• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information

Diagnosis of OME

Bibliographic detailsStudy type and evidence levelPatient characteristicsTest, reference standard, threshold for a positive testResultsReviewer comments
Anteunis (1999)23
{37318}
Diagnostic study

EL = II
Full-term infants (birthweight 2500–4500 gms and gestational age 38–42 weeks) and preterm infants (birthweight under 1500 gms and gestational age under 33 weeks) recruited from newborn ward and intensive care unit respectively at birth, and examined every 3 months until the age of 24 months.
(Full-term infant–parent pairs= 150, preterm infant–parent pairs= 66)
  1. Parental reporting on OME vs OME diagnosed clinically
    Parental reporting about OME assessed by a questionnaire asking questions on the period since preceeding examination.
    OME confirmed clinically by otoscopy and tympanometry (by an otolaryngologist and an audiologist)
  2. Parental reporting on AOM vs AOM diagnosed clinically
  3. Parental reporting on HI vs HI diagnosed clinically
Comparison 1 in Full-term infants
Sensitivity: 16.5%
Specificity: 92.8%
PPV: 67.3%
NPV: 55.2%

When parents informed about OME presence in previous visit
Sensitivity: 19.6%
Specificity: 89.1%
PPV: 73.7%
NPV: 41.7%

When parents informed about OME absence in previous visit
Sensitivity: 12.8%
Specificity: 94.5%
PPV: 56.8%
NPV: 66.0%

Comparison 1 in Preterm infants
Sensitivity: 18.2%
Specificity: 88.3%
PPV: 68.4%
NPV: 43.8%

When parents informed about OME presence in previous visit
Sensitivity: 20.9%
Specificity: 88.2%
PPV: 82.6%
NPV: 29.4%

When parents informed about OME absence in previous visit
Sensitivity: 13.2%
Specificity: 92.7%
PPV: 55.6%
NPV: 60.7%
Unselected population
Validity of questionnaire – not specified
Tests and reference standard adequately described
Tests and reference tests done by trained personnel
Blinding – not specified
Results not given for AOM and HI as not relevant to guideline question
Babonis (1991)41
{37255}
Diagnostic study

EL = I b
Children scheduled for elective myringotomy and pressure equalization tube placement due to persistent MEE for 4 months, recurrent OM (three in previous 6 months or five in a year), or recurrent OM unresponsive to prophylactic antibiotics (n = 120, ears= 240)
Age range: 6 months – 10 years 9 months
139 males
  1. Portable tympanometry by one of the authors vs Myringotomy
    Threshold: Type B
  2. Acoustic otoscopy/reflectometry by one of the authors vs Myringotomy
    Threshold: > 5 RU
Comparison 1 (n = 220)
Prevalence: 53.6% (118/220)
Sensitivity: 78.0% (92/118)
Specificity: 82.3% (84/102)
PPV: 83.6% (92/110)
NPV: 76.4% (84/110)
Selected population
Tests done immediately prior to the reference standard (exact timing not specified)
Adequate description of test and reference standard
Blinding – Yes
Acoustic otoscopy not relevant to the guideline question
Cantekin (1980)37
{37325}
Diagnostic study

EL = II
Patients scheduled for myringotomy and insertion of tympanostomy tubes on the basis of history of recurrent AOM or persistent MEE or both (n = 333, ears= 599)
Age range: 7 months – 15 years
203 males, 130 females
  1. Pneumatic otoscopy by two otolaryngologists (A & B) vs Myringotomy
    Threshold: Present, absent or inflammation without effusion or equivocal
  2. Tympanometry and middle ear (ME) muscle reflex by an audiologist and independently classified by two investigators vs Myringotomy
    Threshold: ME muscle reflex threshold ≤ 105 dB measured using different quantitative criterion (ambient pressure/peak pressure, stimulus frequency 1000/2000)
Comparison 1 excluding equivocal data (Examiner A)
Prevalence: 62.2% (230/370)
Sensitivity: 97.0% (223/230)
Specificity: 81.4% (114/140)
PPV: 89.6% (223/249)
NPV: 94.2% (114/121)

Comparison 1 excluding equivocal data (Examiner B)
Prevalence: 57.4% (201/350)
Sensitivity: 87.6% (176/201)
Specificity: 80.5% (120/149)
PPV: 85.8% (176/205)
NPV: 82.7% (120/145)
Selected population
Test and reference standard done within 1 hour
Adequate description of test and reference standard
Blinding – Not specified
Data not given for tympanometry as combination of thresholds used.
Capper (1987)27
{37279}
Diagnostic study

EL = II
Children presenting with glue ear (n = 125, ears=not specified, Visits= 331)
Age range: not specified (but results given for 4–5, 5–6, and 7– 10 years)

Exclusions: child with known or suspected sensorineural hearing loss, unreliable results on PTA
Tuning fork tests (Rinne and Weber) at 512 Hz by one of the authors vs PTA by an experienced audiologist

Threshold: Rinne negative for a positive test
Weber – lateralized to bad ear
Rinne test (all age groups)
Sensitivity: 87.0%
Specificity: 55.0%

