Introduction: The main objective of the study was to validate the Norwegian translation of the Speech, Spatial and Qualities of Hearing Scale (SSQ) and investigate the SSQ disability profiles in a cochlear implant (CI) user population. Methods: The study involved 152 adult CI users. The mean age at implantation was 55 (standard deviation [SD] = 16), and the mean CI experience was 5 years (SD = 4.8). The cohort was split into three groups depending on the hearing modality: bilateral CIs (BCIs), a unilateral CI (UCI), and bimodal (CI plus contralateral hearing aid; HCI). The SSQ disability profiles of each group were compared with those observed in similar studies using the English version and other translations of the SSQ. Standard values, internal consistency, sensitivity, and floor and ceiling effects were investigated, and the missing-response rates to specific questions were calculated. Relationships to speech perception were measured using monosyllabic word scores and the Norwegian Hearing in Noise Test scores. Results: In the BCI group, the average scores were around 5.0 for the speech and spatial sections and 7.0 for the qualities section (SD ∼2). The average scores of the UCI and HCI groups were about one point lower than those of the BCI group. The SSQ disability profiles were comparable to the profiles in similar studies. The slopes of the linear regression lines measuring the relationships between the SSQ speech and monosyllabic word scores were 0.8 per 10% increase in the monosyllabic word score for the BCI group (explaining 35% of the variation) and 0.4 for the UCI and HCI groups (explaining 22–23% of the variation). Conclusion: The Norwegian version of the SSQ measures hearing disability similar to the original English version, and the internal consistency is good. Differences in the recipients’ pre-implantation variables could explain some variations we observed in the SSQ responses, and such predictors should be investigated. Data aggregation will be possible using the SSQ as a routine clinical assessment in global CI populations. Moreover, pre-implantation variables should be systematically registered so that they can be used in mixed-effects models.

Self-assessment of hearing disability can effectively complement behavioural testing in clinically assessing cochlear implant (CI) users. The Speech, Spatial and Qualities of Hearing Scale (SSQ) is a self-report questionnaire for adults [Gatehouse and Noble, 2004; Noble and Gatehouse, 2004] comprising scored items in three different sections: 14 items on speech-hearing, 17 items on spatial hearing, and 18 items on other functions and qualities. The SSQ questionnaire was initially validated in English by Gatehouse and Noble [2004], who assessed a hearing-impaired (HI) group using hearing aids (HAs). The study group comprised 153 subjects with a mean age of 71 (standard deviation [SD] = 8.1) who had better and worse-ear average hearing thresholds of around 40 and 50 decibels (dB) HL, respectively, averaged over frequencies of 0.5, 1, 2, and 4 kHz. They found that the SSQ showed promise as an instrument for evaluating various interventions, particularly (but not exclusively) to restore binaural function. Furthermore, the mean total SSQ score was 5.5 (SD 1.9), with items in the speech section having generally lower scores than those in the other two sections.

The SSQ has been translated and validated in many languages. For example, Moulin et al. [2015] validated it in French and assessed its reproducibility across different languages (i.e., Dutch, English, and German). They analysed responses from normal-hearing participants and found that the French study group assessed themselves lower than the Dutch and English study groups. However, the authors reported that the differences in the mean values were well below one point and that the “pattern of the items was remarkably similar,” as shown by the correlations between the different language versions of the SSQ. Mertens et al. [2013] compared the SSQ disability profile of the HI group (described in the study by Gatehouse and Noble [2004], see the former section), with that of a CI group consisting of post-lingually deaf adults (n = 54) unilaterally implanted at mean age of 55 years. The comparison indicated that the disability profile of the CI group differed significantly from that of the HI group, with the CI group having a mean speech score of 3.9 (0.5 lower than that of the HI group), a mean spatial score of 3.4 (2.3 lower than that of the HI group), and a mean qualities score of 5.2 (1.1 lower than that of the HI group). There were significant differences in the spatial and qualities scores between both groups. In a multicultural longitudinal observational study by Lenarz et al. [2017], local language versions of the SSQ were used to evaluate pre-implantation of unilateral cochlear implants (UCI) and post-implantation outcomes at 1, 2, and 3 years. The mean age of the cohort was 50 (SD = 17; range = 13–91). One year post-implantation, the SSQ mean values for the speech, spatial, and qualities sections were 4.5, 4.6, and 5.5, respectively, with all values increasing slightly 3 years post-implantation (5.0, 4.9, and 6.3). Furthermore, they found less disability in the younger population (<65 years) than in the older population (≥ 65 years, comprising around 20% of the study population). Other studies have also used SSQ to measure the effects of cochlear implantation [Kelsall et al., 2017; Louza et al., 2017; Hey et al., 2018]. Specifically, Hey et al. [2018] found a statistically significant mean increase of 2.2, 2.0, and 1.9 points for the speech, spatial, and qualities sections, respectively, and reported that the values were close to the multicentre study by Lenarz et al. [2017]. In the study by Kelsall et al. [2017], which reported on the gains of receiving hybrid CIs in individuals with significant residual low-frequency hearing and severe-to-profound high-frequency sensorineural hearing loss, mean increases of 2.4, 1.2, and 1.5 were observed for all three SSQ sections. Furthermore, mean values were 5.6, 5.7, and 6.5 1 year post-implantation. In the study by Louza et al. [2017], the SSQ was used to measure the benefit of CI in 10 patients with single-sided deafness. They found statistically significant mean increases for the speech and spatial sections but not the qualities section. Overall, these studies have shown that SSQ is a useful tool for measuring the benefits of CIs.

Furthermore, some studies have used the SSQ questionnaire to compare the effect of bilateral cochlear implants (BCIs) to that of UCIs. In various longitudinal studies where participants rated their hearing when they had UCI and after receiving BCIs, mean SSQ ratings (speech, spatial, and qualities) improved after receiving BCIs with the greatest improvement in the spatial ratings [Summerfield et al., 2006; Dwyer et al., 2014; Reeder et al., 2014]. In the cross-sectional study by Laske et al. [2009], where groups were matched by age, CI experience, and gender, no statistically significant increase in the mean SSQ ratings was found in the BCI group (n = 29) compared to the UCI group (control group; n = 25). In a randomised controlled study by Smulders et al. [2016], 38 participants were randomised to receive UCIs or BCIs. The BCI group had superior SSQ ratings in all three sections 1 year post-operatively. Other cross-sectional studies [Summerfield et al., 2006; Laske et al., 2009; Smulders et al., 2016] have shown inconsistent between-group effects when comparing the UCI and BCI groups.

