This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
The use of personal computers (PCs) and the Internet to provide health care information and interventions has increased substantially over the past decade. Yet the effectiveness of such an approach is highly dependent upon whether the target population has both access and the skill set required to use this technology. This is particularly relevant in the delivery of hearing health care because most people with hearing loss are over 50 years (average age for initial hearing aid fitting is 74 years). Although PC skill and Internet use by demographic factors have been examined previously, data do not currently exist that examine the effects of hearing difficulties on PC skill or Internet use in older adults.
To explore the effect that hearing difficulty has on PC skill and Internet use in an opportunistic sample of adults aged 50-74 years.
Postal questionnaires about hearing difficulty, PC skill, and Internet use (n=3629) were distributed to adults aged 50-74 years through three family physician practices in Nottingham, United Kingdom. A subsample of 84 respondents completed a second detailed questionnaire on confidence in using a keyboard, mouse, and track pad. Summed scores were termed the “PC confidence index.” The PC confidence index was used to verify the PC skill categories in the postal questionnaire (ie, never used a computer, beginner, and competent).
The postal questionnaire response rate was 36.78% (1298/3529) and 95.15% (1235/1298) of these contained complete information. There was a significant between-category difference for PC skill by PC confidence index (
Hearing health care is of particular relevance to older adults because of the prevalence of age-related hearing loss. Our data show that older adults experiencing slight hearing difficulty have increased odds of greater PC skill and Internet use than those reporting no difficulty. These findings suggest that PC and Internet delivery of hearing screening, information, and intervention is feasible for people between 50-74 years who have hearing loss, but who would not typically present to an audiologist.
The use of personal computers (PCs) and the Internet to provide health care and health-related information to patients and the public has increased substantially over the last decade [
Sensorineural hearing loss (SNHL) is highly associated with age, sex, and socioeconomic status (SES) [
Hearing aids are not the only form of intervention available for hearing loss. Computerized auditory training, such as Listening and Communication Enhancement (LACE) [
Severity of age-related SNHL increases with age, thus the hearing health care requirements of adults are likely to change as they get older. Davis et al [
Recent evidence suggests that the Internet may provide a means to facilitate communication in people with hearing loss because it removes the auditory barrier [
The primary aim of the present study was to explore the relationships among hearing difficulties and both PC skill and Internet use in an older adult population after accounting for the confounding demographic factors of age, SES, and sex. Should hearing difficulties be related to levels of PC skill and Internet use, it is hypothesized that individuals with hearing difficulty will have greater PC skill and Internet use than those reporting no difficulty. Furthermore, those effects may be greater as the degree of hearing difficulty increases.
Level of PC skill was defined in this study by using a three-category scale (ie, never used a computer, beginner, or competent). The category “never used a computer” is clear. However, the distinction between the “beginner” and the “competent” PC user categories is less well defined. As degree of computer use has been demonstrated to affect computer attitude measures including confidence with computers [
The primary measure was a 16-item postal questionnaire (
Three family practice physician offices in Nottingham, United Kingdom, participated in study recruitment. Postal questionnaires were sent to a total of 3529 patients on the family practice patient registers aged between 50-74, together with an invite letter from the family practice physician (GP in the United Kingdom) and the lead researcher (MAF). Patients were invited to complete and return the questionnaire in an enclosed reply-paid envelope. Non-respondents were not followed up; non-response was assumed to indicate a desire not to participate in the survey.
The response rate to the postal questionnaire was 36.77% (1298/3529), which is comparable with response rates for a recent national postal survey of family practice patients [
Respondents’ ages ranged from 50-74 years with a mean of 62.2 years (SD6.6 years). There were more female respondents (54.49%, 673/1235) than male (45.26%, 559/1235), but 3/1235 (0.25%) of respondents failed to report their sex. SES was determined by using the Index of Multiple Deprivation (IMD) score based on respondents’ postal codes. The IMD is a measure of deprivation by area, with higher IMD scores equating to lower SES, or greater deprivation. In our sample, IMD scores ranged between 3.65 and 78.37 with a mean of 27.87 (SD17.10). Nationally, IMD scores range between 0.99 and 84.22 [
PC skill was rated in the postal questionnaire on a three-category scale (ie, never used a computer, beginner, or competent). We aimed to verify this PC skill scale by examining whether individuals’ confidence in using a PC differed significantly among these three categories of PC skill.
