determine the type of responses that will
be elicited.
An unconditioned response procedure such as behavioural observation
audiometry (BOA) does not require voluntarily participation and the results are
based on the elicitation of unconditioned responses such as the startle reflex or
sound localization (Hodgson, 1994).
Thus, it may be assumed that this
procedure can be followed in auditory assessments of populations who are
difficult to condition due to various factors such as cognitive or physical
disabilities.
However,
this
procedure
demonstrates
limited
diagnostic
applicability as it is a test of auditory responsiveness only (Gans & Gans, 1993)
and neither frequency nor ear specific information can be obtained (Northern &
Downs, 2002). In an attempt to compromise for these limitations, conditioned
response procedures are employed.
Conditioned response procedures which include conventional pure tone (PT)
audiometry, visual response audiometry (VRA) and conditioned play audiometry
(CPA) are standard behavioural procedures to determine frequency and ear
42
specific auditory sensitivity (Driscoll et al., 2002; Yantis, 1994). Although the age
as well as the developmental level of the client needs to be considered in order
to administer the most appropriate behavioural procedure (Folsom & Diefendorf,
1999), the term conditioned responses implies that the child's cooperation is
expected in order to obtain reliable PT results, whether it be a head turn or an
eye movement.
Voluntary participation and cooperation to conduct the behavioural procedure
are, however, not always possible.
Voluntary cooperation is especially
challenging with CP clients where a variety of complex behaviours such as motor
disabilities (e.g. poor head control, spasticity or involuntarily movements),
cognitive impairments (e.g. mental retardation), perceptual problems (e.g. short
attention span and/or hyperactivity) and visual impairments (e.g. cortical
blindness) may be present (Mechem, 2002, Cogher et al. 1992; Stanton, 1992).
These associated disabilities may impede voluntary and consistent cooperation
(Vlaskamp & Cuppen-Fonteine, 2007) which may interfere with the administration
of behavioural PT assessments and ultimately poses a potential threat to the
reliability of the audiogram (Vlaskamp & Cuppen-Fonteine, 2007). From Table
2.4 it is clear that associated motor and cognitive impairments were the main
obstacles in the way of obtaining reliable behavioural PT results in children with
CP.
From the perspective of paediatric audiology where reliable audiometric
information is imperative, behavioural audiometry may be unreliable and nonspecific (Picton, 1991). This statement may be especially true for assessments
of difficult-to-test populations such as CP where false negative or positives
responses complicate the establishment of audiometric thresholds.
Routine
behavioural hearing assessments (follow-up assessment, for example every 6
weeks) could be argued for as an alternative approach to obtain the necessary
audiometric thresholds.
43
Table 2.4:
Outcomes of behavioural audiometric techniques in the CP
population
Author
Behavioural audiometric
Outcome
technique
Sano et al. (2005)
Behavioural
PT audiometry
67% of the subjects could not be
and
tested with either behavioural
BOA
audiometric techniques due to
cognitive impairments or motor
dysfunction
Topolska et al.
Behavioural PT audiometry
(2002)
75% of the subjects could not be
tested
due
to
the
cognitive
impairments
Driscoll et al.
Behavioural PT audiometry
(2000)
50%
of
the
moderate
subjects
to
with
severe
developmental retardation could
not be tested
Benham-Dunster &
Dunster (1985)
Behavioural
VRA
and BOA
PT
audiometry,
75%, 20% and 5% of moderately
delayed subjects were tested with
behavioural PT audiometry, VRA
and BOA respectively.
11%, 39% and 50% of profoundly
delayed subjects were tested with
behavioural PT audiometry , VRA
and BOA respectively
Two problems, however, may arise when following this approach. Firstly, some
children might never be able to participate voluntarily due to immaturity of the
central nervous system which is reflected in developmental delays in various
44
areas including motor development (Folsom & Diefendorf, 1999).
Secondly,
obtaining frequency-specific audiometric information (i.e. 0.5 kHz to 8 kHz) may
take several sessions. This suggests that the presence of a hearing loss may
only be identified in the second or third follow-up session, thus delaying the
diagnosis of the impairment. Since the Joint Committee on Infant Hearing (JCIH)
(2007) recommends that infants and children receive intervention as early as
possible, any delay in the identification and diagnosis of a hearing loss can be
regarded as an obstacle in the early intervention process (Yoshinago-Itano,
1998).
Providing appropriate amplification (e.g. hearing aids, cochlear implants) is an
important part of the intervention process and depends largely on accurate
diagnostic audiometric results. Hence, the audiologist working with the difficultto-test population needs a testing instrument to identify and characterize a
hearing loss in this population that either complements behavioural audiometry or
replaces it. Objective audiometric measures are typically relied on in such cases.
2.5.2 Objective audiometry and cerebral palsy
Behavioural conditioning of difficult-to-test populations to sound field auditory
stimuli is not feasible, hence there is a need for objective audiometric procedures
(Folsom & Diefendorf, 1999; Hodgson, 1994).
Immittance measurements
(tympanometry and acoustic reflexes), otoacoustic emissions (OAEs) and
auditory evoked responses (AER) provide the audiologist with important
diagnostic information. These procedures are objective and relatively easy to
perform, enhancing their functionality in auditory assessments of difficult-to-test
populations such as the population with CP (Margolis & Hunter, 2000; Palmu et
al., 1999; Danhauer, 1997; Hall & Mueller, 1997). Each procedure focuses on
the functionality of a specific section of the auditory pathway and therefore is
collectively known as place of lesion tests (Danhauer, 1997; Hall & Mueller,
1997).
45
Tympanometry and acoustic reflexes provide diagnostic information regarding
the status and integrity of the middle ear (Margolis & Hunter, 2000; Palmu et al.,
1999). When assessing auditory functioning in difficult-to-test populations such
as CP these procedures fulfil an important part of differential diagnosis in
particular. Tympanometry is especially useful in identifying the presence of otitis
media with its high incidence in intellectually and multi-handicapped children –
higher than in the normal population (Driscoll et al., 2002; Mechem, 2002). In
addition, acoustic reflex testing objectively predicts frequency-specific pure tone
thresholds (Northern & Gabbard, 1994).
The use of these immittance
measurements is extremely valuable, though it does not provide any information
regarding the integrity of the cochlea.
Otoacoustic emission (OAE) measurements provide physiological information
about the functioning of the cochlea, specifically regarding the integrity of the
outer hair cells of the organ of Corti (Sano et al., 2005; Hood & Berlin, 2002;
Norton & Stover, 1994; Durrant, 1992). Several types of OAEs can be recorded,
though the most commonly used in research and clinical settings are Distortion
Product Otoacoustic Emissions (DPOAEs) and Transient Evoked Otoacoustic
Emissions (TEOAEs) (Danhauer, 1997).
Both DPOAEs and TEOAEs are
sensitive to non-pathological factors such as body movements (Venter, 2000).
Excessive body movements create internal noise that influences the recording of
the OAE negatively (Baer & Hall, 1992). Children with CP display a fair amount
of internal noise, whether it is due to spastic contractures, involuntary or
uncontrollable body movements (Workinger, 2005). Venter (2000) particularly
noted that body movements had a great effect on OAE measurements in children
with CP. Thus, OAE measurements may be difficult to obtain, or might not be
obtainable at all in children with severe involuntary or uncontrollable body
movements.
Despite the fact that OAEs may be difficult to measure due to non-pathological
subject factors, the inclusion of OAE as well as immittance measurements in the
46
diagnostic audiological test battery remains essential. However, the audiologist
can not solely rely on these procedures as hearing levels can not be quantified
and hearing thresholds in dB HL are still required.
Auditory evoked responses (AER) including the auditory brainstem response
(ABR), auditory steady state response (ASSR) and electrocochleargraphy
(EcochG) are objective audiometric procedures that can be employed to estimate
hearing thresholds in non-collaborating populations (Aimoni, Ciorba, Bovo,
Trevisi, Busi & Martini, 2010; Hall, 2007; Luts, Desloovere & Wouters, 2006).
Although the latter may serve as a reliable diagnostic tool in hearing
assessments, the invasiveness of this procedure may compromise its
applicability in difficult to test populations (Aimoni et al., 2010). Additionally, the
ASSR provides the audiologist with frequency-specific information, though this
procedure is more sensitive towards a moderate-profound hearing loss which
implies that a mild-moderate hearing loss may be overlooked (Hall, 2007).
Alternatively, the ABR has long been the preferred choice for auditory
assessment of infant and difficult-to-test populations and the use of this
procedure has been burgeoned during the past several years (Aimoni et al.,
2010; Hall, 2007; JCIH, 2007; Jiang, Andrew & Wilkinson 2006; Folsom &
Diefendorf, 1999; Galambos, Hicks, & Wilson, 1984).
2.6
The auditory brainstem response procedure in auditory assessment
The ABR is a short latency AER that occurs in the first 10-15 milliseconds after
commencement of acoustic stimuli (Hall, 2007).
This response was first
described by Sohmer and Feinmesser in 1967 and since then, has also been
referred to as the BAEP (brainstem auditory evoked potentials) BSEP (brainstem
evoked potential), BAER (brainstem auditory evoked response) or the BSER
(brainstem auditory evoked response) (Hall, 2007; Jiang, Andrew & Wilkinson,
2006; Kolker, 2004; Arnold, 2000; Zafeiriou et al., 2000; Rowe 1981).
47
The ABR characterises the electrical activity of the cochlear part of the eighth
cranial nerve as well as the neural components in the brainstem just below the
inferior colliculus in response to acoustic stimulation as illustrated in Figure 2.3
(Hall, 2007; Arnold, 2000; Rowe, 1981). The electrical activity of the ABR is
visually presented by series of components (I, II, III, IV, V) which, as illustrated in
Figure 2.3, have various anatomical generators (Hall, 2007; Arnold, 2000; Rowe,
1981).
Wave I
Distal end of the eight nerve (near cochlea)
Wave II
Pproximal end of the eight nerve (near the brainstem)
Wave III
not entirely clear, but caudal brain stem near the
Trapezoid body and superior olivary complex could be
responsible
Wave V
Not entirely clear, but lateral lemnicus and the inferior
colliculus could be responsible
Figure 2.3:
Anatomic generators of the different components of the ABR
Adapted from: Hall (2007); Arnold (2000); Rowe (1981)
The ABR is highly dependent on synchronous firing of the neural fibres (Hall,
2007; Arnold, 2000; Weber, 1994).
The most optimal type of stimulus that
enhances neural synchrony is an abrupt click stimulus (e.g. 0.1 milliseconds)
(Hall, 2007; Luts, 2004; Arnold, 2000; Weber, 1994). The ABR recording elicited
by click stimuli is usually referred to as a click-evoked ABR.
The click-evoked ABR has a dual purpose in the clinical setting. It can be used
for neuro-diagnostic purposes as well as for hearing assessments (Hall, 2007;
Arnold, 2000; Musiek, Borenstein, Hall & Schwaber, 1994). The robustness of
the click-evoked ABR responses, that also ensures clear inter-peak and absolute
latencies, contribute to the value of neuro-diagnostic assessments (Hall, 2007;
Arnold, 2000). Any dysfunction or abnormality of the auditory nerve or the lower
brainstem will be observed in delayed inter-peak and absolute latencies, or
48
ultimately, in the absence of the ABR (Hall, 2007; Arnold, 2000). Furthermore,
the click-evoked ABR is a valuable procedure in objective hearing assessments
of difficult-to-test populations such as neonates or children with disabilities (Hall,
2007; Arnold, 2000; Folsom & Diefendorf, 1999). Click-evoked ABR thresholds
can estimate hearing thresholds within 5 dB to 10 dB of behavioural thresholds,
though it is best associated with behavioural thresholds between 2 kHz to 4 kHz
(Hall, 2007; Arnold, 2000; Folsom & Diefendorf, 1999; Gorga, Worthington,
Reiland, Beauchaine, & Goldgar, 1985; Galambos et al., 1984).
The click stimulus encompasses energy over a broad frequency spectrum (Hall,
2007; Oates & Stapells, 1998; Gorga et al., 1993).
This typically reflects
activation of a wide range of the basilar membrane which correlates well with
high frequency hearing in the 2 kHz to 4 kHz area (Hall, 2007; Luts et al., 2004;
Arnold, 2000; Hall & Mueller, 1997; Bergman et al., 1992).
The correlation with hearing in the 2 kHz to 4 kHz area explicitly implies the
limitations of this procedure, i.e. the lack of frequency-specificity and the lack of
low frequency information (Hall, 2007; Marttila & Karikoski, 2005; Luts, 2004;
Purdy & Abbas, 2002; Arnold, 2000; Stapells, Gravel & Martin, 1995). Without
low frequency information and frequency-specific information, a hearing loss can
be overestimated or underestimated (Hall, 2007).
The overestimation or
underestimation of a hearing loss not only affects the validity of the outcome of
diagnostic audiology, but also the habilitation process. These limitations suggest
the importance of the inclusion of a frequency-specific ABR procedure to
complement the click-evoked ABR in diagnostic hearing assessments of difficultto-test populations (Purdy & Abbas, 2002).
In order to provide a more frequency-specific ABR, several types of stimuli such
as filtered clicks, noise stimuli and tone bursts (TB) have been employed (Hall,
III, 2007; Arnold, 2000; Stapells, 2000; Gorga, 1999; Oates & Stapells, 1998).
TB stimuli are the most commonly used to obtain frequency-specific information
49
for recording an ABR (Hood; 1998).
TB stimuli are an attempt to maximize
frequency specificity as well as neural synchrony: These stimuli are brief tones
with a rise and fall time of a few milliseconds and a brief or no plateau duration
(Hall, 2007; Arnold, 2000; Hood, 1998). These types of stimuli have narrower
frequency spectra than clicks, therefore they contain energy at a specific pure
tone frequency (for example 0.5 kHz) (Hall, 2007).
However, the trade-off
between stimulus duration and frequency specificity is well appreciated, since a
TB with a very short onset may consequently produce spectral splatter (Hall,
2007; Purdy & Abbas, 2002). Blackman ramping is the most optimal method and
is included in the stimulus package of most current AER systems (Hall, 2007).
Blackman ramping is an alternative algorithm that refers to the appropriate
shaping or windowing of the rise/fall portion of the TB and attempts to reduce the
spectral splatter and ensure frequency specificity (Hall, 2007; Arnold, 2000).
Frequency-specific TB stimuli can predict reasonably accurate estimates for the
pure tone audiogram in the frequency region 0.5 kHz to 4 kHz (Oates & Stapells,
1998) with threshold differences in the region of 20dB for lower frequencies and
10dB for higher frequencies (Stapells, Gravel & Martin, 1995).
Thus, tone-
evoked ABR can provide the essential information needed for hearing
assessments, diagnosis and further management (e.g. hearing aids) of difficultto-test populations. However, the inclusion of the click-evoked ABR remains
imperative for determining the integrity and the functionality of the auditory
pathway from the acoustic nerve to lower brainstem level (Hall, 2007; Oates &
Stapells, 1998).
The audiologist therefore needs to employ different stimuli to attain the essential
ABR information. Acquisition parameters of the ABR need to be adjusted when
using click and TB stimuli respectively.
Table 2.5 illustrates the different
acquisition parameters for click-evoked and TB ABR recordings.
50
Table 2.5:
Stimuli
Acquisition parameters of click-evoked and TB ABR recordings
Parameter
Suggestion
Rationale
Filter settings
30-3000Hz,
although the
high pass filter
can be
increased to
150Hz
Analysis time
10 -15ms
The major components (wave I, III and V)
can, at least in some cases, be observed
within 5-6ms after the stimuli were
presented. Analysis time of 10-15ms is
recommended to encompass the major
wave components in most patients,
including infants.
Filter settings
30-3000Hz
The tone burst ABR is dominated by low
frequency energy; thus a high pass filter of
30Hz is imperative to encompass the low
frequency information.
Analysis time
15 -20ms
The latency of wave V will increase as the
frequency decreases.
In order to
incorporate wave V in the recording of
lower tone burst ABR recordings (1kHz and
0.5kHz), a longer analysis time is needed.
employed
Click
Tone burst
The click-evoked ABR correlates
high frequency hearing; thus
frequency energy can be cut
150Hz in stead of 30Hz) in
produce a clearer recording.
best with
the low
off (e.g.
order to
Adapted from: Hall (2007)
2.7
The auditory brainstem response as assessment method in the
cerebral palsy population
The objectivity and frequency-specificity (provided TB stimuli are used) of the
ABR make it an ideal procedure for auditory assessments of difficult-to-test
populations such as individuals with CP (Hall, 2007; Arnold, 2000; Sininger 1993;
Galambos et al., 1985). Table 2.6 summarizes a list of studies in which the ABR
was conducted in a population with multiple disabilities, including CP.
51
Research on ABR assessments within the CP population is dominated by studies
that employed click stimuli only, as illustrated in Table 2.6 (Romero et al., 2008;
Sano et al., 2005; Kolker, 2004; Topolska et al., 2002; Zafeiriou et al., 2000;
Sheykholeslami & Kaga, 1999; Benham-Dunster & Dunster, 1985; Stein et al.,
1981). As click stimuli correlate best with hearing in the 2 kHz to 4 kHz region,
the majority of the research provides information regarding hearing in these
frequencies only (Hall, 2007; Gorga et al., 1985). This highlights the lack of low
frequency and frequency-specific ABR information for the CP population. This
information is crucial for the intervention process; especially in the CP population
where information over the entire frequency spectrum is rarely obtainable without
objective procedures.
The objectivity of the ABR favoured this procedure in various research projects
involving the CP population (Sano et al., 2005; Kolker, 2004; Zafeiriou et al.,
2000; Sheykholeslami & Kaga, 1999). Although it is an objective procedure, the
audiologist still needs to take some considerations into account during
assessment of this population as it may influence the outcome of the results and
ultimately, the diagnosis.
One of the considerations that needs to be taken into account when recording an
ABR in the CP population is the fact that children with CP present with an
immature central nervous system CNS) (Donnely et. al., 2007; Workinger, 2005;
Cogher et al., 2002; Hutton & Pharoah, 2002). Immaturity of the CNS causes
prolonged latencies (inter-peak and absolute latencies) of the ABR and these
must be accounted for when utilizing the ABR in this population (Hall, 2007;
Jiang & Wilkinson, 2005).
52
Table 2.6:
Summary of previous studies in which the ABR procedure was
conducted in a population with multiple disabilities
Author
Research aim
Romero, Mendez, Tello &
Torner (2008)
To characterize the ABR differences between children with
perinatal encephalopathy and healthy children
Sano, Kaga, Kitazumi &
Kodama (2005)
To identify the location of causing hearing loss in patients with CP
due to asphyxia and hyperbilirubinemia
Kolker (2004)
To determine hearing function by means of auditory evoked
potentials (ABR and cortical potentials) in the population with
spastic CP
Topolska, HassmannPoznańska, Sołowiej
(2002) **
To assess hearing function in children with infantile CP
Zafeiriou, Andreou &,
Karasavidou (2000)
To determine the utility of the ABR in the population with spastic
CP and determine inter wave latencies of the ABR in this
population
Sheykholeslami & Kaga
(1999)
To localize the pathophysiology of hearing loss in subjects with
hyperbilirubinemia by utilizing OAE and ABR measurements as
well as behavioural audiometric techniques
Benham-Dunster &
Dunster, (1985)
To compare behavioural, acoustic reflexes and ABR procedures in
developmentally delayed population
Stein, Ozdamar &
Schnabel (1981)
To determine the utility of the ABR with multi-handicapped and
suspected deaf-blind children
Current study
To determine the clinical utility of the VS ABR system in
children with CP
** Research report was only available in Polish, thus information was obtained from the English abstract only
Within the CP population, it appears that latency values vary (Kolker, 2004;
Zafeiriou et al., 2000; Sheykholeslami & Kaga, 1999).
Prolonged inter-peak
latencies during ABR recordings of subjects with spastic CP were reported by
53
Kolker (2004) and Zafeiriou, Andreou & Karasavidou (2000). However, normal
inter-peak latencies were also reported (Zafeiriou et al., 2000; Sheykholeslami &
Kaga, 1999). The discrepancy in the latency values may be attributed to the
range of disorders and their severity severities in the CP population and
correlates with the heterogeneity of this population.
The clinical audiologist needs to consider the intactness of the CNS because
important ABR components may appear at later time intervals where the
compromised CNS is compromised.
Thus, by the implementation of, for
example, an increased analysis time e.g. 20ms instead of 15ms, all the ABR
components may be visualized. All ABR components may, longer analysis time
not withstanding, not always be visualized – like in some CP cases where ANSD
is present (Hall, 2007).
Alongside the presence of a cochlear microphonic
response, ANSD will produce either absent or abnormal and poorly defined ABR
recordings at maximum intensity, i.e. 95dB nHL (Rance, Beer, Cone-Wesson,
Shepherd, Dowell, King, Rickards, & Clark, 1999).
Absent or abnormal ABR findings were recorded in some studies within the CP
population (Sano et al., 2005; Kolker, 2004; Sheykholeslami & Kaga, 1999). The
majority of the CP population who presented with abnormal or absent ABR
recordings had a history of hyperbilirubinemia (Sano et al., 2005; Sheykholeslami
& Kaga, 1999). Since hyperbilirubinemia is a high risk factor for ANSD it may be
speculated that the ABR findings were abnormal or absent due to the presence
of ANSD (Hall, 2007; Shapiro, 2003). Unfortunately, ANSD as a topic has not
been documented in any of the previous research reports (Sano, Kaga, Kitazumi
& Kodama, 2005; Kolker, 2004; Sheykholeslami & Kaga, 1999). It remains an
important issue, however, to consider in CP cases where ABR recordings are
strange or abnormal.
Irregular ABR recordings in the CP population may also be attributed to an
inadequate signal to noise ratio (SNR). A poor SNR originates from noise levels
54
exceeding the amplitude of the incoming stimuli (Sanchez & Ganz, 2006; Kurtz &
Steinman, 2005). Any noise produced internally e.g. movements of the body
including the eyes, head or jaw will result in myogenic potentials that increase the
noise levels (Hall, 2007; Sanchez & Ganz, 2006). Within the CP population,
uncontrollable or involuntary movements may be responsible for excessive
myogenic potentials that decrease the SNR and ultimately contaminate the ABR
recording.
Previous research rarely mentions the effect of uncontrollable or involuntary
movements or the presence of ANSD on ABR recordings (Sano et al., 2005;
Kolker, 2004; Zafeiriou et al., 2000; Sheykholeslami & Kaga, 1999).
It is
imperative that both factors be considered since these factors can lead to invalid
ABR recordings and ultimately to erroneous diagnosis.
For example, when
suspecting the presence of ANSD, the polarity of the stimulus (i.e. rarefaction
and condensation) needs to be changed (Hall, 2007). In addition, the effects of
sporadic muscular movements need to be considered as well, since it may
contaminate the SNR which in turn may affect accurate identification of ABR
wave components.
In an attempt to improve the SNR various technical
parameters of the conventional ABR system may be modified (Hall, 2007; Kurtz
& Steinman, 2005).
2.7.1 Improving the signal-to-noise ratio in the conventional auditory
brainstem response system
Modifications of the technical parameters of the conventional ABR system
include alterations to the amplification scale, filtering settings, the amount of
signal averaging and artifact rejection that are used by the ABR system (Hall,
2007; Sanchez & Ganz, 2006).
The theoretical principle of each technique used by the conventional ABR system
is presented in Table 2.7. Basically, the main objective of all the techniques is to
present a well-defined ABR recording. More specifically, in order to record a
55
distinct ABR, the aim of filtering, signal averaging and artifact rejection is to
improve the SNR, while amplification focuses on presenting improved ABR
amplitude.
The amplitude of the evoked response generated by the cochlea, auditory nerve
and brainstem is minute (for wave V of the ABR approximately only 0.5
microvolt); amplification is a critical component (Hall, 2007).
However,
amplification is the first process in conventional ABR technology which implies
that additional responses such as myogenic potentials or electrical interferences
may also be generated and, as a result, also be amplified. When myogenic
potentials or electrical interferences are extensive, the amplitude of these
background noises (e.g. 100 microvolt) may exceed that of the ABR,
contaminating the SNR.
In order to reduce the amplitude of unwanted electrical noise or myogenic
potentials but preserve the actual ABR, band-pass filter settings may selectively
be modified (Hall, 2007). The selected settings of the high pass/low cut and low
pass/high cut filters will determine the electrical activity that will pass through the
filters for the averaging process (Hall, 2007; Arnold, 2000). A high pass/low cut
filter setting of 30 Hz and low pass/high cut filter setting of 1500 Hz (or 3000 Hz)
is generally effective: a setting of 30 Hz will filter out normal EEG as well as
electrical energy below 30 Hz, while settings of 1500 Hz or 3000 Hz will reduce
interferences during ABR measurement due to activity in the higher frequency
range (Hall, 2007).
56
Table 2.7:
Technique
Amplification
Filtering
Signal
averaging
Artifact
rejection
Principles and limitations of conventional ABR technology
Theoretical principle
Limitation
Because the ABR is such a minute response, it has to be amplified
substantially (i.e. up to 100,000 times) before it can be processed
by a signal averaging process and displayed on a computer
screen (Hall, 2007)
Myogenic, magnetic and electric potentials may share a portion of the ABR
frequency spectrum (30Hz-100Hz) (Hall, 2007). Conventional ABR technology
amplifies the broad signal, including these potentials, which may result in a less
identifiable ABR recording.
The implementation of filters attempts to eliminate unwanted
electrical activity (the noise) of the desired electrical activity (the
actual response) (Hall, 2007: 63)
Minimal improvement of the SNR as the amplified broad signal – the ABR and
the myogenic potentials pass through the filters. Furthermore, the frequency
spectrum of the myogenic potentials (50-500Hz) often overlaps with the
frequency spectrum of the ABR (30-3000Hz) (Hall, 2007; Sanchez & Gans,
2006).
The time (signal averaging) needed to record a detectable
response is depends on the size of the signal as well as the
amount of myogenic or electric potentials (noise) within the
recording (Hall, 2007: 63). Thus; signal averaging attempts to
reduce noise by the square root of the number of sweeps in an
averaged response (Sanchez & Gans, 2006).
Less sweeps are required to elicit an adequate SNR if minimal myogenic
potentials are present suggesting a shorter recording time. On the contrary, the
more myogenic potentials detected by the electrodes, the more sweeps are
needed to obtain an ABR recording. This suggests that an ABR recording may
take very long when muscle artifacts are in excess.
An artifact (in AER recording) can be defined as electrical activity
that is not part of the AER and thus should be excluded during the
analysis (Hall, 2007). Artifact rejection is an approach used during
AER recording to minimize the effect of the artifacts on the
recording. This approach evaluates the amplitude of the incoming
noise from the electrodes for each sweep and rejects the sweep
from the averaging process when the noise exceeds a
predetermined microvolt level (Sanchez & Gans, 2006:154).
The inability of this approach to make progress with the averaging process due
to continuous artifact rejection can be responsible for a lengthy ABR recording
time. Furthermore, the obvious contamination of a waveform that is being
averaged with artifact despite the use of this approach is another limitation
(Hall, 2007; Sanchez & Gans, 2006).
57
The filtering process will, however, not obscure all the effects of unwanted
background noise (Hall, 2007; Sokolov, Kurtz, Sokolova, Steinman, Tedesco &
Broome, 2007).
In an attempt to improve the SNR even further, thereby
improving the visibility of the recorded ABR, the signal averaging technique is
applied (Sanchez & Gans, 2006).
The signal averaging process has been described as the heart of the
conventional evoked response system (Hall, 2007). The technique is based on
the assumption that presenting repetitive acoustic stimuli (sweeps) results in a
constant pattern in auditory brain activity within a certain time (Hall, 2007;
Sanchez & Gans, 2006; Arnold, 2000; Sininger, 1993).
However, the time
needed to record a detectable ABR depends largely on the amount of
background interferences such as myogenic potentials during the recoding (Hall,
2007: 63).
In the presence of minimal myogenic potentials the ABR may be recorded within
a short period of time as fewer sweeps are required during the signal averaging
process. On the other hand, the signal averaging process will take longer when
myogenic potentials are in excess, when more sweeps will be presented in an
attempt to improve the SNR.
However, in some instances where the child
displays excessive muscular movements on a sporadic basis, the utilization of
more sweeps, i.e. more signal averaging, may not improve the SNR sufficiently.
In an attempt to record a detectable ABR in the midst of sporadic muscular
movements, artifact rejection is often incorporated in conventional ABR systems
(Sanchez & Gans, 2006). Although there are other techniques of artifact removal
available that focuses specifically on offline averaging e.g. Bayesian weighted
average and artifact rejection equal noise average, the majority of clinical
systems employs the artifact rejection technique (Sanchez & Gans, 2006).
