eFigure 1. MEDLINE Search Strategy
eFigure 2. Study Attrition Through Screening and Eligibility Assessment
eTable. Supporting Quotations for Sub-domains and Items Emphasized in Patient Semi-structured Interviews
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Dixon PR, Feeny D, Tomlinson G, Cushing S, Chen JM, Krahn MD. Health-Related Quality of Life Changes Associated With Hearing Loss. JAMA Otolaryngol Head Neck Surg. 2020;146(7):630–638. doi:10.1001/jamaoto.2020.0674
What changes in health-related quality of life are associated with hearing loss?
In this qualitative study, a systematic literature review of 45 studies including 1036 patients with hearing loss, an expert focus group of 4 audiologists and 4 otologists, and semistructured interviews with 26 individuals with hearing loss were used to identify a comprehensive list of the 29 subdomains of quality of life associated with hearing loss.
Understanding the multifaceted associations of hearing loss with health-related quality of life is essential to studying hearing loss conditions and treatments; the framework presented here is useful for generation or evaluation of hearing-related quality of life measures.
Utility is a single-value, preference-based measure of health-related quality of life that represents the desirability of a health state relative to being dead or in perfect health. Clinical, funding, and policy decisions rely on measured changes in utility. The benefit of hearing loss treatments may be underestimated because existing utility measures fail to capture important changes in quality of life associated with hearing loss.
To develop a comprehensive profile of items that describe how quality of life is associated with hearing loss and its treatments that can be used to generate hearing-related quality of life measures, including a novel utility measure.
Design, Setting, and Participants
This qualitative study, performed from August 1, 2018, to August 1, 2019, in tertiary referral centers, comprised a systematic literature review, focus groups, and semistructured interviews. The systematic review evaluated studies published from 1982 to August 1, 2018. Focus groups included 8 clinical experts experienced in the measurement, diagnosis, treatment, and rehabilitation of hearing loss. Semistructured interviews included 26 adults with hearing loss recruited from an institutional data set and outpatient hearing aid and otology clinics using stratified convenience sampling to include individuals of diverse ages, urban and rural residency, causes of hearing loss, severity of hearing loss, and treatment experience.
Main Outcomes and Measures
A set of items and subdomains that collectively describe the association of hearing loss with health-related quality of life.
The literature search yielded 2779 articles from the MEDLINE, Embase, Cochrane, PsycINFO, and CINAHL databases. Forty-five studies including 1036 individuals (age range, 18-84 years) were included. The focus group included 4 audiologists and 4 otologists. Hour-long semistructured interviews were conducted with 26 individuals (13 women; median age, 54 years; range, 25-83 years) with a broad range of hearing loss causes, configurations, and severities. From all 3 sources, a total of 125 items were generated and organized into 29 subdomains derived from the World Health Organization’s International Classification of Functioning, Disability and Health.
Conclusions and Relevance
The association of hearing loss with quality of life is multidimensional and includes subdomains that are not considered in the estimation of health utility by existing utility measures. The presented comprehensive profile of items can be used to generate or evaluate measures of hearing-related quality of life, including utility measures.
Hearing loss is the fourth-leading cause of disability globally and is associated with social isolation, depression, cognitive decline, and other negative health outcomes.1-4 Measures of health-related quality of life (HRQoL) quantify the ultimate outcome associated with hearing loss and its associated deficits and are therefore important measures of the effectiveness of hearing loss interventions.
Health utility is a single-number summary of HRQoL that represents the desirability of a health state relative to being dead (assigned a utility of 0) or in perfect health (assigned a utility of 1). Utility is used in clinical studies and population health surveys as a global measure of health status. It is also used in cost-effectiveness analyses as a HRQoL weight in the calculation of quality-adjusted life expectancy.5-8 Funding and policy decisions are increasingly reliant on measured changes in utility. The risk in the assessment of hearing loss interventions is that, if available utility measures fail to capture important improvements, beneficial health care resources may be withheld inappropriately.
The validity of utility estimates depends on the content of the questionnaire—if relevant questions about health status are not asked, relevant changes may not be captured. The Health Utilities Index Mark 3 is commonly used to study HRQoL in hearing loss in part because its hearing content is relatively comprehensive; others have more limited conceptualizations or do not include hearing at all.9-11 The Health Utilities Index Mark 3 summarizes the respondents’ ability to understand speech in one-on-one and group settings and their reliance on hearing aids.12 Hearing loss, however, occurs on a spectrum of severities across a range of frequencies and 2 independent ears. Signals are integrated to provide more than detection of sound but an entire perception of an individuals’ position in auditory space. A simplified conceptualization of hearing may fail to capture impairments that are important to individuals with the hearing loss. For example, sound localization and listening effort are impaired in monoaural hearing and are not described in the Health Utilities Index Mark 3 or any other utility instrument. Available utility estimates may therefore underestimate the benefit of interventions designed to achieve a semblance of binaural hearing. Bilateral cochlear implantation is one example, where measured utility change compared with unilateral implantation is minimal, ranging from 0.02 to −0.03.7,13-15 This finding contradicts established performance benefits such as speech recognition and sound localization,13,16 as well as HRQoL improvements measured with other instruments.16-19 A new utility measure that comprehensively characterizes the abilities and disabilities of patients across the spectrum of hearing loss is critical for accurately characterizing the effectiveness of modern hearing loss interventions.
Scientific standards for the development of typical HRQoL instruments are well established,20 and most principles are also applicable to the development of condition-specific utility instruments.10-12 The multiphase process must begin with a valid and comprehensive conceptualization of the target construct. The objective of the present study was to identify independent subdomains of hearing that, when considered collectively, describe how hearing loss and its treatment are associated with HRQoL. This comprehensive profile could be used to generate or evaluate any measure of hearing-related quality of life, including a new hearing-specific utility instrument.
This qualitative study was performed from August 1, 2018, to August 1, 2019. Specific examples of hearing-related function or disability that are associated with changes in HRQoL are referred to as “items.” Multiple sources were consulted for item generation, including a systematic literature review, expert focus group, and semistructured interviews of patients. Written informed consent was obtained from each participant in the focus group and all patients and the study was approved by the Sunnybrook Health Sciences Centre Research Ethics Board. The systematic review is reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline.21
We sought studies that described changes in HRQoL associated with hearing loss. A systematic review published in 2017 reports items relevant to adults with hearing loss and their communication partners.22 Items from that review were extracted to serve as a preliminary item list. Given the underrepresentation of patients with cochlear implants (CIs) in that review, a separate systematic review was conducted to summarize qualitative studies exploring the HRQoL of patients with CIs.
Eligible studies explored the HRQoL of patients with CIs with open-ended or closed-ended questionnaires, interviews, or focus groups. Experimental, observational, and cross-sectional studies along with case reports and autobibliographic reports were eligible. Professional commentaries and web-based forums were excluded.
The target population for the systematic review was adults (≥18 years of age) with hearing loss rehabilitated with a unilateral CI or bilateral CIs. No restriction on the comparison groups was applied.
Eligible studies were published in or after 1982, the year of publication of one of the earliest hearing loss patient-reported outcome measures, the Hearing Handicap Inventory for the Elderly.23 This time frame also approximates the market introduction of the multichannel CI.24 Studies published up to August 1, 2018, were eligible.
Databases searched included MEDLINE, Embase Classic+Embase, Cochrane, PsycINFO, and CINAHL. Conference abstracts were included.
Generated in consultation with a medical librarian, the search consisted of subject headings and text word terms for cochlear implants AND quality of life AND qualitative research (eFigure 1 in the Supplement). Validated search strategies for qualitative studies were applied.25,26 Reference lists of included articles supplemented the search.
Abstract and title screening and full-text review were completed in duplicate. Discrepancies were resolved by a third investigator.
A piloted data collection form was used to extract study year of publication, country of origin, design, qualitative data collection method, and sample size. Participant mean age, proportion female, severity and proportion of bilateral or unilateral hearing impairment, and intervention was extracted for each study. Data were managed with DistillerSR (Evidence Partners).
Item extraction was performed independently by 2 investigators (including P.R.D.). For studies whose qualitative component was a predicate step to generation of an HRQoL instrument, questions included in the preliminary instrument were extracted. The aim of capturing all reported items and subdomains obviated the relevance of a risk of bias assessment for articles in this review.
A group of 8 experts (4 audiologists and 4 otologists) was convened after the systematic review was complete to generate additional items. Broad inclusion of individuals with varying expertise in the medical and surgical treatments and rehabilitation of patients with hearing loss was prioritized. Discussion was led first by open-ended questions about factors that are associated with the quality of life of patients with hearing loss. Once novel items were no longer generated spontaneously, experts were guided through the World Health Organization’s International Classification of Functioning, Disability and Health (ICF) brief core set for hearing loss and then the master item list generated by the systematic review to generate additional discussion.
A total of 26 patients were consulted in individual semistructured interviews for further item generation. Adult (≥18 years of age) English-speaking participants with hearing loss were recruited from those seen in audiology and otolaryngology clinics for various levels of hearing loss, those evaluated for hearing aid fitting, and those evaluated for CI surgery. Stratified convenience sampling was applied to recruit participants with a broad distribution of ages, urban and rural residency, causes of hearing loss, severity of hearing loss, and treatment experience. Block recruitment ensured that at least 1 participant from each of 6 defined hearing categories (mild to moderate hearing loss, single-sided deafness, candidate for CI on both sides, CI with contralateral normal hearing, CI with contralateral hearing loss, and bilateral CIs) was included in each block. Semistructured interviews continued until saturation, defined as no emergence of new items in a participant block.
The interview began with open-ended questions including: “In your personal experience, how has hearing loss impacted your quality of life?” Participants were encouraged to reflect on the past several weeks and then on years prior and other stages of their hearing loss and other treatments they had experienced. When no novel items were spontaneously offered, the master list including items generated from the systematic review and expert focus group was presented to stimulate further discussion.
Items extracted from studies included in the systematic review were initially organized according the World Health Organization’s ICF core sets for hearing loss. Core sets are standardized sets of categories designed to systematically describe health.27-29 New categories were generated for items or subdomains that were not suitable for inclusion in existing hearing loss core set categories. The resulting item list and subdomain framework served as the preliminary master item list.
Items generated by spontaneous and guided discussion in the focus group or semistructured interviews were added to the master item list in real time. The language used to describe items and their categorization were reviewed with and approved by the participant(s). If any participant thought that an item reflected a concept unique from existing subdomains, a new subdomain was generated. Transcripts were reviewed by 2 study investigators (including P.R.D.), organized thematically, and compared with the master item list to ensure that all discussed items were captured. New items and subdomains were added if either investigator considered an identified theme to be unique from those already listed. This approach is adapted from previously reported conceptualizations of condition-specific HRQoL.30,31
The search yielded 2779 unique publications. Full-text review confirmed eligibility of 45 studies (eFigure 2 in the Supplement).32-76 There was broad geographic representation of studies from predominantly developed countries (Table 1). Across all studies, 1036 participants (age range, 18-84 years) were included. Semistructured interviews were the most common data collection method (18 of 45 [40%]) followed by open-ended questionnaires (14 of 45 [31%]).
Focus group participants included 4 audiologists with experience with various methods of hearing rehabilitation including CI and hearing aid fitting, and 4 otolaryngologists with subspecialty interest in otology and neurotology.
Two blocks of participants were recruited, including 26 individuals with hearing loss (Table 2). Saturation of information was achieved with this sample. A broad range of ages (median, 54 years; range, 25-83 years) and equal proportion of sexes (13 women) were included. Sudden idiopathic hearing loss was the most common cause (9 of 26 [35%]).
All 56 items reported in the identified previously published systematic review were extracted.22 Extraction of data from the 45 studies included in our systematic review yielded 111 items. Of those, 56 were represented by items identified in the previous review and 55 were novel. These 111 items were organized according to 23 of the 27 categories in the ICF brief core set for hearing loss as well as 5 additional categories added from the ICF comprehensive core set for hearing loss and 1 additional category from the ICF framework that is not included in either hearing loss core set (Table 3).
The expert focus group discussions generated 14 additional items (Table 3). There was unanimous agreement from focus groups participants that all items discussed had been added to the master item list. The experts acknowledged that additional real-life examples pertaining to a particular item could be conceptualized, but that no further items representing unique concepts could be suggested.
No unique items were generated by semistructured patient interviews. At the conclusion of each interview after structured review of the master item list with the interview participant, all participants agreed that all discussed topics were captured by the existing list. Review of themes abstracted from interview transcripts supported that conclusion.
Several subdomains and items were emphasized as being important to quality of life by interview participants. Additional supporting quotations are provided in the eTable in the Supplement.
Impaired ability to recognize and understand speech was described as being context dependent, varying on the characteristics of the speaker and the environment. Characteristics of the speaker included voice pitch and accent, as well as the presence of multiple talkers: “Women’s voices seems to be a little bit easier to pick up for me” (male, aged 50-59 years, bilateral CIs). The presence of background noise increased the challenge of understanding speech and sometimes led to social avoidance.
An inability or reduced ability to understand what was said over a telephone was particularly notable in work environments, where the ability to use alternative methods of communication were more limited: “I’m a pharmacist. I couldn’t take phone calls. I couldn’t speak to doctors on the phone. I really couldn’t interact with staff” (male, aged 50-59 years, bilateral CIs).
Hearing thresholds were typically discussed in the context of ability to hear and recognize certain environmental sounds. Participants emphasized the safety implications of missing certain warning sounds: “[Before my CI] I wouldn’t know if somebody was at the door. Or, the phone was ringing. Or, the fire alarm was going off…you have no clue” (male, aged 50-59 years, bilateral CIs).
An inability to identify the location of sounds in one’s surrounding environment was a recurring theme, particularly for patients with a unilateral CI or single-sided deafness. The safety implications of misinterpreting the direction of origin of some sounds was emphasized: “An ambulance…it’s not hearing it, it’s pinpointing where it is and where it’s coming from” (male, aged 50-59 years, single-sided deafness).
Participants described missing sounds on their poor-hearing side and the nuisance of having to position themselves strategically to ensure their better-hearing side was directed toward an important signal. Participants also described the troubling sensation of having an incomplete sound field as well as the relative differences in sound quality associated with hearing with 2 ears instead of 1: “…when you have the hearing on both sides…it’s, it’s louder. It’s clearer” (male, aged 50-59 years, bilateral CIs).
The reported association of tinnitus with quality of life was variable. Some participants described an intrusive and annoying sound that was associated with profoundly reduced quality of life, others reported tinnitus but were not bothered by it, and others had not experienced tinnitus. Perception was noted to be dependent on environment and stress: “Tinnitus is always, always, always there…I’m living with the damn noise [of] my tinnitus” (female, aged 50-59 years, single-sided deafness with CI). “A few occasions I hear whooshing, something like that and it stops and I forgot all about it” (male, aged 80-89 years, hearing aids).
The effort and focus required to hear had a negative association with some respondents’ quality of life, predominantly mediated through reduced energy level: “If I had a long day—I would come home and I would be exhausted… Because you’re spending a lot of time focusing” (male, aged 50-59 years, bilateral CIs).
Many participants described a reduced enjoyment of music associated with their hearing impairment, particularly among participants with CI(s): “I would try to put on a CD [compact disc] and oh, it just—it sounds like garbage” (male, aged 50-59 years, unilateral CI).
Some participants described a reduced enjoyment of music because it distracted from one’s ability to understand speech: “I like to drive with music, I can’t anymore…if there’s music, there’s no conversation. If there’s conversation there’s no music” (male, aged 20-29 years, single-sided deafness).
The negative HRQoL implications of assistive hearing devices reported included stigma, comfort, and breakdown: “I have to carry around like the case with me all the time and then I always worry about the battery…I don’t like being reliant on a contraption” (female, aged 30-39 years, hearing aids).
Participants described a reduced willingness to attend social events owing to poor speech recognition and a negative association with existing relationships owing to communication partners’ frustration with making accommodations (eg, frequent need for repetition). Developing new social relationships was described as challenging owing to limited ability to interpret verbal social cues, which led to misunderstandings: “I don’t hear enough of a person to know what they’re like…there’s no interaction between people if you can’t hear them, so there’s no affinities going on, no likes or dislikes” (female, aged 60-69 years, bilateral CIs).
Hearing loss was also described to have an association with educational and vocational achievement: “…there was a major change in my grades, like I would get C’s until my hearing got better, then I would get B’s and A’s” (female, aged 30-39 years, hearing aids).
Some participants described poor speech understanding as a cause of embarrassment, which was associated with an unwillingness to ask for accommodations at school or work: “…in grade 9 I had another bit of a hearing loss and you know how teenagers are, they start getting self-conscious and all that…so I’d just get bad grades…” (female, aged 60-69 years, bilateral CIs).
Work performance and advancement had a negative association with hearing loss through perceived discrimination and avoidance of work that required social engagement.
A data-driven approach to conceptualizing the association between hearing loss and HRQoL was taken whereby examples of how hearing loss is associated with HRQoL were systematically collected from a literature review, clinical experts, and patients. The result is a comprehensive profile of subdomains and items that can be used for the generation and evaluation of tools for measuring HRQoL of individuals with hearing loss.
Our approach differs from the top-down approach applied in the generation of some existing hearing-related quality-of-life instruments. Both the Nijmegen Cochlear Implant Questionnaire and the Speech, Spatial and Qualities of Hearing Scale, for example, are based on a subdomain framework generated solely by expert opinion; items were generated to fit within those frameworks.77,78 Reliance on a single source for subdomain generation introduces the potential for a narrowed conceptualization of the target construct. A comparatively rigorous conceptualization of hearing served as the basis for the recent generation of the Cochlear Implant Quality of Life instruments.34,79,80 Communication, emotion, environmental sounds, independence and work function, listening effort, social isolation and ability to socialize, and sound clarity were themes that McRacken and colleagues34 identified as being relevant to the HRQoL of patients with CIs. All these themes were also identified in our study to be of importance to the broader population of individuals with hearing loss. We also identified additional items—including tinnitus and reliance on products and technology for communication. Observed differences may be explained by differing HRQoL priorities of patients with CIs compared with those of the broader population with hearing loss. Greater diversity of sources for item and subdomain generation applied in our study may also have accounted for identification of additional items and subdomains in our study. For example, expert opinion was systematically included in the form of a focus group. Patient input was captured with individual semistructured interviews rather than focus groups, which may have more reliably captured the opinions of individuals with hearing impairments that challenge participation in group conversation.
Part of the motivation for our work was to generate a comprehensive pool of subdomains and items that could be used to generate a new utility instrument. Design of any HRQoL measure requires an understanding of how HRQoL can be associated with the condition under study and can be conceptualized with a list of subdomains and items. To balance accurate estimation of HRQoL and response burden of the new instrument, the list is systematically shortened by eliminating items that are least important or redundant. Parsimony is particularly important in utility instruments, where the scoring function requires the utility associated with all possible sets of responses (ie, health states) to be empirically estimated or modeled. The number of possible health states can quickly become unmanageable for this labor-intensive exercise: an instrument with 10 questions each with 5 response options has more than 9.7 million response combinations. Consulting patients with hearing loss about the importance of each item and subdomain to their HRQoL is one approach to identifying which of these items and subdomains should be included in a novel hearing utility instrument.30,31 No matter how meticulous the item reduction method, mistakenly omitting an important item from the initial conceptualization guarantees that it will not be included in the final instrument.
This study has some limitations. A challenge of holistically describing the association of hearing loss with HRQoL is that the characteristics of the individuals and their hearing loss modify that association. There are numerous causes of hearing loss and a nearly unlimited spectrum of hearing loss severity, some of which may be associated with unique HRQoL implications. A limited sample of demographics and causes were represented by semistructured interview participants. Setting is also associated with participant selection. Patients and experts were recruited from a tertiary referral center offering a spectrum of treatment including a CI program, implying access to comprehensive care. To mitigate potential selection bias, we purposefully sampled experts who treat all types and severities of hearing loss and patients with varied demographics and hearing impairments at various stages of treatment. Broad international representation in the literature search may also have compensated for bias associated with setting. Sequential consultation of multiple independent sources further reduced risk of selection bias and also served to check the comprehensiveness of previous steps. That no novel items or subdomains were identified in the semistructured interviews supports the robustness of the systematic review and focus group. Nonetheless, the potential for additional items or subdomains to have been identified by broader inclusion of studies in the systematic review or additional focus group or interview participants is acknowledged.
We present a comprehensive set of items and subdomains that describe how hearing loss is associated with HRQoL. This conceptualization can be used to evaluate the appropriateness and comprehensiveness of existing measures of hearing-associated HRQoL. It can also be used to generate new measures of HRQoL, including a utility instrument, specific to populations with hearing loss. Such an instrument would be the logical choice for clinical, population health, and cost-effectiveness studies evaluating HRQoL changes associated with hearing loss.
Accepted for Publication: March 17, 2020.
Corresponding Author: Peter R. Dixon, MD, MSc, Department of Otolaryngology–Head & Neck Surgery, University of Toronto, Six Queen’s Park Crescent W, Ste 120, Toronto, ON M5S 3H2, Canada (firstname.lastname@example.org).
Published Online: May 14, 2020. doi:10.1001/jamaoto.2020.0674
Author Contributions: Dr Dixon had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Dixon, Feeny, Chen, Krahn.
Acquisition, analysis, or interpretation of data: Dixon, Feeny, Tomlinson, Cushing, Krahn.
Drafting of the manuscript: Dixon, Krahn.
Critical revision of the manuscript for important intellectual content: Dixon, Feeny, Tomlinson, Cushing, Chen.
Statistical analysis: Dixon, Tomlinson.
Obtained funding: Dixon, Chen.
Administrative, technical, or material support: Dixon, Krahn.
Supervision: Feeny, Cushing, Chen, Krahn.
Conflict of Interest Disclosures: Dr Feeny reported a patent to Health Utilities Inc issued, licensed, and with royalties paid. Dr Cushing reported serving on the speakers bureau for Cochlear Corporation during the conduct of the study; serving on the speakers bureau for Interacoustics and Plural Publishing outside the submitted work; and having a patent to 7041-0 issued. No other disclosures were reported.
Funding/Support: This work was supported by the Harry Barberian Scholarship.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: Jenna Wolno, BSc, Department of Otolaryngology–Head & Neck Surgery and Cochlear Implant Program, Sunnybrook Health Sciences Centre, contributed to acquisition of data and administrative support for this work. She was employed as a paid research assistant. Joanna Bielecki, BSc, MISt, contributed her expertise as a research librarian to generation of the systematic review search strategy. She was compensated in her role as a paid research librarian.