A 14-Year Prospective Follow-up Study of Children Treated Early in Life With Tympanostomy Tubes: Part 2: Hearing Outcomes

Objective  To determine hearing outcomes in young children receiving early and repeated tympanostomy tube insertion for recurrent acute otitis media or otitis media with effusion.

Design  Prospective 14-year follow-up.

Setting  Central Hospital of Central Finland, a tertiary care hospital.

Patients  Three hundred five consecutive infants and young children with otitis media received initial tympanostomy tube insertion at the age of 5 to 16 months. The final study group comprised 237 patients (77.7%) attending the 14-year checkups.

Main Outcome Measures  At the 14-year checkups, children received clinical examinations and audiometric testing for the determination of bone and air conduction pure-tone thresholds.

Results  The mean pure-tone average of 177 healed ears was 4.3 dB. The mean pure-tone average of all ears was 5.8 dB, with significantly poorer results in ears with abnormal outcomes such as grade II or higher pars tensa retraction, otitis media with effusion, and tympanic membrane perforation. Thirteen (5.5%) of 237 ears had a hearing level worse than 15 dB, and the better ear hearing level was poorer than 15 dB in 3 patients.

Conclusions  The hearing level of healed ears was comparable to that of age-matched normal ears. Hearing losses were infrequent, of slight grade, and, when present, almost exclusively conductive and related to unsuccessful otological outcomes. From the hearing point of view, repeated tympanostomy tube insertion for recurrent acute otitis media or otitis media with effusion early in life is a safe treatment.

Instant improvement of hearing is one of the beneficial effects of tympanostomy tube insertion and aspiration of middle ear effusion (MEE) in primary cases of otitis media with effusion (OME).¹ Insertion of ventilation tubes (VTs) reduces the prevalence of MEE and produces a short-term benefit.² Long-term advantages of VT insertion in reducing the risk of subsequent linguistic, educational, and developmental impairments due to OME appear less uniform in different studies.^3-6 Five-year follow-up results after tympanostomy tube insertion showed that hearing in general was well preserved, but in one third of the ears the otological outcomes were abnormal, necessitating frequent follow-up visits or further treatment.⁷ Because of this finding and the chronic nature of OME, there is a need for longer follow-up studies concerning hearing outcomes after tympanostomy tube insertion. This 14-year follow-up study discusses the audiological findings in an effort to determine the relevance of early and repeated tympanostomy tube insertion in a homogeneous group of young children with OME or recurrent acute otitis media (RAOM).

The details of the study procedures have been described.^7-9 The study commenced in 1983 at the Department of Otorhinolaryngology of Central Hospital of Central Finland. Three hundred five consecutive children with otitis media (OM) were enrolled for early initial tympanostomy tube insertion during 1983 and 1984. Inclusion criteria were (1) a minimum of 2 episodes of RAOM within 3 months (RAOM group) or (2) signs of OME persisting after a minimum of 2 months (OME group). Children with a cleft palate or other major congenital anomalies were excluded, as were those older than 16 months and those with prior tympanostomy tube insertion or adenoidectomy.

During the first 5 years, the children were examined by otorhinolaryngologists at 3-month intervals while VTs were present and at 6-month intervals following their extrusion. All OM episodes and any subsequent VT insertions were recorded; adenoidectomy was performed with the first reinsertion of VT.^7,8

According to the study protocol, the follow-up of these children was continued for a minimum of 14 years and ended with a final checkup and audiological testing. The treatment principles were the same as those during the 5-year follow-up.^7,8 The otological findings are reported in this issue.⁹

The 237 patients who attended the 14-year checkup formed the final study group.⁹ At the 14-year checkup, the children were examined using pneumatic otoscope and otomicroscope by the same senior otorhinolaryngologist (H.V.) according to the same principles as at the 5-year checkup.⁷ Abnormal otological findings (tympanosclerosis or atrophy of the tympanic membrane [TM], retractions of the pars flaccida [PF] and pars tensa [PT], presence of MEE, or perforations of the TM) and the localization of changes in different quadrants of the PT (quadrant 1, anterosuperior; quadrant 2, anteroinferior; quadrant 3, posteroinferior; or quadrant 4, posterosuperior) were recorded. Whenever MEE was suspected, its presence was confirmed by myringotomy and aspiration. The otological findings were classified into one of the following groups: TM perforation, VT in place, OME, retraction of the PT,¹⁰ retraction of the PF,¹¹ or healed. This classification was exclusive, with the order from TM perforation to healed.⁹

