G Model

PEDOT-5922; No. of Pages 5

International Journal of Pediatric Otorhinolaryngology xxx (2011) xxxxxx

Contents lists available at SciVerse ScienceDirect

International Journal of Pediatric Otorhinolaryngology

journal homepage: www.elsevier.com/locate/ijporl

Hearing impairment in otitis media with effusion: A cross-sectional study based in

Pokhara, Nepal
Richard J. Bagshaw a, Elizabeth H. Wall a, George Dowswell b,*, Thomas P.C. Martin c, Michael C.F. Smith d,e
a

College of Medical and Dental Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
Cancer and Chronic Disease Team, Primary Care Clinical Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
c
Addenbrookes Hospital, Hills Rd., Cambridge CB2 0QQ, UK
d
ENT Department, Hereford Hospitals Acute NHS Trust, Union Walk, Hereford HR1 3HB, UK
e
International Nepal Fellowship, PO Box 5, Pokhara, Nepal
b

A R T I C L E I N F O

A B S T R A C T

Article history:
Received 12 April 2011
Received in revised form 14 September 2011
Accepted 19 September 2011
Available online xxx

Objective: Otitis media with effusion (OME) is a major cause of childhood hearing impairment (HI) in the
developing world, but its prevalence has never been quantied in Nepal. This study therefore set out to
determine the proportion of cases of OME complicated by HI and to identify associated factors.
Methods: This was a cross-sectional prevalence survey carried out in rural, urban and Tibetan schools in
and around Pokhara, Nepal. HI was the primary outcome, and was dened as a middle-frequency pure
tone average >25 dB on audiological testing. The study population was dened as children aged four
years and older, attending primary school and with a diagnosis of OME.
Results: One hundred and eleven schoolchildren with a combined total of 172 ears affected by OME
underwent audiometric assessment. HI was most prevalent in the rural Nepali population; 27% (95%CIs
1838%) had HI, with a mean hearing loss of 22 dB (1525 dB). In the Tibetan population, 16% (829%)
had HI, with a mean loss of 17 dB (1222 dB). The urban Nepali population had the least HI; 4% (113%)
were affected, with a mean loss of 16 dB (1519 dB). The difference in prevalence between the urban and
rural Nepali populations was statistically signicant (p > 0.05). Logistic regression analysis did not
identify any associated factors.
Conclusions: HI is a common complication of OME in Nepal. There is hitherto-unreported variation
between populations in the number of cases of OME complicated by HI. This study identied higher rates
of morbidity amongst rural populations but was unable to identify associated factors.
2011 Elsevier Ireland Ltd. All rights reserved.

Keywords:
Otitis media with effusion
Hearing impairment
Prevalence
Risk factors
Nepal

1. Introduction
Globally, an estimated 278 million people live with disabling
hearing impairment (HI), of whom 80% live in low- and middleincome countries, and a quarter develop HI during childhood [1].
Childhood HI, even if transient, can have long-term adverse effects
on behaviour, speech and language development, academic
performance and future employment [24]. Early identication
and treatment of children with HI is therefore imperative. This is
particularly true in low-income countries, where the overall
prevalence of HI in infants is twice that of high-income countries
[5].
Otitis media is the single most common cause of childhood
hearing impairment in low-income countries, where middle ear
disorders tend to be more prevalent than in high-income countries
[5,6]. Of these, otitis media with effusion (OME) is amongst the

* Corresponding author. Tel.: +44 0 121 414 9069; fax: +44 0 121 414 6571.
E-mail address: g.dowswell@bham.ac.uk (G. Dowswell).

most prevalent, as it is linked to poverty and malnutrition [7,8].

Whilst the clinical course of OME is often benign and self-limiting
[9,10], it can cause a persistent, uctuant, conductive hearing loss
in a minority of cases [11] lasting for months or years [12]. Even
this minority of cases of OME constitute a major cause of HI in lowincome countries [1215].
The literature on hearing loss in OME is small and fragmented,
with existing papers employing various methodologies [5].
Consensus exists, however, that hearing loss in OME is typically
mild. Low-frequency pure-tone average (LPTA) thresholds were
reported in one study as averaging 24 dB (decibels), with the
greatest loss in the lowest frequencies [16]. Another study
describes 26.3% of ears with OME failing to hear a 25 dB screening
tone in at least one of four 0.54 kHz frequencies [17]. A further
paper reports hearing loss as ranging from 15 to 40 dB [4]. The mild
(>25 dB) HI caused by OME is less disabling but more insidious
than severe HI, often going unnoticed by children and their parents
but still adversely impacting development [18].
Clinically signicant hearing loss has been dened in several
clinical guidelines as bilateral hearing loss of >2530 dB, averaged