  • In the study by Laske et al. [2009], bilateral implantation showed better mean and median results than unilateral implantation, however, the differences were insignificant. Furthermore, the number of participants in each group was only 15.

  • In the study by Summerfield et al. [2006], a large positive effect of receiving BCIs was recorded in the spatial section of the SSQ. The study group consisted of 24 post-lingually deafened adults randomized to receive a second implant immediately or subsequently after receiving the first CI. The between-group effect on the spatial score was 1.7 and the within-group effect was 2.0. There was a smaller positive effect of receiving BCI on the qualities score; the between-group effect and the within-group effect was around 1.0. For the speech section, the effect of receiving BCIs was only significant in the within-group comparison and the increase in the mean speech score was less than one point.

  • Smulders et al. [2016] found significant differences in the mean values of all three SSQ sections between groups when randomising 38 participants in a group receiving BCIs and a group receiving UCIs. The self-reported hearing abilities were consistent with behavioural test results of speech recognition in noise when the noise came from different directions. Moreover, the BCI group participants were better than the UCI group in localising sounds.

These inconsistent findings indicate that other sources of variability exist and that the number of participants needs to be fairly high to measure the effects on the SSQ section scores. The sources of variability in SSQ scores were investigated by Moulin and Richard [2016] in 216 SSQ responses from HI adults in an ENT department where the mean age was 54 years (SD = 14; age range not reported). They found that hearing loss in the better ear and hearing loss asymmetry were the main predictors of SSQ scores. Regarding the relationship between age and SSQ ratings, they observed that age mostly affected the speech subscale items, though this is not quantified in their study. The Pearson correlation coefficients calculated between age and the three SSQ sections were −0.38, −0.26, and −0.25, respectively (p < 0.01). An age effect was also found in the study by Noble et al. [2009]. They recorded a trend towards higher SSQ ratings in younger individuals with CIs than in older individuals; however, the relationship with age was insignificant.

The SSQ as a self-reported outcome measure, is important as it accounts for the functional consequences of reduced hearing ability. The SSQ measures subjective hearing disability in different situations and aspects of hearing. Studies have investigated the relationship between behavioural tests (e.g., speech recognition and sound localisation) and subjective hearing assessments, and have reported discrepant findings regarding correlations between the behavioural and self-reported measures. Zhang et al. [2015] observed a moderate correlation (r = 0.6, p < 0.01) between the SSQ speech score and word recognition in a group of 19 BCI users. Laske et al. [2009] demonstrated a statistically significant correlation between the localisation performance and averaged SSQ spatial ratings in a study comprising 34 adult patients with BCIs. Other studies have demonstrated no correlation between behavioural tests and self-assessed hearing: Reeder et al. [2014] found no correlation between SSQ ratings and localisation or speech recognition in 21 sequentially bilaterally implanted CI recipients. A discrepant correlation between behavioural test results and self-assessed scores was also observed in a Korean study, which included 14 bimodal listeners [Heo et al., 2013]. They found that individual bimodal benefits varied across tasks and test materials. They suggested clinicians should be careful when predicting subjective bimodal advantages, based on behavioural localisation and speech recognition measures.

Furthermore, the correlations between SSQ scores and behavioural hearing ability tests were investigated in a randomised controlled study of patients receiving unilateral and BCIs [Ramakers et al., 2017]. They found a moderate correlation between the SSQ spatial score and localisation test results and weak-to-moderate correlations between the SSQ and speech perception test results. Interestingly, Ramakers et al. [2017] compared the correlations in BCI and UCI patients and found that all correlations were stronger in the BCI group than in the UCI group, although no correlation differed significantly. They considered that the latter finding could be explained by the small sample sizes, 19 in each group.

The findings of the previously mentioned studies show that the variability in SSQ scores can be extensive even within groups and that a relatively high number of participants that use the SSQ are required to investigate the effects of certain exposure variables, such as investigating the gain of BCIs against UCIs in between-groups comparisons. Furthermore, to compare the results of cohorts from different nations or participate in multinational studies, translated and validated versions of the SSQ must be available. This study aimed to validate the Norwegian translation of the SSQ in a large population of adult CI users with different modalities: BCI, bimodal cochlear implant (CI plus contralateral HA; HCI), and UCI. The validation included investigating the relationship between the SSQ section scores, age, and speech perception in quiet and noise. Reliability in terms of internal consistency was measured using Cronbach’s alpha.

The SSQ Questionnaire

The questionnaire has three different sections identifying three main domains of hearing ability: speech (14 questions), spatial (17 questions), and qualities (18 questions). The speech section addresses speech-hearing issues in different contexts and environments. The spatial section addresses abilities such as sound localisation, including the distance to and movements of sound objects. The qualities section reports the ability to segregate and recognise sounds, clarity of sounds, and listening effort. The SSQ is scored on an analogue rating scale from 0 to 10, where 10 indicates no disability. Section scores are calculated by averaging the ratings for questions in each section, and the total score is calculated by averaging the ratings of all 48 questions.

Akeroyd et al. [2014] reported a factor analysis of the SSQ and found three clear factors of “speech understanding,” “spatial separation,” and “clarity, separation, and identification” which corresponded to the three SSQ sections. Original reports on the SSQ analysed responses to the individual items, leading to quite complex sets of tables. To reduce this complexity, a pragmatic set of subscales was developed by Gatehouse and Akeroyd [2006]. Based on the content and intent of each SSQ item, the subscales were as follows: speech in quiet (SiQ), speech in noise (SiN), speech in speech contexts (SiSCont), multiple speech-stream processing and switching (MultStream), localisation (Loc), distance and movement (DisMov), segregation of sounds (SegSnds), identification of sound and objects (IdSnd), sound quality and naturalness (Qlty), and listening effort (Eff). Table 1 shows the distribution of the SSQ items included in each subscale.

Table 1.

