Keywords
air-bone gap - chronic otitis media - various sites and sizes of perforation
Introduction
Tympanic membrane (TM) serves several functions such as conduction of sound waves
across the middle ear, protection of the middle ear cleft from infections, and shielding
of the round window from direct sound waves, which is referred to as “round window
baffle.”[1] This shield is necessary to create a phase differential across the oval and round
windows. The effect of the enhanced ratio of the TM surface area to that of the oval
window increases the sound pressure by approximately 27 dB, and the lever action of
ossicles contributes to approximately 3 dB.[2]
[3]
TM perforations can be classified based on their size, site, and etiology. TM perforation
can result from many causes such as trauma, acute otitis media, and chronic otitis
media (COM).[4] Chronic suppurative otitis media has been an important cause of middle ear disease
and is the most important cause of hearing loss (HL) in developing countries such
as India. Poor and overcrowded living conditions, poor hygiene, and poor nutrition
have been suggested as a basis for the widespread prevalence of chronic suppurative
otitis media in developing countries.[5] It is an inflammatory process of the mucoperiosteal lining of the middle ear space
and the mastoid.[6] COM is defined as chronic otorrhea through a nonintact (perforated) TM of greater
than 6 weeks’ duration unresponsive to oral and ototopical medications.[5] COM is also classified into active mucosal, inactive mucosal, active squamous, inactive
squamous, and healed type.[6]
The effects of TM perforations on middle ear sound transmission are not well characterized
largely because ears with perforations typically have additional pathological changes.[5] It has been studied that the larger the perforation, the greater is HL. A total
perforation of TM would lead to a loss in the transformer action of the middle ear.[7] The site of the perforation is believed to have a significant effect on magnitude
of HL.[8] For instance, the posterior quadrant perforations are worse than the anterior ones
because of the direct exposure of the round window to the sound waves, and perforations
at or near the site of TM attachments to the manubrium have more severe effects than
those of comparable size at different sites.[9] However, some studies show that there is no significant effect associated with location
of perforation.[7]
[8] [10] COM can lead to conductive HLs of as much as 60 dB, especially when the ossicles
are involved, which constitutes a serious handicap.[11] Perforations cause a loss that depends on frequency, perforation size, and middle
ear air space volume. Perforation-induced losses are greatest at the lowest frequencies
and generally decrease as frequency increases.[5] Perforation size is an important determinant of the loss; larger perforations result
in larger HLs. The volume of the middle ear air space (combined tympanic cavity and
mastoid air volume) is also an important parameter that determines the amount of HL
caused by a perforation.[12]
The idea that middle ear volume affects HL caused by a perforation has several clinical
implications. It provides an explanation for the variability in HL that is observed
clinically in patients with otherwise similar perforations. Also, it can account for
fluctuations in the HL with the same perforation in a given patient during active
disease when the middle ear volume might be expected to vary because of mucosal edema
or accumulation of fluid within the middle ear spaces. This explains the clinical
observation that TM perforations in patients with canal wall down mastoidectomy cavities
have typically greater AB gaps than similar-sized perforations in normal ears, because
ears with canal wall down mastoidectomy cavities have smaller middle ear volumes.
Mehta et al speculated that the importance of location may have resulted from attempts
to explain differences in HL with similar-sized perforations that were caused by inter-ear
differences in middle ear volumes.[13]
The aim of this study was to correlate the location and size of TM perforation and
middle ear air space volume with the magnitude of HL in patients with tubotympanic
or inactive mucosal type of COM.
Materials and Methods
This cross-sectional study was conducted in the Department of Otolaryngology and Head
& Neck Surgery, Post Graduate Institute of Medical Education and Research, Chandigarh,
India. All consecutive patients with COM (tubotympanic or inactive mucosal type) attending
the Otolaryngology services from July 2010 to December 2011 were taken as study population.
A total of 300 ears of patients with chronic otitis were studied. Informed and written
consent of all the participating patients was taken, and the institute ethics committee
approval was sought before commencing the study.
Inclusion criteria included patients of age between 15 and 45 years with tubotympanic
or inactive mucosal type of COM with only mild to moderate conductive HL, that is,
up to 55 dB HL according to the World Health Organization (WHO) grading of hearing
impairment, having good general physical condition, and not having any evidence of
active infection in nose, throat, and paranasal sinuses.
Patients with active ear disease, attic retraction or perforation, presence of cholesteatoma
or granulations, with sensorineural HL and history of trauma to the TM or prior attempts
of its repair were excluded from the study.
Photo documentation with help of an otoendoscope was performed by recording a digital
image of the perforation next to a scale, as shown in [Figs. 1 ] and [2]. Then, the percentage area of the perforation was calculated using java image software.
