記住我
The rapid growth of digital health, sped up by the COVID-19 pandemic and associated lockdowns, brings the promise of improved health care efficiency, empowerment of consumers, and health care equity []. Digital health is the use of information and communication technology to improve health []. eHealth, which is a type of digital health, refers to the use of internet-based technology for health care and can be used by systems, providers, and patients []. At the time of this study (before COVID-19), examples of eHealth used by patients in Australia included searching for web-based health information, booking appointments on the web, participating in online peer-support health forums, using mobile phone health apps (mobile health), emailing health care providers, and patient portals for electronic health records.
Digital health is expected to improve chronic disease management and has already shown great potential in improving chronic disease health outcomes [,]. Just under half of the Australian population (47.3%) has at least 1 chronic disease []. Rates of chronic disease and complications from chronic disease are overrepresented among those with socioeconomic disadvantage []. Therefore, patients with chronic disease and socioeconomic disadvantage have a greater need for the potential benefits of digital health, such as an improvement in their health outcomes. However, there is a risk that those who could benefit most from digital health services are the least likely to receive them, exemplifying the inverse care law in the digital age by Hart [].
Our Current Understanding of the Digital Health DivideWhile the rapid growth of digital health brings the promise of health care equity, it may also intensify existing inequities []. The digital health divide for socioeconomic disadvantage describes a pattern in which patients considered socioeconomically disadvantaged who are already marginalized through poor access to traditional health care are additionally hindered through poor access to digital health []. In Australia, only 67.4% of households in the lowest household income quintile have home internet access, compared to 86% of the general population and 96.9% of households in the highest household income quintile []. Survey-based studies have also shown that even with internet access, effective eHealth use is lower in populations considered disadvantaged, which speaks to broader barriers to digital health access [].
The ongoing COVID-19 global pandemic has sped up digital health transitions with the rapid uptake of telephone and video consultations, e-prescription, and the ongoing rollout of e-mental health in Australia. These have supported the continuation of health care delivery while limiting physical contact and the pandemic spread; however, the early evidence shows that the digital health divide remains problematic. A rapid review identified challenges with reduced digital access and digital literacy among the older adults and racial and ethnic minority groups, which are both groups at greater health risk from COVID-19 infections []. An Australian population study showed that the rapid uptake of telehealth during peak pandemic was not uniform, with the older adults, very young, and those with limited English language proficiency having a lower uptake of general practitioner (GP) telehealth services [].
To ensure that digital health improves health care outcome gaps, it is essential to better understand the nature and nuance of the digital health divide for socioeconomic disadvantage. The nature of the digital health divide for socioeconomic disadvantage has been explored primarily through quantitative survey data, some qualitative papers, a few mixed methods papers, and systematic reviews [,-]. Identified barriers include a lack of physical hardware and adequate internet bandwidth, a reduced inclination to seek out digital health, and a low ability and confidence to use digital health effectively []. The few mixed methods studies that exist on the digital health divide generally triangulate quantitative and qualitative data on a specific disease type or population subgroup to draw a combined conclusion [,]. These studies have found digital health access to be associated with education, ethnicity, and gender as well as trust, complementary face-to-face services, and the desire for alternative sources of information [,].
What This Work AddsThis project sought to extend previous research by using an exploratory mixed methods design to ensure that the first step and driver of our survey of a larger population was primarily shaped by the experiences of our interviewees within primary care. This differs from the triangulation method, which places the qualitative and quantitative data as equal first contributors to the findings and does not allow one type of data to determine the direction of the other []. We qualitatively explored the experience of digital health for patients with socioeconomic disadvantage and chronic disease and then quantitatively measured some of the qualitative findings via a survey of the Australian general practice patient population. Our key objective was to provide comprehensive primary mixed methods data, describing the experience and extent of barriers to accessing digital health and its benefits, with a focus on the digital health divide. We completed this research in a primary care context to investigate a diverse community-based population with conceivable reasons to seek digital help in managing their health. Findings from this mixed methods study were intended to provide health care providers and policy makers with a more detailed understanding of how specific barriers affect different aspects or steps of accessing digital health. Ultimately, understanding digital health access can influence the future design and implementation of digital health services by more effectively avoiding certain barriers or building in enablers to achieve improved digital health access not only for everyone but also especially for those in need.
