Are hearing tests performed in General Practice at the 31/2-year developmental assessment a reliable indicator of the presence of otitis media with effusion?

Abstract

Hypothesis: Hearing tests performed in General Practice at the 3½‐year developmental assessment do not correlate with objective measures of otitis media with effusion (OME: ‘glue ear’). Subjects: One hundred and thirty‐nine consecutive children attending the 3½‐year developmental assessment of four General Practitioners in a single inner‐city practice. Outcome measures: The main outcome measure was the presence or absence of effusion as determined by the ‘gold standard’ of type B and C2 tympanograms bilaterally. Independent variables were standard toy test performed by the four GPs, parental assessment of hearing on the test day, the occurrence of two or more documented episodes of otitis media, presence of a smoking parent in the dwelling place and otoscopic findings (glue/no glue) determined by the GP performing the test. Results: The incidence of bilateral OME, as determined by tympanometry, in this population of 3½‐year olds was 28%. Forward stepwise linear regression analysis indicated that the only statistically significant independent variable to correlate with tympanometric findings of OME was the GP’s hearing test (P = 0.003). Conclusion: Carefully performed GP hearing tests appear to be good predictors of OME in 3½‐year olds, and should continue.