Rinne test (4 – 5 years)
Sensitivity: 80.0%
Specificity: 50.0%

Weber test (all age groups)
Sensitivity: 65.0%
Specificity: 75.0%
Selected population
Time interval between test and reference standard not specified
Test and reference standard described in details
Reference test – not a standard one
Blinding – Yes
Other diagnostic test results unknown as no data provided
Fiellau-Nikolajsen (1980)44
{37281}
Diagnostic study

EL = I b
Children with persistent type B or C tympanogram during 4 screenings done within six month period, and referred for surgery (n = 44, ears= 88)
Age range: 42 – 54 months
23 male, 21 female
Tympanometry (operator not specified) vs Myringotomy

Threshold: Different thresholds used for a positive test – compliance value ≤ 0.1, 0.2, 0.3, 0.4, 0.5, 0.6 or ≥ 0.7, gradient < 0.025, 0.050, 0.075, 0.10, 0.125, 0.150 or ≥ 0.150, and Type B or C2 as abnormal
Compliance < 0.1 as threshold
Prevalence: 52.3% (46/88)
Sensitivity: 19.6% (9/46)
Specificity: 100.0% (42/42)
PPV: 100.0% (9/9)
NPV: 53.2% (42/79)

Compliance < 0.2 as threshold
Prevalence: 52.3% (46/88)
Sensitivity: 45.6% (21/46)
Specificity: 95.2% (40/42)
PPV: 91.3% (21/23)
NPV: 61.5% (40/65)

Compliance < 0.3 as threshold
Prevalence: 52.3% (46/88)
Sensitivity: 34.8% (16/46)
Specificity: 23.8% (10/42)
PPV: 33.3% (16/48)
NPV: 25.0% (10/40)

Gradient < 0.1 as threshold
Prevalence: 52.3% (46/88)
Sensitivity: 91.3% (42/46)
Specificity: 54.8% (23/42)
PPV: 68.8% (42/61)
NPV: 85.2% (23/27)

Type B or C2 as threshold
Prevalence: 52.3% (46/88)
Sensitivity: 91.3% (42/46)
Specificity: 88.1% (37/42)
PPV: 89.4% (42/47)
NPV: 90.2% (37/41)
Selected population
Test done within 30 minutes of the reference standard
Adequate description of test and reference standard
Blinding – Yes
Results calculated from the data given in the study
Grimaldi (1976)33
{37277}
Diagnostic study

EL = III
Children referred by otologists with presumptive diagnosis of MEE, and undergoing myringotomies as an outpatient procedure
(n = 120, ears= 209)
Age group: not specified
  1. Otoscopy by otologists vs Myringotomy
    Threshold: Effusion probable, possible or unlikely
  2. Audiometry by an audiologist vs Myringotomy
    Threshold: not given
  3. Tympanometry by an audiologist vs Myringotomy
    Threshold: not given
Comparison 1 with possible cases as false positive
Prevalence: 73.7% (154/209)
Sensitivity: 85.7% (132/154)
Specificity: 87.3% (48/55)
PPV: 95.0% (132/139)
NPV: 68.6% (48/70)

Comparison 1 with possible cases as true positive
Prevalence: 73.7% (154/209)
Sensitivity: 98.0% (151/154)
Specificity: 36.4% (20/55)
PPV: 81.2% (151/186)
NPV: 87.0% (20/23)
Selected population
Time interval between test and reference standard not specified
Test and Reference test described in details
Blinding – Yes for audiometry and tympanometry, but not specified for otoscopy
Other diagnostic test results (comparison 2 and 3)
unknown as no data provided
Haapaniemi (1997)31
{37297}
Diagnostic study

EL = III
School children of 1, 4, and 8 grades for hearing screening according to the recommendations of the Finnish National Board of Health.
(n = 687, ears=not specified)
Age range: 6–9 years for grade 1, 10–12 years for grade 4, and 13–15 years for grade 8.
  1. Pure tone audiometry (PTA) at 0.25, 0.5, 1, 2, 3, 4, 6, 8 kHz (operator not specified) vs Tympanocentesis
    Threshold: Hearing loss > 15 and 20 dB
  2. Tympanometry and stapedius reflex (operator not specified) vs Tympanocentesis
    Threshold: Different thresholds – peak pressure −100, −150 and −200 daPa, admittance of 0.3 ml, different gradients, and Type B curve.
Comparison 1 with subjects as unit of measure
Threshold > 15 dB
Prevalence: 4.2% (29/687)
Sensitivity: 82.8% (24/29)
Specificity: 82.7% (544/658)
PPV: 17.4% (24/138)
NPV: 99.1% (544/549)
Representative population
Time interval between tests and reference standard not specified
Not clear whether tympanocentesis done in all subjects
Blinding – not specified
Data not extractable for 20 dB threshold on PTA, and different thresholds used for tympanometry.
Harris (2005)38
{37330}
Diagnostic study