A second 11-item PC confidence questionnaire (
Respondents ranged from 50-74 years with a mean age of 63.8 years (SD6.4). A total of 52/84 (62%) respondents were male and 32/84 (38%) were female. IMD scores ranged from 3.84-67.73 with a mean of 25.14 (SD17.45).
For all analyses, alpha<.05 was considered statistically significant. Initial correlations using Spearman rank correlation (ρ) were conducted to identify associations among PC skill, Internet use, better-ear hearing difficulty (BEHD), and demographic factors.
Main analyses assessed levels of PC skill and Internet use by reported hearing difficulty. Associated factors of age, SES, and sex were included within these analyses. Initially, univariate explorations were used to assess the individual relationship between each factor and PC skill and Internet use. All factors were then pooled within multivariate regression analyses to control for any confounding effects between factors. Through a backward elimination process, factors that were considered to be statistically significant (Wald z-statistic;
Secondary analyses assessed the three PC skill levels by using respondents’ PC confidence indexes from the subset of respondents who completed the PC confidence questionnaire (n=84). The PC confidence index for each respondent was compared to their selected PC skill category. Analysis of variance (ANOVA) was used to assess any significant differences between categories. These analyses sought to identify whether self-selected PC skill categories were an accurate reflection of the respondents’ confidence in using a PC.
Although verification of the self-reported PC skill categories was a secondary aim, the results for the PC confidence questionnaire respondents (n=84) are presented first because they form the basis of the main analyses to follow.
Mean PC confidence scores by reported PC skill categories are shown in
The primary objective of this study was to explore the effects of BEHD on PC skill and Internet use for the 1235 postal questionnaire respondents.
Degree of hearing difficulty was categorised on reported hearing difficulties in the better ear: no hearing difficulty, slight difficulty, and moderate or greater difficulties (moderate+ difficulty).
Prevalence of hearing difficulties was 26.80% overall (331/1235), which is comparable to data from a large UK hearing population study [
Over two-thirds of our sample (839/1235, 67.94%) reported being PC users, either beginner or competent, and 45.83% (566/1235) used the Internet. These figures are consistent with existing literature on PC and Internet use in older adults [
Respondents who reported slight hearing difficulty were equally likely to rate their PC skill as “never used a computer” (30.6%, 70/229) as those reporting no hearing difficulty (30.8%, 278/904), and equally likely to rate their PC skill as “competent” (40.2%, 92/229) as those reporting no hearing difficulty (40.0%, 362/904). However, those respondents who reported moderate+ hearing difficulty were more likely to rate their PC skill as “never used a computer” (47.1%, 48/102) and less likely to rate their PC skill as “competent” (24.5%, 25/102).
Similarly, for Internet use, respondents who reported slight hearing difficulty were equally likely to use the Internet (45.9%, 105/229) as those reporting no hearing difficulty (47.5%, 429/904). However, those respondents who reported moderate+ hearing difficulty were less likely to report being an Internet user (31.4%, 32/102).
Mean PC confidence scores (range 0-3; 0=not at all confident to 3=confident) and PC confidence index (range 0-12) by PC skill level for computer confidence questionnaire respondents (n=84).
PC confidence measure | Self-reported PC skills |
Mean confidence |
||
Never | Beginner | Competent | ||
Keyboard | 0.24 (0.56) | 2.03 (0.85) | 2.75 (0.60) | 1.98 (1.17) |
Mouse | 0.64 (1.01) | 2.21 (0.92) | 2.91 (0.28) | 2.25 (1.09) |
Track pad | 0.31 (0.79) | 1.05 (1.25) | 2.19 (0.97) | 1.40 (1.28) |
Overall PC confidence | 0.24 (0.44) | 1.23 (0.77) | 2.49 (0.66) | 1.56 (1.10) |
PC confidence indexa | 1.22 (1.93) | 6.10 (3.01) | 9.94 (2.47) | 6.70 (4.21) |
a sum of four confidence scores
Mean and median PC confidence index by PC skill level for computer confidence questionnaire respondents (n=84).
Better-ear hearing difficulty (BEHD) and demographics of postal questionnaire respondents (n=1235) by PC skill and Internet use.