Two experienced audiometricians performed the pure-tone audiometric measurements of air conduction (AC) and bone conduction (BC) thresholds. The regular ascending-descending techniques and rules of masking were used. The investigation was conducted in a silent chamber with a Madsen OB 822 clinical audiometer (Herlev, Denmark), calibrated in accord with the ISO 389 standards (International Organization for Standardization, Geneva, Switzerland). A set of TDH 39 earphones (Telephonics Corp, Farmingdale, NY) mounted in rubber cushions and insert receiver masking were used. A standard BC vibrator was used. The AC and BC pure-tone hearing thresholds were determined in all children and recorded in decibels for each octave frequency (range, 0.25-4.0 kHz for BC and 0.25-8.0 kHz for AC). The pure-tone average (PTA) for AC was calculated as the mean of the thresholds at 0.5, 1.0, and 2.0 kHz. The BC averages for the same frequencies were used for calculation of the air-bone gaps. In the case of MEE, hearing was retested 1 hour after myringotomy and aspiration of effusion.

At the beginning of the study, the worse ear of each child was selected for analysis of the results. This selection was based on the OM history before the initial tympanostomy tube insertion (the number of RAOM episodes and duration of OME) and on the clinical status of the ear at the initial tympanostomy tube insertion (the amount of MEE and TM abnormalities).

Data analysis and all statistical tests were performed using SPSS for Windows, version 11.5 (SPSS Inc, Chicago, Ill). In the comparisons of means, t tests and analysis of variance with Tukey honestly significant difference post hoc tests were used.

The final study group comprised 237 (77.7%) of 305 children with an initial tympanostomy tube insertion who completed the 14-year follow-up and attended the 14-year checkup. The mean age of these patients (57.0% boys) was 15.1 years (age range, 14.1-15.9 years), and the mean length of follow-up from the initial tympanostomy tube insertion was 14.2 years (range, 14.0-14.8 years). There were no significant differences with respect to any analyzed factors between those who dropped out during the 14-year follow-up and those who completed it.

The mean AC PTAs and air-bone gaps with respect to otological outcomes are presented in Table 1. The mean ± SEM PTA for the 177 healed ears was 4.3 ± 0.3 dB. Of these 177 ears, 111 were normal (mean ± SEM PTA, 4.0 ± 0.3 dB) and 66 had tympanosclerosis or atrophy of the TM as the sole abnormality (mean ± SEM PTA, 4.6 ± 0.5 dB); these differences were not significant (t₁₇₅ = −1.19; 95% confidence interval, −1.80 to 0.45; P = .24; independent samples t test for group differences). For the same subgroups of 111 and 66 ears, the means of AC octave thresholds were also similar (4.2 and 5.1 dB at 0.25 kHz, 4.0 and 4.6 dB at 0.5 kHz, 3.9 and 4.5 dB at 1.0 kHz, 4.1 and 4.8 dB at 2.0 kHz, 6.5 and 6.6 dB at 4.0 kHz, and 8.6 and 9.2 dB at 8.0 kHz). Among the 237 ears, the mean ± SEM AC PTA of 5.8 ± 0.4 dB and air-bone gap of 5.0 ± 0.3 dB were not significantly different (r = 0.8, P<.001; Pearson product moment correlation). The mean BC PTA was 0.8 dB (range, −6.7 to 18.3 dB) and exceeded 10 dB in 3 ears (patients 55, 81, and 291 in Table 2). One of these (patient 81) had a normal TM with a pure sensorineural hearing loss present at the 5-year checkup, while the other 2 had combined conductive and sensorineural impairment.

Table 1 shows that otological outcomes were associated with poorer PTAs in some nonhealed subgroups (F_6,230 = 15.7, P<.001; 1-way analysis of variance). The hearing level of healed ears did not differ from that of ears with grade I PF retraction but was significantly better than that of the ears with grade II or higher PT retraction, OME, or PT perforation. The PTAs of all 4 ears with OME normalized after myringotomy and aspiration of MEE, as the conductive component of the hearing loss disappeared.