EL = II
Children seeking medical treatment for suspected middle ear disease (n = 21, ears= 35)
Age range: 1–10 years
13 boys, 8 girls
  1. Pneumatic otoscopy by otolaryngologist vs Myringotomy
    Threshold: Mobility normal, decreased or no mobility.
  2. Tympanometry at conventional frequency of 226 Hz, and also high frequency 678 and 1000 Hz (test operator not specified) vs Myringotomy
    Threshold: At 226 Hz – Type B and Type B or C.
    At 678 and 1000 Hz – Gelfand criterion.
Comparison 1 with decreased and no mobility as positive test
Prevalence: 62.8% (22/35)
Sensitivity: 90.9% (20/22)
Specificity: 69.2% (9/13)
PPV: 83.3% (20/24)
NPV: 81.8% (9/11)

Comparison 2 at 226 Hz (Threshold - Type B)
Prevalence: 62.8% (22/35)
Sensitivity: 54.5% (12/22)
Specificity: 100.0% (13/13)
PPV: 100.0% (12/12)
NPV: 56.5% (13/23)

Comparison 2 at 226 Hz (Threshold - Type B or C) Prevalence: 62.8% (22/35)
Sensitivity: 81.8% (18/22)
Specificity: 61.5% (8/13)
PPV: 78.3% (18/23)
NPV: 66.7% (8/12)

Comparison 2 at 678 Hz Prevalence: 62.8% (22/35)
Sensitivity: 95.5% (21/22)
Specificity: 53.8% (7/13)
PPV: 77.8% (21/27)
NPV: 87.5% (7/8)

Comparison 2 at 1000 Hz Prevalence: 62.8% (22/35)
Sensitivity: 100.0% (22/22)
Specificity: 53.8% (7/13)
PPV: 78.3% (22/28)
NPV: 100.0% (7/7)
Selected population
Test done immediately before reference standard (exact time not mentioned)
Adequate description of test and reference standard
Blinding – Yes
Jonathan (1989)32
{37519}
Diagnostic study

EL = III
Children admitted for routine myringotomies including in some cases adenoidectomy and/or tonsillectomy
(n = 64, ears= 128)
Age range: 3–14 years
35 boys, 29 girls

A control group also recruited, but findings not relevant to the guideline question
  1. Otoscopy (examiner not specified) vs Myringotomy
    Threshold: Normal or abnormal appearance
  2. PTA (examiner not specified) vs Myringotomy
    Threshold: Hearing loss > 15 dB at all frequencies
  3. Tympanometry (examiner not specified) vs Myringotomy
    Threshold: Flat tympanogram
Comparison 1
Compliance rate: 88.0%
Sensitivity: 100.0% (80/80)
Specificity: 28.0% (9/32)
PPV: 77.7% (80/103)
NPV: 100.0% (9/9)

Comparison 3
Compliance rate: 80.0%
Sensitivity: 90.0%
Specificity: 52.0%

Comparison 2
Compliance rate: 93.0%
Sensitivity: 86.0%
Specificity: 86.0%
Selected population
Time interval between test and reference standard not specified
Test and Reference test described in details
Blinding – Not specified
Other diagnostic test results (comparison 2 and 3)
unknown as no data provided
Karma (1989)39
{37284}
Diagnostic study

EL = II
Children followed for otitis episodes in two urban areas in Finland (n = 2911, Ear related visits= 11804)

Group A: seen by an otolaryngologist in one area (n = 1688, visits= 5949)
Group B: seen by a paediatrician in second area (n = 1223, visits= 5855)
  1. Pneumatic otoscopy by the two examiners vs Myringotomy
Threshold: Different tympanic membrane findings (colour, position, mobility) with and without acute symptoms.
Colour – red, distinctly red, cloudy, abnormal
Position – bulging, retracted, abnormal
Mobility – impaired distinctively or slightly.
Comparisons for findings without acute symptoms

Colour – cloudy (Group A)
Prevalence: 68.8% (408/593)
Sensitivity: 92.9% (379/408)
Specificity: 98.4% (182/185)
PPV: 99.2% (379/382)
NPV: 86.3% (182/211)

Colour – cloudy (Group B)
Prevalence: 69.1% (345/499)
Sensitivity: 69.0% (238/345)
Specificity: 87.7% (135/154)
PPV: 92.6% (238/257)
NPV: 55.8% (135/242)

Colour – abnormal (Group A)
Prevalence: 68.8% (408/593)
Sensitivity: 97.6% (398/408)
Specificity: 92.9% (172/185)
PPV: 99.2% (398/411)
NPV: 94.5% (172/182)

Colour – abnormal (Group B)
Prevalence: 69.1% (345/499)
Sensitivity: 81.2% (280/345)
Specificity: 79.2% (122/154)
PPV: 89.7% (280/312)
NPV: 65.2% (122/187)

Position – bulging (Group A)
Prevalence: 68.8% (408/593)
Sensitivity: 45.1% (184/408)
Specificity: 98.9% (183/185)
PPV: 98.9% (184/186)
NPV: 44.9% (183/407)

Position – bulging (Group B)
Prevalence: 69.1% (345/499)
Sensitivity: 18.3% (63/345)
Specificity: 99.4% (153/154)
PPV: 98.4% (63/64)
NPV: 35.2% (153/435)

Position – abnormal (Group A)
Prevalence: 68.8% (408/593)
Sensitivity: 55.4% (226/408)
Specificity: 94.0% (174/185)
PPV: 95.4% (226/237)
NPV: 48.9% (174/356)