The following equation was generated: P/T × 100% = percentage perforation, where P
is the area (in pixels 2) of the TM perforation and T is the total area (in pixels
2) for the entire TM (including the perforation).[14] Then area in square millimeter was calculated taking mean adult normal TM area as
85 mm2.[1]
[15]
Fig. 1 Otomicroscopic picture of right tympanic membrane with central perforation of area
42 mm2 (large size), with air-bone gap of 18 dB HL in low-volume group.
Fig. 2 Otomicroscopic picture of right tympanic membrane of the patient with posteroinferior
of area 1 mm2 (small size) with no hearing loss.
The perforation was divided into three categories depending on the area of TM perforation:
small (0–8 mm2), medium (9–30 mm2), and large (> 30 mm2). Each patient underwent pure-tone threshold audiometry at frequencies 250, 500,
1,000, 2,000, and 4,000 Hz. Air and bone conduction thresholds were determined. Contralateral
ear was always masked while determining the bone conduction thresholds, whereas for
air conduction thresholds, contralateral ear was masked if the test ear was comparatively
the poorer one and if sounds of ≥ 45 dB were used for the tests. The mean HL was calculated
through the pure-tone average taken at 500, 1,000, and 2,000 Hz for each site of perforation.[1]
[2]
[3]
[4]
[5]
[13]
Thereafter, patients underwent tympanometry to measure the volume of middle ear and
mastoid space of ipsilateral ear, which is the combined volume of external auditory
canal, middle ear, and mastoid air space. Then patients underwent tympanometry of
contralateral ear (with intact TM), which is the volume of the external auditory canal.
Subtracting the volume of ipsilateral ear with that of contralateral ear provided
the volume of middle ear and mastoid space of the ipsilateral ear. For patients with
bilateral perforations, the mean of ear canal volume of unilateral perforation was
used. The ears were divided into two subgroups: small (< 4.3 mL) and large (> 4.3
mL).[13]
The comparison was made between
-
Mean air-bone (AB) gap between small- and large-sized perforations, and small- and
medium-sized perforations at different frequencies.
-
Mean AB gap of small- and large-volume groups.
-
Mean AB gap of perforation of different locations (anterosuperior [AS], posterosuperior
[PS], anteroinferior [AI], posteroinferior [PI], and central) but of similar size
and middle ear volume within this study population.
Results
A total of 246 patients (males 131, females 115) ([Tables 1], [2]) were included in the study with 54 bilateral (n = 54) and 192 unilateral (n = 192) cases, with total number of cases (ears) being 300. Maximum cases (n = 124, 41.6%) were of large-sized perforations (> 30 mm2), and minimum cases (n = 60, 20%) were of small-sized perforations group (0–9 mm2), as shown in [Fig. 3]. Maximum cases were of central perforations (n = 198, 66%), and minimum cases were of AS perforations (n = 9, 3%). Large-sized perforations were found to have maximum mean AB gap of 26.43
dB, and small-sized perforations were found to have minimum mean AB gap of 9.12 dB.
Medium-sized perforations had mean AB gap of 16.13 dB. Maximum cases were of low-volume
group (n = 246, 92%). High-volume group had 8% cases. In the remaining 34 cases, volume could
not be calculated as seal could not be achieved during tympanometry, so they were
excluded from the study of correlation of air space volume with HL.
Table 1
Sex distribution of cases in the study
|
Male
|
Female
|
Total
|
|
Number
|
131
|
115
|
246
|
|
Percentage
|
53.2
|
46.7
|
|
Table 2
Age distribution of the study group
|
Age groups (y)
|
15–25
|
25–35
|
35–45
|
|
Number
|
116
|
64
|
66
|
|
Percentage
|
47
|
26
|
26.8
|
Fig. 3 Pie diagram showing distribution of cases into large size (> 30 mm2), medium size (9–30 mm2), and small size (0–9 mm2).
The correlation of mean AB gap with perforation size is shown in [Fig. 4]. The low-volume group had larger mean AB gap (19.96 dB HL) than the high-volume
group (11.80 dB HL). The t-test was applied, and the difference was found to be statistically significant with
a p-value = 0.004. The mean AB gap was highest in central perforation (23.14 dB HL) and
lowest in AS perforation (7.5 dB HL). On applying ANOVA (analysis of variance) test,
this difference was found to be statistically significant with p = 0.00.
Fig. 4 Line diagram showing correlation of mean air-bone (AB) gap with perforation size.
