J Am Acad Audiol 9 : 165-171 (1998)
Long-Term versus Short-Term
Hearing Aid Benefit
Rauna K. Surr*
Mary T. Cord*
Brian E. Walden*
Abstract
This study compared hearing aid benefit obtained 6 weeks and a minimum of 1 year after
fitting to determine if changes occurred over time . Fifteen individuals with mild-to-moderate
sensorineural hearing losses, who were successful users of linear amplification, were fitted
binaurally with the Resound BT2 Personal Hearing System . These hearing aids are programmable in two frequency bands that provide wide dynamic range compression (WDRC)
amplification. The manufacturer's recommended loudness growth in octave bands (LGOB)
and audiogram programming algorithm and fitting procedures were used . Following an initial 6-week period and again following a minimum of 1 year of use, the Profile of Hearing Aid
Benefit (PHAB) was administered . Similarly, speech recognition performance was tested
using the Connected Speech Test (CST) in a six-talker speech babble at 50 dBA, +10 signalto-noise (S/N); 60 dBA, +5 SNR; and 70 dBA, +2 SNR ; and in quiet with a reverberation time
of 0.78 seconds. Significant aided benefit was shown. These short-term benefit scores for
the PHAB and CST were compared with those obtained after 1 year of full-time use. Results
revealed no significant change in hearing aid benefit with long-term use, suggesting that a
6-week acclimatization period is sufficiently long for clinical trials of this type of WDRC amplification .
Key Words :
Hearing aid benefit, long term, short term, wide dynamic range compression
Abbreviations : cd = critical difference, CST = Connected Speech Test, FDA = Food and
Drug Administration, LGOB = loudness growth in octave bands, PHAB = Profile of Hearing
Aid Benefit (BN = scale and subscale, speech in background noise; ES = scale, environmental sounds ; QT = scale, speech in quiet; RC = scale, speech with reduced cues), PHAP
= Profile of Hearing Aid Performance, PHS = Personal Hearing System, VC = volume control, WDRC = wide dynamic range compression
everal studies of auditory acclimatization have suggested that hearing aid
benefit increases over time (e .g ., Cox and
Alexander, 1992 ; Gatehouse, 1992, 1993),
whereas others have contradicted this finding
(e.g ., Bentler et al, 1993a, b; Humes et al, 1996).
In a recent review of research, Turner et al
(1996) discuss the many procedural variables
that may confound the assessment of acclimatization. Overall, research that shows an increase
in hearing aid benefit over time suggests that
achievement of optimal use of new auditory cues
S
*Audiology & Speech Center, Walter Reed Army
Medical Center, Washington, DC
Reprint requests : Rauna K . Surr, Army Audiology
& Speech Center, Walter Reed Army Medical Center,
Washington, DC 20307-5001
from amplification occurs within the first few
weeks of hearing aid use. Consequently, the
Food and Drug Administration (FDA) has recommended a 4- to 6-week acclimatization period
for clinical trials to support user benefit claims
(FDA, 1994). However, newer hearing aid circuits
with complex nonlinear processing, such as wide
dynamic range compression (WDRC), alter the
speech signal more drastically than frequency
shaping in traditional linear circuits (Yund and
Buckles, 1995). As a result, the time course for
acclimatization may be longer than suggested by
studies that have involved primarily linear
amplification.
This is a follow-up study of 15 of 40 subjects
who participated in a manufacturer-sponsored
clinical trial of the ReSound BT2 Personal Hearing System (PHS) in 1995 (Walden et al, 1998).
All had previous experience with linear am165
Journal of the American Academy of Audiology/Volume 9, Number 3, June 1998
plification . The BT2 employs compression amplification over a wide range of input levels (45-85
dB SPL) with individually programmable
crossover frequency, gain, and compression ratio
(1 :1-3 :1) for low- and high-frequency bands
(Moore et al, 1992). In general, the results of
Walden et al (1998) revealed significant mean
hearing aid benefit. In addition, individual data
showed significant aided benefit for most subjects using both objective speech recognition
measures and subjective field ratings. Further,
aided Profile of Hearing Aid Benefit (PHAB)
ratings for the WDRC instruments were significantly better in comparison to those for the
linear instruments. Connected Speech Test (CST)
measures were not obtained for the linear instruments . Approximately 1 year later, the 28 subjects who had elected to purchase the BT2 PHS
following the original study were contacted by
phone. The 15 subjects who reported full-time
use of the BT2 PHS and were available to participate in this follow-up study were enrolled as
subjects .
