Systematic Review of Quality of Life in Bone Anchored Hearing: Conductive vs. Unilateral Sensorineural Hearing Loss

ABSTRACT

Objective

The purpose of this study was to systematically review the differences in disease‐specific quality of life (QoL) benefits experienced by bone‐anchored hearing implant (BAHI) users between those diagnosed with unilateral sensorineural hearing loss (U‐SNHL) and those with conductive/mixed hearing loss (CHL).

Data Sources

Eligible studies were searched for in Medline (Ovid), Embase (Ovid), CINAHL (Ebsco), Cochrane (Wiley), Global Health (Ovid), Web of Science (Clarivate Analytics), Africa Wide Information (Ebsco) and Global Index Medicus (WHO) from inception to October 23, 2022. Updated searches were performed on November 9, 2023, and July 11, 2024.

Review Methods

There were no restrictions on language. PRISMA standards were followed, and screening was conducted by two independent reviewers in Rayyan, with a third reviewer resolving conflicts. Risk of bias was assessed using RoBANS. Articles were included if patients were implanted with a BAHI and administered a validated, disease‐specific QoL measure.

Results

One thousand, three hundred twelve articles were identified after duplicate removal, with 56 articles meeting the inclusion criteria. Eight different disease‐specific QoL measures were administered. In all, the APHAB's “Global” (p = 0.0002), EC (p < 0.00001), and BN (p = 0.02) scores, as well as the GBI's “Global” (p = 0.0001), “General” (p = 0.002), and “Physical” (p = 0.02) scores were significantly different between U‐SNHL and CHL populations.

Conclusion

These results demonstrated disease‐specific QoL differences between BAHI users with U‐SNHL and CHL. Specifically, patients with CHL reported greater benefits in domains pertaining to communication ease, the clarity of sound, and their overall health and psychosocial status.

The literature demonstrates that bone‐anchored hearing implants have a positive impact on auditory performance and quality of life (QoL), but these outcomes may not always correlate. This systematic review analyzed 56 articles to identify discrepancies in self‐reported, disease‐specific QoL benefits among patients with either unilateral sensorineural hearing loss or conductive/mixed hearing loss. The findings of this review highlight significant differences in specific QoL measures, thereby enhancing our comprehension of the distinctive QoL advantages encountered by different hearing loss populations.

1. Introduction

Since the first bone‐anchored hearing implant (BAHI) procedure in 1977, these implants have played a significant role in restoring hearing function for individuals with conductive hearing loss (CHL), mixed hearing loss (MHL), and unilateral sensorineural hearing loss (U‐SNHL) [1]. By capturing sound waves in the user's environment and transferring vibrations directly to the cochlea through the skull, the external sound processor bypasses the need for the outer and middle ear [2]. The sound processor is connected to an internal titanium implant that is osseointegrated into the temporal bone, enhancing sound transmission. Today, individuals can choose between a percutaneous BAHI (pBAHI), where the sound processor is connected to a protruding abutment, or a transcutaneous BAHI (tBAHI), where the sound processor is magnetically attached to an internal implant.

The audiological benefits of BAHIs have been extensively studied. Research has shown improvements in speech comprehension in both quiet and noisy environments, as well as in word discrimination tasks [3, 4]. Other reviews have reported significant improvements in aided thresholds and functional gain [5]. However, Dornhoffer et al. [6] demonstrated that improvements in audiological performance do not necessarily correlate with an increased perception of quality of life (QoL) benefits. In fact, no audiological tests were able to predict patient‐reported QoL benefits, as measured using the abbreviated profile of hearing aid benefits (APHAB). These findings suggest that disease‐specific QoL measures may capture a dimension of benefit that audiological tests fail to assess. Therefore, understanding the daily hearing challenges faced by BAHI users and how their devices help them overcome these challenges is crucial for evaluating the overall effectiveness of the BAHI.

Previous systematic reviews have shown improvements in the QoL of BAHI users with U‐SNHL, as measured by disease‐specific QoL measures [7, 8]. These findings have also been observed in pediatric populations and in patients with bilateral hearing loss fitted unilaterally or bilaterally with a BAHI [3, 9]. However, no review has specifically examined the differences in QoL outcomes between patients with CHL, MHL, or U‐SNHL.

The objective of our study was to systematically review disease‐specific QoL benefits in BAHI patients with unilateral CHL/MHL, bilateral CHL/MHL, and U‐SNHL. Identifying variations in outcomes across these patient groups will help clinicians provide more informed counseling to prospective patients regarding the QoL benefits they may expect. This information will also facilitate better management of patient expectations concerning the device's performance, potentially resulting in greater long‐term satisfaction and utilization of the BAHI.

