Use of an autoinflation device does not lead to a clinically meaningful change in hearing thresholds in children with otitis media with effusion

Design The objective of this study was to establish whether autoinflation was an effective intervention in a paediatric audiology service. This was a pragmatic retrospective study with historical controls using a paired availability design. Setting The study took place at a single paediatric audiology service in England. Participants All children seen in the clinic over a two-year period who were aged between 3 and 11 years and who had a type B tympanogram in at least one ear were included. The Otovent autoinflation device was available as a treatment option over the second year (Cohort B) but not the first (Cohort A). There were 976 children included in the study: Cohort A comprised 513 children, Cohort B comprised 463 children. Main outcome measures The aims were to evaluate whether there was improvement in hearing thresholds following introduction of an autoinflation device, and whether there was a reduction in further audiology follow-ups, and in referrals to an ear, nose and throat specialist for consideration of ventilation tube insertion. Results and conclusions There was a statistically significant improvement in hearing thresholds in Cohort B compared to Cohort A, however the improvements were clinically minimal with small effect sizes. There was no significant difference in improvement of tympanometry results between the two cohorts. Significantly more children in Cohort B (autoinflation group) were referred to an ear, nose and throat specialist after their second appointment compared to Cohort A. It was feasible to introduce autoinflation into the care pathway, however there was no evidence of clinically meaningful improved outcomes for patients.