ARTICLEINFO ABSTRACT

Introduction: Differences in digital health literacy levels are associated with a lack of access to digital tools, usage patterns,
and the ability to effectively use digital technologies. Although some studies have investigated the impact of
sociodemographic factors on digital health literacy, a comprehensive evaluation of these factors has not been conducted.
Therefore, this study sought to examine the sociodemographic determinants of digital health literacy by conducting a
systematic review of the existing literature.
Methods: A search of four databases was conducted. Data extraction included information on study characteristics,
sociodemographic factors, and the digital health literacy scales used. Meta-analyses for age and sex were conducted using
RStudio software with the metaphor package.
Results: A total of 3922 articles were retrieved, of which 36 were included in this systematic review. Age had a negative effect
on digital health literacy (B = -0.05, 95%CI [-0.06; − 0.04]), particularly among older adults, whereas sex appeared to have no
statistically significant influence among the included studies (B = - 0.17, 95%CI [-0.64; 0.30]). Educational level, higher
income, and social support also appeared to have a positive influence on digital health literacy.
Discussion: This review highlighted the importance of addressing the digital health literacy needs of underprivileged
populations, including immigrants and individuals with low socioeconomic status. It also emphasizes the need for more
research to better understand the influence of sociodemographic, economic, and cultural differences on digital health
literacy.
Conclusions: Overall, this review suggests digital health literacy is dependent on sociodemographic, economic, and cultural
factors, which may require tailored interventions that consider these nuances.

1. Introduction may be more vulnerable to misinformation [7].

Differences in digital health literacy levels between
The development of electronic health records and individuals persist noticeably, as well as in their
telemedicine has led to an increasing use of digital online skills and Internet knowledge, which are
health tools, gaining significant momentum in the ultimately related to socioeconomic status and
2000s with the widely available internet [1]. Today, autonomy in the use of these tools [8,9],
the use of digital health tools has expanded to consequently contributing to social health
include health-related apps, wearable devices, and inequalities and poorer health outcomes [10].
online medical resources to help people manage Currently, there are several approaches to
their health and make informed decisions about address digital inequality. A three-level model for
health care. The digital divide has been presented by van Deursen &
Helsper [3], with the first digital divide level being
growing use of digital tools in the modern world associated with lack of access to digital tools and the
requires quick adaptation, which is ultimately means to access the Internet, the second level to
hindered or facilitated by access to new technologies usage patterns, and the last level being associated
and the skills needed for their adequate use [2,3]. As with the ability to use digital technologies effectively
the proper use of digital health tools, also known as and efficiently to achieve improved outcomes.
e-health or eHealth tools, demands a vast number of Nevertheless, digital divide – and each of these levels
skills, including numeracy, science literacy, present on the three-level model - is influenced by
technology use, health literacy, and the capacity to factors such as socioeconomic status, generation, sex,
critically appraise health information [4], exploring region, and health status, with the first being one of

* Corresponding author at: Department of Medical Sciences, Institute of Biomedicine – iBiMED- University of Aveiro, Campus Universitario de Santiago Agra do ´ Crasto - edifício 30
3810-193 Aveiro, Aveiro, Portugal. E-mail address: [email protected] (M. Estrela). https://doi.org/10.1016/j.ijmedinf.2023.105124
Received 16 March 2023; Received in revised form 18 May 2023; Accepted 5 June 2023 Available online 10 June
2023
1386-5056/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