Thirteen (5.5%) of 237 ears had AC PTAs worse than 15 dB (Table 2); the hearing losses were for the most part related to the conductive components. Large air-bone gaps appeared most frequently in ears with PT retraction or PT perforation of the TM. In 1 ear (patient 133), the sole observed abnormalities were tympanosclerosis (in quadrants 1, 3, and 4) and atrophy (in quadrant 2) of the TM. There were 7 ears with TM retraction as a main abnormal outcome. Three of these had undergone ear surgery, including patient 55, who received a modified radical mastoidectomy and ossicular chain reconstruction, with subsequent grade I PT retraction in quadrants 2 and 3 of the thickened TM; patient 67, who received an early mastoidectomy and 2 revision operations for chronic OM, with subsequent adhesive OM; and patient 108, who received myringoplasty, with subsequent retraction in all quadrants of the PT. In 2 ears (patients 272 and 291), limited PT retraction (in quadrant 2 and in quadrants 2 and 3, respectively) presented with simultaneous grade II PF retraction, and 1 ear (patient 259) had grade II PF retraction as a sole abnormality. Three of the ears with a hearing loss (patients 10, 16, and 27) had PT perforation, with patient 10 having undergone myringoplasty earlier; these perforations were located in a single quadrant of the PT (in quadrants 2, 2, and 3, respectively), with patients 16 and 27 also having tympanosclerosis in the remaining parts of the TM. The last (patient 181) of these 13 ears with PTAs worse than 15 dB had a conductive hearing loss due to OME. Hearing returned to normal (PTA, 5.0 dB) after myringotomy and aspiration of MEE. The hearing results of all 9 ears that underwent surgery during the 14-year follow-up are presented in Table 3. The hearing levels seem to be related to the clinical outcomes, and hearing loss, when present, was mainly conductive. Two (patients 107 and 187) of the 3 ears with an early mastoidectomy performed for chronic discharge had normal hearing, while 1 ear (patient 67) had an adhesive TM and a moderate, mainly conductive hearing loss after 2 revision operations for chronic OM. In 1 ear (patient 55), the mainly conductive hearing loss was associated with limited grade I PT retraction (in quadrants 2 and 3) of a thickened TM after a modified radical mastoidectomy and ossiculoplasty for a retraction cholesteatoma. Two other ears operated on with hearing losses exceeding 15 dB had undergone myringoplasty and at the 14-year checkup had recurrent PT perforation (patient 10) and grade I PT retraction (patient 108).

Analysis of the better ear hearing levels (BEHLs) demonstrated only 3 patients with hearing levels worse than 15 dB (Table 2). These included patient 81 (BEHL, 20.0 dB) with normal outcomes and a nearly symmetric sensorineural hearing loss, patient 259 (BEHL, 16.7 dB) with grade I PF retraction and a mixed hearing loss, and patient 278 (BEHL, 20.0 dB) with PT retraction and a conductive hearing loss.

Previous results showed that at 5 years’ follow-up the hearing level of healed ears in children treated with VTs for OME or RAOM did not differ from that of same-aged children who had not experienced any episodes or had experienced only well-cured episodes of acute OM and had received no VT treatment.⁷ For that reason, we did not recruit a control group to establish the normal age-related hearing levels for this study. In earlier studies,^7,12-14 the PTA of healed ears at the age of 6 years was 7.6 dB and that in the present study at the age of 15.1 years was 4.3 dB, representing normal age-related hearing levels. Moreover, the sensitivity at different frequencies between 0.25 and 8.0 kHz fell within the normal age-related values.^13,14 Among children with OM, de Beer et al¹⁵ found hearing levels to be worse in those with than without VT treatment at the ages of 8 and 18 years. The association of VTs with the observed hearing loss is difficult to evaluate because of the heavier disease load in the patients with VT treatment. In addition, de Beer et al¹⁵ reported elevated BC thresholds among patients with VT treatment, with an increase between 8 and 18 years of age. We did not notice any significant sensorineural hearing loss or its progression, with the mean PTAs for BC of healed ears being −0.6 dB and 0.2 dB at 6 and 15 years of age, respectively. This slight difference is probably due to the larger relative BC receiver for younger and smaller children. In accord with 5-year follow-up results,⁷ among healed ears no difference in hearing levels was observed between the normal ears and the ears with tympanosclerosis or atrophy in the TM as sole abnormalities. Therefore, the hearing levels of healed ears were used to represent normal age-related values for statistical comparisons.

