Whitepaper
Systematic review and meta-analysis of audiological and patient-reported outcomes with the ADHEAR bone conduction device
Technology Assessment, BU Vibrant
29293_1.4 (Nov 2022)
Foreword
Building on extensive experience in hearing implant technology, MED-EL extended its product portfolio with the non-surgical ADHEAR bone conduction device (BCD) in 2017. Clinical evidence reported for ADHEAR up to October 2019 had been summarized in an earlier version of this Whitepaper. Coinciding with the 5th anniversary of ADHEAR, this document is updated to version 1.4, now including peer-reviewed articles published up to May 2022. Also, the content has been transferred to an online format, which makes it accessible anywhere at any time via https://go.medel.pro/OutcomesADHEAR. Designed as a ‘living document’ it will be updated regularly, while the URL will permanently link to the most recent version and can be shared or bookmarked on any digital platform.
Executive Summary
Background:
BCDs have a long history in hearing loss rehabilitation and the last decade has seen major technological advances in bone conduction technology. For those patients seeking a non-surgical solution, treatment options have been traditionally limited, and, most importantly, all relied on constant pressure onto the skin (e.g., BCD on softband, Hearing glasses). In 2017, MED-EL revolutionized this market: with ADHEAR, the need for constant pressure belonged to the past and, for the first time, a non-surgical BCD became a real alternative to implantable solutions.
ADHEAR consists of an adhesive adapter that is placed on the skin behind the ear and an audio processor that is snapped onto the adapter with a simple click (Figure 1.1). The adhesive adapter may be replaced every 3 to 7 days. Within the EU, EFTA and countries accepting the CE mark, ADHEAR is indicated for:
ᐅ Visit MED-EL’s Indications homepage for a detailed list of eligibility criteria.
- Patients with unilateral or bilateral conductive hearing loss, either temporary or permanent, no age limit.
- Patients with severe-to-profound sensorineural hearing loss in the affected ear and normal hearing in the contralateral ear (i.e., single-sided deafness).
Goal:
This Whitepaper summarizes audiological and patient-reported outcomes reported for the ADHEAR BCD that have been published in peer-reviewed medical journals between January 2017 and May 2022. The goal is to provide:
- A comprehensive overview of the scientific literature published on ADHEAR.
- A quantitative analysis of the overall benefit and variation in selected clinical outcomes.
- An online platform to explore summarized outcomes.
Methods: Based on a systematic literature review, 33 publications were identified and of these, clinical outcomes from 27 primary articles (reporting on 500 placements in 458 patients) were included and plotted as raw mean unaided/aided or improvement scores. From a subset of 17 publications, meta-analyses were conducted to estimate pooled mean scores across publications. Mixed models and model selection were used to explore the effects of potential demographic and clinical predictors on the same outcomes, if applicable.
Results:
Both, raw means and pooled estimates from meta-analyses indicated similar average outcomes and levels of variation among studies. Specifically, the results of the meta-analyses were:
- Sound-field hearing thresholds (SF-PTA4; N=13): The average aided threshold was estimated at 27.2dB HL (95% CI: 25.5-28.9). The average benefit (i.e., functional gain; N=20) was estimated at 21.7dB (95% CI: 17.6-25.8).
- Word recognition score at 65dB SPL (WRS65; N=8): The average aided score was estimated at 84.4% (95% CI: 76.5-92.3).
- Speech recognition threshold in noise (SRT50N; N=8): The average improvement was estimated at 2.2dB SNR (95% CI: 1.1-3.2).
- ADHEAR questionnaire (N=10): Most patients reported high satisfaction in the following categories: general satisfaction, adhesive tolerance, sound quality, aesthetics and self confidence.
- Daily use of the audio processor (N=5): The average daily use was estimated at 10 hours (95% CI: 8.4-12).
Overall, none of the predictors age group (adults, pediatrics), type of hearing loss (CHL, SSD), test setup, speech tests or F/U time had a significant effect on any of the outcomes investigated. Both, relatively low sample size and rather uniform patient populations (i.e., children with CHL, same testing conditions), provide reasonable explanations for this result. However, future studies will continue to fill current sampling gaps (especially in the SSD population) and increase sample size in general.
Direct comparative evidence showed that patients performed equally well with ADHEAR compared to patients using BCDs on headband or transcutaneous bone-anchored hearing aids (tBAHA) in terms of audiological outcomes. However, ADHEAR was not able to provide the same amplification as the active implantable BONEBRIDGE device.
Conclusion:
Taken together, the reviewed literature on ADHEAR provides compelling evidence for:
- Hearing improvement by means of hearing thresholds, speech in quiet and speech in noise tests.
- Measurable benefit in patient-reported satisfaction.
- Long daily use due to the absence of pressure.
Please cite this Whitepaper as follows: VIBRANT Technology Assessment Team. Systematic review and meta-analysis of audiological and patient-reported outcomes with the ADHEAR bone conduction device. MED-EL, 2022. Version 1.4 (Nov 2022). URL: https://go.medel.pro/OutcomesADHEAR.html
1 Introduction
ADHEAR (Figure 1.1) is an innovative non-surgical BCD that stands out for exerting no pressure on the skin, making it a comfortable and cosmetically appealing option for a variety of patients. The big advantage of non-surgical BCDs is that the treatment success can be experienced right away, just like with ordinary hearing aids. ADHEAR quickly found its way into clinical practice and ENT research, where it serves as a workhorse for testing bone conduction in hearing implant candidates. More than 30 peer-reviewed publications emerged from this research activity over the last 5 years. This Whitepaper aims at summarizing these results, focusing on audiological and patient-reported outcomes that were published between January 2017 and May 2022. It does not cover the full spectrum of tests and questionnaires used, but provides quantitative analyses of outcomes that were most commonly reported in the scientific community.
Figure 1.1: The ADHEAR bone conduction device (model 701), together with adhesive plaster.
ᐅ The very first patient to be included in a clinical trial with ADHEAR was 8-year-old William (Birmingham, UK). Follow up on his story he recently told in an interview.
While the history of simple bone conduction aids may date back as far as the Roman empire1, the story of ADHEAR started in 2017, when MED-EL launched the first non-surgical solution in its product portfolio. Even with implantable solutions being MED-EL’s core-expertise, MED-EL recognized the need for better non-surgical BCDs. After all, many people are unsuitable or unwilling to undergo surgery. Also, firm pressure onto the skin is still a basic requirement for other non-surgical solutions such BCDs on softband or hearing glasses, often leading to wearing discomfort and reduced patient compliance.