Artifact rejection evaluates the amplitude of the incoming noise from the
electrodes for the individual sweeps and rejects a sweep if the incoming noise
exceeds predetermined microvolt levels, i.e. rejection criteria (Sanchez & Gans,
2006). Although rejection criteria may often range between 10 microvolt and 20
microvolt for example, a more conservative criteria level may be selected during
an ABR recording which is characterized by excessive myogenic potentials
(Sanchez & Gans, 2006).
Some populations, e.g. the population with CP, may display excessive myogenic
potentials due to involuntary muscular movements. The extensiveness of these
muscular movements may have a negative impact on amplification, filtering,
signal averaging as well as artifact rejection techniques of the conventional ABR
system.
2.7.2 Cerebral palsy and conventional auditory brainstem response
technology
The presence of involuntary muscular movements typically displayed in CP may
hinder the feasibility of conventional ABR recording. The feasibility of the ABR
may be compromised as the modified settings of the conventional techniques
may not sufficiently capture the extensiveness of the muscular movements,
which evidently may result in an inadequate SNR.
The SNR may be inadequate in the presence of excessive muscular movements
since myogenic potentials, which are generated by muscular movements, may
share a portion of the ABR frequency spectrum (frequency spectrum of myogenic
potentials: 50 Hz to 500 Hz; frequency spectrum of ABR: 30 Hz to 3000 Hz)
(Hall, 2007).
The fact that there is an overlap in the frequency spectrum
suggests that the SNR might already have been contaminated during the
amplification process since a broad signal, which include the ABR signal as well
as myogenic potentials, is amplified in conventional ABR technology (Hall, 2007;
Kurtz, Sokolova, Steinman, Tedesco & Broome, 2007). Furthermore, the fact
59
that the overlap in the frequency spectrum tends to be specifically in the low
frequencies implies that the low cut filter setting may need to be set to the
maximum in an attempt to filter out the adverse effects of myogenic potentials on
the SNR.
However, when implementing a higher low cut filter setting, the danger arises
that portions of the ABR may be eliminated, which naturally may result in an
inaccurate ABR waveform (Hall, 2007). ABR waveforms elicited by both click
and TB stimuli may be affected, though TB ABR recordings are particularly
vulnerable to severe filtering. TB stimuli depend on low frequency energy, thus
the utilization of a high low pass band filter, e.g. 100 Hz or 150 Hz suggests that
essential information between 30 Hz and 100 Hz/150 Hz may be eliminated.
Clearly, the feasibility of the conventional ABR in the CP population may be
affected by inadequate functioning of the amplification and filtering techniques
due to the influence of myogenic potentials.
The presence of myogenic
potentials may, however, offer unique challenges to the signal averaging and
artifact rejection techniques.
Excessive myogenic potentials may have a strenuous effect on the signal
averaging and artifact rejection techniques which may ultimately affect the
feasibility of the recording (Sanchez & Gans, 2006). The ABR signal may be
impossible to detect when contaminated by undesired potentials, even with the
inclusion of more sweeps, i.e. more signal averaging. In addition, the utilization
of artifact rejection can have detrimental effects on the morphology of the ABR
(Sanchez & Gans, 2006).
Together with the possibility of the ABR being compromised, the period in which
the ABR is recorded may also increase.
excess, more sweeps are required.
When myogenic potentials are in
In addition, the inability of the artifact
rejection technique to make progress with the averaging process due to
60
continuous artifacts being rejected may also lengthen ABR recording time. The
recording time of a hearing test, including the ABR, remains crucial, especially in
difficult-to-test populations such as CP and as much information as possible must
be obtained in the shortest time available (Gorga et al., 2006; Bachmann & Hall,
2001). Thus, the ABR needs to be conducted efficiently and quickly.
Limited research is available regarding the recording time and efficiency of the
ABR procedure with specific reference to the CP population (Sano et al., 2005;
Kolker, 2004; Zafeiriou, Andreou &, Karasavidou, 2000; Sheykholeslami & Kaga,
1999; Benham-Dunster & Dunster, 1985; Stein et al., 1981). Despite the limited
research, it is apparent that the extensiveness of the muscular movements may
have a detrimental effect on conventional ABR techniques and may affect both
the feasibility and the recording time negatively. It is apparent that modifications
to the acquisition parameters of the conventional ABR system alone may not
adequately improve the SNR due to the effects of involuntary muscular
movements. In an attempt to reduce the effects of the muscular movements, e.g.
by enhancing the restfulness of the child, certain patient management techniques
can be implemented (Hall, 2007; Surya, Harkera, Begentb, & Chongc, 2005).
2.7.3 Improving
the
signal-to-noise
ratio
by
implementing
patient
management techniques
The restfulness of the child may enhance low muscular activity which can
contribute to effective recording of the ABR. The audiologist may implement a
few patient management techniques to enhance the required restfulness. These
patient management techniques include natural sleep, sleep deprivation or the
use of melatonin, sedation or general anaesthesia (Hall, 2007; Surya et al.,
2005).
Although these patient management techniques may improve the quiet state of
the child and the ABR recording, there are various disadvantages which can be
summarized as follows:
61
™ Time insufficiency
™ Expenditure
™ Health risks
Natural sleep and sleep deprivation are both cost-effective techniques and,
compared to sedation or general anesthesia imposes the least health risks.
However, these techniques might be impractical as children, especially older
children, may become irritable and take a long time before falling asleep.
Additionally, muscle artifacts are often still present in natural sleep in which case
the signal to noise ratio of the ABR recording will not be improved (Surya et al.,
2005). If natural sleep or sleep deprivation is ineffective, the audiologist may be
forced to make use of sedation or any form of general anesthesia.
However, several researchers have reported that sedation or general
anaesthesia may impose multiple health risks such as sleep apnoea or upper
airway obstruction in severely handicapped children such as those with CP
(Schmidt, Krief, Deuster, Matulat & Zehnoff-Dinnesen, 2007; Wasemer &
Whitehouse, 2002; Elwood, Hansen & Seeley, 2001). According to Elwood et al.
(2001) severely handicapped children are especially at risk for upper airway
obstruction as this population (including children with CP) display a narrower
antero-posterior diameter of the airway at the level of the soft palate as opposed
to children without any developmental delay.
A light sedative in the form of melatonin has been employed in children with
developmental delays and severely handicapped children during procedures
such as CT, PET, MRI or the ABR procedure (Schmidt, Krief, Deuster, Matulat &
Zehnoff-Dinnesen, 2007; Surya et al., 2005).
This sleeping agent has been
proven to enhance sleep in children during these medical procedures (Schmidt et
al., 2007). However, melatonin might be more effective in younger children than
in older ones and side effects such as sleep apnea may still occur in some
children.
Thus, careful observation of the airway during sedation or general
anaesthesia is essential in ensuring safe management (Elwood et al., 2001).
62
Safe management of general anaesthesia is an expensive process in the hospital
setting. It involves the use of highly specialized equipment and trained personnel
(Schmidt et al., 2007). Public health care in South Africa is already challenged
with problems such as poverty and infectious diseases e.g. HIV/AIDS
(Theunissen & Swanepoel, 2008), thus limiting the financial expenditure on
specialized services such as general anaesthesia for an ABR for a non-lifethreatening condition such as a hearing loss. Although the use of sedation may
be viewed as more cost-effective than general anaesthesia, especially in the
public health care system of South Africa, it still requires constant supervision of
the child.
For difficult-to-test populations such as CP the increased risk for
apnoea still remains (Schmidt et al., 2007; Rowe, 1981).
Not only does sedation and general anaesthesia increase the risk for airway
obstruction (especially in difficult-to-test populations), but it may also not be costeffective for the public health care systems of developing countries such as
South Africa. Yet, without sedating or anaesthetizing an uncooperative child, the
ABR may not be reliable due to the negative effects of excessive muscular
movements on the recording. Therefore, the usefulness of the conventional ABR
instrument may be seriously limited, especially in populations where sedation or
general anaesthesia can not be administered.
This also suggests that
conventional ABR technology may limit its applicability for the variety of patients
seen in clinical practice. From the above it becomes clear that audiologists are in
need of ABR technology that is less sensitive for the effects of excessive
muscular movements.
2.7.4 Novel technology for improving ABR signal-to-noise ratio
Recently an ABR system, the Vivosonic Integrity™ (VS), has become available
(Hall, 2007). The VS was first introduced in 2006 and has since been used in
various clinical facilities across the US (Sokolov et al., 2007). What makes this
system particularly appealing to the audiologist is the potential that an ABR may
63
be reliably recorded within the presence of muscular activity (Hall, 2007; Kurtz &
Steinman, 2005). This potential suggests the prospect of eliminating or reducing
the number of cases requiring sedation or general anaesthesia for ABR
assessments.
This supposed advantage also holds significant promise for
assessing auditory functioning in children with CP.
The supposed benefit of the VS system lays with the introduction of three novel
features namely the pre-amplification of the evoked response, Kalman-weighted
averaging and wireless/blue-tooth recording (Hall, 2007).
Each feature
endeavours to address the challenges often found in conventional ABR
recordings. These features as well as the underlying principles and the expected
advantage in ABR recording is presented in Table 2.8. The inclusion of wireless
recording poses a definite advantage over conventional ABR systems. While the
set-up of the conventional ABR system seriously jeopardizes the pureness of the
ABR signal because of electrically conducted noises that may stem from the
power line and the computer, wireless recording eliminates such contamination
of the signal (Hall, 2007).
Wireless recording removes the electric path (i.e.
wires) between the computer and the amplifier, eliminating electrically conducted
interferences (Hall, 2007).
The elimination of electrical interferences will, however, not ensure a well-defined
ABR recording in the presence of myogenic potentials. In order to manage the
effects of the myogenic potentials, remaining features (i.e. alternative filtering and
Kalman averaging) is incorporated. The alternative filtering, also referred to as
pre-amplification, accomplished through the Amplitrode™. The Amplitrode™ is a
miniature, on-site AER amplifier fitted directly on the ABR electrode (Hall, 2007).
The significance of the location of the amplifier lies in the fact that the signal is
filtered prior to amplification (Hall, 2007). This alternative filtering arrangement
suggests that the amplified signal may be less contaminated by myogenic
potentials.
This feature differs from conventional ABR technology where
amplification occurs prior the filtering process.
64
Table 2.8:
Principles and supposed advantages of the VS ABR system
Technique
Underlying principle
Supposed advantage
Alternative
Filtering/ Preamplification
The input signal is filtered prior
amplification by means of the
Amplitrode, the
on-site
AER
amplifier (Hall, 2007)
The amplified response is less
contaminated by unwanted low
frequency myogenic potentials
such as general muscular activity,
EOG, ECG and EEG (Hall, 2007).
Kalman
Averaging
Each
sweep
is
individually
considered during the averaging
process: more weight (value) is
awarded to signals with less noise,
and less weight (value) is awarded
to contaminated/noisier signals
(Steinman & Kurtz, 2005)
This technique promises to reduce
the effects of sporadic noise during
ABR recording (Sokolov et al.,
2007; Steinman & Kurtz, 2005)
Wireless/Blue
tooth recording
Communication
between
the
interface module (VivoLink) and the
computer is performed wirelessly.
Wireless communication is possible
as long as the computer and the
VivoLink are within a 10m distance
of each other.
This feature promises to eliminate
the introduction of electrically
conducted
noises
from
the
computer as well as the power line
(Hall, 2007:90).
Also different to ABR technology is the incorporation of Kalman averaging.
Kalman averaging may allow an improved SNR during significant muscular
activity (Hall, 2007). It constitutes a weighted averaging algorithm where each
sweep is individually considered in order to obtain an ABR with minimum
probability of error, regardless of the patient’s muscular activity (Hall, 2007; Kurtz
& Steinman, 2005). This averaging technique aims at estimating the error in
each sweep based on the measurement variance and continuously updates this
estimate to produces an estimate of the ABR signal (Hall, 2007).
As the
likelihood of error in the amplitude estimate at each latency point is minimized in
the predicted ABR, the effects of sporadic muscular movements on the ABR may
be reduced (Hall, 2007; Kurtz & Steinman, 2005). This suggests that the ABR
may be recorded reliably and accurately, even in the presence of significant
muscular movements (Hall, 2007; Kurts & Steinman, 2005).
65
The information above suggests that the new technology introduced by the VS
ABR system may be less affected by the influence of excessive muscular
movements. This may enhance ABR assessments in the CP population since
involuntary muscular movements are common in this population (Workinger,
2005; Mechem, 2002; Stanton, 1992).
2.8
Conclusion
CP is a diverse condition which may involve varies disabilities including hearing
loss (Stanton, 1992). A hearing loss in a CP child needs to be identified as early
as possible for efficacious intervention to be realized. Identifying an additional
hearing loss in members of the CP population could be challenging when relying
on conventional audiometry procedures such as behavioural audiometry and
OAE measurements (Sano et al., 2005; Topolska et al, 2002; Driscoll et al.,
2000, Venter, 2000). For this reason the audiologist relies on the ABR to provide
an estimate of hearing sensitivity (Hall, 2007). Traditionally, ABR assessment in
difficult-to-test populations such as the CP population was challenged as the
acquisition parameters could not fully compensate for the presence of involuntary
and irregular body movements. Although these sporadic movements can be
manipulated by sedation or general anaesthesia, both of these processes are not
cost-effective and impose the risk of airway obstruction, especially in severely
handicapped children (Schmidt et al., 2007; Johnson et al., 2002). The inclusion
of the VS ABR system’s novel features may reduce the need for sedation or
general anaesthesia which could be especially beneficial for ABR assessments
of severely handicapped children (Kurtz & Steinman, 2005). The incorporation of
pre-amplification, Bluetooth recording and Kalman averaging holds the promise
of a well-defined ABR recording, even in the presence of excessive body
movements.
2.9
Summary
The possibility of obtaining distinctive ABR recordings with the VS system, even
in the presence excessive body movements, suggests that the utilization of
66
sedation or general anaesthesia may become redundant for VS ABR
assessments. This not only favours the applicability of ABR assessments in
public health care hospitals of South Africa, but also the assessment outcomes in
the difficult-to-test CP population, since a hearing loss may be detected much
earlier than previously. If early identification of the hearing loss is followed by
appropriate intervention, e.g. hearing aids and auditory training, the child with CP
may be integrated in the community despite his/her physical and cognitive
constraints.
67
Chapter 3
METHODOLOGY
The aim of this chapter is to describe the methodological approach that
was implemented during the empirical research component of this study.
3.1
Introduction
Research can be described as the process of investigating scientific questions
(Hedge, 2003:24). The motivation for this research and the research question
underlying this project was described in the preceding chapter.
In order to
practically investigate the research question, the process was dependent on a
methodological foundation (Maxwell & Satake, 1997).
The research methodology outlined in this chapter describes the process that
was followed in order to determine the clinical usefulness of the Vivosonic
Integrity Auditory Brainstem Response (VS ABR) system in the auditory
assessment of children with cerebral palsy (CP).
3.2
Aims of the research
The following aims have been specified for this study.
3.2.1 Main aim
The main aim of this project was to evaluate the clinical utility of the Vivosonic
Integrity ABR system in children with cerebral palsy.
3.2.2 Sub-aims
The following sub-aims were formulated to realize the main aim:
68
Sub-aim 1
To describe the feasibility and characteristics of an audiometric test battery for
assessing auditory functioning in children with cerebral palsy.
Sub-aim 2
To compare the VS ABR system with a conventional ABR, in terms of:
™ Feasibility using click and 0.5 kHz tone burst (TB) stimuli;
™ Electrophysiological thresholds using click and 0.5 kHz TB stimuli;
™ Threshold correspondence with behavioural PT thresholds;
™ Recording time using click and 0.5 kHz TB stimuli.
3.3
Research design
A cross-sectional, within-subject comparison design implementing a quantitative
research approach was selected for this study (Leedy & Ormord, 2005; Maxwell
& Satake, 1997; Johnston & Pennypacker, 1993). The study was cross-sectional
in that all the data for each subject was collected at a specific time (Maxwell &
Satake, 1997).
The experimental part of this study was represented by the
within-subject control condition where two ABR systems, a conventional ABR
system and the VS ABR system, were simultaneously conducted in each subject.
The conventional ABR system served as the controlled condition whilst the VS
ABR system served as the experimental condition.
This unique setup was
important in the research as equivalent test conditions in terms of EEG and
environmental conditions had to be ensured for both ABR systems.
All the
subjects were exposed to both the control and the experimental conditions; thus
within-subject comparisons were the outcome of this research.
The conclusions and implications of any experimental design are dependent on
the dependent (measured) and independent (manipulated) variables as well as
on the experimental setting (Johnston & Pennypacker, 1993).
This study
investigated the usefulness of the VS ABR system when assessing auditory
functioning in children with CP. The manipulated variables for this study were
69
the procedure employed to determine hearing thresholds (behavioural PT
audiometry) and the procedures employed to estimate hearing thresholds (a
conventional ABR system and the VS ABR system). The dependent variables
were the thresholds obtained with different stimuli (click and 0.5 kHz) with the VS
and conventional ABR systems. The PT thresholds at 0.5 kHz, 1 kHz, 2 kHz and
4 kHz are also considered dependent variables as these specific frequencies
were manipulated by the researcher.
Behavioural PT thresholds obtained at
these frequencies served as the gold standard (reference hearing threshold)
against which the thresholds of the VS and conventional ABR systems were
compared.
3.4
Ethical considerations
Different auditory evoked potential equipment was used by researcher; thus it is
important to mention that the researcher had no relationship to either Vivosonic
or Bio Logic.
The study was approved by the Ethics Committee of the University of Pretoria
(Appendix A) as well as the Gauteng Department of Education (Appendix B). A
meeting was also scheduled with the principal of the Pretoria School for Children
with Cerebral Palsy as well as the head of the Speech therapy and Audiology
department of the school. Prior to the meeting mentioned above, the principal of
the school was provided with a letter requesting informed consent in which the
details of the research were also explained (Appendix C).
The fundamental principle of ethical research is to preserve and to protect the
rights and welfare of all the subjects involved in a research project (Jenkins,
Price & Straker, 2003). Hence, the following ethical considerations were taken
into account (Maxwell & Satake, 1997):
70
3.4.1 Respect of privacy of research subjects
To respect the privacy of the subject is a fundamental ethical principle (Leedy &
Ormrod, 2005). Confidentiality of the subjects was ensured by not using the
individuals’ names on any data documentation during the research project
(Strydom, 1998).
A specific code was allocated for each subject for data
processing purposes.
This was clearly explained in the letter requesting
informed consent which was mailed to the parents of the subjects (Appendix D).
3.4.2 Informed consent
According to Leedy & Ormrod (2005) and Strydom (1998) obtaining informed
consent entails the following components:
™ Providing the subjects with adequate information regarding the research;
™ Emphasizing voluntary participation;
™ Informing subjects that they could withdraw at any time during the research
Subsequently, the researcher obtained letters granting informed consent from
each subject’s parents (included as Appendix D). This letter was signed by the
parents after they have read the aims, procedures and the possible benefits of
the study. This letter ensured confidentiality and voluntary participation as well
as each subject's right to withdraw at any time during the research (Kidder &
Judd, 1986).
Additionally, verbal consent was also obtained from each subject prior the
auditory assessments. As it was the responsibility of the researcher to convey
the information in such a way that it was understandable to the each subject, the
entire procedure was explained to the subject in the presence of either the
teacher or the speech therapist of the school (Iacono & Murray, 2003). Voluntary
participation was ensured and it was emphasized that the subject had the right to
withdraw at any time during the research (Appendix E).
71
3.4.3 Beneficence and non-malfeasance
When conducting the various tests, acoustic stimuli were presented at a
comfortable listening level and lower intensity levels, therefore not causing any
discomfort to the subject.
The letter requesting informed consent stated the
duration of the sessions and also that the subject were actively involved during
only one test (pure tone audiometry). Any abnormality that was noted during any
auditory assessment was communicated to the speech therapist and the teacher
for further management, thereby rendering a service to the parents and subjects.
The research project posed no medical risks to the subjects (no sedation was
used). The information that was gathered provided useful data for future hearing
assessments in individuals with CP.
3.5
Research sample
15 Subjects were selected from the Pretoria School for Children with Cerebral
Palsy.
Although the sample size was small, comprehensive audiological
assessments were conducted in each subject. Table 3.1 presents the criteria
that subjects had to comply with to participate in the research project.
The
heterogeneity of the research sample might be seen as confounding the quality
of the interpretation of the especially the ABR data; however data were collected
during school hours which limited the availability of possible subjects.
3.5.1 Selection criteria and procedures
A non-probability purposive sampling procedure was used to select subjects for
the research group (Hedge, 2003; Maxwell & Satake, 1997). According to Hedge
(2003:96) purposive sampling can be viewed as 'a method of handpicking
individuals because they have special characteristics that are necessary for the
study'. Although purposive sampling typically limits generality, this method is
specific and useful in clinical research (Hedge, 2003).
72
Table 3.1 presents the selection criteria for the research sample of the study. The
following procedures were followed for selecting the research sample:
™ The aim of this research was discussed during a scheduled meeting with the
head of the targeted school and the head of the Speech Therapy and
Audiology department of the school.
™ The researcher discussed the criteria for subject with designated teachers
and letters requesting informed consent were handed to them.
™ Based upon the age criteria (between 12 and 18 years), teachers distributed
the letters.
Table 3.1:
Selection criteria for the research sample
Criteria
Rationale
Cerebral palsy
Children diagnosed with CP as defined by the medical records of the
Pretoria School for Children with Cerebral Palsy were selected for this
study
Age
Subjects between the ages of 12 and 18 years were selected for this
research. Less modification of standard audiometric procedures is
needed as children get older (Northern & Downs, 1991). For this
reason conventional pure tone audiometry is more likely to be used in
older children than the modified play audiometry.
Normal middle ear
functioning
Middle ear pathology influences the accuracy of the pure tone
thresholds as well as the latency and morphology of the ABR recording
(Hall & Mueller, 1997); Therefore normal middle ear functioning is a
requirement. According to Worthington & Peters (1984) peripheral
effects, such as a conductive component, should be ruled out before
an abnormal ABR can be interpreted.
™ Because of certain time constraints, a cut off date of 2 weeks for returning the
informed consent letters was stipulated.
™ The children from whom letters granting informed consent were received
were scheduled for the hearing tests.
73
™ Specific dates were scheduled with the Speech Therapy and Audiology
department at the school to conduct the testing at that venue.
™ Otoscopy and tympanometry was conducted and if not complying with the set
criteria for participation, the child was eliminated from the study.
™ If the hearing test results were compliant with the selection criteria, the
researcher proceeded with the behavioural PT audiometry and DPOAE
measurements.
3.5.2 Description of the research sample
The relevant biographic details of each subject are illustrated in Table 3.2. The
subjects’ ages were documented as the age at the time of the study. Although
the causes of CP were unknown in the majority of the subjects, it was apparent
through the school medical records that most acquired the condition during or
directly after birth. Only one subject (subject 7) was diagnosed with CP after a
motor vehicle accident at the age of three years. All the subjects were diagnosed
according to a physiological and topographical classification protocol by a
professional team (including a neurologist, physiotherapist, occupational
therapist and speech-language therapist) at the time of their intake by the school.
3.6
Material and apparatus
The material and apparatus used in this research can be divided into apparatus
used for the selection of the research sample and that used during the data
collection. Both of these categories are discussed in the following section.
3.6.1 Material and apparatus for subject selection
The following material and apparatus were used for the selection of the research
sample:
3.6.1.1 Otoscopic examination
Otoscopic examination of the external meatus and the tympanic membrane was
performed with a Heine Mini 2000 Otoscope. A visible light reflex is most often
indicative of a healthy tympanic membrane (Hall & Chandler, 1994).
74
Table 3.2: Description of the research sample
Subject
Age
Gender
Type of CP
Language of
education
Related auditory problems
(previously determined)
1
17
Male
Spastic diplegia
Afrikaans
None
Language learning problems
2
15
Female
Spastic diplegia
Afrikaans
None
Language learning problems
15
Female
Microcephaly
Afrikaans
None
Language learning problems
4
15
Male
Spastic quadriplegia
Afrikaans
None
Non-ambulatory, limited speech
repertoire
5
16
Female
Athetosis
Afrikaans
Bilateral high frequency
hearing loss
Vision problems; limited speech
repertoire
6
14
Female
Right hemiplegia
Afrikaans
None
Language learning problems
7
16
Female
Right hemiplegia
Afrikaans
None
Mental Retardation
8
15
Female
Athetosis
Afrikaans
None
Dysarthia; Language learning
problems
9
16
Male
Spastic Diplegia
Afrikaans
None
Language learning problems
10
16
Male
Right hemiplegia
Afrikaans
None
Language learning problems
11
17
Male
Spastic triplegia
English
None
Language learning problems
12
13
Male
Spastic triplegia
English
None
Limited speech repertoire;
Language learning problems
13
17
Male
Right hemiplegia
English
None
Language learning problems
3
Other disabilities
75
14
17
Male
15
15
Male
Athetosis
Athetosis
Afrikaans
Afrikaans
None
Language learning problems
Hearing loss was suspected; Language learning problems;
though not confirmed
limited speech repertoire
76
3.6.1.2 Tympanometry
Tympanometric evaluation of the middle ear was performed with a GSI 38 Auto
Tymp Middle Ear Analyzer. A type A tympanogram with normative values as
reflected in Table 3.3 was indicative of normal middle ear functioning.
Table 3.3: Normative tympanometric values
Components of tympanometric
Normative values
measurements
Ear canal volume
0.5 -1.5 ml
Compliance
0.3 -1.6 cc
Middle ear pressure
-100 - +100 daPa (adults)
-150 - + 150 daPa (children)
Adapted from: Stach (1998); Hall & Mueller (1997)
3.6.1.3 Case history
Information regarding the subject’s date of birth as well as the diagnosis of the
CP condition was obtained from the medical files of the school.
3.6.2 Material and apparatus used for data collection
The following material and apparatus were used during data collection.
The
specific protocol that was employed for each auditory procedure is also included
in this section.
3.6.2.1 Ipsilateral acoustic reflexes
Ipsilateral acoustic reflex thresholds were obtained using a GSI 38 Auto Tymp
Middle Ear Analyzer. Ipsilateral acoustic reflexes were measured at 0.5 kHz, 1
kHz and 2 kHz with pulsed pure tones. These frequencies were selected in order
to describe the auditory functioning more comprehensively. Acoustic reflexes
that elicited between 70 dB and 90 dB of the behavioural PT threshold were
regarded as being within the normal range. Acoustic reflexes above 90 dB of the
behavioural PT threshold were regarded as being elevated.
77
3.6.2.2 Distortion product otoacoustic emissions
Distortion product otoacoustic emissions (DPOAE) were measured using AuDX
Bio-Logic OAE equipment.
DPOAEs were measured using the Vanderbilt
DPOAE diagnostic protocol in the frequency spectrum 634 Hz to 6347 Hz.
Details of the protocol were as follows:
™
L1 = 65 dB SPL; L2 = 55 dB SPL
™
F1/F2 ratio: 1.2
™
Number of octaves: 4
™
Number of sweeps per set: 25
To be regarded as a DPOAE, the following analyzing strategies were employed:
™ The amplitude of the DPOAE was within the boundaries of the Vanderbilt
65/55 95th percentile reference set as illustrated in Figure 3.1.
™ The noise floor (NF) level did not exceed 3dB SPL (Hall & Mueller, 1997).
™ The difference between the DP emission and the NF level was equal to, or
DP amplitude (dB SPL)
larger than 10 dB (that is DP-NF >10 dB).
20
15
10
5
0
-5
-10
-15
-20
-25
-30
635Hz
808Hz 1001Hz 1586Hz 2002Hz 3174Hz 4003Hz 6347Hz
Frequency (Hz)
Vanderbilt DP 95th percentile
Vanderbilt DP 5th percentile
Noise floor level 95th percentile
Figure 3.1:
Vanderbilt 65/55 95th percentile normative values
In addition, the value of the DP/NF difference was categorized according to the
criteria presented in Table 3.4.
78
Table 3.4: Criteria for DP/NF difference
DP-NF difference
Outcome
< 6dB
Absent DPOAE
6-9dB
Present DPOAE, but abnormal
>10dB
Present DPOAEs
Adapted from Hall & Mueller (1997)
3.6.2.3 Behavioural pure tone audiometry
Behavioural PT air conduction thresholds were obtained using a GSI 68
Diagnostic Audiometer. The acoustic stimuli were presented through TDH 39
supra-aural headphones. The behavioural PT thresholds were determined in the
frequency range 0.5 kHz to 4 kHz.
3.6.2.4 Auditory brainstem response
The state of awareness for each subject was evaluated prior the ABR recordings
and rated according to a rating scale (Appendix F).
The click-evoked and 0.5 kHz tone burst (TB) ABR thresholds were obtained
using the VS ABR system as well as an ABR system with conventional
technology, i.e. the Bio-Logic Navigator Pro system (BL).