SUBJECTS
S
ubjects who participated in the initial 1995
clinical trial of the Resound BT2 (1995) and
who chose to purchase the instruments at the
conclusion of the trial were recruited for followup (Walden et al, 1998). They were experienced
users of linear amplification (a mirdmum of 1
year) prior to the initial BT2 trial. Only subjects
who reported full-time, binaural use of the BT2
were included in this study. Another requirement
for participation was that hearing threshold
levels were within 10 dB of those obtained at the
time of the initial trial and tympanometric and
acoustic reflex data remained within the normal
range. Additionally, the program parameters of
the BT2 were unchanged and the results of electroacoustic analyses of the instruments were
comparable to those obtained at the time of the
initial clinical trial. The subjects used their own
BT2 instruments for this study. Fifteen subjects who met these criteria were included in this
follow-up study.
The mean age of the 15 subjects was 67 .3
years (range : 55-75) . The average lifetime hearing aid experience was 12 .2 years (range : 3-12).
The 1995 and 1997 audiometric data are summarized in Figure 1 . On average, the subjects
had bilaterally symmetric (not shown) moderateto-severe, gradually sloping, sensorineural hearing loss with relatively good word recognition
ability. The average length of time that the
166
0r
O 1995 (n = 30 eors)
" 1997 (n = 30 eors)
20
so
100
Ni-6
1995: MEAN = $2.8% (SO = 3.7)
1997: MEAN 80.1% (SO 7.5)
0.5
2.0
1 .0
FREQUENCY (kHz)
4.0
8.0
Figure 1 Mean audiometric data obtained in 1995 and
1997 for both ears of the 15 subjects . The error bars
show 1 standard deviation.
subjects had used the BT2 was 19 .6 months
(range : 16-22) .
TEST PROCEDURES
T
he manufacturer's recommended loudness
growth in octave bands (LGOB) and audiogram programming algorithm and adjustment
procedures were used to fit the BT2 before the
original clinical trial. These are described in
detail in an earlier report (Walden et al, 1998)
and are summarized briefly here . The algorithm
incorporates results from loudness growth and
pure-tone threshold measures . The crossover
frequency and gain parameters were initially
verified, and adjusted if necessary, based on the
subject's loudness judgments of running speech
presented at soft, conversational, and loud levels. Further "finetuning" adjustments were made
after a 2-week trial period based on the subjects'
experiences with the fitting, including an assessment of their use of a remote volume control
(VC) . Following this, an additional 4-week
acclimatization period, without the VC, was
provided during the original clinical trial before
the administration of the initial test measures .
The dependent measures of the present
follow-up study were the same as those used in
the initial clinical trial in 1995 (Walden, 1997 ;
Walden et al, 1998) and consisted of the PHAB
(Cox and Gilmore, 1990 ; Cox and Rivera, 1992)
and the CST (Cox et al, 1987, 1988). The Walter Reed protocol and rationale for the test materials and methods have been described in detail
elsewhere (Walden, 1997) and will be discussed
only briefly here .
The PHAB is a 66-item questionnaire that
asks subjects to rate their hearing performance
in a variety of everyday listening situations . A
pencil-and-paper format and a 7-point response
Hearing Aid Benefit/Surr et al
scale with assigned percentages (always, 99%;
almost always, 87%; generally, 75%; half-thetime, 50%; occasionally, 25%; seldom, 12%; never,
1%) were used . The subjects rated their performance twice, aided and unaided. The difference
between the aided and unaided rating is the
benefit score .
Items are grouped into four scales representing specific types of listening situations .
The first scale (QT) describes relatively quiet listening situations and the second scale (RC)
assesses performance in listening situations
with reduced cues . The third scale (BN) consists of items dealing with speech understanding in background noise. The fourth scale (ES)
includes statements involving environmental
sounds such as warning sounds, music, voice
quality, and naturalness of environmental
sounds .
The CST consists of 48 passages produced
by a female talker at a natural rate of articulation . There are 10 sentences per passage and 25
key words per passage are scored . A six-talker
babble is recorded on a second track. One practice passage and six test passages (150 test
items) were administered aided and unaided
via one speaker at 0 degrees azimuth in four listening conditions : (1) relative quiet, 50 dBA,
+10 S/N; (2) reverberation, 60 dBA, 0.78-seconds reverberation time ; (3) moderate noise, 60
dBA, +5 SIN; and (4) high noise, 70 dBA, +2 dB
SIN. The conditions were randomized into eight
test sequences, which were presented in a counterbalanced order. The administration of the
CST was partially automated . The passages
and the associated babble were digitized and presented to subjects on line . Computer software
automatically determined the randomization of
the eight test conditions for the subject and the
attenuator settings appropriate to the test condition . Test sentences were presented one at a
time and the audiologist scored subject responses
live . Although the subjects had been given a VC
at the conclusion of the initial trial for use in
daily living, they did not use it during the testing. Incidentally, at the time of the follow-up,
most subjects did not have the VC with them and
reported satisfaction with the adaptive gain
function of the instruments.