We conducted a sequential exploratory mixed methods study to explore a complex phenomenon in depth and then measure its prevalence. We qualitatively explored the experience of digital health for patients with chronic disease and socioeconomic disadvantage in the first phase. Data from the first phase informed a quantitative survey of the phenomenon across a wider population in the second phase []. Both stages of research were conducted before the COVID-19 pandemic in Australia.
RecruitmentQualitative Phase ParticipantsThe eligibility criteria for the qualitative phase were as follows: English-speaking adults aged ≥18 years with at least 1 self-reported chronic disease and 1 marker of socioeconomic disadvantage (indicated by ownership of a Health Care Card or receiving a disability pension, unemployment, or a user of public housing). A chronic disease was defined to potential participants as a diagnosed long-term health condition that had lasted at least 6 months (or is expected to last for at least 6 months; examples are listed in ). The markers of socioeconomic disadvantage we used to identify potential participants were based on criteria typically used by local general practices to determine which patients can have lower or no out-of-pocket expenses. Apart from unemployment, the 3 other criteria to identify socioeconomic disadvantage are means-tested government-allocated public social services []. Qualitative phase participants were recruited from May to July 2019 through 3 general practices and 1 service organization that serve populations considered socioeconomically disadvantaged across urban, regional, and rural regions in the Australian Capital Territory and South Eastern New South Wales. A total of 2 recruitment methods were used in consultation with and as per the choice of the participating organizations. Potential participants were either provided with an opportunity to engage with researchers (KB and MAC) in the general practice waiting room or identified by the practice or organization as suitable for an interview. Interested participants were given a detailed verbal and written description of the project in a private space before providing written consent to be interviewed. All interview participants received an Aus $50 (US $32.68) grocery shopping voucher in acknowledgment of their time.
Quantitative Phase ParticipantsEligibility for the quantitative phase was English-speaking adults aged ≥18 years. The eligibility criteria for the quantitative phase were deliberately broader than those for the qualitative phase to achieve a larger sample size within the limitations of recruitment and with the intention that the factors of socioeconomic disadvantage and having a chronic disease could be compared to the digital health access of a more general population. The quantitative phase participants were recruited from November 2019 to February 2020. Study information and paper-based surveys were distributed and collected through 24 general practices across the Australian Capital Territory and South Eastern New South Wales regions, with an option for web-based completion.
Ethical ConsiderationsQualitative and quantitative phase research protocols, including the participant information sheet, were approved by the Australian Capital Territory Health Human Research Ethics Committee (2019/ETH/00013) and the Australian National University Human Research Ethics Committee (2019/ETH00003). Qualitative phase participants were given a verbal and written explanation of the study, including how and when they could opt out, before they provided written consent. All interview participants received an Aus $50 (US $32.68) grocery shopping voucher in acknowledgment of their time. Quantitative participants were given a written explanation and their informed consent was implied by return of a completed survey. Participants in both phases of the study were told that all their data was deidentified. Consent was implied through the return of a completed survey.
Qualitative Data Collection and AnalysisParticipants were purposively sampled to represent a range in age, gender, degree of socioeconomic disadvantage, and experience of digital health. The sampling and sample size were reviewed regularly by the research team as the interviews were being completed to identify potential thematic saturation.
The interview guide was developed by the research team based on a review of the literature and the patient dimensions of the framework of access by Levesque et al []. The framework by Levesque et al [] is a conceptualization of health care access comprising 5 service and patient dimensions of accessibility and ability. The patient dimensions are as follows: (1) ability to perceive, (2) ability to seek, (3) ability to reach, (4) ability to pay, and (5) ability to engage []. The key interview topics included (1) digital health use and access, including facilitators and barriers; (2) attitudes toward digital health; and (3) self-perception of digital health skills and potential training. The interview guide was reviewed for face and content validity by the whole research team, a patient advocate, a digital inclusion charity representative, and the general practices where recruitment occurred. The questions and guide were iteratively refined by the research team to ensure relevance and support reaching data saturation. The interview guide has been provided as . The interviews, which took 45 minutes on average, were taped and transcribed. An interview summary sheet and reflective journal were completed by the interviewer after each interview to also capture nonverbal cues and tone.
Interview transcriptions were coded and processed by inductive thematic analysis. Data collection and analysis were completed in parallel to support the identification of data saturation. Data saturation was defined as no significant new information arising from new interviews and was identified by discussion with the research team []. The 2 interviewers (MAC and KB) independently coded the first 5 transcripts and reflected on them with another researcher (EAS) to ensure intercoder validity and reliability. The rest of the interviews were coded independently by the 2 interviewers, who regularly met to reflect on emerging themes and thematic saturation. Data saturation was initially indicated after 15 interviews and subsequently confirmed with a total of 19 interviews. Coding disagreements and theme development were discussed with at least 1 other researcher (EAS, ER, and KD). Thematic saturation and the final themes were agreed upon by the entire research team.
