A retrospective review of softband bone conduction hearing device fittings and outcomes from routine clinical care

Bone conduction hearing devices (BCDs) are an established intervention for pediatric conductive hearing loss (CHL), including congenital conditions such as aural atresia. Early fitting on a softband provides auditory access before surgical eligibility, making it a critical early hearing intervention. Recent guidelines have broadened BCD indications to include persistent CHL from otitis media with effusion (OME). With increasing clinical use, evidence on softband BCD outcomes remains limited, particularly for non-congenital etiologies. This study aimed to (I) characterize the demographic and clinical profiles of children fitted with a BCD through a national pediatric hearing care program (Study A) and (II) evaluate clinical outcomes during routine follow-up in a subset of these children with long-term BCD use on a softband (Study B). Study A involved a retrospective review of 5,490 clinical records (1998–2025) from a nationwide Australian pediatric hearing service, extracting etiology, hearing loss laterality and degree, device type, and fitting parameters. Study B involved case file reviews of 200 children, capturing etiology, device use, and outcomes from the parents' evaluation of aural/oral performance of children (PEACH) questionnaire and routine speech perception testing; reported outcomes are limited to the non-surgical softband phase in children with long-term BCD use. Children with CHL associated with microtia/atresia represented 12% of all BCD fittings, while other causes—including canal stenosis, OME, craniofacial anomalies, cleft palate, and unknown etiologies—accounted for 88%. Median age at fitting was 3.6 months for children with microtia/atresia and bilateral hearing loss and 7.2 months for children with microtia/atresia and unilateral hearing loss, with later fitting (approximately 5 years of age) in other etiologic groups. In study B, most children were consistent daily users of their devices. For bilateral CHL, PEACH scores improved significantly with BCD vs. unaided listening (p = 0.003), while differences for unilateral CHL were not significant (p = 0.87). Across children with unilateral and bilateral CHL, routine aided speech scores typically exceeded 80%. Among children with documented longer-term softband BCD use across diverse pediatric etiologies, outcomes showed robust aided speech performance and functional listening ability, particularly in children with bilateral CHL, supporting their clinical effectiveness.

Conductive hearing loss (CHL), caused by pathology in the external or middle ear, affects approximately 0.7% of children worldwide—a prevalence comparable to sensorineural hearing loss and higher than mixed hearing loss (0.2%) (Kesser et al., 2013). In children, CHL commonly arises from chronic middle ear disease or congenital malformations, both of which can effect auditory and communication development. Surgical bone conduction hearing devices (BCDs) have been an effective rehabilitation option for permanent CHL, delivering substantial benefits for speech perception and auditory performance (Snik et al., 2005). For young children not yet suitable for surgery, non-surgical BCDs fitted on a softband provide essential early auditory access for permanent losses (Tjellström et al., 2001). Increasingly, guidelines also recommend their use for persistent or fluctuating CHL due to conditions such as otitis media with effusion (OME) [Leach et al., 2021; National Institute for Health Care Excellence (NICE), 2023a]. In this context, this study examines the evolving clinical application and outcomes of softband BCDs within a large nationwide publicly funded pediatric hearing care program.

BCDs transmit sound through mechanical vibrations via the skull bone directly to the cochlea, bypassing the external and middle ear. They are particularly advantageous when the air-bone gap exceeds approximately 30 decibels (dB), as BCDs tends to provide superior speech perception performance compared with conventional air conduction hearing aids (Mylanus et al., 1998; de Wolf et al., 2011). BCD technology is broadly classified into direct drive systems (percutaneous or active transcutaneous) and skin drive systems (passive transcutaneous). Skin drive systems transmit vibrations through intact skin using an external processor held by a softband or magnetic retention (Reinfeldt et al., 2015). Softband BCDs are therefore valuable both as temporary solutions before osseointegration and as non-surgical option for children with non-permanent CHL. However, skin drive systems experience 5–20 dB attenuation at mid-to-high frequencies compared with direct drive systems, necessitating compensatory gain adjustments during fitting (Zarowski et al., 2011).

Evidence regarding outcomes in children fitted with BCDs on a softband is promising but remains limited to those with permanent congenital CHL, such as microtia/atresia. Early studies in this population have demonstrated measurable improvements in outcomes following softband fitting in infancy. Cywka et al. (2021) reported that children under 2 years of age with congenital microtia/atresia achieved significant improvements in free-field thresholds and near-age-appropriate auditory development, with mean 6-month LittlEARS scores falling within the age-based normative range. Similarly, Wang et al. (2018) evaluated 40 infants aged 3 months to 2 years with bilateral microtia-atresia and found marked gains in infant–toddler meaningful auditory integration scale scores over a 24-month period, alongside a reduction in mean aided thresholds from approximately 77 dB HL unaided to 32 dB HL with the BCD on softband. Extending this evidence base, van Zyl et al. (2024) included a wider age range of children (< 13 years) managed within a pediatric public hospital context, encompassing both congenital and ENT-related conductive pathologies. More than 80% of participants were rated by caregivers and teachers as wearing the device always or often, with functional listening performance on both the parents' and teacher evaluation of aural/oral performance of children (PEACH and TEACH, respectively) instruments aligning with age-expected normative ranges after 1 month of use. Collectively, these findings suggest that BCDs on a softband confer consistent benefits across developmental stages, supporting auditory access and functional communication outcomes.