In this follow-up study, a routine audiometric re-evaluation, verification of the BT2
program parameters, and electroacoustic analyses of the instruments were carried out first.
Next, the PHAB was administered followed by
the CST. Testing was accomplished in one 3hour session.
m
60
K
40
0
w
0
20
z
0
D
0
w
-20
-40
-60
PHAB SCALE
Figure 2 Mean benefit (aided minus unaided success)
PHAB scale scores obtained on the initial (1995) and
the follow-up (1997) evaluations for the 15 subjects . The
error bars show 1 standard deviation.
RESULTS AND DISCUSSION
T
he primary goal of this study was to assess
whether auditory acclimatization occurs
beyond the initial 6 weeks of experience with
hearing aids that employ WDRC . Comparisons
of mean short-term and long-term benefit scores
(BT2 aided minus unaided) for the PHAB scales
are shown in Figure 2. Error bars in this and the
next figure show +1 standard deviation . The
results oftwo-tailed paired t-tests (p > .05) indicated no significant difference in benefit . The
same comparison for the CST scores is presented in Figure 3. CST scores are reported in
rationalized arcsine units (raus; Studebaker,
1985). For these data as well, the paired t-tests
revealed no statistically significant differences
100
80
60
40
20
0
Figure 3 Mean CST benefit scores (aided minus
unaided) for the four test conditions obtained on the initial (1995) and the follow-up (1997) evaluations. The
error bars show 1 standard deviation.
Journal of the American Academy of Audiology/Volume 9, Number 3, June 1998
Table 1 Mean Differences between the 1997 and
1995 Benefit Scores for the PHAB and the CST Data
PHAB Scale
Mean Difference (91o)
QT
1 .7
RC
CST Condition
(-4 .5-7 .9)
50
60
60
70
-4 .8 (-11 .3-1 .8)
BN
ES
0 .2 (-5 .5-5 .9)
2 .2 (-5 .8-10 .2)
dBA,
dBA,
dBA,
dBA,
+10 S/N
Reverberation
+5 S/N
+2 S/N
Mean Difference (rau)
1 .7 (-6 .8-10 .2)
-0 .7 (-10 .1-8 .6)
-0 .8 (-7 .7-6 .1)
1 .8 (-6 .2-9 .8)
The 95 percent confidence intervals are shown in the parentheses,
between the short-term and long-term benefit
scores (p > .05) .
The transformation of the CST scores to
rau is done to minimize possible ceiling and
floor effects. However, a comparable transformation of the PHAB ratings is not routinely performed . Because the aided PHAB ratings were
relatively high, they may have been susceptible to ceiling effects. Therefore, the PHAB data
for eac h subject were expressed in terms of relative benefit according to the following formula :
Relative benefit
Aided rating-Unaided rating
100-Unaided rating
50
k
U_
m
10
W
-10
50
PHAB SCALE OT
30
F
e
. -
10
` --
-10
-30
-50
5o
k
-30
-50
INDIVIDUAL SUBJECT (n = 15)
INDIVIDUAL SUBJECT (n =15)
PHAB SCALE BN
30
959 cd
959 cd
1" -30
-50
PHAB SCALE RC
30
.]
,
The paired t-tests for these transformed data did
not reveal a significant difference in benefit
between the 1995 and 1997 data for any of the
four scales .
Recall that the number of subjects for this
study was limited by the size of the original
pool of subjects . Only 15 of the original 40 subjects met the selection criteria and were available to participate . This relatively small sample
may have limited the power of the statistical
tests to detect significant differences between the
short-term and long-term CST and PHAB data .
Assuming that differences in benefit scores of 10
percent or greater are clinically important on
INDIVIDUAL SUBJECT (n =15)
Figure 4 Difference in long-term and short-term benefit scores (1997-1995) for the four PHAB scales for each subject . The horizontal lines on the graph of the BN scale/subscale data show the 95 percent critical difference (cd) value.
168
Hearing Aid Benefit/Surr et al
50
Q
4
30
tZ
w
z 10
w
m
In -10
I
50
50 dBA/+10 CONDITION
I
h
m
I
-30
-50
w
LU
30
10
-10
- 1111 - 11
-30
-50
INDIVIDUAL SUBJECT (n = 15)
60 dBA/REV CONDITION
INDIVIDUAL SUBJECT to 15)
INDIVIDUAL SUBJECT
50
50
60 dBA/+5 CONDITION
a
30
10
m
L_
IN _11L _
A
z
w
30
10
-10
m
~ -10
-30
-30
-50
Figure 5
subject .