Behavioural electrophysiological thresholds using the VS and the BL ABR
systems were obtained from a group of five normal hearing young adults, aged
between 20 and 28 years, prior to ABR testing in the research sample. These
behavioural electrophysiological click-evoked and 0.5 kHz TB thresholds served
as the reference values for the research sample. The method for selecting these
subjects, the procedures and material implemented as well as the reference
values for both ABR systems are summarized in Table 3. 5.
79
Table 3.5:
Summary of methodology followed to obtain reference values for the VS and BL ABR systems
™ Normal middle ear functioning
Subjects in the normative group were required to have normal middle ear functioning since any middle ear pathology could
influence the accuracy of pure tone thresholds (Hall & Mueller, 1997).
Selection criteria
™ Hearing sensitivity
The normative group was required to present with normal hearing sensitivity (<15dB HL) in the frequency spectrum 0.5 kHz-4
kHz. These frequencies were selected to provide corresponding points in comparing the data to the click-evoked as well as the
0.5 kHz tone burst ABR procedures
™ Constraints in terms of time
™ A convenient sampling process was followed because the sample consisted of acquaintances of the researcher and the
student body of the Department of Communication Pathology at the University of Pretoria (Maxwell & Satake, 2006:96)
™ Selection procedures included:
- The availability and willingness to take part in the study was enquired
- If subject was willing to participate, the informed consent letter was given (Appendix G) and a suitable date and time was
Selection procedures
scheduled for the testing to be done
- At the day of testing otoscopy, tympanometry and behavioural pure tone audiometry was conducted. If subject presented
with normal hearing sensitivity and normal tympanometric results, the researcher proceeded to the behavioural
electrophysiological thresholds obtained with the VS and the BL ABR systems using click stimuli as well as 0.5 kHz tone burst
stimuli.
™ Otoscopy: mini Heine 2000
Apparatus and protocols
™ Tympanometry: GSI 33 middle ear analyzer
(Similar to the research
sample)
™ Behavioural audiometry: GSI 38 Diagnostic
™ VS and BL ABR systems: Similar to that of the research sample
80
™ Protocols: similar to the research sample
™ Normative behavioural thresholds were obtained after behavioural pure tone audiometry was conducted; thus subjects
remained in the soundproof booth. The researcher explained to each subject what was expected of him/her when introduced to
the ABR systems.
The researcher alternated between the ABR systems being employed first. After the ER-3A insert
earphones of the ABR system employed firstly was inserted in the ear canals of the subject, click stimuli was presented
Data collection procedures
monotically at a supra-threshold intensity of 70dB nHL. Stimulation was then gradually descended in steps of 10dB until the
subject indicated that the stimuli were no longer audible. At this stage the intensity was increased in steps of 5dB until the
subject indicated that the stimuli were audible. This level was taken as the behavioural threshold. Click stimulation was
followed by 0.5 kHz tone burst stimuli presentation bilaterally. After behavioural thresholds were obtained bilaterally for click
and 0.5 kHz tone burst stimuli for the one system, the next system was employed following the same procedures.
™ The raw quantitative data was prepared and organized into a data set suitable for analysis (Neuman, 1997). The prepared
Data analysis procedures
data organized on Microsoft Excel XP worksheets were analyzed with statistical measures. The mean behavioural thresholds
using the click and 0.5 kHz stimuli were obtained for both systems.
Click stimuli
VS ABR system: Left ear = 3dB nHL; Right ear = 2 dB nHL
BL ABR system: Left ear = 11 dB nHL; Right ear = 11 dB nHL
Results (mean ABR reference
values in dB nHL)
0.5 kHz TB stimuli
VS ABR system Left ear = 7 dB nHL; Right ear = 8 dB nHL
BL ABR systemLeft ear = 5 dB nHL; Right ear = 5 dB nHL
81
The VS ABR system consisted of the Integrity model V500 (version 4.50, R3
Research code) which was installed on a Dell laptop (Latitude, D 520; Windows
XP operating system). The interface module (Integrity V500) has been designed
for the acquisition and analysis of AER, including the click and the tone-evoked
ABR.
The communication between the data collecting module, the Vivolink
(calibrated February 2008; serial number: VP 0187) happened wirelessly (via
bluetooth) through the D-Link (model number DBT-122) which was inserted in
the USB-port of the laptop. The new generation of pre-amplifiers, the Amplitrode
(serial number ENG 9963) was connected to the Vivolink and consisted of an
integrated pre-amplifier and electrode clip in a combined unit. This was affixed to
Ambu Neuroline 720 disposable pre-jelled, foam-backed electrodes on the
subject’s forehead (Fz), and both mastoids. Insert earphones (ER-3A; Integrity
insert earphones serial numbers: Right: 44747, Left: 44745) were connected to
the Vivolink.
The BL ABR equipment consisted of a specialized hardware component
(Navigator Pro Bio-Logic) that was connected to a laptop.
The system was
operated by a software package (BL Auditory Evoked Potentials version 2.3.0)
specifically designed for the acquisition and analysis of auditory evoked
responses (AER), including the click and tone-evoked ABR. Insert earphones
(ER-3A; Bio Logic insert earphones serial numbers: Right: 35086, Left: 35095)
were connected to the hardware component.
The acquisition parameters of the VS and BL ABR systems were identical,
though technical differences regarding the filtering and the signal processing
occurred. The specifications for the VS ABR system and the BL ABR system are
presented in Table 3.6.
82
Table 3.6:
Acquisition parameters for the VS and BL ABR systems
Specifications
VS ABR system
BL ABR system
Amplification
x100 000
x100 000
Filters :
Settings
100-3000Hz
100-3000Hz
Signal processing algorithm
Kalman Weighted
Signal Averaging
Number of sweeps
2 runs of 2000 or 1 run of
2 runs of 2000 sweeps
4000 sweeps
Analysis time:
Click stimuli
16ms
16ms
0.5 kHz tone burst stimuli
21.3ms
21.3ms
Artifact rejection
Not applicable
23.3uV
Stimuli parameters for the click-evoked and 0.5 kHz TB recorded stimuli were
identical for both ABR systems. The specifications for the click and the 0.5 kHz
TB stimulus are presented in Table 3.7.
3.7 Procedures
The audiological test battery and the ABR procedures were conducted by the
researcher whom has acquired a B.Communication Pathology degree with
specialization in Speech Therapy and Audiology.
The data that was obtained through the audiometric test battery, i.e. the
tympanograms, ipsilateral acoustic reflexes, DPOAEs as well as behavioural PT
audiometry results, were collected prior to the data collection of the ABR
measurements. As there was a 2 month time difference between the two data
collection dates, tympanometry was conducted prior to ABR measurement of
83
each subject to ensure normal middle ear functioning and to exclude subjects
with any middle ear condition that could possibly influence the ABR results.
Table 3.7:
Stimuli parameters for VS and BL ABR systems
Click-evoked ABR
Parameter
0.5 kHz TB ABR
Settings
Parameter
Settings
Synchronism
Internal
Synchronism
Internal
Stimulus
Click
Stimulus
Tone burst
Stimulus rate
37.7/s
Stimulus rate
37.7/s
Ear tested
Left or Right
Frequency
0.5kHz
Polarity
Rarefaction
Duration
2ms (2-0-2 cycles)
Intensity scale
dB HL
Envelope (ramping)
Blackman
Intensity
Starting at 70dB nHL
Ear tested
Left or Right
Polarity
Alternating
Intensity scale Intensity
dB HL
Intensity
Starting at 70dB nHL
3.7.1 Data collection procedures: immittance, distortion product
otoacoustic emissions and behavioural pure tone audiometry
Immittance, DPOAE measurements and behavioural PT audiometry were
conducted in a double walled soundproof booth at the Speech Therapy
department of the Pretoria School for Children with Cerebral Palsy.
3.7.1.1 Immittance
After careful probe fitting in the subject’s ear canal, tympanometry was first
conducted with immediately following ipsilateral acoustic reflexes.
3.7.1.2 Distortion product otoacoustic emissions
DPOAE measurements were conducted after immittance measurements. After
the probe was fitted in the subject’s ear, he/she was encouraged to remain as
quiet as possible.
84
3.7.1.3 Behavioural pure tone audiometry
PT stimuli were presented monotically, commencing at 40 dB HL if the subject’s
hearing status was unknown.
In some cases the previous records of PT
audiometry results were available to the researcher, in which cases PT
audiometry commenced 20 dB above the determined PT thresholds.
Behavioural PT thresholds were determined descending intensity steps of 10dB
and ascending steps of 5 dB and defined by a 50% response rate at a specific
intensity level.
3.7.2 Data collection procedures: auditory brainstem response
ABR measurements were conducted in a quiet room at the Pretoria School for
children with cerebral palsy. Prior to each subject’s assessment, the noise level
was measured and monitored with a sound level meter (Extech instruments
407730, serial number 9590505). The ABR recording was also preceded by
tympanometry in order to ensure that no middle ear pathology developed since
auditory functioning was assessed 2 months earlier.
Simultaneous click-evoked recordings using the VS and BL ABR systems were
conducted followed by simultaneous 0.5 kHz TB recordings with both systems.
The recordings of the different ABR systems were conducted simultaneously in
order to evaluate both systems as objectively as possible. As discussed in the
research design section, this unique setup was imperative in order to ensure and
maintain an equivalent test condition in terms of EEG, myogenic activity and
environmental noises for both systems. This set-up is illustrated in Photos 3.2 to
3.4.
Furthermore, the researcher alternated the VS and BL ABR systems
between the left and right ears of the subjects to eliminate the order effect.
85
Figure 3.2: Position of the ABR systems (VS ABR system to the left, BL ABR
system to the right)
Figure 3.3 Position of the ABR electrodes as seen from behind (Inverting
electrode of the BL ABR system on the right ear of the subject; ground
and the Invert electrodes of the VS ABR system on the right and left
ear of the subject respectively)
86
Figure 3.4 Position of the ABR electrodes as seen from the front (Fz electrode of
the VS ABR system beneath the Fz electrode of the BL ABR system;
Vivolink attached to subject via cord)
The acquisition parameters of both ABR systems were presented in Table 3.6.
Although the VS and BL ABR systems utilized different technology for the
amplification and filtering processes, similar settings were employed. The VS
system utilized Kalman filtering (averaging) and 2000 to 4000 sweeps were
averaged before the researcher manually stopped the averaging process. The
BL system employed signal averaging and 2000 were averaged and
automatically stopped. Artifact rejection was included in the signal averaging
process of the BL system and not in the VS system because the effect of the
Kalman filtering was investigated.
Prior the ABR recording the subject was asked to sit in a comfortable position on
a chair or to remain in the wheelchair. An age-appropriate DVD was provided to
the subjects during the ABR recording which the subject was encouraged to
watch. The screen of the laptop from which the DVD was played was on a
87
comfortable height for each subject. During the ABR recording, the DVD was
muted.
Prior the ABR recording, each subject’s skin was prepared using alcohol prep
swipes. Following the preparation of the skin, electrode discs were fixed to the
forehead Fz (non-inverting), mastoid ipsilateral (inverting) and mastoid contralateral (ground) for each ABR system separately. The mastoid ipsilateral and
mastoid contra-lateral data collection procedures were switched between
reference and ground depending on the test side since it was a single channel
recording. This setup is illustrated in Figure 3.5 and Figure 3.6 respectively. The
impedance values were kept below 5 kΩ, with less than 3 kΩ difference between
the electrodes. ER 3A -insert earphones of the VS and the BL ABR system were
then placed in the subject’s left and right ear canals.
Inverting electrode of the BL ABR
system
Ground electrode of the VS ABR
system
Figure 3.5: Position of the inverting and ground electrodes of the VS and BL ABR
systems as seen from the side
88
Fz electrodes of the VS ABR system ABR system
(above) and the BL ABR system (below)
ER–3A Insert earphone: VS ABR system
ER-3A insert earphone: BL ABR system
Vivolink of the VS ABR system
Figure 3.6: Position of the Fz electrodes, as well as the ER- 3A insert earphones each
ABR system
ABR recording commenced as soon as the ER 3A-insert earphones of both
systems were inserted in the left and right ear canals respectively. The click
stimulus was first employed with both ABR systems. The 0.5 kHz TB stimulus
was presented directly after a click-evoked threshold was determined by a
specific ABR system.
Stimulation (click and 0.5 kHz tone burst stimuli) was
presented bilaterally at an above threshold intensity of 70 dB nHL. In cases
where the behavioural PT indicated a moderate or severe hearing loss,
stimulation was presented at an above threshold level of 80 dB nHL or 90 dB
nHL. In order to establish ABR thresholds, the researcher descended in intensity
steps of 10 dB and ascended in steps of 5 dB. However, in cases where there
was a time constraint or the subject presented with consistent muscular activity,
the researcher descended in intensity steps of 10 dB only to obtain the ABR
threshold.
The specific descending method of 10 dB and on occasion the
ascending method of 5 dB was kept consistent between the click and 0.5 kHz TB
recordings of each subject.
89
3.7.3 Procedures for analysis of auditory brainstem response
Repeatability of at least two ABR waveforms had to be recorded in succession
with the same measurement conditions, i.e. no change in stimuli, intensity, rate
or polarity in one ear to be regarded as an ABR response (Hall, 2007). The
presence of wave V of the ABR waveform was also confirmed by replication of
the waveform. Wave V had to be observed in the same latency region in at least
two separately averaged waveforms (Hall, 2007). The ABR threshold was taken
at the last intensity level at which the ABR response was successfully repeated
and the wave V was still identifiable. Furthermore, the recordings were analyzed
offline by two clinical audiologists who are experienced in ABR recordings. This
objective analysis was an attempt to ensure a minimum of problems in
interpreting the responses.
3.7.4 Procedures for analysis of the recording time of the auditory
brainstem response recordings
The recording time of both systems was determined offline.
In order to
determine the recording time of the VS system the total amount of sweeps per
recording (click-evoked or 0.5 kHz tone burst recording) up to one intensity lower
than where the threshold were calculated and then divided by the rate of the
stimuli (37.7s). The recording time of the BL system was calculated in a similar
way, although the total artifact rejections per recording were also considered
when calculating the total number of sweeps.
3.8.
Data processing
The raw data were organized on Microsoft XP Worksheets and were analyzed
with statistical measures including descriptive and inferential statistics by the
Department Statistics at the University of Pretoria. Whilst descriptive statistics
includes the sorting, ordering and summarizing of data by means of graphs and
tables, inferential statistics evaluates, contemplates and draws conclusions about
the population from which the sample was drawn (Leedy & Ormrod, 2005;
Neuman, 1997).
90
3.9
Data analysis
In order to evaluate the clinical value of the Vivosonic Integrity ABR system in the
auditory functioning of children with CP, each subject’s individual performance
was described for each procedure. The results obtained during acoustic reflex
and DPOAE measurements assisted the researcher in describing the auditory
status of the subject more comprehensively. The collective results for all the
subjects were taken into account in analyzing and processing the behavioural
PT, click-evoked and 0.5 kHz TB thresholds (both ABR systems).
The focus of this study was to compare the VS ABR system to the BL ABR
system in terms of feasibility, threshold correspondence and recording time using
click and 0.5 kHz TB stimuli. In terms of the threshold correspondence, the
difference between the thresholds of the ABR systems and behavioural PT
thresholds provided an indication of how close to each other the PT threshold the
ABR thresholds were.
All subject data were collectively analyzed using
descriptive statistics. The analysis of the data was done at the Department of
Statistics at the University of Pretoria. The data analysis included the following:
3.9.1 Describing the feasibility and characteristics of an audiometric test
battery
™ Determining the number of subjects in which ipsilateral acoustic reflexes,
DPOAE measurements and behavioural PT audiometry were feasible.
™ Determining the amount of present, elevated and absent ipsilateral reflexes
per ear and per frequency for subjects with spastic CP, athetiod CP and
microcephaly.
™ Determining the amount of present, abnormally reduced and absent DPOAE
by calculating the difference between the amplitude of the DPOAE and the
noise floor level.
™ Calculating the distribution of thresholds (mean, standard deviation and
range) for 30 ears (n=15) per stimulus frequency as recorded by behavioural
PT audiometry.
91
3.9.2 Comparing the VS ABR system with a conventional ABR system
The VS ABR system was compared to the BL ABR system in terms of feasibility,
electrophysiological thresholds, correspondence to behavioural PT thresholds
and recording time when using click and 0.5 kHz TB stimuli. The data analysis
for these comparisons is discussed under separate headings:
3.9.2.1 Feasibility of the VS and BL ABR systems using click and 0.5 kHz
tone burst (TB) stimuli
™ Comparing the successful VS ABR recordings using click and 0.5 kHz TB
stimuli with the successful BL ABR recording using corresponding stimuli by
using the Fisher exact p one-tailed test (Steyn, Smit, Du Toit & Strasheim,
1998).
3.9.2.2 Electrophysiological thresholds of the VS and the BL ABR systems
using click and 0.5 kHz TB stimuli
™ Comparing the electrophysiological thresholds obtained with the VS ABR
system to the click-evoked and 0.5 kHz TB thresholds obtained BL ABR
system in terms of the mean and range.
3.9.2.3 Threshold correspondence of the VS and BL ABR systems to
behavioural PT thresholds
The organization of the threshold correspondence data for statistical analysis is
presented in Figure 3.7.
™ Determining the difference between the click-evoked threshold of the VS and
BL ABR systems and behavioural thresholds for the corresponding ear at 2
kHz, 4 kHz and the average of 2 kHz and 4 kHz, for each subject for whom
PT thresholds were obtained.
™ Determining the normal distribution (mean, standard deviation and range) of
the differences for the VS and BL ABR systems at 2 kHz, 4 kHz and the
average of 2 kHz and 4 kHz.
92
Manipulated variable
Dependent variable
VS ABR system
Click stimuli
Outcome
Click-evoked
threshold
Di
Difference
PT audiometry
2 and 4 kHz Pure tone stimuli
2 kHz threshold
4 kHz threshold
Average threshold
(2 and 4 kHz)
Compare
Difference
BL ABR system
Click stimuli
VS ABR system
0.5 kHz Tone burst stimuli
Click-evoked
threshold
0.5 kHz TB
threshold
Difference
PT audiometry
0.5 kHz Pure tone stimuli
0.5 kHz threshold
Compare
Difference
BL ABR system
Figure 3.7:
0.5 kHz Tone burst stimuli
0.5 kHz TB
threshold
Organization of threshold difference data for statistical analysis
93
™ Determine the difference between the 0.5 kHz TB threshold of the VS ABR
and the behavioural PT threshold of the corresponding ear at 0.5 kHz, as well
as the 0.5 kHz TB threshold of the BL system and the behavioural PT
threshold of the corresponding ear at 0.5 kHz for each subject.
™ Determining the normal distribution (mean, standard deviation and range) of
the differences for the VS (15 ears) and BL ABR (15 ears) systems at 0.5
kHz.
™ Establishing the statistical significance of the differences between the
averages of the thresholds for 15 points (15 thresholds of the VS ABR system
and 15 thresholds of the BL ABR system) per stimulus frequency by
implementing inferential statistics.
Two-sample comparisons were made
between the differences obtained with the various measures as illustrated in
Figure 3.2. The comparisons were done by using the Wilcoxon Signed RankTest (Mawell & Satake, 1997). This test was used because of the small
sample size (Delport, 1998).
It is a powerful test in the sense that the
determined p-value can indicate the magnitude and the direction (positive or
negative) of the differences between the various data sets (Delport, 1998;
Johnstone & Pennypacker, 1993).
3.9.2.4 Recording time of the VS and BL ABR systems using click and 0.5
kHz TB stimuli
™ Determining the recording time per subject and per measurement (i.e. per
ear) using click and 0.5 kHz TB stimuli.
™ Determining the normal distribution (mean, standard deviation and range) for
the recording time of the VS and BL systems using click stimuli.
™ Determining the average, standard deviation and range of normal deviation
for the recording time of the VS and BL systems using 0.5 kHz TB stimuli
™ Calculating the difference between the average recording times of the VS and
BL systems using click and 0.5 kHz TB stimuli.
™ Establishing the statistical significance of the differences between the mean
recording times per stimulus frequency through inferential statistics.
94
3.10
Reliability and validity
Scientific research endeavors to draw valid and reliable conclusions about
underlying relations among variables based on the empirical test procedures
(Maxwell & Satake, 2007:40).
Reliability refers to the consistency of a
measurement (Drummond, 2003).
In this research various procedures
(behavioural PT thresholds and ABR thresholds) were used to measure the
same phenomenon. The procedures were conducted with the same subjects
using the same protocol.
Validity refers to the extent to which the instrument measures what it is supposed
to measure and includes internal and external validity. Whereas internal validity
relates to the issue whether independent variables were responsible for
variations in the dependent variable, external validity pertains to the extent to
which results can be generalized to one or more populations (Maxwell & Satake,
1997:44).
In this study the independent variables – behavioural PT audiometry and ABR
measurements with the VS and BL ABR systems – were kept consistent by using
the similar protocol throughout the research. Furthermore, as described in the
research design, controlled variables were applied to the experimental setting.
These controlled variables were applied to ensure a stable context for clear
visualization of effects on the independent variables. The factors controlled by
the researcher were:
™ Disability: Subjects diagnosed with CP were selected for this study
™ Age: Subjects between the ages of 12 to 18 years were selected for this study
™ Middle ear functioning: Subjects were required to have normal middle ear
functioning as defined by a Type A tympanogram
External validity in this study would particularly entail the problem of
interpretation. A potential threat to this research lies in the comparison between
the VS and BL ABR systems in the research sample. In order for the comparison
95
to be valid, both systems were subjected to similar test conditions. Therefore,
different ears were used while the ABR recordings were conducted
simultaneously in each subject.
Additionally, the threat to interpretation was reduced by introducing additional
researchers to analyze ABR results objectively.
Both of the additional
researchers were well acquainted with electrophysiological auditory procedures,
including the ABR.
3.11
Conclusion
This chapter described the research process that was followed to determine the
clinical value of the VS ABR system when assessing children with CP. Each
procedure in the audiometric test battery was explained. It was also explained
that data obtained from the auditory test battery procedures was exploited for
descriptive purposes. Additionally, the protocols and procedures of ABR clickevoked and 0.5 kHz TB recordings utilizing different ABR systems were
described. The importance of simultaneous ABR recordings of the different ABR
systems was emphasized.
3.12
Summary
The main aim and sub aims of the study were set out in this chapter, followed by
a summary of the organization of the research process. The methodological
framework for obtaining electrophysiological behavioural thresholds from a
normative sample was described and was concluded by the inclusion of the
normative values (click and 0.5 kHz TB stimuli) for the VS and the BL ABR
systems. The methodological framework for the research sample was described,
commencing with a discussion of the research design and ethical considerations.
The criteria and procedures for selecting the subjects were discussed and a
description of each subject was presented in table format.
The apparatus,
material and protocols that were implemented were described. Subsequently,
96
the procedures for analyzing and processing the data were discussed.
The
chapter concluded with a review of reliability and validity in relation to this study.
97
Chapter 4
RESULTS
The aim of this chapter is to present the results of the empirical
research.
4.1
Introduction
To determine the clinical value of the Vivosonic Integrity (VS) auditory
brainstem response (ABR) system when assessing auditory functioning in
children with cerebral palsy (CP) different sets of data collection and analysis
procedures were required. This chapter provides the results of the empirical
research according to the sub-aims as discussed in Chapter 3.
4.2
Sub-aim 1 results: Characteristics and feasibility of an
audiometric test battery in children with cerebral palsy
The first sub-aim focused on the applicability of the auditory test battery in a
small sample of children with CP. The test battery in this study consisted of
immittance and distortion product otoacoustic emissions (DPOAE), and
behavioural pure tone (PT) audiometry. In the subsequent text, the results of
each auditory procedure are provided according to its feasibility and
characteristics in this sample. Table 4.1 provides a summary of the feasibility
of each auditory procedure in the sample of 15 subjects.
Within Table 4.1 results obtained from each procedure have been categorized
according to “reliable results obtained”; “unreliable results obtained” and “no
results obtained”. For tympanometry measurements the term “reliable results”
suggested that a tympanogram with clear indication of the middle ear
pressure, compliance and middle ear volume was obtainable from the subject.
For acoustic reflexes the term “reliable results” implied that acoustic reflexes
were elicited at each test frequency, regardless of the intensity at which the
reflex was elicited. In cases where acoustic reflexes could not be elicited due
98
to excessive muscle movements or due to a profound hearing loss, results
were noted as “no results obtained”. For DPOAE measurements “reliable
results” suggested that emissions could have been elicited and that emissions
were present with normal or abnormal amplitudes. In cases where emissions
could not have been elicited due to excessive muscular movements or due to
a profound hearing loss results were noted as “no results obtained”.
For
behavioural PT audiometry “reliable results obtained” suggested that subjects
responded consistently to pure tone stimuli.
In cases where subjects
responded inconsistently to PT stimuli, results were noted as unreliable.
The feasibility of the click-evoked and 0.5 kHz tone burst (TB) ABR recordings
by means of the VS ABR and the Bio Logic Navigator Pro (BL) ABR systems
is also included in this table. Reliable results obtained with the click-evoked
and 0.5 kHz TB ABR procedures indicate that the ABR wave V was with
repetition obtainable at the lowest intensity level that was regarded as the
threshold in nHL. In cases where the ABR wave V was absent or where
muscle activity was excessive, results were noted as “no results obtained”.
4.2.1 Immittance measurements
Immittance measurements in this study consisted of tympanometry and
ipsilateral acoustic reflexes. As demonstrated in Table 4.1, tympanometry
was conducted and measurable in 15/15 of the subjects. All the subjects
(100%) presented with normal middle ear functioning as defined by a type A
tympanogram.
Tympanometry was followed by ipsilateral acoustic reflex measurements.
Although it was conducted in 15 subjects, Subject 15 was not testable due to
persistent involuntary movements which caused continuous high levels of
internal noise.
Acoustic reflex measurements were therefore obtained in
14/15 subjects.
99
Table 4.1:
A summary of the feasibility of the various auditory procedures employed in the current study
Auditory procedures
Subject
Type of
CP
Tympanometry
Acoustic reflexes
DPOAE
Behavioural PT
audiometry
1
Spastic
Y
Y
Y
Y
Y
Y
Y
Y
2
Spastic
Y
Y
Y
Y
Y
Y
Y
Y
3
Y
Y
Y
Y
Y
Y
0
Y
4
Microcepha
ly
Spastic
Y
Y
Y
Y
Y
Y
Y
Y
5
Athetosis
Y
Y
Y
Y
Y
Y
Y
Y
6
Spastic
Y
Y
Y
X
Y
Y
Y
Y
7
Spastic
Y
Y
Y
X
Y
Y
Y
Y
8
Athetosis
Y
Y
Y
X
Y
Y
Y
Y
9
Spastic
Y
Y
Y
X
Y
Y
Y
Y
10
Spastic
Y
Y
Y
Y
Y
Y
Y
Y
11
Spastic
Y
Y
Y
Y
Y
Y
Y
Y
12
Spastic
Y
Y
Y
X
0
Y
0
Y
13
Spastic
Y
Y
Y
X
0
Y
0
Y
14
Athetosis
Y
Y
Y
X
0
0
0
0
15
Athetosis
Y
0
0
X
0
0
0
0
Click-evoked ABR
BL ABR
VS ABR
0.5kHz TB ABR
BL ABR
VS ABR
Key: Y = Reliable results obtained (= X = unreliable/inconsistent results; 0 = no results obtaine
100
Frequency and ear specific acoustic reflex data were categorized in terms of
normal, elevated and absent reflexes as discussed in Chapter 3.
As this
procedure was unsuccessful in Subject 15, another category, 'Could not test',
was created. Right and left ear data are presented according to these categories
Subjects
in Figures 4.1 and 4.2 respectively.
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Normal acoustic
reflexes
Elevated acoustic
reflexes
Acoustic reflex at 0.5kHz
Figure 4.1:
Absent acoustic
reflexes
Acoustic reflex at 1kHz
Could not test
Acoustic reflex at 2kHz
Distribution of right ipsilateral acoustic reflexes
As shown in Figures 4.1 and 4.2 the majority of subjects (9/15 and 7/15)
Subjects
presented with normal acoustic reflexes at 0.5 kHz bilaterally.
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Normal acoustic
reflexes
Elevated acoustic
reflexes
Acoustic reflex at 0.5kHz
Figure 4.2:
Absent acoustic
reflexes
Acoustic reflex at 1kHz
Could not test
Acoustic reflex at 2kHz
Distribution of left ipsilateral acoustic reflexes
101
Five subject (n=15) presented with elevated acoustic reflexes at 1 kHz in the left
and the right ears whilst 6/15 and 5/15 subjects presented with elevated acoustic
reflexes at 2 kHz. Two subjects presented with absent acoustic reflexes at 0.5
kHz, 1 kHz and 2 kHz in the right ear (Subject 5 and Subject 14). In the left ear,
four subjects (Subjects 5, 6, 12 and 14) presented with absent acoustic reflexes
at 1 kHz and 2 kHz, whilst the acoustic reflexes at 0.5 kHz were absent in five
subjects (Subjects 3, 5, 6, 12 and 14).