Sections and subscales of the Speech, Spatial and Qualities of Hearing Scale (SSQ)

SSQ sectionSSQ subscaleSSQ items included in the subscales
Speech (items 1–14) Speech in quiet (SiQ) 2, 3 
Speech in noise (SiN) 1, 4–6 
Speech in speech contexts (SiSCont) 7–9, 11 
 Multiple speech streams (MultStream) 10, 12, 14 
Spatial (items 1–17) Localisation (Loc) 1–6 
 Distance and movement (DisMov) 7–13, 15, 16 
Qualities (items 1–18) Segregation of sound (SegSnds) 1–3 
Identification of sound and objects (IdSnd) 4–7, 13 
Sound/quality/naturalness (Qlty) 8–12 
Listening effort (Eff) 14, 18 
SSQ sectionSSQ subscaleSSQ items included in the subscales
Speech (items 1–14) Speech in quiet (SiQ) 2, 3 
Speech in noise (SiN) 1, 4–6 
Speech in speech contexts (SiSCont) 7–9, 11 
 Multiple speech streams (MultStream) 10, 12, 14 
Spatial (items 1–17) Localisation (Loc) 1–6 
 Distance and movement (DisMov) 7–13, 15, 16 
Qualities (items 1–18) Segregation of sound (SegSnds) 1–3 
Identification of sound and objects (IdSnd) 4–7, 13 
Sound/quality/naturalness (Qlty) 8–12 
Listening effort (Eff) 14, 18 

As suggested by Gatehouse and Akeroyd [2006], the following items were excluded from the subscales – speech section: item 13; spatial section: items 14 and 17; qualities section: items 15–17.

Norwegian SSQ

The SSQ was translated from English into Norwegian and independently translated back into English by Amesto Translations. Noble, the co-author of the original article by Gatehouse and Noble [2004], checked the back-translation of the SSQ. After receiving feedback on the items needing modification, seven of nine items were edited in the original Norwegian translation. The translations of the remaining two items were related to the back-translation, and their initial translations were found to be adequate. The final Norwegian SSQ was the edited version, based on feedback from the back-translation.

The SSQ was originally validated using an interview format [Gatehouse and Noble, 2004]; however, self-administration has since been used in several studies [Noble et al., 2008, 2009; Laske et al., 2009; Banh et al., 2012; Dwyer et al., 2014]. The reliability of the interview and self-administration formats is comparable [Singh and Pichora-Fuller, 2010]. Self-administration is preferred in clinical settings because the interview format is considered time-consuming. Therefore, in this study, participants completed the SSQ in a self-reporting format using paper and pencil in the clinic during follow-up visits or at home, returning the questionnaire by mail. Speech perception tests were administered during the same follow-up visit.

Ratings for the different items were averaged to yield the scores for the three sections (speech, spatial, and qualities). In addition, the values of the 10 subscales (SiQ, SiN, SiSCont, MultStream, Loc, DisMov, SegSnds, IdSnd, Qlty, and Eff) were calculated by averaging the items, as described in Table 1. The participants were not required to respond to every question. Sometimes, they responded to a question by saying “not applicable” or otherwise did not respond. The average “not applicable” responses per questionnaire were 1.5 (SD = 2.7), ranging from 0 to 15. The average missing-response rate was 3.0% (SD = 2.6%), with values ranging from 0% to 9.3%, except for question 16 in the qualities section, which had a 13.9% missing-response rate. This question had up to 45.2% missing responses in a study by Akeroyd et al. [2014].

Participants

The participants were recruited during follow-up visits at the hospital. Adult participants with post-lingual severe-to-profound hearing loss were asked to participate and complete the SSQ questionnaire during these CI follow-up visits at the hospital. The study included participants who were at least 18 years old when they received CIs and had used their CIs for at least 1 year. The participants’ demographics are presented in Table 2. For patients in the UCI and HCI groups, the speech perception test scores using only the CI were required to be equal to or better than when using only the contralateral HA. For participants in the HCI group, the hearing sensitivity thresholds, calculated as the pure tone average (PTA) of four frequencies (0.5, 1, 2, and 4 kHz) for the non-implanted ear, were all above 70 dB HL, except for three individuals with PTAs of 53, 67, and 68 dB HL. A 120 dB HL threshold was used in the PTA calculations when thresholds were not obtained. The mean PTA of the non-implanted ears was 97 (SD = 14) dB HL. When the participants in the HCI group were tested with only the HA, the mean speech recognition score on monosyllables was 16% (SD = 21%; 95% confidence interval: 6–27%); the range was from 0% to 46%, with one outlier at 82%.

Table 2.

Demographics of the 152 participants

BCI (n = 29)HCI (n = 56)UCI (n = 67)
Age at 1st CI, years 
 Mean (SD) 46.2 (14.2) 60.7 (15.1) 54.0 (16.5) 
 Median (range) 44.7 (20.6–78.2) 60.3 (26.4–90.1) 53.3 (23.9–88.6) 
Age at 2nd CI, years 
 Mean (SD) 51.0 (15.0)   
 Median (range) 48.7 (21.6–80.9)   
Duration of CI usage, years 
 Mean (SD) 8.0 (5.2) 3.3 (2.9) 5.3 (5.3) 
 Median (range) 6.0 (2.4–23.4) 2.2 (1.0–13.1) 3.7 (1.0–24.2) 
Duration of 2nd CI usage, years 
 Mean (SD) 3.3 (2.3)   
 Median (range) 2.2 (1.0–8.1)   
Gender, n (% female) 21 (72) 38 (68) 40 (60) 
Implant type (n) (Nucleus-MED-EL-Advanced Bionics) n = (9–16–4) n = (28–23–5) n = (46–20–1) 
BCI (n = 29)HCI (n = 56)UCI (n = 67)
Age at 1st CI, years 
 Mean (SD) 46.2 (14.2) 60.7 (15.1) 54.0 (16.5) 
 Median (range) 44.7 (20.6–78.2) 60.3 (26.4–90.1) 53.3 (23.9–88.6) 
Age at 2nd CI, years 
 Mean (SD) 51.0 (15.0)   
 Median (range) 48.7 (21.6–80.9)   
Duration of CI usage, years 
 Mean (SD) 8.0 (5.2) 3.3 (2.9) 5.3 (5.3) 
 Median (range) 6.0 (2.4–23.4) 2.2 (1.0–13.1) 3.7 (1.0–24.2) 
Duration of 2nd CI usage, years 
 Mean (SD) 3.3 (2.3)   
 Median (range) 2.2 (1.0–8.1)   
Gender, n (% female) 21 (72) 38 (68) 40 (60) 
Implant type (n) (Nucleus-MED-EL-Advanced Bionics) n = (9–16–4) n = (28–23–5) n = (46–20–1) 

BCI, bilateral cochlear implant recipient; CI, cochlear implant; HCI, HA in one ear and CI in the contralateral ear; SD, standard deviation; UCI, unilateral cochlear implant recipient.