Recent clinical guidelines have broadened the candidacy for BCDs, recommending amplification consultation for children with persistent CHL associated with conditions such as OME [National Institute for Health Care Excellence (NICE), 2023a]. In Australia, the 2020 Otitis Media Guidelines recommend referral for a hearing device consultation when the average hearing level, calculated using available audiometric thresholds, exceeds 30 dB HL in the better ear in cases of persistent OME, chronic suppurative otitis media (CSOM), or dry tympanic membrane perforation—particularly when surgical management is unavailable, delayed beyond 6 months, or unsuccessful (Leach et al., 2021). While these guidelines do not specify the type of device, the National Institute for Health Care Excellence (NICE) (2023a) guidelines suggest that BCDs may offer advantages over air conduction hearing aids for children with OME, as they remain effective when hearing levels fluctuate and are suitable for those with ear discharge or narrow ear canals, where air conduction aids may be ineffective or challenging to use. Consequently, these evolving clinical recommendations are likely to increase the consideration of BCDs for children with OME-related CHL; however, the evidence regarding their benefits in this population remains limited [National Institute for Health Care Excellence (NICE), 2023b]. Existing studies are constrained by small sample sizes, heterogeneous populations, and variable outcome measures (Brough et al., 2025; Backous et al., 2022). For instance, Ramakrishnan et al. (2006) evaluated BCDs on softband in eleven children aged 6–15 years with persistent OME whose parents declined grommet surgery. Device use varied considerably: 27% wore the device for more than 6 h per day, 55% for up to 6 h, and one child discontinued use. Despite variable use, caregivers and teachers reported perceived improvements in hearing and classroom listening performance with the device.

In summary, the evolving recommendations and candidacy criteria for BCDs highlight the need for a better understanding of current fitting practices and outcomes in children using non-surgical, BCDs. To this end, this study had two main objectives: to describe the demographic and clinical characteristics of children fitted with BCDs during routine care within Australia's nation-wide pediatric hearing care system, serving as an example of a publicly-funded healthcare model (Study A); and

to examine outcomes, as measured during standard clinical practice, in a subset of these children with long-term BCD use on a softband (Study B).

This study employed retrospective data collection methods. The study protocol was approved by the Hearing Australia Human Research Ethics Committee (NHMRC application number HAHREC 2024-36). Detailed methods specific to sub-studies are described below.

Study A

Participants

Clinical data from 5,490 children (under 18 years at first BCD fitting) who had been fitted between 3rd February 1998 and 31st January 2025 were analyzed. Devices included processors from several manufacturers, including Ponto (Oticon Medical AB, Gothenburg, Sweden), Baha (Cochlear BAS, Gothenburg, Sweden), SAMBA or ADHEAR (MED-EL, Innsbruck, Austria), Extreme 121 or PB675 (bone conductor transducer) or AudioFlex120 (bone conductor transducer; Bernafon, Bern, Switzerland), Contact Mini (BHM, Berlin, Germany), and non-descript “bone conduction” devices labeled as “Generic.” The predominance of Ponto fittings reflects the clinic's procurement arrangement with Oticon Medical, under which the Ponto device became the standard subsidized option for all pediatric fittings from 2019 onwards. No comparisons were made between device brands or models in this study.

Variables extracted

Demographic data such as current age of the child at the time of data extraction, gender, etiology and age at first device fit were extracted. Data were de-identified prior to analysis. The database provided three etiology categories: microtia/atresia, conductive (other), and other/unknown. Based on clinical notes, conductive (other) likely included canal stenosis, persistent CHL due to OME, middle ear involvement from craniofacial structures, and cleft palate; other/unknown encompassed CHL of unknown etiology. Audiometric data was available for 3,847 children in the cohort. Masked vs. unmasked BC thresholds were unable to be reliably differentiated in the available audiometric data. Hearing loss classification was based on the 4-frequency pure tone average (PTA4) of unaided air conduction thresholds measured at 500, 1,000, 2,000, and 4,000 Hz for each ear. Children with PTA4 above 20 dB HL in only one ear were classified as having unilateral hearing loss, while those exceeding 20 dB HL in both ears were classified as having bilateral hearing loss (Olusanya et al., 2019).