Position – abnormal (Group B)
Prevalence: 69.1% (345/499)
Sensitivity: 50.4% (174/345)
Specificity: 90.4% (138/154)
PPV: 91.6% (174/190)
NPV: 44.7% (138/309)

Mobility – abnormal(Group A)
Prevalence: 68.8% (408/593)
Sensitivity: 98.8% (403/408)
Specificity: 90.3% (167/185)
PPV: 95.7% (403/421)
NPV: 97.1% (167/172)

Mobility – abnormal(Group B)
Prevalence: 69.1% (345/499)
Sensitivity: 93.6% (323/345)
Specificity: 71.4% (110/154)
PPV: 88.0% (323/367)
NPV: 83.3% (110/132)
Unselected population
Test done before the reference standard (exact time not mentioned)
Adequate description of test and reference standard
Blinding – No
Results calculated from the data given in the study
Lo (2006)25
{37296}
Diagnostic (Nested case-control) study

EL = III
Subjects taken from population- based OME screening survey of schoolchildren – positive screens and random sample of negative screens re-examined after 2–3 wks (n = 276)
Inclusion criterion: cases and controls with
  1. parental consent;
  2. parental response to questionnaire;
  3. of Chinese descendants;
  4. 6–7 years of age; and
  5. with PTA results
Cases: positive screen subjects with effusion on microscopy or abnormal tympanometry with average air-bone gap of 10 dB in at least one of the ears (n = 117, 59.8% bilateral OME, 69 boys and 48 girls)
Controls: negative screen with normal otoscopy and tympanogram during re- examination (n = 159, 91 boys and 68 girls)
Self-administered questionnaire sent to parents prior to screening, and a binary choice question asking about hearing impairment.

Otoscopy and tympanometry performed for screening, and re-examination included history, microscopy, repeated tympanometry and stapedius reflex testing, and standard PTA.
OME diagnosed during re- examination
  1. Parental suspicion of hearing loss vs OME
  2. Parental suspicion of hearing loss vs actual hearing loss obtained from PTA
    PTA threshold for hearing loss > 25 dB
Average PTA hearing threshold level in cases = 17 dB (range 3.8–40.0 dB)
Children diagnosed with MEE = 117
Children with average PTA threshold > 25 dB = 17

Comparison 1
Prevalence: 42.4% (117/276)
Sensitivity: 19.7% (23/117)
Specificity: 96.9% (154/159)
PPV: 82.1% (23/28)
NPV: 62.1% (154/248)

P < 0.001 for chi-square test parental suspicion vs OME

Comparison 2
Prevalence: 6.2% (17/276)
Sensitivity: 11.8% (2/17)
Specificity: 90.0% (233/259)
PPV: 7.1% (2/28)
NPV: 94.0% (233/248)

P < 0.686 for chi-square test parental suspicion vs PTA findings
Questionnaire – not validated
Test and reference standard performed by trained personnel
Blinding – not specified
Adequate description of tests and reference standard
Mitchell (1990)30
{37314}
Diagnostic study

EL = III
Consecutive admissions of children with suspected glue ear (n = 50, ears= 100)
Age range: 6 months – 14 years
  1. Pure tone audiometry (PTA) at 500, 1 kHz and 2 kHz (operator not specified) vs Myringotomy
    Threshold: Hearing loss ≥ 20 dB
  2. Tympanometry (operator not specified) vs Myringotomy
    Threshold: Type B
Comparison 1 (n = 67)
Prevalence: 67.5% (51/67)
Sensitivity: 80.4% (41/51)
Specificity: 68.7% (11/16)
PPV: 89.1% (41/46)
NPV: 52.4% (11/21)

Comparison 2 (n = 84)
Prevalence: 77.4% (65/84)
Sensitivity: 87.7% (57/65)
Specificity: 52.6% (10/19)
PPV: 86.4% (57/66)
NPV: 55.6% (10/18)
Selected population
Tests and reference standard done within 24 hours
Tests and reference standard not described in details.
Blinding – not specified
Nozza (1992)45
{37303}
Diagnostic study

EL = I b
Children admitted to the same-day surgery unit of a children’s hospital for myringotomy and tube placement (n = 61, ears= 111)
Age range: 1 – 8 years

Data not given for second part of this study as comparison with non- reference standard (comparison of tympanometry with pneumatic otoscopy in an unselected group of children)
Tympanometry by an audiologist vs Myringotomy

Threshold: Different thresholds used alone and in combination – acoustic reflex present/absent, gradient ≤ 0.1 or 0.2, and peak admittance ≤ 0.1, 0.2, 0.3 or 0.4
Acoustic reflex absent (n = 103)
Prevalence: 73.8% (76/103)
Sensitivity: 88.2% (67/76)
Specificity: 85.2% (23/27)
PPV: 94.4% (67/71)
NPV: 71.9% (23/32)

Gradient0.2
Prevalence: 73.0% (81/111)
Sensitivity: 91.4% (74/81)
Specificity: 70.0% (21/30)
PPV: 89.2% (74/83)
NPV: 75.0% (21/28)

Peak admittance0.2
Prevalence: 73.0% (81/111)
Sensitivity: 55.6% (45/81)
Specificity: 93.3% (28/30)
PPV: 95.7% (45/47)
NPV: 43.8% (28/64)