2. Methods

The findings of this study are reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) statement. The protocol for this systematic review was registered in a publicly accessible database (PROSPERO ID: CRD42022367259) prior to searching the databases.

2.1. Information Sources

A senior medical librarian searched the following databases Medline (Ovid), Embase (Ovid), CINAHL (Ebsco), Cochrane (Wiley), Global Health (Ovid), Web of Science (Clarivate Analytics), Africa Wide Information (Ebsco) and Global Index Medicus (WHO) from inception until November 9, 2023. The search strategy used variations in text words found in the title, abstract, or keyword fields and relevant subject headings to retrieve articles looking broadly at bone anchored hearing implants and patient reported quality of life, outcome, or experience measures, with no language restriction. The full search strategy for all databases, as well as the PRISMA literature search extension (PRISMA‐S) and PRISMA checklists, can be found in Supporting Information (Tables S1–S3).

2.2. Inclusion Criteria

The following inclusion criteria were considered: (a) Population: studies involving patients of any age diagnosed with unilateral or bilateral CHL/MHL, (b) Intervention: participants implanted with either a percutaneous, passive transcutaneous, or active transcutaneous osseointegrated BAHI, (c) Comparison: studies involving patients of any age diagnosed with U‐SNHL, (d) Outcomes: studies that reported QoL outcomes, by means of a validated, disease‐specific outcome measure, and other relevant outcomes related to the type of hearing loss, treatment outcomes, or adverse events, were included, (e) Study design: randomized controlled trials (RCTs), non‐randomized comparative studies, cohort studies, case–control studies, cross‐sectional studies, and retrospective questionnaire studies. The validation of a measure is characterized by its prior evaluation through preliminary pilot testing and psychometric analysis to determine the measure's reliability and validity.

2.3. Exclusion Criteria

Studies where patients used any other type of hearing aids, such as air conduction hearing aids, cochlear implants, dental fixtures, or middle ear implants, were excluded. Studies that investigated QoL outcomes in patients using a non‐implantable BAHI (BAHI on a headband or with adhesive) were also excluded. Additionally, studies that used a non‐validated, disease‐specific QoL measure or whose translation had not been validated were excluded from the review. Finally, studies that measured general health QoL were not included.

2.4. Study Selection

The titles and abstracts of identified studies were screened on Rayyan by two independent reviewers (K.T. and J.D.) for relevance based on the inclusion criteria [10]. Articles that passed the first screening had their full texts retrieved and assessed for eligibility. Any disagreements were resolved through discussion or consultation with a third reviewer (S.J.D.). References of the included articles were manually searched to identify any other possible articles. The primary reasons for study exclusion were documented in an Excel spreadsheet.

2.5. Data Extraction

The data from the included articles was extracted independently by two authors (K.T. and J.D.) using a standardized data extraction form in Excel (Microsoft Office 365, Windows). Extracted information included author, year of publication, country of study, study design, and number of participants that filled out a QoL measure. Patient demographics were also noted, such as participant's type of hearing loss, age, and sex (male or female). If available, the BAHI device and/or the nature of the device (percutaneous, passive transcutaneous or active transcutaneous) used by the participants was recorded. The QoL measure used by the authors was identified and the reported scores and follow‐up periods were extracted. For simplicity, patients with CHL or MHL were combined into a single group named “CHL.” The QoL scores were organized into one of four groups: U‐SNHL, U‐CHL, B‐CHL, or “U‐CHL + B‐CHL.”. The “U‐CHL + B‐CHL” group is for all participant data where the laterality of the hearing loss was not specified or reported. If necessary, a robust imputation strategy was employed in order to include the highest number of articles in the analysis. Articles that reported their results by means of graphs or bar plots rather than with numerical values did not have their data extracted. This system was implemented to enhance the reliability of the results.

2.6. Risk of Bias Assessment

The Risk of Bias (RoB) Assessment tool for Non‐Randomized Studies (RoBANS) was utilized to assess the methodological quality of the included studies and evaluate their RoB [11]. Two authors (K.T. and J.D.) independently assessed the risk of bias, with any disagreements being resolved by a third reviewer (S.J.D.). RoBANS consists of six variants that determine whether there is a low, high, or unclear risk of bias: (1) selection of participants, (2) confounding variables, (3) measurement of exposure, (4) blinding of outcome assessments, (5) incomplete outcome data, and (6) selective outcome reporting. The fourth variant was deemed irrelevant for the risk of bias evaluation of the included articles and therefore was not utilized. An article is deemed to have a low RoB if the majority of the variants were scored as “Low risk” and a high RoB if the majority of the variants were scored as “High risk.” An article is considered to have an unclear RoB if two variants scored as “Low risk,” two as “High risk” and one as “Unclear risk.”