Characteristics | Total |
PC skill |
Internet use |
||||
Never |
Beginner |
Competent |
No |
Yes |
|||
|
None | 904 (73.20) | 278 (70.2) | 264 (73.3) | 362 (75.6) | 475 (71.0) | 429 (75.8) |
Slight | 229 (18.54) | 70 (17.7) | 67 (18.6) | 92 (19.2) | 124 (18.5) | 105 (18.6) | |
Moderate+ | 102 (8.26) | 48 (12.1) | 29 (8.1) | 25 (5.2) | 70 (10.5) | 32 (5.6) | |
|
50-54 | 195 (15.79) | 30 (7.6) | 54 (15.0) | 111 (23.2) | 67 (10.0) | 128 (22.6) |
55-59 | 237 (19.19) | 44 (11.1) | 69 (19.2) | 124 (25.9) | 93 (13.9) | 144 (25.5) | |
60-64 | 330 (26.72) | 87 (21.9) | 102 (28.3) | 141 (29.4) | 163 (24.4) | 167 (29.5) | |
65-69 | 261 (21.11) | 100 (25.3) | 87 (24.2) | 74 (15.4) | 171 (25.5) | 90 (15.9) | |
70-74 | 212 (17.17) | 135 (34.1) | 48 (13.3) | 29 (6.1) | 175 (26.2) | 37 (6.5) | |
|
0-20 | 492 (39.84) | 116 (29.3) | 152 (42.2) | 224 (46.8) | 223 (33.3) | 269 (47.5) |
21-40 | 393 (31.82) | 132 (33.3) | 109 (30.3) | 152 (31.7) | 217 (32.4) | 176 (31.1) | |
41-60 | 297 (24.05) | 121 (30.6) | 80 (22.2) | 96 (20.0) | 190 (28.4) | 107 (18.9) | |
61-80 | 53 (4.29) | 27 (6.8) | 19 (5.3) | 7 (1.5) | 39 (5.8) | 14 (2.5) | |
|
Male | 559 (45.26) | 160 (40.4) | 165 (45.8) | 234 (48.9) | 272 (40.7) | 287 (50.7) |
Female | 673 (54.49) | 235 (59.3) | 193 (53.6) | 245 (51.1) | 394 (58.9) | 279 (49.3) |
a 0.25% (3/1235) of respondents failed to report their sex
Percentage of postal questionnaire respondents (n = 1235), by PC skill level and Internet use, by better-ear hearing difficulty (BEHD) and demographic factors.
Although PC skill was shown to be significantly associated with Internet use in our sample (ρ=0.74, n=1235,
Univariate ordinal logistic regression analyses were conducted to examine any significant effects of BEHD, age, SES, and sex on reported PC skill. Results revealed a significant effect of moderate+ hearing difficulty (odds ratio [OR]=0.49, 95% confidence interval [CI] 0.34-0.72,
All factors were combined within a multivariate ordinal logistic regression model to assess the effects of hearing difficulty on PC skill while controlling for any confounding effects attributable to age, SES, and sex. To aid interpretation, PC skill was categorized as either PC “use” (those with beginner or competent PC skills) or “non-use” (respondents who have never used a computer). Within the categories of BEHD and sex, “none” (no difficulty) and “male” were used as the baselines for comparison, respectively. Results are presented in
Logistic regression coefficients (beta) and odds ratio estimates from the multivariate ordinal logistic regression models for factors affecting PC skill level.