Speech Recognition

Speech recognition was measured using (1) percentage scores on correctly repeated monosyllables in phonetically balanced word lists of 50 words presented at 65 dBA [Øygarden, 2009]; (2) percentage scores of correctly repeated words in sentences presented at 65 dBA in quiet, using the sentence lists from the Norwegian Hearing in Noise Test (HINT) [Myhrum and Moen, 2008; Soli and Wong, 2008; Myhrum et al., 2016]; and (3) HINT speech reception thresholds (SRTs) in 60 dBA speech spectrum noise under the test condition, in which speech and noise originated from a speaker in front of the patient. The HINT is an adaptive sentence recognition test that estimates the signal-to-noise ratio at which the listener obtains 50% correctly repeated sentences. The HINT score is an SRT expressed as a signal-to-noise ratio in dB, and increasing SRT values reflect reduced speech recognition in noise. Since correct responses require all words in the sentence to be correctly repeated, the HINT was only applied to participants with at least 75% correctly repeated words in sentences presented in quiet.

The speech recognition tests were administered during the CI follow-up visits at the hospital, and the best-aided monosyllabic word and HINT scores from the latest visit were used in the study. Overall, 123 participants (29 participants missing) completed the monosyllabic word recognition test, and 138 participants (14 participants missing) completed the sentences in the quiet test. Further, 85 participants had HINT scores above 75% and, as such, possessed a word score in quiet that was sufficiently high, according to the clinical test protocol, to use the HINT to find a HINT SRT in noise. The HINT SRT test was administered to 78 participants.

Data Analysis

Boxplots were used to illustrate the distributions of the SSQ subscale scores. The descriptives (mean, median, SD, and range) of the SSQ subscale scores were calculated for each participant group (BCI, HCI, and UCI). The data were considered normally distributed, and parametric tests were applied to determine if significant differences existed between the HCI and BCI groups and between the UCI and BCI groups. Multiple t tests were conducted, and the significance level was set to α = 0.01. Thus, instead of reporting the exact p value, the significance levels of α = 0.01 and α = 0.05 were denoted in Table 3 for readability. The age differences between groups were adjusted for using multiple linear regression models for SSQ subscale scores with age as a continuous exposure variable and the group as a binary exposure variable (using two models, one with binary variable HCI and BCI and one with binary variable UCI and BCI). Internal consistency was determined using item-to-total correlations and Cronbach’s alpha coefficients of the three SSQ section scores.

Table 3.

Mean SSQ section scores and subscale scores for BCI, HCI, and UCI participants

BCI (n = 29)HCI (n = 56)UCI (n = 67)
SSQ section score: median; mean (SD) 
 Speech (part 1) 4.9; 4.9 (2.2) 3.7; 3.9 (1.9)* 3.9; 4.2 (1.8) 
 Spatial (part 2) 5.3; 5.0 (2.4) 3.9; 4.0 (2.2)* 3.4; 3.6 (2.1)** 
 Qualities (part 3) 7.4; 7.0 (2.0) 5.8; 5.5 (2.1)** 6.2; 5.9 (1.9)* 
SSQ subscale score: median; mean (SD) 
 Speech in quiet (SiQ) 8.0; 7.9 (2.2) 7.0; 6.8 (2.2)* 7.3; 6.9 (2.2) 
 Speech in noise (SiN) 4.3; 4.3 (2.3) 3.3; 3.5 (2.1) 3.5; 3.7 (2.0) 
 Speech in speech contexts (SiSCont) 5.3; 5.3 (2.8) 3.8; 3.9 (2.2)* 4.3; 4.3 (2.3) 
 Multiple speech streams (MultStream) 2.3; 2.7 (2.2) 1.9; 2.4 (1.8) 2.0; 2.4 (1.9) 
 Localisation (Loc) 6.3; 5.3 (2.8) 3.7; 3.9 (2.5)* 3.4; 3.5 (2.7)** 
 Distance and movement (DisMov) 5.0; 4.6 (2.6) 3.9; 3.9 (2.1) 3.0; 3.6 (2.1)* 
 Segregation of sound (SegSnds) 8.3; 7.7 (2.2) 6.3; 5.9 (2.8)** 7.0; 6.5 (2.3)* 
 Identification of sound and objects (IdSnd) 7.4; 6.9 (2.4) 5.8; 5.6 (2.2) 6.6; 6.1 (2.1) 
 Sound/quality/naturalness (Qlty) 8.5; 7.9 (2.3) 6.6; 6.1 (2.3)** 7.3; 6.8 (2.3)* 
 Listening effort (Eff) 5.1; 5.0 (2.7) 3.5; 4.0 (2.5) 4.0; 4.4 (2.4) 
BCI (n = 29)HCI (n = 56)UCI (n = 67)
SSQ section score: median; mean (SD) 
 Speech (part 1) 4.9; 4.9 (2.2) 3.7; 3.9 (1.9)* 3.9; 4.2 (1.8) 
 Spatial (part 2) 5.3; 5.0 (2.4) 3.9; 4.0 (2.2)* 3.4; 3.6 (2.1)** 
 Qualities (part 3) 7.4; 7.0 (2.0) 5.8; 5.5 (2.1)** 6.2; 5.9 (1.9)* 
SSQ subscale score: median; mean (SD) 
 Speech in quiet (SiQ) 8.0; 7.9 (2.2) 7.0; 6.8 (2.2)* 7.3; 6.9 (2.2) 
 Speech in noise (SiN) 4.3; 4.3 (2.3) 3.3; 3.5 (2.1) 3.5; 3.7 (2.0) 
 Speech in speech contexts (SiSCont) 5.3; 5.3 (2.8) 3.8; 3.9 (2.2)* 4.3; 4.3 (2.3) 
 Multiple speech streams (MultStream) 2.3; 2.7 (2.2) 1.9; 2.4 (1.8) 2.0; 2.4 (1.9) 
 Localisation (Loc) 6.3; 5.3 (2.8) 3.7; 3.9 (2.5)* 3.4; 3.5 (2.7)** 
 Distance and movement (DisMov) 5.0; 4.6 (2.6) 3.9; 3.9 (2.1) 3.0; 3.6 (2.1)* 
 Segregation of sound (SegSnds) 8.3; 7.7 (2.2) 6.3; 5.9 (2.8)** 7.0; 6.5 (2.3)* 
 Identification of sound and objects (IdSnd) 7.4; 6.9 (2.4) 5.8; 5.6 (2.2) 6.6; 6.1 (2.1) 
 Sound/quality/naturalness (Qlty) 8.5; 7.9 (2.3) 6.6; 6.1 (2.3)** 7.3; 6.8 (2.3)* 
 Listening effort (Eff) 5.1; 5.0 (2.7) 3.5; 4.0 (2.5) 4.0; 4.4 (2.4) 

CI, cochlear implant; BCI, bilateral CI recipients; HCI, CI in one ear and HA in the contralateral ear; SD, standard deviation; SSQ, Speech, Spatial and Qualities of Hearing Scale; UCI, unilateral CI recipients.

