Download PDF

How accurate are bedside hearing tests?
D. F. Boatman, D. L. Miglioretti, C. Eberwein, M. Alidoost and S. G. Reich
Neurology 2007;68;1311-1314
DOI: 10.1212/01.wnl.0000259524.08148.16
This information is current as of April 30, 2007
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://www.neurology.org/cgi/content/full/68/16/1311
Neurology is the official journal of AAN Enterprises, Inc. A bi-monthly publication, it has been
published continuously since 1951. Copyright © 2007 by AAN Enterprises, Inc. All rights reserved.
Print ISSN: 0028-3878. Online ISSN: 1526-632X.
Downloaded from www.neurology.org at Johns Hopkins University on April 30, 2007
How accurate are bedside
hearing tests?
D.F. Boatman, PhD, CCC-A; D.L. Miglioretti, PhD; C. Eberwein, MS, CCC-A; M. Alidoost, MS;
and S.G. Reich, MD
Abstract—The accuracy of five bedside hearing tests was evaluated in 107 consecutive adults, using pure-tone audiometry as the standard reference. Bedside tests had poor sensitivity (ⱕ0.60), relatively good specificity (ⱖ0.74), and variable
positive predictive value (0.24 to 1.0) for detecting hearing loss. Sensitivity improved when bedside tests were combined
with case history. The diagnostic utility of bedside tests routinely administered by neurologists to detect hearing loss in
adults requires further study.
NEUROLOGY 2007;68:1311–1314
Hearing loss affects 23% to 40% of individuals aged
older than 65 years,1,2 impacting adversely on physical, cognitive, and social functioning.3,4 To screen for
hearing loss, neurologists traditionally use bedside
tests such as finger rub, whispered speech, watch
tick, tuning forks, and self-report.5 Evidence-based
reviews have questioned the reliability of screening
measures, citing lack of test standardization and validation.6,7 For this study, we sought to determine the
sensitivity, specificity, and positive predictive value
of bedside hearing tests and a self-assessment questionnaire, using pure-tone audiometry as the standard reference.
Methods. We studied 107 consecutive adults, aged 50 to 88
(mean 66) years, over 2 months. Fifty-four were patients attending a movement disorders clinic, and 53 were spouses or family
members (table 1). We excluded subjects with known hearing loss,
stroke, or clinically diagnosed dementia. All participants provided
written informed consent.
Tests included finger rub, whispered speech, watch tick, and
the Rinne and Weber tuning fork tests. The same neurologist
(S.G.R.) performed the testing, with subjects seated in an examination room (mean ambient noise 58 dB SPL). Each ear was
tested while subjects occluded the nontest ear. Finger rub and
watch tick were administered six times each, 6 in from the ear.
Results were abnormal if subjects did not respond to more than
one sound presentation. For whispered speech, two sets of three
words were whispered 2 ft from each ear (soup, pen, apple; soap,
pot, orange).7 More than one error (incorrect, no response) was
considered abnormal.
The Weber and Rinne tests used tuning forks of 128, 256, and
512 Hz.5 For the Weber test, the vibrating tuning fork was placed
midline on the forehead. Normal listeners detect no interaural
loudness differences. Lateralizing responses indicate unilateral
hearing loss. For the Rinne test, the tuning fork was alternately
placed on the mastoid and at the ear. Normal listeners and individuals with sensorineural hearing loss hear the sound louder at
the ear (positive Rinne test result) because air conduction is more
effective than bone conduction. A negative Rinne test result occurs
when sound is heard louder at the mastoid, consistent with conductive hearing loss.
Subjects completed an eight-item, self-assessment questionnaire about their hearing (table 2). Questions were developed
as a potential screening tool and not to elicit perceived social or
emotional handicap secondary to hearing loss, as in other questionnaires including the Hearing Handicap Inventory for the
Elderly.8
An audiologist (C.E.), blinded to the bedside test results, performed pure-tone audiometry the same day. Hearing thresholds
were established for each ear at four octave frequencies (500 to
4,000 Hz) under headphones using a two-channel audiometer.
Hearing loss was defined as thresholds ⬎ 25 dB at one or more
frequency in either ear and classified as mild (26 to 40 dB), moderate (41 to 55 dB), or severe (⬎55 dB).9 Three-frequency puretone averages (500 to 2,000 Hz) were computed. Otoscopy and
tympanometry were performed to identify external or middle ear
abnormalities.