Correlation of Air-Bone Gap with Site of Perforation in Low Volume, Medium Area of
Perforation
In low-volume and medium-area category, the maximum mean AB gap was in central perforation
(19.6 dB) and minimum AB gap was in PS perforations, that is, 11 dB, where it was
13 dB in AI, AS, and PI segments. ANOVA test was applied, which showed that the differences
were statistically significant with p = 0.005.
Correlation of Air-Bone Gap with Site of Perforation in Low Volume, Small Area of
Perforation
In low-volume, small perforation area, the mean AB gap was highest in PS perforation
(15 dB HL) and lowest in PI perforations (0.5 dB HL). It was 9.5 dB HL in central
perforation cases, 8 in AS, and 13 in AI cases. On applying ANOVA test, the difference
was found to be statistically significant with p = 0.001.
Correlation of Air-Bone Gap with Site of Perforation in Low Volume, Large Area Group
In low-volume, large-area category, 106 cases were of central perforations and only
1 case was of PI perforation. No cases of other site of perforation were seen in this
group. Mean AB gap was 26 dB in central perforation cases and 18 in one PI case.
On applying ANOVA test, it was found that the correlation was statistically nonsignificant
with p = 0.499. The groups were incomplete in high-volume low-area, high-volume medium-area,
and high-volume large-area subgroups. Therefore, comparison could not be done in these
subgroups.
Frequency Distribution of Air-Bone Gaps
AB gaps were maximum at lower frequencies and decreased with increase in frequencies
except at 4,000 Hz. It was 56.9 dB HL at 250 Hz, 42.6 at 500 Hz, 41.5 at 1,000 Hz,
32.4 at 2,000 Hz, and 49.5 at 4,000 Hz.
Discussion
The primary mechanism of HL by a TM perforation is a reduction in middle ear sound
transmission caused by a loss in sound pressure difference across the TM. On analyzing
various aspects of perforation, we found that the HL depends on the size, site, and
middle ear air space volume.
Effect of Size of Perforation
All ears with perforation in our study were divided into three groups: (I) small-sized
perforation (0–9 mm2), (II) medium-sized perforation (9–30 mm2), and (III) large-sized perforation (30 mm2). Maximum number of ears were found in group III (n = 124, 41.6%), followed by group II (n = 116, 30%). Minimum number of ears were in group I (n = 60, 20%). On the contrary, Mehta et al[13] reported maximum cases 30 (48.35%) in small-area group and minimum cases 7 (11.2%)
in large-area group, out of the total of 62 ears, which is not in agreement with our
study. As per a study by Pannu et al,[16] 47% patients were categorized to group I, 34% to group II, and remaining 19% to
group III. The mean AB gap in small-, medium-, and large-sized perforations were 9,
16, and 26 dB, respectively. The difference was found to be statistically significant
with p = 0.00. As the size of perforation increased, the AB gap also increased.
Similarly, Ibekwe et al[3] also found a positive correlation of perforation size with the magnitude of HL (p = 0.01). Similarly, Mehta et al[13] reported that the difference in AB gaps in small size (0–9 mm2) and large size (> 30 mm2) and small size and medium size (9–30 mm2) were significant (p < 0.05) at 1,000, 2,000, and 4,000 Hz. They also concluded that AB gap increased
with increase in the size of TM perforations. Pannu et al[16] also found that mean AB gap increased statistically significantly as the perforation
size increased. Nepal et al[17] also found that the hearing is directly proportional to the perforation size. Matsuda
et al[18] in their study also found that the larger the area of a traumatic perforation, the
greater the conduction disturbance.
Mills et al[19] also found a significant relationship between HL and perforation size, with a strong
trend for HL to increase as the perforation size increased (p = 0.01). This correlation is due to decrease in sound pressure difference across
the TM with the increase in size of the TM perforation.
Effect of Middle Ear Air Space Volume
In our study, low-volume group had 92% cases and high-volume group had 8% cases. It
was found that HL decreased as the volume of middle ear air space increased with p = 0.004. Low-volume group had mean AB gap of 19.9 dB, and high-volume group had mean
AB gap of 11.8 dB, which was found to be statistically significant with p = 0.004. Mehta et al[13] also found that the mean AB gaps in the small-volume group were larger by 10 to
20 dB compared with those in the large-volume group. These differences were statistically
significant (p < 0.05, Student’s t-test) at 250, 500, 1,000, and 4,000 Hz.
Similarly, Voss et al[10] found that perforations in two different ears that appeared identical could have
losses that differed by up to 30 dB, if the middle ear air space volumes differed
substantially. The reason why middle ear volume should affect HL caused by a perforation
is that for a given perforation, the resulting sound pressure within the middle ear
space will vary inversely with the middle ear volume.