0165-5876/$ see front matter 2011 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.ijporl.2011.09.013

Please cite this article in press as: R.J. Bagshaw, et al., Hearing impairment in otitis media with effusion: A cross-sectional study based in
Pokhara, Nepal, Int. J. Pediatr. Otorhinolaryngol. (2011), doi:10.1016/j.ijporl.2011.09.013

G Model

PEDOT-5922; No. of Pages 5

2

R.J. Bagshaw et al. / International Journal of Pediatric Otorhinolaryngology xxx (2011) xxxxxx

across 0.5, 1, 2 and 4 kHz [1921]. So far as we are aware, there are
no published estimates of the proportion of people with bilateral
OME who also suffer signicant hearing loss. Treatment of OME is
by the insertion of ventilation tubes into the tympanic membrane
[19]. Weighing the risks of this procedure against the adverse
developmental effects of HI has led to recommendations that
surgical treatment of OME should be reserved for cases complicated by HI for three months or longer [1921].
Given OMEs association with poverty and Nepals status as
one of the poorest and least-developed countries in the world
[22], it is unsurprising that the prevalence of OME in Nepal is
reported to be high; one study of schoolchildren aged 515
reports a prevalence of 24.5% [23]. However, as the majority of
cases of OME do not require treatment, and the reported degree
of HI is quite variable [16,17,24], the scale of the public health
problem in Nepal can only be estimated when both prevalence
and rates of HI are known.
We identied only one published study of good methodological
quality on hearing loss and middle ear disease in Nepal [25]. Whilst
Schmitz et al. report that abnormal tympanometry increases the
risk of HI compared to that of the general population (OR = 18),
they do not quantify the hearing loss, or report the proportion of
ears with both abnormal tympanometry and HI. Additionally, this
study examined only a rural population, and only participants aged
1523. Our study therefore set out to examine the proportion of
cases of childhood OME complicated by HI, to establish the range
and median level of hearing loss seen in cases of OME, and to
identify environmental and demographic factors associated with
HI.
2. Methodology
This was a cross-sectional point prevalence survey nested within
a survey of the prevalence of OME. The primary outcome measure
was the presence of hearing impairment (HI). In line with World
Health Organisation [26] and UK National Institute for Health and
Clinical Excellence [19] denitions, HI was dened in this study as a
middle-frequency pure-tone average (MPTA) threshold of >25 dB in
the frequencies 500 Hz, 1 kHz, 2 kHz and 4 kHz. Data on other
relevant audiological outcomes are also provided.
The study took place in primary schools in Pokhara District,
Kaski, Western Region Nepal. Data was collected from the pupils of
four schools in the rural area around Pokhara, one urban school
within Pokhara, and two schools in Tibetan migrant settlements.
Despite being located in rural areas, the Tibetan compounds had a
population density and infrastructure comparable to urban
Pokhara. The study population was dened as children attending
primary school, aged over four years and with a diagnosis of
unilateral or bilateral OME.
2.1. Inclusion criteria
 Only children present at school on the day of data collection were
included in the study.
 Children were included if a diagnosis of OME was made, based on
clinical, otoscopic and tympanometric ndings. The latter two
tests were chosen as they are recommended by clinical guidelines [20] as the best tests for OME which can be carried out by
non-specialist staff. Based on existing clinical denitions of OME
[7], we dened OME as:
The presence of a Type B tympanogram, indicative of a middle
ear effusion.
The absence of ear pain and an inamed tympanic membrane,
indicative of acute otitis media.
 Participants with OME and excess but unimpacted cerumen
(diagnosed by the presence of view-obstructing cerumen, a low

auditory canal volume and a low compliance peak in the

tympanogram) were not excluded, as previous studies have
demonstrated that the prevalence of HI is not signicantly
greater in this group than in the general population [5,25].

