The Journal of Laryngology and Otology July 1992, Vol. 106, pp. 600-602
The Welch Allyn Audioscope and Microtymp: their accuracy and that of pneumatic otoscopy, tympanometry and pure tone audiometry as predictors of otitis media with effusion R.
VAUGHAN-JONES,
F.R.C.S.Ed., F.R.C.S., R. P. MILLS, M.Phil., F.R.C.S. (Dundee)
Abstract Results of standard audiometry and tympanometry were compared with the Welch Allyn 'Audioscope' and 'Microtymp', in the diagnosis of secretory otitis media, in 100 children. Standard pure tone audiometry had a specificity of 92 per cent and sensitivity of 51.6 per cent. The Welch Allyn 'Audioscope' a specificity of 84.2 per cent and sensitivity of 57.5 per cent; standard tympanometry a specificity of 71 per cent and sensitivity of 88 per cent; the Welch Allyn 'Microtymp' a specificity of 63 per cent and sensitivity of 90 per cent. In view of cost, portability, and speed of testing the Welch Allyn instruments would be particularly suitable for community screening. Pure tone audiometry with a pass threshold of 25 dB, at 2 and 4 kHz is a poor indicator of effusion.
Introduction Previous studies (Grimaldi, 1976) have evaluated methods for the diagnosis of secretory otitis media both separately and in combination to attempt to increase accuracy of diagnosis. Recently two devices have been introduced, the Welch Allyn 'Audioscope' and 'Microtymp'. Advantages over standard audiometric test being portability, speed of use and cost. The object of this study was to assess their accuracy in predicting otitis media with effusion (OME) as compared to existing instruments. The accuracy of pneumatic otoscopy, pure tone audiometry, and tympanometry was also assessed. An attempt has been made to combine test results so as to maximize their predictive value.
(1 daPa = 1.04 mm H2O), (with a 226 Hz probe tone and pump speed of 200 ±20 daPa/second). Tympanometric peak pressures were recorded and assigned using the A, Cl, C2, and B classification of Jerger (1970), 'A' representing a tympanogram peak between +100 to -100 mm H 2 O,'Cl',-100to-200 mm H 2 O,'C2',-200 to -400 mm H2O and 'B' a flat tympanogram. Subjects had air conduction audiometry over four frequencies 500 Hz, 1 kHz and 4 kHz, using a Madsen OB822 audiometer in a sound proof booth. Failure at 25 dB was considered pathological. Tympanometry was performed using a American Electric Model 85 tympanometer over a pressure change of+200 to -400 mm H2O with a 256 Hz probe. Myringotomies were performed under general anaesthesia within 24 hours of otoscopy and audiometric testing. Only patients undergoing adenoidectomy were intubated. Anterior myringotomies were performed and a fine sucker passed into the middle ear, any amount of fluid found was considered to indicate an effusion. Logistical regression was used to combine different test systems in an attempt to increase sensitivity and specificity.