Subject 5 and Subject 14 were the only ones whom presented with absent
acoustic reflexes bilaterally.
Both these subjects were diagnosed with the
athetoid type of CP. Table 4.2 presents a summary of the acoustic reflexes
obtained in the different CP sub-groups of the research sample.
Table 4.2:
Summary of ipsilateral acoustic reflexes obtained in different CP
sub-groups of the research sample
Spastic CP (n=10; 20 ears)
Normal
Elevated
Absent
Could not test
Left ear (30 acoustic reflexes)
Right ear (30 acoustic reflexes)
15/30
9/30
6/30
-
19/30
11/30
-
Left ear (12 acoustic reflexes)
Right ear (12 acoustic reflexes)
2/12
1/12
6/12
3/12
3/12
6/12
3/12
Left ear (3 acoustic reflexes)
Right ear (3 acoustic reflexes)
2/3
1/3
-
3/3
-
Athetosis CP (n=4; 8 ears)
Normal
Elevated
Absent
Could not test
Microcephaly (n=1; 2 ears)
Normal
Elevated
Absent
Could not test
102
4.2.2 Distortion product otoacoustic emissions
DPOAE measurements were conducted after the acoustic reflex procedure.
Although DPOAE were administered in all the subjects (n=15), results were only
obtained in 14/15 subjects as shown in Table 4.1. Figure 4.3 illustrates that,
whilst acceptable noise floor levels (i.e. below 3 dB SPL) were attained during
the DPOAE measurements in the majority of the research sample (n=14), the
noise floor levels exceeded 3 dB SPL in one subject (Subject 15). Subject 15
presented with consistent involuntary movements of especially the neck and jaw.
Nois e floor
le ve l be low 3
dB SPL
(n= 14)
Nois e floor
le ve l above 3 dB
SPL
(n= 1)
Subjects
Figure 4.3:
Noise floor levels attained during DPOAE measurements (n=15)
A total of 224 DPOAE frequencies were measured for 28 ears (n=14). Figure 4.4
provides an illustration of measured DPOAEs that were present, abnormally
reduced and absent.
As shown in Figure 4.4, 35 of 224 emissions (16%) were absent. There was a
greater preponderance of absent emissions in the 2002 Hz to 6346 Hz frequency
region (46%) followed by 31% and 23% in the low (635 Hz to 808 Hz) and mid
frequencies (1001 Hz to 1586 Hz) respectively. Two subjects (Subject 5 and
Subject 14) presented with absent emissions across the frequency spectrum.
Both of these subjects presented with the athetiod type of CP as shown in Table
4.1.
103
Absent DPOAEs
16%
Present but
abnormal DPOAEs
5%
Present DPOAEs
79%
Figure 4.4:
Distribution of DPOAE in the current research as specified by
Distortion product (DP) – Noise Floor (NF) criteria (DP-NF < 6dB =
absent DPOAE; DP-NF = 6-9dB = present, but abnormally reduced
DPOAE; DP-NF > 10dB = present DPOAE)
Only a small portion (11 out of 224 or 5%) of the emissions was present, though
abnormally reduced. The majority of the reduced emissions (45%) were noted in
the high frequencies (2002 Hz to 6346 Hz) followed by 36% and 18% in the low
(635 Hz to 808 Hz) and mid (1001 Hz to 1586 Hz) frequencies respectively. The
majority of the DPOAEs (79%) occurred within the normal amplitude range as
specified by the Vanderbilt criteria.
There was, however, a percentage of
emissions with amplitudes (11% in the left ear and 8% in the right ear) that
exceeded the Vanderbilt 95th percentile values at certain frequencies.
This
occurred in six subjects as illustrated in Figure 4.5.
Interestingly, elevated DPOAEs were only obtained from subjects diagnosed with
spastic CP. Elevated DPOAE amplitudes were obtained from 6/10 of spastic CP
subjects.
As depicted in Figure 4.5, the amplitude of the emissions was
especially elevated between 635 Hz and 1586 Hz bilaterally for Subject 7,
Subject 11 and Subject 13. The amplitude of the emissions was elevated in
104
either the left or the right ear at similar frequencies for Subject 2, Subject 4 and
Right ear: DP amplitude (dB SPL)
Subject 10.
25
Vanderbilt DP 95th
percentile
20
Vanderbilt DP 5th percentile
15
Subject 2
10
5
Subject 4
0
Subject 7
-5
-10
Subject 10
-15
-20
Subject 11
-25
635Hz
808Hz 1001Hz 1586Hz 2002Hz 3174Hz 4003Hz 6347Hz
Subject 13
Frequency (Hz)
Left ear: DP amplitude (dB SPL)
25
Vanderbilt DP 95th
percentile
20
Vanderbilt DP 5th percentile
15
10
Subject 2
5
Subject 4
0
-5
Subject 7
-10
-15
Subject 10
-20
Subject 11
-25
635Hz 808Hz 1001Hz 1586Hz 2002Hz 3174Hz 4003Hz 6347Hz
Subject 13
Frequency (Hz)
Figure 4.5:
DPOAEs with elevated amplitudes in at least one ear of a subject
To investigate the extent of the increased DPOAE amplitudes obtained in these
subjects, the difference between the amplitudes of the DPOAE and the 95th
percentile of the Vanderbilt criteria was calculated for each test frequency and is
displayed in Table 4.3 and Table 4.4. As shown in Table 4.3 and Table 4.4 the
amplitude differences were larger for the low and mid frequencies (635 Hz to
1586 Hz) for these subjects than the differences for the higher frequencies (2002
Hz to 6347 Hz).
105
Table 4.3: The difference between the DPOAE amplitude (dB SPL) of the left ear and the 95th percentile of the Vanderbilt criteria at
various test frequencies of subjects in whom elevated emissions were obtained
Frequency
635Hz
808Hz
1001Hz
1586Hz
2002Hz
3174Hz
4003Hz
6347Hz
Key:
95th
Percentile
11.16
11.05
12.28
12.76
8.9
7.75
7.91
3.95
Subject 2
Subject 4
Subject 7
Subject 10
Subject 11
DP
Amplitude
Difference
DP
Amplitude
Difference
DP
Amplitude
Difference
DP
Amplitude
Difference
DP
Amplitude
Difference
11
11
16
16
9
3
-6
-9
-0.16
-0.05
3.72
3.24
0.1
-4.75
-13.91
-12.95
18
18
17
16
11
10
4
4
6.84
6.95
4.72
3.24
2.1
2.25
-3.91
0.05
21
22
22
12
12
8
1
-17
9.84
10.95
9.72
-0.76
3.1
0.25
-6.91
-20.95
15
15
14
9
5
8
-3
-7
3.84
3.95
1.72
-3.76
-3.9
0.25
-10.91
-10.95
15
15
17
12
8
9
2
-3
3.84
3.95
4.72
-0.76
-0.9
1.25
-5.91
-6.95
th
Indicates DP amplitudes (in dB SPL) that exceeded the 95 percentile of the Vanderbilt criteria and the extent of the difference
106
Table 4.4: The difference between the DPOAE amplitudes (dB SPL) of the right ear and the 95th percentile of the Vanderbilt criteria
at various test frequencies of subjects in whom elevated emissions were obtained
Frequency
635Hz
808Hz
1001Hz
1586Hz
2002Hz
3174Hz
4003Hz
6347Hz
Key:
95th
Percentile
11.16
11.05
12.28
12.76
8.9
7.75
7.91
3.95
Subject 2
Subject 4
DP
Amplitude
Difference
DP
Amplitude
Difference
5
5
5
5
2
6
-4
-9
-6.16
-6.05
-7.28
-7.76
-6.9
-1.75
-11.91
-12.95
-3
-3
1
17
6
0
-2
-9
6.84
6.95
4.72
4.24
-2.9
-7.75
-9.91
-12.95
Subject 7
Subject 10
Subject 11
DP
Amplitude
Difference
DP
Amplitude
Difference
DP
Amplitude
Difference
19
19
22
19
13
13
1
1
7.85
7.95
9.72
6.24
4.1
4.25
-6.91
-2.95
11
11
16
11
4
5
-8
-20
0.16
0.5
3.72
-1.76
-4.9
-2.25
-15.91
-23.95
17
17
14
10
0
1
-13
-20
5.84
5.95
1.72
-12.76
-8.9
-6.75
-20.91
-23.95
th
Indicates DP amplitudes (in dB SPL) that exceeded the 95 percentile of the Vanderbilt criteria and the extent of the difference
107
4.2.3 Behavioural pure tone audiometry
Behavioural pure tone (PT) audiometry was conducted between 0.5 kHz to 4
kHz. Table 4.5 provides behavioural PT thresholds (left and right ears) for each
subject (n=15 subjects, 30 ears).
Table 4.5:
Subject
Behavioural PT thresholds (in dB HL) for each subject (n=15)
Left ear PT threshold (dB HL)
0.5 kHz
Spastic CP
1
2
4
6
7
9
10
11
12
13
Athetotic CP
5
8
14
15
Microcephaly
3
Right ear PT threshold (dB HL)
1 kHz
2 kHz
4 kHz
0.5 kHz
5
5
0
5
10
0
0
0
10
0
10
0
0
0
15
10
5
0
15
10
5
5
0
10
5
10
5
10
15
5
10
5
0
15
5
5
10
10
5
5
5
5
0
10
10
0
0
5
5
5
25
5
0
55
40
0
5
55
60
10
5
55
65
5
0
55
10
0
5
10
1 kHz
2 kHz
4 kHz
10
5
0
5
10
0
10
5
5
0
15
5
0
10
15
5
0
0
20
0
20
10
0
0
15
5
5
5
5
0
25
5
10
40
40
5
10
40
60
5
5
65
65
10
5
70
10
5
10
10
The consistency of each subject’s responses to the PT stimuli played an integral
role in the study, especially in determining the correspondence of the
electrophysiological thresholds of each ABR system to the PT stimuli.
Accordingly, the subjects were categorized according to the consistency/reliability
of their behavioural responses as shown in Table 4.6.
As depicted in Table 4.6 inconsistent behavioural responses were obtained from
three subjects with athetosis with only Subject 5 responding reliably. Behavioural
responses of the subjects with spastic CP were equally distributed between
Group A (5 subjects) and B (5 subjects). The only subject with microcephaly,
Subject 3, responded consistently to the PT stimuli.
108
Table 4.6:
Consistency of behavioural PT responses (0.5 kHz to 4 kHz)
(n=15 subjects)
Subject
Spastic CP
1
2
4
6
7
9
10
11
12
13
Athetotic CP
5
8
14
15
Microcephaly
3
GROUP A
GROUP B
(Subjects with consistent
(Subjects with inconsistent
responses)
responses)
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
The mean behavioural PT thresholds of subjects in Group A are presented in
Figure 4.6.
Investigating the mean behavioural PT thresholds of subjects in
Group A revealed that the inter-aural differences were slight and ranged between
0 dB and 3 dB.
The majority of the behavioural thresholds obtained from
subjects in Group A were equal to, or less than 10 dB HL with the exception of
Subject 5. This subject presented with a moderate to severe high frequency
hearing loss (25 dB HL to 65 dB HL) bilaterally. Subsequently, the SD of the
mean PT thresholds across the frequency spectrum varied between 13 dB HL to
19 dB HL and 11 dB HL to 22 dB HL in both ears.
Although the subjects in Group B responded inconsistently to behavioural PT
stimuli, thresholds (or rather minimal response levels) were still obtained and are
reported below. The mean behavioural thresholds of subjects in Group B are
presented in Figure 4.7. In this group, the inter-aural differences ranged between
0 dB and 5 dB. The majority of the thresholds were equal to, or less than 15 dB
109
HL, with the exception of Subject 15. This subject presented with a moderate to
severe high frequency hearing loss (mean behavioural thresholds of 53 dB HL
and 48 dB HL in the left and right ear respectively).
Intensity (dB HL)
55
50
45
40
35
30
25
20
15
10
5
0
S1
S2
S3
S4
S5
S10
S11
Subjects
Left PT thresholds
Figure 4.6:
Right PT thresholds
Subjects in Group A: Mean behavioural PT thresholds for 0.5 kHz to
Intensity (dB HL)
4 kHz
60
55
50
45
40
35
30
25
20
15
10
5
0
S6
S7
S8
S9
S12
S13
S14
S15
Subjects
Left PT thresholds
Figure 4.7:
Right PT thresholds
Subjects in Group B: Mean behavioural PT thresholds for 0.5 kHz to
4 kHz
110
4.3
Sub-aim 2 results: Comparing the Vivosonic Integrity ABR
system with a conventional ABR system
The Vivosonic Integrity ABR system (VS) was compared to a conventional ABR
system, in this instance the Bio Logic Navigator Pro system. The VS system was
compared to the BL ABR system in terms of the feasibility, the threshold
correspondence to behavioural PT thresholds as well as the recording time. The
results of each of these comparisons are presented separately in the following
section.
Prior to the presentation of these results, information regarding the
state of each subject’s awareness, the sound level that was measured prior to
each ABR recording and the inter-aural differences of the ABR components (that
is the latency differences between the wave I, wave V and wave I-V for each ear
i.e. for each ABR system) is presented in Table 4.7.
As shown in Table 4.7, all the subjects were awake during the ABR recordings.
Whilst the majority of the subjects presented with sporadic muscular movements,
Subject 12, Subject 13 and Subject 14 displayed constant muscular movements.
The environmental noise was minimal (below 40 dB A) prior to each ABR
recording as measured by the sound level meter. The sound level meter did not
test below 40 dB A.
Because the VS and the BL ABR systems were compared to each other by
comparing the left and right ears, it was important to determine inter-aural
differences of the ABR wave I, wave V and the absolute latency (wave I-V) for
each subject. From Table 4.7 it is apparent that the mean inter-aural differences
for wave I, wave V and wave I-V were slightly larger when compared to the mean
inter-aural differences obtained from normal hearing subjects with no CP (wave I
= -0.02 ±0.08; wave V = 0.00 ±0.11; wave I-V = 0.00 ±0.11) (Hall, 2007).
4.3.1 Feasibility of the VS and BL ABR systems using click and 0.5 kHz TB
stimuli
111
As stated in Chapter 3, the VS and BL ABR systems were simultaneously
conducted in all 15 subjects. The VS ABR system was implemented for 9/15
right and 6/15 left ears whereas 6/15 right ears and 9/15 left ears were tested
with the BL system.
Table 4.7:
Information related to the state of awareness, sound level and interaural latency differences (n=15 subjects)
Inter-aural difference
(ms)
Subject
State of awareness
1
Awake, sporadic
movements
Awake, sporadic
movements
Awake, sporadic
movements
Awake, sporadic
movements
Awake, sporadic
movements
Awake, sporadic
movements
Awake, sporadic
movements
Awake, sporadic
movements
Awake, sporadic
movements
Awake, sporadic
movements
Awake, sporadic
movements
Awake, constant
movements
Awake, constant
movements
Awake, sporadic
movements
Awake, constant
movements
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Sound level
(dB A)
<40
Wave I
Wave V
Wave I-V
0.1
1
1
<40
0.1
0.1
0
<40
0.6
0.1
0.5
<40
**
0.2
**
<40
0.2
0.3
0.1
<40
0
0
0.1
<40
0.1
0.2
0.4
<40
0
0
0
<40
0.1
0.1
0.1
<40
0.1
0.1
0.1
<40
0.1
0.3
0.2
<40
**
**
**
<40
**
**
**
<40
**
0
**
<40
**
**
**
Mean ±SD:
0.14±0.2
(n= 10)
Mean ±SD: 0.2±0.3
(n=12)
Mean ±SD:
0.3±0.3
(n=10)
** Indicates that the ABR wave I or wave V was not obtainable with one or both of the ABR systems, with the result that
latencies could not be determined
112
The number of successful click-evoked and 0.5 kHz tone burst (TB) recordings
using the VS and BL ABR systems is illustrated in Figure 4.8. In general, both
ABR systems were more successful with click stimuli than with 0.5 kHz TB
stimuli. Assessment of the success rate of the systems individually, showed that
more subjects were testable with the VS ABR system than with the BL ABR
system.
The VS ABR system was successful in 14/15 and 13/15 subjects using click and
0.5 kHz TB stimuli respectively.
In the case of the BL ABR system, it was
possible to obtain electrophysiological hearing thresholds for 12/15 subjects
(using click stimuli) and for 11/15 subjects (using 0.5 kHz TB stimuli). These
findings are illustrated in Figure 4.8.
Total subjects
15
0
Vivosonic Integrity Bio Logic Navigator
device
Pro device
Click stimuli
Figure 4.8:
0.5 kHz tone burst stimuli
Successful click-evoked and 0.5 kHz TB ABR recordings using the
VS and BL ABR systems
The VS system was successful in more cases than the BL system using click and
0.5 kHz TB stimuli. However, to determine whether the VS ABR system showed
a statistically significant better success rate than the BL system (for click and 0.5
kHz TB stimuli), the Fisher Exact P One Tailed test was used. Although results
113
showed no statistically significant value for click (p=.1121) and 0.5 kHz TB stimuli
(p=.1648), there is a tendency towards the 95% confidence level in both cases.
The strong tendency towards a statistically significant difference between the VS
and BL successful ABR recordings (using click and 0.5 kHz TB stimuli) suggests
that the VS ABR system may produce a statistically significant success rate for
click as well as for 0.5 kHz TB stimuli, provided a larger sample is tested.
A number of unsuccessful ABR recordings occurred in the study. The cases in
which ABR recordings were unsuccessful are presented in Table 4.8. The VS
ABR system was unsuccessful in one subject for the click-evoked ABR and in
two subjects for the 0.5 kHz TB ABR. The BL ABR system proved successful in
three subjects for the click-evoked ABR and in four subjects for the 0.5 kHz tone
burst ABR.
Table 4.8:
Subjects with unsuccessful ABR recordings
VS ABR system
Subjects
Type of CP
Click stimuli
0.5 kHz TB
BL ABR system
Click stimuli
stimuli
3
Microcephaly
13
Right
0.5 kHz TB
stimuli
X
X
X
X
X
X
X
X
X
hemiplegia
14
Athetosis
15
Athetosis
X
As shown in Table 4.8 the BL ABR system, using click stimuli, was unsuccessful
in Subject 13, Subject 14 and Subject 15, whilst the 0.5 kHz TB BL ABR
recordings were inconclusive in Subject 3, Subject 13, Subject 14 and Subject
15. The VS ABR system was unsuccessful in Subject 14 (using 0.5 kHz TB
stimuli) and Subject 15 (using 0.5 kHz TB and click stimuli).
114
Noteworthy was the percentages of artifact rejection by the BL ABR system for
three of the four subjects mentioned above.
The acceptable percentage of
artifact rejection for ABR recordings is an upper limit of 20% (Hall, 2007). The
artifact rejection percentage in Subject 3, Subject 12, Subject 13 and Subject 15
exceeded this value in click-evoked and/or 0.5 kHz TB recordings as shown in
Figure 4.9.
These subjects displayed sporadic involuntary body movements,
especially head, neck and jaw movements which probably caused high levels of
internal noise.
100
% Artifact rejection
90
80
70
60
50
40
30
20
10
Click stim uli
Figure 4.9:
je
ct
10
Su
bj
ec
t1
1
Su
bj
ec
t1
2
Su
bj
ec
t1
3
Su
bj
ec
t1
4
Su
bj
ec
t1
5
je
ct
9
Su
b
je
ct
8
Su
b
Su
b
je
ct
7
Su
b
je
ct
6
je
ct
5
Su
b
je
ct
4
Su
b
je
ct
3
Su
b
Su
b
je
ct
2
Su
b
Su
b
je
ct
1
0
0.5 kHz Tone burst stim uli"
Percentage of rejected sweeps in the BL ABR system in subjects
with unsuccessful ABR recordings using click and 0.5 kHz TB
stimuli
The high percentage of artifact rejection might have contributed to the
unsuccessful ABR recordings in Subject 3, Subject 13 and Subject 15. However,
the percentage of artifact rejection was well below the acceptable percentage of
20% in Subject 14 (2.68% and 5.93%).
4.3.2 Electrophysiological thresholds of the VS and BL ABR systems using
click and 0.5 kHz TB stimuli
Table 4.9 provides the mean thresholds, standard deviations, and the minimum
and maximum thresholds (in dB nHL) for both systems (See Chapter 3 paragraph
115
3.6.2.4 and Table 3.5 for the calculation of the dB nHL reference values for click
and 0.5 kHz TB stimuli for both systems).
The average click-evoked threshold obtained with the VS system was 28 dB nHL
with a SD of ±18 dB nHL. The large SD consequently leads to a wide range of
normal deviation between 10 dB nHL to 46 dB nHL. The mean threshold of the
BL system for click stimuli was 21 dB nHL with a SD of ±14 dB nHL. The range
of normal deviation for the click-evoked thresholds obtained with the BL system
was therefore 7 dB nHL to 35 dB nHL.
Table 4.9:
The distribution of the ABR thresholds (dB nHL) of the VS ABR and
BL ABR systems using click and 0.5 kHz TB stimuli
Sample
Mean
SD
Minimum
Maximum
(N)
Click stimuli
VS system
14
28
±18
7
67
BL system
12
21
±14
-1
49
0.5 kHz TB
stimuli
VS system
13
34
±15
17
63
BL system
11
37
±18
15
75
The mean 0.5 kHz TB threshold for the VS system was 34 dB nHL with a SD of
±15 dB nHL. Hence, the normal range of deviation for the 0.5 kHz TB recording
with the VS system was between 19 dB nHL to 53 dB nHL. The average 0.5 kHz
tone burst threshold for the BL system was 37 dB nHL with a SD of ±18 dB nHL.
The large standard deviations obtained with 0.5 kHz TB recordings of the VS and
BL ABR systems led to a wide range of normal deviation of between 19 dB nHL
to 53 dB nHL and 19 dB nHL to 55 dB nHL respectively.
The thresholds obtained with the VS system were more widely distributed across
the intensity scale (minimum = 7 dB nHL; maximum = 67 dB nHL) when
116
compared to that of the BL system (minimum = -1 dB nHL; maximum = 49 dB
nHL). Although the wide range of the ABR thresholds could be responsible for
the large standard deviation, the small research sample could also have
contributed to the large SD (Stein, Smit, Du Toit & Strasheim, 1998).
The higher mean electrophysiological thresholds obtained with the VS ABR
system could be attributed to the fact that the VS ABR system was successful for
more subjects than the BL ABR system. Thus, to compare the thresholds of the
two systems, only the thresholds of the subjects in which both systems were
successful needed to be considered. Table 4.10 presents the mean and SD
values for the thresholds that were obtained with both ABR systems.
Table 4.10:
The mean ABR thresholds and SD (dB nHL) of the VS ABR and BL
ABR systems using click stimuli (n=12) and 0.5 kHz TB stimuli
(n=11)
Sample
Mean
SD
(N)
Click stimuli
VS system
12
25
±16
BL system
12
21
±14
0.5 kHz Tone burst
stimuli
VS system
11
32
±16
BL system
11
37
±18
4.3.3 Threshold correspondence of the VS and BL ABR systems to
behavioural PT thresholds
The mean difference between the two ABR systems and the behavioural PT
thresholds were determined by comparing the click-evoked threshold of each
system with the 2 kHz and 4 kHz PT thresholds, and the 0.5 kHz TB ABR
threshold of each system with the 0.5 kHz PT threshold. For the results of this
section it needs to be considered that ABR stimuli used for the purposes of this
117
study were calibrated in dB nHL for a group of normal listeners as discussed in
Chapter 3 (see paragraph 3.6.2.4 and Table 3.5). The deviation between the
click stimuli of both ABR systems did not significantly differ from the 0 dB HL
standard for behavioural pure tone thresholds at 2 kHz and 4 kHz as specified by
ANSI (S3.6-1996). Thus, the click stimuli thresholds of both systems (in dB nHL)
and the behavioural PT thresholds in dB HL are comparable without any
significance.
The deviation between the 0.5 kHz TB stimuli for both ABR
systems was also not significantly different from the 0 dB HL standard for the
behavioural PT thresholds at 0.5 kHz as specified by ANSI (S3.6-1996). The 0.5
kHz TB stimuli thresholds (in dB nHL) of both ABR systems were therefore
comparable to the 0.5 kHz behavioural PT thresholds (in dB HL) without any
significant deviation. As there was no significant difference between the dB nHL
intensity scale for the ABR stimuli of both systems and the dB HL intensity scale
of the behavioural PT, thresholds were compared in dB HL.
Another important factor was the fact that the consistency of the behavioural PT
responses varied from subject to subject. It was mentioned earlier in this chapter
(paragraph 4.2.3 and Table 4.6) that 7/15 subjects responded consistently to the
PT stimuli (Group A), whilst the behavioural PT thresholds of the remaining 8
subjects were unreliable (Group B). Therefore, for purposes of accuracy, the
differences between the ABR systems and the behavioural PT thresholds of
Group A are provided individually prior to a presentation of the similar results
which include the total sample, i.e. subjects of Group A and subjects of Group B.
Furthermore, Subject 13, Subject 14 and Subject 15 were excluded from the
calculation of threshold differences between the click-evoked ABR and 2 kHz and
4 kHz PT as the BL ABR procedure was not feasible for any of these subjects.
The sample of 15 subjects were also reduced to 11 for determining the mean
threshold differences between the 0.5 kHz TB thresholds and the 0.5 kHz PT
thresholds, because the BL system proved not feasible for these subjects as well
as Subject 3.
118
The threshold differences for the two ABR systems are individually presented in
the subsequent text.
4.3.3.1 Threshold correspondence at 2 kHz, 4 kHz and the average of 2 kHz
and 4 kHz
Click-evoked threshold differences of the VS and BL ABR systems are
represented by a normal distribution in Table 4.11.
Differences between the VS and BL ABR systems were apparent for the subjects
in the different groups (subjects in Group A and subjects in Group A and B).
Table 4.11 illustrates that the VS ABR threshold differences were within 15 dB or
less of the behavioural PT thresholds in 6/7 subjects in Group A. For the same
group, however, the results of the BL ABR system were slightly different in that
the threshold differences were within 20 dB or less of the behavioural PT
thresholds in 6/7 subjects.
Thus, it seemed that the VS ABR system
corresponded better with high frequency PT thresholds (2 kHz, 4 kHz, and the
average of 2 kHz and 4 kHz) than the BL ABR system.
Upon inclusion of
behavioural PT thresholds of subjects in Group B (unreliable behavioural PT
thresholds) the BL ABR system illustrated a better correspondence with the
behavioural PT than the VS ABR system. As shown in Table 5.8, the BL ABR
system threshold differences were within 20 dB or less of the behavioural PT
thresholds in 11/12, 10/12 and 10/12 subjects as opposed to 7/12, 8/12 and 8/12
of the subjects tested with the VS ABR system at 2 kHz, 4 kHz and the average
of 2 kHz and 4 kHz respectively.
119
Table 4.11:
Distributions of threshold differences (click-evoked ABR threshold – PT threshold) for the VS and BL ABR
systems at 2 kHz, 4 kHz and the average of 2 kHz and 4 kHz
VS ABR system
Threshold
difference
BL ABR system
Subjects of Group A
Subjects of Group A and B
Subjects of Group A
(Reliable PT thresholds)
(Reliable and unreliable PT
thresholds)
(Reliable PT thresholds)
2kHz
N=7
4 kHz
4/7
N=12
Average
2-4 kHz
2kHz
4 kHz
Average
2-4 kHz
2kHz
N=7
4 kHz
6/7
6/7
6/12
8/12
8/12
4/7
6/7
6/7
6/7
7/12
8/12
8/12
6/7
6/7
6/7
7/12
8/12
8/12
Subjects of Group A and B
(Reliable and unreliable PT
thresholds
N=12
Average
2-4 kHz
2kHz
4 kHz
Average
2-4 kHz
4/7
5/7
8/12
8/12
7/12
4/7
5/7
5/7
8/12
7/12
7/12
6/7
6/7
6/7
11/12
10/12
10/12
Less than or
equal to
10 dB
Less than or
equal to
15 dB
Less than or
equal to
20 dB
120
In order to determine whether there was a statistically significant difference
between the ABR thresholds of each system and the PT thresholds, collective
data had to be considered.
The collective results for the mean threshold
differences of subjects in Group A (n=7) at 2 kHz, 4 kHz and the average of 2
kHz and 4 kHz are presented in Table 4.12, whilst collective results for mean
threshold differences for subjects in Group A and Group B (i.e. the total sample)
(n=12) are presented in Table 4.13.
