Is Automated Auditory Brainstem Response Hearing Screening ...

Maced J Med Sci electronic publication ahead of print,
published on February 15, 2010 as doi:10.3889/MJMS.1857-5773.2010.0085
Brinsmead TL. Sudden Infant Death Syndrome
Macedonian Journal of Medical Sciences. 2010 Mar 15; 3(1):XXX-XXX.
doi:10.3889/MJMS.1857-5773.2010.0085
Clinical Science
OPENACCESS
Is Automated Auditory Brainstem Response Hearing
Screening Predictive of Sudden Infant Death Syndrome?
Tammy Lee Brinsmead, FRACP, Luke Jardine, FRACP, Delene Thomas, RN EM
Grantley Stable Neonatal Unit, Royal Brisbane & Women’s Hospital, Brisbane, Queensland, Australia
Abstract
Citation: Brinsmead TL, Jardine L, Thomas D. Is
Automated Auditory Brainstem Response Hearing
Screening Predictive of Sudden Infant Death
Syndrome? Maced J Med Sci.2010,3(1)XXX-
XXX,doi.10.3889/MJMS.1957-5773.2010.0085.
Key words: Neonatal; infant; hearing screening;
automated auditory brainstem response; AABR;
Sudden Infant Death Syndrome; SIDS.
Correspondence: Dr Tammy Brinsmead
Paediatric Registrar, Royal Children’s Hospital,
Brisbane, Queensland, Australia 4029
E-mail (h): tammybrinsmead@yahoo.com.au
Received: 30-Nov-2009; Revised: 18-Jan-2010;
Accepted: 28-Jan-2010; Online first: 15-Feb-2010
Copyright: © 2010 Brinsmead TL. This is an openaccess
article distributed under the terms of the
Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any
medium, provided the original author and source are
credited.
Objective. To determine the incidence of unilateral deficits on newborn hearing screening in babies
who have died of SIDS.
Methods. Records of universal newborn hearing screening outcomes for the state of Queensland,
from October 1, 2004 to December 31, 2006, were accessed. A list of all sudden, unexplained infant
deaths in the period January 1, 2005 to December 31, 2006 was acquired from the Queensland State
Coroner. For each child whose death was attributed to SIDS, individual hearing screening results were
reviewed. The screening process in Queensland utilises Automated Auditory Brainstem Response
(AABR) equipment.
Results. Between January 1, 2005 and December 31, 2006, 44 deaths in Queensland were attributed
to SIDS. Only 18 of the 44 babies who died with SIDS had had their hearing screened. All 18 passed
the hearing screen. Of the remaining 26 babies who died with SIDS who did not have their hearing
screened: 25 were born in places that did not practice routine screening at the time of their births; one
infant of the list from the coroner could not be located on the lists of births.
Conclusion. In Queensland there is no association with failing a hearing screening test and dying
with SIDS.
Competing Interests: The authors have declared
that no competing interests exist.
Introduction
Sudden Infant Death Syndrome (SIDS) is the
most common cause of postneonatal (age over one month
and less than one year) infant death in Australia [1]. Its
aetiology remains unknown.
The number of deaths attributed to SIDS has
consistently declined since ‘Reduce the Risks‘ campaigns
in Australia in the 1990s [2,3]. These campaigns informed
the public and health care workers of a possible association
Maced J Med Sci. 2010 Mar 15; 3(1):XXX-XXX.
between SIDS and a number of factors, such as infants
sleeping in a prone position, and exposure to smoking
either in utero or at birth. The result was an almost 40%
decrease in the number of deaths reported as SIDS when
pre- and post-campaign periods were compared [2].
In recent decades, studies have attempted to
identify infants at risk of SIDS using various investigations.
These have included electrocardiography,
polysomnography and hypoxia challenge [4]. An
investigation, or group of investigations, reliably predicting
1

Clinical Science
which infants are at risk of dying of SIDS has not been
established [4].