the digital health literacy context constitutes a the main predictors of Internet access and associated
complex challenge. skills, directly influencing competent Internet use
Digital health literacy is defined as the ability to [3,11,12]. Furthermore, digital health literacy can be
find, understand, and use health information from influenced by other factors such as technology
digital sources [5], such as the Internet and mobile readiness, attitudes towards technology, and Internet
devices, and is strongly related to the frequency with use patterns [13–16]. Although some studies have
which people use different health and digital reported the influence of sociodemographic factors
resources. These resources include online video on digital health literacy, a systematic appraisal of
consultations, digital health records, social networks, these factors is lacking. Thus, this study aimed to
and other health-related applications aimed at analyze the sociodemographic determinants of digital
promoting and improving patient health [6]. The health literacy through a systematic review and meta-
burden of digital health illiteracy is significant, as analysis of the available studies on the topic.
those with difficulty navigating health information
M. Estrela et al. International Journal of Medical Informatics 177 (2023)
105124
2. Methods of the aforementioned sociodemographic
characteristics on digital health literacy levels, and
2.1. Screening and study selection the secondary outcome was the scale used to analyze
digital health literacy. The results were summarized
A search was conducted on November 24, 2021, qualitatively and quantitatively. This systematic
on MedLINE- PubMed, Scopus, Web of Knowledge, review and meta-analysis followed the PRISMA [19]
and EMBASE databases. To update the results and MOOSE [20] guidelines. Further
obtained, a new search was conducted on April 12, recommendations for conducting meta-analyses of
2022, using the same databases. The screening of the observational studies were retrieved from a study by
obtained articles was conducted by title and abstract Mueller [21].
by two independent researchers (ME and GS), and
the search strategy was primarily designed to identify 2.5. Statistical analysis
relevant studies that analyze the influencing factors
on digital health literacy, and identify which scales All statistical analyses were conducted using
were used in these studies to measure digital health RStudio software (v. 4.2.2), and the packages metafor
literacy. The keywords used to search the [22], dplyr, and readxl. Regression coefficients were
aforementioned databases are as follows: estimated with 95% confidence intervals using a
random-effects model with Hedges and Olkin’s
(digital health OR e-health OR ehealth) AND estimator [23,24]. Heterogeneity due to differences
literacy AND (determinants OR factors OR between studies was assessed using Cochran’s Q and
sociodemographic OR demographic OR scale) I2-statistic [25,26]. Forest plots were used to visually
This systematic review was registered in the represent the presence of heterogeneity. Publication
PROSPERO database (CRD42022325207) [17]. bias was assessed using funnel plots [27] – see
supplementary figures S2). Sensitivity for age was
2.2. Selection criteria assessed through subgroup analyses of young adults,
adults, and older adults. As there were no evident
The inclusion criteria accepted studies that subgroups for sex, sensitivity analyses for this
analyzed the influence of sociodemographic factors, variable were conducted using the leave-one-out
such as sex, age, income, geographic region, and method. 3. Results
social status. There was no time restriction and the
languages of the included studies were English, 3.1. Screening
Portuguese, or Spanish. Conference abstracts,
systematic reviews, reviews, meta-analyses, A total of 3922 articles were retrieved from
editorials, study protocols, scale design and Pubmed, Scopus, WoS, and EMBASE databases, of
validation studies, correspondence papers, and which 1886 were duplicates (Fig. 1). Screening by title
studies that were not within the scope of our study and abstract was conducted, and 1926 records were
were excluded. excluded as they did not agree with the inclusion
All titles and abstracts obtained from the searches criteria, achieving a Cohen’s kappa of 0.623,
were independently reviewed by two researchers. corresponding to substantial agreement [28]; 110
The inclusion and exclusion criteria were applied by reports were analyzed by full text to check eligibility.
ME and GS and validated by a third researcher (TH) Of these, 36 were included in this systematic review.
when there was no agreement. Inter-rater agreement Cohen’s kappa of 0.861 was obtained, corresponding
was calculated using the Cohen’s kappa coefficient. to an almost perfect agreement [28] between the
Full-text articles were selected using the same researchers.
approach.
3.2. Quality analysis
2.3. Quality assessment
The quality analysis was conducted by two
The quality of the included studies was assessed independent researchers. From the thirty-seven
using the Joanna Briggs Institute Checklist for Cross- studies included after full-text screening, one study
sectional studies [18]. For each study, the risk of bias was removed for inadequate reporting of results.
was assessed separately by two researchers (ME and Although the remaining studies had overall good
GS). Similar to the screening process, a third reviewer quality, some criteria were classified as “unclear” for
(TH) acted as a referee to reach a consensus in case some studies, especially regarding inclusion criteria
of disagreement. of the sample, exposure measurement, and
confounding factor identification/management. The
2.4. Data analysis results of the quality analysis are shown in
supplementary material - Table S2.
Data extraction retrieved information on authors,
year, country, study design, study population,
response rate, average digital health literacy score,
and a brief description of the main digital health
literacy determinants, namely sociodemographic
characteristics. The primary outcome was the impact
2
M. Estrela et al. International Journal of Medical Informatics 177 (2023)
105124

Fig. 1. PRISMA flowchart with screening results.

3.3. Studies’ characteristics

All studies used the eHealth Literacy Scale

(eHEALS), except for Zakar’s study [29], which used
the Digital Health Literacy Instrument. The weighted
average eHEALS score among the studies was 30.4 ±
2.4. Table 1 presents the main characteristics of the
studies included. The data extracted from the
regression analyses are presented in Table S1.
Table 2 summarizes the characteristics analyzed in
each study. Almost all the studies analyzed age and
sex. The results of the effect of each variable on
digital health literacy scores are included in the
supplementary material. 3.4. Study design and
setting

All studies had a cross-sectional design.

Seventeen studies were conducted on-site [13,16,30–
44], either through interviews or the distribution of a
questionnaire; nine studies were conducted online
[29,45–52], and five studies were conducted through
telephone (CATI) [53–57]. The remaining five studies

3
M. Estrela et al. International Journal of Medical Informatics 177 (2023)
105124
adopted a mixed approach [58–62], using two of the adults were the target population in six studies
aforementioned data collection strategies. [16,31,36,39,49,55], while one study included
younger and older adults [59]. Of the remaining
3.5. Location twenty-two studies, eight targeted populations with
specific diseases [36,43,44,50,51,54,60,61], one
The included studies are distributed around the targeted at healthcare workers [46], and the
globe: almost half of the studies (14/36) were remaining had patients in general or the general
conducted in Asian countries [13,29–35,45– population as the target population.
48,58,59]. Twelve studies were conducted in North
America [16,36–39,49–51,53–55,60], six in Europe 3.6.1. Digital health literacy levels
[40,41,56,57,61,62], three in Africa [42,43,52] and From studies that presented average eHEALS
one in Australia [44]. scores among respondents, it was observed that
healthcare workers [46] and online health consumers
[52] presented higher levels of digital health literacy.
3.6. Population characteristics

Seven studies were conducted among high school

or college students [29,30,32,35,41,42,47]. Older

Table 1 Characteristics of the included studies.