2.7. Statistical Analysis

Statistical analysis was conducted to assess and synthesize the data collected by the QoL measures. Mean and standard deviation (SD) values for QoL scores were extracted from all included articles. The means and SDs of the benefit scores were calculated, with the latter following the formula provided in Chapter 6.5.2.8 of the Cochrane Handbook for Systematic Reviews of Interventions (2023). The correlation coefficient used to compute the benefit SD was 0.59, as determined by Balk et al. [8].

A random effect meta‐analysis (Review Manager Version 5.4, The Cochrane Collaboration, 2020) was performed on the calculated benefit scores of QoL measures that had 4 or more studies for each type of hearing loss. Overall effect and subgroup differences were noted. High heterogeneity was defined as an I ² > 50% [12]. Publication bias was also evaluated by producing funnel plots when appropriate. If a study had 2 or more follow‐up periods, the QoL scores from the longer follow‐up period were included in the meta‐analysis to control for patients' enthusiasm bias [13]. As described by Hol et al. [13], enthusiasm bias refers to the phenomenon where patients may potentially respond more favorably to reflect their satisfaction with the BAHI following their initial fitting. This bias could result in misleading findings that do not accurately reflect patients' true outcomes. Therefore, only the latter follow‐up period will be considered in order to account for the impact of this effect.

All meta‐analyses performed underwent subsequent sensitivity analyses. During the sensitivity analyses, articles were excluded if authors presented aggregate results that contained both adult and pediatric patients; only articles with adult‐only populations and results were included. This decision was made to account for the fact that the QoL measures administered had not yet been validated for use in pediatric populations, therefore rendering any pediatric results invalid. Furthermore, articles were excluded if they required imputation techniques to calculate pre‐operative QoL scores, had a retrospective study design, or were rated as having an “Unclear” RoB score.

3. Results

3.1. Study Selection

The results of the search are outlined in the PRISMA flowchart (Figure 1). Due to the low number of pediatric articles identified during the search, this population was removed from this systematic review. Out of the 1612 articles identified, 300 duplicates were removed and 1312 assessed for eligibility. Among those assessed, 308 articles met the eligibility criteria and underwent full‐text review, from which 55 articles were included in this systematic review. An additional article was identified by manually screening references and was included in the review, bringing the total to 56 articles. Out of the 56 articles included in this systematic review, 6 were added during updated searches performed on November 9, 2023, and July 11, 2024.

FIGURE 1.

The literature demonstrates that bone‐anchored hearing implants have a positive impact on auditory performance and quality of life (QoL), but these outcomes may not always correlate. This systematic review analyzed 56 articles to identify discrepancies in self‐reported, disease‐specific QoL benefits among patients with either unilateral sensorineural hearing loss or conductive/mixed hearing loss. The findings of this review highlight significant differences in specific QoL measures, thereby enhancing our comprehension of the distinctive QoL advantages encountered by different hearing loss populations. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]

3.2. Study Characteristics

Demographic data of the included articles as well as hearing loss etiologies can be found in Table 1. A table describing the study characteristics of the included articles can be found in Supporting Information (Table S4) [14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69]. A total of 42 studies were conducted exclusively in adult populations, while 14 studies included both adult and pediatric patients. These studies were conducted in 19 different countries, with 8 of them adopting an international multicenter design. The articles included in this analysis involved a total of 511 patients with U‐SNHL and 1408 patients with CHL.

TABLE 1.

Demographic and etiological data of included studies.

	Type of hearing loss
Demographics	CHL	U‐SNHL
n	1408	511
Mean age (Range)	44.94 (4–93)	49.54 (7.5–79)
Female:Male ratio	309:263	229:178
Etiology (n [%])
Chronic otitis media	237 (30.66)
Unilateral or bilateral aural atresia	170 (21.99)
Unilateral or bilateral CMA	104 (13.45)
Discharging mastoid cavity	65 (8.41)
Otosclerosis	49 (6.34)
Acoustic neuroma		128 (35.75)
Sudden SNHL		64 (17.88)
Unknown		20 (5.59)
Congenital SNHL		23 (6.42)
Tinnitus		20 (5.59)
Other	148 (19.15)	103 (28.77)

Note: Data provided in this table is dependent on the disclosure of the figures of interest in the included articles. As such, this table does not comprise the data of all the included studies.