Quantitative Survey DevelopmentThe final themes derived in the qualitative phase of the project guided the specific quantitative phase research questions. The final themes were a list of ordered cumulative barriers experienced by participants in accessing digital health and its benefits (). The quantitative survey was designed to test the association between barriers to access and the frequency of use of digital health as a proxy measure for digital health access.
Figure 1. Barriers to digital health access learned in phase 1. In the survey, the participants were asked about their demographic details, health and chronic diseases, knowledge, use and experience of digital health tools, internet access, perception of digital resource affordability, trust in digital health and traditional health services, perceived capability, health care empowerment, eHealth literacy, and relationship with their GP.
Existing scales and questions from the literature and standardized Australian-based surveys were used whenever possible. We used selected questions and scales from the Australian Bureau of Statistics standards, the eHealth Literacy Scale (eHEALS), the eHealth Literacy Questionnaire, and the Southgate Institute for Health Society and Equity [,-]. We adapted other scales from the ICEpop Capability Measure for Adults, the Health Care Empowerment Inventory (HCEI), the Patient-Doctor Relationship Questionnaire, and the Chao continuity questionnaire [,-]. Where an existing scale to measure a barrier or theme did not exist, the research team designed the questions based on the literature. Our questions around the frequency of digital health use were informed by multiple existing Australian-based surveys on general technology use [,]. Most of the questions used a Likert scale. Every choice regarding the design, adaptation, or copy of questions for the survey was influenced by the qualitative findings and decided on by full agreement among the 2 researchers who completed and coded the interviews. A complete copy of the survey is provided in .
Pilot-testing of the survey was completed with 5 patients, 2 experts on digital inclusion, and 3 local GPs for both the paper surveys and web-based surveys via Qualtrics Core XM (Qualtrics LLC). The resulting feedback on face and content validity, functionality of the survey logic, and feasibility of questionnaire completion was incorporated into the final version of the survey.
The survey was offered on paper with a participant information sheet, which gave the patients the option to complete the web-based survey. The survey was handed out to every patient on paper to avoid sampling bias through the exclusion of participants who could not complete the web-based survey [].
Quantitative Data Treatment and AnalysisData were exported from Qualtrics Core XM to an SPSS (version 26; IBM Corp) data set. Data cleaning and screening were undertaken (KB and KO).
Descriptive statistics (number and percentage) were used to summarize participant characteristics, preference measures, and frequency of eHealth use. Significance testing was conducted using chi-square tests, with a threshold of P<.05; effect sizes were measured by the φ coefficient for 2×2 comparisons and Cramer V statistic for all others. Where the cells sizes were too small, the categories were collapsed for the purposes of significance testing. The interpretation of effect sizes was as per the study by Cohen []. The analysis was conducted in SPSS and SAS (version 9.4; SAS Institute).
Participant CharacteristicsParticipants’ self-reported characteristics included gender, indigenous status, income category, highest level of education, marital status, and language spoken at home.
Age was derived from participant-reported year of birth and year of survey completion as of 2019 and stratified into age groups. The state or territory of residence was derived from the participant-reported postcode. The remoteness area was derived using the postcode reported by the participants and mapped to a modified concordance from the Australian Bureau of Statistics. Occupation-free text responses were coded using the Australian Bureau of Statistics Census statistics level 1 and 2 descriptors. The country of birth was mapped to Australia, other Organisation for Economic Cooperation and Development countries, and non–Organisation for Economic Cooperation and Development countries.
Frequency of eHealth UseA summary measure of the frequency of eHealth use was derived from the questions on the use of different types of eHealth.
Specifically, respondents were asked if they had ever used any form of web-based health (“eHealth“) and, if so, to rate how often (never, at least once, every now and then, and most days) against 6 types of “eHealth” (searching for health information online, booking appointments online, emailing health care providers, using health-related mobile phone apps, accessing My Health Record, and accessing online health forums). The frequency of eHealth use was then classified as follows:
High user: answered “most days” to at least 1 question on eHealth use OR answered “every now and then” to at least 2 questions on eHealth useNever user: answered “no” to having ever used any form of eHealth OR “never” to all 6 questions on eHealth useLow or medium user: all other respondents.The frequency of eHealth use was reported as unweighted descriptive statistics (counts and percentages) against demographic characteristics and for the elements of each of the themes identified in phase 1.