*
Average (SD) per item.
**
Median (IQR).
***
Minimum-maximum normalization.

Table 2
Sociodemographic variables studied per study.
Author Age Sex Education Socioeconomic status Employment Ethnicity, race and Language spoken Household composition, social support, and residence

Abdulai, A ✓ ✓ ✓ ⨯ ⨯ ⨯ ⨯
Alhuwail, D ✓ ✓ ✓ ⨯ ⨯ ⨯ ⨯
Arcury, T ⨯ ⨯ ⨯ ⨯ ⨯ ✓ ⨯
Bergman, L ✓ ✓ ✓ ⨯ ⨯ ✓ ⨯
Berkowsky, R ✓ ✓ ✓ ✓ ✓ ✓ ✓
Cherid, C ✓ ✓ ✓ ⨯ ⨯ ⨯ ⨯
Choi, N ✓ ✓ ⨯ ✓ ⨯ ✓ ✓
De Santis, K ✓ ✓ ✓ ✓ ⨯ ⨯ ⨯
Do, B ✓ ✓ ⨯ ✓ ⨯ ⨯ ✓
Gazibara, T ✓ ✓ ✓ ✓ ⨯ ⨯ ✓*
Guo, Z ✓ ✓ ✓ ✓ ✓ ⨯ ✓
Kim, S Knapp, ✓ ⨯ ⨯ ⨯ ⨯ ⨯ ⨯
C ✓** ✓* ✓* ✓ ⨯ ✓* ✓*
Lee, O ✓ ✓ ✓ ⨯ ⨯ ⨯ ⨯
Lee, W ✓ ✓ ✓ ✓ ✓ ✓ ⨯
Makowsky, M ✓ ✓ ✓ ⨯ ⨯ ✓ ✓
Maroney, K ✓ ✓ ✓ ⨯ ✓ ✓ ✓
Mengestie, N ⨯ ✓ ⨯ ⨯ ⨯ ⨯ ✓
Moon, Z ✓ ⨯ ✓ ✓ ⨯ ✓ ⨯
Morton, E ✓ ✓ ✓ ⨯ ⨯ ⨯ ⨯
Nguyen, L ✓ ✓ ⨯ ✓ ⨯ ⨯ ⨯
Richtering, S ✓ ✓ ✓ ✓ ⨯ ⨯ ⨯
Schrauben, S ✓ ⨯ ✓ ✓ ⨯ ✓ ⨯
Schrauben, S ✓ ⨯ ✓ ✓ ⨯ ✓ ⨯
Shiferaw, K ⨯ ⨯ ✓ ✓ ✓ ⨯ ✓
Stellefson, M ✓ ✓ ✓ ⨯ ⨯ ⨯ ✓
Tennant, B ✓ ✓ ✓ ✓ ⨯ ✓ ✓
Tran, H ✓ ✓ ⨯ ✓ ⨯ ⨯ ⨯
Vicente, M ✓ ✓ ✓ ✓ ✓ ⨯ ✓
Wong, D ✓ ✓ ✓ ⨯ ✓ ⨯ ⨯
Xesfingi, S ✓ ✓ ✓ ✓ ⨯ ⨯ ✓
Xu, R ✓ ✓ ✓ ✓ ✓ ⨯ ✓
Yang, E ✓ ⨯ ✓ ✓ ⨯ ⨯ ✓
Zakar, R ✓ ✓ ⨯ ✓ ⨯ ⨯ ⨯
Zhou, J ✓ ✓ ✓ ⨯ ⨯ ⨯ ✓
Yang, S ✓ ⨯ ⨯ ⨯ ⨯ ⨯ ⨯
**
-✓Parents and children. students belonged to the studies with the lowest
*
-✓Parents. average eHEALS scores [31,32]. It was also observed
that most studies presenting higher levels of digital
However, although nursing students from Tran’s health literacy were conducted online and/or
study [47] presented high scores on eHEALS on through phone. In studies comparing two distinct
average, other groups of nursing and medical groups, while no significant differences were noted

4
M. Estrela et al. International Journal of Medical Informatics 177 (2023)
105124
among young adults and older adults [59],
statistically significant differences were observed
among Arabic and Swedish speakers [40], those
under and above 60 years of age [60], and US and
South Korean older adults [31].