*Denotes significantly different mean (p < 0.05), compared with the BCI group, based on the t test.

**Denotes significantly different mean (p < 0.01), compared with the BCI group, based on the t test.

To investigate the relationships between the SSQ and monosyllabic word scores, regression lines were calculated and plotted in scatter plots of SSQ subscale scores and monosyllabic word scores. The goodness-of-fit or R-squared of the regression lines was used to measure the amount of variance in the SSQ subscales explained by the monosyllabic word score. To adjust for age, three multiple linear regression models were employed, with the three SSQ subscales as dependent variables and the monosyllabic word score and age as independent variables. Additionally, the relationships between the SSQ subscale and HINT scores were investigated. The HINT requires substantial hearing ability, which is expected to produce improved SSQ scores. Therefore, participants who obtained HINT SRT scores are expected to have higher SSQ ratings than those who did not. Consequently, the differences between the means values of the SSQ section scores were compared for those who obtained HINT SRT scores and those who did not.

All statistical analyses were conducted using IBM SPSS Statistics software, version 27.0 (IBM SPSS, Chicago, IL, USA).

SSQ Results

Figure 1 shows boxplots of the three SSQ scores for each modality group (BCI, HCI, and UCI), and Table 3 provides the group-wise descriptives of the three SSQ section scores and the subscale scores. The speech and spatial sections had the lowest SSQ scores (highest disability), which were more than 2.0 points below the mean qualities score on average. Furthermore, the subscales concerning speech perception in noise (SiN, SiSCont, and MultStream) and sound location (Loc and DisMov) had the lowest SSQ scores.

Fig. 1.

Boxplot of ratings for the mean SSQ scores for the UCI, HCI, and BCI. The boxes represent the first to third quartiles, and the whiskers represent the range from the smallest to the largest non-outlier. An outlier is considered to be more than 1.5 times the interquartile range.

Fig. 1.

Boxplot of ratings for the mean SSQ scores for the UCI, HCI, and BCI. The boxes represent the first to third quartiles, and the whiskers represent the range from the smallest to the largest non-outlier. An outlier is considered to be more than 1.5 times the interquartile range.

Close modal

The mean SSQ speech scores among the groups were not significantly different. However, we observed significantly higher spatial and qualities scores in the BCI group than in the UCI and HCI groups (Table 3). Group-wise age descriptives are reported in Table 4. There was a significant difference in the mean age between the BCI and HCI groups (p = 0.004) but no significant mean differences between any of the other groups (BCI and UCI, and UCI and HCI, p > 0.05). Since the participants in the BCI group were younger than those in the UCI and HCI groups on average, age was adjusted for using group and age as factors in multiple linear regression models. Consequently, we observed that the Loc and SegSnds subscales remained higher in the BCI group than in the UCI group (p < 0.01).

Table 4.

Age distribution and speech perception test results for BCI, HCI, and UCI participants

BCIHCIUCI
Age, years 
 Mean (SD) 54.2 (15.1) 63.9 (14.1)** 59.2 (16.5) 
 Median (range) 51.5 (28.1–82) 63.3 (32.4–91.2) 60.9 (26.5–90.1) 
Monosyllabic score, % 
 Mean (SD) 70.0 (17) 61.4 (19.0) 66.0 (18.8) 
 Median (range) 74.0 (32–92) n = 27 64.0 (15–96) n = 47 70.0 (20–92) n = 52 
HINT in quiet, % 
 Mean (SD) 88.4 (14.0) 82.4 (16.9) 88.7 (12.7) 
 Median (range) 95.0 (55–100) n = 25 89.5 (38–100) n = 35 94.0 (48–100) n = 39 
HINT SRT (dB SNR) 
 Mean (SD) 5.6 (4.7) 5.6 (3.9) 6.6 (3.7) 
 Median (range) 3.6 (−1.4–16) n = 20 4.8 (0.1–16.5) n = 27 6.6 (0.6–13.6) n = 32 
BCIHCIUCI
Age, years 
 Mean (SD) 54.2 (15.1) 63.9 (14.1)** 59.2 (16.5) 
 Median (range) 51.5 (28.1–82) 63.3 (32.4–91.2) 60.9 (26.5–90.1) 
Monosyllabic score, % 
 Mean (SD) 70.0 (17) 61.4 (19.0) 66.0 (18.8) 
 Median (range) 74.0 (32–92) n = 27 64.0 (15–96) n = 47 70.0 (20–92) n = 52 
HINT in quiet, % 
 Mean (SD) 88.4 (14.0) 82.4 (16.9) 88.7 (12.7) 
 Median (range) 95.0 (55–100) n = 25 89.5 (38–100) n = 35 94.0 (48–100) n = 39 
HINT SRT (dB SNR) 
 Mean (SD) 5.6 (4.7) 5.6 (3.9) 6.6 (3.7) 
 Median (range) 3.6 (−1.4–16) n = 20 4.8 (0.1–16.5) n = 27 6.6 (0.6–13.6) n = 32 

BCI, bilateral cochlear implant recipient; CI, cochlear implant; HCI, HA in one ear and CI in the contralateral ear; dB, decibels; HINT, Hearing in Noise Test; SD, standard deviation; SNR, signal-to-noise ratio; SRT, speech reception threshold; SSQ, Speech, Spatial and Qualities of Hearing Scale; UCI, unilateral cochlear implant recipient.

*Denotes significantly different mean (p < 0.05), compared with the BCI group, based on the t test.