Table 4.12:
Mean threshold differences at 2 kHz, 4 kHz and the average for 2 kHz
and 4 kHz for subjects in Group A (n=7)
Mean threshold difference (mean ± SD)
(dB)
2 kHz
VS ABR system
BL ABR system
9±10
11±13
0.8
p-value
4 kHz
6±11
8±13
1.0
p-value
Average for 2 kHz and 4 kHz
8±10
p-value
9±13
0.8
The mean threshold differences for the BL ABR system at 2 kHz, 4 kHz and the
average of 2 kHz and 4 kHz were 11 dB, 8 dB and 9 dB respectively. The SD of
the mean threshold differences remained constant at 2 kHz, 4 kHz and the
average of 2 kHz and 4 kHz (±13 dB). The normal range of deviation for the
121
threshold differences of the BL ABR system was therefore -2 dB to 24 dB, -5 dB
to 21 dB and -4 dB to 22 dB respectively.
Table 4.13:
Mean threshold differences at 2 kHz, 4 kHz and the average of 2 kHz
and 4 kHz for subjects in Group A and B sample (n=12)
Mean threshold difference (mean ± SD)
(dB)
2 kHz
VS ABR system
BL ABR system
13±12
9±12
0.23
p-value
4 kHz
12±17
9±1
0.23
p-value
Average for 2 kHz and 4 kHz
13±15
p-value
9±13
0.21
The mean threshold differences, SD as well as the normal range of deviation,
were visibly smaller for the VS ABR system, though statistically the threshold
differences at 2 kHz, 4 kHz and the average of 2 kHz and 4 kHz were not
significant (p=0.8; p>0.05) as determined by the Wilcoxon Signed-Rank test. The
small sample size (n=7) could have contributed to the inconclusive statistical
findings. Upon inclusion of the inconsistent behavioural PT thresholds, i.e. for
the subjects of Group B (n=12), the Wilcoxon Signed-Rank test also indicated no
significant difference of threshold differences between the two ABR systems (p =
0.21; p=0.23; p>0.05). Although there was no statistically significant difference,
the threshold differences were noticeably larger for the VS ABR system than for
the BL ABR system.
122
In addition to the mean threshold differences and the standard deviations, the
range of threshold differences for each ABR system is presented in Table 4.14.
Subjects in Group A and the subjects in Group A and B are dealt with separately.
From the table it is apparent that the inclusion of the behavioural PT data
obtained from subjects in Group B affected the maximum threshold of the ABR
systems, in particular the VS system.
Table 4.14:
Range of threshold differences of ABR systems using click stimuli
for subjects in Group A and subjects in Group A and B
VS ABR system
Difference at 2 kHz (dB nHL)
Difference at 4 kHz dB nHL)
Difference at the average of
2 kHz and 4 kHz (dB nHL)
Key:
BL ABR system
Minimum
Maximum
Minimum
Maximum
-1
-3
-8
4
-6
-6
27
32
27
42
27
37
-11
-11
-16
-1
-14
-9
29
19
24
34
27
27
Minimum and maximum threshold differences of subjects in Group A
Minimum and maximum threshold differences of subjects in Group A and Group B, i.e. the total sample
Threshold differences for individual subjects were determined for all subjects with
comparable thresholds. Figures 4.10, 4.11, and 4.12 illustrate the differences
between the click-evoked threshold of each ABR system and the PT thresholds
at 2 kHz, 4 kHz and the average for 2 kHz and 4 kHz respectively for each
subject. Although the behavioural PT thresholds of all the subjects (n=12) were
plotted on the graphs, different colours and symbols were used to distinguish the
subjects with reliable behavioural PT thresholds (n=7) from the subjects who
responded unreliably to behavioural stimuli (n=5).
123
80
Click-evoked ABR threshold - 2kHz
behavioural PT threshold
70
60
50
40
30
20
10
0
-10
-20
S1
S2
S3
S4
S5
S6
S7
S8
S9
S10
S11
S12
Subjects
Subjects in Group A: Threshold difference betw een click-evoked VS ABR and 2kHz PT threshold
Subjects in Group A: Threshold difference betw een click-evoked BL ABR and 2 kHz PT threshold
Subjects in Group B: Threshold difference betw een click-evoked VS ABR and 2 kHz PT threshold
Subjects in Group B: Threshold difference betw een click-evoked BL ABR and 2 kHz PT threshold
Figure 4.10: Threshold differences between the click-evoked thresholds (VS and
BL ABR systems) and the 2 kHz behavioural PT threshold. Group A subjects who responded consistently during behavioural PT
audiometry; Group B - subjects who responded inconsistently
during behavioural PT audiometry.
Figure 4.10 shows that threshold differences equal to, or less than 10 dB were
present in 6 subjects using the VS ABR, compared to 7 subjects using the BL
ABR system. Threshold differences between 11 dB to 20 dB were present in two
subjects implementing the VS ABR system and four subjects using the BL ABR
system. Furthermore, there were a number of subjects with threshold differences
equal to or greater than 25 dB between both ABR systems and the 2 kHz PT
threshold. In particular, the difference between the click-evoked VS ABR system
and the 2 kHz behavioural PT thresholds for Subject 6, Subject 9 and Subject 12
were 27 dB, 27 dB and 32 dB respectively. It must be noted that these subjects
responded inconsistently to the PT stimuli. The largest difference between the
click-evoked thresholds of the BL ABR system and the 2 kHz PT threshold was
124
found in Subject 1. This difference was 29 dB whereas the difference between
the click-evoked threshold of the VS system and the PT threshold was 12 dB.
Click-evoked ABR threshols - behavioural
4 kHz PT thresholds
80
70
60
50
40
30
20
10
0
-10
-20
S1
S2
S3
S4
S5
S6
S7
S8
S9
S10
S11
S12
Subjects
Subjects in Group A: Threshold difference betw een click-evoked VS ABR and 4kHz PT threhold
Subjects in Group A: Threshold difference betw een click-evoked BL ABR and 4 kHz PT threshold
Subjects in Group B: Threshold difference betw een click-evoked VS ABR and 4 kHz PT threshold
Subjects in Group B: Threshold difference betw een click-evoked BL ABR and 4 kHz PT threshold
Figure 4.11: Threshold differences between the click-evoked thresholds
(VS and BL ABR systems) and the 4 kHz behavioural PT
threshold. Group A - subjects who responded consistently during
behavioural PT audiometry; Group B - subjects who responded
inconsistently during behavioural PT audiometry.
Results for threshold differences between the click-evoked thresholds of each
ABR system and the behavioural PT threshold of the average for 2 kHz and 4
kHz were very similar to the threshold differences at 2 kHz and 4 kHz
respectively. Eight subjects presented with a threshold difference equal to, or
smaller than 10 dB with the VS and the BL ABR systems. Threshold differences
between 11 dB and 25 dB occurred in two subjects using the BL ABR system,
whilst differences equal to or larger than 25 dB occurred in four subjects using
the VS ABR system and two subjects using the BL ABR system. These results
are illustrated in Figure 4.12. The largest threshold differences while using the
VS ABR system occurred in Subjects 6 (32 dB), Subject 9 (27 dB) and Subject
125
12 (37 dB). The largest differences between the thresholds while using the BL
ABR system and 4 kHz PT threshold were found in Subject 1 and Subject 12
Click-evoked ABR threshold - the average of 2 and 4
kHz behavioural PT thresholds
(both 27 dB).
80
70
60
50
40
30
20
10
0
-10
-20
S1
S2
S3
S4
S5
S6
S7
S8
S9
S10
S11
S12
Subjects
Subjects in Group A: Threshold difference between click-evoked VS ABR and 4kHz PT threhold
Subjects in Group A: Threshold difference between click-evoked BL ABR and 4 kHz PT threshold
Subjects in Group B: Threshold difference between click-evoked VS ABR and 4 kHz PT threshold
Subjects in Group B: Threshold difference between click-evoked BL ABR and 4 kHz PT threshold
Figure 4.12: Threshold differences between the click-evoked thresholds
(VS and BL ABR systems) and the average of the 2 kHz and 4
kHz behavioural PT thresholds. Group A - subjects who
responded consistently during behavioural PT audiometry;
Group B -subjects who responded inconsistently during
behavioural PT audiometry)
4.3.3.2 Threshold correspondence at 0.5 kHz
The mean differences between the 0.5 kHz TB threshold using the VS ABR
system and the 0.5 kHz behavioural PT threshold, as well as the 0.5 kHz TB
threshold rendered by the BL ABR system and the 0.5 kHz behavioural PT
thresholds are presented in this section. As in the case of the 2 kHz and 4 kHz
threshold differences, the mean differences of the thresholds of each ABR
system and the behavioural PT thresholds of Group A, are presented in isolation
followed by the inclusion of the behavioural PT thresholds of the subjects in
Group B.
126
Table 4.15:
Normal distributions of threshold differences (0.5 kHz TB ABR
threshold – behavioural PT threshold at 0.5 kHz) for the VS and BL
ABR systems at 0.5 kHz
Threshold
VS ABR system
BL ABR system
difference
Subjects in
Subjects in Group
Subjects in
Subjects in
Group A
A and B
Group A
Group A and B
(Reliable PT
(Reliable and
(Reliable PT
(Reliable and
thresholds)
unreliable PT
thresholds)
unreliable PT
thresholds)
thresholds)
(N=6)
(N=11)
(N=6)
(N=11)
4/6
7/11
5/6
7/11
5/6
9/11
6/6
11/11
Less than or
equal to
30 dB
Less than or
equal to
35 dB
6/6
9/11
Less than or
equal to 55 dB
6/6
11/11
As illustrated in Table 4.15, TB thresholds were within 30dB or less of the 0.5
kHz behavioural PT threshold for 4/6 and 5/6 (using the VS and BL ABR systems
respectively) of the subjects in Group A and 7/11 of subjects in Group A and B.
Furthermore, TB thresholds were within 55 dB or less of the 0.5 kHz PT
thresholds in all subjects of both groups.
The mean difference for both ABR systems at 0.5 kHz is presented in Table 4.16.
The mean difference between the 0.5 kHz tone burst VS ABR thresholds and the
0.5 kHz behavioural PT were similar (26 dB ± 16 dB) for subjects in Group A and
subjects in Group A and B (i.e. the total sample). On average, the BL system
presented with a smaller threshold difference (24 dB ± 6 dB) than the VS ABR
system for subjects in Group A.
However, when the unreliable behavioural
127
responses were included, the mean threshold difference (31 dB) was 5 dB larger
than the mean threshold difference of the VS ABR system (26 dB). As shown in
Table 4.16, the standard deviations were large for both systems, except for the
BL system in Group A which presented with a SD of ± 6 dB. The normal range of
deviation was 10 dB to 42 dB for the VS ABR system. The BL system yielded a
smaller normal range of deviation for the subjects in Group A (18 dB to 30 dB)
than for the subjects of Group A and B (13 dB to 52 dB). The differences for the
subjects of Group A, and the subjects of Group A and B were not significant (p >
0.05) as determined by the Wilcoxon Signed-Rank Test.
Table 4.16: Mean threshold differences at 0.5 KHz for subjects in Group A as well
as subjects in Group A and B
Mean threshold difference (mean ±SD)
(dB)
Subjects in Group A
(n=5)
VS ABR system
BL ABR system
26±16
24±6
0.26
p-value
Subjects in Group A and B
26±16
31±18
(n=11)
p-value
0.25
In addition to the mean threshold differences and the standard deviations, the
range of threshold differences for each ABR system is presented in Table 4.17.
Subjects in Group A and the subjects in Group A and B are dealt with separately.
From the table it is apparent that the inclusion of the behavioural PT data
obtained from subjects in Group B affected the maximum threshold difference of
the ABR systems, in particular the BL system (70 dB).
128
Table 4.17:
Range of threshold differences of ABR systems using 0.5 kHz TB
stimuli for subjects in Group A and subjects in Group A and B
VS ABR system
BL ABR system
Minimum
Maximum
Minimum
Maximum
8
33
20
35
12
53
10
70
Difference at 0.5 kHz (dB)
Key:
Minimum and Maximum threshold differences of subjects in Group A
Minimum and maximum threshold differences of subjects in Group A and Group B i.e. the total sample
In five subjects (Subjects 5, 7, 8, 10 and 11) both ABR systems rendered a
threshold difference of equal to or smaller than 25 dB. The threshold differences
obtained with both ABR systems for Subjects 4, 6, 9 and 12 can be seen as
outliers. The difference between the 0.5 kHz TB threshold of the VS system and
the 0.5 kHz behavioural PT for Subject 4 was 53 dB. The threshold differences
indicated by the BL system for Subjects 6 and Subjects 12 were 60 dB and 70 dB
respectively. These results are illustrated in Figure 4.13.
0.5 kHz tone burst ABR threshold - 0.5
kHz PT threshold
80
70
60
50
40
30
20
10
0
-10
-20
S1
S2
S4
S5
S6
S7
S8
S9
S10
S11
S12
Subjects
Subjects in Group A: Threshold difference betw een 0.5 kHz tone burst VS and 0.5kHz PT threhold
Subjects in Group A: Threshold difference betw een 0.5 kHz BL ABR and 0.5 kHz PT threshold
Subjects in Group B: Threshold difference betw een 0.5 kHz tone burst VS ABR and 0.5 kHz PT threshold
Subjects in Group B: Threshold difference betw een 0.5 kHz tone burst BL ABR and 0.5 kHz PT threshold
Figure 4.13: Threshold differences between the 0.5 kHz TB thresholds (VS
129
and BL systems) and the 0.5 kHz behavioural PT threshold. Group A
- subjects who responded consistently during behavioural
PT audiometry; Group B - subjects who responded
inconsistently during behavioural PT audiometry.
4.3.4 Recording time of the VS and BL ABR systems using click and 0.5
kHz TB stimuli
The subjects for whom ABR recordings were unsuccessful – in only one system
or both ABR systems – were excluded when the mean recording time per ear
was determined. As it was still possible to obtain electrophysiological thresholds
for Subjects 3, Subject 13 and Subject 14 using the VS ABR system, the
recording times were calculated and are presented in Table 4.18.
Table 4.18:
VS ABR recording time per ear for subjects in whom ABR
assessments using the BL ABR device were not successful
Stimuli
Recording time per ear
(minutes)
Click
Subject 3
Subject 13
Subject 14
Threshold obtained;
7
2
6
Threshold was not
therefore included in
mean recording time
0.5 kHz TB
11
obtainable (i.e.
unsuccessful ABR
recording)
The recording time can be defined as the amount of time (in minutes) needed to
determine the electrophysiological threshold per ear. The recording time of the
ABR measurements was classified into three broad time limits (equal to, or less
than 6 minutes, 7 to 10 minutes and equal to, or more than 11 minutes). The
number of recordings that was possible in 1 to 6 minutes, 7 to 10 minutes and 11
130
minutes or more is shown Figures 4.14 (click stimuli) and Figure 4.15 (0.5 kHz
TB stimuli).
12
11
10
9
8
7
Subjects 6
5
4
3
2
1
0
< 6 min
7-10 min
11 min >
Recording time (minutes)
Vivosonic Integrity ABR device
Bio Logic Navigator Pro ABR device
Figure 4.14: Recording time per ear for VS and BL ABR systems
using click stimuli
Figure 4.14 illustrates the recording time of both systems using the click stimulus.
From this figure it is apparent that in 7 subjects (n=12) the recording time for one
ear was 1 to 6 minutes using the BL ABR system. Contrarily, using the VS ABR
device, recordings of 3 subjects (n=12) were possible within the same time
constraints. Within the 7 to 10 minutes time limit, ABR recordings of 6 subjects
and 2 subjects (n=12) were administered with the VS system and BL system
respectively. In 3 subjects an extended recording time (i.e. 11 minutes or more)
was needed for both ABR systems before a threshold was obtained.
For the 0.5 kHz TB recordings it was apparent that both systems required a
longer recording time per ear. As shown in Figure 4.15 there was only 1 subject
in whom an ABR threshold was obtained in 6 minutes or less with the VS as well
as the BL ABR system. Within the 7 to 10 minutes time limit, ABR recordings of
3 subjects and 7 subjects (n=11) were administered with the VS and BL ABR
systems respectively. An extended recording time (i.e. 11 minutes or more) was
required in 7 subjects and 3 subjects for the VS and BL systems respectively
before an electrophysiological 0.5 kHz TB threshold was obtained.
131
11
10
9
8
7
6
Subjects
5
4
3
2
1
0
< 6 min
7-10 min
11 min >
Recording time (minutes)
Vivosonic Integrity ABR device
Figure 4.15:
Bio Logic Navigator Pro ABR device
Recording time per ear for VS and BL ABR systems using 0.5 kHz
TB stimuli
The mean and SD recording time per ear was 9±4 minutes for the VS and 9±5
minutes for the BL ABR systems using click stimuli. The normal recording time
per ear for click-evoked thresholds therefore ranged between 5 and 13 minutes
for the VS ABR system and between 4 and 14 minutes for the BL ABR system.
Using 0.5 kHz TB stimuli, the mean and SD recording time per ear was 9±5
minutes for the BL ABR system. The normal range of deviation for recording 0.5
kHz TB stimuli with the BL ABR system was between 5 and 13 minutes. The
mean and SD recording time per ear was 11±3 minutes for the VS ABR system.
The normal range of deviation for recording 0.5 kHz TB stimuli with the VS ABR
system was between 8 and 14 minutes. The mean recording time per ear for
both ABR systems using click and 0.5 kHz TB stimuli is presented in Figure 4.16.
132
25
20
15
Time (minutes)
11
9
10
5
9
9
0
Vivosonic
Integrity ABR
device
Click stim uli
Bio Logic
Navigator Pro
ABR device
0.5 kHz tone burst stim uli
Figure 4.16: Mean recording time per ear for VS and BL ABR systems using click
and 0.5 kHz TB stimuli
The Wilcoxon Signed-Rank Test was implemented to determine whether there
was a statistical difference between the mean recording times of the ABR
systems. No significant difference was found between the recording times of the
VS and BL ABR system for click stimuli (p = 0.13; thus p > 0.05). Although the
recording times of the VS and BL ABR systems for 0.5 kHz TB stimuli also did
not differ significantly, the BL ABR system neared the 90% confidence level of p
= 0.13.
4.4 Conclusion
Each auditory procedure that was conducted yielded its own set of results. The
results of the first sub-aim were of a descriptive nature and served to provide
information regarding the middle ear functioning, cochlear functioning and
hearing sensitivity of each subject. Results of the second sub-aim were to an
extent descriptive in nature and, although inferential statistics were used,
inconclusive findings were obtained, possibly due to the small research sample.
133
4.5 Summary
Chapter 4 presented the results of each auditory procedure that was conducted.
As explained in Chapter 3, the results of the empirical research were presented
according to the sub-aims. As the first-sub aim focused on the feasibility and
characteristics of the procedures within the auditory test battery, the results of the
immittance audiometry, DPOAE measurements and behavioural PT audiometry
were presented and described.
In order to realize the second sub-aim a
comparison between the VS and BL ABR systems were made in terms of their
feasibility, the correspondence to the 0.5 kHz, 2 kHz, 4 kHz PT thresholds and
the recording time per ear.
134
Chapter 5
DISCUSSION
The aim of this chapter is to introduce the research question, to provide the
rationale for the study, to explain the terminology used, and to present an
overview of the content and the organization of the study.
5.1
Introduction
Chapters 1 and 2 discussed auditory assessments and its challenges in
populations with multiple disabilities, in particular the population with cerebral
palsy (CP).
The importance of the identification of a hearing loss was
emphasized and the difficulty to detect and diagnose the hearing loss, especially
in the CP population, was discussed extensively.
Detection and diagnosis of a hearing loss in the CP population is challenging due
to the complexity of the condition: it is characterized by a variety of additional
disabilities, i.e. physical, cognitive and perceptual disabilities ranging from mild to
severe (Donnelly, Parkes, McDowell & Duffy, 2007; Workinger, 2005; Beckung &
Hagberg, 2002; Cogher, Savage & Smith, 1992, Newton, 1977).
These
additional disabilities may mask the presence of a hearing loss in a child with CP
and may ultimately lead to erroneous information regarding the auditory system.
Additional disabilities may mask the presence of a hearing loss in such a way
that the feasibility of auditory procedures as well as the reliability of the results
obtained is compromised (Workinger, 2005; Cogher et al., 2002; McDonald,
1987; Newton, 1977).
The reliability of the results obtained from subjective
auditory procedures, e.g. behavioural pure tone (PT) audiometry, is especially at
risk since the child with CP may respond inconsistently to PT stimuli.
135
In cases where behavioural responses to PT stimuli are inconsistent, the
objective auditory brainstem response (ABR) procedure can be implemented to
determine the integrity of the auditory system up to the level of the brainstem
level and also for predicting behavioural PT thresholds by using click and
frequency-specific stimuli, i.e. tone bursts (TB), respectively (Hall, 2007; Gorga,
Johnson, Kaminski, Beauchaine, Garner & Neely, 2006; Folsom & Diefendorf,
1999; Galambos, Hicks & Jo Wilson, 1984).
The ABR is highly valued in the clinical context and has been successfully
implemented in difficult-to-test populations such as infants (Hall, 2007; JCIH,
2007; Jiang, Andrew & Wilkinson 2006; Folsom & Diefendorf, 1999). The clinical
value of the ABR in difficult-to-test populations not only includes the identification
and diagnosis of a hearing loss; its significance also extends to the management
and intervention of auditory disorders.
In the current research the auditory functioning of a sample of children with CP
was determined using an audiometric test battery. The results of the test battery
were provided in Chapter 4 and are discussed in this section. The ABR was
conducted on the selected sample of CP children using two different ABR
systems, namely the Vivosonic Integrity (VS) ABR system and the Bio Logic
Navigator Pro (BL) ABR system.
The rationale for employing different ABR systems was to determine if new
features, i.e. Kalman filtering and pre-amplification by the VS ABR system are
clinically practical in auditory assessments of this population with.
In the
following section the results obtained in this study are discussed and compared
with existing relevant literature.
136
5.2
Sub-aim 1 discussion: Characteristics and feasibility
of an audiological test battery in children with cerebral palsy
In this section the results obtained from the immittance measurements, OAE
measurements as well as behavioural pure tone (PT) audiometry are
discussed. Table 5.1 summarizes different procedures conducted by various
authors when assessing the auditory functioning in children with multiple
disabilities including CP.
5.2.1 Immittance measurements
Immittance measurements, consisting of tympanometry and acoustic reflexes,
reveal essential information about the auditory system (Block & Wiley, 1994). In
the current study these measurements were performed as part of the test battery
and the results are discussed in terms of the feasibility of the procedures as well
as characteristics that were noted during its administration.
As shown in Table 5.1, the inclusion of tympanometry was not mentioned in the
majority of the research reports (Romero, Mendez, Tello & Torner, 2008;
Topolska, Hassmann-Poznańska & Sołowiej, 2002; Sano, Kaga, Kitazumi &
Kodama, 2005; Kolker, 2004; Zafeiriou, Andreou &, Karasavidou, 2000;
Sheykholeslami & Kaga, 1999; Benham-Dunster & Dunster, 1985; Stein,
Ozdamar & Schnabel, 1981).
It therefore remains unknown whether this
procedure was included in the data selection or data collection procedures.
137
Table 5.1:
Summary of previous assessment procedures employed in difficult-to-test populations
Author
Research sample
Sample
size
Mean age
of sample
Subjective auditory
procedures
Behavioural
audiometry (PTA,
VRA, BOA)
Objective auditory procedures
Tympanometry
Acoustic
reflexes
OAE
AEP
9 (Clickevoked ABR)
Stein, Ozdamar
& Schnabel
(1981)
Developmentally
delayed including
CP
82
11-14 years
BenhamDunster &
Dunster, (1985)
Developmentally
delayed including
CP
164
29-38 years
9
Not known
Sheykholeslami
& Kaga (1999)
Children who
suffered neonatal
hyperbilirubinemia,
2 of them
developed
athetotic CP
3
15 years
9
Not known
Palmu,
Puhakka, Rahko
& Takala (1999)
Infants
58
2 – 11
months
9
Zafeiriou,
Andreou &,
Karasavidou
(2000)
Children with
spastic form of CP
75
6 years
Not known
9(Clickevoked ABR)
Kolker (2004)
Children with
spastic form of CP
126
1-14 years
Not known
9(Clickevoked
Not known
9
9(Clickevoked ABR)
9 (DPOAE and
TEOAE)
9(Clickevoked ABR)
138
ABR and
CEP)
9
Sano, Kaga,
Kitazumi &
Kodama (2005)
Various types of
CP
6
18.5 years
Driscoll, Kei,
Bates &
McPherson
(2002)
Children with
various
impairments e.g.
intellectual and
multiple
impairments
489
9.6 years
Topolska,
HassmannPoznańska,
Sołowiej
(2002)**
Various types of
CP
32
Not known
Romero,
Mendez, Tello &
Torner (2008)
Children with
perinatal
encephalopathy
including CP
135
Less than 1
year of age
Current study
Various types of
CP
15
15.6 years
Not known
9 (DPOAE)
9
9 (TEOAE)
9
9
9(DPOAE)
9
9
9(Clickevoked
ABR)
Not known
9
9(Clickevoked ABR)
9
9 (DPOAE)
9(Clickevoked
ABR and 0.5
kHz TB ABR)
** The English abstract of this report was used as the original report was only available in Polish
139
The possible exclusion of tympanometry in previous research is surprising,
especially since literature indicates that difficult-to-test populations, including
children with CP, display a high risk of middle ear diseases such as otitis media
(Driscoll et al., 2002).
It is well appreciated that middle ear diseases may influence the results of nearly
all
audiometric
procedures
including
behavioural
PT
audiometry,
OAE
measurements as well as ABR (Hall, 2007; Palmu, Puhakka, Rahko & Takala,
1999). In an attempt to avoid false audiometric results due to a compromised
middle ear functioning, all the subjects in the current study had to comply with the
selection criterion of normal middle ear functioning.
Tympanometry was feasible in all the CP subjects of the current research project.
Although tympanometry is generally expected to be viable for all patient
populations including babies and children with multiple disabilities, a low success
rate (74%) was reported for a group of children with developmental and cognitive
disabilities (Driscoll et al., 2002; Palmu et al., 1999). It seems that the feasibility
of this procedure, although it is an objective procedure, is still dependent on a
certain level of cooperation by the patient (e.g. being quiet and accepting a probe
in the ear). When the child does not cooperate to this level for various reasons
including immaturity (too young) or cognitive disabilities, the feasibility of this
procedure may be compromised.
It is satisfying to report that the 100% feasibility rate obtained in the current
research can be attributed to the compliant behaviour of the subjects. The formal
education that subjects received for at least six years prior to testing as well as
the ages of the subjects (12 years to 18 years) suggested a familiarity with
instructions, enhancing compliant behaviour and the viability of the procedure.
Tympanometry measurements were followed by ipsilateral acoustic reflexes.
Published research report highlight the clinical value of the acoustic reflex in
140
special populations (Hall & Mueller, 1997; Northern & Gabbard, 1994; BenhamDunster & Dunster, 1985). In particular, the sensitivity prediction acoustic reflex
(SPAR) may be included as a valuable procedure for predicting hearing
thresholds
in
difficult-to-test
populations,
although
there
is
a
risk
of
overestimating the hearing loss when relying on this procedure (Benham-Dunster
& Dunster, 1985).
Although the current study did not use the SPAR to predict hearing thresholds,
ipsilateral acoustic reflex testing was done at 0.5 kHz, 1 kHz and 2 kHz. The
research sample of the current study was small (n=15), making it difficult to
obtain a specific pattern in the resulting acoustic reflexes.
It was apparent,
however, that the majority of normal acoustic reflexes were elicited from the
subjects with spastic CP whilst the majority of absent acoustic reflexes were
obtained from subjects with athetiod CP. In the athetoid group there was also
one subject (Subject 15) for whom this procedure was not feasible because of
consistent muscular movements.
5.2.2 Distortion product otoacoustic emissions
The advantages of OAE measurements, such as its objectivity and brevity,
contribute to the wide spread implementation of this procedure especially in the
assessment of auditory function in difficult-to-test populations (Driscoll et al.,
2004; Longsbury-Martin, McCoy, Whitehead & Martin, 1992). In previous studies
an additional advantage of the OAE procedure, namely site of lesion specificity,
adds particular value to differential diagnosis of a hearing loss in the CP
population (Sano et al., 2005; Sheykholeslami, & Kaga, 1999).
However, in terms of the CP population limited research regarding OAEs is
available. The available reports are characterized by small research samples,
reducing generalization of the findings within this population. Table 5.2 provides
a summary of the available research reports involving the CP population.
141
Table 5.2:
Summary of OAE research involving the CP population
Author
Sample
size
(n)
Sample type
Type of OAE
Sano et al.
