Defining Hearing Loss Severity Based on Pure Tone Audiometry and Self-Reported Perceived Hearing Difficulty, National Health and Nutrition Examination Survey.

Abstract

There is a well-known metric to describe average/normal vision, 20/20, but the same agreed upon standard does not exist for hearing. The pure tone average has been advocated for such a metric. We aimed to use a data driven approach to inform a universal metric for hearing status based on pure tone audiometry and perceived hearing difficulty (PHD). Cross-sectional national representative survey of the civilian non-institutionalized population in the United States. Data from the 2011-2012 and 2015-2016 cycles of the National Health and Nutrition Examination Survey (NHANES) was used in our analysis. Of 9,444 participants aged 20-69 years old from the 2011-2012 and 2015-2016 cycles, we excluded those with missing self-reported hearing difficulty (n=8) and pure tone audiometry data (n=1,361). The main analysis sample, therefore, included 8,075 participants. We completed a sub-analysis limited to participants with "normal" hearing based on the WHO standard (pure tone average, PTA of 500, 1000, 2000, 4000 Hz < 20 dBHL). Descriptive analyses to calculate means and proportions were used to describe characteristics of the analysis sample across PHD levels relative to PTA. Four PTAs were compared, low frequency (LF-PTA, 500, 1000, 2000 Hz), four frequency PTA (PTA4, 500, 1000, 2000, 4000 Hz), high frequency (HF-PTA, 4000, 6000, 8000 Hz) and all frequency (AF-PTA, 500, 1000, 2000, 4000, 6000, 8000 Hz). Differences between groups were tested using Rao-Scott χ2 tests for categorical variables and F tests for continuous variables. Logistic regression was used to plot receiver operating characteristic (ROC) curves with PHD as a function of PTA. The sensitivity and specificity for each PTA and PHD was also calculated. We found that 19.61% of adults aged 20-69 years reported PHD, with only 1.41% reporting greater than moderate PHD. The prevalence of reported PHD increased with higher decibel hearing levels (dBHL) categories reaching statistical significance (p < 0.05 with Bonferroni correction) at 6-10 dBHL for PTAs limited to lower frequencies (LF-PTA and PTA4) and 16-20 dBHL when limited to higher frequencies (HF-PTA). The prevalence of greater than moderate PHD reached statistical significance at 21-30 dBHL when limited to lower frequencies (LF-PTA) and 41-55 dBHL when limited to higher frequencies (HF-PTA). Forty percent of the sample had high frequency loss with normal low frequency hearing, representing nearly 70% of hearing loss configurations. The diagnostic accuracy of the PTAs for reported PHD was poor to sufficient (< 0.70), however the HF-PTA had the highest sensitivity (0.81). We provide three basic recommendations for clinical application based on our analysis. 1). A PTA based metric for hearing ability should include frequencies above 4000 Hz. 2). The data driven cutoff for any PHD/normal hearing is 15 dBHL. When considering greater than moderate PHD, the data driven cutoffs were more variable but estimated at 20-30 dBHL for LF-PTA, 30-35 for PTA4, 25-50 for AF-PTA, and 40-65 for HF-PTA. 3). Clinical recommendations and legislative agendas should include consideration beyond pure tone audiometry such as functional assessment of hearing and PHD.