**Denotes significantly different mean (p < 0.01), compared with the BCI group, based on the t test.

The linear regression analyses, with age as the exposure variable and SSQ scores as dependent variables, showed that age significantly predicts the SSQ speech and qualities scores. The correlation coefficients of the SSQ speech and qualities scores were −0.286 and −0.314, while the regression coefficients were −0.036 and −0.041, respectively (p < 0.001). This means the SSQ speech and qualities scores are reduced by around 0.4 per age decade in the study group. When the binary group variable is added to the model (coding BCI and HCI as a binary variable in one model and BCI and UCI as a binary variable in the other), these multiple linear regression models explain a greater amount of the variation in the SSQ scores, indicated by slightly larger R-squared values. Additionally, the age exposure variable coefficients remained between −0.035 and −0.046 for the SSQ speech and qualities scores. Typically, the coefficients of the binary modality variables (BCI vs. HCI and BCI vs. UCI) were lower when age was included in the model, than when the mean values were compared without correcting for age (simple linear regression model using only binary groups as exposure variable, which is equivalent to the t tests).

Speech Perception and Relationships in SSQ Ratings

Table 4 shows that differences in mean values among the groups were not significant for the monosyllabic word scores, the sentences in quiet scores, or HINT SRT results. Scatter plots were used to illustrate the relationships between speech perception and SSQ and to analyse these relationships. Figure 2a shows a scatter plot of the SSQ speech score against the monosyllabic word score. The scatter points were differentiated with respect to the BCI, HCI, and UCI groups. All three regression lines in Figure 2a have statistically significant slopes, with slope values of 0.8 (BCI), 0.4 (HCI), and 0.4 (UCI). The slope values are measured in units of SSQ speech score per 10% increase in monosyllabic word score, meaning that when the monosyllabic word score increased by 10%, the speech-hearing ability increased by 0.8 on average (illustrated by the regression line of the BCI group). Furthermore, the monosyllabic word score explained 35% of the variation in the speech score in the BCI group (R2 = 0.35), 23% in the UCI group (R2 = 0.23), and 22% in the HCI group (R2 = 0.22). Similar scatter plots of the SSQ spatial and qualities scores versus the monosyllabic word scores are provided in Figure 2b and c. The monosyllabic word score explained 31% of the variation in the spatial score in the BCI group (slope = 0.8 per 10% monosyllabic word score). However, it was not correlated to the spatial score in the HCI and UCI groups. For the SSQ qualities score, the monosyllabic word score explained 43% of the variation in the BCI group, 23% in the UCI group, and 13% in the HCI group (the slopes were 0.7, 0.5, and 0.3, respectively; the result of the HCI group was not significant [p = 0.057]).

Fig. 2.

Scatter plot of SSQ speech scores (a), SSQ spatial scores (b), and SSQ qualities scores (c) versus the assessed monosyllabic word recognition score. The scatter points were differentiated with respect to the UCI, HCI, and BCI participant groups.

Fig. 2.

Scatter plot of SSQ speech scores (a), SSQ spatial scores (b), and SSQ qualities scores (c) versus the assessed monosyllabic word recognition score. The scatter points were differentiated with respect to the UCI, HCI, and BCI participant groups.

Close modal

The correlation between the monosyllabic word scores and the HINT word scores of sentences was r = 0.67 (R2 = 45%). We found a significant correlation between the HINT word scores and each SSQ section score in the BCI group (R2 = 14%, 37%, and 56%). However, no significant correlations were found between the HINT word scores and any of the mean SSQ scores for the UCI or HCI groups, except for a mild relationship (low correlation coefficient) between the HINT word scores and the SSQ speech scores in the UCI group (R2 = 14%; p = 0.022). Thus, the word scores on HINT sentences presented in quiet had a lower association with the SSQ speech score in the UCI and HCI groups, than with the monosyllabic word score. We found a similar pattern of relationships when investigating the relationship between HINT SRT and SSQ: a strong relationship was found between the HINT SRT and each SSQ section score in the BCI group (R2 = 35%, 27%, and 41%, respectively), but no significant correlations were found in the UCI and HCI groups (p > 0.233 for all variable pairs used to calculate the correlation between the HINT SRT and the averaged SSQ scores).

The correlation analysis of speech perception and the 10 SSQ subscales showed that, for the BCI group, the correlation coefficients between the HINT SRT scores and the SSQ subscales were r < −0.389 for all variable pairs, except between the HINT SRT scores and the SegSnds subscale (p = 0.885). For the latter non-significant correlation, all participants with HINT SRT scores had scored themselves high on questions related to sound segregation, resulting in a ceiling effect for the SegSnds variable. Furthermore, for the UCI and HCI groups, no significant correlations existed between HINT SRT scores and any of the SSQ subscales. When we analysed the differences between the mean SSQ section scores for participants who had obtained HINT SRT scores (n = 78) and those who had not (n = 74), we found consistently superior mean SSQ scores for participants with a HINT SRT score. This difference was significantly increased for the mean speech and qualities scores (p < 0.01). For example, the mean SSQ speech score was 3.3 for participants without HINT SRT scores and 5.0 for participants with HINT SRT scores. For the mean spatial score, there was no significant difference (p = 0.095).

Psychometric Properties

Cronbach’s alpha for the speech, spatial, and qualities sections was 0.94, 0.95, and 0.94, respectively, indicating a strong correlation among the items in each SSQ section. All SSQ section and subscale scores were significantly correlated, with coefficients ranging from 0.3 (between the SiQ and Loc) to 0.8 (between the SiN and SiSCont scores) between the subscales. The Pearson correlation coefficient was 0.61 between the speech and spatial section scores, 0.78 between speech and qualities section scores, and 0.59 between spatial and qualities section scores (p < 0.01). The Pearson correlation coefficients between each section score and the total score were 0.90, 0.85, 0.90 (p < 0.01). All coefficients of item-to-total correlations (within the three sections) ranged from 0.52 to 0.81 for the speech section, 0.43–0.85 for spatial section, and 0.56 and 0.77 for the qualities section. Question 14 in the spatial section (“Do the sounds of things you are able to hear seem to be inside your head rather than out there in the world?”) had an item-to-total correlation score below 0.5, which indicates that the response of this question is somewhat inconsistent with the average behaviour of the other questions’ responses.

