Real-life efficacy and reliability of training a hearing aid

Commercial trainable hearing aids (HA) (i.e., devices that for a period are adjusted by the user in different acoustic environments and that subsequently with changing environments automatically adapt to the user's preferred settings), are readily available; however, little information exists about the efficacy of training a HA. The purpose of this study was to investigate the efficacy and reliability of training a HA in everyday environments.

The participants were 26 hearing-impaired volunteers with a median age of 79 years and an average pure-tone average of 53 dB HL. Test devices were commercial, multimemory, prototype devices that enabled training of the compression characteristics in four frequency bands and in six sound classes. Participants wore the National Acoustic Laboratories nonlinear fitting procedure version 2 prescription for 3 weeks and trained the devices from the prescribed response for 3 weeks, before comparing their trained response with the prescription for 2 weeks. The devices were reset to the prescription, and 19 participants repeated the training and comparison trials. During the comparison trial, participants made daily diary ratings of their satisfaction with the programs, and a structured interview was completed at the end of the comparison trial.

The participants displayed different needs for changing the prescription, with more daily adjustments leading to training across more sound classes. Unreliable observations were obtained from 8 participants after each of the test and retest comparison trials. Of the 10 participants who made sufficient changes to the prescription during the first trial, 80% preferred their trained response. The 8 "low trainers" reported no preference, and also reported lower overall satisfaction with the device. Fewer adjustments were made during the repeat trial, resulting in less training. Significant correlations between trained variations were seen for 63% of 19 participants. Of the 10 participants who provided valid data after both comparison trials, those who trained the device consistently generally showed consistent preferences, and vice versa.

For those who wanted a change to the prescription, training was mostly effective. Limited data on reliability showed reasonable consistency in training outcomes and preferences. Findings, in particular on reliability, should be verified in larger populations. A guideline on how to clinically manage training with clients is presented.