ThemesOverview of Key ThemesData were reported against the 6 themes from the phase 1 results of preference, trust, cost, structural access, capacity to engage, and self-efficacy. Where the components of trust, cost, capacity to engage, and self-efficacy had missing data (for less than half of the components only), mean imputation was used to minimize data loss. For each theme, the analysis excluded those for whom the frequency of eHealth use was unknown.
PreferencePreference measures (survey section D1 parts 1 to 3) asked participants to report against measures with a 4-point Likert scale (strongly disagree, disagree, agree, and strongly agree). Chi-square tests were conducted after the categories were condensed into 2 by combining strongly disagree and as well as combining strongly agree and agree.
TrustSummary measures for trust were created in 4 domains: trust from the eHealth Literacy Questionnaire (survey section D1 parts 4 to 8), trust from Southgate—GPs, specialists, or allied health (survey section D2 parts 1 to 5), trust from Southgate—digital health (survey section D2 parts 6, 7, 9, and 10), and trust from Southgate—books or pamphlets (survey section D2 part 8). The data were grouped as low, moderate, and high trust based on the assigned scores from the component data. Chi-square tests were conducted comparing low-to-moderate trust against high trust for GP, specialists, or allied health and comparing low trust against moderate-to-high trust for book or pamphlet.
CostSummary measures for cost were created from survey item C10. To measure cost, participants were asked about whether they considered certain items or services to be affordable. These included cost items mentioned in the qualitative phase interviews relating to mobile phones (1 that connects to the internet, 1 with enough memory space to download apps, downloads or apps requiring payment, repairs, and maintenance costs), having an iPad or tablet with internet connectivity, a home computer or laptop (owning, repairs, and maintenance), home fixed internet access, and an adequate monthly data allowance. These 9 items were scored as “yes definitely”=1 or 0 otherwise. Chi-square tests were conducted with never and low or medium eHealth users combined.
Structural AccessStructural access included asking where the internet is used by participants (survey section C8) and factors relating to internet access (survey section C8 parts 1-3) reporting against a 4-point Likert scale (strongly disagree, disagree, agree, and strongly agree). Chi-square tests were conducted with strongly disagree, disagree, agree, or strongly agree, and never, low, or medium eHealth use combined.
Capacity to EngageSummary measures for capacity to engage were created from survey section E1. To measure the capacity to engage, participants were asked about feeling “settled and secure,” “being independent,” and “achievement and progress” as an adaptation of the ICEpop Capability Measure for Adults [], reporting against a 4-point Likert-like scale. Responses were scored from 1 (“I am unable to feel settled and secure in any areas of my life”) to 4 (“I am able to feel settled and secure in all areas of my life”).
The summary capacity measure was derived by the summation of responses across the 3 questions, which were classified into 4 groups, A to D, based on these scores. Where fewer than half of the responses were missing, mean imputation was used; otherwise, the record was excluded. Groups A and B were combined for significance testing.
Self-EfficacySummary measures for self-efficacy were adapted from the eHEALS (E3) and the HCEI (E2) [,].
Survey section E3—eHEALS—comprised 8 questions, with participants reporting against a 5-point Likert scale for each (strongly disagree, disagree, neither, agree, and strongly agree). These responses were assigned 1 to 5 points, respectively. The summary eHEALS measure was derived by the summation of responses across the 8 questions, which were classified into 5 groups, A to E, based on these scores. Where fewer than half of the responses were missing, mean imputation was used; otherwise, the record was excluded. Groups A to C and D to E were combined for significance testing.
Survey section E2—HCEI—comprised 5 questions, with participants reporting against a 5-point Likert scale for each (strongly disagree, disagree, neither, agree, and strongly agree). Strongly disagree and disagree and neither were combined, and similarly agree and strongly agree were combined for significance testing.
The demographic characteristics of the patients that we interviewed are presented in .
The key barriers found to accessing digital health included (1) strong patient preference for human-based health services; (2) low trust in digital health services; (3) high financial costs of necessary tools, maintenance, and repairs; (4) poor publicly available internet access options; (5) reduced capacity to engage due to increased life pressures; and (6) low self-efficacy and confidence in using digital health.