A recent report by Rubens et al., 2008 suggested
an association between right sided unilateral hearing
deficits as detected by newborn transient evoked
otoacoustic emissions and the risk of SIDS [5]. Sininger
et al., 2006 and Berninger, 2007 have previously reported
larger amplitude waves on the right side on auditory
brainstem response (ABR) testing in normal neonates
[6,7]. An increased venous pressure effect on the right
side causing disruption to the inner ear hair cells is a
proposed mechanism for the unilateral difference [5].
Stimulation of the vestibular system is known to cause a
change in the respiratory pattern [8,9]. Vestibular stimulation
can re-establish respiratory effort after apnoea [8-10]. An
abnormality in respiratory control is a proposed mechanism
for SIDS [5]. Inner ear insults can disrupt cochlear and
vestibular function and it is possible that such an injury
may occur in the perinatal period [5].
The transient evoked otoacoustic emission
technique described by Rubens et al [5] uses a small
probe placed in the baby’s ear and the administration of a
click stimulus at four frequencies in each ear (1500, 2000,
3000 & 4000 decibels [dB]). A signal/noise ratio (SNR) is
generated for each ear at each frequency. The SNR is the
difference between the hearing measurement in decibels
and the level of background noise. To have a pass then the
following conditions need to be met:
2
1. SNR ≥ 4dB at 2,3 & 4 kilohertz;
2. a click stimulus level between 74 dB and 86 dB;
3. ≥ 50 sweeps of stimulus delivery in which the
detected noise floor falls below the software’s
predetermined acceptance/rejection level.
The Royal Brisbane and Women’s Hospital has
undertaken a universal newborn screening program
(Healthy Hearing) since October 2004. This has gradually
been established in other Queensland centres, and since
December 2006 has been available to all birthing facilities
in the state.
Australia is a wealthy country, with generally high
standards of medical care readily available. Its system of
government has evolved from liberal democratic tradition,
with influence from British and North American models
[11]. Based on per capita gross domestic product, Australia
ranks among the top twenty nations in the world [11]. Life
expectancy for males and females is seventy-nine and
eighty-four years respectively [11]. Queensland’s
population, which represents 21% of Australia’s total, is
estimated at 4.47 million [12]. 92% of the Australian
population is Caucasian, Asians consistute 7%, and less
the 1% is made up of Aboriginals (indigenous Australians)
and other ethnic groups [13]. It is well known that health
outcomes are consistently poorer in the indigenous
population. Although Australia is an arid environment [14],
88.6% of the population reside in urban areas [11], which
are typically clustered around the more moderate, coastal
regions.
The aim of this study was to determine the
incidence of unilateral deficits on newborn hearing screening
in babies who have died of SIDS.
Methods
Records of universal newborn hearing screening
outcomes for the state of Queensland, from October 2004
to December 2006 inclusive, were accessed. A list of all
sudden, unexplained infant deaths in the period January 1,
2005 to December 31, 2006 inclusive was acquired from
the Queensland State Coroner. For each child whose
death was attributed to SIDS, individual hearing screening
results were reviewed if available. To improve the likelihood
of identifying hearing screening results on each baby,
searches were conducted under the child’s surname at
the time of death, and maternal surname at the time of
birth.
Hearing screening requires detecting the presence
or absence of a response at a preset signal intensity level
for the purpose of identifying which infants need further
evaluation. The screening process in Queensland utilises
Automated Auditory Brainstem Response (AABR)
equipment. The goal of AABR is to detect the presence or
absence of a signal, not estimate its characteristics [5].
The screening equipment used by Queensland Health is
the Natus ALGO 3® and the ALGO 3i® (Natus, San
Carlo). One of the unique features of the ALGO® response
detection method is the use of a binary signal detector.
The equipment automatically records whether the baby
has passed the screen (‘pass result’) or whether a referral
for further screening (‘refer result’) is required. If a baby
receives a ‘refer result’ on either or both ears a second
screen of both ears is conducted at a later time to confirm
the result. A ‘refer result’ does not necessarily mean that
the baby has a hearing loss. Some common reasons for
needing a second screen are: the baby was unsettled at
the time of the first screen; there was background noise
when the test was carried out and/or; the baby had fluid or
a temporary blockage in their ear after birth. If the baby
http://www.mjms.ukim.edu.mk

Brinsmead TL. Sudden Infant Death Syndrome
receives a ‘refer result’ on either or both ears on the second
screen the baby is referred for diagnostic audiology
assessment.