3.6.2. Age
The weighted average age of all participants was
40.92 ± 15.37 years old. Considering the statistically
significant outcomes, all articles, except Morton [50],
reported a negative association between age and
eHEALS scores. People aged over 75 years are up to
four times more likely to have lower levels of digital
health literacy [36]. When conducting a meta-
analysis, a statistically significant negative effect of
age on eHEALS scores was observed (B = -0.05, 95%CI
[-0.06;-0.04]) (Fig. 2). After conducting subgroup
analyses (Supplementary figure S1), a significant
effect of age on the eHEALS scores among older
adults was observed. The young adults’ subgroup
presented no heterogeneity but no significant results,
whereas the adult group also presented non-
significant results and substantial heterogeneity.

3.6.3. Sex Fig. 2. Meta-analysis results for age.

Regarding sex, five studies had more than 80%
females as participants, and only seven studies had
over 50% males as a study population. From the 28
studies that analyzed the influence of sex on eHEALS
scores, only ten studies presented significant results
[29,32,41,42,45–48,52,57]. However, although
significant, these results present high heterogeneity,
with females presenting higher levels of digital health
literacy in some studies and lower levels in others.
Fig. 3 presents the forest plot for the studies included
in the meta-analysis, with an effect size of –0.17,
95%CI [-0.64;0.30], confirming the high
heterogeneity between the studies and the non-
significance of the pooled results. A table with a
sensitivity analysis is presented in the Supplementary
Material.

3.6.4. Education
Among the 27 studies that analyzed the influence
of education, every study with statistically significant
results presented a positive influence of educational
level on eHEALS scores
[13,31,36,38,40,43,45,48,50,51,55,56,58,62], with
those with a college degree or higher being Fig. 3. Meta-analysis results for sex.
particularly predisposed to present higher digital
health literacy. Gazibara’s study [41] analyzed 3.6.5. Socioeconomic status
parents’ highest educational attainment, with no Regarding income, five studies presented
statistically significant results. Moon’s study [61] significant results [38,43,56,58,62], with higher
presents a positive influence of the age at which income associated with higher eHEALS levels. Two of
respondents left full-time education on eHEALS these studies only presented significant results for
scores. populations in the highest income bracket [43,58].
Other studies measured the income-to-needs ratio
[60], financial and social status [29,46,59], economic
condition [32], ability to pay for medication [47],
healthcare insurance scheme [34], and Index of
Multideprivation quintile [61], where individuals
belonging to the 3rd quintile presented lower digital