Rather than being an equal list of factors, our interviewees described these barriers as a stepped series of cumulative hurdles, which is illustrated in . Initial issues of preference and trust were foundational to a person even when considering the option of digital health, while digital health confidence and literacy were barriers to full engagement with and optimal use of digital health. Alternatively, interviewees who did use digital health had been enabled by the same factors that were barriers to others.
Table 1. Phase 1 participant characteristics (N=19).CharacteristicValue, n (%)GenderaGP: general practitioner.
bMultiple answers per respondent.
Strong Patient Preference for Human-Based Health ServicesSome patients expressed a strong preference for human-based health services rather than digital health services. In answer to a question about how digital health services could be improved, a patient said the following:
Well, having an option where you can actually bypass actually having to go through the app and actually talk directly to someone.For some patients, this preference for human-based health services appeared to be related to a lack of exposure to eHealth. These patients were not at all interested in or had never thought about digital health options. A participant responded the following to the interviewer’s questions:
Interviewer: So when...something feels not right, how do you find out what’s going on?For other patients, their preference for human-based health care stemmed from negative experiences with technology. These patients reported actively disliking computers and technology in general and were generally frustrated with what they saw as the pitfalls of technology. A patient stated the following:
If computers and internet weren’t so frigging slow because everything is on like the slowest speed network ever and there’s ads blocking everything. Ads, (expletive) ads.A patient felt that he was pushed out of the workforce due his inability to keep up with technology-based changes and thus made a decision to never own a computer:
But, you know, in those days when I was a lot younger those sorts of things weren’t about and they’re just going ahead in leaps and bounds and that’s one of the reasons why I retired early. I retired at 63 because it was just moving too fast and it’s all computers and all those sorts of things and I just couldn’t keep up.Several patients described low trust levels for digital and internet-based technology in general. Their low trust was generally based on stories they had heard of other people’s negative experiences. A patient said the following:
I don’t trust the internet to be quite honest. You hear all these stories about people getting ripped off and I’ve worked too hard to get what I’ve got rather than let some clown get it on the internet for me.Some of this distrust was specific to eHealth. For example, some patients were highly suspicious of the government’s motives with regard to digital health and were concerned about the privacy of their health information, which made them hesitant about the concept of a universal electronic health record. In response to the interviewer’s question, a participant said the following:
Interviewer: Are there any other ways you think that eHealth might help you?Another participant said the following:
I just would run a mile from it because I just wouldn’t trust it. It wouldn’t be used to, as I said, for insurance or job information.A wide variety of patients described affordability issues across several different aspects of the costs involved in digital health. They expressed difficulty in paying for the following items: a mobile phone that could connect to the internet, a mobile phone with enough memory space to download apps, mobile phone apps requiring extra payment without advertisements, mobile phone repair costs such as a broken screen, a computer or laptop, home internet access, and adequate monthly data allowance and speeds to functionally use the internet. Current popular payment systems, such as plans, were not feasible for some patients. A participant stated the following:
I don’t have a computer...I’m not in the income bracket to own a computer really. Like I could, if I got one on a plan kind of thing or if I saved up for x-amount of time. But then like if I was going on the plan I’d be paying interest for having it on like lay-buy kind of thing, paying it off, and if it ever got lost or stolen I would still have to repay that off, which is always a hassle. And yeah. Yeah, I’m like financially not in the state where I’m able to...own a computer right now as I’m kind of paying off a number of debts.Some patients described struggling without home internet access. While they noted some cost-free public internet access points, such as libraries, hotel bars, and restaurants, they often found these to be inconvenient, lacking in privacy, and constituting low-quality options for digital health. A patient stated the following:
...it’s incredibly slow at the library. And I know why...a friend I went to school with used to belong to the council and the way they set it up, they just got the raw end of the stick and it is really, really slow. It’s bizarre but you can go to the X Hotel and it’s heaps quicker.In response to the interviewer's question, a participant said the following:
Interviewer: And do you feel comfortable doing private stuff on computers at the library...?When discussing why they were not using digital health or why they had stopped using digital health, patients often described significant competing priorities and life pressures that affected their capacity to engage. An unemployed patient mentioned that his time and energy on the internet were focused primarily on finding work and that he barely had time to focus on his health in general, let alone engage in digital health.