Peak admittance0.3
Prevalence: 73.0% (81/111)
Sensitivity: 72.8% (59/81)
Specificity: 80.0% (24/30)
PPV: 90.8% (59/65)
NPV: 52.2% (24/46)

Peak admittance0.4
Prevalence: 73.0% (81/111)
Sensitivity: 81.5% (66/81)
Specificity: 63.3% (19/30)
PPV: 85.7% (66/77)
NPV: 55.9% (19/34)
Selected population
Test done within 30 minutes of the reference standard
Adequate description of test and reference standard
Blinding – Yes
Results calculated from the data given in the study
Nozza (1994)34
{37304}
Diagnostic study

EL = I b
Children admitted to the same-day surgery unit of a children’s hospital with history of chronic or recurrent middle ear disease. (n = 171, ears= 249)
Age range: 1 – 12 years
  1. Pneumatic otoscopy by a trained Paediatric Nurse Practitioner (whose findings had been validated earlier) vs Myringotomy
    Threshold: Present or absent
  2. Tympanometry by a trained and certified audiologist vs Myringotomy.
    Threshold: Different thresholds used alone and in combination – acoustic reflex present/absent, gradient ≤ 0, 0.1, 0.2 or 0.3, peak admittance ≤ 0, 0.1, 0.2, 0.3 or 0.4, and tympanometric width > 150, 200, 250, 275, 300, 325, 350 or 400 daPa.
Comparison 1
Prevalence: 55.0% (137/249)
Sensitivity: 84.7% (116/137)
Specificity: 71.4% (80/112)
PPV: 78.4% (116/148)
NPV: 79.2% (80/101)

Comparison 2

Acoustic reflex absent (n = 218)
Prevalence: 56.9% (124/218)
Sensitivity: 85.5% (106/124)
Specificity: 64.9% (61/94)
PPV: 76.3% (106/139)
NPV: 77.2% (61/79)

Gradient0.3
Prevalence: 55.0% (137/249)
Sensitivity: 92.7% (127/137)
Specificity: 38.4% (43/112)
PPV: 64.8% (127/196)
NPV: 79.2% (43/53)

Peak admittance0.2
Prevalence: 55.0% (137/249)
Sensitivity: 46.0% (63/137)
Specificity: 91.9% (103/112)
PPV: 87.5% (63/72)
NPV: 58.2% (103/177)

Peak admittance0.3
Prevalence: 55.0% (137/249)
Sensitivity: 70.1% (96/137)
Specificity: 80.4% (90/112)
PPV: 81.4% (96/118)
NPV: 68.7% (90/131)

Peak admittance0.4
Prevalence: 55.0% (137/249)
Sensitivity: 83.2% (114/137)
Specificity: 68.7% (77/112)
PPV: 76.5% (114/149)
NPV: 77.0% (77/100)

Tympanometric width > 300 daPa
Prevalence: 55.0% (137/249)
Sensitivity: 76.6% (105/137)
Specificity: 84.8% (95/112)
PPV: 86.1% (105/122)
NPV: 74.8% (95/127)
Selected population
Test done within 1 hour of the reference standard
Adequate description of test and reference standard
Blinding – Yes
Results calculated from the data given in the study
Ovesen (1993)42
{37329}
Diagnostic study

EL = I b
Children with unilateral or bilateral secretory OM fulfilling two of the following three criterion for surgical intervention – otomicroscopic findings consistent with SOM during 3 months, hearing impairment below 20 Db, and/or adenoid symptoms.
(n = 220, ears= 440)
Age range: 0.8 – 14.8 years
60% males, 40% females
Portable tympanometry by an ENT physician vs Myringotomy

Threshold: Type B and Type B or C2

Results also compared with otomicroscopy – but reference test not a standard one.
Type B as threshold
Prevalence: 87.0% (342/393)
Sensitivity: 90.6% (310/342)
Specificity: 72.6% (37/51)
PPV: 95.7% (310/324)
NPV: 53.6% (37/69)

Type B or C2 as threshold
Prevalence: 87.0% (342/393)
Sensitivity: 94.4% (323/342)
Specificity: 52.9% (27/51)
PPV: 93.1% (323/347)
NPV: 58.7% (27/46)
Selected population
Tests done immediately before the reference standard (exact timing not specified)
Adequate description of test and reference standard
Blinding – Yes
Paradise (1976)35
{37246}
Diagnostic study

EL = I b
Infants and children scheduled by physicians other than authors for myringotomy and insertion of tympanostomy tubes because of recurrent AOM or persistent MEE or both (n = 107, ears= 214)
Age range: 10 days – 5 years 11 month
62 males, 35 females
  1. Pneumatic otoscopy by a paediatrician vs Myringotomy
    Threshold: Present, absent or suspected OME
  2. Tympanometry by audiologist vs Myringotomy
    Threshold: Not defined
Comparison 1 (a) – ‘fluid suspected’ with OME as TP, and ‘fluid suspected’ without OME as FP
Prevalence: 64.9% (139/214)
Sensitivity: 97.8% (136/139)
Specificity: 74.7% (56/75)
PPV: 87.7% (136/155)
NPV: 94.9% (56/59)

