Identifying factors associated with spontaneous restoration of hearing in children with otitis media with effusion

Objectives: To identify predictors of acceptable hearing at 5 weeks, 6 and 12 months in children with bilateral otitis media with effusion (OME). Design and setting: Secondary analysis of OSTRICH data, conducted in hospital ear, nose and throat (ENT) and paediatric audiology and audiovestibular medicine departments across Wales and England. Participants: The OSTRICH study included 389 children aged 2-8 years with bilateral hearing loss attributable toOME for at least 3 months. Main outcome measures: Baseline, 5-week, 6- and 12-month audiology measurements were collected and logistic regression models used to identify pre-randomisation baseline variables that predicted return of acceptable hearing, which was defined as less than or equal to 20 dB hearing loss averaged within the frequencies of 0.5, 1, 2 and 4 kHz in at least one ear in children assessed by pure tone audiometry, ear-specific insert visual reinforcement audiometry or ear-specific play audiometry less than or equal to 25 dB hearing loss averaged within the frequencies of 0.5, 1, 2 and 4 kHz in children assessed by sound-field visual reinforcement audiometry or sound-field performance/play audiometry, based on national guidelines. Results: Less severe baseline hearing loss across both ears most consistently predicted acceptable hearing at 5 weeks (adjusted odds ratio [aOR] 0.91, 95% CI 0.87-0.95), 6 months (0.94 [0.90-0.98]) and 12 months (0.93 [0.89-0.97]). Negative history of atopy (2.05 [1.16-3.61]), never using hearing aids (aOR 2.16 [1.04-4.48]), and being male (1.75 [1.02-2.99]) were significant at 6 months, but not at 12 months. Symptom duration was a predictor at 5 weeks, but not at 6 or 12 months. Conclusions: Milder baseline hearing loss most consistently predicts acceptable hearing at 5 weeks, 6 and 12 months in children with chronic OME. Negative history of atopy, never using hearing aids, and male gender are associated with better prognosis. These predictors can be used to identify children that may not require treatment.