5
M. Estrela et al. International Journal of Medical Informatics 177 (2023)
105124
health literacy when compared to their least- subgroup analyses, studies conducted among young
deprived counterparts. adults showed no significant impact of age on the
DHL levels. Studies of the general population are
3.6.6. Employment highly heterogeneous; thus, they lack sufficient
Seven studies analyzed the influence of consistency to draw conclusions. However, the
employment status on eHEALS scores, with only Xu’s included studies on older adult subgroups presented
study presenting a significant difference in digital a significant negative impact of age on eHEALS
health literacy scores between employed and scores. Still, these results should be carefully
unemployed individuals [34]. considered, as only two studies were included in this
subgroup. Thus, our results suggest that age may not
3.6.7. Ethnicity, race, and languages spoken be a strong predictor of digital health literacy on its
Ethnicity appeared to have no effect on digital own.
health literacy in the included studies, with the The review included studies that investigated the
exception of Lee’s study [13] conducted in Malaysia, direct impact of socioeconomic factors such as
where Malaysian Chinese people presented lower education, income, and employment status on digital
levels of eHEALS scores. Tennant et al. [55] evaluated health literacy. As expected, the results showed that
the influence of race and achieved non-significant individuals with higher levels of education tended to
results. Choi [60] and Bergman [40] analyzed the have higher digital health literacy than those with
influence of the languages spoken on eHEALS scores. lower levels of education, thus reinforcing its role in
In Choi’s study, being Spanish-speaking in the US had digital health literacy. These results agree with the
no influence on eHEALS scores, while being an Arabic previously published literature on the impact of
native speaker in Sweden was associated with a educational level on health literacy, digital literacy,
higher probability of presenting low digital health and digital health literacy [64,65]. Socioeconomic
literacy. status also seemed to influence the level of digital
health literacy. As low socioeconomic status is
3.6.8. Household composition, social support, and associated with suboptimal use of health resources
residence and health status [64], it is only expected that it is
None of the studies that analyzed the influence of also reflected in the ability to acquire adequate
marital status on eHEALS scores showed statistically health information from digital sources. Furthermore,
significant differences between the groups. Four access to the Internet and digital health tools may
studies analyzed the impact of residence on digital also be severely hindered by individuals’ economic
health literacy, all with no statistically significant status, thus highlighting the importance of
results. While having children appeared to have no reinforcing digital health interventions among those
influence [34], living alone and lack of social support who are the most underprivileged [3,11,12]. This
were associated with lower eHEALS [34,51]. review also highlights the influence of social support
on digital health literacy, suggesting that individuals
4. Discussion
with more social support tend to have higher health
literacy than those with less social support.
The role of various socioeconomic and
While ethnicity by itself appears to have no effect
demographic factors in determining DHL has been a
on digital health literacy levels on most of the
subject of interest in the research community. This
included studies, and appear to be very dependent
systematic review and meta-analysis discusses the
on the context, being a native speaker may constitute
influence of sociodemographic determinants on DHL,
an advantage for higher digital health literacy levels
an important aspect of healthcare that involves an
when compared to their immigrant counterparts. As
individual’s ability to access, understand, and use
immigrants are at a higher risk of social exclusion,
health information from digital sources. Overall, our
which consequently hinders access to healthcare
results suggest that there are some factors that may
services [66], it is also important to target DHL
directly influence the digital health literacy levels,
interventions for these populations, ultimately
such as age, education, and social support.
helping them navigate the country’s health systems.
Most studies have analyzed the impact of age and
One of the limitations of this review was the high
sex on DHL levels. One interesting finding of this
heterogeneity and variability of the included studies.
systematic review and meta-analysis is that sex does
This was expected, as the included studies were
not appear to be a significant determinant of digital
cross-sectional, each representing only a specific
health literacy. While sex is often associated with
population, and being at risk for a higher selection
disparities in health outcomes, studies demonstrate
bias of participants. Furthermore, it is observed that
that these differences are also highly influenced by
those studies conducted through an online survey
other variables, such as cultural context, marital
tend to present higher average levels of digital health
status, and socioeconomic conditions [63]. Whereas
literacy, which might be result of an exclusion of
the impact of sex tends to be heterogeneous among
those individuals who are unable to use digital tools.
studies, a negative relationship between age and DHL
Additionally, this review only included articles that
levels appears to exist. Though it may seem obvious
analyzed the direct influence of socioeconomic
that older individuals may have lower digital health
variables, and studies on differences between groups
literacy, the review found significant results only in
were excluded. However, the review had some
subgroup analyses. However, when conducting
6
M. Estrela et al. International Journal of Medical Informatics 177 (2023)
105124
strengths, including the inclusion of studies from Validation, Writing – review & editing, Visualization,
around the world, the reliability of the scale used, Supervision.
and lack of publication bias. The eHEALS [4] is the
most widely used scale to measure digital health
Declaration of Competing Interest
literacy, presenting high levels of validity and
consistency [67,68]. However, while digital health The authors declare that they have no known
literacy may have not changed since its development, competing financial interests or personal
the context in which digital health skills are applied relationships that could have appeared to influence