Method One hundred and twelve children were admitted with a diagnosis of otitis media with effusion; 12 were excluded due to an inadequate view of the tympanic membrane or variable audiometric responses. Of the 100 children included, 56 were male (mean age 6.3 years) and 44 female (mean age 6.2 years). The examiner not knowing the results of audiometry and tympanometry, examined ears using a Welch Allyn pneumatic otoscope. They were designated to have either effusions or aerated middle ears. The Welch Allyn 'Audioscope' was tested in a quiet ward side room. A screening level of 25 dB was used over the frequencies of 500 Hz, 1 kHz, 2 kHz and 4 kHz. Patients were asked to respond verbally or by using gross motor responses. The Welch Allyn 'Microtymp' was tested over a pressure change of +200 to -300 daPa
Results One hundred and thirty-five (67.5 per cent) ears were found to contain effusions and 65 (32.5 per cent) were dry at myringotomy. Pneumatic otoscopy correctly predicted myringotomy results in 170 (85 percent) ears, giving a specificity of 75 per cent and sensitivity of 90 per cent (Table I). Standard pure tone audiometry, correctly predicted myringotomy results at 500 Hz in 147 (73.5 per cent), at 1 kHz in 141 (70.5 per cent), at 2 kHz in 106 (53 percent),
Accepted for publication: 28 March 1992. 600
601
THE WELCH ALLYN AUDIOSCOPE AND MICROTYMP
TABLE I OPERATIVE FINDING VS DIAGNOSIS USING PNEUMATIC OTOSCOPY
Pre-operative prediction by pneumatic otoscopy Operative rinding
Aerated
Effusion
49 14
16 121
Dry Effusion
and at 4 kHz in 126 (63 per cent) of ears. This gave a specificity of 92 per cent and sensitivity of 52 per cent (Table II). The 'Audioscope' correctly predicted myringotomy results at 500 Hz in 150 (75 per cent), at 1 kHz in 148 (74 per cent), at 2 kHz in 106 (53 per cent), and at 4 kHz in 126 (63 per cent) of ears. This gave a specificity of 84 per cent and sensitivity of 57 per cent. Standard tympanometry gave correct predictions of myringotomy when type A and Cl were considered to indicate dry ears, and C2 and B tympanograms to indicate effusions, in 165 (82.5 per cent) of ears. This gave a specificity of 71 per cent and sensitivity of 88 per cent (Table III). The Welch Allyn 'Microtymp' gave a correct result in 161 (81 per cent) of ears, with a specificity of 63 per cent and sensitivity of 90 per cent. Using logistic regression analysis to combine results from each test system, the inclusion of pure tone thresholds at either 500 Hz or 4 kHz with pneumatic otoscopy and tympanometry increased specificity to 83 per cent; however sensitivity remained at 90 per cent. Discussion Two kinds of error can occur in testing. False positive which lead to alarm and unnecessary treatment, and false negative which lead to failure to diagnose a disease when present. The sensitivity of a test reflects true positives correctly diagnosed. The specificity of a test indicates the proportion of true negatives correctly identified. Pneumatic Otoscopy Sensitivity rates of 94 per cent (detection of fluid when fluid is present), and specificity rates (detection of no fluid when there is no fluid present) of between 74-78 per cent have been given for experienced otologists in the diagnosis of middle ear effusions using pneumatic otoscopy alone (Cantekin et al., 1980). In this study pneumatic otoscopy was better than either pure tone audiometry or tympanometry in predicting effusions. One hundred and TABLE II OPERATIVE FINDING VS PREOPERATIVE PURE TONE AUDIOMETRIC THRESHOLD AT 0.5, 1, 2 AND 4 K H Z . ( > 2 6 DB CONSIDERED TO INDICATE EFFUSIONS)
seventy ears (85 per cent) were correctly diagnosed, giving a sensitivity of 89.6 per cent and specificity of 75.3 per cent. Amongst the 16 false positives, eight had pure tone thresholds better than 25 dB at all frequencies and type A or Cl tympanograms. It is unlikely that these eight ears would have been subjected to myringotomy had audiometric data been reviewed. In six ears and results of both audiometry and tympanometry agreed with the false impression of effusion given by pneumatic otoscopy. Thresholds greater than 25 dB occured in, one at one frequency, two at three frequencies (0.5, 1 and 4 kHz) and three at all frequencies. Amongst this group one type C2 and five type B tympanograms were found. Explanations for false positive results include displacement of fluid into the Eustachian