Study B

Participants

From the original cohort described in Study A, a subset of 200 children was selected for file review by two experienced pediatric audiologists. Children were selected by prioritizing those with the longest documented duration between the first BCD fitting date and the data extraction date. This included children (n = 14) who later received an abutment, for whom only softband outcomes were analyzed. This approach prioritized identifying children with a higher likelihood of documented outcomes and thus possibly biased the subcohort toward those with sustained records, rather than representing all children initially fitted with a BCD. For each child, clinical records from a range of appointment types were reviewed, including initial consultations prior to the first fitting, device fitting, follow-up, refit and adjustment visits.

Variables extracted

Demographic and audiological data were collected, including age at first fitting, gender, hearing thresholds and granular etiology per ear, comorbidities, fitting configuration, and device type for each ear. Fitting configuration was categorized as unilateral (one device worn on one side) or bilateral (two devices worn on both sides). Due to the details available, three categories of etiologies were used in this study: microtia/atresia, conductive, and conductive (OME). Conductive included missing middle ear structures, permanent tympanic membrane perforation (per ENT reports), canal stenosis, cholesteatoma, mastoidectomy, and unknown (if no etiology was identified from audiology or ENT notes). Conductive (OME) included persistent CHL due to OME. Per-ear etiologies were combined for each child: bilateral children were classified as microtia/atresia(B), conductive(B), conductive OME(B), or mixed(B) (microtia/atresia in one ear, conductive in the other); unilateral children as microtia/atresia(U), conductive(U), or conductive OME(U). Unaided air conduction (AC) and bone conduction (BC) thresholds were collected immediately prior to fitting (baseline audiogram). However, complete audiograms were often unavailable at this time due to factors such as limited cooperation of the child and/or unreliable results. To address this, the review expanded the window around the fitting date to identify the most reliable audiological data. Eligible audiograms were those demonstrating reliable responses, and AC and BC thresholds across a broad frequency range. The PTA4 of each ear was calculated from thresholds at 500, 1,000, 2,000, and 4,000 Hz; if one or more frequencies were missing, the average was based on the available data. Thresholds included both behavioral hearing levels (HL) and electrophysiological estimates (estimated HL i.e. eHL), depending on data availability. When both behavioral and electrophysiological thresholds were available, the former were used. Outcome measures collected included device use, parent-reported functional listening, and in-clinic speech perception tests.

Device fitting

Devices were fitted as part of routine clinical care, with bilateral BCD fittings typically reserved for children with bilateral permanent CHL who met local implantable device candidacy criteria. Unilateral fittings were standard for children with milder, unilateral, or non-permanent CHL. Fitting practices generally prioritized observation of aided speech perception and parent/clinician report over aided thresholds in sound field.

Device use

Device use was classified primarily from clinical case notes, with Genie Medical BAHS software datalogging used as a secondary source when needed, due to limited availability (e.g., lost processors, connectivity issues) and potential overestimation of logged usage time (e.g., muted but unworn processors, devices removed without being switched off). For each child, all visit-level comments describing real-world use were coded into three categories: wearing regularly, wearing irregularly, or unclear. When comments clearly indicated consistent or near-daily wear (e.g., “wearing all day,” “regular use,” “wears at school”), the child was classified as wearing regularly; when comments indicated infrequent, inconsistent, or minimal use (e.g., “takes off,” “rarely worn”), the child was classified as wearing irregularly. Each child was assigned a single modal usage category based on all available visits, with ties resolved in favor of a more informative category (e.g., prioritizing wearing regularly over wearing irregularly, and either over unclear) (Marnane and Ching, 2015). Where clinical comments were unavailable or insufficient, datalogging determined classification: children averaging ≥8 h/day (across ears and all visits) were classified as wearing regularly; those < 8 h/day were classified as wearing irregularly (Walker et al., 2015). Using this approach, classification was achieved for all 200 children (138 from clinical notes alone, 62 from datalogging alone). Note that for children fitted bilaterally, wearing regularly indicates at least one device was worn consistently (not necessarily both).

PEACH

The PEACH questionnaire (Ching and Hill, 2007) assessed everyday listening (communication) abilities in quiet and noise. Since PEACH scores increase with age, age-corrected total scores were calculated for comparisons with and without a device to control for developmental effects (Ching and Hill, 2007). PEACH data were retrospectively collected and analyzed from 95 children. For analysis, PEACH total scores were expressed as a percentage of the age-corrected total score by calculating the ratio of the observed total score to the age-expected score multiplied by 100.

