Predicting hearing aid use in adults: the Beaver Dam Offspring Study.

The sample was restricted to people with treatable hearing loss, so the analysis was relatively small and could not examine all hypothesized factors. The models only used participants with complete data for every variable, so variables with many missing values were not examined. Unmeasured factors such as social support, expectations, self-efficacy and hearing aid fitting details could have influenced the results. The study could not separate hearing aid acquisition, actual hearing aid use and hearing health care seeking. Hearing aid use was only self-reported and did not include hours per day or changes in hearing aid technology over the follow-up period. Participants were ethnically homogenous non-Hispanic white individuals, so the models may not generalize to other populations. The models were not externally validated, so they are not yet ready-to-use clinical risk scores. The clinical model is not representative of all audiologic evaluations because information collected varies by patient and clinician.

The authors declared no other funding, financial relationships or conflicts of interest. The article is listed as supported by NIH extramural research.

Do not treat these models as a ready-to-use clinical risk score. The paper states that they need external validation first, and the expanded model did not significantly improve prediction over the clinical model.