Abbreviations: CHL, conductive hearing loss; CMA, congenital microtia‐atresia; SNHL, sensorineural hearing loss; U‐SNHL, unilateral sensorineural hearing loss.

3.3. Risk of Bias

There were no concerns for a high RoB in any of the included articles. However, 6 articles were assessed as having an “Unclear risk” and were excluded from further statistical analyses. A summary plot of the 5 RoBANS variants and the distribution of the judgments was created with the web application Robvis (Figure 2) [70]. Individual ratings for each article can be found in Supporting Information (Figure S1).

FIGURE 2.

Summary plot of the risk of bias assessment for included studies. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]

3.4. QoL Measures

In this review, 7 QoL measures were utilized: Abbreviated Profile of Hearing Aid Benefit (APHAB) [71, 72], Glasgow Benefit Inventory (GBI) [73, 74], Speech, Spatial, and Qualities of Hearing Scale (SSQ) [75], SSQ12, SSQ Benefit (SSQ‐B), International Outcome Inventory for Hearing Aids (IOI‐HA) [76, 77], and the Bern Benefit Single‐Sided Deafness (BBSSD) [78] (Table 2).

TABLE 2.

Summary of the disease‐specific quality of life measures administered in the included studies of this systematic review.

Questionnaire	Description	Scoring
Abbreviated profile of hearing aid benefit (APHAB)	A 24‐item questionnaire that evaluates the patient's speech understanding in everyday situations and the presence of unwanted noises in the environment [71, 72]. It is composed of 4, 6‐item subscales: ease of communication (EC), background noise (BN), reverberation (RV) and aversiveness (AV).	Scale responses range from 1 (Never) to 99% (Always) Global score = average of EC + BN + RV subscales Lower scores represent fewer problems in listening situations
Glasgow benefit inventory (GBI)	A post‐intervention questionnaire that aims to assess patient benefit following an otorhinolaryngology procedure [73, 74]. It consists of 18 questions divided into 3 subscales: general (12 questions), social (3 questions) and physical (3 questions) benefits.	Item scored on 5‐point Likert scale, with the middle indicating “No change” Scores converted to a − 100 to +100 benefit scale
Speech, spatial, and qualities of hearing scale (SSQ)	A 49‐item questionnaire examining the listener's ability to comprehend speech in scenarios with competing noises and various spatial characteristics, as well as assessing the clarity and naturalness of sounds [75]. Three subscales are evaluated pre‐ and post‐BAHI fitting: speech (14 questions), spatial characteristics (17 questions) and qualities of sound (18 questions).	Items scored on a scale from 0 to 10 Standard anchor terms of “Not at all” (0) to “Perfectly” (10)
SSQ12	An abbreviated, 12‐item questionnaire derived from the SSQ scale. 5 questions were associated with “Speech,” 3 questions regarding “Spatial,” and 4 questions for “Qualities”.	Same scoring and anchoring terms as the SSQ scale
SSQ benefit (SSQ‐B)	A modified SSQ scale that tasks the patient with comparing their hearing experience now to how it was before being fitted with their hearing device.	Items scored on a scale from −5 to +5, where the midpoint represents “Unchanged”
International outcome inventory for hearing aids (IOI‐HA)	This questionnaire was developed to supplement existing health‐related QoL outcome measures and to be accessible to international communities [76]. It measures the effectiveness of the patients' hearing aid by covering 7 domains: daily use, benefit, residual activity limitations, satisfaction, residual participation restrictions, impact on other, and QoL [77].	Items scored on a scale of 1 to 5
Bern benefit single‐sided deafness (BBSSD)	This 10‐item questionnaire uses visual analogue scales (VAS), ranging from 0 to 10 points, to measure the patient's perceived benefit from their BAHI or contralateral routing of signal (CROS) device in aiding them understand speech in various listening settings [78].

3.5. Clinically Meaningful Differences

It is important to analyze QoL benefit scores in terms of their clinical relevance rather than relying solely on their statistical significance. A clinically meaningful difference (CMD) can be defined as the threshold at which the patient and/or clinician consider the intervention to have a meaningful impact [79]. In the context of this review, we will examine whether the patient perceives their BAHI as having a meaningful impact on their lifestyle. The CMD values for the APHAB, GBI, and SSQ scales are presented below:

• APHAB: 10‐point change in the global benefit score was considered clinically meaningful [71, 80]. Additionally, it was observed that patients experienced a true difference when a 10‐point change was observed in all of the benefit scores of the EC, RV, and BN subscales, or if there was a 22‐point increase in any of these scores [71].

