Tympanometry and otoacoustic emissions in a cohort of special care neonates.

Otoacoustic emission (OAE) screening and oto-admittance testing (678 Hz probe tone) were performed on both ears of 84 special care neonates, as part of a larger study of middle-ear effusion in neonates and infants. OAE results, tympanometry, and acoustic reflex results are all strongly and significantly associated. No evidence was found of any maturational effects in the results. Based on the findings, a tentative classification scheme for neonatal tympanograms is suggested. We conclude that 678 Hz tympanometry is a useful indicator of middle-ear status in very young babies, and that middle-ear effusion does strongly affect OAEs in neonates. OAEs are also strongly affected by negative middle-ear pressure (MEP), and mean MEP in ears failing OAE screens was significantly more negative than in those passing. The prevalence of abnormal tympanometry, which may indicate middle-ear effusion or dysfunction, was 20% of ears (29% of babies) in this group. It appears that middle-ear effusion could account for about half of the ears failing an OAE screen on the special care baby unit. We also find that length of stay on the special care baby unit is an important risk factor for development of middle-ear effusion: those on the unit for over 30 days have about four times the risk of bilateral abnormal tympanometry.

Acoustic correlates of tonal tinnitus.

A sensitive microphone has been developed which can pick up tonal signals (spontaneous acoustic emissions) in the sealed ear-canal of certain subjects. Various properties of these frequency components suggest that they arise from an active, frequency-selective self-limiting feedback process within the cochlea and that they rely on internal reflection from the middle ear. An external tone can synchronize, frequency-lock, suppress of frequency-shift the acoustic component. These interactions are frequency-dependent in a way suggestive of cochlear tuning properties. Positive or negative middle-ear pressure can also influence the components by increasing their frequency and in some cases can enhance one component at the expense of a neighbouring one. Some subjects hear these components as tinnitus and can report on the measured changes. Other subjects do not hear the measured signals, which otherwise behave similarly. A third group of subjects have tinnitus but no objective sound can be detected. In this last group there are, nevertheless, sometimes notches or other discontinuities in the audiogram which correspond to their tinnitus pitch-matches. It appears likely that the recordable type of tinnitus is essentially non-pathological and represents hypersensitivity of the system, whereas the non-recordable type might be associated with local pathological changes at the end-organ or more centrally.