Characteristics | All (50-74 years) |
Younger (50-62 years) |
Older (63-74 years) |
|||||||
beta | Odds ratio (95% CI) |
|
beta | Odds ratio (95% CI) |
|
beta | Odds ratio (95% CI) |
|
||
|
None | – | 1.00 |
– | – | 1.00 |
– | – | 1.00 |
– |
Slight | .13 | 1.14 |
.37 | ‑.24 | 0.79 |
.26 | .45 | 1.57 |
.02 | |
Moderate+ | ‑.54 | 0.58 |
.008 | ‑.74 | 0.47 |
.008 | ‑.36 | 0.70 |
.20 | |
|
‑.11 | 0.90 |
<.001 | ‑.04 | 0.96 |
.05 | ‑.16 | 0.85 |
<.001 | |
|
‑.02 | 0.98 |
<.001 | ‑.02 | 0.98 |
<.001 | ‑.02 | 0.98 |
<.001 | |
|
Male | – | 1.00 |
– | – | 1.00 |
– | – | 1.00 |
– |
Female | ‑.28 | 0.75 |
.01 | ‑.19 | 0.83 |
.23 | ‑.39 | 0.68 |
.02 |
Results from the multivariate analyses revealed a significant effect of moderate+ hearing difficulty on PC skill level when no difficulty was used as the baseline measure (OR=0.58, 95% CI 0.39-0.87,
Despite the slight hearing difficulty category showing no significant association with PC skill, a comparison using the likelihood ratio test (LRT) on AIC estimates of a model with and without BEHD revealed a significant difference between models (LRT=8.71, df=2,
There was a significant effect of moderate+ hearing difficulty (OR=0.47, 95% CI 0.26-0.86,
Although age and SES were significant factors related to PC skill, there was no difference between levels of PC skill for males and females (OR=0.83, 95% CI 0.61-1.12,
The final model showed a significant effect of moderate+ hearing difficulty (OR=0.49, 95% CI 0.27-0.89;
For the older group, there were no significant effects of moderate+ hearing difficulty (OR=0.70, CI 0.40-1.20,
Univariate logistic regression analyses revealed a significant effect of moderate+ hearing difficulty (OR=0.51, 95% CI 0.32-0.78,
All factors were combined within a multivariate ordinal logistic regression model to assess the effects of hearing difficulty on Internet use while controlling for any confounding effects of age, SES, and sex. Results are presented in
Logistic regression coefficients (beta) and odds ratio estimates from the multivariate logistic regression model for factors affecting Internet use.
Characteristics | All (50-74 years) |
Younger (50-62 years) |
Older (63-74 years) |
|||||||
beta | Odds ratio (95% CI) |
|
beta | Odds ratio (95% CI) |
|
beta | Odds ratio (95% CI) |
|
||
|
None | – | 1.00 |
– | – | 1.00 |
– | – | 1.00 |
– |
Slight | .02 | 1.02 |
.90 | ‑.35 | 0.70 |
.12 | .45 | 1.57 |
.05 | |
Moderate+ | ‑.55 | 0.58 |
.02 | ‑.74 | 0.32 |
.001 | .05 | 1.05 |
.88 | |
|
‑.11 | 0.89 |
<.001 | ‑.04 | 0.96 |
.04 | ‑.13 | 0.87 |
<.001 | |
|
‑.02 | 0.98 |
<.001 | ‑.02 | 0.98 |
<.001 | ‑.03 | 0.97 |
<.001 | |
|
Male | – | 1.00 |
– | – | 1.00 |
– | – | 1.00 |
– |
Female | ‑.47 | 0.62 |
<.001 | ‑.45 | 0.64 |
.009 | ‑.51 | 0.60 |
.007 |
There was a significant effect of moderate+ hearing difficulty on Internet use (OR=0.58, 95% CI 0.36-0.92,
Further analysis by using error sum of squares (SSE) on AIC estimates revealed that a model including BEHD did not perform significantly better than a model with this factor removed (SSE=5.49, df=2,
Results from the final model show a significant effect of age (OR=0.89, 95% CI 0.88-0.91,
There was a significant effect of moderate+ hearing difficulty (OR=0.32, 95% CI 0.16-0.62,
For the older group, there was a significant effect of slight hearing difficulty (OR=1.57, 95% CI 0.99-2.47,
Hearing-related interventions are of particular relevance to older adults because of the increasing prevalence of hearing loss with age, particularly beyond 50 years. Our postal questionnaire survey identified that for the younger group (aged 50-62 years), levels of PC and Internet use were high (81.0% and 60.9%, respectively), whereas for the older group (63-74 years), PC and Internet use was considerably less (54.0% and 29.8%).