nowadays has, and eHEALS focuses only on the work reported in this paper.
information gathering (Health 1.0 skills), disregarding
interactivity on the Web (Health 2.0) [69]. The ever- Acknowledgements
growing influence of social media on health decisions
and the risk of health misinformation [70] are We would like to thank Ms. Joana Antao and Mr.
prominent problems that were not as present as they Guilherme Rodri˜ gues, from the Institute of
are now. Furthermore, as there is so much Biomedicine, Portugal, for their help with the
misinformation circulating on the Internet, people statistical analysis.
with low digital health literacy may also be more
susceptible to cognitive bias, overestimating their Appendix A. Supplementary material
knowledge base [71]. As the eHEALS is a scale
comprising self-reporting questions, it may not Supplementary data to this article can be found
entirely reflect the actual digital health literacy levels online at https://doi.
of the population. Thus, it is also important to assess org/10.1016/j.ijmedinf.2023.105124.
actual digital health literacy through performance-
based items, possibly with reference to those References
developed by van der Vaart and Weiss [69,72].
[1] J. Jagarapu, R.C. Savani, A brief history of telemedicine and
the evolution of teleneonatology, Semin. Perinatol. 45 (5)
5. Conclusions (2021), 151416.
[2] A. Karnoe, D. Furstrand, K.B. Christensen, O. Norgaard, L.
In conclusion, this review provides valuable Kayser, Assessing competencies needed to engage with
digital health services: Development of the eHealth
insights into the influence of socioeconomic and
literacy assessment toolkit, J. Med. Internet Res. 20 (5)
demographic factors on digital health literacy. These (2018) 1–14.
findings suggest that DHL is multifactorial and may be [3] D.A.J.A.M. van, E.J. Helsper, The third-level digital divide:
influenced by cultural and contextual factors. Who benefits most from being online?, in: Communication
and information technologies annual 10 Emerald Group
Furthermore, educational level, social support, and Publishing Limited, 2015, pp. 29–52.
socioeconomic status may be key factors for [4] C.D. Norman, H.A. Skinner, eHealth literacy: Essential skills
improving digital health literacy. Nevertheless, this for consumer health in a networked world, J. Med.
Internet Res. 8 (2) (2006 Jun 16) e9.
review also highlights the need for more research to [5] WHO Regional Office for Europe, Report on the WHO
better understand the influence of symposium on the future of digital health systems in the
sociodemographic, economic, and cultural European region, WHO Regional Office for Europe,
Copenhagen, 2019.
differences on digital health literacy. Thus, we [6] The HLS19 Consortium of the WHO Action Network M-
recommend targeting digital health interventions POH. International report on the methodology, results,
considering these nuances, to improve their and recommendations of the European Health Literacy
Population Survey 2019-2021 (HLS19) of M-POHL
effectiveness - which is highly dependent on several [Internet]. Austrian National Public Health Institute,
underlying factors -, as they may not be one-size-fits- Vienna, 2021 [cited 2022 Apr 22]. Available from:
https://m-pohl.net/sites/m-pohl.net/files/inline-files/HLS1
all.
9_International% 20Report%20%28002%29_0.pdf.
[7] A.W.K. Yeung, A. Tosevska, E. Klager, F. Eibensteiner, C.
Funding Tsagkaris, E.D. Parvanov, et al., Medical and health-related
misinformation on social media: Bibliometric study of the
scientific literature, J. Med. Internet. Res. 24 (1) (2022)
This research was funded by an individual grant e28152.
by FCT (ref 2021.05141.BD). [8] E. Hargittai, A.M. Piper, M.R. Morris, From internet access
to internet skills: digital inequality among older adults,
Univ. Access Inf. Soc. 18 (4) (2019) 881–890.
CRediT authorship contribution statement
[9] I. Kickbusch, Health Literacy, the solid facts, World Health
Organization, Geneva, 2013, p. 73.
Marta Estrela: Conceptualization, Methodology, [10] K. Latulippe, C. Hamel, D. Giroux, Social health inequalities
Software, Validation, Formal analysis, Investigation, and eHealth: A literature review with qualitative synthesis
of theoretical and empirical studies, J. Med. Internet Res.
Data curation, Writing – original draft, Writing – 19 (4) (2017) e136.
review & editing, Visualization, Project [11] A. Cornejo Müller, B. Wachtler, T. Lampert, Digital divide-
administration, Funding acquisition. Guilherme social inequalities in the utilisation of digital healthcare,
Bundesgesundheitsblatt Gesundheitsforschung
Semedo: Formal analysis, Investigation, Data Gesundheitsschutz. 63 (2) (2020) 185–191.
curation, Writing – review & editing, Visualization. [12] M.C. Kim, J.K. Kim, Digital Divide: Conceptual Discussions
Fatima ´ Roque: Validation, Writing – review & and Prospect, in: W. Kim, T.W. Ling, Y.J. Lee, S.S. Park (Eds.),
the Human Society and the Internet Internet-Related
editing, Visualization, Supervision. Pedro Lopes
Socio-Economic Issues, Springer, Berlin, Heidelberg, 2001,
Ferreira: Validation, Writing – review & editing, pp. 78–91. Lecture Notes in Computer Science.
Visualization, Supervision. Maria Teresa Herdeiro:
7
M. Estrela et al. International Journal of Medical Informatics 177 (2023)
105124
[13] W.L. Lee, Z.J. Lim, L.Y. Tang, N.A. Yahya, K.D. Varathan, S.M. [34] R.H. Xu, L.M. Zhou, E.L.Y. Wong, D. Wang, the association
Ludin, Patients’ technology readiness and ehealth literacy: between patients’ ehealth literacy and satisfaction with
implications for adoption and deployment of eHealth in shared decision-making and well-being: Multicenter cross-
the COVID-19 era and beyond, Comput Inform Nurs. 40 sectional study, J. Med. Internet. Res. 23 (9) (2021)
(4) (2021) 244–250. e26721.
[14] S. Yusif, A. Hafeez-Baig, J. Soar, e-Health readiness [35] S.C. Yang, Y.F. Luo, C.H. Chiang, The associations among
assessment factors and measuring tools: A systematic individual factors, eHealth literacy, and health-promoting
review, Int. J. Med. Inf. 107 (2017) 56–64. lifestyles among college students, J. Med. Internet Res. 19
[15] E.V. Estacio, R. Whittle, J. Protheroe, The digital divide: (1) (2017) e15.
Examining socio- demographic factors associated with [36] C. Cherid, A. Baghdadli, M. Wall, N.E. Mayo, G. Berry, E.J.
health literacy, access and use of internet to seek health Harvey, et al., Current level of technology use, health and
information, J Health Psychol. 24 (12) (2019) 1668–1675. eHealth literacy in older Canadians with a recent fracture-a
[16] T.A. Arcury, J.C. Sandberg, K.P. Melius, S.A. Quandt, X. survey in orthopedic clinics, Osteoporos. Int. 31 (7) (2020)
Leng, C. Latulipe, et al., Older adult internet use and 1333–1340.