Statistical tests were done using GraphPad Prism
version 3.02 for Windows (GraphPad Software, San Diego
California USA, www.graphpad.com). To test for differences
in proportions the Chi-squared test was used.
Ethical approval was sought and granted by the
Executive Director of Medical Services at the Royal
Brisbane Hospital.
Results
From October 1, 2004 to December 31, 2006
(inclusive), there were 117,029 births in Queensland [15].
Of those, 75,161 births were in hospitals where babies
could have their hearing screened [16]. 74,146 (ie.,
63.36% of births and 97.79% of those born in hospitals
that screen) of these babies actually underwent hearing
screening. 1,015 of the 75,161 were not screened: 362
were offered hearing screening but their parents or guardians
declined, the reason the remaining 653 were not screened
is unknown.
Between January 1, 2005 and December 31,
2006, 44 deaths in Queensland were attributed to SIDS
[17]. Only 18 of the 44 babies who died with SIDS had had
their hearing screened. All 18 passed the hearing screen.
Of the remaining 26 babies who died with SIDS who did not
have their hearing screened: 25 were born in places that
did not practice routine screening at the time of their births
[16]; one infant of the list from the coroner could not be
located on the lists of births.
The two by two table for failing the hearing screen
by whether they died with SIDS or not is shown in Table
1. The difference in proportions is not statistically significant
(Chi-squared test, P value = 0.49).
Table 1: Number of babies who passed hearing screening
vs. failed hearing screening, and number of deaths attributed
to Sudden Infant Death Syndrome.
Two-sided Chi-squared test, P value = 0.4878.
Maced J Med Sci. 2010 Mar 15; 3(1):XXX-XXX.
Discussion
None of the babies who died with SIDS failed to
pass a hearing screening test and no baby that failed to
pass a hearing screen died with SIDS. Therefore, using
Queensland data, we have shown that there is no
association with failing a hearing screening test and dying
with SIDS. We have therefore failed to confirm the findings
of Rubens et al., 2008 [5].
This must be interpreted in context, however. The
number of children who died with SIDS in our study was
very small. In addition, only 41% of these deaths were in
babies who had hearing screening, reducing the power of
our study significantly.
There are important differences between how
Rubens et al screened hearing and our screening method.
Otoacoustic emissions, as used in Rubens et al’s study
[5] are obtained from the ear canal by using a sensitive
microphone within a probe assembly that records cochlear
responses to acoustic stimuli [18]. Thus, otoacoustic
emissions reflect the status of the peripheral auditory
system extending to the cochlear outer hair cells. In
contrast, auditory brainstem response measurements are
obtained from surface electrodes that record neural activity
generated in the cochlea, auditory nerve, and brainstem in
response to acoustic stimuli delivered via an earphone
[18]. Automated auditory brainstem response
measurements reflect the status of the peripheral auditory
system, the eighth nerve, and the brainstem auditory
pathway. Both otoacoustic emission and auditory brainstem
response screening technologies can be used to detect
sensory (cochlear) hearing loss; however, both
technologies may be affected by outer or middle-ear
dysfunction. Consequently, transient conditions of the
outer and middle ear may result in a “failed” screening-test
result in the presence of normal cochlear and/or neural
function. Moreover, because otoacoustic emissions are
generated within the cochlea, otoacoustic emissions
technology cannot be used to detect neural (eighth nerve
or auditory brainstem pathway) dysfunction. Thus, neural
conduction disorders or auditory neuropathy/dyssynchrony
without concomitant sensory dysfunction will not be
detected by otoacoustic emissions testing [18].
