Hearing aids in the real world: typical automatic behavior of expansion, directionality, and noise management

Automatic DSP (digital signal processing) features, widely available in hearing aids today, are useful because they alleviate the need for the hearing aid wearer to manually adjust the hearing aid as listening conditions change. Although the theoretical basis for the design of these features may be sound, little is known about their behavior in the real world. Data logging offers a glimpse into the life of the individual hearing aid wearer, but there are no published data to date that provide a frame of reference for the interpretation of this information. Further, data logging in hearing aids provides only aggregate summaries for individual features, ignoring complex interactions including the differences between the left and right sides of a bilateral pair.

The purpose of this study was to determine the typical behavior of three automatic DSP hearing aid features-expansion, directionality, and noise management-in daily life.

Ten individuals with hearing impairment were fitted bilaterally with BTE (behind the ear) hearing aids. The hearing aids were programmed for the individual's hearing loss with expansion, directionality, and noise management set to activate automatically. A PDA (personal digital assistant) logged the input level and status of expansion, directionality, and noise management from both devices at 5 sec intervals. Data were gathered in this manner over a period of 4-5 wk.

A total of 741 hr of hearing aid use were logged, 50% of which were spent in environments no louder than 50 dB SPL. Expansion, directionality, and noise management were active 45, 10, and 21% of the time, respectively; the median amount of gain reduction for noise management was ∼1 dB. Although expansion and noise management were always active at the low and high input levels, respectively, activation of directionality never exceeded 50%. Expansion and noise management were sometimes active simultaneously, as were directionality and noise management. Bilateral agreement in feature activation typically exceeded 80%, except when the input level was at the cusp of a threshold for activation of a specific feature and at high input levels.