Speech perception

Speech perception data were retrospectively extracted from clinical notes within the pediatric audiology service operating across a geographically diverse network. Testing was conducted in the free field using recorded or live-voice materials in quiet and in noise, typically via the MedRx Avant Audiometer test suite, under conditions that reflected routine clinical practice. Speech and noise were usually presented from spatially separated loudspeakers, including portable calibrated speakers where required, with presentation levels documented between 35 and 75 dB SPL and signal-to-noise ratios between +20 and −20 dB. Calibration and sound-level verification followed local clinical procedures and, where available, incorporated sound level meter checks for both recorded and live-voice stimuli. Contralateral masking for unilateral hearing loss was applied according to clinician judgement. Across the cohort, 198 speech assessments were documented for 117 children (59% of the sample). For this study, auditory-only assessments with a specified test, recorded masking status, SNR and test score (all reported as % correct) were included, yielding a final dataset of 173 observations from 104 children. The most frequently used materials were the Bamford–Kowal–Bench (Australian Version; BKB) sentences (n = 43 children; Bench et al., 1979; Bench and Doyle, 1979), the Kendall Toy Test (KTT; n = 33 children; Kendall, 1953; Antognelli, 1986), and the speech perception in noise (SPIN) test (n = 9 children; Kalikow et al., 1977; Hutcherson et al., 1979), with smaller contributions from the Northwestern University Children's Perception of Speech (NU-CHIPs) (Elliott and Katz, 1980), phonetically balanced lists of two-syllable words (PLOTT 3/4) (Plant and Westcott, 1983), Arthur Boothroyd words (AB Words) (Boothroyd, 1968), high-frequency discrimination (HF Dis) (Gardner, 1971), word intelligibility by picture identification (WIPI) (Ross and Lerman, 1970), and the Helen test (Plant et al., 1982).

Statistics

Data were stored in Microsoft Excel and analyzed using R version 4.5.2 (R Core Team, 2025) within the RStudio integrated development environment (Posit, Boston). Descriptive statistics summarized demographic, PEACH, and speech perception data; where distributions were skewed, medians were reported instead of means. Figures were generated with the ggplot2 package (Wickham, 2016); paired t -tests compared conditions within children using the rstatix package (Kassambara, 2023), with Welch's correction applied when equality of variances was violated; Pearson product–moment correlations were used where assumptions were met; and chi-square tests assessed device use differences across etiology groups. Linear mixed-effects models (Wiley and Rapp, 2018), implemented with the R lmerTest package (Kuznetsova et al., 2017), were used to analyze PEACH scores as a function of condition (e.g., etiology). Type III ANOVA assessed the omnibus fixed effects.

Study A

Fitted device models over time

As shown in Figure 1A, the number of fittings increased steadily over time, with a noticeable uptick in 2019 that likely corresponded to the introduction of local Australian guidelines (Leach et al., 2021) and clinic-wide initiatives aimed at improving the detection and management of persistent CHL related to OME. From 2019 onward, the Ponto (Ponto 3; Ponto 5 from 2022) was the primary device fitted, consistent with its availability as the subsidized option. SuperPower variants (fitting range up to 65 dB HL BC thresholds) accounted for 82% of all Ponto fittings, consistent with clinical recommendations for softband coupling to compensate for sound attenuation through the skin (Zarowski et al., 2011).

Variability in the age of first BCD fit by etiology, hearing loss configuration, and gender

Figures 1B, C illustrates the age at first BCD fitting by etiology and gender. Most fittings occurred by 6–7 years of age, with most children with microtia/atresia receiving their first device within the first year of life. The age of first fitting did not differ substantially by gender.

Figure 1D shows variation in the median age at first fitting across etiologies and laterality of hearing loss. Children with microtia/atresia and bilateral or unilateral hearing loss had the lowest median fitting ages (3.6–7.2 months, respectively), although children with microtia/atresia and unilateral hearing loss exhibited greater variability. In contrast, children with other or unknown etiologies had median fitting ages between 4.9 and 5.5 years, with comparable spread across children with unilateral and bilateral hearing loss.

Study B

Cohort characteristics

Table 1 shows demographic characteristics of the 200 children chosen for case file review, stratified by etiology within bilateral and unilateral hearing loss subgroups. Bilateral “mixed” refers to microtia/atresia in one ear and other CHL in the contralateral ear. Conductive pathology (excluding microtia/atresia) was most common in bilateral hearing loss, while microtia/atresia was most common in children with unilateral hearing loss. Notably, 33 children (17%) were fitted for persistent CHL due to OME.

Consistent with Study A, children with microtia/atresia had the youngest age at first BCD fitting in both bilateral and unilateral hearing loss groups, with no gender bias observed. All children with unilateral hearing loss received unilateral fittings; among children with bilateral hearing loss, bilateral fitting rates were 97% (microtia/atresia), 57% (other conductive), and 28% (conductive OME).