2.2. Exclusion criteria

 Children younger than four were excluded, as standard
audiological testing in this age range does not produce reliable
results [27].
 Children with impacted cerumen were excluded, as this
condition causes a conductive hearing loss [5]. Impacted
cerumen was dened as view-obstructing cerumen and a at
tympanometry trace.
 Participants with incidental sensorineural hearing loss (indicated by an absent boneair gap) were excluded, as their hearing
loss was not attributable to OME.
Participants were recruited from the prevalence survey of OME
within which this study was nested. The prevalence survey used a
convenience sample drawn from local schools. Audiological data
on air- and bone-conduction thresholds for 500 Hz, 1 kHz, 2 kHz
and 4 kHz tones were collected. Audiology was conducted in a
quiet room, by a trained audiometrician using a recently calibrated
Amplivox 260 portable audiometer and ambient noise-attenuating
TDH-39 audiocups [28]. Data on age, gender and environmental
risk factors were collected from participants in a structured
interview by an interviewer uent in Nepali. Environmental risk
factors were proposed on the basis of their association with OME;
known risk factors include exposure to other children, low socioeconomic status, and environmental pollutants such as cigarette
smoke [2932]. Data were collected on population of origin (rural
Nepali, urban Nepali or Tibetan), parental smoking status, number
of rooms at home (intended as a proxy for family income), number
of siblings and fuel burnt at home (wood or gas). Clinical ndings
were also used to characterise OME by laterality (unilateral or
bilateral) and side affected (left or right).
Statistical analysis was carried out using SPSS Version 17. As per
protocol, the population as a whole was analysed and owing to
unforeseen heterogeneity between subpopulations rural Nepali,
urban Nepali and Tibetan populations were analysed separately.
Analysis was performed by case (i.e. ears affected by OME, meaning
children with bilateral OME were analysed as two separate cases).
Tests of normality were carried out on continuous data. Medians,
inter-quartile ranges (IQR) and ranges were calculated for nonnormal data and presented as summary statistics. The proportion
of cases with mild (MTPA >25 dB) and moderate (MTPA >40 dB) HI
was calculated, with 95%CIs, for each population and for the whole
sample. A binary logistic regression was performed to demonstrate
the association between HI and demographic, environmental and
pathological variables.
Prior to the start of data collection, institutional ethical
approval for this study was gained from the BMedSci Internal
Ethics Review Committee at the University of Birmingham, UK
[Reference: BMedSc/IERC/010220100]. Verbal consent was sought
from participants according to their competency. However, due to
the age of our study population many children lacked the
competency to provide full informed consent. Written consent
was gained instead from the headteacher of the school, who acts as
the guardian for the children during the daytime. Parental
acquiescence was assumed but parents were free to request that
their children opt out of the study. Written parental consent was
felt to be not appropriate, as fewer than two thirds of adults in
Nepal are literate [22] and written forms are not the Nepali cultural
norm.

Please cite this article in press as: R.J. Bagshaw, et al., Hearing impairment in otitis media with effusion: A cross-sectional study based in
Pokhara, Nepal, Int. J. Pediatr. Otorhinolaryngol. (2011), doi:10.1016/j.ijporl.2011.09.013

G Model

PEDOT-5922; No. of Pages 5

R.J. Bagshaw et al. / International Journal of Pediatric Otorhinolaryngology xxx (2011) xxxxxx
Table 1
Case characteristics.
Agea
n = 166
Gender
n = 164c
Population
n = 166
Parental smoking
n = 160
Number of roomsa
n = 141
Number of siblingsa
n = 164
Fuel burnt at home
n = 160
Laterality
n = 157
Side
n = 166
Excess cerumen
n = 164
a
b
c

45%
Median
8
Male
66 (40%)
Urban Nepali
51 (31%)
Yes
17 (11%)
Median
2
Median
2
Wood
66 (41%)
Unilateral
44 (28%)
Right
83 (50%)
Yes
23 (14%)

IQRb
610
Female
98 (60%)
Rural Nepali
70 (42%)
No
143 (89%)
IQR
23
IQR
13
Gas
94 (59%)
Bilateral
113 (72%)
Left
83 (50%)
No
141 (86%)

Range
413

Tibetan
45 (27%)

Mild (>25dB) Hearing

Impairment
40%
Moderate (>40dB) Hearing
Impairment
35%

30%
Range
111
Range
06

25%

20%

15%

10%

Data are not normally distributed.