tube by anaesthetic gases, or positional displacement into the mastoid cortex. Both mechanisms may be possible with serous effusions though seem less likely in the more common mucoid effusion. Further suggestions include loculation within the middle ear; a more likely possibility is that these ears represent either Eustachian tube dysfunction diagnosed prior to the onset of effusion, or ears subjected to surgery during recovery from effusion. The latter explanation may account for nine ears predicted dry by pneumatic otoscopy, with thresholds better than 25 dB and type A or C1 tympanograms, but found to have effusions. Cantekin et al. (1980) found inclusion of tympanometry with results of pneumatic otoscopy increased sensitivity to 97 per cent and specificity to 90 per cent. Mills (1987) confirmed the complementary value of both techniques resulting in a sensitivity of 92.6 per cent and specificity of 95.1 percent. In contrast Toner and Mains (1990) found no increase in predictive value when results of tympanometry and pneumatic otoscopy were included together. In this study using logistic regression analysis, the combination of both tympanometry and pneumatic otoscopy increased sensitivity to 90 per cent and specificity to 80 per cent. Pure tone Audiometry Pure tone audiometry had a specificity of 92 per cent and sensitivity of 51.6 per cent. Of ears with effusions, thresholds better than 25 dB occurred in, 43 (32 per cent) at 500 Hz, 55 (41 per cent) at 1 kHz, 72 (53 per cent) at 4 kHz and 91 (67 per cent) at 2 kHz. At 2 and 4 kHz pure tone audiometry is a poor indicator of middle ear effusions. Sensitivity was greatest at 500 Hz, unfortunately a frequency most adversely affected by background noise. This may explain the poorer specificity (84 per cent) seen with the Welch Allyn 'Audioscope', which was tested in a ward environment. In a similar study, Bluestone et al. (1973) found that despite proven middle ear effusions, half of the children TABLE III
Puretone threshold 500 Hz Operative rinding Dry Effusion
26 26 26 26 4 63
Standard tympanometry Operative rinding Dry Effusion
23
6
Cl
C2
23 JO
15 28
4 91
602 tested with standard puretone audiometry passed the 25 dB threshold. Brooks (1974) also highlighted deficiencies of pure tone audiometry, finding tympanometry to detected probable conductive pathology in 14 per cent of cases passed by pure tone audiometry. In this study, 43 (32 per cent) ears with effusions had thresholds equal or better than 25 dB at three frequencies, 31 of these ears had type C2 or B tympanograms. Many children have better hearing than the calibrated audiometric zero and thus pass audiometric screening, 25 dB may therefore be too high a pass level to reliably detect middle ear effusions. An advantage afforded to the 'Audioscope' is its speed, testing all four frequencies in 12 seconds at either 20,25 or 40 dB, this is useful in children with poor attention spans. In view of the poorer reliability of pure tone in predicting effusions, this less sophisticated method of testing may be more valuable in combination with tympanometry as a screening tool. Tympanometry Tympanometry agreed with myringotomy in 165 (83 per cent) of ears, inaccuracies arise with Cl and C2 tympanograms. From this study it apears that C2 tympanogram should be taken to indicate effusion. Forty-three ears had C2 tracings, 28 (65 per cent) had middle ear effusions. Of the 33 ears with Cl tympanograms, 10 (30 per cent) had effusions. This result contradicts Toner and Mains (1990) who included all type C tympanograms with type A tympanograms, and still found a predictive rate for tympanometry of 89 per cent. In this study inclusion of both type Cl and C2 tympanograms would reduce sensitivity to 67 per cent but raise specificity to 94 per cent. Other methods to improve the predictive value of tympanometry have been attempted. Paradise et al. (1976) described 15 tympanometric profiles by taking into account tympanometric gradients. Rounded curves being more likely associated with effusions than sharply peaked traces. Gates et al. (1988) used this more complicated classification but still found a false positive rate of 17 per cent and were unable to give a satisfactory explanation for their results. The 'Microtymp' had a lower specificity (63 per cent) compared to standard tympanometry (71 per cent), this was due tofivefalse positive B tracings, no definite reason has been established for this error. Tympanometric gradients are displayed on the 'Microtymp', increased accuracy may have been obtained had these results been used. Further advantages offered by both machines are portability and convenience in view of their small size. They