6
Mixed type of
DPOAE
(2005)
CP;
Results
Absent OAEs bilaterally in
83% of the subjects; normal in
Ataxia;
17%
Athetosis
Topolska et
Unknown
Unknown
DPOAE
al. (2002) **
37.5% of children with
extrapyramidal CP (i.e.
athetiod CP) presented with
hearing loss
Sheykholesl
ami & Kaga
(1999)
3
Athetosis
DPOAE and
Absent or abnormal OAEs
TEOAE
were measured bilaterally in
all the subjects
Current
research
15
Athetosis
Spastic
Microcephaly
DPOAE
Absent, abnormal and normal
OAEs were measured in 16%,
5% and 79% of the subjects,
respectively
** The English abstract of this report was used as the original report was only available in Polish
In contradiction to previous research reports, the majority of the DPOAEs (79%)
obtained in the current research were within the normal range as specified by the
Vanderbilt criteria. These normal OAE measurements were all obtained from
subjects with the spastic form of CP. Interestingly, there was a percentage of
normal OAEs (11% and 8% in the left and right ears respectively) that exceeded
the amplitude range of the 95th percentile of the Vanderbilt criteria, predominantly
in the low and mid frequencies (635 Hz to1586 Hz). Compared to the Vanderbilt
95th percentile, the extent of the amplitudes obtained in 6 of the spastic CP
subjects (n=10) ranged between 0.1 and 10.95 dB SPL.
142
Reasons for the increased amplitudes of the DPOAEs in the six subjects with
spastic CP remain unclear.
Interestingly, research indicated that increased
DPOAE amplitudes in the lower frequencies are often displayed in subjects with
sickle cell disease (SCD) (Stuart, Jones & Walker, 2006; Downs, Stuart &
Holbert, 2000). Additionally, statistically significant associations were revealed
between this condition and developmental disabilities such as CP (Ashley-Koch,
Murphy, Khoury & Boyle, 2001; Downs et. al., 2000). Although it could not be
inferred that the subjects in the current study also presented with SCD, it remains
interesting and creates an opportunity for further research.
Previous research conducted within the CP population did not note an increase
in the amplitudes of DPOAEs, though reported absent or abnormal OAEs. As
shown in Table 5.2, the majority of OAEs recorded in CP subjects in previous
studies was either abnormal or absent (Sano et al., 2005; Sheykholeslami &
Kaga, 1999). The OAE results obtained in the current study contradicts previous
research: in all subjects where OAE-testing was possible (i.e. emissions were
reliable and noise levels were low) (n=14), only 5% of the OAEs were absent or
abnormally reduced.
Absent or abnormally reduced OAEs are usually an indication of malfunctioning
of the outer hair cells (OHC) of the cochlea supporting the results of previous
studies in terms of the prevalence of a sensorineural hearing loss in children with
CP (McDonald, 1987; Newton, 1977; Durrant, 1992; Longsbury-Martin et al.,
1992). Despite the findings of these studies, none of the previous reports noted
the effects of excessive muscular movements, typically displayed by the CP
population, on the OAE measurements.
Excessive muscular movements create high levels of internal noise, thereby
creating a high noise floor level during OAE testing. These high noise floor levels
affect the signal-to-noise ratio (SNR). Accurate and reliable detection of the OAE
relies on an adequate SNR where the level of the noise should not exceed that of
143
the signal (OAE) (Baer & Hall, 1992). Thus, a high noise floor level will influence
the detection of reliable emissions leading to inaccurate results and ultimately to
an erroneous conclusion regarding cochlear functioning.
Two of the reports
listed in Table 5.2 (Sano et al., 2005; Sheykholeslami & Kaga, 1999) indicated a
large percentage of absent OAE measurements, but no mention is made of the
possible adverse effect of high levels of internal noise typical of patients with CP.
Thus, the question can be raised whether the OAEs were absent or abnormal
purely due to OHC failure or were the feasibility of the OAE measurements
compromised by excessive muscular movements.
As theory states the importance of a low noise level, in the current study the
noise levels for each OAE measurement (i.e. for each subject) were taken into
account. In Chapter 4 it was reported that the noise floor level was below 3 dB
SPL in 14/15 subjects. To be regarded as a DPOAE, the noise floor level should
not exceed 3 dB SPL (Hall & Mueller, 1997). There was, however, one subject
(Subject 15) in whom the noise floor levels exceeded 3 dB SPL.
Subject 15 was diagnosed with athetoid CP and presented with excessive
muscular movements throughout the administration of the OAE measurements.
The consistency and excessiveness of the muscular movements produced
consistently high noise floor levels which made the reliable detection of the OAE
impossible.
Interestingly, Subject 5, Subject 8 and Subject 14 were also diagnosed with the
athetoid type of CP and did present with some muscular movements during OAE
measurements, but the noise floor levels were consistently below 3 dB SPL.
Whilst Subject 5 and Subject 14 presented with absent OAE measurements,
OAEs obtained from Subject 8 were within the normal range as specified by the
Vanderbilt criteria. The variety of the OAE results in just one sub-group of the
sample (the athetoid group) emphasized the variability within the CP population
point to the problem of generalizing OAE results in this population.
144
5.2.3 Behavioural pure tone audiometry
Behavioural PT audiometry remains fundamental within the diagnostic
audiometric process. However, the administration and ultimately the feasibility of
this procedure may become a challenge when confronted with various factors
(e.g. age of the individual, the level of formal education, the severity of cognitive,
physical or perceptual disabilities) that may reduce the consistency of responding
to pure tone stimuli (Workinger, 2005; Folsom & Diefendorf, 1999; Hodgson,
1994; Benham-Dunster & Dunster, 1985).
Considering the various factors that may influence the administration of
behavioural PT audiometry, it is not surprising that in the current study the
consistency of responses to PT stimuli varied between CP subjects. Although
behavioural PT audiometry was feasible in all the subjects, there was a
percentage of subjects with athetoid CP (3/5; 60%) and spastic CP (5/10; 50%)
that responded inconsistently to the PT stimuli. This was illustrated in Table 4.6.
The researcher distinguished between subjects who responded reliably to PT
stimuli (Group A) and subjects who responded inconsistently to PT stimuli (Group
B). Although this distinction was made, the mean behavioural PT thresholds of
the subjects in Group A and Group B were very similar as illustrated in Chapter
4.
In the majority of the subjects in Group A and Group B behavioural PT
thresholds within the normal range (0d B HL to 20 dB HL) were obtained. Only
two subjects (Subject 5 in Group A and Subject 15 in Group B) presented with
elevated behavioural PT thresholds (between 40 dB HL to 90 dB HL). Both
these
subjects
showed
sensorineural loss.
a
ski-slope
configuration
moderate
to
severe
The presence of the sensorineural hearing loss in these
subjects, whom were both diagnosed with athetoid CP, correlates with literature
that proposes the presence of a sensorineural hearing loss specifically in the
athetoid group (Sano et al., 2005; Northern & Downs, 1991; McDonald, 1987,
Newton, 1977).
145
5.3.
Sub-aim 2 discussion: Comparing the Vivosonic Integrity ABR
system with a conventional ABR system
The Vivosonic Integrity ABR system (VS) was compared to an ABR system with
conventional technology, in this instance the Bio Logic Navigator Pro system.
The VS system was compared to the BL ABR system in terms of the feasibility,
the threshold correspondence to behavioural PT thresholds as well as the
recording time. The results of each of these comparisons were presented
separately in Chapter 4 and are also individually discussed in the following
section.
5.3.1 Feasibility of the VS and BL ABR systems using click and 0.5 kHz TB
stimuli
The feasibility of the ABR procedure is particularly significant for the identification
of a hearing loss or auditory neuropathy in those populations in which
behavioural PT thresholds are unreliable or unobtainable (Hall, 2007).
Therefore, a high clinical value in terms of the success rate of the click-evoked
ABR is essential for assessments of the difficult-to-test populations such as CP.
Illustrations of feasible ABR recordings conducted with the VS and BL ABR
systems are presented in Figure 5.1 and Figure 5.2 respectively.
Figure 5.1:
An example of a feasible ABR recording using click stimuli
with the VS ABR system
146
Figure 5.2: An example of a feasible ABR recording using click stimuli with the
BL ABR system
For the CP population limited research is available that reveals the success rate
of ABR assessments. Stein et al. (1981) reported that ABR recordings were
successful in 96% of their research sample, whilst Benham-Dunster & Dunster
(1985) reported a success rate of 73% to 74% (profoundly delayed subjects) and
89% to 91% (moderately delayed subjects). It is interesting to note that, as
shown in Table 5.3, in both these investigations sedation was used during the
ABR recording.
The higher success rates of the ABR recording in the previous studies compared
to the success rates of the current study may be partially explained by the use of
sedation. Sedatives limit muscular movements, thus enhancing the restfulness
of the child which is a requirement for an acceptable SNR in that the amplitude of
the noise (i.e. myogenic potentials caused by muscular movements) does not
exceed the amplitude of the ABR (Hall, 2007). Subjects in the current study were
awake and sitting on a chair watching a silent movie.
147
Table 5.3:
A review of the percentage successful ABR recordings conducted in
CP populations
Author
Stimuli used in ABR
% Successful ABR
recording
recordings
Click
96%
Click
Moderately delayed subjects:
89-91%
Profoundly delayed subjects:
73-74%
Click
VS ABR system: 93%
BL ABR system: 80%
Stein et al.
Research
(1981)
conducted
sedation
Benham-Dunster
& Dunster (1985)
Research
Current
conducted
research
without sedation
0.5 kHz TB
VS ABR system: 80%
BL ABR system: 73%
Although the SNR can effectively be improved by the utilization of sedation or
general anaesthesia, CP children display a high risk for developing upper
respiratory obstruction e.g. sleep apnea when the sleeping pattern is
manipulated (Schmidt, Krief, Deuster, Matulat & Zehnoff-Dinnesen, 2007; Surya,
Harkera, Begentb, & Chongc, 2005; Johnson, Page, Williams, Wassemer &
Whitehouse, 2002; Sanchez, Zaldivar, Padilla & Morales, 2002; Elwood, Hansen
& Seeley, 2001). As the manipulation of the sleeping pattern remains dangerous
in the CP population, alternative techniques such as natural sleep or the sleeping
agent melatonin may be administered to improve the SNR during ABR recording
(Schmidt et al., 2007; Surya et al., 2005).
Whilst research showed that melatonin can be used effectively in children with
multiple disabilities (Schmidt et al., 2007), at least one previous study indicated
that the ABR can be recorded in CP infants during natural sleep (Romero et al.,
2008). Although natural sleep has been used in CP subjects, this technique
could be extremely time-consuming within the clinical setting as the audiologist
148
because the assessment can only commence once the child falls asleep.
Another disadvantage of natural sleep is that the presence of involuntary and
uncontrollable muscular movements may still occur during natural sleep, implying
that this method may be less effective, especially in the population with CP
(Surya et al., 2005).
Seeing that the click-evoked and 0.5 kHz TB ABR recordings were feasible in
more than 70% in the current study employing the VS and BL ABR systems
without the use of sedatives, melatonin or natural sleep, it seems possible to
obtain an ABR with CP children while they are awake. Two factors that could
have contributed to the high success rates of ABR assessments in the current
study include the age of the subjects (12 years to 18 years) and the fact that the
subjects have been in a formal educational setting for at least six years and have
received additional rehabilitative services. The subjects in the research sample
were therefore familiar with the instructions given to them prior to ABR
assessments and displayed adequate cooperation.
An additional factor –
providing entertainment by means of a silent movie – might also have contributed
to the success rates of the ABR recordings. By watching the silent movie the
subject was kept occupied during the ABR recordings, which possibly served to
enhance restfulness of the subject and reduce involuntary movements.
Although the restfulness of the child contributes largely to an acceptable SNR,
the appropriate parameter settings, i.e. acquisition parameters of the ABR also
add to the improvement of the SNR (Hall, 2007). As discussed in previously,
SNR in the CP population may be compromised by the presence of involuntary
muscular movements. In an attempt to improve the SNR, acquisition parameters
e.g. analysis time, the number of sweeps and filter settings can be modified to
obtain improved ABR recordings in the CP population.
Table 5.4 illustrates
potential modifications of the acquisition parameters in this population.
149
As depicted in Table 5.4, literature suggests that, in an attempt to improve the
SNR, more sweeps are used (i.e. 2000 sweeps instead of 1000 sweeps) and that
the recordings be repeated (Hall, 2007). However, the SNR may not effectively
be improved by increasing the number of sweeps since unwanted neuromuscular
energy still interferes with the ABR signal because it shares a portion of the
frequency spectrum of the ABR (Hall, 2007).
Table 5.4:
Potential modifications of the acquisition parameters during ABR
assessments in the CP population
Acquisition parameters
Possible modifications of
Rationale for possible
usually employed in the ABR
acquisition parameters
modifications of the
protocol (click and tone burst
specifically for the CP
acquisition parameters
stimuli)
population
Filter settings
Low pass filter: 30Hz-100Hz
Low pass filter: 100Hz
(click and tone burst stimuli)
Neuromuscular activity shares
a
portion
of
the
ABR
(or 200Hz for click stimuli)
frequency.
High pass filter: 2000Hz-
obtain a purer ABR signal a
3000Hz
higher low pass filter setting,
e.g.
In an attempt to
100Hz,
is
therefore
suggested.
Analysis time
15ms to 20ms
10ms to 15ms
Because the child with CP
presents with an immature
central
nervous
system
(CNS), a longer analysis time
may be needed to incorporate
all the ABR components.
Repetitions/sweeps
2 runs of 2000 (at least)
Since the child with CP may
OR whatever is needed for
display excessive muscular
a sufficient SNR
movements that will interfere
with the ABR recording, more
sweeps may be needed in an
attempt to obtain an adequate
SNR.
Adapted from: Hall (2007)
150
In an attempt to eliminate some of the redundant myogenic potentials that share
a portion of the frequency spectrum of the ABR, the filter settings of ABR
systems, in particular the low pass filter can be adjusted. Table 5.4 suggests a
low pass filter setting of 100Hz to reduce segments of unwanted myogenic
potentials.
From Table 5.5, which summarizes ABR research previously
conducted with the population with multiple disabilities including CP, it is clear
that the majority of the studies, including the current study, used a protocol with a
higher low pass filter setting of 100Hz (Romero et al., 2008;
Sheykholeslami &
Kaga, 1999; Stein et al., 1981).
Table 5.5 also shows that some studies used an even higher low pass filter of
150Hz and 200Hz (Zafeiriou, Andreou & Karasavidou, 2000; Benham-Dunster &
Dunster, 1985).
Although low cut filter settings of 150Hz to 300Hz are still
acceptable for click-evoked ABR recordings since this type of stimulus display a
broad spectral frequency, Hall (2007) cautions against excessive filtering.
Excessive filtering may not only result in the elimination of important portions of
the ABR, but may also contribute to the formation of a distortion product in the
waveform which can be falsely identified as a response component (Hall, 2007).
In the current study excessive filtering of low frequency energy might have
contributed to the lower success rates obtained with both systems utilizing 0.5
kHz TB stimuli. Literature clearly states that TB stimuli are dominated mainly by
low frequency energy (Hall, 2007). This implies that the use of a low pass filter
such as 30 Hz is more appropriate since it encompasses the low frequency
energy of the TB stimulus. On the other hand, using a higher low pass filter
setting, e.g. 100Hz, the risk of eliminating essential components of the TB ABR is
increased. Hence, it could be argued that the using the 100Hz low pass filter for
the 0.5 kHz TB recordings could have been a drawback in the current study.
However, as stated in the previous chapters, the rationale for using this particular
low pass filter setting was an attempt to obtain 0.5 kHz TB recordings even in the
151
presence of neuromuscular movements typically encountered in the CP
population without sedation.
The audiologist consequently faces a challenge when conducting frequencyspecific ABR recordings in the CP population in that an increase in the low pass
filter setting may filter out some of the large myogenic potentials, but at the same
time may be responsible for unreliable recordings of the ABR (Hall, 2007). This
dilemma may explain the lack of research specifically using tone burst stimuli in
populations with multiple disabilities. However, the importance of incorporating
frequency-specific ABR recordings remains critical for the diagnostic audiology
process since intervention services, i.e. the fitting of hearing aids for children with
CP, highly depend on it.
Although the majority of the VS and BL ABR assessments in the present study
were successful (as discussed in the previous text), there was a number of
subjects in whom click-evoked and 0.5 kHz TB recordings were unsuccessful as
illustrated in Table 4.8.
Unsuccessful ABR recordings suggest that the ABR components (i.e. wave I,
wave III, wave V) are not identifiable nor repeatable (Hall, 2007). Unsuccessful
ABR recordings may be attributed to various factors including the presence of a
profound sensory hearing loss or conditions of a compromised central nervous
system (CNS), such as auditory neuropathy (Hall, 2007; Topolska et al., 2002;
Rance et al., 1999). Furthermore, additional symptoms of a compromised CNS,
e.g. excessive involuntary and uncontrollable muscular movement, may also
affect the feasibility of the recording of the ABR because excessive muscular
movements may be responsible for large myogenic potentials which directly
influence the ABR recording negatively (Hall, 2007). The recording is negatively
affected when the amplitude of the myogenic potentials exceeds the amplitude of
the underlying ABR signal which may result in an undesirable SNR.
152
Table 5.5:
A summary of different ABR systems and parameters implemented in previous studies in populations with multiple
disabilities
Author
Stein,
Ozdamar &
Schnabel
(1981)
ABR system
1st Phase:
Grason-Stadler
471-1
Sedation/
general
anaesthesia
Yes
Acquisition parameters
Filter settings
1st Phase: 1003000Hz
2nd Phase: 1003000Hz
nd
2 Phase:
Nicolet model
1074
Repetitions
(Sweeps)
Analysis time
Type
Polarity
1st Phase:50ms
2048
3rd Phase: 1002000Hz
3rd Phase:
Grason Stadler
1216A
Stimulus parameters
2nd Phase:
20ms
Rate
1st Phase: 20.sec
Click
Rarefaction
2nd Phase: 20/sec
3rd Phase: Varied
between 1033/sec
rd
3 Phase:
20ms
BenhamDunster &
Dunster,
(1985)
Nicolet CA
1000
Yes
150-3000Hz
2000
10ms
Click
Rarefaction
19.1/sec
Sheykholesl
ami & Kaga
(1999)
Neuropack
System
Not known
100-3000Hz
Not known
Not known
Click
Rarefaction
Not known
Zafeiriou,
Andreou &
Karasavidou
(2000)
Nihon-Kohden
Neurpack 4
system
Not known
200-2000Hz
1024
Not known
Click
Rarefaction
10/sec
153
Kolker
(2004)
Not known
Not known
50-100Hz
1000
10ms
Click
Rarefaction
20/sec
Sano, Kaga,
Kitazumi &
Kodama
(2005)
Nihon-Kohden
Neuropack
System
Not known
Not known
Not known
Not known
Click
Not known
Not known
Romero,
Mendez,
Tello &
Torner
(2008)
Amplaid MK 15
No - subjects
tested during
natural sleep
100-2500Hz
2000
15ms
Click
Rarefaction
and
condensation
11/sec
Current
study
Two systems:
1. Vivosonic
Integrity
2. Bio Logic
Navigator Pro
Both
systems:
Both systems:
100-3000Hz
VS system: 20004000 sweeps
recording (click
and 0.5 kHz TB)
BL system: 2
recordings of 2000
sweeps ( click and
0.5 kHz TB)
Both systems:
Click ABR :
15ms
0.5 kHz TB
ABR: 21.2ms
Both
systems
: Click
and 0.5
kHz TB
Both
systems:
Click:
Rarefaction
0.5 kHz TB:
Alternating
Both systems:
Click and 0.5 kHz
TB:
37.7/sec
No - subjects
awake
154
A poor SNR may ultimately influence the feasibility of the ABR (Hall, 2007).
Figure 5.3 and Figure 5.4 provide examples of unsuccessful ABR recordings
in the current research, implementing the VS and the BL ABR systems.
(a)
(b)
Figure 5.3: An example of an unsuccessful ABR recording using click stimuli
(a) and 0.5 kHz TB stimuli (b) with the BL ABR system
155
(a)
(b)
Figure 5.4: An example of an unsuccessful ABR recording using click stimuli
(a) and 0.5 kHz TB stimuli (b) with the VS ABR system
In this investigation the VS ABR system illustrated the highest rate of
successful ABR recordings using click and 0.5 kHz TB stimuli.
This is
noteworthy when considering that both ABR systems were exposed to similar
test conditions and were employed simultaneously. Thus, it could be argued
that the novel features of the VS ABR system, i.e. pre-amplification and
Kalman filtering, contributed to ABR components (e.g. wave V) being more
readily identifiable in the presence of a poor SNR.
156
As previously discussed (Chapter 2), the SNR should be improved because
pre-amplification holds the promise of reducing the permeation of unwanted
myogenic potentials prior to the amplification process, whereas Kalman
filtering attempts to reduce the effects of sporadic noise during the ABR
recording (Hall, 2007; Steinmann & Kurtz, 2005).
Although the BL ABR system was less successful in more subjects, the
implementation of the artifact rejection technique in the present study
rendered valuable information. As seen in Chapter 4 (Figure 4.10) there were
a number of subjects in whom a high artifact rejection percentage occurred.
However, only one subject (Subject 15) and three subjects (Subject 3, Subject
13 and Subject 15) displayed unsuccessful recordings when using click and
0.5 kHz TB stimuli respectively. Click-evoked and 0.5 kHz TB thresholds
were therefore obtained from the remaining subjects who also presented with
high artifact rejection percentages.
Although excessive muscular movements might have contributed to the some
of the recordings being unsuccessful, this is not true for Subject 14.
Recordings with both ABR systems indicated absent wave components in the
presence of minimal artifacts as illustrated by the percentage of artifact
rejection obtained with the BL ABR system (2.68% during click-evoked
recordings and 5.93% during 0.5 kHz TB recordings).
Subject 14 also
presented with absent ipsilateral acoustic reflexes and DPOAEs bilaterally, yet
the behavioural PT thresholds were within the normal range. Although it could
be argued that this subject might have displayed clinical symptoms of auditory
neuropathy spectrum disorder (ANSD) based upon audiological findings, the
fact that the subject acquired spoken language and is an effective
communicator reduces the likelihood of the presence of this condition (Rance,
Beer, Cone-Wesson, Shepherd, Dowell, King, Rickards, Clark, (1999).
5.3.2 Electrophysiological thresholds of the VS and BL ABR systems
using click and 0.5 kHz TB stimuli
As previously discussed, behavioural PT thresholds are often not obtainable
or might be unreliable in difficult-to-test populations such as CP.
Hence,
157
obtaining reliable electrophysiological thresholds remains essential in the
diagnostic audiometric process.
Surprisingly, literature offers limited available data that includes the actual
ABR thresholds in populations with multiple disabilities such as CP (Sano et
al., 2005; Sheykolsami & Kaga, 2000 Benham-Dunster & Dunster, 1985; Stein
et al., 1981). Table 5.6 provides a summary of studies in which ABR results
were provided. However, the ABR results were presented differently in the
various reports, making comparison of the results difficult.
Table 5.6:
Summary of ABR results in previous studies
Author
How ABR results were presented
Results
Sano et al.
ABR thresholds presented in dB
No response at 95 dB nHL (maximum of
(2005)
nHL
the ABR)
Sheykholeslami
ABR thresholds presented in dB
No response at 95 dB nHL (maximum of
& Kaga (2000)
nHL
the ABR)
Benham-
ABR thresholds presented in dB
Moderately delayed subjects presented
Dunster &
nHL
with a mean ABR threshold of 33.9 dB
Dunster (1985)
nHL
Profoundly delayed subjects presented
with a mean ABR threshold of 38.9 dB
nHL
Stein et al.
ABR thresholds presented
43% of the subjects illustrated with
(1981)
according to diagnostic categories
hearing sensitivity between 0-30dB HL
(0-30 dB HL and 40-70dB HL)
19.5% of the subjects presented with
hearing sensitivity between 40-70dB HL
Current
ABR thresholds presented
Mean click-evoked ABR threshold of VS
research
according to intensity level (dB
system:
nHL)
25 dB nHL (Calculated for 12 Subjects)
Mean click-evoked ABR threshold of BL
158
system:
21 dB nHL (Calculated for 12 Subjects)
Mean 0.5kHz ABR threshold of VS
system:
32 dB nHL (Calculated for 11 Subjects)
Mean 0.5kHz ABR threshold of BL
system:
37dB nHL (Calculated for 11 Subjects)
As depicted in Table 5.6, the ABR thresholds presented in available research
reports are diverse
The diversity is, however, not surprising as the
heterogeneity of the CP population suggests that the presence of a hearing
loss may vary greatly between individuals (Workinger, 2005; Cogher et al.,
2002). The results of the current study also showed variability in the ABR
thresholds as measured by the VS and BL ABR systems.
When investigating the ABR results of Subject 5, and comparing the
electrophysiological thresholds of both systems to the behavioural PT
thresholds presented in Figure 5.5, it becomes evident that the click-evoked
thresholds of both ABR systems actually underestimated the high frequency
hearing loss. This disagreement between the click-evoked ABR thresholds
and the behavioural PT thresholds may, however, be expected in the
presence of a steeply sloping high frequency hearing loss as in the case of
Subject 5 (Hall, 2007; Gorga, Johnson, Kaminski, Beauchaine, Garner, Neely,
2006).
159
Frequency (kHz)
0.5 kHz
1kHz
2 kHz
4 kHz
0
Intensity (dB HL)
10
20
30
40
50
60
70
80
Left behavioural PT thresholds
Right behavioural PT thresholds
VS device (0.5 kHz tone burst and click thresholds)
BL device (0.5 kHz tone burst and click thresholds)
Figure 5.5:
Illustration of the behavioural PT thresholds and ABR thresholds
obtained (VS and BL ABR systems) obtained in Subject 5
Obtaining TB thresholds is imperative during ABR recordings of the difficult-totest population. However, the identification of the TB thresholds might be
challenging in difficult-to-test populations such as CP because the excessive
muscular movements often displayed in this population could cause an
undesirable SNR (Hall, 2007). The implementation of a higher low pass filter
setting (e.g. 100 Hz) in an attempt to account for the undesirable SNR may
hinder the identification of the ABR wave V of the TB recording even more.
In the current research project it was also apparent that it was more difficult to
identify the ABR wave V when using 0.5 kHz TB stimuli than when using click
stimuli. As depicted in Table 5.6, 0.5 kHz TB ABR thresholds were obtained
at higher intensities than those of the click-evoked recordings. In fact, there is
a 7 dB and 16 dB average difference between the 0.5 kHz TB thresholds and
the click-evoked thresholds of the VS and BL ABR systems respectively.
By investigating the 0.5 kHz TB thresholds of each ABR system, it was clear
that, contrary to the results of the click-evoked ABR recordings, the VS ABR
160
system displayed a higher percentage of electrophysiological thresholds that
fell within the normal hearing level range than the BL ABR system.
5.3.3 Threshold correspondence of the VS and BL ABR systems to
behavioural pure tones
The main aim of the ABR measurements in difficult-to-test populations is to
provide estimated behavioural PT thresholds (Hall, 2007; Gorga et al., 2006;
Gorga et al., 1993; Picton, 1991).
Thus, the difference between ABR
thresholds and the behavioural PT thresholds provides an indication of the
proximity of the electrophysiological thresholds to the gold standard of
behavioural audiometry thresholds.
In the current study, the difference between the electrophysiological
thresholds of both ABR systems and the behavioural PT thresholds (0.5 kHz,
2 kHz, 4 kHz, the average of 2k Hz and 4 kHz) provided an indication of how
accurate each ABR system estimated hearing sensitivity. These results were
reported in the previous chapter and are discussed below.
The reliability of the behavioural PT thresholds was important in the
determination of the threshold differences. The reliability of these thresholds
for some subjects were questionable, therefore the subjects who responded
reliably during behavioural PT audiometry (Group A) were distinguished from
the subjects who responded inconsistently to the stimuli (Group B).
Furthermore, threshold differences of subjects in Group A, and threshold
differences of the total sample (subjects in Group A and B) were compared to
the pure tone thresholds separately. As reported in Chapter 4, Group A and
Group B consisted of 7/15 and 8/15 subjects respectively.
5.3.3.1 Threshold correspondence at 2 kHz, 4 kHz and the average of 2
kHz and 4 kHz
In the current study the mean click-evoked threshold differences for both ABR
systems (subjects in Group A and the total sample i.e. subjects in Group A
and B) corresponded with findings reported in the literature.
Hall (2007)
reported an average range of agreement of 5 dB to 15 dB between click161
evoked ABR thresholds and that of the behavioural PT thresholds at 2 kHz
and 4 kHz whilst Hood (1995) reported similar ranges (6 dB to 20 dB) for
normal hearing subjects between click-evoked ABR thresholds and the
behavioural thresholds at 2 kHz and 4 kHz.