Effect of Site of Perforation
In our study, 32 (10.6%) perforations were in AI segment, 9 (3%) were in AS segment,
18 (6%) in PS segment, and 41 (13.6%) in PI segment. Two hundred (66.6%) involved
the central compartment. Maharajan et al[5] found that large central perforation was seen in 60.50% of cases, close to the result
of our study. Perforation in PS quadrant was the least common finding in their study,
whereas in our study, the least common site of perforation was AS quadrant. Pannu
et al[16] reported 45 (38%) perforations in the anterior quadrant, 35 (29%) in the posterior
quadrant, and 40 (33%) in the multiple quadrant. Similarly, Mehta et al[13] found that 27% perforations were anterior, 30% were posterior, and 43% affected
both halves of the TM. Therefore, it can be concluded that central perforations are
the most common site of TM perforations.
On analyzing the effect of location of perforation on hearing, it was found that the
mean AB gap was maximum in central perforations (23 dB) and minimum in PI perforations
(9.2 dB). The difference in mean AB gap in central perforations was statistically
significant with all other sites. However, it was also found that 60% of central perforations
were of large size. In low-volume, medium-area group, a total of 91 cases were observed.
The mean AB gap of AI, AS, and PI were found to be similar, that is, 13 dB, and PS
group had mean AB gap of 11 dB. The values were found to be statistically significant
(p = 0.005). This contradicts a widely held clinical view that perforations over the
round window, that is, PI perforations, result in a greater HL due to the sound pressure
acting at the round window, which diminishes the cochlear responses by the “phase
cancellation effect.” The reason for this observation is that the wavelengths of sounds
at < 4 kHz are larger than the middle ear dimensions, so the phase cancellation effect
should theoretically be the same for all sites of perforations. This is in agreement
with the experimental works of Voss et al[10] and Mehta et al,[13] who found no effect of location of perforations on the pressure difference at oval
and round window and their clinical studies, also showed that there was no difference
in HL in anterior and posterior perforations having the same middle ear volume. These
observations are contrary to the results of Pannu et al,[16] Bhusal and Shrivastav,[20] Durko and Latkowski,[15] Nepal et al,[17] and Yung.[21]
However, in our study, the medium-sized central perforations were found to have 7
to 9 dB greater mean HL than anterior or posterior perforations. This can be explained
by the fact that medium-sized central perforation involves the handle of malleus.
This concept was initially proposed by Ahmad and Ramani[8] who observed that malleolar perforations had greater HL than that of similar-sized
nonmalleolar perforations. In low-volume, small-area group, a total of 48 cases were
studied. PI site was found to have less mean AB gap than AI and PS sites by 12 to
14 dB, which was found to be statistically significant (p = 0.001). Similar to the observations in low-volume, medium-area group, this also
showed that there was no round window phase cancellation effect on HL at different
locations in small-sized perforations. It can thus be concluded that the round window
phase cancellation effect causing greater HL in PI perforations does not exist. It
was also found that PS group had maximum mean AB gap of 15 dB, which was 4 to 7 dB
more than those in the AS and AI groups. However, the difference was found to be statistically
nonsignificant (p > 0.05).
The most probable explanation is that the PS perforations are more prone, although
not exclusively to the diseases such as ossicular discontinuity and cholesteatoma.[22] This is in agreement with earlier observations by Ahmad and Ramani.[8] It also confirms the findings by Ibekwe et al[3] who found a statistically significant correlation in HL at PS perforations in chronic
cases. The third subgroup that was studied was the low-volume large-area group. In
this subgroup, there were 106 cases of central perforations and only 1 case of PI
perforation with mean AB gap of 26 and 18 dB, respectively. Therefore, correlations
could not be done in this subgroup. The p-values were found to be nonsignificant.
Limitations of our study were that we did not evaluate the status of ossicles and
the Eustachian tube functioning as we excluded the patients with HL more than the
moderate conductive HL, which is suggestive of ossicular chain involvement by the
disease.
Conclusion
Hearing loss increases as the area of TM perforation increases, and it is greater
at lower frequencies and less at higher frequencies. However, in small- and medium-sized
perforations, the phase cancellation effect of round window causing HL in PI perforations
does not exist. There is an inverse relationship between HL and middle ear air space
volume. On comparing small-sized perforations with low middle ear volume at various
sites, PS perforations were found to have greater hearing impairment than the others,
which was statistically insignificant in our study and would require further studies
with a greater number of ears with PS perforations to establish its significance on
HL.
Note
This study was cleared by institution ethics committee.
Funding
None.