Softband coupling was the predominant fitting method in this cohort. The 14 children with abutment fittings all had microtia/atresia affecting at least one ear. Notably, all children with bilateral or unilateral microtia/atresia were over six years of age at the time of the study; however, only 6 of the 115 children with microtia/atresia had an abutment fitting. Comorbidities were more prevalent among children with bilateral than unilateral hearing loss. Across bilateral hearing loss groups, between 38 and 79% of children had reported comorbidities, whereas rates of comorbidities ranged from 15 to 33% among those with unilateral hearing loss.

Association between etiology, hearing levels, and device fitting configuration

Figure 2 depicts the variability in hearing levels by ear, etiology, and hearing aid fitting configuration. Bilateral fittings were most common among children with PTA4 thresholds in the moderately severe to severe range, who frequently presented with microtia/atresia (red squares). In contrast, unilateral fittings were more prevalent among children with mild to moderate hearing loss in at least one ear (blue triangles and circles), most often associated with other conductive or mixed (microtia/atresia one ear and other conductive in other ear) etiologies.

Among those children with bilateral hearing loss, an independent t -test comparing the better ear PTA4 of unilaterally and bilaterally fitted children revealed significantly worse (mean difference = 35.6 dB HL) PTA4 in bilaterally fit children [ t (106.1) = 18.1, p < 0.001]. The analogous comparison for the worse ear PTA4 showed a small but significant difference [mean difference = 4.9 dB HL; t (170.3) = −2.3, p = 0.03]. This suggests that bilateral fittings were more common in those with bilateral hearing loss with worse degrees of hearing loss in the better ear.

Outcomes

Consistency in device use

Most children (n = 148, 74%) were classified as wearing regularly, with 45 (22.5%) wearing irregularly and seven (3.5%) unclear. Overall, regular use appeared less frequent in children with unilateral hearing loss vs. bilateral (66% vs. 80%), although this trend did not reach statistical significance [χ 2 (2) = 5.87, p = 0.053]. Figure 3 presents device use categories by current age (Figure 3A), age at first fitting (Figure 3B), and etiology for children with bilateral and unilateral hearing loss (Figures 3C, D). Usage patterns did not differ by current age [χ 2 (6) = 2.15, p = 0.905] or age at first fitting [χ 2 (8) = 8.57, p = 0.380], though children aged 12+ years trended toward irregular use (standardized residual = −2.1). Usage also did not vary across etiologies within bilateral [χ 2 (6) = 9.96, p = 0.126] or unilateral [χ 2 (4) = 1.94, p = 0.747] hearing loss groups.

PEACH

Figure 4 illustrates the percent of age-corrected PEACH scores with and without the BCD, among children with bilateral hearing loss (n = 40 children, n = 48 observations) and unilateral hearing loss (n = 45 children, n = 54 observations), as a function of one or two BCDs and etiology. For children with bilateral hearing loss, the mean [standard deviation (SD)] percent of age-corrected PEACH scores were 46.42% (28.29) unaided and 82.36% (20.13) aided. For children with unilateral hearing loss, the corresponding mean (SD) scores were 86.32% (28.10) unaided and 97.98% (20.63) aided.

In the bilateral hearing loss group, the mean (SD) percent of age-corrected PEACH scores were 88.00% (21.09) for microtia/atresia, 85.43% (44.44) for conductive, 76.89% (14.70) for conductive (OME), and 79.92% (15.05) for mixed etiologies. A linear mixed-effects model indicated no significant effect of etiology on PEACH performance, F (3, 31.74) = 0.65, p = 0.59. The same analysis was not performed for children with unilateral hearing loss due to limited available data. As no clear patterns were observed across etiologies, data were pooled across etiology groups for subsequent analyses.

For children with bilateral hearing who had completed PEACH during unaided and aided periods (n = 7), a paired t -test revealed a significant increase in percent of age-corrected PEACH scores with BCD, with an average increase of 43.5% [ t (6) = −4.8, p = 0.003, 95% CI (−65.7, −21.4)], representing a large effect size, d = 1.82. Across the broader bilateral hearing loss cohort, PEACH performance was comparable between children using one vs. two BCDs. Those fitted with one device (n = 24) had a mean (SD) PEACH score of 82.61% (16.51), while those using two devices (n = 12) had a mean (SD) score of 83.41% (26.90). An independent t -test revealed no statistically significant difference between these groups [ t (15.3) = −0.1, p = 0.93, 95% CI (−18.8, 17.2)].

For children with unilateral hearing loss who completed the PEACH during periods with and without BCDs (n = 9), a paired t -test revealed no statistically significant difference between the two periods [mean difference = −1.3%; t (8) = −0.2, p = 0.87, 95% CI (−19, 16.4)].

