The predictive value of tympanometry in otitis media with effusion : a prospective cross-sectional study of 92 ears in Hospital Universiti Sains Malaysia Kota Bharu, Kelantan, Malaysia June 1999 - May 2001
Loading...
Date
2002-05
Authors
Gazali, Norzi
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Sains Malaysia
Abstract
This study was aimed to determine the correlation between tympanogram types
and myringotomy findings, the predictive value of various tympanogram types in
otitis media with effusion, to observe the correlation between otoscopy and some
predisposing factors of otitis media with effusion with myringotomy findings.
A prospective cross-sectional study was conducted on 92 ears presumed to have
otitis media with effusion from history, otoscopy, tympanometry and audiometric
assessment among patients attending the Otorhinolaryngology clinic Hospital
Universiti Sains Malaysia from June 1999 till May 2001. All ears fulfilled at least
two out of five criteria i.e. symptoms of reduced hearing, fullness, recurrent otitis
media or nasal congestion, otoscopic appearance of dullness, retraction or fluid
level, reduced tympanic membrane mobility, conductive hearing loss from
audiogram and type B (flat) tympanogram. A repeat tympanometry was done 24
hours prior to myringotomy. Myringotomy procedure was performed under local
or general anaesthesia.
The highest number of cases (72. 8%) was in the age group of 4 to 13 years old
with mean age of 15.5. Majority was Malay (93.5%) while 72.8% (67) of cases
were males. All symptoms of reduced hearing, ear fullness, chronic nasa
congestion and recurrent otitis media showed slightly higher risk of fluid in the
middle ear with Odds ratio of 1.83, 1.41, 1.67 and 1.47 respectively. Cigarette
smoking (OR =.1.65), passive smoking (OR= 1.65) and allergy (OR= 2.22) also
showed higher risk of fluid in the middle ear although not statistically significant.
In this study, tympanic membrane retraction was found to be highly and
significantly associated with presence of fluid at myringotomy with Odds ratio of
3.35 and P = 0. 046. There was no statistically significant association of otoscopic
appearance of bubbles or fluid level with fluid at myringotomy (P = 0.20) most
likely due to the small samples having these features. Tympanic membrane
immobility showed no significant association with presence of fluid in this study
(P = 0.20). Adenoid hypertrophy was found to have higher risk of fluid in the
middle ear (OR = 2. 77).
Audiogram with conductive hearing loss was found to be significantly correlated
with presence of fluid, with Odds ratio of 3. 89 and P value 0. 044.
Type B tympanogram was found to be significantly associated with presence of
fluid at myringotomy. (OR = 3.63, P = 0.035). The positive predictive value of
type B tympanogram was 86.5% with sensitivity of 78.4% and specificity of 50%.
Type A tympanogram had a high specificity (94%) but low sensitivity (22.2%) in
predicting a dry middle ear and this correlation was statistically significant (OR =
0.20, P = 0.044). Type C1 tympanogram had less risk of fluid in the middle ear
(OR = 0.23) and this was statistically significant (P = 0.036). Type C2
tympanogram although not statistically significant (P = 0.59), had slightly higher
risk of fluid in the middle ear (OR = 1.27). The combination of type B and C2
tympanogram improved the positive predictive value (87. 7%) and sensitivity
(86.5%) in predicting presence of middle ear fluid. Type A tympanogram was
found in all ears without sign and symptoms of otitis media with effusion.
In conclusion, this study showed a reasonable correlation of tympanogram types
with myringotomy findings but the use in screening and diagnosis of otitis media
with effusion should be in combination with other parameters i.e. symptoms,
otoscopy and audiogram.
Description
Keywords
Tympanogram types and myringotomy findings,