• GBI: Since the GBI is a post‐intervention scale, it is already designed to capture the positive or negative impact of an otorhinolaryngology procedure on a patient's life. Therefore, its global and subscale scores would already reflect a clinically relevant change.

• SSQ: There is limited consensus on the CMD for the SSQ scale. A 2022 systematic review mentioned that it could range from a 0.7‐ to 1.3‐unit decrease, indicating an improvement in the patient's QoL, while another article suggests that a 1.0‐unit decrease would be clinically relevant [8, 75, 81]. For this systematic review, a 1.0‐unit decrease in subscale score was established as the CMD. However, we acknowledge that further research is necessary to establish a validated value.

3.6. Meta‐Analysis Results

Insufficient studies reported QoL scores specifically for U‐CHL patients, making it impossible to conduct a subgroup analysis between U‐SNHL, U‐CHL, B‐CHL, and “U‐CHL + B‐CHL” populations. Consequently, we combined all CHL patients into one group, regardless of the laterality of their hearing loss, to compare subgroup differences between U‐SNHL and CHL patients.

The QoL measures that were subjected to meta‐analyses are the APHAB, GBI, and SSQ scales. Data were collected from the relevant studies which focused either exclusively on adult patients or included a combination of adult and pediatric patients. An overview of all the meta‐analyses results is provided in Table 3. Patients with U‐SNHL reported significant benefits in most QoL measures, except in the APHAB's AV subscale (p = 0.09) and the GBI's “Physical” subscale (p = 0.26). In the CHL group, there were significant improvements in all QoL measures compared to baseline measurements. In most meta‐analyses, heterogeneity was substantially high. The funnel plots for the APHAB's BN and RV subscales, as well as the GBI's “General” subscale, all demonstrate asymmetry, while the other plots demonstrate a proper inverted funnel shape. The forest plots for both the “Global” scores of the APHAB and GBI can be viewed in Figures 3 and 4, respectively. The forest plots for all other subscales, along with the funnel plots, can be found in Supporting Information (Figures S2–S7).

TABLE 3.

Summary of meta‐analysis results of disease‐specific quality of life measures from patients with unilateral sensorineural hearing loss versus conductive hearing loss.

	U‐SNHL			CHL			Subgroup differences
	Mean (95% CI)	p	I 2 (%)	Mean (95% CI)	p	I 2 (%)	p	I 2 (%)
APHAB
Global	13.80 (8.72–18.89)	< 0.00001	79	28.95 (22.64–35.26)	< 0.00001	94	0.0002	92.6
EC	14.14 (10.18–18.09)	< 0.00001	74	28.49 (24.65–32.33)	< 0.00001	83	< 0.00001	96.2
BN	20.45 (13.64–27.25)	< 0.00001	94	30.51 (25.79–35.24)	< 0.00001	90	0.02	82.4
RV	16.58 (12.02–21.15)	< 0.00001	80	26.60 (14.22–38.98)	< 0.0001	100	0.14	54.9
AV	−5.27 (−11.45–0.91)	0.09	84	−4.27 (−8.32 to −0.22)	0.04	88	0.79	0
GBI
Global	22.63 (16.62–28.64)	< 0.00001	61	36.78 (32.94–40.61)	< 0.00001	81	0.0001	93.4
General	30.02 (19.11–40.93)	< 0.00001	77	48.63 (44.07–53.19)	< 0.00001	67	0.002	89.5
Social	11.85 (4.21–19.49)	0.002	61	18.95 (8.66–29.24)	0.0003	91	0.28	15.1
Physical	1.55 (−1.18–4.28)	0.26	0	12.67 (3.55–21.79)	0.006	89	0.02	80.9
SSQ
Speech	1.78 (1.19–2.38)	< 0.00001	70	3.32 (1.62–5.01)	0.0001	97	0.09	64.3
Spatial	1.19 (0.54–1.84)	0.0003	74	2.81 (0.94–4.68)	0.003	96	0.11	61.1
Qualities	1.11 (0.44–1.77)	0.001	79	2.83 (1.20–4.46)	0.0006	98	0.05	72.9

Note: Bold values indicates p < 0.05.

Abbreviations: APHAB, abbreviated profile of hearing aid benefit; AV, aversiveness; BN, background noise; CHL, conductive hearing loss; EC, ease of communication; GBI, Glasgow Benefit Inventory; RV, reverberation; SSQ, speech spatial and qualities of hearing scale; U‐SNHL, unilateral sensorineural hearing loss.

FIGURE 3.