Speech perception

Descriptive analyses are presented for the BKB, KTT, and SPIN tests, which contributed the majority of available speech perception data. More detailed results from these tests, together with results from other speech perception measures that were not a primary focus of the analysis, are provided in Supplementary Tables S1–S3.

BKB

Figures 5A, B show BKB data (n = 41 children, n = 80 observations) by hearing loss status and fitting configuration where better-ear masking was not applied for unilateral hearing loss. All scores were documented as using live voice and without better-ear masking for children with unilateral hearing loss. The median age of testing was 8.3 years (IQR = 7.1–10.4 years). Experience with BCDs varied across children (median 1.9 years, IQR 0.6–3.8), but no evidence of an association was found between aided scores and duration of device use [ r (63) = −0.13, p = 0.3]. In aided conditions, scores were near or at ceiling; 91.2% (SD = 9.3%) for bilateral hearing loss and 91.9% (SD = 13.0%) for unilateral hearing loss. In unaided conditions, mean scores were lower; 70.0% (SD = 28.3%) for bilateral hearing loss and 79.8% (SD = 19.4%) for unilateral hearing loss.

For children with unilateral hearing loss tested on the BKB with better-ear masking applied, aided scores ranged from 65% to 100% (mean = 89.4%, SD = 13.2%; n = 5 children, 7 observations). Of these, two children had corresponding unaided data, in which aided performance exceeded unaided by 68–70 percentage points (i.e., 26% vs. 94% and 26% vs. 96%).

SPIN

Figures 5C, D show SPIN data (n = 9 children, n = 15 observations) by hearing loss status and fitting configuration where better-ear masking was not applied for unilateral hearing loss. The median age at testing was 12.2 years (IQR = 10.7–13.6 years). Experience with BCDs varied across children (median 0.1 years, IQR 0.05–1.9), but no evidence of association was found between aided SPIN scores and duration of device use [ r (9) = 0.34, p = 0.3]. Overall, aided SPIN performance was near ceiling levels; children with bilateral hearing loss obtained a mean score of 92.5% (SD = 8.7%), and those with unilateral hearing loss obtained a mean score of 93.1% (SD = 6.2%). Scores were relatively lower in unaided conditions. A single child with bilateral hearing loss scored 48%, while three children with unilateral hearing loss had a mean unaided score of 78.3% (SD = 2.89).

Three children with unilateral hearing loss were assessed in both unaided and aided conditions under comparable within-subject test conditions (no better-ear masking applied). Aided scores ranged from 84 to 100% (mean = 91.3%, SD = 8.5%), compared with unaided scores of 75 to 80% (mean = 78.3%, SD = 2.9%). Aiding improved SPIN scores by 4 to 20% (mean benefit = 13.0%, SD = 8.2%).

KTT

Figures 5E, F shows KTT data (n = 31 children, n = 41 observations) by hearing loss status and fitting configuration where better-ear masking was not applied in children with unilateral hearing loss. The median age at testing was 4.9 years (IQR = 3.9–7.4 years). Experience with BCDs varied across children (median 3.4 years, IQR 1.2–4.6), but no evidence of a correlation was found between aided KTT scores and duration of device use [ r (36) = 0.19, p = 0.26]. Overall, KTT performance was at or near ceiling. Among children with bilateral hearing loss, aided scores averaged 97.1% (SD = 4.5%). For children with unilateral hearing loss, aided scores were similarly high, averaging 97.3% (SD = 5.9%). In contrast, the small number of unaided observations showed greater variability; one child with bilateral hearing loss scored 50%, whereas all children with unilateral hearing loss scored 100%, likely reflecting contributions from the normal-hearing ear in the absence of better-ear masking.

Two children were assessed both unaided and aided under comparable test conditions. Aided scores were 100% across both children, whereas unaided performance ranged from 50% (bilateral hearing loss) to 100% (unilateral hearing loss). Two additional children with unilateral hearing loss were evaluated only in aided conditions with better-ear masking, and both achieved scores of 100%.

Overall, across different types of tests, these speech recognition results demonstrate strong aided performance (typically >80%).

The present study aimed to (i) characterize the demographic and clinical profile of children fitted with BCDs within a publicly funded pediatric hearing care system and (ii) evaluate functional outcomes among a subset of these children with long-term BCD use on a softband gathered during routine clinical care.

Fitting configuration: bilateral vs. unilateral

In Study B, bilateral fittings were more common among children with poorer better-ear thresholds, many of whom had bilateral microtia/atresia (see Figure 3). This reflects the clinical protocol prioritizing bilateral fittings for children with bilateral permanent CHL (see Section 2). The predominance of unilateral fittings in Study A is therefore expected, given the relatively low proportion of children with microtia/atresia in that larger cohort (see Figure 1B).