Inter-quartile range.
n < 166 due to missing data.

5%

Data were anonymised and stored securely to ensure that

participant condentiality was maintained. Referrals to secondary
care were arranged when incidental serious ear pathology was
detected.

0%
Urban Nepali

Rural Nepali

Tibetan

Overall

Fig. 1. Prevalence of hearing impairment.

30

From a total of 233 ears diagnosed with OME, audiological data

were collected for 172 (74.2%; 111 schoolchildren). Data were not
collected for all eligible ears owing to time constraints. Six ears
(3.5%) were excluded due to the presence of sensorineural HI,
leaving data on 166 ears eligible for analysis. Information on the
demographic, environmental and pathological characteristics of
cases is presented in Table 1.
The audiological ndings for each population and for the overall
sample are summarised in Table 2. Prevalence of mild (MPTA
>25 dB) hearing impairment (HI) was high in the rural Nepali
population with OME (27%, 95%CIs 1838%) and low in the urban
Nepali population with OME (4%, 95%CIs 113%) compared to the
overall rate (17%, 95%CIs 1223%). There was a statistically
signicant difference between rates of mild HI in urban and rural
Nepali populations (p > 0.05). Ten percent of rural Nepali cases (4%
of the total) had a moderate (MPTA >40 dB) HI, but none was seen
in the Tibetan or urban Nepali populations.
The overall MPTA in OME ranged from normal (OdB) to a
moderate impairment of 47 dB, with a median threshold of 17 dB.
Median hearing thresholds appeared higher in the rural Nepali
population (22 dB, 95%CIs 1525 dB) than in the Tibetan (17 dB,
95%CIs 1222 dB) and urban Nepali (16 dB, 95%CIs 1519 dB)
populations. None of the differences between populations were
statically signicant (see Fig. 1).
Pearsons Chi-squared test demonstrated no signicant association between the presence of non-impacted wax and HI in our
sample (p = 0.064).

25

Median MPTA* (dB)

3. Results

20
15
10
5
0
Urban Nepali

Rural Nepali

Tibetan

Overall

*MPTA = Middle-frequency pure-tone average

Fig. 2. Degree of hearing loss. *MPTA, middle-frequency pure-tone average.

One hundred and twenty-two cases (73%) had no missing data

and were therefore included in the regression analysis. Forwards
and backwards binary logistic regressions were performed with HI
as the dependent variable. In both regressions age, gender, parental
smoking status, number of rooms, number of siblings, fuel burnt at
home, laterality and side affected were all found not to be
associated with HI. Population of origin, however, was found to be
correlated with the risk of HI, with a signicantly increased risk for
the rural Nepali population (OR = 10.9, p = 0.003) and the Tibetan
population (OR = 5.13, p = 0.05) compared to the urban Nepali
population. This accounted for 16.7% of all variability (RN = 0.167).

Table 2
Audiological ndings.

MPTAa median (dB)

MPTA range (dB)
Mild (>25 dB) hearing impairment
Moderate (>40 dB) hearing impairment
a
b

Urban Nepali (n = 51)

Rural Nepali (n = 70)

Tibetan (n = 45)

Overall (n = 166)

16 (1519)b
527
2 (4%)
0 (0%)

22 (1525)
047
19 (27%)
7 (10%)

17 (1222)
536
7 (16%)
0 (0%)

17 (1620)
047
28 (17%)
7 (4%)

Middle-frequency pure-tone average.

95% condence intervals.

Please cite this article in press as: R.J. Bagshaw, et al., Hearing impairment in otitis media with effusion: A cross-sectional study based in
Pokhara, Nepal, Int. J. Pediatr. Otorhinolaryngol. (2011), doi:10.1016/j.ijporl.2011.09.013

G Model

PEDOT-5922; No. of Pages 5

4

R.J. Bagshaw et al. / International Journal of Pediatric Otorhinolaryngology xxx (2011) xxxxxx