Using the latest generation of bone conduction technology, ADHEAR is an ideal device for patients looking for a comfortable and non-surgical solution. ADHEAR consists of an audio processor that connects to the skin with an adhesive adapter (Figure 1.1). The adhesive adapter is placed on the hairless area of the mastoid behind the ear. The audio processor simply clicks into place on the adapter and might be removed as desired throughout the day. The adhesive adapter is typically replaced after some days (watch video How does ADHEAR work?).
The audiological indication of ADHEAR includes two target groups: first, patients suffering from conductive hearing loss (Figure 1.2A) with bone conduction threshold better than or equal to 25 dB HL. Second, patients suffering from severe-to-profound sensorineural hearing loss in one ear and normal hearing in the contralateral ear (single-sided deafness, Figure 1.2B). ADHEAR is indicated for all ages, but is especially suitable for use in babies, toddlers and kids due to the lack of surgery.
While still in its first generation, ADHEAR received a lockable battery door, following requests from the field. To find out more about ADHEAR visit the ADHEAR product page or explore even more free content on the MED-EL Professionals page.
Figure 1.2: Audiological criteria for (A) CHL and (B) SSD indications.
2 Methods
2.1 Systematic review
A search strategy (see Appendix A.1) for searching the National Library of Medicine literature database (PubMed) was set up to identify articles investigating the ADHEAR BCD. Using these search terms, weekly PubMed searches were conducted since January 2017 and new articles were added to the pool of papers continuously. Articles not related to ADHEAR were excluded and not further categorized. Additional articles identified by manual search were added to the pool as well. Articles identified up to May 2022 were included for analyses in this report.
Titles, abstracts and full texts were continuously screened against inclusion/exclusion criteria as specified in Table 2.1. All screening steps were conducted by two independent reviewers and discrepancies were resolved upon discussion. The list of included publications is reported in Appendix A.2. The following parameters were extracted: 1) Demographic parameters, including sample size, mean age, type of hearing loss, uni-/bilateral implantation. 2) Study design parameters, including time to outcome measurements (F/U time), cohorts, subgroups and comparator device(-s). 3) Audiological outcomes, including sound-field hearing thresholds (SF-PTA4), speech recognition thresholds in noise (SRT50N) and word recognition scores in quiet at 65dB (WRS65). 4) Patient-reported outcome measures (PROMs), including hearing-related or general quality of life questionnaires. Hearing thresholds were extracted as pure-tone average over 4 frequencies (PTA4). Continuous numerical parameters were extracted as mean (+/-SD) for overall cohort and/or subgroup cohorts if available. In some cases, data were extracted from figures using WebPlotDigitizer v4.4.2 If raw data were available, mean values (+/-SD) were calculated. In some cases, mean values (+/-SD) were provided by the authors upon request. Data was extracted by one internal reviewer and checked by a second. In cases of disagreement, the respective paper was checked by a third internal reviewer and consensus was achieved through discussion.
| Inclusion criteria | Exclusion criteria | |
|---|---|---|
| Population | All ages, etiologies or hearing loss types (including simulated hearing loss) | Preclinical (animal) or temporal bone studies |
| Intervention | ADHEAR | None |
| Comparators | Unaided or aided with any other hearing device | None |
| Outcome | Audiometric tests; Patient-reported outcomes; Daily use; Safety | Technical outcomes |
| Study design | All publications reporting on primary data | Publications reporting on secondary data, reviews, letters, commentaries |
2.2 Data analysis
Data extracted from publications was tabulated into Excel spreadsheets and further processed within the R computational environment3 and RStudio IDE.4 Random-effects meta-analysis was performed when two or more samples (cohorts) reported mean and variance (SD or CI) for any specific outcome. Meta-analyses were performed with the metafor5,6 package for R. Separate meta-analyses were conducted for unaided, aided or improvement scores. For aided scores, raw means were used as effect size. For improvement scores, raw mean change was used as effect size. The Restricted Maximum Likelihood Estimator (RMLE) was used to estimate the amount of heterogeneity. Funnel plots and regression test for funnel plot asymmetry were used to detect signs of heterogeneity and potential publication bias. Influential case diagnostics based on Cook’s distance criterion was applied to identify outlier studies that had a significant effect on the model fit. Once detected, influential studies were excluded to achieve a more robust model fit whenever 10 or more samples (cohorts) remained in the model after exclusion. When the sample size fell below 10 studies, the power of influential case diagnostics was considered low and all studies were kept in the model.
Model selection based on the corrected Akaike information criterion (AICc) and multi-model inference was used to investigate the relative importance of demographic and clinical predictors (confounding variables) on the respective outcome under investigation. Specifically, for n predictors, separate models including combinations of 0 to n predictors were calculated and ranked by their AICc value using the R package glmulti.7 For each predictor, the model-averaged importance (relative importance) corresponds to the sum of weigths for the models in which the predictor appears.8
Subgroup meta-analyses were performed by including categorical predictors into the respective model one at a time. Only cohorts reporting on pre-defined subgroup levels were included into the model. Cohorts reporting on pooled means for two or more levels (e.g. mean outcome pooled for CHL and MHL patients) were excluded to prevent masking of the true effect. Meta-regression was used to investigate the effect of numerical predictors (e.g. F/U time) on the respective outcome.
The goal of systematic literature reviews is to derive general conclusions based on the aggregation of available data. For quantitative data, meta-analysis is the appropriate statistical tool to summarize outcomes from multiple studies. In short, meta-analysis can quantify the overall (pooled) outcome, the variability among studies and the potential influence of other variables (i.e., confounders) on the outcome under investigation.
Random Effects Model: Is one type of meta-analysis that is particularly suitable to analyze data that were not necessarily sampled from one single population.
Forest Plot: Is the preferred graph type for displaying the results of meta-analyses. It allows for an intuitive overview of outcomes from single studies and highlights the pooled outcome in the bottom line.
Confidence Interval: Margins of confidence (95%) for capturing the true mean.
Prediction Interval: Tells you that the mean outcome of any further study will fall within these limits in 95 out of 100 cases.2.3 Figures and layout
This document was written in R Markdown language using the R packages bookdown9 and knitr.10 The following R packages were used for processing data and creating figures: bibtex,11 ggbeeswarm,12 ggplot2,13 ggpubr,14 glmulti,7 kableExtra,15 lubridate,16 metafor,6 magrittr,17 plotly,18 purrr,19 reactable,20 readxl,21 RefManageR,22 stringr,23 tidyverse.24 Forest plots were generated with a function from the R package meta.25 Flowchart renderings were created with Adobe Illustrator v25.4.1.