It is not uncommon for a newborn to fail hearing
screens of this nature in the neonatal period. In an Italian
study of 32,502 newborns in 13 regional birth centres,
1.05% had an abnormal otoacoustic emissions result
(156 children unilaterally, 179 bilaterally). Of those,
however, 82.37% had a normal auditory brainstem response
result [19]. Similarly, a Swedish study of 14,287 newborns
3

Clinical Science
at two maternity wards noted 1.35% of babies had a
unilateral abnormality on otoacoustic emissions after two
tests [20]. Families participating in neonatal hearing
screening should be prepared for the possibility of false
positives, and should be supported during the potentially
anxious waiting period between screening and more
definitive investigation.
It is encouraging to note the high percentage of
Queensland babies undergoing neonatal hearing screening
tests. One of the limitations of this study, however, is that
not all children identified as dying from SIDS in the time
frame reviewed had hearing screening at birth.
Our list of deaths attributed to SIDS is also likely
to be incomplete. We were unable to account for children
born in Queensland who may have died from SIDS interstate
and internationally.
It is interesting to note that a post hoc analysis of
our data suggests a baby that does not have hearing
screening is more likely to die of SIDS than a baby who has
hearing screening – see Table 2. This association would
need further exploration to draw any conclusions. It would
be useful to establish if this trend has continued now that
screening is universally available in Queensland.
Table 2: Post hoc analysis of hearing screening vs no
hearing screening and deaths attributed to Sudden Infant
Death Syndrome.
vulnerable to a Type I error [22].
SIDS remains a complex and socially relevant
area of study. There are vast opportunities for research,
but the results must be filtered to the public with sensitivity
and a sense of perspective. We would be wise to heed the
advice of Krous and Byard, and Hamill and Lim. Highlighting
the limitations of Rubens’ study, they remind us to protect
families from misinterpreting its implications. First, we
must prevent a sense of false reassurance in families with
normal hearing screening results. An emphasis should
remain on the principles of the Back to Sleep Campaign,
the evidence for which is more reliable [21]. At the same
time, we must prevent families being falsely alarmed if
their baby fails a neonatal hearing screen [22]. The
difference between the hearing results of those in Rubens’
study who did, and did not, die of SIDS are subtle. This
has further implications for anxiety in the context of an
abnormal hearing screening test. In order to detect a
majority (95%) of babies at risk of SIDS, an enormous
number of false positives would occur (specifically, 87% of
normal infants) if Rubens’ predictions are accurate [22].
The fact that we found no association between
failed hearing screening and SIDS, and given that in
Queensland we use auditory brainstem response to screen
hearing, means that Rubens et al’s results have little
bearing or relevance in our population. Further study,
involving a much larger number of subjects (particularly
those dying with SIDS), is necessary to confirm the
validity of this study’s findings.
Two-sided Chi-squared test, P value = 0.0033.
SIDS will always be a difficult area to research. It
probably encompasses a range of causes, and once it
occurs, further study is limited by the loss of the child’s
life. Most observations must be made retrospectively.
Critics caution that Rubens’ study [5] was
retrospective, the number of SIDS cases small, and
cochlear function was used as an unproven representative
of vestibular function [21]. There is also a lack of clarity in
the paper’s definition of SIDS, a common problem in this
area of research [21]. In addition, the paper’s methods are
controversial. Its use of matched controls, as compared
to independent samples testing, makes it much more
Conclusions
This limited study suggests that in Queensland
there is no association between failing a hearing screening
test and dying with SIDS. Of the babies who died of SIDS
between January 1, 2005 and January 2007, hearing
screening was not predictive.
References
1. Australian Bureau of Statistics. SIDS in Australia 1981-
2000 A Statistical Overview. 2003, August. p1. Accessed
online (31/3/08). http://www.sidsandkids.org/researchstatistics.html
2. Tursan d’Espaignet E, Bulsara M, Wolfenden L, Byard
RW, Stanley FJ. Trends in sudden infant death syndrome in
Australia from 1980 to 2002. Forensic Sci Med Pathol.
2008;4(2):83-90. doi:10.1007/s12024-007-9011-y
PMID:19291477.