R. VAUGHAN-JONES, R. P. MILLS
are easy to operate and calibrate, the cost of both 'Audioscope' and 'Microtymp' including recharger and graph plotter is just over £2,000, which compare well with standard tympanometers and audiometers. Conclusion Pneumatic otoscopy gave the greatest accuracy for the diagnosis of otitis media with effusion. The Welch Allyn 'Audioscope' and 'Microtymp' compare well with standard audiometric equipment. Pure tone audiometry particularly at 2 and 4 kHz using a threshold of 25 dB is a poor indicator of middle ear effusion. Acknowledgements The assistance of the staff at the audiology department at Ninewells Hospital is greatly acknowledged, as is that of Mr S. Ogden at the Department of Medical Computing for his help with the statistics. References Bluestone, C, Berry, Q. C, Paradise, J. L. (1973). Audiometry and tympanometry in reaction to middle ear effusions in children. Laryngoscope, 83: 594-604. Brooks, D. N. (1974) Impedance bridge studies on normal hearing and hearing impaired children. Acta Otorhinolaryngolica Belgica, 28: 140-145. Cantekin, E. L., Bluestone, C. D., Fria, T. J., Stool, S. E., Berry, Q. C, Sabo, D. L. (1980) Identification of otitis media with effusion in children. Annals of Otology, Rhinology and Laryngology, 89 (Supplement 68), 190-195. Gates, A. G., Avery, C, Cooper, J. C, Hearne, E. M., Holt, R. G. (1986) Predictive value of tympanometry in middle ear effusion. Annals of Otology, Rhinology and Laryngology, 95: 46-50. Grimaldi, P. M. (1976) The value of impedance testing in diagnosis of middle ear effusion. Journal of Laryngology and Otology, 90: 141-152. Jerger, J. (1970) Clinical experience with impedance audiometry. Archives of Otolaryngology, 92: 311-324. Mills, R. P. (1987) Persistent middle ear effusions in children with recurrent acute otitis media. Clinical Otolaryngology, 12: 97-101. Paradise, J. L., Smith, G. C, Bluestone, C. D. (1976) Tympanometric detection of middle ear effusion in infants and children. Pediatrics, 58: 198-210. Toner, J. G., Mains, B. (1990) Pneumatic otoscopy and tympanometry int he detection of middle ear effusion. Clinical Otolaryngology, 15: 121-123. Address for correspondence: R. Vaughan-Jones, Department of Otolaryngology, Ninewells Hospital, Dundee DD1 9SY.
Key words: Otitis media with effusion; Audiometry, hearing tests, acoustic impedance tests
The Welch Allyn Audioscope and Microtymp: their accuracy and that of pneumatic otoscopy, tympanometry and pure tone audiometry as predictors of otitis media with effusion R.
VAUGHAN-JONES,
F.R.C.S.Ed., F.R.C.S., R. P. MILLS, M.Phil., F.R.C.S. (Dundee)
Abstract Results of standard audiometry and tympanometry were compared with the Welch Allyn 'Audioscope' and 'Microtymp', in the diagnosis of secretory otitis media, in 100 children. Standard pure tone audiometry had a specificity of 92 per cent and sensitivity of 51.6 per cent. The Welch Allyn 'Audioscope' a specificity of 84.2 per cent and sensitivity of 57.5 per cent; standard tympanometry a specificity of 71 per cent and sensitivity of 88 per cent; the Welch Allyn 'Microtymp' a specificity of 63 per cent and sensitivity of 90 per cent. In view of cost, portability, and speed of testing the Welch Allyn instruments would be particularly suitable for community screening. Pure tone audiometry with a pass threshold of 25 dB, at 2 and 4 kHz is a poor indicator of effusion.
Introduction Previous studies (Grimaldi, 1976) have evaluated methods for the diagnosis of secretory otitis media both separately and in combination to attempt to increase accuracy of diagnosis. Recently two devices have been introduced, the Welch Allyn 'Audioscope' and 'Microtymp'. Advantages over standard audiometric test being portability, speed of use and cost. The object of this study was to assess their accuracy in predicting otitis media with effusion (OME) as compared to existing instruments. The accuracy of pneumatic otoscopy, pure tone audiometry, and tympanometry was also assessed. An attempt has been made to combine test results so as to maximize their predictive value.