3 Results
3.1 Publication overview
Figure 3.1: Flowchart illustrating the number of articles throughout the screening process. ‘-’ = not reported.
The literature search identified 33 articles in which ADHEAR was the target of primary research, published between January 2017 and May 2022. Articles in which ADHEAR was mentioned but not the primary target of interest were not considered. Twenty seven articles reported on clinical outcomes in patients with hearing loss that had been treated with ADHEAR. Six articles reported on methodological, technical or financial topics relating to ADHEAR or reviewed primary literature and were therefore not further analyzed. The number of articles reporting on target outcomes as well as the number of those included in meta-analyses is summarized in Figure 3.1. The proportion of articles qualifying for meta-analysis (i.e., sample size equal or larger than 3 patients AND variance (either SD or CI) reported was similar across outcomes (SF-PTA4: 56.5%, WRS65: 57.1%, SRT50N: 57.1% and Daily use: 55.6%), indicating that in about half of the publications, the mean effects were reported without a measure of variation. Among all publications, 30.3% reported outcomes from the device-specific ADHEAR questionnaire. All clinical studies were observational, either retrospective chart reviews (N=2) or prospective cohort studies (N=25).
ᐅ Mouse-over columns to see
more details
Figure 3.2: Absolute (per year) and cumulative number of publications investigating ADHEAR between January 2017 and May 2022. Please note that some publications that were available as preprints in 2021 were officially published in 2022.
3.2 Patient demographics
The 27 publications reporting clinical outcomes were based on studies conducted in 19 countries in Europe, North- and South America, Asia and Australia. These studies included a total of 458 patients (95 in chart reviews, 363 in prospective cohort studies) with 500 ADHEAR devices (102 in chart reviews, 398 in prospective cohort studies).
Most publications reported outcomes for pediatric cohorts, usually children between 5 and 17 years of age. One study26 also included 6 young children (below 5 years of age) in their pediatric cohort, but no separate summary statistics were reported for this cohort. One study reported outcomes in adult patients.
Most patients had pure conductive hearing loss (CHL) or single-sided deafness (SSD), but in one publication27 the main patient cohort had mixed hearing loss (MHL). Figures 3.3 and 3.4 summarize the pooled number of patients across age groups and types of hearing loss in included publications. Two publications28,29 specifically addressed outcomes in patients with bilateral microtia-atresia. However, outcomes with bilateral ADHEAR were addressed occasionally in 7 studies26,27,30–34 in a total of 50 patients across all publications.
By age group
Figure 3.3: Number of patients across all included studies, by age group.
Of the 458 patients for whom clinical outcomes have been reported, 141 were adults and 221 were under 18 years of age. For 96 patients, results were either reported in pooled cohorts, including both adults and pediatric patients or no age was reported at all. Six children aged below 5 years were included in 1 study but, due to the low sample size, these were pooled with older children and adolescents.
Back to top
By type of hearing loss
Figure 3.4: Number of patients across all included studies, by type of hearing loss
Of the 458 patients for whom clinical outcomes have been reported, 381 had conductive hearing loss, 13 had mixed hearing loss, 30 had single-sided-deafness and for 34 patients no specific type of hearing loss was reported.
Back to top
3.3 Audiological outcomes
3.3.1 SF-PTA4
Aided SF thresholds (SF-PTA4) were reported in 23 patient cohorts (273 ADHEAR devices in 19 publications). Out of these, 15 cohorts (195 ADHEAR devices in 13 publications) were eligible for quantitative analysis. Mean aided SF-PTA4 values varied considerably among studies, ranging from 9dB HL to 31.1dB HL. One cohort from Urik et al.34 was identified as outlier and excluded from subsequent analyses. As shown in Figure 3.5, meta-analysis indicated a mean aided SF-PTA4 of 27.2dB HL (95% CI: 25.5-28.9) and a mean functional gain of 21.7dB (95% CI: 17.6-25.8) when pooling all studies together. Forest plots summarizing results of single studies can be found in Appendix B.1 and Appendix B.2, respectively.
ᐅ Use buttons on top of each figure to zoom in/out or to export figures
ᐅ Click on legend symbols to hide/show single categories
Figure 3.5: Meta-analysis of combined aided and unaided PTA4 SF-thresholds.
Figure 3.6: Relative importance of predictors to SF-PTA4, based on multi-model inference. Predictors with values greater than 0.8 are considered to have a significant effect on the outcome under investigation.
To explore potential causes of variation in aided SF-PTA4 among publications, the following predictors were included into the meta-analytic model: age group (Adults, Pediatrics) and F/U time (in months). Type of hearing loss was not included since only CHL cohorts were available for this analysis. None of the predictors seemed to have a big effect on mean aided SF thresholds (Figure 3.6), but see Figure 3.7 below for subgroup-specific values.
Independent of age, aided SF thresholds (PTA4) did not change significantly over F/U times of up to 1 year (see subgroup analysis below).
Subgroup analysis was performed to quantify potential differences among subgroup levels. Mixed levels (e.g. CMHL, Unspecified) were excluded from subgroup analyses to reduce masking effects. Average values (including confidence and prediction intervals) and average differences in aided PTA4 among subgroups can be explored in the tabbed section below (Figures 3.7 through 3.8). Please note that sample sizes may differ from the base model and among predictors.
By age group
Figure 3.7: Effect of age group on mean aided SF-PTA4 thresholds. Box plots indicate outcomes of separate meta-analyses, i.e. overall estimated mean (black bar), confidence interval (boxes) and prediction interval (whiskers). Alongside each box plot mean values are plotted for single cohorts included in meta-analyses (filled circles) and not included due to missing standard deviation (open circles).
For 18 cohorts (214 placements in 14 studies) specific information on age group was available. Data from 11 cohorts (155 placements in 10 studies) qualified for inclusion in a subgroup meta-analysis (Adults: 3 cohorts, Pediatrics: 8 cohorts). On average, pediatric studies reported slighty higher (worse) aided SF thresholds (-1.83dB) than studies on adult patients, but this difference was not statistically significant (p=0.43).
By F/U time
Figure 3.8: Effect of F/U time on mean aided SF-PTA4 thresholds. Mean values are plotted for single cohorts included in meta-analyses (filled circles) and not included due to missing standard deviation (open circles). The dashed line indicates values as predicted by the meta-analysis, thus not including open circles. The grey area indicates the corresponding 95% confidence interval.