eHealth literacy, J. Appl. Gerontol. 39 (2) (2020) [37] M.J. Makowsky, S. Davachi, C.A. Jones, eHealth literacy in a
141–150. sample of South Asian adults in Edmonton, Alberta,
[17] Estrela M, Semedo G, Roque F, Lopes Ferreira P, Herdeiro Canada: Subanalysis of a 2014 community-based survey,
MT. Socioeconomic determinants for digital health literacy: JMIR Form. Res. 6 (3) (2022) e29955.
a systematic review. PROSPERO 2022 CRD42022325207 [38] S.J. Schrauben, L. Appel, E. Rivera, C.M. Lora, J.P. Lash, J.
[Internet]. Available from: Chen, et al., Mobile health (mHealth) technology:
https://www.crd.york.ac.uk/prosp Assessment of availability, acceptability, and use in CKD,
ero/display_record.php?RecordID=325207. Am J Kidney Dis. 77 (6) (2021) 941–950.e1.
[18] Joanna Briggs Institute. © Joanna Briggs Institute 2017 [39] L. Zibrik, S. Khan, N. Bangar, E. Stacy, H. Novak Lauscher, K.
Critical Appraisal Checklist for Analytical Cross Sectional Ho, Patient and community centered eHealth: Exploring
Studies, 2017; Available from: https://jbi.global/sit eHealth barriers and facilitators for chronic disease self-
es/default/files/2019-05/JBI_Critical_Appraisal- management within British Columbia’s immigrant Chinese
Checklist_for_Analytical_Cross and Punjabi seniors, Health Policy Technol. 4 (4) (2015)
_Sectional_Studies2017_0.pdf. 348–356.
[19] M.J. Page, J.E. McKenzie, P.M. Bossuyt, I. Boutron, T.C. [40] L. Bergman, U. Nilsson, K. Dahlberg, M. Jaensson, J.
Hoffmann, C.D. Mulrow, et al., The PRISMA 2020 Wångdahl, Health literacy and e-health literacy among
statement: an updated guideline for reporting systematic Arabic-speaking migrants in Sweden: a cross-sectional
reviews, BMJ 29 (372) (2021), n71. study, BMC Public Health 21 (1) (2021) 2165.
[20] D.F. Stroup, J.A. Berlin, S.C. Morton, I. Olkin, G.D. [41] T. Gazibara, J. Cakic, M. Cakic, T. Pekmezovic, A. Grgurevic,
Williamson, D. Rennie, et al., Meta-analysis of eHealth and adolescents in Serbia: psychometric
observational studies in epidemiology: a proposal for properties of eHeals questionnaire and contributing
reporting. Meta-analysis of observational studies in factors to better online health literacy, HealthPromot Int.
epidemiology (MOOSE) group, JAMA 283 (15) (2000) 34 (4)
2008–2012. (2019) 770–778.
[21] M. Mueller, M. D’Addario, M. Egger, M. Cevallos, O. [42] N.D. Mengestie, T.M. Yilma, M.A. Beshir, G.K. Paulos,
Dekkers, C. Mugglin, et al., Methods to systematically eHealth literacy of medical and health science students
review and meta-analyse observational studies: a and factors affecting eHealth literacy in an Ethiopian
systematic scoping review of recommendations, BMC University: A cross-sectional study, Appl. Clin. Inform. 12
Med. Res. Method. 18 (1) (2018) 44. (2) (2021) 301–309.
[22] W. Viechtbauer, Conducting meta-analyses in R with the [43] K.B. Shiferaw, B.C. Tilahun, B.F. Endehabtu, M.K. Gullslett,
metafor package, J Stat Soft. 36 (3) (2010). S.A. Mengiste, E-health literacy and associated factors
[23] L.V. Hedges, Estimation of effect size from a series of among chronic patients in a low-income country: a cross-
independent experiments, sectional survey, BMC Med. Inform. Decis. Mak. 20 (1)
Psychol. Bull. 92 (2) (1982) 490–499. (2020) 181.
[24] L.V. Hedges, I. Olkin, CHAPTER 9 - Random effects models [44] S.S. Richtering, K. Hyun, L. Neubeck, G. Coorey, J.
for effect sizes, in: L. Chalmers, T. Usherwood, et al., eHealth Literacy:
V. Hedges, I. Olkin (Eds.), Statistical Methods for Meta- Predictors in a population with moderate-to-high
Analysis, Academic Press, San Diego, 1985, pp. 189–203. cardiovascular risk, JMIR Hum. Factors. 4 (1) (2017) e4.
[25] M. Borenstein, In a meta-analysis, the I-squared statistic [45] D. Alhuwail, Y. Abdulsalam, Assessing electronic health
does not tell us how much the effect size varies, J. Clin. literacy in the state of Kuwait: Survey of internet users
Epidemiol. 9 (152) (2022) 281–284. from an Arab State, J Med. Internet Res. 21 (5) (2019)
[26] T.B. Huedo-Medina, J. Sanchez-Meca, F. Marín-Martínez, J. e11174.
Botella, Assessing´ heterogeneity in meta-analysis: Q [46] B.N. Do, T.V. Tran, D.T. Phan, H.C. Nguyen, T.T.P. Nguyen,
statistic or I2 index? Psychol. Methods 11 (2) H.C. Nguyen, et al., Health literacy, eHealth Literacy,
(2006) 193–206. adherence to infection prevention and control procedures,
[27] M. Egger, G.D. Smith, M. Schneider, C. Minder, Bias in lifestyle changes, and suspected COVID-19 symptoms
meta-analysis detected by a simple, graphical test, BMJ among health care workers during lockdown: Online
315 (7109) (1997) 629–634. survey, J. Med. Internet Res. 22 (11) (2020) e22894.
[28] M.L. McHugh, Interrater reliability: the kappa statistic, [47] H.T.T. Tran, M.H. Nguyen, T.T.M. Pham, G.B. Kim, H.T.
Biochem. Med. (Zagreb). 22 (3) (2012) 276–282. Nguyen, N.M. Nguyen, et al., Predictors of eHealth literacy
[29] R. Zakar, S. Iqbal, M.Z. Zakar, F. Fischer, COVID-19 and and its associations with preventive behaviors, fear of
health information seeking behavior: Digital health literacy COVID-19, anxiety, and depression among undergraduate
survey amongst university students in Pakistan, Int. J. nursing students: A cross-sectional survey, Int. J. Environ.
Environ. Res. Public Health. 18 (8) (2021) 4009. Res. Public Health. 19 (7) (2022) 3766.
[30] S. Kim, J. Jeon, Factors influencing eHealth literacy among [48] J. Zhou, C. Wang, Improving cancer survivors’ e-health
Korean nursing students: A cross-sectional study, Nurs. literacy via online health communities (OHCs): a social
Health Sci. 22 (3) (2020) 667–674. support perspective, J. Cancer Surviv. 14 (2) (2020)
[31] O.E.K. Lee, D.H. Kim, K.A. Beum, Factors affecting 244–252.
information and communication technology use and [49] R.W. Berkowsky, Exploring predictors of eHealth literacy
eHealth literacy among older adults in the US and South among older adults: Findings from the 2020 CALSPEAKS
Korea, Educ. Gerontol. 46 (9) (2020) 575–586. survey, Gerontol. Geriatr. Med. 7 (2021),
[32] L.H. Nguyen, T.B.T. Le, E-Health literacy of medical students 23337214211064228.
at a university in Central Vietnam, Indian J. Public Health [50] E. Morton, K. Ho, S.J. Barnes, E.E. Michalak, Digital health
Res. Develop. 11 (2) (2020) 1299–1304. literacy in bipolar disorder: International web-based
[33] D.K.K. Wong, M.K. Cheung, Online health information survey, JMIR Ment. Health. 8 (10) (2021) e29764.
seeking and ehealth literacy among patients attending a [51] M.L. Stellefson, J.J. Shuster, B.H. Chaney, S.R. Paige, J.M.
primary care clinic in Hong Kong: A cross-sectional survey, Alber, J.D. Chaney, et al., Web-based health information
J. Med. Internet Res. 21 (3) (2019) e10831. seeking and ehealth literacy among patients living with