(1 daPa = 1.04 mm H2O), (with a 226 Hz probe tone and pump speed of 200 ±20 daPa/second). Tympanometric peak pressures were recorded and assigned using the A, Cl, C2, and B classification of Jerger (1970), 'A' representing a tympanogram peak between +100 to -100 mm H 2 O,'Cl',-100to-200 mm H 2 O,'C2',-200 to -400 mm H2O and 'B' a flat tympanogram. Subjects had air conduction audiometry over four frequencies 500 Hz, 1 kHz and 4 kHz, using a Madsen OB822 audiometer in a sound proof booth. Failure at 25 dB was considered pathological. Tympanometry was performed using a American Electric Model 85 tympanometer over a pressure change of+200 to -400 mm H2O with a 256 Hz probe. Myringotomies were performed under general anaesthesia within 24 hours of otoscopy and audiometric testing. Only patients undergoing adenoidectomy were intubated. Anterior myringotomies were performed and a fine sucker passed into the middle ear, any amount of fluid found was considered to indicate an effusion. Logistical regression was used to combine different test systems in an attempt to increase sensitivity and specificity.
Method One hundred and twelve children were admitted with a diagnosis of otitis media with effusion; 12 were excluded due to an inadequate view of the tympanic membrane or variable audiometric responses. Of the 100 children included, 56 were male (mean age 6.3 years) and 44 female (mean age 6.2 years). The examiner not knowing the results of audiometry and tympanometry, examined ears using a Welch Allyn pneumatic otoscope. They were designated to have either effusions or aerated middle ears. The Welch Allyn 'Audioscope' was tested in a quiet ward side room. A screening level of 25 dB was used over the frequencies of 500 Hz, 1 kHz, 2 kHz and 4 kHz. Patients were asked to respond verbally or by using gross motor responses. The Welch Allyn 'Microtymp' was tested over a pressure change of +200 to -300 daPa
Results One hundred and thirty-five (67.5 per cent) ears were found to contain effusions and 65 (32.5 per cent) were dry at myringotomy. Pneumatic otoscopy correctly predicted myringotomy results in 170 (85 percent) ears, giving a specificity of 75 per cent and sensitivity of 90 per cent (Table I). Standard pure tone audiometry, correctly predicted myringotomy results at 500 Hz in 147 (73.5 per cent), at 1 kHz in 141 (70.5 per cent), at 2 kHz in 106 (53 percent),
Accepted for publication: 28 March 1992. 600
601
THE WELCH ALLYN AUDIOSCOPE AND MICROTYMP
TABLE I OPERATIVE FINDING VS DIAGNOSIS USING PNEUMATIC OTOSCOPY
Pre-operative prediction by pneumatic otoscopy Operative rinding
Aerated
Effusion
49 14
16 121
Dry Effusion
and at 4 kHz in 126 (63 per cent) of ears. This gave a specificity of 92 per cent and sensitivity of 52 per cent (Table II). The 'Audioscope' correctly predicted myringotomy results at 500 Hz in 150 (75 per cent), at 1 kHz in 148 (74 per cent), at 2 kHz in 106 (53 per cent), and at 4 kHz in 126 (63 per cent) of ears. This gave a specificity of 84 per cent and sensitivity of 57 per cent. Standard tympanometry gave correct predictions of myringotomy when type A and Cl were considered to indicate dry ears, and C2 and B tympanograms to indicate effusions, in 165 (82.5 per cent) of ears. This gave a specificity of 71 per cent and sensitivity of 88 per cent (Table III). The Welch Allyn 'Microtymp' gave a correct result in 161 (81 per cent) of ears, with a specificity of 63 per cent and sensitivity of 90 per cent. Using logistic regression analysis to combine results from each test system, the inclusion of pure tone thresholds at either 500 Hz or 4 kHz with pneumatic otoscopy and tympanometry increased specificity to 83 per cent; however sensitivity remained at 90 per cent. Discussion Two kinds of error can occur in testing. False positive which lead to alarm and unnecessary treatment, and false negative which lead to failure to diagnose a disease when present. The sensitivity of a test reflects true positives correctly diagnosed. The specificity of a test indicates the proportion of true negatives correctly identified. Pneumatic Otoscopy Sensitivity rates of 94 per cent (detection of fluid when fluid is present), and specificity rates (detection of no fluid when there is no fluid present) of between 74-78 per cent have been given for experienced otologists in the diagnosis of middle ear effusions using pneumatic otoscopy alone (Cantekin et al., 1980). In this study pneumatic otoscopy was better than either pure tone audiometry or tympanometry in predicting effusions. One hundred and TABLE II OPERATIVE FINDING VS PREOPERATIVE PURE TONE AUDIOMETRIC THRESHOLD AT 0.5, 1, 2 AND 4 K H Z . ( > 2 6 DB CONSIDERED TO INDICATE EFFUSIONS)