Forest plot demonstrating mean difference in APHAB “Global” scores in U‐SNHL and CHL populations. APHAB, abbreviated profile of hearing aid benefit; CHL, conductive hearing loss; U‐SNHL, unilateral sensorineural hearing loss. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]

FIGURE 4.

Forest plot demonstrating mean difference in GBI “Global” scores in U‐SNHL and CHL populations. CHL, conductive hearing loss; GBI, glasgow benefit inventory; U‐SNHL, unilateral sensorineural hearing loss. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]

There were 6 subscale scores that were significantly different between the subgroups: the APHAB's “Global” score (I ² = 92.6%, p = 0.0002), EC subscale (I ² = 96.2%, p < 0.00001), and BN subscale (I ² = 82.4%, p = 0.02), and the GBI's “Global” score (I ² = 93.4%, p = 0.0001), and both its “General” (I ² = 89.5%, p = 0.002) and “Physical” subscales (I ² = 80.9%, p = 0.02).

3.7. Summary Statistics of IOI‐HA, BBSSD, SSQ12 & SSQ‐B

Of the included studies, 4 articles described their results with the IOI‐HA scale [18, 38, 42, 68], and 2 included articles used the BBSSD questionnaire [54, 63]. Additionally, 6 included articles administered the SSQ12 scale [38, 42, 48, 55, 60, 68] and 2 others reported results with the SSQ‐B [51, 54]. Descriptive statistics for these scales can be retrieved in Supporting Information (Tables S5–S8).

3.8. Sensitivity Analysis

The sensitivity analyses conducted yielded mixed results regarding heterogeneity differences before and after the sensitivity analyses. While certain subscales demonstrated a decrease in heterogeneity within the U‐SNHL and CHL populations, others exhibited increased levels of heterogeneity, with some remaining unchanged. The variations in heterogeneity before and after the sensitivity analysis are detailed in the Supporting Information (Table S9). Nevertheless, the overall heterogeneity levels for subgroup differences improved across all subscales in the three questionnaires evaluated. Additionally, the statistical significance of the GBI's global score, along with the “General” and “Physical” subscales, was altered as a consequence of the sensitivity analysis, rendering them statistically insignificant (Global: p = 0.06, General: p = 0.13, Physical: p = 0.18).

4. Discussion

4.1. Results From Meta‐Analyses

The findings of this systematic review and meta‐analyses show that QoL improvements for BAHI users vary based on their classification of hearing loss. Significant subgroup differences were found in the scores of 3 disease‐specific QoL measures: the APHAB, GBI, and SSQ.

Several subscales showed statistically significant differences in QoL scores between patients with U‐SNHL and CHL. These subscales include the APHAB's “Global,” EC, and BN subscales, as well as the GBI's “Global,” “General,” and “Physical” subscales. In fact, CHL patients scored higher in all these subscales, indicating that they reported greater benefits compared to those with U‐SNHL. Therefore, CHL patients appear to have an easier time following conversations in environments with varying numbers of speakers and with background noise present, while also having a better overall health and psychosocial well‐being. However, all GBI subscales became statistically insignificant following sensitivity analyses. Additionally, the funnel plot of the “General” subscale from the GBI questionnaire was asymmetric, revealing possible publication bias or differences in methodological quality, particularly in smaller studies, as described in Chapter 13.3.4.3 of the Cochrane Handbook for Systematic Reviews of Interventions (2023). It is worth noting that all benefit scores for both populations were clinically meaningful, as demonstrated by their APHAB and SSQ scores.

One hypothesis that could explain the discrepancy in the experienced QoL benefits in these different hearing loss populations is the presence of the head shadow effect in patients with U‐SNHL [7]. The head shadow effect causes an attenuation of the intensity of sound signals, particularly those above 1000 hertz (Hz), as the head reflects sound waves away from the ear with better hearing [15]. Although this effect can be alleviated with the help of a BAHI [47, 82], situations where unwanted noise is presented to the hearing device and a speaker is positioned in front of the patient can remain bothersome [7, 31].

Alternatively, a predictive model has identified factors such as young age, a shorter period of follow‐up after fitting the BAHI, and a higher pure‐tone average in the better hearing ear as crucial in identifying who may experience greater benefit from their hearing aids [83]. However, a two‐sample t‐test assuming unequal variance revealed that the mean age for both the CHL and the U‐SNHL populations was not statistically significant. We are therefore unable to comment or provide evidence regarding younger age resulting in greater hearing aid benefit.