For 22 cohorts (249 devices in 18 studies) specific information on the F/U time was available. Data from 14 patient cohorts (171 devices in 12 studies) qualified for inclusion in a meta-regression analysis. Only results for the longest F/U time were included if studies reported aided thresholds at multiple testing intervals. F/U time did not have a significant effect on the mean aided SF threshold (p=0.525). It should be pointed out, though, that most studies covered F/U times up to 12 months only. On the other hand, only the study by Zernotti et al. (2021) reported longer F/U (52 months). Until studies with a wider range of F/U times will become available, these results need to be handled with care.
3.3.2 Comparative sound-field PTA4
Eight studies used a sequential trial design and reported SF thresholds with both ADHEAR and BAHA on softband consequently in the same patient cohort. The overall mean difference between the devices was -0.42dB HL with 95% CI spanning from -2.20 to 1.35, indicating statistically equivalent results in both devices (Figure 3.9).
Figure 3.9: Forest plot of mean difference among sequential ADHEAR-aided and Comparator-aided SF-PTA4. Each row represents one study. For each study, mean difference (black dots) among the sequential treatments and respective confidence intervals (red bars) are plotted on the x-axis. The red diamond in the bottom line represents the pooled estimate of the meta-analysis (center of the diamond) and its confidence interval (width of the diamond). The dashed line indicates the prediction interval. Studies are ordered by age group and hearing loss type. N = sample size; ad = adults; ped = pediatrics; HL = hearing loss; CHL = conductive hearing loss; CI = confidence interval; dB HL = decibel hearing level; RE = random effects; ‘-’ = not reported.
3.3.3 WRS at 65dB SPL
Aided word recognition score measured with signal from the front (S0) and at 65dB SPL (WRS65) was reported for 10 patient cohorts (113 ADHEAR devices in 9 publications). Out of these, 9 cohorts (100 ADHEAR devices in 8 publications) were eligible for quantitative analysis. Aided mean WRS65 varied considerably among studies, ranging from 59% to 100% correct understanding. Meta-analysis resulted in a pooled aided WRS65 of 84.4% with a narrow confidence interval (95% CI: 76.5-92.3) indicating a good estimate of the true mean (Figure 3.10). The forest plot showing results of single studies can be found in Appendix B.3.
Figure 3.10: Meta-analysis of combined aided and unaided WRS at 65dB SPL.
Figure 3.11: Relative importance of predictors to WRS65, based on multi-model inference. Predictors with values greater than 0.8 are considered to have a significant effect on the outcome under investigation.
To explore potential causes of variation in aided WRS65 among publications, the following predictors were included into the meta-analytic model: age group (Adults, Pediatrics) and F/U time. type of hearing loss (CHL, SSD) and speech test (Monosyllables, Bisyllables, Sentences) were not included as predictors, since all cohorts available for this analysis included only CHL patients and used monosyllabic speech tests. Relative importance was very low for both age group and F/U time (Figure 3.11), indicating that there are no measurable effects of age group and F/U time on mean aided WRS65.
Subgroup analysis was performed to quantify differences among subgroup levels. Mixed levels (e.g. CMHL, Unspecified) were excluded from analyses to reduce masking effects. Average scores (including confidence and prediction intervals) and average differences in aided WRS65 among subgroups can be explored in the tabbed section below (Figures 3.12 through 3.14). Please note that sample sizes may differ from the base model and among predictors.
By age group
Figure 3.12: Effect of age group on mean aided word recognition score at 65dB SPL. Box plots indicate outcomes of separate meta-analyses, i.e. overall estimated mean (black bar), confidence interval (boxes) and prediction interval (whiskers). Alongside each box plot mean values are plotted for single cohorts included in meta-analyses (filled circles) and not included due to missing standard deviation (open circles).
For 6 cohorts (66 placements in 5 studies) specific information on age group was available and all of these qualified for inclusion in a subgroup meta-analysis (Adults: 3 cohorts, Pediatrics: 3 cohorts). On average, pediatric studies reported lower (worse) aided WRS (-1.24%) compared to studies on adult patients, but this difference was not statistically significant (p=0.879).
By speech test
Figure 3.13: Effect of speech test material on mean aided WRS at 65dB SPL. Box plots indicate outcomes of separate meta-analyses, i.e. overall estimated mean (black bar), confidence interval (boxes) and prediction interval (whiskers). Alongside each box plot mean values are plotted for single cohorts included in meta-analyses (filled circles) and not included due to missing standard deviation (open circles).
The speech material used to measure WRS65 in quiet may consist of sentences, words (mono- or bisyllabic) or numbers (polysyllabic). For 8 cohorts (94 placements in 8 studies) specific information on the type of speech test used was available. All 6 cohorts (66 placements in 6 studies) that qualified for meta-analysis reported monosyllabic speech tests and therefore no subgroup analysis was performed. However, some studies reported results from bi- or polysyllabic speech tests that were not eligible for meta-analysis. These results are also visualized in Figure 3.13.
By F/U time
Figure 3.14: Effect of F/U time on mean WRS65. Mean values are plotted for single cohorts included in meta-analyses (filled circles) and not included due to missing standard deviation (open circles). The dashed line indicates values as predicted by the meta-analysis, thus not including open circles. The grey area indicates the corresponding 95% confidence interval.
For 10 cohorts (113 placements in 9 studies) specific information on the F/U time was available. Data from 9 patient cohorts (100 placements in 8 studies) qualified for inclusion in a meta-regression analysis. Only results for the longest F/U time were included if studies reported WRS65 at multiple testing intervals. F/U time did not have a significant effect on the average aided WRS65 (p=0.525).
3.3.4 Comparative WRS65
Several studies measured WRS65 to compare ADHEAR results to alternative hearing solutions. Five studies compared ADHEAR to BAHA on softband in sequential trial designs, but only three qualified for meta-analysis. All studies used the same test design with signal presentation from front (S0) at 65dB SPL. The overall mean difference between ADHEAR-aided and softband-aided conditions was 1.47% with 95% CI spanning from -5.79 to 8.73, indicating comparable results with these treatment options (see Figure 3.15).
Each one study compared ADHEAR to either BONEBRIDGE35 or transcutaneous BAHA36. In these studies, patients used the ADHEAR during a trial period before implantation of the respective transcutaneous device. Skarzysnki et al. (2019) measured WRS65 from the front and found equivalent outcomes with ADHEAR and transcutaneous BAHA. Dahm et al. (2021) measured WRS65 with signal presented from the ipsilateral side and found significantly higher benefit with BONEBRIDGE compared to ADHEAR.