When further forwards and backwards regressions were

performed without inclusion of the population variable, fuel burnt
at home was found to be associated with HI. People burning wood
at home were at signicantly increased risk of HI (OR = 4.11,
p = 0.004) compared to people burning gas. However, as almost all
Tibetan (97.8%) and urban Nepali (100%) cases burnt gas, and all
rural Nepali cases burnt wood, this association is likely to be due to
confounding by population rather than causal (see Fig. 2).
4. Discussion
Our study found that the overall prevalence of HI in OME was
17% (95%CIs 1223%). A new nding of our study is that there
appears to be a difference in rates of HI in OME between urban and
rural populations. The highest prevalence was found in the rural
Nepali population (27%) and the lowest in the urban Nepali
population (4%).
The main strength of our study is that it is the rst to quantify
the prevalence of HI in OME in urban Nepali and Tibetan
populations. Furthermore, comparison of these populations
demonstrated variation not observed in previous studies; which
treated all participants as members of a homogenous population.
However, our study also has several methodological weaknesses.
The major weakness is that audiological data were collected in suboptimal conditions; in a quiet rather than a soundproof room.
Ambient noise was an issue, especially in the rural schools, where
testing in a separate room to clinical examination was not possible.
Higher levels of ambient noise in these schools than at others may
have caused false positives, and consequently the high level of HI
described in the rural Nepali population may be partly artefactual.
Unfortunately, a decibel meter was not used to quantify ambient
noise during testing, and hence its inuence on the results is
uncertain. We did, however, use TDH 39 audiocups rather than
standard headphones, which attenuate background noise by 25
40 dB in the frequencies used by this study [28].
The other major weakness of our study is that the convenience
sample used reduces the reliability with which our ndings can be
generalised to all schoolchildren in Nepal. This sampling strategy
was guided by resource availability. Our results should also not be
generalised to all children in Nepal, as children who attend school
are not representative of all children of school age. For instance,
24% of rural Nepali children do not attend primary school [25]. An
additional weakness of our study is that by conducting a crosssectional study, we are unable to estimate the proportion of cases
of HI sustained for long enough to merit treatment [19]. Given the
uctuant nature of hearing loss in OME [11], a prospective design
would have produced more valuable results, but was not feasible
given our resources. A nal weakness of our study concerns the
internal validity of our data on environmental variables. The
accuracy of young childrens knowledge is questionable, with
perhaps the most-affected appearing to be data on parental
smoking. Just 11% of participants reported that at least one parent
smoked, despite an estimated prevalence of 29% [22]. Our results
should therefore be interpreted with caution.
Our research supports the view that, in Nepal as in other lowincome countries [5,15], OME accounts for a large proportion of
childhood HI. Given the adverse effects of childhood HI [24], and
the relative ease with which OME can be treated [19], we can
conclude that OME constitutes an important but soluble public
health problem in Nepal, especially for the rural poor. The variation
in rates of HI that we observed also highlight the importance of
measuring not only the prevalence of OME but the rate of HI it
causes in any given population. Generalising from other populations may produce inaccurate estimates of the burden of disease.
Comparison of ours and others results suggests that the
heterogeneity observed within our study has been repeated

between studies. We observed hearing thresholds in OME ranging

from normal (OdB) to a moderate loss of 47 dB, broader than the
1540 dB range reported by Klein [4]. The overall median hearing
threshold of 17 dB (95%CI 1620 dB) reported in this study is lower
than the 24 dB hearing threshold reported by Fria et al. [16],
although this may be partially explained by our use of a middlefrequency pure-tone average (MPTA), where Fria et al. used the
low-frequency pure-tone average (LPTA). The overall prevalence
we reported of a moderate (>40 dB) HI was 4% (95%CI 28%),
slightly lower than Roberts et al.s reported estimate of 510% [24].
Our ndings beg the question what characteristics of these
populations cause variation in the rate of HI? Our study is unable to
offer an explanation, as none of the variables we examined were
associated with a risk of HI. Much of our data on environmental
variables is also of uncertain quality, highlighting the need for
further research into the effect of such variables on the risk of HI.
We can, however, offer a hypothesis to be tested in future research.
The prevalence of OME is linked to poverty, with overcrowding,
malnutrition and lack of running water all contributing to the
ongoing carriage of pathogenic bacteria in the respiratory tract [7].
Risk of developing OME also varies with genetic and immunological factors, exposure to environmental pollutants such as tobacco
smoke and wood smoke, and anatomical or pathological disturbance of Eustachian tube function [4]. All of these factors can vary
between populations. We suggest that some or all of these factors
exert an inuence not only on the risk of acquiring OME but on its
natural history, leading to variation in the risk of HI.
Future research could use a larger sample size and include a
broader range of variables in analysis, in order to determine
possible causes of the variation in HI seen in the different
populations we studied. Future projects wishing to improve
external validity could also sample by village rather than school,
and employ a random selection process for participant recruitment. Future studies conducting audiology in the eld would be
also well-advised not only to use noise-attenuating headphones
but to monitor ambient noise, as did Schmitz et al. [25]. This would
allow the effect of ambient noise to be controlled for at the analysis
stage.Lastly, pilot studies are needed to ensure treatment with
ventilation tubes or hearing aids is as effective and well-tolerated it
has been shown to be in developed countries such as the UK [19].
Poor ear hygiene and an increased risk of infection in the rural
population may produce higher rates of complications for
ventilation tubes, whilst maintaining working hearing aids may
also prove difcult and costly for rural populations.
Funding
This study was funded by the University of Birmingham and the
Arthur Thompson Trust.
Author contributions
MS and TM conceived the project. Write-up of the protocol and
data collection were carried out by RB and EW. Analysis of the data
and write-up of the rst draft was done by RB. Methodological
input was provided by GD, MS and TM. All authors have
contributed to and approved the nal document.
Declaration of interest
The authors report no conicts of interest.
Acknowledgements
We gratefully acknowledge the assistance of Eka Dev Dekota,
Camps Director at the International Nepal Fellowship, Robin Lama,