INDIVIDUAL SUBJECT (n =15)
-50
INDIVIDUAL SUBJECT (n =15)
Difference in long-term and short-term benefit scores (1997-1995) for the four CST test conditions for each
both the PHAB and the CST, the probability
that the true differences between short-term
and long-term benefit scores on these two measures exceed 10 percent can be assessed by computing 95 percent confidence intervals for the
observed differences (Goodman and Berlin,
1994). These data are summarized in Table 1 and
suggest that there is a low probability (<_ 0.05%a)
that long-term benefit exceeds short-term benefit by a clinically significant amount . This is also
consistent with the individual subject data presented in Figures 4 and 5.
Figure 4 shows the differences for each of the
PHAB scales between follow-up (1997) and the
initial (1995) benefit scores for each subject. A
positive value indicates an increase and a negative value a decrease in benefit with long-term
use. Examination of these graphs reveals no
tendency for hearing aid benefit to change toward
one direction with long-term use, with the exception of the RC scale. For this scale, the long-term
ratings for most subjects tended to indicate less
benefit in comparison to the initial ratings. The
horizontal lines for the BN scale (which is the
same as the BN subscale) show the 95 percent
critical differences (cds) reported by Cox and
Rivera (1992) for PHAB subscales . It is apparent that the difference scores did not approach
significance for any of the subjects . The cd values for benefit scores for the other full scales have
not been reported to our knowledge, but they can
be expected to be higher than those reported for
Profile of Hearing Aid Performance (PHAP)
scores-appropriate for comparing two unaided
or two aided ratings (Cox and Rivera, 1992).
These range from 16 .1 percent to 22 .7 percent
for the different scales . Thus, the differences in
PHAB ratings obtained in this study after 6
weeks and 1 year of hearing aid use generally
appear attributable to measurement error, rather
than to a true change in benefit over time .
Figure 5 shows the differences between follow-up and initial benefit scores for the four
CST test conditions for each subject. Again, no
trend for either greater or lesser long-term benefit is evident. Ninety-five percent cd values for
CST benefit scores have not been reported . The
95 percent cd value for comparing two performance scores (e .g ., aided vs unaided, two aided
scores, etc.) for the 150-item test is 12 .2 rau
169
Journal of the American Academy of Audiology/Volume 9, Number 3, June 1998
50
SUBJECT #5
1
30
LV
m
10
~ -10
0
IL
g', -30
-50
OT
RC
BN
PHAB SCALE
50
4
D:
SUBJECT #14
30
~
10
m
-10
I
-30
-50
OT
RC
BN
ES
PHAB SCALE
50
SUBJECT #10
30
SUBJECT #12
30
H
U_
rn
K
ES
F
w
m
50
10
-10
-30
-50
1
50 dBA
+10 S/N
U_
m
I
60 dBA
REV.
60 dBA
+5 S/N
70 dBA
+2 S/N
CST CONDITION
10
-10
-30
-50
50 dBA
SIN
10
60 dBA
.
REV
V.
65
S/N
CST CONDITION
70
+2 S/N
Figure 6 The difference scores (1997-1995) for two subjects across the PHAB scales and for two subjects across the
CST conditions .
(Cox et al, 1989) . When comparing two CST
benefit scores, the measurement error is compounded because each score is derived from two
scores similarly to the differences in critical difference values between the PHAB and the PHAP
(Cox and Rivera, 1992). Therefore, it is evident
that the difference in short-term versus longterm benefit is likely to be significant for only a
few subjects . This conclusion was supported by
examination of difference scores for individual
subjects across PHAB scales and across CST
conditions . Figure 6 illustrates these findings for
four subjects who were selected based on a high
difference value on at least one scale or test
condition . Generally, a high difference score
occurred on only one PHAB scale or one CST test
condition for any subject. Inconsistencies in the
direction of change are also evident . Further, no
specific PHAB scale or CST test condition tended
to show large differences . Thus, again we conclude that differences in benefit scores may
reflect measurement error rather than a true
change in benefit.