Statistical analyses. Rate of hearing loss was compared by
sex and patient status using ␹2 tests. Test sensitivity and specificity were estimated using logistic regression fitted with generalized
estimating equations assuming an exchangeable correlation structure to account for possible within-subject correlations between
ear measurements.10 Sensitivity was not calculated for the Rinne
test because audiometry did not differentiate conductive vs sensorineural hearing loss. Positive predictive values were determined from test sensitivity and specificity and published
estimates of the minimum prevalence of hearing loss in the
ages tested: 45 to 64 years (13.7%), 65 to 74 years (22.9%), and
⬎74 years (31.9%).1 Positive predictive values were not calculated for the Weber test because unilateral hearing loss rates in
these ages are unknown.
Results. Pure-tone audiometry identified 52 of 107 subjects (48%) with either unilateral (10) or bilateral (42)
hearing loss. Of these, 37 (71%) had only high-frequency
hearing loss (⬎1,000 Hz) and 34 (65%) had moderate or
greater hearing loss. More men (67%) than women (31%)
had hearing loss (p ⫽ 0.0002), likely reflecting the older
ages of our male subjects. Rate of hearing loss did not
differ for patients vs nonpatients (p ⫽ 0.29). Twenty-six
subjects (24%) had pure-tone averages ⬎ 25 dB in their
From the Departments of Neurology and Otolaryngology, Johns Hopkins School of Medicine, Baltimore, MD (D.F.B., C.E., M.A.); Center for Health Studies,
Group Health Cooperative, Seattle, WA (D.L.M.); and Department of Neurology, University of Maryland, Baltimore, MD (S.G.R.).
Supported by NIH grants R01-DC005645 and R21-DC007490.
Disclosure: The authors report no conflicts of interest.
Received March 21, 2006. Accepted in final form December 19, 2006.
Address correspondence and reprint requests to Dr. Dana Boatman, Department of Neurology, 600 North Wolfe Street/Meyer 2-147, Baltimore, MD 21287;
e-mail: [email protected]
Copyright © 2007 by AAN Enterprises, Inc.
Downloaded from www.neurology.org at Johns Hopkins University on April 30, 2007
1311
Table 1 Subject demographics
Participants
Number
Age, years
Male
52
51 to 88 (mean 67)
Female
55
50 to 81 (mean 65)
Patient
54
50 to 88 (mean 66)
Nonpatient
53
51 to 88 (mean 66)
50 to 64 years
40
NA
65 to 74 years
52
NA
⬎74 years
15
NA
Sex
Status
Age
NA ⫽ not applicable.
better ear. Of the 214 ears tested, 135 (63%) had hearing
loss at 4,000 Hz. Otoscopy was normal in 87 subjects
(81%); tympanometry was normal (0 ⫾ 1,000 daPa) in 89
subjects (83%).
Test sensitivity for detecting hearing loss ranged from
0.05 to 0.60 (table 3). Watch tick had the highest sensitivity: 0.44 for hearing loss ⬎ 25 dB and 0.60 for hearing loss
⬎ 40 dB. The Weber test for unilateral hearing loss had
the lowest sensitivity (0.05 to 0.30). Combining tests improved sensitivity slightly (0.64). In contrast, bedside tests
showed good specificity (0.74 to 1.0) with positive predictive values varying from 0.24 to 1.0.
Subjects’ assessment of their hearing abilities by questionnaire had poor sensitivity (0.01 to 0.51) for detecting
hearing loss (table 2). Combining questions about hearing
in background noise (Question 7, 0.51) and the perceptions
of others (Question 3, 0.30) with non–tuning fork tests
yielded the highest sensitivity: 0.80 for hearing loss ⬎ 25
dB and 0.87 for hearing loss ⬎ 40 dB. However, this combination had lower specificity (0.53) than individual tests
(0.74 to 1.0). Reported difficulty hearing women’s or children’s voices (Questions 4 and 6) and perceived benefits of
hearing aids (Question 8) had good specificity and positive
predictive value (1.0).
Discussion. Our results demonstrate that bedside
hearing tests have poor sensitivity for detecting
hearing loss in older adults, yielding a high rate of
false negatives. Test sensitivity was greatest (0.64)
when bedside tests were combined, although still below acceptable levels (i.e., ⱖ0.80). Conversely, bedside tests have good specificity, suggesting that
false positives are not a limitation. Using audiometric pure-tone averages for comparison, the
prevalence of hearing loss in our subjects is consistent with epidemiologic studies.1,2 The most common hearing loss identified was moderate or
greater bilateral, and high frequency. Individuals
with this pattern of hearing loss typically have
communication difficulties, especially in noisy environments, 3,4 and are candidates for hearing
aids.4,5,9 Given the high prevalence of hearing loss
in older adults, referral for audiologic testing is
recommended if hearing loss is suspected, even if
bedside tests are normal.
