Keywords
hearing loss - otolaryngology - otology
Introduction
Otitis media (OM) is a broad term referring to inflammation of the middle ear cleft,
with or without effusion. Acute otitis media (AOM) is characterized by rapid onset
of signs and symptoms, such as otalgia and fever. Likewise, otitis media with effusion
(OME) refers to nonpurulent fluid in the middle ear space, often occurring as a result
of previous infection when there are no clinical features of acute infection and without
perforation of the tympanic membrane.[1] Hippocrates in 450 BC was first to describe OM and it continues to present itself
even today as one of the most inexplicable universally observed medical problems of
childhood and a leading cause of hearing loss among them.[2]
OM is estimated to affect over 90% of children before their second birthday.[1] Within the Unites States, half of the children are diagnosed with OM before their
first birthday, and 9 out of 10 children are diagnosed by their fifth birthday.[3] OME is most prevalent under the age of 7 years.[4]
A wide variety of risk factors have been documented for causation of OM. These are
characterized into three groups: host factors (sex, age, racial, social, cultural,
hygiene, nutrition, family history, immunological, physiological and pathological
differences), environmental factors (climatic, seasonal, housing, crowding, and schooling),
and etiological factors (microorganisms, toxic substances, and allergens).[4]
Exposure to smoke is a potential risk factor in the development of OM and has gained
considerable attention in the recent times. Smoking either in active or passive form
is a potential risk factor and certainly increases the risk of respiratory tract infections
and few invasive diseases mostly among adults.[5] Among children, passive smoking is mostly associated with upper and lower respiratory
tract infections, such as AOM, pneumonia, and respiratory bronchitis.[6] Several studies have shown that passive smoking is associated with increased prevalence
of OM.[7] Various pathophysiologies have been explained for passive smoking and most explainable
one is smoke either from active or passive form increases bacteria adherence to the
respiratory epithelium, depresses local immune function, decreases mucociliary action
of respiratory epithelium, and hence is a potential risk factor for OM development.[8]
The clinical course of OM, recurrent OM and its squeal is varied. There may be spontaneous
resolution of symptoms without any adverse effect, while few children may eventually
require some medical attention, be it in form of medications or surgical intervention
(myringotomy and tympanostomy tubes).[9] Unfortunately, many children will still experience recurrent and more severe episodes
of the disease that eventually lead to decreased hearing.[10]
Among many health-related challenges, ear disease and decreased hearing still remain
major health concerns in most of the developing countries. In 1993, hearing impairment
and ear pathology assessment was done in eastern and mid-western part of Nepal in
which 16.6% of the study population had hearing impairment and 7.4% had ear drum pathology,
equivalent to, respectively, 2.71 and 1.48 million people extrapolated to the whole
of Nepal.[11] Likewise, a prospective cross-sectional study in the year 2009 among the children
of age group 5 to 13 showed prevalence of chronic suppurative otitis media to be 7.6%
and OME to be 4.7%.[12]
All studies do not establish an association between environmental tobacco smoke (ETS)
and middle ear disease. Kitchens did a prospective cohort study to determine if ETS
is an etiological factor in OM. Its results revealed that the case group commonly
had exposure to ETS (p = 0.04). However, there was no significant difference in the clinical course of the
children who were exposed to ETS and those who were not.[13]
Despite the high prevalence of OM in the pediatric population and its underlying risk
factors, no studies have been conducted in Nepal on the association of passive smoking
and OM. Changing the study setting from clinics to the community will help in broader
and better understanding of the disease and its associated factors. Complete understanding
of the disease burden in community settings will help in better planning and formulation
of community level interventions.
Methods
A community-based cross-sectional study was conducted in the Montessori and primary
schools of a submetropolitan city in Eastern part of Nepal. The data for this study
was collected in the period between February 2018 and February 2019. Ethical approval
was obtained from the Institutional Ethical Review Board (Ref No: 310/074/075-IRC).
Informed and written consent was taken prior to study.