Hearing difficulties were shown to be significantly associated with PC skill and Internet use after controlling for effects of age, sex, and SES, all of which are significant factors in hearing loss. The findings can be categorized into two main effects. First, those with slight hearing difficulties had significantly increased PC and Internet use compared with those with no hearing difficulties in the older group only. Second, those with moderate+ hearing difficulty (ie, moderate/great/cannot hear at all) had lower PC and Internet use than those with no hearing difficulties, both for the whole sample and within the younger group. Although literature exists that examines the effects of age on PC and Internet use [
The postal questionnaire sample was similar to those in other studies for response rate [
PC skill and Internet use differed according to age and hearing difficulty in our sample. Respondents with slight hearing difficulties in the older group had increased odds of PC use (OR=1.57, 95% CI 1.06-2.30,
There are 10 million people in the United Kingdom with significant hearing loss, yet only 2 million have hearing aids and just 1.4 million use their hearing aids regularly [
Two potential health care approaches for the older group with slight hearing difficulties are delivery of effective information and hearing screening. Information and advice to educate these older adults about the effects of hearing loss and the benefits of hearing aids could be delivered in the form of short, easy-to-use video tutorials [
Remote hearing screening through the Internet is another highly relevant intervention for this cohort, and it has been shown to provide early benefits [
Finally, for those with mild hearing losses where hearing aids are unlikely to offer substantial benefit compared to those with greater levels of hearing loss, PCs and the Internet could be used to deliver alternative interventions such as auditory training to help alleviate the difficulties associated with hearing loss [
In the younger age group (50-62 years), in which PC and Internet use was highest, there was no difference in PC or Internet use between respondents with slight hearing difficulty and those with no hearing difficulty. We suggest that this is a result of there being a high level of PC and Internet use in this group already, which masks specific differences due to hearing difficulties. The relatively high skill set and Internet use in this age group suggests that online delivery of hearing health care is feasible for this younger age group with hearing loss. Appropriate online hearing health care for these adults within this age range may include online information and advice on the detrimental effects of untreated hearing loss, with a view to promote earlier awareness of hearing-related interventions including auditory training.
For respondents in the older age group (63-74 years), in which PC and Internet use was lowest, those with moderate+ hearing difficulties were equally likely to use PCs and the Internet as those with no hearing difficulties. This suggests that online hearing-related interventions may not be particularly effective if targeting this group as a whole. Subsequently, hearing health care tailored for those with moderate+ hearing losses, which is most likely to be hearing aid provision and information or advice relating to amplification, would be best supplemented with additional methods of information support (eg, printed materials or video tutorials delivered through DVD for those who do not or cannot access PC and Internet technology). Of course, for the 30% in the older age group who do use the Internet, information and advice would still be a valid option. The relatively low PC and Internet use in this group does not mean that online delivery of hearing health care will not be suitable for most of this cohort in the future. As PC and Internet use is becoming more prevalent in older adults over time, online delivery of hearing health care may provide a cost-effective, efficient method of providing hearing health care for older adults with a moderate+ hearing difficulties in future years [
Some limitations of the present study should be highlighted. First, our sample is limited to a small demographic (adults aged 50-74 years living in Nottingham, United Kingdom) because we were specifically interested in the pre-hearing aid user population. Nevertheless, our respondents have been shown to be representative of published data in terms of the prevalence of hearing loss and PC and Internet use. The World Health Organization lists adult-onset hearing loss as the most common cause of disability worldwide, with presbycusis (age-related hearing loss) the leading cause of adult-onset hearing loss [
In the present digital era, delivery of health care information and intervention through PCs and the Internet is common and the traditional method of clinical or medical health care delivery is supplemented increasingly by online information and support [
Postal questionnaire (hearing questions).
Computer confidence questionnaire.
Akaike information criterion
analysis of variance
better-ear hearing difficulty
confidence interval
general practitioner
hearing level
Index of Multiple Deprivation
likelihood ratio test
National Health Service
National Institute for Health Research
odds ratio
personal computer
socioeconomic status
short message service (text message for mobile phones)
sensorineural hearing loss
error sum of squares
We would like to thank the practice managers of the general practices who took part in this study for their assistance with the postal survey: Linda Lawton, practice manager, Leen View Surgery, Bulwell, Nottingham, United Kingdom; Jonathon Cummins, practice manager, Torkard Hill Medical Centre, Hucknall, Nottingham, United Kingdom; and Sayed Rashanmoniri, practice manager, Wollaton Vale Health Centre, Wollaton, Nottingham, United Kingdom.
This research was funded by the National Institute for Health Research (NIHR).
None declared.