A limitation of this study is the inclusion of only
one examiner. Although this allowed us to control
for interexaminer variability,6,7 our results cannot
be generalized across examiners. Other neurologists performing the same bedside hearing tests
might obtain different levels of test accuracy due
to variations in the quality and loudness of their
voices, finger rubs, or watch ticks. Nonetheless, these results are relevant to neurologists
who learn to administer bedside hearing tests
during clinical training, using recommended
procedures from standard textbooks, and who administer these tests as part of a neurologic
examination.
Observed intertest differences likely reflect
acoustic properties of the sounds used. Watch
ticks are brief, low-intensity, clicklike sounds
covering a wide frequency range, increasing sensitivity to hearing loss. Conversely, whispered
speech has attenuated low-frequency components,
due to absence of vocal cord vibration (voicing),
Table 2 Self-assessment questionnaire
Sensitivity
(95% CI)
Specificity
(95% CI)
Positive
predictive value
1. Do you think you have difficulty hearing?
0.27 (0.19–0.37)
0.89 (0.66–0.97)
0.29
2. Do you hear better in one ear than the other?
0.23 (1.6–0.32)
0.92 (0.72–0.98)
0.30
3. Do family members or friends think you have
difficulty hearing?
0.30 (0.21–0.40)
0.74 (0.50–0.89)
0.15
4. Do you have difficulty hearing women’s voices?
0.05 (0.02–0.11)
1.0
1.0
5. Do you have difficulty hearing men’s voices?
0.01 (0.00–0.08)
0.95 (0.71–0.99)
0.03
6. Do you have difficulty hearing children’s voices?
0.02 (0.01–0.09)
1.0
1.0
7. Do you have difficulty hearing in noisy environments,
such as a restaurant or a party?
0.51 (0.41–0.61)
0.79 (0.55–0.92)
0.28
8. Do you think you would benefit from hearing aids?
0.19 (0.12–0.29)
1.0
1.0
Question
1312
NEUROLOGY 68
April 17, 2007
Downloaded from www.neurology.org at Johns Hopkins University on April 30, 2007
Table 3 Diagnostic accuracy of five bedside tests for detecting hearing loss
Bedside hearing test
Sensitivity (95% CI)
Specificity (95% CI)
Positive predictive value
HL ⬎ 25 dB
0.27 (0.20–0.36)
0.98 (0.85–1.00)
0.68
HL ⬎ 40 dB
0.35 (0.26–0.46)
0.97 (0.90–0.99)
0.62
HL ⬎ 25 dB
0.44 (0.35–0.53)
1.0
1.0
HL ⬎ 40 dB
0.60 (0.50–0.69)
0.99 (0.92–1.00)
0.89
HL ⬎ 25 dB
0.40 (0.32–0.49)
0.83 (0.68–0.91)
0.24
HL ⬎ 40 dB
0.46 (0.36–0.56)
0.78 (0.68–0.86)
0.25
HL ⬎ 25 dB
NA
1.0
1.0
HL ⬎ 40 dB
NA
1.0
1.0
HL ⬎ 25 dB
NA
1.0
1.0
HL ⬎ 40 dB
NA
1.0
1.0
HL ⬎ 25 dB
NA
1.0
1.0
HL ⬎ 40 dB
NA
1.0
1.0
HL ⬎ 25 dB
0.20 (0.05–0.54)
0.79 (0.70–0.86)
NA
HL ⬎ 40 dB
0.21 (0.08–0.45)
0.80 (0.70–0.87)
NA
HL ⬎ 25 dB
0.30 (0.10–0.62)
0.74 (0.65–0.82)
NA
HL ⬎ 40 dB
0.26 (0.11–0.50)
0.74 (0.64–0.82)
NA
HL ⬎ 25 dB
0.30 (0.10–0.62)
0.87 (0.78–0.92)
NA
HL ⬎ 40 dB
0.05 (0.007–0.29)
0.83 (0.74–0.89)
NA
Finger rub
Watch tick
Whispered speech
Rinne test
128 Hz
256 Hz
512 Hz
Weber test
128 Hz
256 Hz
512 Hz
HL ⫽ hearing loss; NA ⫽ not applicable.
better suited for detecting low-frequency hearing
loss.