Figure 3.15: Forest plot of mean difference among sequential ADHEAR-aided and ‘BAHA on softband’-aided WRS65. Each row represents one study. For each study, mean difference (black dots) among the sequential treatments and respective confidence intervals (red bars) are plotted on the x-axis. The red diamond in the bottom line represents the pooled estimate of the meta-analysis (center of the diamond) and its confidence interval (width of the diamond). The dashed line indicates the prediction interval. Studies are ordered by age group and hearing loss type. N = sample size; ad = adults; ped = pediatrics; HL = hearing loss; CHL = conductive hearing loss; CI = confidence interval; dB HL = decibel hearing level; RE = random effects; ‘-’ = not reported.
3.3.5 SRT in noise (SRT50N)
Speech recognition threshold in noise (SRT50N) was reported for 25 patient cohorts (312 ADHEAR devices in 13 publications). Out of these, 14 cohorts (156 ADHEAR devices in 8 publications) were eligible for quantitative analysis. Mean aided values varied among studies to such a degree (5.5dB SNR to -7.89dB SNR) that comparison among studies was questionable. Meta-analysis was therefore run on improvement scores, resulting in a pooled mean improvement of 2.2dB SNR from the respective unaided value (95% CI: 1.1-3.2; Figure 3.16). A forest plot showing results of single studies can be found in Appendix B.4.
ᐅ SRT in noise is analyzed in terms of improvement rather than absolute unaided/aided scores, as indicated by a different colour scheme.
Figure 3.16: Meta-analysis of combined SRT50 improvement in noise.
Figure 3.17: Relative importance of potential predictors to SRT50 in noise, based on multi-model inference. Predictors with values greater than 0.8 are considered to have a significant effect on the outcome under investigation.
To explore potential causes of variation in SRT50N improvement among publications, the following predictors were included into the meta-analytic model: age group (Adults, Pediatrics, All ages), type of hearing loss (CHL, SSD), speech test setup (S0N0, S90N270) and F/U time. Analysis of potential predictors indicated that F/U time had the biggest effect on SRT50N improvement (Figure 3.17), but none of the effects was statistically significant.
Subgroup analysis was performed to quantify differences among subgroup levels. Mixed levels (e.g. All Ages) were excluded from analyses to reduce masking effects. Average values (including confidence and prediction intervals) and average differences among subgroups can be explored in the tabbed section below (Figures 3.18 through 3.21). Please note that sample sizes may differ from the base model and among predictors.
By age group
Figure 3.18: Effect of age group on improvement in SRT50N. Box plots indicate outcomes of separate meta-analyses, i.e., overall estimated mean (black bar), confidence interval (boxes) and prediction interval (whiskers). Alongside each box plot mean improvement values of single studies are plotted (circles), including those not included in meta-analyses due to missing standard deviation.
For 20 cohorts (265 placements in 9 studies) specific information on age group was available. Data from 12 cohorts (132 placements in 6 studies) qualified for inclusion in a subgroup meta-analysis (Adults: 8 cohorts, Children: 4 cohorts). SRT50N improvement did not differ significantly among adult and pediatric populations (p=0.223).
By type of HL
Figure 3.19: Effect of hearing loss type on improvement in SRT50N. Box plots indicate outcomes of separate meta-analyses, i.e. overall estimated mean (black bar), confidence interval (boxes) and prediction interval (whiskers). Alongside each box plot mean improvement values of single studies are plotted (circles), including those not included in meta-analyses due to missing standard deviation.
For 24 cohorts (306 placements in 13 studies) specific information on type of hearing loss was available. Data from 14 cohorts (156 placements in 8 studies) qualified for inclusion in a subgroup meta-analysis (CHL: 11 cohorts, SSD: 2 cohorts). The Type of hearing loss did not have a significant effect on SRT50N improvement (p=0.608). However, the number of studies investigating SRT50N improvement in SSD patients is currently under-sampled and more results would be needed for a more reliable pooled estimate.
By Test setup
Figure 3.20: Effect of test setup on improvement in SRT50N. Box plots indicate outcomes of separate meta-analyses, i.e. overall estimated mean (black bar), confidence interval (boxes) and prediction interval (whiskers). Alongside each box plot mean improvement values of single studies are plotted (circles), including those not included in meta-analyses due to missing standard deviation.
For 21 cohorts (265 placements in 10 studies) specific information on the Test setup was available, including the following testing conditions: S0N0, S90N270 and S270N90. Data from 8 cohorts (76 placements in 6 studies) qualified for inclusion in a subgroup meta-analysis, but only 2 subgroup levels were represented among these (S0N0: 6 cohorts, S90N270: 2 cohorts). Test setup did not have a significant effect on SRT50N improvement (p=0.084). However, the number od studies investigating SRT50N improvement in a S90N270 test setup is currently under-sampled and more results would be needed for a more reliable pooled estimate.
By F/U time
Figure 3.21: Effect of F/U time on mean improvement in SRT50N. Mean values are plotted for single cohorts included in meta-analyses (filled circles) and not included due to missing standard deviation (open circles). The dashed line indicates values as predicted by the meta-analytic model, thus not including open circles. The grey area indicates the corresponding 95% confidence interval.
For 25 cohorts (312 placements in 13 studies) specific information on the F/U time was available. Data from 14 cohorts (156 placements in 8 studies) qualified for inclusion in a meta-regression analysis. Only results for the longest F/U time were included if studies reported SRT50N improvement at multiple testing intervals. F/U time did not have a significant effect on SRT50N improvement (p=0.290), but only F/U times up to 8 weeks were represented in this dataset.
Back to top
3.4 Patient-reported outcomes
Outcome measures that are collected from patients directly via questionnaires (i.e., Patient-Reported Outcome Measures, PROMs) have been increasingly reported in the medical literature in recent years. Publications included in this literature review have used the following questionnaires as a direct measure of patient satisfaction or quality of life (QoL) after using ADHEAR: ADHEAR Questionnaire, APSQ, AQoL-8D, GCBI, BBSS (modified), HISQUI29, IOI-HA, and SSQ. Out of these, only the ADHEAR Questionnaire has been reported in a significant number of publications. Daily use (in hours per day) is another parameter that is increasingly reported in the literature on hearing devices, either as a standalone question or within custom-designed questionnaires.
3.4.1 ADHEAR Questionnaire
This is a non-validated questionnaire developed by MED-EL, designed to collect ADHEAR-specific feedback from audiologists and clinicians. As such, it includes an open set of questions that may be administered to patients only after an ADHEAR trial period. The questions can be answered on Likert scales with different levels and thus, each question needs to be analyzed separately. Five questions have been used across different studies and will be summarized below, dealing with general satisfaction, adhesive tolerance, sound quality, aesthetics and self confidence.