SSQ Results

The lowest section scores were the speech and spatial scores, which were around 2.0 points lower than the qualities score on average. Thus, increased disability ratings were observed in questions regarding speech and spatial hearing. More specifically, the average scores were 4.9, 5.0, and 7.0 for the BCI group and 4.2, 3.6, and 5.9 for the UCI group (speech, spatial, and qualities scores, respectively). Similar values of 5.0, 4.9, and 6.3 were reported by Lenarz et al. [2017], described in the introduction section for SSQ assessed 3 years post-implantation (in a study population with a mean age of 50 years). Thus, section scores in our study had a less than one point difference from the scores in the study by Lenarz et al. [2017], except for a spatial score in the UCI group in our study (3.6 compared to 4.9). Lenarz et al. [2017] reported a statistically significant effect of age on the speech score when comparing participants younger and older than 65. We observed a moderate age effect on the speech and qualities scores (correlation coefficient ∼0.3): the SSQ speech and qualities scores were reduced by around 0.4 for every decade-increased age (regression coefficients were −0.36 and −0.41, respectively, per decade increase). The participants in the UCI group in our study were 9 years older than the participants in the study by Lenarz et al. [2017] on average, and age might be an explanatory variable for the spatial score being lower in our study than in the study by Lenarz et al. [2017].

The subscales concerning speech perception in noise (SiN, SiSCont, and MultStream) and sound location (Loc and DisMov) had the lowest SSQ scores (Table 3), which is reasonable. There was a significant group mean difference in the Loc subscale between the BCI and UCI groups, even after correcting for age group differences. Thus, the BCI advantage is measured in the Loc subscale, and the spatial section score. To investigate whether the disability profiles of the Norwegian SSQ are comparable to those of other SSQ language versions, we compared the SSQ disability profiles (the three section scores and 10 subscale scores) in our study with the profiles reported in other studies. We found that the SSQ disability profile of the UCI group differed by less than one scale point from similar profiles obtained from UCI groups in the study by Mertens et al. [2013] and Dwyer et al. [2014]. The only exception was the SSQ Loc subscale in the present UCI group, which was 1.3 points higher than the score in the UCI group in the study by Dwyer et al. [2014] (3.5 compared to 2.2). Still, it was only 0.4 points higher than the value in the UCI study group in the study by Mertens et al. [2013]. When investigating the disability profile of the BCI group, we observed that the profile variation was similar to that of the BCI group in the study by Dwyer et al. [2014]; however, these scores in the current BCI group were generally lower. The differences between the SSQ scoring of the present BCI group compared to those in the BCI group in the study by Dwyer et al. [2014] were especially large (mean difference [d] was equal to or larger than 1.5) for the speech score (d = 1.5) and the SiN (d = 1.9), MultStream (d = 2.1), and Eff (d = 1.6) subscales. However, since the mean scores of the speech section and the SiN, MultStream, and Eff subscales in the present UCI group are comparable to those in the UCI groups in the study by Dwyer et al. [2014] and Mertens et al. [2013], we have no reason to believe that the participants understood the questions in the Norwegian SSQ as more difficult listening situations than the corresponding questions in the English and Dutch versions.

Associations of Speech Perception

We found a significant correlation between the monosyllabic word score and the SSQ speech score for all modalities (BCI, HCI, and UCI) which is that the monosyllabic word score explained around 20–30% of the variation in the SSQ speech score. Correlations between the monosyllabic and SSQ spatial scores were observed only in the BCI group. A strong correlation existed between the monosyllabic word score and the SSQ qualities score in the BCI group, while a weak correlation existed in the UCI group dataset; however, no correlation existed between the monosyllabic word score and the SSQ qualities score in the HCI group. The strong relationships between the monosyllabic word score and SSQ section scores in the BCI group would be an interesting objective for further studies.

Moderate correlations (r < −0.4) were found between the HINT SRT scores and all SSQ subscale scores, except for the SegSnds subscale, which was not correlated with the HINT SRT scores. However, to measure a correlation, there needs to be some range of variation in both variables, and the SegSnds responses varied within only a small range, as participants responded that they had little disability regarding sound segregation. Furthermore, for the UCI and HCI groups, no correlations were observed between the HINT SRT scores and any of the three SSQ section scores. Since the adaptive HINT test (HINT SRT) is a difficult test, it can only be applied to individuals with relatively good speech perception in quiet, implying a relatively narrow range of outcomes, which limits the possibility of detecting relationships. Therefore, to overcome the relatively small variation in HINT SRT scores and to include all participants in the analysis of the relationship of SSQ scores and speech recognition in noise, the SSQ results of participants tested with HINT were compared with those of participants not tested with HINT. We found that the group mean SSQ scores for the speech and qualities sections were significantly better for participants capable of being tested with HINT; however, they did not assess themselves with statistically significantly less disability in the spatial section questions. We know that the spatial listening abilities when measured with localisation tests usually are correlated to speech recognition scores, but we did not observe a relationship between the SSQ spatial scores and HINT SRT yes/no.

As previously discussed, speech perception test results can explain some variations in the SSQ scores. Other behavioural tests might explain more of the variation in SSQ scores, such as localisation and speech perception in noise tests with speech and noise presented from different directions. For example, Laske et al. [2009] found that localisation test results were correlated with the SSQ spatial scores, though not quantified.

The bilateral advantage was not observed when comparing mean speech perception scores among the groups but was observed in the SSQ spatial section and Loc subscale scores, as reported in the previous section. Furthermore, the relationship between speech perception and SSQ section scores was evident in the BCI group but inconsistently observed in the UCI and HCI groups. Thus, speech perception was a weak or absent predictor of SSQ self-assessment in the UCI and HCI groups.

The Norwegian SSQ measures hearing disability similar to the original English version, and the internal consistency is good. The SSQ responses show large variations within the CI groups, indicating large variation in the subjective benefit or hearing disability experienced by CI users. We found positive relationships between the SSQ scores and speech perception, as well as negative relationships between SSQ scores and age. However, these relationships were weak and inconsistent. Thus, we neither focused on the relationships between SSQ scores and the recipients’ pre-implantation variables nor compared pre-implant SSQ scores with post-implant scores. Data aggregation will be possible using the SSQ as a routine clinical assessment pre- and post-implantation in global CI populations. Furthermore, pre-implantation variables such as speech perception and duration of deafness, should be registered so that they can be used in mixed-effects models.

We thank Dr. William Noble at the MRC Institute of Hearing Research for checking the back-translated version of the Norwegian SSQ, which was modified according to his feedback.

The study followed Norwegian legislation and the internal regulations at Oslo University Hospital (Oslo, Norway). Approval was obtained from the privacy and data protection officer without any objections. The study protocol was reviewed and approved by the Regional Committees for Medical and Health Research Ethics, with approval number 2010/17217. Written informed consent was obtained from all the participants.

The authors have no conflicts of interest related to this publication.

No funding was received for this study.