Bilateral device fitting is standard clinical practice for children with bilateral hearing loss, particularly for air-conduction hearing aids and cochlear implants, where binaural benefits in sound localization and speech understanding in noise are well established (Häusler et al., 1983; Bronkhorst and Plomp, 1989; Byrne and Noble, 1998; Forli et al., 2011). Although the minimal interaural attenuation in skull-transmitted sound has traditionally supported the notion that unilateral BCD fittings are sufficient (Eeg-Olofsson et al., 2008; Stenfelt, 2012), binaural processing can be achieved with bilateral BCDs, and is more pronounced when BC thresholds are symmetrical, rather than asymmetrical (Stenfelt et al., 2024). Several studies in adults have demonstrated measurable binaural advantages from bilateral BCD fittings, including improved sound localization and enhanced speech understanding in both quiet and noisy environments (Brassington et al., 2022; Priwin et al., 2005; Dun et al., 2013; Snapp et al., 2020; Caspers et al., 2021; Canale et al., 2022). There is also evidence of binaural processing when sound is delivered bilaterally via bone-conduction stimulation (Stenfelt and Zeitooni, 2013; Zeitooni et al., 2016). Improved lateralization with bilateral vs. unilateral fittings has also been observed in children with congenital bilateral CHL, despite sound localization remaining suboptimal (Den Besten et al., 2020). In our sample, improvements in functional listening (see Figure 4A) between unilateral and bilateral fittings in children with bilateral hearing loss were modest. This may be due to the small sample size, between-subject comparisons, the limited sensitivity of the outcome measures, and/or limited additional benefit from the second device. Future research employing within-subject comparisons and outcome measures that specifically probe spatial hearing is needed to more clearly identify and quantify bilateral benefits. Future research should also aim to quantify the relative benefits and cost-effectiveness of bilateral vs. unilateral BCD fittings, particularly in children whose hearing loss may be transient and for whom a BCD serves as a temporary solution.

Fitting ages vary with etiology and hearing loss configuration (unilateral/bilateral)

Children identified with microtia/atresia were fitted between three and 7 months of age (see Figure 1D), consistent with Early Hearing Detection and Intervention (EHDI) 1–3–6 guidelines for congenital hearing loss, which recommends hearing screening no later than 1 month of age, diagnosis no later than 3 months of age for infants not passing screening, and enrollment into intervention no later than 6 months of age [Joint Committee on Infant Hearing (JCIH), 2000, 2019]. However, greater variability was observed in the fitting ages of children with unilateral microtia/atresia, a finding consistent with Patel et al. (2023), who reported that children with unilateral hearing loss were on average first fitted closer to school age, with a wide range in age at first fit. This variability may reflect a lower perceived benefit of amplification in children with unilateral hearing loss—particularly in low demand listening environments—which has been reported as a contributing factor to delayed fitting (Alexander et al., 2020; Kanzara et al., 2020; Nelissen et al., 2015; Purcell et al., 2016). Nonetheless, our findings indicate encouraging trends toward early fitting and consistent BCD use within this population.

Children with etiologies other than microtia/atresia, such as OME, were typically first fitted with BCDs closer to school entry, for both bilateral and unilateral hearing loss (see Figure 1D; Table 1). This later fitting likely reflects the acquired nature of the hearing loss, the influence of school-entry hearing screening programs on detection, and clinical pathways that prioritize medical management of middle ear disease before device provision. The comparatively older fitting age in this group—consistent with previous reports indicating a mean fitting age of approximately 7 years among children with ear, nose, and throat (ENT)-related pathologies (van Zyl et al., 2024)—may also reflect family decisions to pursue amplification as children begin formal schooling. Future studies incorporating age at diagnosis would help clarify the relative contribution of these factors, as this variable was not available in the present dataset.

Device benefit for children with both bilateral and unilateral hearing loss

Aided speech perception scores obtained during routine clinical testing were generally high for children with bilateral (see Figures 5A, C, E) and unilateral hearing loss (see Figures 5B, D, F). For the unilateral group, routine clinical assessments were typically conducted without masking the contralateral (better) ear. This approach likely reflects both a clinical preference to evaluate performance under listening conditions that approximate real-world listening (i.e., binaural input), and the practical difficulties of implementing reliable masking procedures in pediatric BCD users. However, in the small subset of children for whom masking was applied, substantial aided benefit was observed when comparing unaided and aided speech scores. This outcome is expected, because masking the better ear isolates the device's true contribution to speech perception, whereas unmasked results reflect combined input from both ears.

In our sample, overall device use was relatively high across all etiologies, with expectedly lower—but still encouraging—device use among the children with unilateral hearing loss. Consistent with our findings, Patel et al. (2023) also reported promising rates of device use among children with unilateral hearing loss, where most wore either a BCD or contralateral routing of signal (CROS) aid for the duration of the school day, and Easwar et al. (2024) reported similarly high parent-rated daily device use in preschool-aged children with unilateral hearing loss.