4
http://www.mjms.ukim.edu.mk

Brinsmead TL. Sudden Infant Death Syndrome
3. 3303.0 Causes of Death, Australia, 2006. pp 35-6.
4. Bentele KH, Albani M. Are there tests predictive for
prolonged apnoea and SIDS? A review of epidemiological
and functional studies. Acta Paediatr Scand Suppl.
1988;342:1-21. doi:10.1111/j.1651-2227.1988.tb10792.x
PMID:3291551.
5. Rubens DD, Vohr BR, Tucker R, O'Neil CA, Chung W.
Newborn oto-acoustic emission hearing screening tests:
preliminary evidence for a marker of susceptibility to SIDS.
Early Hum Dev. 2008;84(4):225-9. doi:10.1016/
j.earlhumdev.2007.06.001 PMID:17614220.
6. Sininger YS, Cone-Wesson B. Lateral asymmetry in the
ABR of neonates: evidence and mechanisms. Hear Res.
2006;212(1-2):203-11. doi:10.1016/j.heares.2005.12.003
PMID:16439078.
7. Berninger E. Characteristics of normal newborn transientevoked
otoacoustic emissions: ear asymmetries and sex
effects. Int J Audiol. 2007;46(11):661-9. doi:10.1080/
14992020701438797 PMID:17978948.
8. Siniaia MS, Miller AD. Vestibular effects on upper airway
musculature. Brain Research. 1996;736(1-2):160-4.
doi:10.1016/0006-8993(96)00674-9 PMID:8930320.
9. Mori RL, Bergsman AE, Holmes MJ, Yates BJ. Role of the
medial medullary reticular formation in relaying vestibular
signals to the diaphragm and abdominal muscles. Brain
Research. 2001;902(1):82-91. doi:10.1016/S0006-
8993(01)02370-8 PMID:11376597.
10. Farrimond T. Sudden infant death syndrome and
possible relation to vestibular function. Percept Mot Skills.
1990;71(2):419-23. PMID:2251079.
11. http://www.wpro.who.int/countries/2009/aus/ - accessed
13 th January, 2010.
12. http://www.oesr.gov.au – accessed 17 th January, 2010.
13. CIA World Factbook. http://www.indexmundi.com/
australia/ethnic_groups.html - accessed 17th January, 2010.
14. http://www.bom.gov.au/lam/climate/levelthree/ausclim/
ausclim.htm - accessed 17th January, 2010.
15. Personal correspondence from the Department of Public
Affairs, Queensland State Office.
16. Queensland Health, Healthy Hearing Program Hospital
Databases. Electronic record of universal newborn hearing
screening outcomes in Queensland, public and private
birthing facilities.
17. Personal correspondence from the Queensland Office
of the State Coroner.
18. Hearing JCoI. Year 2007 Position Statement: Principles
and Guidelines for Early Hearing Detection and Intervention
Programs. Pediatrics. 2007;120:898-921. doi:10.1542/
peds.2007-2333 PMID:17908777.
19. Calevo MG, Mezzano P, Zullino E, Padovani P, Serra G,
STERN Group. Ligurian experience on neonatal hearing
screening: clinical and epidemiological aspects. Acta
Paediatr. 2007;96(11):1592-9. doi:10.1111/j.1651-
2227.2007.00475.x PMID:17937684.
20. Hergils L. Analysis of measurements from the first
Swedish universal neonatal hearing screening program.
Int J Audiol. 2007;46(11):680-5. doi:10.1080/
14992020701459868 PMID:17978950.
21. Krous HF, Byard RW. Newborn hearing screens and
SIDS. Early Hum Dev. 2008;84(6):371. doi:10.1016/
j.earlhumdev.2007.10.003 PMID:17996406.
22. Hamill T, Lim G. Otoacoustic emissions does not
currently have ability to detect SIDS. Early Hum Dev.
2008;84(6):373. doi:10.1016/j.earlhumdev.2007.10.004
PMID:18096330.
Maced J Med Sci. 2010 Mar 15; 3(1):XXX-XXX.
5