Perforation of the tympanic membrane and its effect on the real-ear-to-coupler difference acoustic transform function

It is not always possible to undertake extensive real-ear measurements, especially in infants and young children. An alternative approach is to estimate the real-ear SPL by use of an acoustic transform function such as the real-ear-to-coupler difference (RECD). This may be used to estimate the real-ear sound pressure level (SPL) obtained from an insert transducer or a hearing instrument. The aim of the present study was to investigate the effects of tympanic membrane perforation on the RECD transform function. Subjects in the study comprised two groups of 12 individuals aged between nine and 65 years. One group of subjects had a tympanic membrane perforation and was recruited to the study before admission for myringoplasty. There was no evidence of middle ear pathology in the remaining subjects who comprised the control group. An RECD transform function for an insert transducer was measured on each subject using the standard clinical protocol on the Audioscan RM500 real-ear measurement system. There was a statistically significant difference between the two groups; mean RECD transform value of the perforation group was 9-12 dB lower than the corresponding value in the control group at audiometric frequencies below 1.5 kHz. This difference is probably due to the perforation acting as a vent and allowing low-frequency acoustic energy to escape into the middle ear cavity. Use of an average RECD transform function to estimate real-ear SPL in subjects with a perforation will overestimate the SPL reaching the tympanic membrane. As a result, the derived real-ear SPL obtained by use of either an insert transducer or a hearing instrument will be overestimated. This has implications for the selection and verification of a hearing instrument. The difference in the mean RECD transform function between the control group and subjects with a tympanic membrane perforation supports the use of individually measured RECD values wherever possible.