Although the threshold
differences seem to be within a broad range as specified by literature, the
large range of threshold differences may provide an indication that both
systems are less consistent during ABR recordings in the CP population.
In particular, the inclusion of the inconsistent PT data affected the range of the
threshold differences for both systems. Although the range of the threshold
differences was also large for subjects in Group A, the range of particularly
the VS ABR system increased upon including the PT data of subjects in
Group B.
This may be attributed to the elevated click-evoked VS ABR
thresholds obtained in mainly three subjects (Subject 6, Subject 9 and Subject
12). Figure 5.6, Figure 5.7 and Figure 5.8 clearly show that in all but one of
these subjects (Subject 12) the click-evoked thresholds of the VS ABR system
exceeded 20 dB.
Frequency (kHz)
Intensity (dB HL)
2 kHz
4 kHz
0
10
20
30
40
50
60
70
80
Left behavioural PT thresholds
Right behavioural PT thresholds
Click-evoked VS ABR thresholds (comparable to the PT threshold of the right ear)
Click-evoked BL ABR thresholds (comparable to the PT threshold of the left ear)
Figure 5.6:
Behavioural PT thresholds (2 kHz and 4 kHz) and click-evoked
thresholds (VS and BL ABR systems) obtained in Subject 9
162
Frequency (kHz)
2 kHz
4 kHz
Intensity (dB HL)
0
10
20
30
40
50
60
70
80
Left behavioural PT
Right behavioural PT
Click-evoked VS ABR thresholds (comparable to the PT threshold of the right ear)
Click-evoked BL ABR thresholds (comparable to the PT threshold of the left ear)
Figure 5.7:
Behavioural PT thresholds (2 kHz and 4 kHz) and click-evoked
thresholds (VS and BL ABR systems) obtained in Subject 6
Frequency (kHz)
Intensity (dB HL)
2 kHz
4 kHz
0
10
20
30
40
50
60
70
80
Left behavioural PT thresholds
Right behavioural PT thresholds
Click-evoked VS thresholds (comparable to PT threshold of the left ear)
Click-evoked BL thresholds (comparable to PT threshold of the right ear)
Figure 5.8:
Behavioural PT thresholds (2 kHz and 4 kHz) and click-evoked
thresholds (VS and BL ABR systems) obtained in Subject 12
163
5.3.3.2 Threshold correspondence at 0.5 kHz
As there is a need in diagnostic audiometry to provide PT threshold estimates
for frequencies other than 2 kHz and 4 kHz, the use of frequency-specific ABR
recordings can be seen as inevitable (Hall, 2007; Gorga et al., 2006).
Although several techniques have been implemented to obtain frequencyspecificity, research indicated that TB stimuli can be used to predict the
magnitude as well as the configuration of a hearing loss reliably (Hall, 2007;
Gorga et al., 2006; Purdy & Abbas, 2002; Stapells, 2000).
In the current research 0.5 kHz TB stimuli were used to obtain information
regarding low frequency hearing (e.g. at 0.5 kHz) in a CP sample. Although
there are currently no research reports available of ABR recordings using TB
stimuli in the CP population, literature indicates that the 0.5 kHz TB thresholds
are generally within 10 dB to 30 dB of the behavioural PT threshold at 0.5 kHz
(Stapell, 2000; Stapells et al., 1995; Hall, 1992; Stapells et al., 1990).
The results of the current study indicated that the mean threshold differences
for the VS and BL ABR systems for subjects in both groups (subjects in Group
A as well as subjects in Group A and B) were within a broad range of
normality as specified by relevant literature (Stapells, 2000; Stapells et al.,
1995; Hall, 1992; Stapells et al., 1990). However, a large range of threshold
differences obtained with both system may indicate that the variability of the
ABR using 0.5 kHz TB stimuli in this population.
The range of threshold differences was particular influenced by large
individual threshold differences obtained from Subject 4, Subject 6 and
Subject 12.
Interestingly, Subject 4, Subject 6 and Subject 12 were all
diagnosed with a spastic form of CP. Whilst Subject 4 was diagnosed with
quadriplegia, Subject 6 and Subject 12 presented with right hemiplegia and
triplegia respectively.
Taking into account that Subject 4 and Subject 12
displayed high artifact rejection level in BL ABR recordings of 70% and 97%
(click and 0.5 kHz TB stimuli) and 55% and 65% (click and 0.5 kHz TB
stimuli) respectively as illustrated in Chapter 4 (Figure 4.10), it could be
argued that the severity of the sporadic spasms and/or reflexes displayed in
164
these subjects contributed to a poor SNR which ultimately challenged the
identification of the ABR wave V at lower intensities, i.e. near the actual
behavioural PT threshold.
This may imply that the threshold differences
obtained from both the ABR systems could probably have been smaller than
the ones reported in this study.
As in the case of the click-evoked ABR, it could be argued that the
incorporation of a cost-effective and relatively safe sleeping agent such as
melatonin could improve the SNR.
The improvement of the SNR could
ultimately lead to a better correspondence between the 0.5 kHz TB ABR and
the 0.5 kHz behavioural PT threshold.
5.3.4 Recording time of the VS and BL ABR systems using click and 0.5
kHz TB stimuli
When conducting a hearing test the general aim is to obtain as much
information as possible in the shortest possible time.
Together with the
feasibility rate of the ABR system and the correspondence to behavioural PT
thresholds, the recording time of the ABR provides an indication of the
usefulness of this procedure when assessing difficult-to-test populations such
as children with CP (Gorga et al., 2006; Bachmann & Hall, 2001).
In the current study the VS and BL ABR system were simultaneously
conducted in each subject. This suggests that the recording times for each
ABR system obtained were valid for only one ear. As reported in Chapter 4,
the recording times per ear for the two ABR systems did not differ
significantly: a mean recording time for a click-evoked ABR recording was 9
minutes per ear for both ABR systems, whilst the 0.5 kHz TB mean recording
time was 9 minutes and 11 minutes per ear using the BL and VS ABR
systems respectively.
Although the fact that recording times were only obtainable from one ear could
be seen as a drawback in this study, the results may provide some useful
information.
Recording time of one ear can provide an indication of the
projected recording time for the ABR assessment of both ears. Additionally,
as showed in Table 5.7, this projected recording time can provide an
165
indication of the time frame in which an ABR protocol, using the VS and BL
ABR system, might be completed specifically in the CP population.
Table 5.7: The actual and suggested recording time for VS and BL ABR
systems using click and 0.5 kHz TB stimuli
ABR protocol
Click stimuli
0.5 kHz TB stimuli
using click and
0.5 kHz TB
stimuli
Recording
Suggested
Recording
Suggested
Suggested
time
recording time
time
recording time
recording time
(1 ear)
for both ears
(1 ear)
for both ears
(both ears)
VS ABR
9 minutes
18 minutes
11 minutes
22 minutes
40 minutes
BL ABR
9 minutes
18 minutes
9 minutes
18 minutes
36 minutes
The administration of an ABR protocol i.e. a click-evoked recording and a
frequency-specific TB recording remains imperative in diagnostic audiology.
Thus, the time needed for the ABR protocol to be completed needs to be
taken into consideration.
As shown in Table 5.7, the projected time for
completing the ABR protocol is 40 minutes and 36 minutes using the VS and
BL ABR systems respectively.
It is clear that the ABR protocol using the BL ABR system favours the protocol
using the VS ABR system by 4 minutes. The 4 minute delay originates from
the extended recording time obtained during 0.5 kHz TB recordings. An
explanation for the slightly better recording time obtained from the BL system
when using the 0.5 kHz TB stimuli might relate to the inclusion of artifact
rejection. It might be argued that the inclusion of artifact rejection ensured
that the appropriate SNR was achieved faster in comparison to the
implementation of Kalman filtering used by the VS system. Kalman filtering
attempts to compensate for the contaminated sweeps by waiting until the
subject was restful before continuing with the averaging process (Kurtz &
166
Steinman, 2005). The averaging process of the VS system might have been
prolonged because few sporadic moments occurred during the ABR recording
where the subjects did not display muscular movements.
The presence of excessive muscular movements may directly have an effect
on the recording time of the ABR. Although it seems that artifact rejection
might be slightly more effective in terms of time efficiency when using 0.5 kHz
TB stimuli, the detrimental effects this process has on the morphology of the
ABR needs to be considered.
It could be argued that, the only way to
effectively reduce the recording time of the ABR assessment, yet maintain
clinical validity, is to ensure that an appropriate SNR is sustained throughout
the recording. Within a difficult-to-test population such as CP the only way to
sustain an appropriate SNR might be when excessive muscular movements
are reduced by using a light sedative such as melatonin (Schmidt, Krief,
Deuster, Matulat & Zehnoff-Dinnesen, 2007).
5.4
Conclusion
In this chapter the results of the two sub-aims were discussed separately.
The discussion of the results of Sub-aim 1 aimed to provide a general view
regarding the feasibility and characteristics of immittance, DPOAEs and
behavioural PT audiometry. The uniqueness and the complexity of CP were
emphasized through the variability of the results, not only between the
subjects, but also when compared to previous research. The variability of
results is in accordance with the heterogeneity of the CP condition and
stressed the importance of conducting a diagnostic audiological test battery
whilst taking into account the uniqueness of each child being assessed.
The discussion of sub-aim 2 was directed towards comparing the VS ABR
system to an ABR system with conventional technology (the BL ABR system)
in terms of its feasibility, threshold correspondence and recording time when
assessing auditory functioning in children with CP. Table 5.8 provides a short
summary of the conclusions based on the results obtained. Throughout the
discussion it was apparent that the size of the research sample affected the
results.
The results of the threshold correspondence were further
167
compromised because the consistency of the subjects’ responses to PT
stimuli was taken into account and on that basis the sample was divided into
two groups.
5.5
Summary
Chapter 5 provided a critical discussion of the results of the current research
in the light of existing literature. The results of the auditory procedures that
were conducted were discussed separately. Implications for future research
were indicated throughout the discussion and the limitations of the current and
previous research projects were identified.
168
Chapter 6
CONCLUSIONS AND RECOMMENDATIONS
The aim of this chapter is to infer general conclusions and implications
from the research, to critically evaluate the findings and make
recommendations for future research
6.1
Introduction
The main aim of this research project was to determine the clinical utility of the
Vivosonic Intergrity (VS) auditory brainstem response (ABR) system in children
with cerebral palsy (CP). The results obtained were presented and discussed in
the previous chapters.
Chapter 6 serves as the closing of the report.
Conclusions drawn from the reported results are presented in this chapter and
the research process is critically reviewed. Furthermore, recommendations and
implications for further research are presented in this chapter.
6.2
Conclusions
The research process described in this report was primarily aimed at determining
the clinical usefulness of the VS ABR system when assessing auditory
functioning in the CP population.
In order to realize the main aim various
procedures in the audiological test battery were administered on each CP subject
followed by simultaneous ABR measurements using the VS system as well as a
conventional ABR system. Throughout the research it was apparent that the
small sample size influenced the results of the research project. For this reason
significant differences by means of inferential statistics between the ABR
systems could not be determined.
169
Conclusions drawn from this project can be divided into two main sections,
namely audiological tests in children with CP and ABR assessments in the CP
population using the VS ABR system. Conclusions within each section are
presented accordingly.
6.2.1 Audiological tests in children with cerebral palsy
‰ The variability of the audiological test battery results – between the subjects
and when compared to previous research – emphasized the heterogeneity of
the CP population. This variability also stressed the importance of evaluating
each CP child’s auditory status carefully with a battery of tests to cross-check
findings and to identify a hearing loss or an auditory dysfunction such as
auditory neuropathy spectrum disorder (ANSD) appropriately.
‰ The severity of the physical impairment and of any additional impairments
such as mental retardation may influence the consistency and therefore also
the feasibility of behavioural pure tone (PT) audiometry. This was apparent
since eight subjects (n=15) responded inconsistently to behavioural PT
stimuli. The inconsistent behavioural PT data also stressed the necessity for
including frequency-specific ABR assessments when determining auditory
functioning in children with CP.
‰ Underlying conditions such as sickle cell disease (SCD) may be present in
children with CP (Ashley-Koch, Murphy, Khoury & Boyle, 2001). Such cooccurring conditions may affect the results of specifically distortion product
otoacoustic emissions (DPOAEs).
6.2.2 ABR assessments in the CP population using the Vivosonic Integrity
system
‰ Higher success rates obtained with the VS ABR system may suggest that this
system was feasible in a wider variety of subjects using click as well as 0.5
kHz TB stimuli. The findings of the current research showed that the VS ABR
system illustrated high success rates of ABR recordings using click and 0.5
kHz TB stimuli within a small CP sample which consisted of subjects with
170
spastic CP, athetoid CP as well as microcephaly. Success rates obtained
with the VS ABR system were higher than those obtained with the BL ABR
system (80% and 73% utilizing click and 0.5 kHz TB respectively).
‰ Technology employed in the VS ABR system (Kalman filtering) and the BL
ABR system (artifact rejection) may both be useful methods of limiting noise
such as myogenic potentials during an ABR recording. This conclusion may
be founded on the results obtained in this research that threshold differences
between the two ABR systems were not significant (p>0.05).
Threshold
differences obtained with the VS ABR system and the BL ABR system to
behavioural PT thresholds at 2 kHz, 4 kHz and 0.5 kHz fell within a broad
range of normality as indicated by literature (click-evoked thresholds: within 5
dB to 15 dB; 0.5 kHz TB thresholds: within 10 dB to 30 dB; Hall, 2007; Hood,
1998; Stapells, 2000; Stapells et al., 1995; Hall, 1992).
‰ The large range of threshold differences obtained may suggest increased
variability of both ABR systems to provide reliable PT estimates within the CP
population.
Although the threshold correspondence results did not offer
significant findings, the range of threshold differences was used in an attempt
to provide a better indication of the consistency of the systems. It was clear
that both ABR systems displayed a large range of differences between
electrophysiological click-evoked thresholds and behavioural PT thresholds
(i.e. at 2 kHz, 4 kHz and the average of 2 kHz and 4 kHz the threshold
differences of the VS system ranged between -1 dB and 32 dB, 2 dB and 42
dB and between -6 dB and 37 dB respectively whilst threshold differences of
the BL system varied between -1 dB and 34 dB, -11 dB and 29 dB and
between -14 dB and 27 dB at 2 kHz, 4 kHz and the average of 2 kHz and 4
kHz respectively).
‰ The recording time per ear of the VS and BL ABR systems was not
significantly different (p> 0.05), suggesting that both systems worked
reasonably well. Although the recording time per ear of the VS and BL ABR
systems was not significantly different, the projected recording time for an
ABR protocol, i.e. click-evoked and 0.5 kHz TB recordings, favours the BL
171
ABR system by 4 minutes. This could imply that the artifact rejection process
achieved the appropriate SNR within a shorter period compared to the
Kalman filtering method. As Kalman filtering attempts to compensate for the
contaminated sweeps by waiting until the subject was restful before
continuing with the averaging process, it could be suggested that the
averaging process of the VS system might have been prolonged because few
sporadic movements occurred during the ABR recording where the subjects
did not display muscular movements (Kurtz & Steinman, 2005).
6.3
Implications of the findings
The success rates obtained with both systems suggest that ABR recordings
using click- as well as TB stimuli might be more readily conducted in the CP
population. The success rates obtained with the BL ABR system (80% and 73%
utilizing click and 0.5 kHz TB stimuli respectively) in particular might suggest that
the inclusion of artifact rejection remains a reasonably efficient tool to improve
SNR during ABR recording, though it was applicable in less individual cases as
illustrated by the higher success rates of the VS ABR system.
The high success rates of click-evoked and 0.5 kHz TB recordings attained with
the VS ABR system without the use of any form of sedation have various
implications in terms of the application of the ABR on different test populations.
Since ABR recordings were successful without the use of sedation in a subject
sample characterized by excessive muscular movements, it implies that the VS
ABR may also be applicable and feasible in other difficult-to-test populations
such as infants. Furthermore, the fact that the recordings were conducted in
awake and alert subjects suggests that the VS ABR system might be useful in
infants who are awake, sleeping or being breastfed. It could be concluded that
on quiet awake subjects, including infants, ABR assessments may successfully
be recorded using the VS ABR system. Thus, this system may be applicable for
screening programs.
172
The large range of threshold differences that were obtained with the VS ABR
system (click: -8 dB to 42 dB; 0.5 kHz TB: 8 dB to 53 dB) and BL ABR system
(click: -16 dB to 34 dB; 0.5 kHz TB: 20 dB to 70 dB) may suggest less
consistency using both ABR systems. This may imply that, regardless of the
technology used (e.g. Kalman filtering or artifact rejection) diagnostic ABR
assessments,
i.e.
determining
electrophysiological
thresholds,
remain
challenging within the CP population. Determining ABR thresholds within this
population may be regarded as a challenge due to the presence of various
symptoms of central nervous system dysfunction most notably of which are
inconsistent muscular movements that may cause large myogenic potentials
during the ABR recording.
As there were subjects in whom ABR recordings were not feasible with the VS
and BL ABR systems, and/or subjects in whom the threshold differences
obtained with both systems were large (i.e. 42 dB) it could be argued that both
artifact rejection and Kalman filtering only partly improved the SNR during the
ABR recordings. This implies that, in order to effectively improve the SNR, CP
children need to remain in a relaxed and calm state displaying minimal muscular
movements.
As CP children have limited or no control over muscular
movements, a ‘relaxed and calm state’ might only be obtained by the utilization of
a light sedative (Workinger, 2005; Mechem, 2002; Cogher et al., 2001). A light
sedative for ABR assessments in the CP population might be found in the form of
melatonin. Research indicates that melatonin poses minimal risks to children
with multiple disabilities including CP (Schmidt, Krief, Deuster, Matulat & ZehnoffDinnesen, 2007).
Improving the SNR by implementing a patient management technique, i.e. using
a light sedative, may directly assist the audiologist in obtaining closer estimates
of ABR thresholds to the behavioural PT thresholds. Furthermore, it may also
reduce the recording time per ear which may add to the usefulness of ABR
assessments in general within the CP population.
173
6.4
Critical evaluation of the study
In this section the procedures and protocols used in the study are critically
evaluated according to their strengths and limitations.
The following strengths of this study have been identified:
‰ The administration of a diagnostic audiometric battery can be seen as a
strength of this study. The variety of measurements for auditory functioning
assisted the researcher in a comparison of the results.
Additionally, it
emphasized the importance of the cross-check principle particularly in
children with CP.
‰ A within-subject condition was used for evaluating both ABR systems
simultaneously. This meant that both systems were exposed to similar test
conditions in terms of muscular activity (EEG) and environmental noise during
the testing.
‰ The fact that TB stimuli were used during ABR recordings can also be seen
as a strong point of the study. There are currently no research reports that
reveal information regarding ABR recordings using TB stimuli in children with
CP. The inclusion of TB stimuli can be seen as an integral part of the ABR
assessment because frequency-specific estimates of PT thresholds provide a
comprehensive picture of the audiogram.
‰ Within the current study ABR recordings were conducted without the use of
sedatives.
This could be identified as an additional strength since ABR
recordings in previous research were conducted with the use of sedatives.
However, because children with multiple disabilities illustrate a high risk for
airway obstruction during sedation or during natural sleep, sedatives was
avoided in this study (Schmidt, et al., 2007; Elwood, Hansen, Seely, 2001;
Benham-Dunster & Dunster, 1985; Stein et al., 1981).
The following limitations of the study have been identified:
‰ Because of the small sample size (n=15), in-depth categorization of certain
sets of data, e.g. ipsilateral acoustic reflexes and DPOAEs, was not possible.
174
Furthermore, the size of the research sample affected the inferential part of
the statistics which influenced the second sub-aim of the project, which was
to compare the VS ABR system to a conventional ABR system. Although
there were some tendencies towards a 95% confidence level for some
aspects such as the feasibility of the systems, it was not possible to
determine significant differences between the VS ABR and BL ABR systems.
‰ The administration of the VS and BL ABR systems simultaneously could also
be seen as a limitation of the study. As previously explained, this set-up was
necessary to ensure similar test conditions when conducting ABR recordings
using different systems.
However, this set-up also caused that the
electrophysiological thresholds obtained with the different ABR systems could
not be directly compared to each other since they were recorded from
different ears in the same subject.
As a result the electrophysiological
thresholds were compared indirectly to each other in terms of the threshold
correspondences to behavioural PT thresholds. The comparison between the
two ABR systems in terms of the threshold correspondences to behavioural
PT thresholds were further complicated as some of the subjects (n=7)
responded inconsistently to PT stimuli.
‰ The use of a 100 Hz low pass filter during testing could be seen as a
limitation of the study as this setting might have affected 0.5 kHz TB
recordings of both ABR systems. It is well known that TB recordings depend
on low frequency energy (Hall, 2007). The use of a high low pass filter setting
such as 100 Hz therefore may compromise recordings to such an extent that
elevated electrophysiological thresholds are obtained (Hall, 2007).
‰ The use of TB stimuli at only one frequency could be seen as a further
limitation of the study. Frequency-specific ABR recording at 1 kHz, 2 kHz and
4 kHz may provide valuable insight not only in terms of the configuration of
hearing loss in the CP population, but may also improve the rehabilitative
services rendered in this population, such as fitting and verifying hearing aids.
‰ Another limitation of the study was that click-evoked recordings were
recorded using rarefaction polarity only.
Using a condensation polarity
175
recording after a rarefaction polarity could not only have increased the
reliability of the presence of the ABR thresholds, but it could also have
indicated subjects with possible auditory neuropathy spectrum disorder
(ANSD) by means of the presence of the cochlear microphonic.
6.5
Recommendations for further research
‰ The presence of increased DPOAE amplitudes in the CP population needs to
be investigated. Within the current sample 6 out of 10 subjects with spastic
CP presented with enlarged DPOAE amplitudes mainly in the lower and mid
frequencies (635 Hz – 1586 Hz). Although a possible relationship between
SCD and CP in these subjects has been mentioned, it remains purely
speculative.
Further research needs to be conducted to determine the
prevalence of increased amplitudes in the various sub-groups of CP, i.e.
spastic CP, mixed CP and athetoid CP. Future research projects could also
focus on the amplitudes of subjects with CP in different age-groups
comparing it to control groups of similar ages. Additionally, research could be
directed towards the presence of DPOAE as well as the amplitudes of the
DPOAE of CP children with confirmed SCD.
‰ The current study could be replicated, but with a larger research sample than
in the current study where only 15 subjects were included.
Increasing the
size of the research sample may contribute to more significant values
obtained with inferential statistics, which may lead to more specific findings.
‰ As mentioned earlier, the simultaneous administration of the two ABR
systems caused that the ABR thresholds could not be directly compared to
each other. Developing a method of directly comparing electrophysiological
thresholds of the two ABR systems in the same ear therefore seems
necessary. This could be achieved by monitoring the EEG during each ABR
recording and ensuring that the average EEG is comparable when two
successive measurements are obtained in the same ear at the same
frequency with the two ABR systems.
176
‰ ABR recordings using the VS ABR system should be conducted on CP
infants with and without the use of sedation. This type of study may involve
high expenditures as trained personnel will need to be incorporated while
conducting the ABR with sedation on the CP infant. However, completing a
project like this and proving that the VS ABR system can be administrated
with success without the use of sedation or general anaesthesia on this
population could be significant since it may enhance the applicability of the
VS ABR system.
‰ Tone burst ABR recordings at 1 kHz, 2 kHz and 4 kHz need to be conducted
in the CP population. Currently there are no reports available which focused
on TB recordings in this population. Considering that frequency-specific ABR
thresholds are needed to provide optimal rehabilitative services such as
verifying hearing aids, TB recordings remain imperative in a population which
display a higher risk for hearing loss (Hall, 2007; Sano et al., 2005; Topolska
et. al., 2002; Sheykholeslami & Kaga, 1999; McDonald, 1987; Newton, 1977).
‰ Since the CP population may display a high risk for ANSD, the prevalence of
ANSD needs to be investigated by the implementation of a diagnostic
audiological protocol (Shapiro, 2003). Research indicated that the presence
of ANSD can be determined by a battery of tests including DPOAEs, acoustic
reflexes as well as ABR assessments.
In the latter, the presence of a
cochlear microphonic within a certain time period remains deterministic for the
diagnosis of ANSD, yet is not well researched within the CP population ().
6.6
Final conclusion
The CP condition is a universal phenomenon (Andersen, 2008; Donnely et al.,
2007; Fawke, 2007; Couper, 2005; Winter, Autry, Boyle & Yeargin-Allsopp, 2002;
Arens & Molteno, 1989).
As this vulnerable population also presents with a
higher risk for a hearing loss, the audiologist in South Africa is not only in need of
sufficient knowledge and clinical skills, but also need audiometric and
electrophysiological equipment to assess auditory functioning effectively and
efficiently.
177
The variability of the audiological test battery results obtained in the current
research emphasizes the heterogeneity of the CP population and stresses the
importance of evaluating each CP child’s auditory status carefully with a battery
of tests to cross-check findings. However, behavioural PT audiometry, which is
regarded as the gold standard of audiometry, is not always viable or reliable in
this population (Folsom & Diefendorf, 1999). Thus, the ABR remains an integral
part of the auditory test battery for timeous identification of a hearing loss.
Within the current study the VS and BL ABR systems illustrated high success
rates during click-evoked and 0.5 kHz TB evoked recording. However, the VS
ABR system was successful across a wider range of subjects during click-evoked
and 0.5 kHz TB recordings, which may increase its clinical usefulness, especially
in terms of hearing screening in the CP population. It seemed that excessive
muscular movements during the recordings influenced not only the VS ABR’s,
but also the BL ABR’s threshold correspondences to PT thresholds as well as the
recording time of the measurements. Hence, it appears that the use of a light
sedative to reduce excessive muscular movements may still be necessary to
increase the clinical usefulness of the VS ABR system in the CP population in
general.
178
REFERENCES
ANSI (1996). American national standard specification for audiometers. ANSI S.
3.6 – 1996.
Ashley-Koch, A., Murphy, C.C., Khoury, M.J. MD, Boyle, C.A. (2001).
Contribution of sickle cell disease to the occurrence of developmental disabilities:
A population-based study. Genetics in Medicine, 3(3), pp. 181-186.
Amstrong, R.W. (2007). Definition and Classification of Cerebral Palsy.
Developmental Medicine & Child Neurology, 49, pp.166.
Andersen, G.L., Irgens, L. M., Haagaas, I., Skranes, J.S., Meberg, A.E. & Vik, T.
(2008). Cerebral Palsy in Norway: Prevalence, Subtypes and Severity. European
Journal of Paediatric Neurology, Vol. 12 (1), pp. 4-13.
Angulo, C.M., Blanco, N.A., Teran, J.G., Aledo, A.G. & Quintela, J.R. (2006).
Sensorineural
Hearing
Loss
in
Cerebral
Palsy
Patients.
Acta
Otorrinolaringologica Espanola, Vol. 57, pp. 300-302.
Arens, L.J. & Molteno, C.D. (1989). A Comparative Study of PostnatallyAcquired Cerebral Palsy in Cape Town. Developmental Medicine and Child
Neurology, 31, pp. 246 -254.
Arnold, S.A. The Auditory Brainstem Response (2000). In: Audiology Diagnosis.
New York:Thieme. pp. 451-470.
Bachmann, K.R. & Hall, J.W. III (2001). Pediatric auditory brainstem response
assessment: The cross-check principle twenty years later. Seminars in Hearing,
19(1), pp. 41-60.
179
Baer, J. E. & Hall III, J.W. (1992). Effects of Nonpathological Factors on
Otoacoustic Emissions. The Hearing Journal, 45 (11), pp. 17-23.
Beckung, E. & Hagberg, G. (2002). Neuroimpairments, Activity Limitations, and
Participation Restrictions in Children with Cerebral Palsy. Developmental
Medicine & Child Neurology, 44, pp. 309-316.
Benham-Dunster,
R.A.
&
Dunster,
J.R.
(1985).
Hearing
loss
in
the
Developmentally Handicapped: A Comparison of Three Audiometric Procedures.
The Journal of Auditory Research, 25, pp.175-190.
Bergman, B.M., Beauchaine, K.L. & Gorga, M.P. (1992). Application of the
Auditory Brainstem Response in Pediatric Audiology. The Hearing Journal, 45
(9), pp.19-25.
Bess, F.H. (1993). Early Identification of Hearing Loss: A Review of the Whys,
Hows and Whens. The Hearing Journal, 46(6), pp. 22-25.
Block, M.G. & Wiley, T.L. (1994). Overview and Basic Principles of Acoustic
Immittance. In: Handbook of Clinical Audiology, (4th ed.). Philadelphia: Lippencott
Williams & Wilkins. pp. 271-282.