Please cite this article in press as: R.J. Bagshaw, et al., Hearing impairment in otitis media with effusion: A cross-sectional study based in
Pokhara, Nepal, Int. J. Pediatr. Otorhinolaryngol. (2011), doi:10.1016/j.ijporl.2011.09.013

G Model

PEDOT-5922; No. of Pages 5

R.J. Bagshaw et al. / International Journal of Pediatric Otorhinolaryngology xxx (2011) xxxxxx

our audiometrician, and Sandha Paudel, our interviewer. Without

their hard work this study would not have been possible. We also
thank Sarah Riches at Aston University, for her training and advice
prior to data collection, Dr Lesley Roberts and Dr Sayeed Haque at the
University of Birmingham, for their unwavering support and advice
on statistical matters. Lastly, we extend our gratitude to the
schoolchildren of Pokhara District for their participation in this
study.

References
[1] WHO, Deafness and hearing impairment, World Health Factsheet #300, 2006.
Available
from:
http://www.who.int/mediacentre/factsheets/fs300/en/
index.html (accessed 20.03.11).
[2] R.J. Ruben, Redening the survival of the ttest: communication disorders in the
21st century, Int. J. Pediatr. Otorhinolaryngol. 49 (Suppl. 1) (1999) S37S38.
[3] J.E. Roberts, R.M. Rosenfeld, S.A. Zeisel, Otitis media and speech and language: a
meta-analysis of prospective studies, Pediatrics 113 (3 (Pt 1)) (2004) e238e248.
[4] J.O. Klein, The burden of otitis media, Vaccine 19 (Suppl. 1) (2000) S2S8.
[5] J.A. Czechowicz, A.H. Messner, E. Alarcon-Matutti, et al., Hearing impairment and
poverty: the epidemiology of ear disease in Peruvian schoolchildren, Otolaryngol.
Head Neck Surg. 142 (2) (2010) 272277.
[6] A.D. Olusesi, Otitis media as a cause of signicant hearing loss among Nigerians,
Int. J. Pediatr. Otorhinolaryngol. 72 (6) (2008) 787792.
[7] K. Kong, H.L. Coates, Natural history, denitions, risk factors and burden of otitis
media, Med. J. Aust. 191 (9) (2009) S39S43.
[8] J. Engel, L. Anteunis, A. Volovics, et al., Risk factors of otitis media with effusion
during infancy, Int. J. Pediatr. Otorhinolaryngol. 48 (3) (1999) 239249.
[9] R.M. Rosenfeld, D. Kay, Natural history of untreated otitis media, Laryngoscope
113 (10) (2003) 16451657.
[10] G.A. Zielhuis, G.H. Rach, P. van den Broek, The natural course of otitis media with
effusion in preschool children, Eur. Arch. Otorhinolaryngol. 247 (4) (1990) 215
221.
[11] Otitis media with effusion, Pediatrics 113 (5) (2004) 14121429.
[12] I.G. Williamson, J. Dunleavey, J. Bain, et al., The natural history of otitis media with
effusiona three-year study of the incidence and prevalence of abnormal tympanograms in four South West Hampshire infant and rst schools, J. Laryngol.
Otol. 108 (11) (1994) 930934.
[13] F. Martines, D. Bentivegna, P.F. Di, et al., The point prevalence of otitis media with
effusion among primary school children in Western Sicily, Eur. Arch. Otorhinolaryngol. (2009).