As previously mentioned, the degree of hearing impairment in the ear with better hearing may also influence the perception of improvement attributed to the BAHI. Most patients in the combined CHL group had poorer hearing in their better ears, while patients with U‐SNHL typically have normal or near‐normal hearing in their better ear. Patients with bilateral hearing impairment may perceive greater sound amplification from the BAHI, as they may not have experienced normal binaural hearing. On the other hand, patients with low audiological thresholds in their better ear may not notice a significant difference in sound amplification and therefore may not use their hearing device. Nonetheless, patients with normal contralateral hearing may still benefit from the device in subtle ways, such as improved speech understanding in noisy situations. The same argument can be applied to patients with congenital and acquired hearing loss; those with acquired hearing loss may be less satisfied with the BAHI if its performance does not match their previous experience of normal hearing.

4.2. Results in the Context of Other Evidence

The MDs from the APHAB and SSQ questionnaires of our meta‐analyses were similar to the ones calculated in the systematic review by Hampton et al. [8], lending strength to our analyses and results. As demonstrated in Table 3, the global MD for APHAB in the U‐SNHL population was 13.80, compared to Hampton's 15.5. The results for the APHAB and SSQ subscales were also similar. Unfortunately, we could not find any other reviews that reported benefit scores from disease‐specific QoL measures in CHL populations.

4.3. Limitations of the Evidence

Our evidence is limited due to significant methodological and clinical heterogeneity in most meta‐analyses.

Regarding methodological heterogeneity, the majority of the included articles did not specify the method of scale administration, limiting the validity of our results and possibly leading to erroneous patient responses. This is demonstrated by the large number of articles rated as having a high RoB in the “Measurement of exposure” item of the RoBANS. This issue applied to both face‐to‐face interviews and at‐home completion of the questionnaires. Compared to at‐home completion, face‐to‐face interviews were found to significantly improve the test–retest reliability of questionnaires among BAHI users, resulting in more accurate responses and less cognitive load [84]. Additionally, face‐to‐face interviews were deemed the optimal method for assessing the effectiveness of an intervention [84]. The methodological limitations of this systematic review are also emphasized by the substantial number of articles excluded during sensitivity analyses, attributable to their risk of bias assessments, retrospective design, or the combination of data from both children and adults in their aggregate questionnaire results, despite the age‐specific validation of the questionnaires. Consequently, a majority of the subscales that initially exhibited significant differences between populations with hearing loss were rendered statistically insignificant during sensitivity analyses. This observation reinforces the argument for the necessity of more robust methodologies in the collection of QoL data.

Regarding clinical heterogeneity, there was considerable variability in patient demographics both within and between the included studies. Firstly, the included studies involved participants with a wide age range, indicating a significant variation in their cognitive load capacity. This variation may have influenced their ability to sustain attention and focus on multiple questions over an extended duration. Secondly, there were multiple factors that we could not control for, such as surgical techniques, post‐operative complications, abutment type, cosmetic appearance, and sound processor technology. It is possible that the patients expressed their dissatisfaction with any of these outcomes by rating the device's performance poorly. Conversely, individuals may overlook inadequate sound processor performance if they have had a favorable post‐operative experience or to show appreciation to their clinicians, which may not accurately reflect their actual hearing capabilities. Thirdly, the nature of the BAHI (pBAHI or tBAHI) has been shown to affect QoL outcomes, with tBAHI users reporting greater QoL benefits compared to those who are fitted with a pBAHI [85]. Finally, due to the limited reporting of individual audiological thresholds, we grouped patients based on their type of hearing loss, regardless of severity, symmetry, and whether they were fitted with BAHIs unilaterally or bilaterally. A study by Noble & Gatehouse [86] identified correlations between audiological thresholds and SSQ scores, which demonstrated that patients with an asymmetrical hearing loss greater than 10 dB rated their hearing abilities lower in all SSQ subscales. A subsequent study also noted significant differences in perceived benefits between patients who were unilaterally or bilaterally fitted with hearing aids [87].

4.4. Implications for Practice and Future Directions

One of the strengths of this systematic review is the inclusion of a large number of articles and the computation of meta‐analyses for multiple disease‐specific QoL measures. This enabled a more comprehensive understanding and appreciation of the differences in QoL among BAHI patients with different types of hearing loss. Providing correct pre‐operative counseling and setting realistic expectations regarding the performance of the sound processor are crucial steps in ensuring long‐term benefits for BAHI patients [88]. It is now evident that QoL scales provide information on patient benefit that is independent of audiological tests and should be integrated into regular follow‐up procedures.