Functional benefit, as measured by the PEACH, was observed for children with bilateral hearing loss, but we were not able to confirm a similar effect for children with unilateral hearing loss. This finding aligns with previous research (Cywka et al., 2021), which reported that the functional benefit—quantified using the LittlEARS questionnaire—of softband BCD fittings was significantly greater in children with bilateral hearing loss compared to unilateral hearing loss, and suggests that, although the PEACH can be sensitive to difficulties in noise even in unilateral hearing loss (Easwar et al., 2024), improvements can be difficult to detect.

Study B demonstrated that children with persistent CHL due to OME are being fitted with BCDs in routine clinical practice, with encouraging device usage (see Figures 3C, D). However, many of these children had comorbidities that increase susceptibility to OME (e.g., Trisomy 21) and/or relatively stable hearing thresholds. Among those with available PEACH data, analyses showed no significant differences in age-corrected PEACH scores across bilateral etiological groups, indicating that children with OME-related CHL demonstrated similar functional benefit to those with other etiologies. However, further data are needed to clarify BCD outcomes in children with OME-related CHL and to inform evidence-based clinical decision-making for this population.

Progression from softband to abutment in age-eligible children

Study B indicated that children with permanent CHL due to microtia/atresia frequently continued using softband devices beyond age five—the minimum age of eligibility for surgical intervention (Roman et al., 2011; see Table 1). Although many children in this cohort were age-eligible (5 years or older) for surgical implantation (mean age 7.4 years), information about surgical candidacy and discussions was unavailable. Sustained softband use in this group may therefore reflect family preferences, perceived surgical risks, reimbursement challenges (Koenigs et al., 2022), clinical contraindications, or other contextual factors. The comparatively younger age of our microtia/atresia cohort may also partly explain the relatively small number of children with abutments (13/131) observed in our cohort, as other studies report older average ages at implantation, such as 9.3 years (Fan et al., 2019) and 8.9 years (Lippmann et al., 2018) in children with atresia. As abutment coupling enhances word recognition, lexical decision performance, and word learning speed in children over softband (Pittman, 2019), investigating barriers and facilitators to the transition to surgical BCDs could inform strategies to promote optimal auditory and linguistic outcomes.

Limitations

First, the large cohort study (Study A) relied on analyzing BCDs registered to children in the clinical database. This limited the ability to draw firm conclusions about how many children were active device users, as it is possible that a child was no longer actively using the device, or that the aid was lost or not returned, despite the child remaining registered as having one. Nevertheless, the case file review in Study B provided evidence of promising device use albeit in a smaller cohort. Second, speech perception outcomes were obtained under routine clinical conditions across multiple clinic sites, with variability in presentation mode (live vs. recorded speech), presentation levels, SNRs, spatial configurations, and better-ear masking. These results therefore represent pragmatic, real-world speech testing outcomes. In addition, device verification (e.g., text box measures) or audibility-based outcomes (aided sound field thresholds) were not consistently documented and therefore could not be systematically analyzed. However, the generally favorable speech perception and functional outcome scores suggest that, at a group level, insufficient audibility is unlikely to have been a major contributor to poorer performance in this cohort. Third, Study B is an outcome-focused sub-study involving children with documented longer term softband use. As the cohort was chosen based on the longest elapsed time since first fitting, it is likely subject to selection bias, with a tendency to favor children who continued device use and may not be representative of all children initially fitted. Findings should therefore be interpreted within this context.

Clinical implications

This study represents the first large-scale investigation of pediatric BCD fitting practices. Regular device use across a range of etiologies and hearing loss lateralities is encouraging, particularly among children with unilateral hearing loss and those with persistent CHL secondary to OME. Aided functional outcomes, as measured by the PEACH, were comparable for children with bilateral OME and those with other bilateral etiologies, suggesting similar perceived benefit. Variability in outcome measures underscores the need for assessment tools suited to children at different developmental stages in real-world clinical settings. However, greater standardization of testing conditions and adoption of more sensitive measurement tools would facilitate consistent data across clinics and studies, thereby enabling robust inferential statistical analyses.

This study described the demographic and audiological characteristics of children fitted with a BCD during routine clinical care in a large national audiological center and examined clinical outcomes in a subset of these children with long-term BCD use on a softband. A diverse range of etiologies was represented, including persistent CHL due to OME. Children with bilateral hearing loss demonstrated regular device use and significant improvement on the PEACH when aided, supported by strong aided speech perception results. Among children with unilateral hearing loss, device use was also encouraging, and speech testing suggested potential benefit under specific listening conditions. Overall, findings highlight the positive outcomes achievable with BCDs and support their ongoing role in managing both bilateral and unilateral CHL in children.