All authors have made substantial, direct, and intellectual contributions to the work and approved it for publication. M.M., M.G.H., and O.E.T. had the most substantial contributions in organising the conceptual design of the study, with contributions from A.K.R. and G.E.J. The statistical analysis was done by M.M., with contributions mainly from M.G.H., O.E.T., and A.K.R. M.M. drafted the manuscript, and M.G.H., S.R.K., O.E.T., A.K.R., and G.E.J. revised the versions and approved the final version before submission of the final draft.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Further enquiries can be directed to the corresponding author.

1.
Akeroyd MA, Guy FH, Harrison DL, Suller SL. A factor analysis of the SSQ (Speech, Spatial, and Qualities of hearing scale). Int J Audiol. 2014;53(2):101–14.
2.
Banh J, Singh G, Pichora-Fuller MK. Age affects responses on the Speech, Spatial, and Qualities of Hearing Scale (SSQ) by adults with minimal audiometric loss. J Am Acad Audiol. 2012;23(2):81–91; quiz 139-140.
3.
Dwyer NY, Firszt JB, Reeder RM. Effects of unilateral input and mode of hearing in the better ear: self-reported performance using the speech, spatial and qualities of hearing scale. Ear Hear. 2014;35(1):126–36.
4.
Gatehouse S, Akeroyd M. Two-eared listening in dynamic situations. Int J Audiol. 2006;45(Suppl 1):S120–4.
5.
Gatehouse S, Noble W. The Speech, Spatial and Qualities of Hearing Scale (SSQ). Int J Audiol. 2004;43(2):85–99.
6.
Heo JH, Lee JH, Lee WS. Bimodal benefits on objective and subjective outcomes for adult cochlear implant users. Korean J Audiol. 2013;17(2):65–73.
7.
Hey M, Wesarg T, Mewes A, Helbig S, Hornung J, Lenarz T, et al. Objective, audiological and quality of life measures with the CI532 slim modiolar electrode. Cochlear Implants Int. 2018;20(2):80–90.
8.
Kelsall DC, Arnold RJG, Lionnet L. Patient-reported outcomes from the United States clinical trial for a hybrid cochlear implant. Otol Neurotol. 2017;38(9):1251–61.
9.
Laske RD, Veraguth D, Dillier N, Binkert A, Holzmann D, Huber AM. Subjective and objective results after bilateral cochlear implantation in adults. Otol Neurotol. 2009;30(3):313–8.
10.
Lenarz T, Muller L, Czerniejewska-Wolska H, Valles Varela H, Orus Dotu C, Durko M, et al. Patient-related benefits for adults with cochlear implantation: a multicultural longitudinal observational study. Audiol Neurootol. 2017;22(2):61–73.
11.
Louza J, Hempel JM, Krause E, Berghaus A, Müller J, Braun T. Patient benefit from Cochlear implantation in single-sided deafness: a 1-year follow-up. Eur Arch Oto-Rhino-Laryngol. 2017;274(6):2405–9.
12.
Mertens G, Punte AK, Van de Heyning P. Self-assessment of hearing disabilities in cochlear implant users using the SSQ and the reduced SSQ5 version. Otol Neurotol. 2013;34(9):1622–9.
13.
Moulin A, Pauzie A, Richard C. Validation of a French translation of the Speech, Spatial, and Qualities of Hearing Scale (SSQ) and comparison with other language versions. Int J Audiol. 2015;54(12):889–98.
14.
Moulin A, Richard C. Sources of variability of Speech, Spatial, and Qualities of Hearing Scale (SSQ) scores in normal-hearing and hearing-impaired populations. Int J Audiol. 2016;55(2):101–9.
15.
Myhrum M, Moen I. The Norwegian hearing in noise test. Int J Audiol. 2008;47(6):377–8.
16.
Myhrum M, Tvete OE, Heldahl MG, Moen I, Soli SD. The Norwegian hearing in noise test for children. Ear Hear. 2016;37(1):80–92.
17.
Noble W, Gatehouse S. Interaural asymmetry of hearing loss, Speech, Spatial and Qualities of Hearing Scale (SSQ) disabilities, and handicap. Int J Audiol. 2004;43(2):100–14.
18.
Noble W, Tyler R, Dunn C, Bhullar N. Unilateral and bilateral cochlear implants and the implant-plus-hearing-aid profile: comparing self-assessed and measured abilities. Int J Audiol. 2008;47(8):505–14.
19.
Noble W, Tyler RS, Dunn CC, Bhullar N. Younger- and older-age adults with unilateral and bilateral cochlear implants: speech and spatial hearing self-ratings and performance. Otol Neurotol. 2009;30(7):921–9.
20.
Øygarden J. Norwegian speech audiometry (PhD Thesis). Norway: Norwegian University of Science and Technology; 2009.
21.
Ramakers GGJ, Smulders YE, van Zon A, Van Zanten GA, Grolman W, Stegeman I. Correlation between subjective and objective hearing tests after unilateral and bilateral cochlear implantation. BMC Ear Nose Throat Disord. 2017;17:10.
22.
Reeder RM, Firszt JB, Holden LK, Strube MJ. A longitudinal study in adults with sequential bilateral cochlear implants: time course for individual ear and bilateral performance. J Speech Lang Hear Res. 2014;57(3):1108–26.
23.
Singh G, Kathleen Pichora-Fuller M. Older adults’ performance on the Speech, Spatial, and Qualities of Hearing Scale (SSQ): test-retest reliability and a comparison of interview and self-administration methods. Int J Audiol. 2010;49(10):733–40.
24.
Smulders YE, van Zon A, Stegeman I, Rinia AB, Van Zanten GA, Stokroos RJ, et al. Comparison of bilateral and unilateral cochlear implantation in adults: a randomized clinical trial. JAMA Otolaryngol Head Neck Surg. 2016;142(3):249–56.
25.
Soli SD, Wong LLN. Assessment of speech intelligibility in noise with the hearing in noise test. Int J Audiol. 2008;47(6):356–61.
26.
Quentin Summerfield A, Barton GR, Toner J, McAnallen C, Proops D, Harries C, et al. Self-reported benefits from successive bilateral cochlear implantation in post-lingually deafened adults: randomised controlled trial. Int J Audiol. 2006;45(Suppl 1):S99–07.
27.
Zhang J, Tyler R, Ji H, Dunn C, Wang N, Hansen M, et al. Speech, Spatial and Qualities of Hearing Scale (SSQ) and Spatial Hearing Questionnaire (SHQ) changes over time in adults with simultaneous cochlear implants. Am J Audiol. 2015;24(3):384–97.