seventy ears (85 per cent) were correctly diagnosed, giving a sensitivity of 89.6 per cent and specificity of 75.3 per cent. Amongst the 16 false positives, eight had pure tone thresholds better than 25 dB at all frequencies and type A or Cl tympanograms. It is unlikely that these eight ears would have been subjected to myringotomy had audiometric data been reviewed. In six ears and results of both audiometry and tympanometry agreed with the false impression of effusion given by pneumatic otoscopy. Thresholds greater than 25 dB occured in, one at one frequency, two at three frequencies (0.5, 1 and 4 kHz) and three at all frequencies. Amongst this group one type C2 and five type B tympanograms were found. Explanations for false positive results include displacement of fluid into the Eustachian tube by anaesthetic gases, or positional displacement into the mastoid cortex. Both mechanisms may be possible with serous effusions though seem less likely in the more common mucoid effusion. Further suggestions include loculation within the middle ear; a more likely possibility is that these ears represent either Eustachian tube dysfunction diagnosed prior to the onset of effusion, or ears subjected to surgery during recovery from effusion. The latter explanation may account for nine ears predicted dry by pneumatic otoscopy, with thresholds better than 25 dB and type A or C1 tympanograms, but found to have effusions. Cantekin et al. (1980) found inclusion of tympanometry with results of pneumatic otoscopy increased sensitivity to 97 per cent and specificity to 90 per cent. Mills (1987) confirmed the complementary value of both techniques resulting in a sensitivity of 92.6 per cent and specificity of 95.1 percent. In contrast Toner and Mains (1990) found no increase in predictive value when results of tympanometry and pneumatic otoscopy were included together. In this study using logistic regression analysis, the combination of both tympanometry and pneumatic otoscopy increased sensitivity to 90 per cent and specificity to 80 per cent. Pure tone Audiometry Pure tone audiometry had a specificity of 92 per cent and sensitivity of 51.6 per cent. Of ears with effusions, thresholds better than 25 dB occurred in, 43 (32 per cent) at 500 Hz, 55 (41 per cent) at 1 kHz, 72 (53 per cent) at 4 kHz and 91 (67 per cent) at 2 kHz. At 2 and 4 kHz pure tone audiometry is a poor indicator of middle ear effusions. Sensitivity was greatest at 500 Hz, unfortunately a frequency most adversely affected by background noise. This may explain the poorer specificity (84 per cent) seen with the Welch Allyn 'Audioscope', which was tested in a ward environment. In a similar study, Bluestone et al. (1973) found that despite proven middle ear effusions, half of the children TABLE III
Puretone threshold 500 Hz Operative rinding Dry Effusion
26 26 26 26 4 63
Standard tympanometry Operative rinding Dry Effusion
23
6
Cl
C2
23 JO
15 28
4 91
602 tested with standard puretone audiometry passed the 25 dB threshold. Brooks (1974) also highlighted deficiencies of pure tone audiometry, finding tympanometry to detected probable conductive pathology in 14 per cent of cases passed by pure tone audiometry. In this study, 43 (32 per cent) ears with effusions had thresholds equal or better than 25 dB at three frequencies, 31 of these ears had type C2 or B tympanograms. Many children have better hearing than the calibrated audiometric zero and thus pass audiometric screening, 25 dB may therefore be too high a pass level to reliably detect middle ear effusions. An advantage afforded to the 'Audioscope' is its speed, testing all four frequencies in 12 seconds at either 20,25 or 40 