Nevertheless, further research is necessary to fully comprehend the complex nature of hearing‐related QoL and its association with hearing loss and BAHIs. More studies are needed to investigate underreported patient populations, such as pediatric and unilateral CHL/MHL patients. Furthermore, additional subgroup analyses are required to explore how patient demographics, sound processor technology, severity, and symmetry of hearing loss can impact QoL benefits in BAHI patients. Lastly, the disease‐specific QoL measures used in this systematic review were originally developed and validated for traditional air conduction hearing aid users, not for BAHI users [72, 73, 75]. Therefore, it is important to develop updated score percentiles, norms, and CMDs specific to the BAHI population in order to enhance the confidence and reliability of our findings.

5. Conclusion

This systematic review and meta‐analyses have uncovered noteworthy disease‐specific differences in QoL among BAHI patients with U‐SNHL and CHL. The findings indicate that CHL patients encounter fewer difficulties in noisy environments, perceive sounds with greater clarity, and experience better overall and psychosocial well‐being after BAHI fitting. Conducting subgroup analyses would further elucidate the impact of specific characteristics on QoL benefits. Further research is warranted to gather data on underrepresented populations and validate the norms of disease‐specific QoL measures for BAHI patients.

Conflicts of Interest

Previous research grant from Oticon for a multi‐centric clinical trial in MIPS.

Supporting information

Table S1. Search strategy.

Table S2. PRISMA 2020 checklist.

Table S3. PRISMA‐S checklist.

Table S4. Study characteristics of included articles.

Table S5. Descriptive statistics of IOI‐HA scores.

Table S6. Descriptive statistics of BBSSD scores.

Table S7. Descriptive statistics of SSQ12 scores.

Table S8. Descriptive statistics of SSQ‐B scores.

Table S9. Heterogenenity (I ²%) differences pre‐ and post‐sensitivity analysis.

Figure S1. Individual risk of bias assessment of the included studies using the Risk of Bias Assessment tool for Non‐Randomized Studies (RoBANS).

Figure S2. Forest plots showing mean differences in APHAB scores in U‐SNHL and CHL populations. Forest plots of the mean differences for the (A) global scale score, (B) ease of communication subscale, (C) background noise subscale, (D) reverberation subscale, and (E) aversiveness subscale for U‐SNHL and CHL populations. APHAB, Abbreviated Profile of Hearing Aid Benefit; CHL, conductive hearing loss; U‐SNHL, unilateral sensorineural hearing loss.

Figure S3. Forest plots showing mean differences in GBI scores in U‐SNHL and CHL populations. Forest plots of the mean differences for the (A) global scale score, (B) general subscale, (C) social subscale, and (D) physical subscale for U‐SNHL and CHL populations. CHL, conductive hearing loss; GBI, Glasgow benefit inventory; U‐SNHL, unilateral sensorineural hearing loss.

Figure S4. Forest plots showing mean differences in SSQ scores in U‐SNHL and CHL populations. Forest plots of the mean differences for the (A) speech subscale, (B) spatial subscale, and (C) qualities subscale for U‐SNHL and CHL populations. CHL, conductive hearing loss; SSQ, Speech, spatial, and qualities of hearing scale; U‐SNHL, unilateral sensorineural hearing loss.

Figure S5. Funnel plots for the global scale and subscales of the APHAB meta‐analysis articles. Funnel plots for the (A) global scale score, (B) ease of communication subscale, (C) background noise subscale, (D) reverberation subscale, and (E) aversiveness subscale of the APHAB measure for U‐SNHL and CHL populations. APHAB, Abbreviated Profile of Hearing Aid Benefit; CHL, conductive hearing loss; U‐SNHL, unilateral sensorineural hearing loss.

Figure S6. Funnel plots for the global scale and subscales of the GBI meta‐analysis articles. Funnel plots for the (A) global scale score, (B) general subscale, (C) social subscale, and (D) physical subscale of the GBI measure for U‐SNHL and CHL populations. CHL, conductive hearing loss; GBI, Glasgow benefit inventory; U‐SNHL, unilateral sensorineural hearing loss.

Figure S7. Funnel plots for the subscales of the SSQ meta‐analysis articles. Funnel plots for the (A) speech subscale, (B) spatial subscale, and (C) qualities subscale of the SSQ measure for U‐SNHL and CHL populations. CHL, conductive hearing loss; SSQ, Speech, spatial, and qualities of hearing scale; U‐SNHL, unilateral sensorineural hearing loss.

Théorêt K., Deng J., da Silva S. D., Guadagno E., and Daniel S. J., “Systematic Review of Quality of Life in Bone Anchored Hearing: Conductive vs. Unilateral Sensorineural Hearing Loss,” The Laryngoscope 135, no. 10 (2025): 3472–3484, 10.1002/lary.32229.