Predicting Individual Hearing-Aid Preference From Self-Reported Listening Experiences in Daily Life.
Jeppe H Christensen, Johanne Rumley, Juan Camilo Gil-Carvajal and 3 others
PMID 38783420WHAT IT FOUND
Hearing-aid preference could be predicted from daily listening ratings, especially when the app also recorded what the person was listening to and the surrounding sound.
Retrospective questionnaire and in-the-moment reports each helped, and together they predicted most choices.
Key findings
01Participants rated the hearing aid they chose to keep higher on the SSQ than the model they did not choose; six of the 49 SSQ items were significantly higher, and the largest item differences were on the Spatial and Qualities scales.
02All EMA rating items except Q1 showed significant rating differences between preferred and nonpreferred hearing aids; sound satisfaction had the largest difference and spatial localization the smallest.
03Combining SSQ and EMA information correctly classified the final hearing-aid choice for 92.5% of participants; the EMA model that also used self-reported listening activity and data-logged sound environment reached 93.8% accuracy.
STILL TO COME
How it was doneWhat they foundWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
Only two Oticon miniRITE models were compared, and their main difference was noise reduction, so the findings may not generalize to other hearing aids or features. All participants were experienced hearing-aid users, so new users may respond differently to subtle processing differences. The study was conducted remotely during COVID-19. Hearing aids were programmed from audiograms completed within the prior 3 years and real-ear verification was not used, so fittings may have been suboptimal. Pandemic restrictions and face masks likely limited outside listening situations. This reduced the variety of real-life contexts. Two participants retrieved hearing-aid model information from iPhone settings, so blinding was not complete. SSQ items often had N/A responses, and one participant was removed from SSQ analysis for outlier ratings, so questionnaire coverage was incomplete. The two hearing-aid models estimated sound levels differently, although individual tertile binning was used to control for this. Some participants chose a model that was not predicted by either SSQ or EMA, and the discussion notes that preference was not always the same as the highest EMA rating.
Declared interests
The supplied text does not report a conflict-of-interest declaration. Publication types indicate non-U.S. government research support. The hearing aids studied were Oticon devices.
The easy way to misread this
Do not read the high prediction accuracies as proof that either hearing aid improves communication. The study predicted which of two Oticon models a person chose to keep, and some participants' choices were not predicted by either the SSQ or EMA.