The poor diagnostic accuracy of tuning fork tests
is consistent with previous reports.6,7 The Rinne and
Weber tests are often used incorrectly to screen for
any type of hearing loss, whereas they were designed
to identify low-frequency (ⱕ512 Hz) unilateral (Weber) or conductive (Rinne) hearing loss. These tests
would likely have missed subjects with highfrequency, bilateral, and presumably sensorineural
hearing loss, based on normal tympanometry and
otoscopy.
Case history and self-assessment are poor
screening tools.2,7 Sensitivity increased by combining questions about difficulty hearing in noise and
concerns raised by others, together with non–
tuning fork tests. In contrast to recommendations
that older adults be screened for hearing loss by
first asking whether they have difficulty hearing
and then administering one bedside test (whispered speech),7 our results suggest that hearing
loss is more likely to be identified if multiple bedside tests are combined with questions about difficulty hearing in noise and the perceptions of
others. Additional studies are needed to compare
newer screening tools, including the audioscope
April 17, 2007 NEUROLOGY 68 1313
Downloaded from www.neurology.org at Johns Hopkins University on April 30, 2007
(pure-tone screener) and otoacoustic emissions (cochlear function test), and to develop more sensitive
bedside hearing tests.
Acknowledgment
The authors thank Ms. Jenna Los for assistance and Dr. William
J. Weiner for helpful comments.
References
1. Cruickshanks KJ, Wiley TL, Tweed TS, et al. Prevalence of hearing loss
in older adults in Beaver Dam, Wisconsin, Am J Epidemiol 1998:148:
879 – 886.
2. Sindhusake D, Mitchell P, Smith W, et al. Validation of self-reported
hearing loss: the Blue Mountains Hearing Study. Int J Epidemiol 2001;
30:1371–1378.
3. Herbst KG, Humphrey C. Hearing impairment and mental state in the
elderly living at home. BMJ 1980;281:903–905.
4. Mulrow CD, Aguilar C, Endicott JE, et al. Quality-of-life changes and
hearing impairment. Ann Intern Med 1990;113:188–194.
5. Nuwer MR, Sigsbee B. The Health Care Financing Administration’s
new examination documentation criteria: minimum auditing standards
for the neurologic examination to be used by Medicare and other
payors: report from the American Academy of Neurology Medical Economics and Management Subcommittee. Neurology 1998;50:497–500.
6. Yueh B, Shapiro N, MacLean CH, Shekelle PG. Screening and management of adult hearing loss in primary care: scientific review. JAMA
2003;289:1976–1985.
7. Bagai A, Thavendiranathan P, Detsky AS. Does this patient have hearing impairment? JAMA 2006;295:416–428.
8. Weinstein BE. Validity of a screening protocol for identifying elderly
people with hearing problems. ASHA 1986;28:41–45.
9. American Speech-Language-Hearing Association. Guidelines for manual pure-tone audiometry. ASHA 1978;20:297–301.
10. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13–22.
RESIDENT AND FELLOW PAGE
Call for teaching videos
The Neurology Resident section is featured online at www.neurology.org. The Editorial Team of this section is seeking
teaching videos that will illustrate classic or uncommon findings on movement disorders. Such videos will aid in the
recognition of such disorders. Instructions for formatting videos can be found in the Information for Authors at
www.neurology.org.
1314
NEUROLOGY 68
April 17, 2007
Downloaded from www.neurology.org at Johns Hopkins University on April 30, 2007
How accurate are bedside hearing tests?
D. F. Boatman, D. L. Miglioretti, C. Eberwein, M. Alidoost and S. G. Reich
Neurology 2007;68;1311-1314
DOI: 10.1212/01.wnl.0000259524.08148.16
This information is current as of April 30, 2007
Updated Information
& Services
including high-resolution figures, can be found at:
http://www.neurology.org/cgi/content/full/68/16/1311
Related Articles
A related article has been published:
http://www.neurology.org/cgi/content/full/68/16/1243
Subspecialty Collections
This article, along with others on similar topics, appears in the
following collection(s):
Audition
http://www.neurology.org/cgi/collection/audition All Clinical
Neurology
http://www.neurology.org/cgi/collection/all_clinical_neurology
Permissions & Licensing
Information about reproducing this article in parts (figures, tables)
or in its entirety can be found online at:
http://www.neurology.org/misc/Permissions.shtml
Reprints
Information about ordering reprints can be found online:
http://www.neurology.org/misc/reprints.shtml
Downloaded from www.neurology.org at Johns Hopkins University on April 30, 2007