General satisfaction
Question: Were the hearing device and the adapter a valuable aid for you?
In all but one publication the majority of patients found ADHEAR to be a valuable hearing aid. In only one publication, the majority of patients reported ADHEAR to be not valuable or partially valuable.
Figure 3.22: Likert plot showing the proportions of patients that voted for specific answering options. Each line represents one publication (y-axis). Positive answering options extend from the (neutral) centre to the right and are color-coded with warm colors. Negative answering options extend from the (neutral) centre to the left and are color-coded with cold colors.
Adhesive tolerance
Question: Did you suffer from skin problems or irritation from the ADHEAR adhesive adapter?
In five out of 8 studies, the majority of patients never experienced any skin problems, while the rest found skin problems to be a little bothersome. In the remaining three studies, the majortity of patients found skin problems to be only a little bothersome, while the rest never experienced any problems. Overall, only one out of 101 patients reported skin problems to be very bothersome.
Figure 3.23: Likert plot showing the proportions of patients that voted for specific answering options. Each line represents one publication (y-axis). Positive answering options extend from the (neutral) centre to the right and are color-coded with warm colors. Negative answering options extend from the (neutral) centre to the left and are color-coded with cold colors.
Sound quality
Question: How did you rate the sound quality from the device?
In all but one study, the majority of patients rated sound quality of ADHEAR to be good or very good. Overall, only one out of 101 patients rated sound quality as very bad, and six out of 101 patients rated sound quality as bad.
Figure 3.24: Likert plot showing the proportions of patients that voted for specific answering options. Each line represents one publication (y-axis). Positive answering options extend from the (neutral) centre to the right and are color-coded with warm colors. Negative answering options extend from the (neutral) centre to the left and are color-coded with cold colors.
Aesthetics
Question: How did you experience the aesthetics wit the ADHEAR adhesive asapter and the audio processor?
Most patients rated the aesthetics of ADHEAR to be good or very good. While about 30% of all patients found the aesthetic aspect acceptable, only 8 out of 84 patients gave a bad or very bad rating.
Figure 3.25: Likert plot showing the proportions of patients that voted for specific answering options. Each line represents one publication (y-axis). Positive answering options extend from the (neutral) centre to the right and are color-coded with warm colors. Negative answering options extend from the (neutral) centre to the left and are color-coded with cold colors.
Self confidence
Question: How confident did you feel when wearing the processor?
Most patients reported to feel confident or very confident when wearing ADHEAR. While about 30% felt neutral in this regard, only four out of 67 patients felt not so confident when wearing ADHEAR.
Figure 3.26: Likert plot showing the proportions of patients that voted for specific answering options. Each line represents one publication (y-axis). Positive answering options extend from the (neutral) centre to the right and are color-coded with warm colors. Negative answering options extend from the (neutral) centre to the left and are color-coded with cold colors.
3.4.2 Daily use
ᐅ The gray area indicates the timeframe between typical working hours (8 hours) and typical waking hours (16 hours) in the general population.
Figure 3.27: Meta-analysis and raw means of daily use [hrs/day] with ADHEAR
The amount of time that patients choose to use their hearing device is a potential indicator for patient satisfaction.37,38 Different methods exist to measure device usage, ranging from interviews and questionnaires to automated datalogging. The wide range of available options has hindered pooled analyses across studies in the hearing aid literature39 and a consensus on standardized reporting is still lacking to date. Most publications so far collected daily use data via traditional questionnaires.
Some of the publications reporting outcomes with ADHEAR included data on daily use and if so they reported mean usage time in hours per day, with or without standard deviation. Specifically, 9 cohorts (116 placements in 9 publications) reported mean daily use, 5 of which were eligible for meta-analysis.
Due to the small sample size, no predictors were included in meta-analyses and no subgroup analyses were conducted. Overall, mean daily use was estimated at 10.2 hours per day, while reported means varied between 6.9 and 12 hours per day.
4 Conclusion
Data from 27 publications reporting on hearing outcomes in patients using ADHEAR were available up to May 2022. Taken together, this body of literature provides compelling evidence for successful hearing rehabilitation with ADHEAR. Specifically, aided hearing thresholds, speech in quiet and speech in noise tests all indicate good hearing rehabilitation in different listening situations, independent from age group (adults, pediatrics), type of hearing loss (CHL, MHL, SSD) or short-term F/U time. Moreover, the ADHEAR-specific questionnaire and daily usage data show that patients perceived a clear benefit and used the device throughout the day.
Contact us
Do you have questions about the content of this Whitepaper? Send your question via e-mail directly to our VIBRANT Technology Assessment Team.
Would you like to start using ADHEAR at your center? Please use our contact form and you will be directed to a MED-EL representative in your area.