Bottos, M., Feliciangeli, A., Sciuto, L., Gericke, C. & Vianello, A. (2001).
Functional Status of Adults with Cerebral Palsy and Implications for Treatment of
Children. Developmental Medicine & Child Neurology, 43, pp. 516-528.
Byers, R.K., Paine, R.S. & Crothers, B. (1955). Extrapyramidal Cerebral Palsy
with Hearing Loss Following Erythroblastosis. Pediatrics, 15, pp. 248-254.
Christianson, A.L., Zwane, M.E., Manga, P., Rosen, E., Venter, A., Downs, D. &
Kromberg, J.G.R. (2002). Children with Intellectual Disability in Rural South
180
Africa: Prevalence and Associated Disability. Journal of Intellectual Disability
Research, 46 (2), pp. 179-186.
Cogher, L., Savage, E. & Smith, M.F. (1992). Management of Disability Series.
Cerebral Palsy. The child and the younger person. London: Chapman & Hall.
Cooper, S., Lyall, H, Walters, S., Tudor-Williams, G., Habibi, P., De Munter, C.,
Britto, J. & Nadel, S. (2004). Children with Human Immunodeficiency Virus
Admitted to a Paediatric Intensive Care Unit in the United Kingdom over a 10year Period. Intensive Care Medicine, 30, pp.113-118.
Coté, C.J. & Wilson, S. (2006). Guidelines for Monitoring and Management of
Pediatric Patients During and After Sedation for Diagnostic and Therapeutic
Procedures: An Update. Pediatrics, 118(6), pp. 2587-2602.
Couper, J. (2005). Prevalence of Childhood Disability in Rural KwaZulu-Natal.
South African Medical Journal, 92(7), pp. 549-552.
Cone-Wesson, B, Vohr, B.R., Sininger, Y.S., Widen, J.E., Folsom, R.C., Gorga,
M.P. & Norton, S.J. (2000). Identification of Neonatal Hearing Impairment: Infants
with Hearing Loss. Ear & Hearing, 21(5), pp. 488-507.
Danhauer, J.L. (1997). How Otoacoustic Emissions Testing Can Change Your
Audiology Practice. The Hearing Journal, 50(4), pp. 62-68.
Delport, C.S.L. (1998). Quantatative data collection methods. In: Research at
Grassroots. Pretoria: Van Schaik Publishers.
181
Diefendorf, A.O. (2003). Behavioral Hearing Assessment: Considerations for the
Young Child with Developmental Disabilities. Seminars in Hearing, 24(3), pp.189199.
Donnelly, C., Parkes, J., McDowell, B. & Duffy, C. (2007). Lifestyle Limitations of
Children and Young People with Severe Cerebral Palsy: A Population Study
Protocol. Journal of Advanced Nursing, 61(5), pp.557-569.
Downs, M.P. & Yoshinaga-Itano, C. (1999). The Efficacy of Early Identification
and Intervention for Children with Hearing Impairment. Pediatric Clinics of North
America, 46(1), pp. 79-87.
Downs, C.R., Stuart, A. & Holbert, D. (2000). Distortion product otoacoustic
emissions in normal-hearing children with homozygous sickle cell disease.
Journal of Communication Disorders, 33, pp.111-129.
Driscoll, C., Kei, J., Bates, D., & McPherson, B. (2002). Transient Evoked
Otoacoustic Emissions in Children Studying in Special Schools. International
Journal of Pediatric Otorhinolaryngology, 64(1), pp.51- 60.
Drummond, A. (2003). Research Methods for Therapists. United Kingdom:
Nelson Thornes Ltd.
Durrant, J.D. (1992). Distortion-Product OAE Analysis: Is It Ready For Broad
Clinical Use? The Hearing Journal, 45(11), pp. 42-45.
Eggermont, J.J., Herrmann, B.S., Thornton, A.R. & Hyde, M.L. (1991). Peer
Commentary on “Clinical Usefulness of Auditory Evoked Potentials: A Critical
Evaluation” by T.W. Picton. JSPLA, 15(1), pp.19-26.
182
Elwood, T., Hansen, L.D., Seely, J.M. (2001). Oropharyngeal Airway Diameter
During Sedation in Children With and Without Developmental Delay.
Journal of Clinical Anesthesia, 13, pp. 482–485.
Emmer, M.B. & Silman, S. (2003). The Prediction of Hearing Loss in Persons
with Cerebral Palsy Using Contralateral Acoustic Reflex Threshold for Broadband
Noise. American Journal of Audiology, 12, pp. 91-95.
Fawke, J. (2007). Neurological Outcomes Following Preterm Birth. Seminars in
Fetal and Neonatal Medicine, 12(5), pp. 374-382.
Ferraro, J.A., & Durrant, J.D. (1994). Auditory Evoked Potentials: Overview and
Basic Principles. In: Handbook of Clinical Audiology (4th ed.), Philadelphia:
Lippencott Williams & Wilkins, pp. 317-338.
Folsom, R.C. & Diefendorf, A. (1999). Physiologic and Behavioural Approaches
to Pediatric Hearing Assessment. Pediatric Clinics of North America, 46(1). pp.
107-120.
Galambos, R., Hicks, G.E. & Wilson, M.J. (1984). The Auditory Brainstem
Response Reliably Predicts Hearing Loss in Graduates of a Tertiary Intensive
Care Nursery. Ear and Hearing, 5(4). pp. 254-260.
Gans D. & Gans, K.D. (1993). Development of a Hearing Test Protocol for
Profoundly Involved Multi-Handicapped Children. Ear and Hearing, 14(2), pp.128
-140.
Ginsberg, I.A. & White, T.P. (1994). Otologic Disorders and Examination. In:
Handbook of Clinical Audiology (4th ed.). Philadelphia: Lippencott Williams &
Wilkins. pp. 6-24.
183
Gorga, M.P. Worthington, D.W. Reiland, J.K., Beauchaine, K.A. & Goldgar, D.E.
(1985). Some Comparisons between Auditory Brain Stem Response Thresholds,
Latencies, and the Pure-Tone Audiogram. Ear and Hearing, 6(2), pp. 105-112.
Gorga, M.P., Kaminski, J.K., Beauchaine, K.L., Bergman, B.M. (1993). A
Comparison of Auditory Brainstem Latencies Elicited by Air-and Bone-Conducted
Stimuli. Ear and Hearing, 14(2), pp. 85-94.
Gorga, M.P., Johnson, T.A. Kaminski, J.K., Beauchaine, K.L., Garner, C.A. &
Neely, S.T. (2006). Using a Combination of Click- and Tone burst-evoked
Auditory Brainstem Response Measurements to Estimate Pure-Tone Thresholds.
Ear and Hearing, 27(1), pp. 60-74.
Hall III, J. W. (2007). New Handbook of Auditory Evoked Responses. Boston:
Pearson Education, Inc.
Hall III, J.W. (1988). Auditory Evoked Potentials in the Management of Acutely
Brain-Injured Children and Adults. The American Journal of Otology. Vol. 9, (Dec
suppl.). pp.36-46.
Hall III, J.W. & Chandler, (1994). Tympanometry in Clinical Audiology. In:
Handbook of Clinical Audiology (4th ed.). Philadelphia: Lippencott Williams &
Wilkins. pp. 282-299.
Hall III, J.W. & Mueller, H.G. (1997). Audiologists’ Desk Reference Volume 1.
Diagnostic Audiology Principles, procedures and practices. San Diego: Singular
Publishing Group, Inc.
Hannley, M. (1986). Basic Principles of Auditory Assessment. San Diego:
College-Hill Press, Inc.
184
Harrison, M. & Roush, J. (1996). Age of Suspicion, Identification, and
Intervention for Infants and Young Children with Hearing Loss: A National Study.
Ear and Hearing, 17(1), pp. 55-62.
Hedge, M.N. (2003). Clinical Research in Communicative Disorders: Principles
and Strategies (3rd ed). Texas: Pro-ed.
Hemming, K., Hutton, J.L., Colver, A. & Platt, M. (2005). Regional Variation in
Survival of People with Cerebral Palsy in the United Kingdom. Pediatrics, 116
(6), pp.1381-1390.
Hodgson, W.R. (1994). Evaluating Infants and Children. In: Handbook of Clinical
Audiology (4th ed). Philadelphia: Lippencott Williams & Wilkins. pp. 465-475.
Hood, L.J. (1998). Clinical Applications of the Auditory Brainstem Response. San
Diego: Singular Publishing Group.
Hood, L.J. (1995). Estimating Auditory Function with Auditory Evoked Potentials.
The Hearing Journal, 48 (10), pp. 32-42.
Hood, L.J. & Berlin, C.I. (2002). Clinical Applications of Otoacoustic Emissions.
In: Hair Cell Micromechanics and Otoacoustic Emissions. New York: Thomson
Delmar Learning. pp. 121-137.
Hutton, J.L., & Pharoah, P.O.D. (2002). Effects of Cognitive, Motor and Sensory
Disabilities on Survival in Cerebral Palsy. Archives of Disease in Childhood, 86
(2), pp. 84-89.
Iacono, T. & Murray, V. (2003). Issues of Informed Consent in Conducting
Medical Research Involving People with intelectual disabilities. Journal of
AppliedResearch in Intellectual Disabilities, 16(1), pp. 41-51.
185
Jamieson, J.R. (1994). The Impact of a Hearing Impairment. In: Handbook of
Clinical Audiology (4th ed). Philadelphia: Lippencott Williams & Wilkins. pp. 596615.
Jaseja, H. (2008). Cerebral Palsy: Is it truly absolutely non-progressive in nature?
Clinical Neurology and Neurosurgery, 110(2), pp. 211-212.
Jenkins, S., Price, C.J. & Straker, L. (2003). The researching Therapist.
Edinburgh: Churchill Livingstone.
Jiang Z. D. & Wilkonson, A.R. (2006). Does Peripheral Auditory Threshold
Correlate with Brainstem Auditory Function at Term in Preterm Infants? Acta OtoLaryngologica, 126, pp. 824-827.
Joint Committee on Infant Hearing (2007). Year 2007 Position Statement:
Principles and Guidelines For Early Hearing Detection and Intervention
Programs. Pediatrics, 120(4), pp. 898-921.
John A. Ferraro, J.A., Durrant, J.D., Sininger, Y. & Campbell, K. (1996).
Recommended Guidelines for Reporting AEP Specifications. American Journal
of Audiology, 5(3). pp. 35-37.
Johnstone, J.M. & Pennypacker, H.S. (1993). Strategies and Tactics of
Behavioral Research (2nd ed.). New Jersey: Lawrence Erlbaum.
Kennes, J., Rosenbaum, P., Hanna, S.E., Walter, S, Russell, D., Raina, P.,
Bartlett, D. & Galuppi, B. (2002). Health Status of School –Aged Children with
Cerebral Palsy: Information from a Population-Based Sample. Developmental
Medicine & Child Neurology, 44, pp. 240-247.
Kidder, L. H., Judd, C.M. (1986). Research Methods in Social Relations (5th ed.).
New York: CBS College Publishing.
186
Klein, J.O. (1977). Epidemiology of Otitis Media. In: Impedance Screening for
Middle Ear Disease in Children. New York: Harcourt Brace Jovanovich,
Publishers.
Kolker, I.A. (2004). Hearing Function and Auditory Evoked Potentials in Children
with Spastic Forms of Cerebral Palsy. Neurophysiology, 23(4), pp.270- 275.
Kulekci, S., Terlemez, S, Ciprut, A. & Akdas, F. (2007). 500Hz Logon versus
Click ABR Maturation. International Journal of Pediatric Otorhinolaryngology, 71,
pp. 775-779.
Kurtz, I. & Steinman, A. (2005). Kalman filtering in recording Auditory Evoked
Potentials (219). Presented at the 28th Midwinter Meeting of the Association of
Otolaryngology, New Orleans, LA. February. pp.19-24.
Lawson, R.D. & Badawi, N. (2003). Etiology of Cerebral Palsy. Hand Clinics, 19.
pp. 547-556.
Leedy, P.D. & Ormrod, J.E. (2005). Practical Research Planning & Designing
(8th ed). New Jearsy: Pearson Education, Inc.
Lonsbury-Martin, B.L., McCoy, M.J., Whitehead, M.L. & Martin, G.K. (1992).
Otoacoustic Emissions: Future Directions for Research and Clinical Applications.
The Hearing Journal, 45(11). pp. 47-52.
Louw, D.A., Van Ede, D.M. & Louw, A.E. (1998). Menslike Ontwikkeling (3rd ed.).
Pretoria: Kagiso Tersier.
Luterman, D., Kurtzer-White, E. & Seewald, R.C. (1999). The Young Deaf Child.
Baltimore: York Press, Inc.
187
Luts, H., Desloovere, C., Kumar, A., Vandermeersch, E. & Wouters, J. (2004).
Objective assessment of frequency-specific hearing thresholds in babies.
International Journal of Pediatric Otorhinolaryngology, 68(7), pp. 915-926.
MacCarthy, A. & Connell, J. (1984). Audiological Screening and assessment. In:
Screening for children with special needs Multidisciplinary Approaches. London:
Croom Helm Ltd.
MacDonald, E.T. (1987). Treating Cerebral Palsy. Austin, Texas: Pro-ed.
Madden,C., Rutter,M., Hilbert,L., Greinwald,J.H., Choo,D.I. (2002). Clinical and
Audiological Features in Auditory Neuropathy. Archives of Otolaryngology Head
Neck Surgery, 128, pp.1026 -1030.
Martilla, T.I & Karikoski, J.O. (2005). Comparison Between Audiometric and ABR
thresholds in Children: Contradictory Findings. European Archives of Oto-RhinoLaryngology, 263, pp. 399-403.
Mauk, G.W., Cags, MA, White, K.R., Mortensen, L.B., Behrens, T.R. (1991). The
Effectiveness of Screening Programs Based on High-Risk Characteristics in
Early Identification of Hearing Impairment. Ear and Hearing, 12(5), pp. 312-319.
Maxwell, D.L., & Satake, E. (1997). Research and Statistical Methods in
Communication Disorders. Williams & Wilkins, Baltimore.
Mecham, M.J. (2002). Cerebral Palsy (3rd ed.). Austin, Texas: Pro-ed.
Minear, W.L. (1956). Special Article: A Classification of Cerebral Palsy.
Pediatrics, 18, pp.841-852.
188
Mitchell, W. (2001). Neurological and Developmental Effects of HIV and AIDS in
Children and Adolescents. Mental Retardation and Developmental Disabilities
Research Reviews, 7, pp. 211-216.
Musiek, F.E., Borenstein, S.P., Hall, J.W. III & Schwaber, M.K. (1994). Auditory
Brainstem Response: Neurodiagnostic and Intraoperative Applications. In:
Handbook of Clinical Audiology (4th ed). Philadelphia: Lippencott Williams &
Wilkins. pp. 351-374.
Nakamura, H., Takada, S. Shimabuku, R. Matsuo, M. Matsuo, T. & Negishi, H.
(1985). Auditory Nerve and Brainstem Responses in Newborn Infants with
Hyperbilirubinemia. Pediatrics, 75, pp.703-708.
Nekahm, D., Weichbold, V. & Welzl-Muller, K. (2001). Epidemiology of
permanent childhood hearing impairment in the Tyrol, 1980 – 94. Scandinavian
Audiology, 30, pp. 197-202.
Neuman, W.L. (1997). Social Research Methods: Qualitative and Quantative
Approaches, (3rd ed). Boston: Allyn and Bacon.
Newell, M. (1998). Mechanisms and Timing of Mother-to-Child Transmission of
HIV-1 (Editorial Review). AIDS, 12, pp. 831–837.
Ngo, R.Y.S., Tan, H.K.K., Balakrishnan, A., Lim, S.B. & Lazaroo, D.T. (2006).
Auditory Neuropathy/Auditory Dys-Synchrony Detected by Universal Newborn
Hearing Screening. International Journal of Pediatric Otorhinolaryngology, 70
(7), pp. 1299-1306.
Northern, J. L. & Downs, M. P. (1991). Hearing in Children, (4thed). Baltimore:
Williams & Williams.
189
Northern, J.L. & Gabbard, S.A. (1994). The Acoustic Reflex. In: Handbook of
Clinical Audiology (4th ed). Philadelphia: Lippencott Williams & Wilkins. pp. 300316.
Norton, S. J. & Stover, L.J. (1994). Otoacoustic Emissions: An Emerging Clinical
Tool. In: Handbook of Clinical Audiology (4th ed). Philadelphia: Lippencott
Williams & Wilkins. pp. 448-462.
Oates, P. & Stapells, D.R. (1998). Auditory Brainstem Response Estimates Of
The Pure – Tone Audiogram: Current Status. Seminars in Hearing, 19(1), pp. 6185.
Oguro, K. & Hojo, H.A.H. (2001). Different responses to auditory and
somaesthetic stimulation in patients with an excessive startle: a report of
pediatric experience. Clinical Neurophysiology,112, pp.1266-1272.
Olusanya, B.O. (2008). Priorities for early hearing detection and intervention in
sub-Saharan Africa. International Journal of Audiology, 47 (suppl. 1), pp.3-13.
Orlin, M.N., Effgen, S.K. & Handler, S.D. (1997). Effect of Otitis Media with
Effusion on Gross Motor Ability in Preschool-aged Children: Preliminary Findings.
Pediatrics, 99, pp. 334-337.
O’Shea, M. (2008). Cerebral Palsy. Seminars in Perinatology, 53, pp. 35-41.
Palmu, A., Puhakka, H., Rahko, T., Takala, A.K. (1999). Diagnostic Value of
Tympanometry in Infants in Clinical Practice. International Journal of Pediatric
Otorhinolaryngology, 49, pp.207-213.
190
Parker, J.N., Parker, P. M. (2002) The Official Parent’s Sourcebook on Cerebral
Palsy. San Diego: ICON Heath Publications Inc.
Picton, T.W. (1991). Clinical usefulness of Auditory Evoked Potentials: A Critical
Evaluation. JSLPA, 15 (1), pp. 3-17.
Purdy, S.C., Abbas, P.J. (2002). ABR Thresholds to Tonebursts Gated with
Blackman and Linear Windows in Adults with High-Frequency Sensorineural
Hearing Loss. Ear and Hearing, 23 (4), pp. 358-368.
Rance, G., Beer, D. E., Cone-Wesson, B., Shepherd, R.K., Dowell, R.C., King, A.M.,
Rickards, F.W., Clark, G.M. (1999). Clinical Findings for a Group of Infants and Young
Children with Auditory Neuropathy. Ear and Hearing, 20(3), p.238.
Roeser, R.J., Valente, M. & Hosfurd-Dunn, H. (2000). Diagnostic Procedures in
the Profession of Audiology. In: Audiology Diagnosis. New York: Thieme Medical
Publishers. pp. 1-18.
Romero, G., Mendez, I., Tello, A. & Torner, C. (2008). Auditory Brainstem
Responses as a Clinical Tool in Children After Perinatal Encephalopathy.
International Journal of Pediatric Otorhinolaryngology, 72 (2), pp.193- 201.
Rossetti, L.M. (1996). Communication Intervention: Birth to Three. San Diego:
Singular Publishing group, Inc.
Rowe III, M.J. (1981). The Brainstem Auditory Evoked Response in Neurological
Disease: A Review. Ear and Hearing, 2(1), pp. 41-51.
Russman, B.S. & Ashwal, S. (2004). Evaluation of the Child With Cerebral Palsy.
Seminars in Pediatric Neurology, 11(1), 47-57.
191
Sanchez, J.T. & Gans, D. (2006). Effects of Artifact Rejection and Bayesian
Weighting on the Auditory Brainstem Response During Quiet and Active
Behavioural Conditions. American Journal of Audiology,15, pp. 154-163.
Sano, M., Kaga, K., Kitazumi, E., Kodama, K. (2005). Sensorineural Hearing
Loss in Patients with Cerebral Palsy After Asphyxia and Hyperbilirubinemia.
International Journal of Pediatric Otorhinolaryngology, 69 (9), pp. 1211-1217
Scherzer, A.L. (2001). History, definition and classification of cerebral palsy. In:
Early Diagnosis and Intervention Therapy in Cerebral Palsy: An interdisciplinary
age-focused approach. (3rd ed). New York: Marcel Dekker Inc. pp.1-26.
Schmidt, C-M., Knief, A., Deuster, D., Matulat, P. & Zehnoff-Dinnesen, A.G.
(2007). Melatonin is a Useful Alternative to Sedation in Children Undergoing
Brainstem Audiometry with an Age-Dependant Success Rate – A Field Study of
250 Investigations. Neuropediatrics, 38, pp. 2-4.
Sininger, Y.S. (1993). Auditory Brainstem Response for Objective Measures of
Hearing. Ear and Hearing, 14(1), pp. 23-30.
Sininger, Y.S. (2003). Audiologic Assessment in Infants. Current Opinion in
Otolaryngology Head and Neck Surgery, 11(5), pp. 378-382.
Sininger, Y.S., Doyle, K.J., Moore, J.K. (1999). The Case for Early Identification
of Hearing Loss in Children. . Pediatric Clinics of North America, 46 (1), pp. 1-14.
Shapiro, S.M. (2003). Bilirubin Toxicity in the Developing Nervous System.
Pediatric Neurology, 29, pp. 410-421.
192
Sheykholeslami, K. & Kaga, K. (2000). Otoacoustic Emissions and Auditory
Brainstem Responses After Neonatal Hyperbilirubinemia. International Journal of
Pediatric Otorhinolaryngology,52(1), pp. 65 -73.
Shoup, A.G. & Roeser, R.J. (2000). Audiologic Evaluation of Special Population.
In: Audiology Diagnosis. New York: Thieme Medical Publishers. pp. 311-329.
Stach, B.A. (1998). Clinical Audiology: An Introduction. San Diego: Singular
Publishing Group.
Stanley, F., Blair, E. & Alberman, E. (2000) Cerebral Palsies: Epidemiology &
Causal Pathways. London: Mac Keith Press.
Stanton, M. (1992). Cerebral Palsy. A Practical Guide. London: Optima, Little
Brown Ltd.
Stapells, D.R. (2000). Threshold Estimation by the Tone-Evoked ABR: A
Literature Meta-Analysis. JSLPA. 24(2), pp. 74-83.
Stapells, D. R., Gravel, J.S. & Martin, B.A. (1995). Thresholds for Auditory Brain
Stem Responses To Tones in Notched Noise from Infants and Young Children
with Normal Hearing or Sensorineural Hearing Loss. Ear and Hearing, 16(4), pp
361-371.
Stapells, D.R. & Picton, T.W. (1981). Technical Aspects of Brainstem Evoked
Potential Audiometry Using Tones. Ear and Hearing, 2, pp. 20-29.
Stein, A.G.W., Smit, C.F., Du Toit, S.H.C. & Strasheim, C. (1998). Moderne
Statistiek vir die Praktyk (6th ed.). Pretoria: Van Schaik Publishers.
Stein, L. K. & Boyer, K. M. (1994). Progress in the Prevention of Hearing Loss in
Infants. Ear and Hearing, 15, pp. 116 -125.
193
Stein, L.K., Ozdamar, O. & Schabel, M. (1981). Auditory Brainstem Responses
(ABR) with Suspected Deaf-Blind Children. Ear and Hearing, 2(1), pp. 30-40.
Strydom, H. (1998). Ethical aspects of research in the social sciences and
human service professions. In: Research at Grass Roots, (2nd ed). Pretoria: Van
Schaik Publishers.
Stuart, A. Jones, S.M., Walker, L.J. (2006). Insights into elevated distortion
product otoacoustic emissions in sickle cell disease: Comparisons of
hydroxyurea-treated and non-treated young children. Hearing Research, 212, pp.
83-89.
Sokolov, Y., Kurtz, I., Sokolova, O., Steinman, A., Tedesco & P., Broome, R.
(2007). Freedom From Sedation: A New Technology for ABRs. The Hearing
Review, 14(4), pp. 46-55.
South African Pocket Oxford Dictionary. (2002). (3rd ed) edited by Soanes, C.
Cape Town: Oxford University Press.
Surya, M.R.J., Harker, H., Begent, J. & Chong, W.K. (2005). The management of
infants and children for painless imaging. Clinical Radiology, 60, pp. 731-741.
Swanepoel, D. & Störbeck, C. (2008). EHDI Africa:Advocating for infants with
hearing loss in Africa. International Journal of Audiology, 47(suppl. 1), pp.1-2.
Theunissen, M. & Swanepoel, D. (2008). Early Hearing Detection and
Intervention Services in the Public Health Sector in South Africa. International
Journal of Audiology, 47(Suppl. 1), pp. 23-29.
194
Topolska, M.M., Hassmann-Poznanska, E. & Solowiej, E. (2002). Assessment of
Hearing in Children with Infantile Cerebral Palsy. Comparison of Psychophysical
and electrophysical examination. Otolaryngologia Polska, 56(4), pp.467-474.
United Nation Convention. (1989). Convention on the Rights of the Child.
Retrieved April 7 2009 from
http://www.dcsf.gov.uk/everychildmatters/strategy/strategyandgovernance/uncrc/
unitednationsarticles/uncrcarticles/
Van der Reijden, C.S., Mens, L.H.M. & Snik, F.M. (2006). Frequency-Specific
Objective Audiometry: Tone-evoked Brainstem Responses and Steady-State
Responses to 40Hz and 90Hz Amplitude Modulated Stimuli. International Journal
of Audiology, 45, pp. 40-45.
Venter, A. (2000). Distortion Product Otoacoustic Emissions as a Screening tool
for Children with Cerebral Palsy. Unpublished Thesis. University of Pretoria.
Verhaert, N., Willems, M., Van Kerschaver, E. & Desloovere (2008). Impact of
Early Screening and Treatment on Language Development and Educational
Level: Evaluation of 6 Years of Universal Hearing Screening (ALGO®) in
Flanders, Belgium. International Journal of Pediatric Otorhinolaryngology, 72(5),
pp. 599-608.
Vlaskamp, C. & Cuppen-Fonteine, H. (2007). Reliability of assessing the sensory
perception of children with profound intellectual and multiple disabilities: a case
study. Child: Care, Health and Development, 33(5), pp. 547-551.
Wassmer, E. & Whitehouse, W.P. (2006). Melatonin and sleep in children with
neurodevelopmental disabilities and sleep disorders. Current Paediatrics, 16(2),
132-138.
195
Weber, B. (1994). Auditory Brainstem Response. In: Handbook of Clinical
Audiology, (4th ed). Philadelphia: Lippincott Williams & Wilkins. pp. 376-386.
Wesley, R.W.
& Thompson, G. (1984). Behavioral Audiometry. In: Pediatric
Audiology. San Diego: College – Hill Press, Inc.
Wilson- Jones, M., Morgan, E. & Shelton, J.E. (2007). Cerebral Palsy:
Introduction and Diagnosis (Part I). Journal of Pediatric Health Care, 21, pp. 146152.
Willoughby, R.E. Jr., Nelson, K.B. (2002). Chorioamnionitis and brain injury.
Clinical Perinatology, 29(4), pp.603-621
Winskel, H. (2006). The Effects of an Early History of Otitis Media on Children’s
Language and Literacy Skill Development. British Journal of Educational
Psychology, 76, pp. 727–744.
Winter, S., Autry, A., Boyle C., & Yeargin-Allsopp, M. (2002). Trends in the
Prevalence of Cerebral Palsy in a Population-Based Study. Pediatrics, 110, pp.
1220-1225.
Workinger, M.S. (2005). Cerebral Palsy: Resource guide for speech – language
pathologists. New York: Thomson Delmar Learning.
Worhington, D.W. & Peters, J.R. (1984). Electrophysiologic Audiometry. In:
Pediatric Audiology, San Diego: College-Hill Press. pp. 95-124.
Yang, E.Y., Stuart, A., Mencher, G.T., Mencher, L.S. & Vincer, M.J. (1993).
Auditory Brain Stem Responses to Air-and Bone-Conducted Clicks in the
Audiological Assessment of At-Risk Infants. Ear and Hearing, 14(3), 175-180.
196
Yantis, P.A. (1994). Pure tone air-conduction threshold testing. In: Handbook of
Clinical Audiology, (4th ed). Philadelphia: Lippincott Williams & Wilkins. pp. 97108.
Yoshinaga-Itano, C., Sedey, A.L., Coulter, D.K. & Mehl, A.L. (1998). Language of
early-and Later-identified Children with Hearing Loss. Pediatrics, 102(5), pp.
1161-1181.
Young, C. V. (1994). Developmental Disabilities. In: Handbook of Clinical
Audiology (4th ed). Philadelphia: Lippencott: Williams & Wilkins. pp. 521-533.
Zafeiriou, D.I., Andreou, A., & Karasavidou, K. (2000). Utility of Brainstem
Auditory Evoked Potenstials in Children with Spastic Cerebral Palsy. Acta
Paediatrica, 89, pp.194-197.
197
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement