Background

Hearing loss can have a negative impact on an adults’ health and social well-being. Globally, in 2019, an estimated 1.57 billion people had some form of hearing loss, and this figure is estimated to increase to 2.45 billion by the year 2050 [1]. Hearing loss can lead to more frequent use of inpatient or outpatient healthcare services [2, 3], increased fall risk in healthcare facilities [4] and other locations [5], poor communication with providers when using healthcare services [6] that can affect a health consumer’s satisfaction [7] with healthcare delivery and utilisation [8]. Hearing loss is also associated with negative social outcomes, such as reduced academic performance, lower chance of progressing to higher education or undertaking training [9,10,11], unemployment [10, 11], poor personal relationships [12], feelings of inadequacy and low self-esteem [12, 13], social isolation and loneliness [14], and higher rates of depression and low QoL [15,16,17]. Integrated approaches that tackle barriers and social outcomes could mitigate some of these effects for both CI users and their health providers [18]. For example, strategies such as noise reduction approaches, and acting on hearing healthcare, are all likely to ameliorate some of the negative aspects of hearing loss [1, 3, 19].

Cochlear implantation is the surgical insertion of an electrode within the inner most part of the ear to transmit sounds from an externally worn device. Cochlear implants are suitable for individuals with a severe to profound sensorineural hearing loss, who do not benefit from standard hearing aids [20, 21]. A cochlear implanted device can improve a person’s ability to understand speech through providing improved access to speech sounds [22,23,24]. Compared to their preoperative hearing with the use of hearing aids, a majority of CI users demonstrate improvement in speech recognition [25, 26]; however, the magnitude of improvement varies considerably across individual CI users [27,28,29,30]. Factors contributing to this variation are unclear, but different studies have suggested that this could be influenced by: age at implantation, age at onset of hearing loss, duration of implant use (up to 1–2 years), duration and cause of hearing loss, placement of implant in the cochlea, integrity of the cochlear nerve, as well as learning ability of the individual living with hearing loss [27, 31, 32].

Several studies have measured hearing outcomes in CI users [25, 33,34,35]. However, the potential impact of CI on health service utilisation and social outcomes are less understood. To our knowledge, no review has been found that comprehensively synthesized impacts of CI on health service utilisation such as hospital admissions, ED presentations, GP visits and prescription medication use. In addition, there have been few scoping reviews that have examined social outcomes such as work, autonomy and participation post-CI [36]. This review offers some important insights into improve service delivery, health service use and social outcome trajectories of the CI users. Therefore, this systematic review aimed to synthesise the evidence of the impact of CI on health service utilisation and social outcomes in adult CI users.

Methods

Eligibility criteria

This systematic review adhered to the Preferred Reporting Items for Systematic Review and Meta-analyses (PRISMA) statement [37] and the protocol was registered with PROSPERO (CRD42023392131). This review included studies reporting on health service utilisation or social outcomes of adults aged ≥ 18 years with considerable hearing loss who received a CI. The comparison was performed either between CI users and non-users with hearing loss but not implanted or within CI users, based on their pre- and post-operative outcomes. Hearing loss was defined according to the World Health Organization’s (WHO) definition into moderate: 41 to 60 dB, severe: 61 to 80 dB, profound or complete deafness: ≥ 81 dB [38, 39]. CI users could be fitted either unilaterally, bilaterally, bimodally (using a hearing aid on the other ear) or by electric-acoustic stimulation (only part of the cochlea is stimulated with a cochlear implant). Articles were excluded if they solely reported on children or individuals with prelingual hearing loss and did not separately report results for adults. Articles were excluded if they were reviews, editorials or opinion pieces, single case report, study protocols, or conference abstracts. This review included English-language articles that were published in a peer-review journal.

The main outcomes of interest are health service utilisation (e.g., hospital admissions, ED presentations, GP visits, CI revision surgery and pharmaceutical use); and social outcomes (e.g., circumstances relating to education, autonomy, social participation, training, disability, social housing, social welfare benefits, occupation, employment, income level, anxiety, depression, QoL, communication abilities and cognition).

Health service utilisation or social outcomes could either be in the short (e.g., < 6 months), medium (e.g., 6–12 months), or long-term (e.g., > 12 months). Health service utilisation included hospital admissions, ED presentations, GP visits, and prescription medication use. Social outcomes included information or circumstances relating to education, autonomy, social participation, training, disability, social welfare (e.g., social housing, welfare benefits), occupation, employment, or income level. Autonomy or independence was defined as the capability of living the way an individual wished to without being reliant on a third person to control, cope and make personal decisions on their life [36, 40]. Whereas the capability of participating in different social situations or activities without limitation due to hearing loss was defined as social participation [36, 41].

Study selection process

Five databases were searched, including MEDLINE and PsycINFO using the Ovid portal, Scopus, CINAHL using the EBSCOhost portal, and ERIC using the ProQuest portal. The search was conducted from 1 January 2000 to 16 January 2023 for the four common databases. On the recommendation of an educational expert, we added a search of the ERIC database in May 2023. The search strategy was developed in consultation with a university librarian. The full search strategy is provided (see Additional file 1). Snowballing of reference lists from the articles was conducted to identify any potential articles not previously identified. Title and abstract screening involved importing the title, abstract and citation information for each article identified during the database searching into EndNote X20. Duplicates were removed using EndNote. Following duplicate removal, the titles and abstracts of identified articles were screened and assessed for inclusion. Abstracts were excluded if they did not report on health service utilisation or social outcomes post-CI in adults. Uncertainties regarding the inclusion or exclusion of articles based on title and abstract were discussed and consensus was obtained. Full-text screening was done by assessing each article against inclusion criteria. Following the full-text screening, a data extraction form was created and tested on five studies.

Data extraction

For studies that met the inclusion criteria, key characteristics of each study were extracted, including: authors and publication year; study objective or aim; study type; country/study setting and data collection timeframe; study population (e.g., mean age, sex, and sample size); information on health service utilisation and social outcomes. Data extraction was initially performed by one reviewer and subsequently verified for accuracy by two reviewers and any disagreements were discussed between reviewers and consensus was obtained.

Data synthesis

Information extracted from the included studies was narratively synthesised by one reviewer and appraised by two reviewers. The narrative synthesis involved tabulating and summarising health service utilisation and social outcomes.

Quality assessment

The methodological quality of articles was assessed using the Critical Appraisal Skills Programme (CASP) cohort [42] or qualitative [43] study checklists as applicable. Quality assessment was initially performed by one reviewer and independently verified by two reviewers. Any disagreement regarding methodological quality were discussed between reviewers. The checklist consists of 12 questions for cohort, 10 questions for each cross-sectional and qualitative studies. Responses were recorded for each question based on the level of adequate information provided at design and analysis stages in the study. For example, “Was the outcome accurately measured to minimise bias?” ‘Yes’ was recorded for adequate information by looking for measurement or classification bias such as the use of subjective or objective measurements, do the measurement accurately measured what they intend to measure (validated vs. unvalidated tools), reliable method for detecting all cases, similarity of measurement methods in different groups and blinding of outcome assessor or subjects to exposure. Whereas ‘No’ was recorded for missing of one or more information and ‘Can’t tell’ was recorded if the information sought is not applicable. The overall quality of studies was judged based on ‘Yes’ or ‘No’ response to questions regarding relevance, reliability, validity and applicability.

Results

Description of studies

The search identified 2093 articles through searching of electronic databases and snowballing search methods. There were 2058 articles identified via the five databases, 541 were removed as duplicates. 1517 articles underwent title and abstract screening. Based on titles and abstracts, 1424 articles were excluded due to not meeting the inclusion criteria. The full text of the remaining 93 articles were assessed for eligibility. Following full text review, 49 articles were excluded due to an irrelevant target population or absence of the primary outcomes. The remaining 44 articles were intensively appraised and met the inclusion criteria (Fig. 1).

Fig. 1
figure 1

PRISMA flow diagram

Full size image

There were 20 (45.5%) cohort studies (11 prospective and 9 retrospective), 18 (40.9%) cross-sectional studies and six (13.6%) qualitative studies. Almost all (93.2%, n = 41) of the included studies were conducted in high-income countries. Nine (20.5%) studies were conducted in United States of America (USA), six (13.6%) conducted in New Zealand, four (9.1%) conducted in Canada, three (6.8%) studies each were conducted in Germany and Norway, two (4.5%) studies each were conducted in Australia, Belgium, China, Finland, France, Norway and Poland and one study each was conducted in Brazil, Denmark, Netherlands, Saudi Arabia, Spain, Scotland, Sweden and the United Kingdom (UK). The sample size ranged from 125 to 5,130 participants for health service utilisation and 6 to 637 participants for social outcomes.

All studies included adults with severe to profound post-lingual hearing loss. Three studies [44,45,46] also included a few participants with prelingual hearing loss but were predominantly focussed on post-lingual hearing loss. Participants implanted with unilateral, bilateral, or bimodal CI were assessed within all studies. The mean duration of hearing loss was not often recorded in the included studies. Where reported (n = 11), the mean duration ranged from 2.5 to 32.7 years. The mean duration of follow-up from the date of CI ranged from 3 months to 23 years. The age of participants ranged from 18 to 101 years at the time of the study. However, one study had one CI user aged 17 years and thus was included in the sample [47]. The mean age of participants was not recorded in nine studies (20.5%), but where recorded, ranged from 21.9 to 80.9 years. Studies assessed outcomes by comparing CI users and non-users with hearing loss but not implanted (18.2%, n = 8) or within CI users, based on their pre- and post-operative outcomes (81.8%, n = 36).

Health service utilisation

Nine studies (20.5%) reported on health service utilisation (Table 1). Type of health service utilisation examined in each study varied and included readmission for CI revision (n = 5), post-operative surgical and audiological visits (n = 3), pneumococcal vaccination uptake (n = 1), extended hospital length of stay (n = 1), non-home destinations post-CI surgery (n = 1), and medication use (n = 1) (Table 2). The rate of CI revision surgery (with or without reimplantation) ranged from 1.1 to 3.8%. The average time from first implant to revision surgery ranged from 29 months to 7.8 years. Almost all studies indicated an ongoing need for CI revision surgery by users who had experienced complications post-CI. Where complications were reported, soft or hard device failure (n = 5) and flap skin infection, surgical or falls (n = 4) were the main reasons for the revision surgery. The mean number of health service visits ranged from 1.9 to 4.3 times a year. For example, one study [48] compared the health service visits of older (≥ 80 years) and younger adults (aged 60–79 years) within the CI group and found no difference in number of visits between the age groups, with the average number of visits decreasing for both age groups in the second year of post-operative follow-up. Pneumococcal vaccination uptake increased in CI users after a follow-up reminder in another study [49]. The study that examined medication use, identified that prescribed medication was used on average for 1.8 illnesses among CI users and for 3.1 illnesses in a non-user group [50].

Table 1 Characteristics of included studies
Full size table
Table 2 Characteristics of included studies that examined health service utilisation of CI users
Full size table

Social outcomes

Thirty-five (79.5%) studies reported on social outcomes, 14 (40.0%) were cohort, 15 (42.9%) were cross-sectional and six (17.1%) were qualitative studies. The majority (71.4%, n = 25) of these studies reported social outcomes as a primary objective of the study (Table 3). Of the 35 studies, 21 (60.0%) reported on work or employment, 15 (42.9%) on social participation, nine (25.7%) on autonomy or independence, eight (22.9%) on education, six (17.1%) on perceived hearing disability, three (8.6%) on income, and two (5.7%) on safety and welfare.

Table 3 Characteristics of included studies that examined social outcomes of CI users
Full size table

Of the 21 studies that reported on work or employment, the majority (85.7%, n = 18) of studies found positive improvements on work or employment status post-CI. However, three studies (14.3%) did not find an additional benefit of CI on work or employment status. For example, one study conducted by Ross and Lyon [45], found CI users experienced difficulties in their workplace as employer expectations of hearing improvement post-CI were unrealistic.

Of the 15 studies that reported on social participation, improvement in social participation was observed in 14 (93.3%) studies post-CI. Only one study found no additional benefits in social participation post-CI [51]. The Chapman et al. 2017 [51] found that the CI cohort showed increased feelings of being limited due to their hearing loss, and that they participated less in mainstream organisational activities than the non-CI group. However, after categorising age into ≤ 25 and > 25 years, the authors found no statistical difference by age in participation in mainstream organisational activities. Also, the CI users in the older age group were found to socialise more with their hearing friends compared with non-users.

Independence or autonomy was reported in nine studies. Eight (88.9%) studies found substantial improvement in independence or autonomy following a CI. Only one study found no improvement in independence or autonomy measured by a subscale of quality of life [52]. Of the studies that found improvements in autonomy post-CI, Sonnet et al. 2017 [53], found improvement at 12-months post-CI. While another study [54] found a significant difference between pre- and post-implantation on autonomy at 6 months post-implantation, but this result did not persist at the 12-month follow-up.

Education or training was measured in nine studies. All studies that measured education or training found benefits post-CI. For example, Goh et al. 2016, found that 76% of CI users reported that the CI enabled them to access learning opportunities and gain tertiary qualifications [44]. Another study found that CI helped adults in retaining and developing professional abilities, for example two CI users were promoted to management positions and two moved to jobs requiring a higher level of skills [55].

Six studies measured perceived hearing disability. Five of the six studies found some improvements among respondents post-CI. For instance, one qualitative study [56] indicated participants perceived that the CI provided them a ‘new life’. CI was found to be associated with type of identity, such as deaf or hearing identity, type and quality of friendships, social activities, and feelings of limitation previously attributed to hearing loss [51].

Three studies reported on income and two on safety and welfare benefits. The three studies that reported on income, all found increases in the recipient’s income level post-CI. For example, one study [57] found a 31% increase in income bracket after a mean 6.6 years post-CI. Montero and colleagues [58] also found a significant increase in median annual income post-CI compared with preimplantation (CAD $42,672 vs. CAD $30,432). However, none of the identified studies found a positive change in welfare benefits. For instance, Mo et al. 2004 found no significant difference between CI and non-CI users in terms of safety, and welfare [59]. Again, the authors found that after 12 and 15 months post-CI, welfare and safety were not significantly improved [60].

Communication, anxiety, depression, quality of life and cognition

Of the 35 studies that reported on societal outcomes, ability to communicate was reported in 27 (77.1%), QoL in 15 (42.9%), and six (17.1%) studies each reported anxiety or depression and cognition. Of the studies that reported on communication abilities and QoL, all found improvement post-CI. Four of the six studies that reported on anxiety or depression found a reduced level of anxiety or depression post-CI. For example, Mo et al. 2005 [60] found the reduced mean score of anxiety and depression between pre- and post-implantation (-0.10 vs. -0.19). While two of the six studies found no change for anxiety or depression post-CI [53, 61]. There were mixed results on the effects of CI on anxiety, depression and cognitive function over time. For instance, Claes et al. 2018 found a decreased level of anxiety and depression at six months post-CI, however, the decrease was not sustained at a 12 month follow-up [62]. In contrast, another study found a decreased level of anxiety and depression after 12 months and 15 months post-CI [60] that was associated with gain in QoL. On the other hand, in the remaining study at 6 and 12 months post-CI, anxiety, depression levels and cognitive function remained stable [53].

Quality assessment

The overall quality of included studies was deemed low. None of the included studies had a low risk of bias due to inadequate methods to minimise the effects of confounding factors at design and analysis stage as defined by the CASP criteria. Few studies (36.4%, n = 16) scored ‘Yes’ for questions related to minimising the effects of confounding factors (see Additional file 2).

Discussion

This systematic review has synthesised evidence found in 44 studies regarding health service utilisation and social outcomes in adult CI users. The review identified limited research on health service utilisation post-CI. A systematic review that incorporated quantitative pooling of prospective studies to produce overall effect size is imperative. Despite a small number of studies examined health service utilisation, more than half found benefits of a CI. Most included studies (77.1%) have reported improvements for at least one social outcome post-CI.

Relatively small number of CI users who experienced complications required CI revision surgery. The review found that device failure (soft and hard) and medical-related predominantly skin flap-related infections were common reasons for the revision surgery. The current finding supports the need to maintain long-term follow-up post-CI to identify and manage any potential complications. Also, this study supports the importance of counselling for users about realistic expectations post-CI surgery. Prior research recommends that CI users receive a lifetime follow-up to identify and monitor any long-term complications [63, 64].

The current review found that CI users took prescribed medications for a lower number of illnesses than individuals on a wait list for a CI [50]. Similarly, prior studies have indicated that wearing hearing aids or cochlear implants improved communication abilities that further translated into improved health conditions and reduced unnecessary self-medication [19, 65]. However, it has been shown that waiting for medical intervention increases anxiety, depression and may reduce QoL [66]. As such, an individual’s position on a waiting list for CI may lead to negative emotions and associated physiological responses [50].

The current review found substantial improvement in work or employment status post-CI. These findings are consistent with a prior review that indicated evidence of improvement in work performance and employment status post-CI [36]. Only a few studies did not find an additional benefit of CI on work or employment. For instance, CI users reported challenges in their workplace because employers expected that CI can fully restore their normal hearing [45]. Further research that examines employers’ knowledge and expectations post-CI in the workplace may be of benefit.

Improvements were observed in social participation and in autonomy in almost all studies post-CI. Cochlea implant was associated with improved quality of life and speech perception which led to a demonstrated improvement in social participation [54, 67, 68]. CI was also found to improve independence in the adult population. Improvements in QoL may primarily be responsible for the increased feelings of autonomy or independence [69]. Similarly, a scoping review that examined the effects of CI on autonomy, participation and work found similar improvements post-CI [36]. The current review identified inconsistent definitions for both social participation and autonomy as well as inconsistency in the tools used to assess these constructs across studies. Within the included studies, there was an overlap between the examination of social participation and interrelated definitions, such as self-esteem, independence, activity limitations or QoL. Likewise, there were differences in the measurement of autonomy or independence across studies. Although the current review was inclusive of all definitions, further research is suggested to develop validated definitions for social participation and for autonomy or independence for CI users.

All studies that measured effects of CI on education or training found benefits post-CI. Compared with the general population, young adult CI users had a higher rate of attendance at post-secondary education programs and also reported being able to achieve their academic and personal goals [70]. Similarly, a review of five randomised controlled trials in adults who had received a hearing aid found that hearing aids were viewed as having improved educational opportunities [71].

Three studies reported on a CI recipient’s income [57, 58, 72] and all found a higher income level post-CI. However, none of the studies reporting on welfare found additional benefits post-CI. For example, Mo et al. 2004 compared CI with non-CI users and found no significant differences in terms of welfare or safety [59], even at 12 and 15 months post-CI [60]. Large scale and well-designed epidemiological studies are needed to examine the long-term association of CI on income or welfare benefits.

Studies that examined QoL found substantial improvement post-CI using different assessment tools [52,53,54, 59, 60, 62, 67,68,69, 73,74,75,76]. The current review findings are in line with previous reviews that found improvement in QoL post-CI [77,78,79]. For example, Andries et al. 2021, found improvements in health-related quality of life (HRQoL) pre- and post-CI. Andries et al. study examined HRQoL in older adults only, but the current review examined general QoL and HRQoL, health service utilisation and other social outcomes both in younger and older adults. It seems that CI users felt they confidently communicated which resulted in improvements in health-related or general QoL. Prior research has also indicated improved communicative ability through rehabilitation resulted in improved QoL in older adults [80, 81].

The current review found a link between CI and anxiety or depression, with most included studies identifying improvements [59, 60, 62]. The current review is consistent with prior research which found improvements in internalising mental health conditions [82,83,84,85,86]. Improvements in mental health from pre-CI were seen in the first 6 months but diminished after a long-term follow-up at 12 months post-CI in a study of 20 older adults [62]. Several factors could explain these results including users may relate their expectations to unrealistic outcomes, such as device limitations and its maximum benefits post-CI [87].

In this review, CI was found to improve cognitive functions such as immediate memory, attention, and delayed memory subdomains at 12 months post-CI. In older adults, CI was associated with improved cognitive functions mainly by improving the attention domain [54, 61, 62]. These results corroborate previous research findings that found CI was associated with improvement in cognitive function in older adults [80, 82, 88, 89].

Strengths and limitations

The strengths of this review were that it followed the PRISMA guidelines, the search strategy was developed by consulting a university librarian, and dual screening and data extraction was conducted. Despite these strengths, there are limitations for this review. First, most studies reported results based on small sample sizes, which made it difficult to generalise findings to larger populations. Second, the validity of questionnaires for measuring outcomes such as social participation, autonomy or perceived hearing disability in included studies were not known. Studies have used different measurement tools for the same social outcomes, this may lead to variation in patient outcomes or inaccuracy of measured outcomes. Third, post-operative complications of care may not have been observed, or may be underreported, in the included studies. For example, these studies may have only been conducted with CI users who use their cochlear implant, limiting the knowledge that could be gained from CI users who may have stopped using their devices.

Conclusions

Despite identifying small body of evidence regarding health service utilisation, this review found benefits of CI in improving adults’ health service utilisation and social outcomes. Improvement in hearing and communication ability was shown to enhance social interactions and working life, and also to support independence in everyday life. However, the review highlights the need for large scale and well-designed epidemiological studies to measure health and social outcomes.