The 14-year results show that the hearing loss in these patients is related to unsuccessful otological outcomes in a conductive component. There was only one case of symmetric pure sensorineural hearing loss, of slight grade and present at the 5-year checkup. The etiology of this impairment remains obscure. In addition, 2 ears had combined conductive and sensorineural hearing loss, with a slight and unexplained decrease in BC sensitivity. Therefore, even repeated tympanostomy tube insertion is safe, from the hearing point of view, with no marked harmful effects on the inner ear. In general, hearing was well preserved, with a mean PTA of 5.8 dB for the whole group, and hearing levels can be regarded according to the classification by Clark¹⁶ as normal (hearing level, ≤15 dB) in 224 (94.5%) of 237 ears, as a slight loss (hearing level, 16-25 dB) in 10 ears (4.2%), as a mild loss (hearing level, 26-40 dB) in 2 ears (0.8%), and as a moderate loss (hearing level, 41-60 dB) in 1 ear (0.4%). The slight hearing losses, with a maximum hearing level of 21.7 dB, were associated with TM retractions and perforations. The cause of a slight conductive hearing loss in 1 ear with tympanosclerosis and atrophy of the TM as the only visible abnormalities remains uncertain but may be due to postinflammatory changes in the attic limiting the normal movements of the ossicular chain. In one of the 2 ears with a mild hearing loss, the cause was OME and the conductive loss disappeared immediately after myringotomy and aspiration of the MEE. The other ear had undergone a modified radical mastoidectomy and ossicular chain reconstruction for retraction cholesteatoma, with a mild hearing impairment as a result. The only case of moderate, nearly complete conductive hearing loss occurred in an ear that had become adhesive, despite several surgical treatments for chronic OM.

In the case of very young children, especially those younger than 2 years, OM is often bilateral, as in our study. We performed tympanostomy tube insertion bilaterally in all but 3 children to resolve OME and to prevent RAOM. It is essential to preserve normal hearing in at least 1 ear. We succeeded in all but 3 patients, in whom a slight hearing loss (BEHL, ≤20 dB) resulted, one with a symmetric sensorineural hearing loss probably not related to OM or its treatment, another with a mixed hearing loss, and a third with a conductive hearing loss.

Even a slight hearing loss in childhood, such as that caused by bilateral OME, is considered a risk factor for subsequent linguistic and educational difficulties.^3,4 Tympanostomy tube insertion with aspiration of MEE immediately restores normal hearing, unless permanent damage has been caused by OM.^1,2 We adopted a strategy of bilateral treatment with VTs for OME and RAOM early in life, which was repeated if indicated. Our long-term results show that from the hearing point of view early tympanostomy tube insertion is justified and should not be unnecessarily delayed if conservative treatment fails. Most of the hearing losses in the present series were minor and related to unsuccessful otological outcomes, not to the tympanostomy tube insertion. Ear surgery, which sometimes becomes necessary, is not hazardous for hearing, as such, but in the case of chronic middle ear infection may be insufficient to retain normal hearing or restore it to normal levels. Based on our results, we conclude that tympanostomy tube insertion early in life, and repeatedly if necessary, for OME or RAOM is a safe and useful treatment method.

Correspondence: Hannu Valtonen, MD, PhD, Department of Otorhinolaryngology, Kuopio University Hospital, PO Box 1777, FIN-70211 Kuopio, Finland (hannu.valtonen@kuh.fi).

Submitted for Publication: August 17, 2004; final revision received December 8, 2004; accepted December 16, 2004.

Acknowledgment: We thank Pirjo Halonen, MSc, for assistance with statistical analyses.

Financial Disclosure: None.