Comparison 1 (b) – ‘fluid suspected’ with OME as FN, and ‘fluid suspected’ without OME as FP
Prevalence: 64.9% (139/214)
Sensitivity: 91.4% (127/139)
Specificity: 74.7% (56/75)
PPV: 87.0% (127/146)
NPV: 82.4% (56/68)
Selected population
Test and reference standard done within 2 hours
Adequate description of test and reference standard
Blinding – Yes
Data not extractable for tympanometry
Rosenfeld (1998)24
{37319}
Prospective study

EL = II
Consecutive children referred by paediatricians and family practitioners, and attending a hospital ENT practice with inclusion criterion:
  1. age – 6 months to 12 years;
  2. chronic otitis media (MEE in 1 or both ears for 3 months or longer) or recurrent otitis media (3 or more episodes of AOM in past 12 months);
  3. child accompanied by parent or primary caregiver; and
  4. child able to complete age- appropriate audiometry with good reliability (n = 186)
Age range: 6 months – 12 years
62% male
76% enrolled in managed care plans
  1. 6-item quality-of-life questionnaire survey (concerning perceived hearing status of children over past 4 weeks) completed by parents/caregiver vs correlation with Hearing loss evaluated through age-appropriate PTA (500, 1000, 2000 Hz) by a trained audiologist
  2. Change in caregiver assessment of hearing status after treatment vs correlation with change in PTA findings
  3. PTA findings (normal hearing with PTA average < 20 dB for better hearing ear) vs Middle ear status using a validated 4-point clinical profile based on otoscopy (TM grey, translucent and without fibrosis as normal); admittance (> 0.2 millimho as normal), and tympanometric gradient (< 150 daPa as normal)
4-point scale (for middle ear profile)
Level 1 – all 3 normal
Level 2 – normal otoscopy with one or both (admittance & gradient) as abnormal
Level 3 – abnormal otoscopy with both normal or 1 abnormal
Level 4 – all 3 abnormal
Comparison 1
Parent estimate of hearing vs median (range) hearing loss in dB on PTA
No problem – 23 (3–45)
Hardly a problem – 21 (3–45)
Somewhat a problem – 23 (5–47)
Moderate problem – 18 (2–35)
Quite a problem – 22 (3–50)
Very much a problem – 18 (3–40)
Extreme problem – 31 (12–52)

Spearman correlation(R) −0.13, P = 0.09

Comparison 2 (n = 50)
Median (range) change in parent response vs median (range) change in PTA
2 units (0–6) vs 11 dB (−23 to −35 dB)

Spearman correlation(R) 0.07, P = 0.65

Comparison 3
Level 1 vs normal hearing
Sensitivity: 17%
Specificity: 96%
PPV: 76%

Level 4 vs abnormal hearing
Sensitivity: 66%
Specificity: 82%
PPV: 84%
Selected population
Questionnaire – validated
Outcome assessed by trained personnel
Blinding – not specified
Sassen (1994)47
{37309}
Diagnostic study

EL = II
Hospital A: children undergoing insertion of ventilation tubes (indication – chronic OME i.e ≥ 3 months or recurrent OME, ears= 273)
Hospital B: children undergoing adeno-tonsillectomy with myringotomy (indication – recurrent URI or OME, ears= 242)
(total n = 266, total ears= 515)
Age range: 5 months – 11 years 5 months
Tympanometry (operator not specified) vs Myringotomy

Two different tympanometers used and interchanged between the hospitals after 6 weeks.

Threshold: Type B and Type B or C2
Type B as threshold
Combined results (n = 488)
Prevalence: 70.1% (342/488)
Sensitivity: 82.7% (283/342)
Specificity: 63.0% (92/146)
PPV: 84.0% (283/337)
NPV: 60.9% (92/151)

Age: 5 months – 2 years (n = 67)
Prevalence: 77.6% (52/67)
Sensitivity: 90.4% (47/52)
Specificity: 66.7% (10/15)
PPV: 90.4% (47/52)
NPV: 66.7% (10/15)

Age: 2 – 12 years (n = 421)
Prevalence: 68.9% (290/421)
Sensitivity: 81.4% (236/290)
Specificity: 62.6% (82/131)
PPV: 82.8% (236/285)
NPV: 60.3% (82/136)

Type B or C2 as threshold
Prevalence: 70.1% (342/488)
Sensitivity: 94.4% (323/342)
Specificity: 87.0% (127/146)
PPV: 94.4% (323/342)
NPV: 87.0% (127/146)
Selected population but different selection criterion followed in two hospitals
Test done within 1 hour of the reference standard
Adequate description of test and reference standard
Blinding – Yes
Data not extractable for different age groups with Type B or C2 as threshold
Shiao (2005)36
{37291}
Diagnostic study

EL = I b
Patients under 12 years of age admitted to the ward for VT insertion based on the presumptive diagnosis of OME or atelectasis of the eardrum (n = 104, ears= 201)
Age range: 1.5 – 12 years
69 boys, 35 girls
  1. Pneumatic otoscopy by an otolaryngologist vs Myringotomy
    Threshold: Presence or absence of OME
  2. Tympanometry by an audiologist vs Myringotomy
    Threshold: Type B
Comparison 1
Prevalence: 89.1% (179/201)
Sensitivity: 90.5% (162/179)
Specificity: 77.3% (17/22)
PPV: 97.0% (162/167)
NPV: 50.0% (17/34)