Discover more about ADHEAR on medel.pro
Abbreviations
| - | not reported |
| AC | Air Conduction |
| ad | adults |
| AICc | corrected Akaike Information Criterion |
| BAHA Softband | Bone-anchored hearing aid worn on softband |
| BC | Bone Conduction |
| BCD | Bone Conduction Device |
| CE | Conformité Européenne |
| CHL | Conductive Hearing Loss |
| CI | Confidence Interval |
| dB HL | decibel Hearing Level |
| dB SNR | decibel Signal-to-Noise Ratio |
| dB SPL | decibel Sound-Pressure Level |
| EFTA | European Free Trade Association |
| EU | European Union |
| F/U | Follow-Up |
| GCBI | Glasgow Children Benefit Inventory |
| HL | Hearing Loss |
| hrs | hours |
| IDE | Integrated Development Environment |
| IOI-HA | International Outcome Inventory for Hearing Aids |
| MHL | Mixed Hearing Loss |
| N | sample size |
| ped | pediatrics |
| PROMs | Patient-Reported Outcome Measures |
| PTA4 | Pure-Tone Average over 4 frequencies (0.5, 1, 2 and 4 kHz) |
| RE | Random Effects (meta-analysis) |
| RMLE | Restricted Maximum Likelihood Estimator |
| S0N0 | both Sound and Noise from front |
| S0N270 | Sound from front, Noise from contralateral side |
| S0N90 | Sound from front, Noise from treated side |
| S90N0 | Sound from treated side, Noise from front |
| S90N270 | Sound from treated side, Noise from contralateral side |
| S90N90 | both Sound and Noise from treated side |
| SD | Standard Deviation |
| SF | Sound-Field |
| SRT50N | Speech Reception Threshold in Noise |
| SSD | Single-Sided Deafness |
| URL | Uniform Resource Locator |
| WRS65 | Word Recognition Score measured at 65dB |
References
Appendix
A Methods
A.1 Search terms
Search terms targeting articles via combination of hearing-related keywords:
Adhear OR Adjoin OR plaster OR patch OR adhesive
AND
SSD OR single-sided deafness OR CHL OR conductive hearing loss OR MHL OR mixed hearing loss OR hearing loss[MeSh] OR hearing impairment
AND
bone conduction hearing device OR bone conduction hearing system OR bone conduction hearing solution OR skin drive device OR non-implantable hearing solution OR non-implantable hearing aid* OR non-implantable hearing system OR non-implantable hearing device OR BCHA
Go back to Methods: Search terms
A.2 Publications included
| ID | Reference |
|---|---|
| 33 | Liu Y, Zhao C, Yang J, et al. Audiological and subjective benefit with a new adhesive bone conduction hearing aid in children with congenital unilateral microtia and atresia. European archives of oto-rhino-laryngology : official journal of the European Federation of Oto-Rhino-Laryngological Societies (EUFOS) : affiliated with the German Society for Oto-Rhino-Laryngology - Head and Neck Surgery. 2021. |
| 32 | Ellsperman SE, Nairn EM, Stucken EZ. Review of Bone Conduction Hearing Devices. Audiology research. 2021;11(2):207-219. |
| 31 | Potier M, Seldran F, Sonthonnax M, et al. Evaluation of a New Bone Conduction Device for the Rehabilitation of Single-Sided Deafness: Effects on Speech Understanding in Noise. Otol Neurotol. 2022;43(1):105-112. |
| 30 | Dobrev I, Farahmandi TS, Huber AM, Roosli C. Experimental Evaluation of the Adhear, a Novel Transcutaneous Bone Conduction Hearing Aid. Laryngo- rhino- otologie. 2021;100(10):811-817. |
| 29 | Muzzi E, Gambacorta V, Lapenna R, et al. Audiological Performance of ADHEAR Systems in Simulated Conductive Hearing Loss: A Case Series with a Review of the Existing Literature. Audiology research. 2021;11(4):537-546. |
| 28 | Shiraishi K. Sound Localization and Lateralization by Bilateral Bone Conduction Devices, Middle Ear Implants, and Cartilage Conduction Hearing Aids. Audiology research. 2021;11(4):508-523. |
| 27 | Zernotti ME, Alvarado E, Zernotti M, Claveria N, Di Gregorio MF. One-Year Follow-Up in Children with Conductive Hearing Loss Using ADHEAR. Audiology & neuro-otology. 2021;26(6):435-444. |
| 26 | Ren LJ, Duan YS, Yu JC, Xie YZ, Zhang TY. Instant auditory benefit of an adhesive BCHD on children with bilateral congenital microtia. Clinical otolaryngology : official journal of ENT-UK ; official journal of Netherlands Society for Oto-Rhino-Laryngology & Cervico-Facial Surgery. 2021;46(5):1089-1094. |
| 25 | Dahm V, Traxler S, Liepins R, et al. Performance With an Adhesive Bone Conduction Device in Active Transcutaneous Bone Conduction Implant Users. Otol Neurotol. 2021;42(4):510-516. |
| 24 | Hirth D, Weiss R, Stover T, Kramer S. Audiological benefit and subjective satisfaction with the ADHEAR hearing system in children with unilateral conductive hearing loss. European archives of oto-rhino-laryngology : official journal of the European Federation of Oto-Rhino-Laryngological Societies (EUFOS) : affiliated with the German Society for Oto-Rhino-Laryngology - Head and Neck Surgery. 2021;278(8):2781-2788. |
| 23 | Osborne MS, Child-Hymas A, McDermott AL. Longitudinal study of use of the pressure free, adhesive bone conducting hearing system in children at a tertiary centre. Int J Pediatr Otorhinolaryngol. 2020;138:110307. |
| 22 | Di Stadio A, Dipietro L, De Lucia A, et al. A Novel Bone Conduction Hearing System May Improve Memory Function in Children with Single Side Hearing loss: A Case-Control Study. J Int Adv Otol. 2020;16(2):158-164. |
| 21 | Almuhawas F, Alzhrani F, Saleh S, Alsanosi A, Yousef M. Auditory Performance and Subjective Satisfaction with the ADHEAR System. Audiology & neuro-otology. 2021;26(1):1-10. |
| 20 | Fan X, Ping L, Yang T, et al. Comparative effects of unilateral and bilateral bone conduction hearing devices on functional hearing and sound localization abilities in patients with bilateral microtia-atresia. Acta oto-laryngologica. 2020;140(7):575-582. |
| 19 | Kuthubutheen J, Broadbent C, Marino R, Tavora-Vieira D. The Use of a Novel, Nonsurgical Bone Conduction Hearing Aid System for the Treatment of Conductive Hearing Loss. Otol Neurotol. 2020;41(7):948-955. |
| 18 | Moteki H, Kitoh R, Usami SI. The availability of an adhesive bone conduction hearing device: a preliminary report of a single-center experience. Acta oto-laryngologica. 2020;140(4):319-326. |
| 17 | Cho YS, Park YK, Seol HY, Lim JH, Hong SH, Moon IJ. Efficacy of non-invasive treatment options for single-sided deafness: A prospective study of 20 patients. Clinical otolaryngology : official journal of ENT-UK ; official journal of Netherlands Society for Oto-Rhino-Laryngology & Cervico-Facial Surgery. 2020;45(3):409-413. |
| 16 | Snapp H, Vogt K, Agterberg MJH. Bilateral bone conduction stimulation provides reliable binaural cues for localization. Hear Res. 2020;388:107881. |