Study Population
This is a prospective cross-sectional study among school children (Montessori, private,
and government) aged between 3 and 12 years. Children with URTI at the time of examination,
syndromic children, otitis externa, osteoma, ear wax, and environmental smoke exposure
were excluded from the study.
Sampling Technique
Stratified random sampling was done to obtain representative sample. The proportion
was based on total number of Montessori (preschool), private, and government schools.
A list of all Montessori, primary, and government school was obtained from district
education office. The list consisted of the name of the schools and their codes. Then,
using random method, two government schools, two private schools, and two Montessori
schools were selected.
A prior appointment with the head of the selected school was arranged who was then
explained about the study. Once the permission was granted, a visit to the school
was then made on the day of study. The questionaries with the informed consent were
distributed to the students and asked to be filled up by the parents and on the following
day school was revisited and examinations of the students were done. Appropriate tools
and techniques were used to collect the data on the day. In case the data collection
was incomplete, subsequent visits to the school were made on convenient time.
In case of abnormal otoscopic findings suggestive of acute or chronic OM, OME, and/or
hearing loss, the children were further evaluated and investigated in the hospital
clinic with play audiometry (3–5 years of age) or pure tone audiometry (> 5 years
of age) (Madsen Orbiter 922, GN Otometrics A/S, Copenhagen, Denmark) and impedance
tympanometry (Madsen Zodiac 901, GN Otometrics A/S, Copenhagen, Denmark).
Identification of Children with Otitis Media
AOM was considered when there was redness, bulging or perforation with or without
discharge in the tympanic membrane.
Chronic OM was considered when there was perforation, retraction, with or without
ear discharge, or cholesteatoma.
OME was considered when tympanic membrane was dull lusterless, foreshortened handle
of malleus, prominent short process or bluish tympanic membrane, and type B curve
on impedance tympanometry.
Confirmed OM was considered when OM (AOM, OME) in either form was confirmed with type
B curve in impedance audiometry.
Tools of Passive Smoking
Smoking by family member residing in the same house and having close contact with
child in any form (bidi, pipe, cigar, and cigarette).
Audiogram and Impedance Audiometry
All the children were evaluated on the day of study after reviewing the questionaries
filled by their parents. Children were then categorized accordingly if they were normal
or having OM on the basis of handheld otoscope. In case of abnormal otoscopic findings
suggestive of OM, chronic OM, OME, and/or hearing loss, the children were further
evaluated in ENT OPD and were further evaluated with play audiometry (3–5 years of
age) or pure tone audiometry (> 5 years of age) (Madsen Orbiter 922) and impedance
tympanometry (Madsen Zodiac 901) in the required cases.
Data Management and Statistical Analysis
Data was entered in Microsoft Excel and converted into Statistical Package for Social
Sciences for Statistical Analysis (11.5). Alpha numerical codes were used. For descriptive
statistics, data were presented in percentage, mean, and standard deviation (SD) and
summarized using frequency distribution tables and graphical methods of presentation.
For inferential statistics, bivariate analysis was done using appropriate test of
significance (chi-square test, Student’s t-test). Univariate binary logistic regression was done to calculate the crude odds
ratio for the independent factors. Multivariate binary logistic regression was done
to find the adjusted odds ratio. Statistical significance was tested with 95% confidence
interval and p-value < 0.05 was considered as significant.
Result
The study included 663 students studying in Montessori, private, and government school.
The age ranged from 3 to 12 years and the mean age was 7.84 years (SD = 2.772). Male
students (50.7%) slightly outnumbered the females (49.2%) in this study ([Table 1]). Equal number of students (221) were selected from the categorized three different
schools (Montessori, private, and government).
Table 1
Distribution of participants according to age groups, sex (n = 663)
|
Characteristics
|
Category
|
Frequency
|
Percentage
|
|
Abbreviation: SD, standard deviation.
|
|
Age
|
< 6 years
|
166
|
25.0
|
|
≥ 6 years
|
497
|
75.0
|
|
Mean ± SD (min, max)
|
7.84 ± 2.722 (3,12)
|
|
Sex
|
Female
|
326
|
49.2
|
|
Male
|
337
|
50.8
|
Nearly one-fifth of the fathers smoked (20.2%), 2.3% of mothers smoked and few other
family members smoked (8.7%) like grandparents, uncle, and aunts living together as
joint family. The member smoked from a maximum of 7 to a minimum of 1 cigarette per
day ([Fig. 1]).