Other patients reported that they often felt that their ability to learn about and spend time on digital health was taken up by caring for sick family members, paying basic bills, or learning English. Some patients said that the time they would have spent learning digital skills when they were growing up had been lost to adverse life circumstances such as being in jail:
So we didn’t have computers in the house when I was growing up. And I didn’t know I’ve never...I’ve been in and out of jail for 28 odd years so it sort of takes away from learning from this cause it’s a whole different… it’s a whole different way of using a telephone from a prison.Some patients had a pervasive self-perception of being slow learners and being unable to use technology. Their stories of being unconfident learners seemed to stem from the fact that they had been told throughout their lives that they were intellectually behind. A patient said the following:
The computer people...wouldn’t take my calls because I’ve always been dumb with that sort of stuff. Like I only found out this later on in life, but I’m actually severely numerically dyslexic. Like I have to triple-check everything with numbers.Another patient stated the following:
I like went to two English classes like a normal English class with all the kids and then another English class with about seven kids in there because I just couldn’t I don’t know maybe because I spoke another language at home and they sort of like know I was a bit backward.These patients and others had multiple missing pieces of information that they felt made it harder to engage in digital health compared to “easier” human-based services. A patient said the following:
Yeah I’ve heard of booking online but I just I don’t know I find it easier just to ring up. And I’ll answer an email from a health care provider but I wouldn’t know where to start to look for their email address.In contrast, the patients who did connect with digital health described themselves as independent question askers and proactive people. Even when they did not know how to use a specific digital health tool, they were confident in attempting to and asking for help when they needed it. A patient said the following:
I’m a “I will find my way through this, no matter how long it takes me” kind of person. So maybe it’s more my personality...If I have to ask for help from somewhere, wherever it is, I will definitely do that.A total of 487 valid survey responses were received from participants across 24 general practices. The participant characteristics are presented in detail in Table S1 in .
The mean age of the participants was approximately 50 years (females 48.9, SD 19.4 years; males 52.8, SD 20.0 years), and 68.2% (332/487) of the participants identified as female. Overall, 34.3% (151/439) of respondents reported never using eHealth, and 53.8% (236/439) reported high eHealth use.
There were statistically significant (P<.05) differences in the frequency of eHealth use in terms of age group, gender, state, remoteness, highest level of education, employment status, occupation group, marital status, and language spoken at home, with effect sizes being small to medium. Specifically, high eHealth characteristics were associated with younger age, being female, living in an urban area, and being employed.
Preferencepresents the frequency of eHealth use against 3 internet preference questions.
Preference for using the internet and technology in general and for health needs in particular were significantly related to the frequency of eHealth use (P<.05 for each), with the effect sizes being small to medium.
Table 2. Preference measures.aExcludes those for whom frequency of eHealth use is unknown.
bChi-square tests conducted with strongly disagree and disagree combined, and agree and strongly agree combined.
Trustpresents the frequency of eHealth use against 4 measures of trust.
The degree of trust was not statistically significantly different for the frequency of eHealth use for any of the domains.
Table 3. Trust measures.aExcludes those for whom frequency of eHealth use is unknown.
beHLQ: eHealth Literacy Questionnaire.
cDerived from survey question D1, parts 4 to 8. Mean imputation used where ≤2 responses were missing. If >2 responses were missing, the records were excluded.
dDerived from survey question D2, parts 1 to 5. Mean imputation used where ≤2 responses were missing. If >2 responses were missing, the records were excluded.
eChi-square test conducted comparing low-to-moderate trust against high trust.
fDerived from survey question D2, parts 6, 7, 9, and 10. Mean imputation used where ≤2 responses were missing. If >2 responses were missing, the records were excluded.
gDerived from survey question D2 part 8.
hChi-square test conducted comparing low trust against moderate-to-high trust.
CostAffordability of items and services was reported as No cost difficulty or Cost difficulty. eHealth frequency of use responses were available for 273 participants; among those with no cost difficulty, 1% (2/204) were never users, 14.2% (29/204) were low or medium users, and 84.8% (173/204) were high users of eHealth; among those with cost difficulty, 1% (1/69) were never users, 26% (18/69) were low or medium users, and 73% (50/69) were high users. There was a statistically significant difference in the presence of cost as a barrier between never and low or medium eHealth users compared to high users (χ21=5.25; P=.02), although the effect size was small.
Structural Accesspresents the frequency of eHealth use for elements of structural access.
Quality of internet access and feeling limited in access to the internet were significantly associated with frequency of eHealth use (P<.05), although the effect sizes were small.
Table 4. Structural access.
留言 (0)