Comparison 2
Prevalence: 89.1% (179/201)
Sensitivity: 89.4% (160/179)
Specificity: 81.8% (18/22)
PPV: 97.6% (160/164)
NPV: 48.6% (18/37)
Selected population
Test and reference standard done within 48 hours
Adequate description of test and reference standard
Blinding – Yes
Stankiewicz (1979)26
{37520}
Diagnostic study

EL = II
Randomly selected patients from a clinic population complaining of hearing loss, tinnitus and/or vertigo.
(n and ears=variable for each test)
Age range: not specified
  1. Otoscopy by one of the authors vs PTA + Tympanometry done by second author as the reference standard
    Threshold: Normal or abnormal examination
  2. Tuning fork tests (Rinne and Weber) at 256, 512 and 1024 Hz by one of the authors vs Otoscopy + PTA + Tympanometry done by second author as the reference standard
    Threshold: Rinne negative for a positive test
    Weber – lateralized to bad ear for unilateral conductive loss
Comparison 1
Prevalence: 36.2% (58/160)
Sensitivity: 77.6% (45/58)
Specificity: 95.1% (97/102)
PPV: 90.0% (45/50)
NPV: 88.2% (97/110)

Comparison 2 (results for conductive deafness only)
Rinne test at 256 Hz
Prevalence: 29.2% (56/192)
Sensitivity: 42.9% (24/56)
Specificity: 99.3% (135/136)
PPV: 96.0% (24/25)
NPV: 80.8% (135/167)

Rinne test at 512 Hz
Prevalence: 29.2% (56/192)
Sensitivity: 16.1% (9/56)
Specificity: 99.3% (135/136)
PPV: 90.0% (9/10)
NPV: 74.2% (135/182)

Rinne test at 1024 Hz
Prevalence: 29.2% (56/192)
Sensitivity: 19.6% (11/56)
Specificity: 99.3% (135/136)
PPV: 91.7% (11/12)
NPV: 75.0% (135/180)

Weber test at 256 Hz (n = 28) (unilateral conductive loss)
Bad ear: 43%
Good ear: 25%
Mid-line: 32%

Weber test at 512 Hz (unilateral conductive loss)
Bad ear: 54%
Good ear: 21%
Mid-line: 25%

Weber test at 1024 Hz (unilateral conductive loss)
Bad ear: 46%
Good ear: 25%
Mid-line: 29%
Unselected population but age not specified
Tests and reference standard done immediately (exact time not specified)
Reference test – not a standard one
Blinding – Yes
Results calculated from the data given in the study
Tom (1994)46
{37241}
Diagnostic study

EL = I b
Patients scheduled to undergo myringotomies with pressure equalization tube insertion for either OME refractory to medical management or frequent recurrent OME
(n = 109, ears= 213)
Age range: 5 months – 11 years 5 months
62 male, 47 female

Exclusions: ears with small perforations
Tympanometry by a certified audiologist vs Myringotomy

Threshold: Type B and Type B or C2
Type B as threshold
Prevalence: 71.8% (153/213)
Sensitivity: 65.4% (100/153)
Specificity: 78.3% (47/60)
PPV: 88.5% (100/113)
NPV: 47.0% (47/100)

Type B or C2 as threshold
Prevalence: 71.8% (153/213)
Sensitivity: 94.8% (145/153)
Specificity: 23.3% (14/60)
PPV: 75.9% (145/191)
NPV: 63.6% (14/22)
Selected population
Test and reference standard done within 2 hours
Adequate description of test and reference standard
Blinding – Yes
Toner (1990)40
{37308}
Diagnostic study

EL = II
Patients admitted for myringotomy with indication in majority being clinically persistent MEE
(n = 121, ears= 222)
Age range: 18 months – 12 years

Exclusions: where both procedures could not be performed due to lack of cooperation
  1. Pneumatic otosopy by one of authors vs Myringotomy
    Threshold: Immobility for a positive test
  2. Tympanometry (operator not specified) vs Myringotomy
    Threshold: Type B
Comparison 1
Prevalence: 55.9% (124/222)
Sensitivity: 87.1% (108/124)
Specificity: 88.8% (87/98)
PPV: 90.7% (108/119)
NPV: 84.5% (87/103)

Comparison 2
Prevalence: 55.9% (124/222)
Sensitivity: 86.3% (107/124)
Specificity: 92.9% (91/98)
PPV: 93.9% (107/114)
NPV: 84.3% (91/108)
Selected population
Tests and reference standard done within 24 hours
Blinding – Not specified
Tests not described in details
van Balen (1994)43
{37286}
Diagnostic study

EL = I b
Children referred by GP’s for unior bilateral myringotomy and/or tympanostomy tube insertion.
(n = 142, ears= 284)
Age range: 7 months – 12 years

Exclusions: Children where tympanograms could not be performed or where surgery results not registered
Portable tympanometry by one of the authors vs Myringotomy