Fig. 1 Distribution of the smoking characteristics among the family members.
The ears of all the students were examined with a handheld otoscope. Among the four
characteristics of the OM (redness, discharge, perforation, and retraction), maximum
students had retraction in both of the ears (right—13.3%, left—12.2%) ([Table 2]).
Table 2
Distribution of otoscopic findings in both ear (n = 663)
|
Characteristics
|
Category
|
Right ear
|
Left ear
|
|
Frequency
|
Percentage
|
Frequency
|
Percentage
|
|
Redness
|
Present
|
4
|
0.6
|
5
|
0.8
|
|
Absent
|
659
|
99.4
|
658
|
99.2
|
|
Discharge
|
Present
|
4
|
0.6
|
6
|
0.9
|
|
Absent
|
659
|
99.4
|
657
|
99.1
|
|
Perforation
|
Present
|
8
|
1.2
|
18
|
2.7
|
|
Absent
|
655
|
98.8
|
645
|
97.3
|
|
Retraction
|
Present
|
88
|
13.3
|
81
|
12.2
|
|
Absent
|
575
|
86.7
|
582
|
87.8
|
Based on the otoscopic findings, 106 students were investigated with pure tone audiometry
and/or tympanometry as indicated.
Among the investigated ears, conductive hearing loss was detected among 9.5% on right
ear and 10.4% on left ear. Frequency of normal hearing among the investigated ear
was 6.5% and 5.6%, respectively, on right and left ear.
On performing impedance audiometry, B curve was seen in 7.8% (right) and 7.1% (left)
of the investigated ears suggesting OME among the clinically suspected cases of the
OM. Likewise, A curve was seen on 4.7% (right) and 5.6% (left) and C curve was seen
on 0.6% right and 0.5% left ears ([Table 3]; [Figs. 2] and [3]).
Table 3
Results of audiological Investigation among students (n = 663)
|
Investigations
|
Category
|
Right ear
|
Left ear
|
|
Frequency
|
Percentage
|
Frequency
|
Percentage
|
|
Impedance
|
A curve
|
31
|
4.7
|
37
|
5.6
|
|
B curve
|
52
|
7.8
|
47
|
7.1
|
|
C curve
|
4
|
0.6
|
3
|
0.5
|
|
Total
|
87
|
13.1
|
87
|
13.1
|
|
Pure tone audiometry
|
Normal
|
43
|
6.5
|
37
|
5.6
|
|
Conductive
|
63
|
9.5
|
69
|
10.4
|
|
Total
|
106
|
16.0
|
106
|
16.0
|
Fig. 2 Distribution of the impedance audiometry among the students in both ears.
Fig. 3 Distribution of pure tone audiometry (PTA) among the students in the clinically suspected
cases of otitis media.
OM was seen in 12.7% of those younger than 6 years and 16.5% in those aged 6 years
and older. Similarly, males had more OM (15.1%) compared to females (16.0%). Those
studying in government school had highest prevalence of OM (20.8%) in compared to
children in Montessori schools (15.4%), while 10.4% of private school students had
prevalence of OM. OM was found to have statistically significant association with
type of schools.
The prevalence of OM was 36.7% when smoking was present in any of the family members.
The prevalence of the OM was 73.3% when there was smoking habit in the mother, whereas
prevalence was 35.8% when smoking habit was present in father and comparatively less
when smoking habit in other family member (34.5%). OM was found to have statistically
significant association with passive smoking among parents and family members as shown
in [Tables 4] and [5].