| 15 | Weiss R, Loth A, Leinung M, et al. A new adhesive bone conduction hearing system as a treatment option for transient hearing loss after middle ear surgery. European archives of oto-rhino-laryngology : official journal of the European Federation of Oto-Rhino-Laryngological Societies (EUFOS) : affiliated with the German Society for Oto-Rhino-Laryngology - Head and Neck Surgery. 2020;277(3):751-759. |
| 14 | Ryberg A-CWNRS, H.-C-; Caye-Thomasen, P. Knogleforankrede høreapparater og aktive mellemøreimplantater. Ugeskr Laeger. 2019;181(17):5. |
| 13 | Chang Y, Stenfelt S. Characteristics of Bone-Conduction Devices Simulated in a Finite-Element Model of a Whole Human Head. Trends Hear. 2019;23:2331216519836053. |
| 12 | Favoreel A, Heuninck E, Mansbach AL. Audiological benefit and subjective satisfaction of children with the ADHEAR audio processor and adhesive adapter. Int J Pediatr Otorhinolaryngol. 2020;129(2020):109729. |
| 11 | Osborne MS, Child-Hymas A, Gill J, Lloyd MS, McDermott AL. First Pediatric Experience With a Novel, Adhesive Adapter Retained, Bone Conduction Hearing Aid System. Otol Neurotol. 2019;40(9):1199-1207. |
| 10 | Urik M, Hosnova D, Slapak I, et al. First experiences with a new adhesive bone conduction hearing device in children. Int J Pediatr Otorhinolaryngol. 2019;126:109614. |
| 9 | Skarzynski PH, Ratuszniak A, Osinska K, et al. A Comparative Study of a Novel Adhesive Bone Conduction Device and Conventional Treatment Options for Conductive Hearing Loss. Otol Neurotol. 2019;40(7):858-864. |
| 8 | Brill IT, Brill S, Stark T. [New options for rehabilitation of conductive hearing loss : Tests on normal-hearing subjects with simulated hearing loss]. Hno. 2019;67(9):698-705. |
| 7 | Canale A, Boggio V, Albera A, et al. A new bone conduction hearing aid to predict hearing outcome with an active implanted device. European archives of oto-rhino-laryngology : official journal of the European Federation of Oto-Rhino-Laryngological Societies (EUFOS) : affiliated with the German Society for Oto-Rhino-Laryngology - Head and Neck Surgery. 2019;276(8):2165-2170. |
| 6 | Neumann K, Thomas JP, Voelter C, Dazert S. A new adhesive bone conduction hearing system effectively treats conductive hearing loss in children. Int J Pediatr Otorhinolaryngol. 2019;122:117-125. |
| 5 | Dahm V, Auinger AB, Liepins R, Baumgartner WD, Riss D, Arnoldner C. A Randomized Cross-over Trial Comparing a Pressure-free, Adhesive to a Conventional Bone Conduction Hearing Device. Otol Neurotol. 2019;40(5):571-577. |
| 4 | Mertens G, Gilles A, Bouzegta R, Van de Heyning P. A Prospective Randomized Crossover Study in Single Sided Deafness on the New Non-Invasive Adhesive Bone Conduction Hearing System. Otol Neurotol. 2018;39(8):940-949. |
| 3 | Gawliczek T, Munzinger F, Anschuetz L, Caversaccio M, Kompis M, Wimmer W. Unilateral and Bilateral Audiological Benefit With an Adhesively Attached, Noninvasive Bone Conduction Hearing System. Otol Neurotol. 2018;39(8):1025-1030. |
| 2 | Westerkull P. An Adhesive Bone Conduction System, Adhear, a New Treatment Option for Conductive Hearing Losses. Journal of Hearing Science. 2018;8(2):35-43. |
| 1 | Dahm V, Baumgartner WD, Liepins R, Arnoldner C, Riss D. First Results With a New, Pressure-free, Adhesive Bone Conduction Hearing Aid. Otol Neurotol. 2018;39(6):748-754. |
Go back to Methods: Systematic Review
B Forest plots
B.1 SF-PTA4
Figure B.1: Forest plot of mean unaided and aided SF-PTA4. Each row on the y-axis represents one cohort. For each cohort, mean aided and unaided thresholds (black and grey dots) and respective confidence intervals (red and hollow bars) are plotted on the x-axis. Straight lines indicate the mean benefit in each cohort. The red diamond in the bottom line represents the pooled estimate of the meta-analysis (center of the diamond) and its confidence interval (width of the diamond). The dashed line indicates the prediction interval. Studies are ordered by age group and hearing loss type. N = sample size; ad = adults; ped = pediatrics; HL = hearing loss; CHL = conductive hearing loss; CI = confidence interval; dB HL = decibel hearing level; RE = random effects; ‘-’ = not reported.
Go back to Results: SF Thresholds
B.2 Functional gain
Figure B.2: Forest plot of functional gain. Each row on the y-axis represents one cohort. For each cohort, the mean benefit (i.e. functional gain) and its confidence interval (red bars) are plotted on the x-axis. The red diamond in the bottom line represents the pooled estimate of the meta-analysis (center of the diamond) and its confidence interval (width of the diamond). The dashed line indicates the prediction interval. Studies are ordered by age group and hearing loss type. N = sample size; ad = adults; ped = pediatrics; HL = hearing loss; CHL = conductive hearing loss; CMHL = conductive and mixed hearing loss; CI = confidence interval; dB = decibel; RE = random effects; SSD = single-sided deafness;‘-’ = not reported.
Go back to Results: SF Thresholds
B.3 WRS at 65dB
Figure B.3: Forest plot of mean unaided and aided word recognition score (WRS65) in quiet. Each row on the y-axis represents one cohort. For each cohort, mean aided and unaided thresholds (black and grey dots) and respective confidence intervals (red and hollow bars) are plotted on the x-axis. Straight lines indicate the mean benefit in each cohort. The red diamond in the bottom line represents the pooled estimate of the meta-analysis (center of the diamond) and its confidence interval (width of the diamond). The dashed line indicates the prediction interval. Studies are ordered by age group and hearing loss type. N = sample size; ad = adults; ped = pediatrics; HL = hearing loss; CHL = conductive hearing loss; MHL = mixed hearing loss; SSD = single-sided deafness; CMHL = conductive and mixed hearing loss; CI = confidence interval; dB HL = decibel hearing level; RE = random effects; MS = monosyllables; BS = bisyllables; PS = polysyllables; ‘-’ = not reported.
Go back to Results: WRS in quiet
B.4 SRT50 in noise
Figure B.4: Forest plot of mean improvement in SRT50 in noise. Each row on the y-axis represents one cohort. For each cohort, the mean benefit and its confidence interval (turquoise bars) are plotted on the x-axis. The turquoise diamond in the bottom line represents the pooled estimate of the meta-analysis (center of the diamond) and its confidence interval (width of the diamond). The dashed line indicates the prediction interval. Studies are ordered by age group, hearing loss type and test setup. N = sample size; ad = adults; ped = pediatrics; HL = hearing loss; CHL = conductive hearing loss; SSD = single-sided deafness; CMHL = conductive and mixed hearing loss; CI = confidence interval; dB HL = decibel hearing level; RE = random effects; ‘-’ = not reported.
Go back to Results: SRT50 in noise