Threshold: Type B or C2 as positive test

Results also compared with tympanometry (professional) – but reference test not a standard one.
n = 233
Prevalence: 66.9% (156/233)
Sensitivity: 94.2% (147/156)
Specificity: 48.1% (37/77)
PPV: 78.6% (147/187)
NPV: 80.4% (37/46)
Selected population
Tests and reference standard done within 1 hour
Adequate description of test and reference standard
Blinding – Yes
Vaughan-Jones (1992)29
{37280}
Diagnostic study

EL = II
Children admitted for myringotomies with a diagnosis of OME
(n = 100, ears= 200)
56 male, 44 female
Mean age male – 6.3 years
Mean age female – 6.2 years
  1. Pneumatic otoscopy (operator not specified) vs Myringotomy
    Threshold: Effusion or aerated
  2. Pure tone air audiometry (PTA) at 500 Hz, 1 kHz, 2 kHz and 4 kHz (operator not specified) vs Myringotomy
    Threshold: Hearing loss ≥ 25 dB
  3. Tympanometry (operator not specified) vs Myringotomy
    Threshold: Type B or Type B/C2 as positive test
  4. Portable tympanometry (operator not specified) vs Myringotomy
    Threshold: Type B as positive test
Comparison 1
Prevalence: 67.5% (135/200)
Sensitivity: 89.6% (121/135)
Specificity: 75.4% (49/65)
PPV: 88.3% (121/137)
NPV: 77.8% (49/63)

Comparison 2 (at 500 Hz)
Prevalence: 67.5% (135/200)
Sensitivity: 68.2% (92/135)
Specificity: 84.6% (55/65)
PPV: 90.2% (92/102)
NPV: 56.1% (55/98)

Comparison 2 (at 1 kHz)
Prevalence: 67.5% (135/200)
Sensitivity: 59.3% (80/135)
Specificity: 93.8% (61/65)
PPV: 95.2% (80/84)
NPV: 52.6% (61/116)

Comparison 2 (at 2 kHz)
Prevalence: 67.5% (135/200)
Sensitivity: 32.6% (44/135)
Specificity: 95.4% (62/65)
PPV: 93.6% (44/47)
NPV: 40.5% (62/153)

Comparison 2 (at 4 kHz)
Prevalence: 67.5% (135/200)
Sensitivity: 46.7% (63/135)
Specificity: 93.8% (61/65)
PPV: 94.0% (63/67)
NPV: 45.9% (61/133)

Comparison 3 (Type B as threshold)
Prevalence: 67.5% (135/200)
Sensitivity: 67.4% (91/135)
Specificity: 93.8% (61/65)
PPV: 95.8% (91/95)
NPV: 58.1% (61/105)

Comparison 3 (Type B/C2 as threshold)
Prevalence: 67.5% (135/200)
Sensitivity: 88.9% (120/135)
Specificity: 63.1% (41/65)
PPV: 83.3% (120/144)
NPV: 73.2% (41/56))
Selected population
All tests done within 24 hours of the reference standard.
Data not extractable for portable tympanometry and acoustic otoscopy Blinding – Yes for pneumatic otoscopy, and not specified for others.
Tests not described in details.
Watters (1997)48
{37310}
Diagnostic study

EL = II
Children undergoing surgery for suspected MEE
(n = 501, ears= 955)

Exclusions: children whose surgery was cancelled due to normal tympanograms
Tympanometry by a paediatric audiologist vs Myringotomy

Threshold: Type B
Prevalence: 78.0% (745/955)
Sensitivity: 91.1% (679/745)
Specificity: 79.0% (166/210)
PPV: 93.9% (679/723)
NPV: 71.6% (166/232
Selected population
Test and reference standard done within 2 hours
Adequate description of test and reference standard
Blinding – Not specified
Yung (1981)28
{37317}
Diagnostic study

EL = III
Children admitted for myringotomy
(n = 100, ears=not specified)
Age range: 2 – 12 years
Tuning fork tests (Rinne and Weber) at 512 Hz
(operator not specified) vs Myringotomy

Threshold: Rinne negative for a positive test.
Weber – referred or not referred
Rinne test – results for both unilateral and bilateral effusion
Prevalence: 88.3% (83/94)
Sensitivity: 89.2% (74/83)
Specificity: 72.7% (8/11)
PPV: 96.1% (74/77)
NPV: 47.1% (8/17)

Weber test – results for unilateral effusion(n = 40)
Prevalence: 72.5% (29/40)
Sensitivity: 79.3% (23/29)
Specificity: 90.9% (10/11)
PPV: 95.8% (23/24)
NPV: 62.5% (10/16)
Selected population
Time interval between test and reference standard not specified
Test and Reference test – not described in details
Blinding – Not specified
Results calculated from the data given in the study

From: Evidence tables

Cover of Surgical Management of Otitis Media with Effusion in Children
Surgical Management of Otitis Media with Effusion in Children.
NICE Clinical Guidelines, No. 60.
National Collaborating Centre for Women’s and Children’s Health (UK).
London: RCOG Press; 2008 Feb.
Copyright © 2008, National Collaborating Centre for Women’s and Children’s Health.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK [www.cla.co.uk]. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.