Table 4
Association of otitis media with sociodemographic characters
|
Characteristics
|
Category
|
Otitis media
|
Total
|
p-Value
|
|
Yes (%)
|
No (%)
|
|
Age
|
< 6 y
|
21 (12.7)
|
145 (87.3)
|
166
|
0.236
|
|
≥ 6 y
|
82 (16.5)
|
415 (83.5)
|
497
|
|
Gender
|
Male
|
51 (15.1)
|
286 (84.9)
|
337
|
0.771
|
|
Female
|
52 (16.0)
|
274 (84.0)
|
326
|
|
School
|
Montessori
|
34 (15.4)
|
187 (84.6)
|
221
|
0.010
|
|
Private
|
23 (10.4)
|
198 (89.6)
|
221
|
|
Government
|
46 (20.8)
|
175 (79.2)
|
221
|
Table 5
Association of passive smoking with otitis media (n = 663)
|
Characteristics
|
Category
|
Otitis media
|
Total
|
p-Value
|
|
Yes (%)
|
No (%)
|
|
aFisher’s exact test applied.
|
|
Smoking in any members
|
Yes
|
65 (36.7)
|
112 (63.3)
|
177
|
< 0.001
|
|
No
|
38 (7.8)
|
448 (92.2)
|
486
|
|
Smoking mother
|
Yes
|
11 (73.3)
|
4 (26.7)
|
15
|
< 0.001a
|
|
No
|
92 (14.2)
|
556 (85.8)
|
648
|
|
Smoking father
|
Yes
|
48 (35.8)
|
86 (64.2)
|
134
|
< 0.001
|
|
No
|
55 (10.4)
|
474 (89.6)
|
529
|
|
Smoking other members
|
Yes
|
20 (34.5)
|
38 (65.5)
|
58
|
< 0.001
|
|
No
|
83 (13.7)
|
522 (86.3)
|
605
|
Multivariable Analysis of the Associated Factors
The independent variables that were found to be associated with OM at p-value less than 0.1 were taken for multivariable analysis. Binomial logistic regression
was done using ENTER method. There were five variables associated at p < 0.1, which were school type, mother’s literacy, smoking in mother, smoking in father,
and smoking in other members. These five variables were included as covariate in logistic
regression, with confirmed OM as the dependent factor.
Even after adjustment, in government school, smoking in mother, smoking in father,
and smoking in other members were found to be associated significantly (p < 0.05). The different variables, with their respective β-coefficient, adjusted odd’s
ratio along with its 95% confidence interval have been mentioned in [Table 6].
Table 6
Multivariable analysis of the associated factors
|
Variables
|
Categories
|
Beta coefficient
|
Standard error
|
p-Value
|
Adjusted odds ratio
|
95% CI for aOR
|
|
Lower
|
Upper
|
|
Abbreviations: aOR, adjusted odds ratio; CI, confidence interval.
|
|
School
|
Private
|
Reference
|
|
|
|
Montessori
|
0.209
|
0.333
|
0.529
|
1.233
|
0.642
|
2.367
|
|
Government
|
0.686
|
0.349
|
0.049
|
1.987
|
1.002
|
3.939
|
|
Literacy of mother
|
Illiterate
|
Reference
|
|
|
|
|
Literate
|
−0.143
|
0.281
|
0.612
|
0.867
|
0.500
|
1.504
|
|
Smoking in mother
|
No
|
Reference
|
|
|
|
|
Yes
|
2.410
|
0.635
|
0.000
|
11.134
|
3.206
|
38.668
|
|
Smoking in father
|
No
|
Reference
|
|
|
|
Yes
|
1.343
|
0.245
|
0.000
|
3.829
|
2.368
|
6.193
|
|
Smoking in other members
|
No
|
Reference
|
|
|
|
Yes
|
1.015
|
0.327
|
0.002
|
2.758
|
1.453
|
5.235
|
|
Constant
|
|
−2.573
|
0.246
|
0.000
|
0.076
|
|
|
Discussion
OM is a common disease among the children and adults. Various etiologies and risk
factors have been identified and studied for its causation. Among many risk factors
passive smoking is also considered to be one of them. However, there has been conflicting
results among the various studies in the past for it to be considered as a risk factor.