dB, this is useful in children with poor attention spans. In view of the poorer reliability of pure tone in predicting effusions, this less sophisticated method of testing may be more valuable in combination with tympanometry as a screening tool. Tympanometry Tympanometry agreed with myringotomy in 165 (83 per cent) of ears, inaccuracies arise with Cl and C2 tympanograms. From this study it apears that C2 tympanogram should be taken to indicate effusion. Forty-three ears had C2 tracings, 28 (65 per cent) had middle ear effusions. Of the 33 ears with Cl tympanograms, 10 (30 per cent) had effusions. This result contradicts Toner and Mains (1990) who included all type C tympanograms with type A tympanograms, and still found a predictive rate for tympanometry of 89 per cent. In this study inclusion of both type Cl and C2 tympanograms would reduce sensitivity to 67 per cent but raise specificity to 94 per cent. Other methods to improve the predictive value of tympanometry have been attempted. Paradise et al. (1976) described 15 tympanometric profiles by taking into account tympanometric gradients. Rounded curves being more likely associated with effusions than sharply peaked traces. Gates et al. (1988) used this more complicated classification but still found a false positive rate of 17 per cent and were unable to give a satisfactory explanation for their results. The 'Microtymp' had a lower specificity (63 per cent) compared to standard tympanometry (71 per cent), this was due tofivefalse positive B tracings, no definite reason has been established for this error. Tympanometric gradients are displayed on the 'Microtymp', increased accuracy may have been obtained had these results been used. Further advantages offered by both machines are portability and convenience in view of their small size. They
R. VAUGHAN-JONES, R. P. MILLS
are easy to operate and calibrate, the cost of both 'Audioscope' and 'Microtymp' including recharger and graph plotter is just over £2,000, which compare well with standard tympanometers and audiometers. Conclusion Pneumatic otoscopy gave the greatest accuracy for the diagnosis of otitis media with effusion. The Welch Allyn 'Audioscope' and 'Microtymp' compare well with standard audiometric equipment. Pure tone audiometry particularly at 2 and 4 kHz using a threshold of 25 dB is a poor indicator of middle ear effusion. Acknowledgements The assistance of the staff at the audiology department at Ninewells Hospital is greatly acknowledged, as is that of Mr S. Ogden at the Department of Medical Computing for his help with the statistics. References Bluestone, C, Berry, Q. C, Paradise, J. L. (1973). Audiometry and tympanometry in reaction to middle ear effusions in children. Laryngoscope, 83: 594-604. Brooks, D. N. (1974) Impedance bridge studies on normal hearing and hearing impaired children. Acta Otorhinolaryngolica Belgica, 28: 140-145. Cantekin, E. L., Bluestone, C. D., Fria, T. J., Stool, S. E., Berry, Q. C, Sabo, D. L. (1980) Identification of otitis media with effusion in children. Annals of Otology, Rhinology and Laryngology, 89 (Supplement 68), 190-195. Gates, A. G., Avery, C, Cooper, J. C, Hearne, E. M., Holt, R. G. (1986) Predictive value of tympanometry in middle ear effusion. Annals of Otology, Rhinology and Laryngology, 95: 46-50. Grimaldi, P. M. (1976) The value of impedance testing in diagnosis of middle ear effusion. Journal of Laryngology and Otology, 90: 141-152. Jerger, J. (1970) Clinical experience with impedance audiometry. Archives of Otolaryngology, 92: 311-324. Mills, R. P. (1987) Persistent middle ear effusions in children with recurrent acute otitis media. Clinical Otolaryngology, 12: 97-101. Paradise, J. L., Smith, G. C, Bluestone, C. D. (1976) Tympanometric detection of middle ear effusion in infants and children. Pediatrics, 58: 198-210. Toner, J. G., Mains, B. (1990) Pneumatic otoscopy and tympanometry int he detection of middle ear effusion. Clinical Otolaryngology, 15: 121-123. Address for correspondence: R. Vaughan-Jones, Department of Otolaryngology, Ninewells Hospital, Dundee DD1 9SY.
Key words: Otitis media with effusion; Audiometry, hearing tests, acoustic impedance tests