170
SUMMARY AND CONCLUSIONS
he objective of the study was to assess posT sible differences in short-term versus longterm benefit derived from WDRC amplification
for subjects who were previous users of linear
amplification. It was assumed that the hearing-impaired persons who were accustomed to
using linear amplification would be most likely
to show acclimatization effects when switched
to this unique nonlinear form of signal processing. The short-term measures were obtained
after 6-weeks of use in the context of a manufacturer-sponsored clinical trial . The follow-up
long-term measures were taken about l'2 years
later. Only subjects whose audiometric results
were stable and who reported full-time, binaural BT2 use with unchanged settings were
included in this study. No significant increase in
benefit was shown in group or individual data
within this period . The results suggest that the
FDA-recommended 6-week period of hearing
aid use in clinical trials is sufficiently long for
Hearing Aid Benefit/Surr et al
evaluation of hearing aid benefit from the type
of circuitry, subjects, and dependent measures
used in this study. These results cannot answer
the question whether a shorter than 6-week
trial period would suffice.
Acknowledgment . This work received support from
the ReSound Corporation, Redwood City, CA through a
Cooperative Research and Development Agreement with
the Clinical Investigation Regulatory Office, U.S . Army
Medical Department, Ft . Sam Houston, TX . Additional
support was received from the Department of Clinical
Investigation, Walter Reed Army Medical Center, Washington, DC under Work Unit #2564. The opinions and
assertions presented are the private views ofthe authors
and are not to be construed as official or as necessarily
reflecting the views of the Department of the Army, the
Department of Defense, or ReSound Corporation.
Portions of this paper were presented at the Second
Biennial Hearing Aid Research and Development Conference, September 1997, National Institutes of Health,
Bethesda, MD .
REFERENCES
Bentler RA, Niebuhr DP, Getta JP, Anderson CV (1993a).
Longitudinal study of hearing aid effectiveness. I.
Objective measures . J Speech Hear Res 36 :808-819 .
Bender RA, Niebuhr DP, Getta JP, Anderson CV (1993b).
Longitudinal study of hearing aid effectiveness . II .
Subjective measures . J Speech Hear Res 36 :820-831 .
Cox RM, Gilmore C. (1990) . Development of the Profile
of Hearing Aid Performance (PHAP) . JSpeech Hear Res
33 :343-355 .
Cox RM, Rivera IM . (1992) . Predictability and reliability of hearing aid benefit measured using the PHAB . J
Am Acad Audiol 3:242-254 .
Food and Drug Administration . (1994) . Guidance to
HearingAid Manufacturers for Substantiation of Claims .
Rockville, MD : FDA.
Gatehouse S. (1992) . The time course and magnitude of
perceptual acclimatization to frequency responses : evidence from monaural fitting of hearing aids . JAcoust
Soc Am 92 :1258-1268 .
Gatehouse S . (1993) . Role of perceptual acclimatization
in the selection of frequency responses for hearing aids .
JAm Acad Audiol 4 :296-306 .
Goodman SN, Berlin JA . (1994) . The use of predicted
confidence intervals when planning experiments and the
misuse of power when interpreting results. Ann Intern
Med 121:200-206 .
Humes L, Halling D, Coughlin M. (1996) . Reliability and
stability of various hearing-aid outcome measures in a
group of elderly hearing-aid wearers. J Speech Hear Res
39 :923-935 .
Moore BCJ, Johnson JS, Clark TM, Pluvinage V (1992) .
Evaluation of a dual-channel full dynamic range compression system for people with sensorineural hearing
loss . Ear Hear 13 :349-370 .
Studebaker GA . (1985) . A "rationalized" arcsine transform . J Speech Hear Res 28 :455-462 .
Cox RM, Alexander GC . (1992) . Maturation of hearing
aid benefit: objective and subjective measurements . Ear
Hear 13 :131-141 .
Turner CW Humes LE, Bentler RA, Cox RM . (1996) . A
review of past research in hearing aid benefit over time .
Ear Hear 17 :14S-28S .
Cox RM, Alexander GC, Gilmore C. (1987) . Development
of the connected speech test (CST). Ear Hear
8(Suppl) :119S-126S .
Walden BE . (1997) . Toward a model clinical-trials protocol for substantiating hearing aid user-benefit claims .
Am JAudiol 6:13-24 .
Cox RM, Alexander GC, Gilmore C, Pusakulich KM .
(1988) . Use of the connected speech test (CST) with hearing-impaired listeners. Ear Hear 9:198-207 .
Walden BE, Surr RK, Cord MT. (1998) . Aclinical trial of
the ReSound BT2 Personal Hearing System. Am JAudiol,
in press.
Cox RM, Alexander GC, Gilmore C, Pusakulich KM .
(1989) . The Connected Speech Test Version 3: audiovisual administration . Ear Hear 10 :29-32 .
Yund EW Buckles KM . (1995) . Discrimination of multichannel compressed speech in noise: long-term learning
in hearing-impaired subjects . Ear Hear 16 :417-427 .