In our study, OM was clinically diagnosed among 126 students (19%) and confirmed with
investigations in 103 students (15.5%). Smoking among mother, father, and other family
members were seen in 15 (2.3%), 134 (20.2%), and 58 (8.7%) of the total students,
respectively. When smoking history was present in one of the parent and/or family
member, OM was seen in 65 (36.7%), which showed statistically significant association
(p < 0.001). However, OM was seen in 7.8% of students when there was no associated smoking
history in either parent and or a family member.
Sociodemographic Characteristics
Age Group
In this study, mean age of the participants was 7.84 years (SD = 2.72) and it ranged
from 3 to 12 years of age. Various literatures in which school-based studies were
done to determine the prevalence of OM included the similar age group as of the present
study. In the present study, OM was more prevalent among children more than 6 years
of age group. However, in a prospective study conducted among the school-going children
in Kenya by Simões et al the prevalence of AOM and OME was highest in children under
the age of 6 years and steadily declined as children got older.[14]
Gender
There was higher proportion of males (50.8%) than females (49.2%) in this study that
was similar to a cross-sectional community-based survey conducted among school-going
children in Yemen, which had higher proportion of males (54.4%).[15] However, the prevalence of OM was more in the female children (16.0%) compared to
male children (15.1%) in our study that was similar to the cross-sectional study conducted
in Bangladesh where girls were relatively more sufferer than boys (6.6 vs. 4.5%),
which may be due to higher female population in the study group in comparison to the
male.[16]
School Distribution
In our study, we had equal division of the school (i.e., two Montessori, two private,
and two governments schools). The prevalence of OM was higher in the government school
(20.8%) in comparison to the Montessori (15.4%) and private school (10.4%). Contradictory
to our study, a prospective cross-sectional study done in urban private school of
Nepal showed that the prevalence of CSOM in children studying in urban private schools
of Nepal was 5.0%. This study showed that the prevalence of CSOM was lower in government
school children of Nepal contrary to our study probably due to the study done in the
urban setting where there is a good socioeconomic status, good sanitation, better
housing condition, and better health care facilities as compared with the government
school in rural settings.[12]
Smoking and Otitis Media
The present study had a prevalence of OM of 36.7% among children if any member smoked
and it was statistically significant (p ≤ 0.001).
The prevalence of OM when only mother, father and other family members smoked was
73.3, 35.8, and 34.5%, respectively, and all of it were statistically significant
(p ≤ 0.001). The result of this study is similar to that of Jacoby et al who had conducted
a prospective cohort study on the effect of passive smoking on the risk of OM in Aboriginal
and non-Aboriginal children in western Australia and had also supported that passive
smoking is a risk factor of OM. Sixty-four percent of Aboriginal children and 40%
of non-Aboriginal children were exposed to ETS. Tympanometry was performed on 87 Aboriginal
and 168 non-Aboriginal children; a type B tympanogram (suggesting fluid in the middle
ear) was also associated with passive smoking in Aboriginal children.[17]
Likewise, various studies have supported a parental smoking as a risk factor for OM
among children. Similarly, an epidemiological prospective study was done in a day
care institution to see the association of indoor environment and middle ear effusion
(MEE). With the exception of parental smoking, none of the environmental factors (CO2, temperature, and relative humidity) showed a tendency to influence the prevalence
of MEE.[18]
However, in the present study even after adjustment, government school, smoking in
mother, smoking in father, and smoking in other members were found to be associated
significantly (p < 0.05).
In contrast, Kero and Piekkela investigated AOM and recurrent otitis media in 5,356
children and did not find an association between parental smoking and middle ear disease.
However, they studied children aged 1 year and younger. Furthermore, they studied
not only OME but also AOM and OME together.[19] Perhaps the time taken for the mucociliary dysfunction to the pathogenesis of OM
is longer.
Conclusion
There was a high prevalence of OM (15.5%) in school-going children in the eastern
part. A higher prevalence of OM was seen in the children who were exposed to the passive
smoking as compared with children not exposed to passive smoking. Therefore, an association
can also be drawn between the OM and those exposed to passive smoking. A higher prevalence
of OM was seen among the school children attending government school as compared to
those attending the Montessori and private school.