[14] A. Saim, L. Saim, S. Saim, et al., Prevalence of otitis media with effusion amongst
pre-school children in Malaysia, Int. J. Pediatr. Otorhinolaryngol. 41 (1) (1997)
2128.
[15] A. Jacob, V. Rupa, A. Job, et al., Hearing impairment and otitis media in a rural
primary school in south India, Int. J. Pediatr. Otorhinolaryngol. 39 (2) (1997) 133
138.
[16] T.J. Fria, E.I. Cantekin, J.A. Eichler, Hearing acuity of children with otitis media with
effusion, Arch. Otolaryngol. 111 (1) (1985) 1016.
[17] N. Yockel, A comparison of audiometry and audiometry with tympanometry to
determine middle ear status in school-age children, J. Sch. Nurs. 18 (2002) 287.
[18] P. Watkin, D. McCann, C. Law, et al., Language ability in children with permanent
hearing impairment: the inuence of early management and family participation,
Pediatrics 120 (3) (2007) e694e701.
[19] NICE, Surgical Management of Otitis Media with Effusion in Children, 2008.
[20] SIGN, Diagnosis and Management of Childhood Otitis Media in Primary Care: A
National Clinical Guideline, Scottish Intercollegiate Guidelines Network, 2003.
[21] N. Freemantle, A. Long, J. Mason, The treatment of persistent glue ear in children,
Eff. Health Care 4 (1992) 115.
[22] World Bank World Development Indicators, 2010. Available from: http://data.worldbank.org/data-catalog/world-development-indicators (accessed 20.03.2
011).
[23] M. Maharjan, S. Bhandari, I. Singh, et al., Prevalence of otitis media in school going
children in Eastern Nepal, Kathmandu Univ. Med. J. (KUMJ) 4 (4) (2006) 479482.
[24] J. Roberts, L. Hunter, J. Gravel, et al., Otitis media, hearing loss, and language
learning: controversies and current research, J. Dev. Behav. Pediatr. 25 (2) (2004)
110122.
[25] J. Schmitz, J.P. Pillion, S.C. Leclerq, et al., Prevalence of hearing loss and ear
morbidity among adolescents and young adults in rural southern Nepal, Int. J.
Audiol. (2010).
[26] WHO, Ear and hearing disorders survey. Protocol for a population-based survey of
prevalence and causes of deafness and hearing impairment and other ear disease,
Prevention of blindness and deafness, 1999.
[27] M. Cunningham, E.O. Cox, Hearing assessments in infants and children: recommendations beyond neonatal screening, Pediatrics 111 (2) (2003) 436440.
[28] Amplivox, TDH-39 Audiocups, 2010. Available from: http://www.amplivox.ltd.uk/PDF%20les/audiocups2.pdf (accessed 20.03.11).
[29] D.P. Strachan, D.G. Cook, Health effects of passive smoking. 4. Parental smoking,
middle ear disease and adenotonsillectomy in children, Thorax 53 (1) (1998) 50
56.
[30] I. Williamson, Otitis media with effusion, Clin. Evid. 12 (2004) 764774.
[31] D.W. Teele, J.O. Klein, B. Rosner, Epidemiology of otitis media during the rst
seven years of life in children in greater Boston: a prospective, cohort study, J.
Infect. Dis. 160 (1) (1989) 8394.
[32] J.L. Paradise, H.E. Rockette, D.K. Colborn, et al., Otitis media in 2253 Pittsburgharea infants: prevalence and risk factors during the rst two years of life,
Pediatrics 99 (3) (1997) 318333.

Please cite this article in press as: R.J. Bagshaw, et al., Hearing impairment in otitis media with effusion: A cross-sectional study based in
Pokhara, Nepal, Int. J. Pediatr. Otorhinolaryngol. (2011), doi:10.1016/j.ijporl.2011.09.013