8
M. Estrela et al. International Journal of Medical Informatics 177 (2023)
105124
chronic obstructive pulmonary disease (COPD), eHealth literacy, and attitudes toward computer/ Internet
HealthCommun. 33 use, J. Med. Internet. Res. 15 (5) (2013) e93.
(12) (2018) 1410–1424. [61] Z. Moon, M. Zuchowski, R. Moss-Morris, M.S. Hunter, S.
[52] A.F. Abdulai, A.H. Tiffere, F. Adam, M.M. Kabanunye, Norton, L.D. Hughes, Disparities in access to mobile
COVID-19 information- related digital literacy among devices and e-health literacy among breast cancer
online health consumers in a low-income country, Int. J. survivors, Support Care Cancer 30 (1) (2022) 117–126.
Med. Inform. 145 (2021), 104322. [62] S. Xesfingi, A. Vozikis, eHealth literacy: In the quest of the
[53] C. Knapp, V. Madden, H. Wang, P. Sloyer, E. Shenkman, contributing factors, Interact. J. Med. Res. 5 (2) (2016) e16.
Internet use and eHealth literacy of low-income parents [63] C. Vlassoff, Gender differences in determinants and
whose children have special health care needs, J. Med. consequences of health and illness, J. Health Popul. Nutr.
Internet Res. 13 (3) (2011) e75. 25 (1) (2007) 47–61.
[54] K. Maroney, L.M. Curtis, L. Opsasnick, K.D. Smith, M.R. [64] T. Jansen, J. Rademakers, G. Waverijn, R. Verheij, R.
Eifler, A. Moore, et al., eHealth literacy and web-based Osborne, M. Heijmans, The role of health literacy in
patient portal usage among kidney and liver transplant explaining the association between educational
recipients, Clin. Transplant. 35 (2) (2021) e14184. attainment and the use of out-of-hours primary care
[55] B. Tennant, M. Stellefson, V. Dodd, B. Chaney, D. Chaney, S. services in chronically ill people: a survey study, BMC
Paige, et al., eHealth literacy and Web 2.0 health Health Serv. Res. 18 (1) (2018) 394.
information seeking behaviors among baby boomers and [65] K.A. Holt, D. Overgaard, L.V. Engel, L. Kayser, Health
older adults, J. Med. Internet Res. 17 (3) (2015), e70. literacy, digital literacy and eHealth literacy in Danish
[56] K.K. De Santis, T. Jahnel, E. Sina, J. Wienert, H. Zeeb, nursing students at entry and graduate level: a cross
Digitization and health in Germany: Cross-sectional sectional study, BMC Nurs. 19 (2020) 22.
nationwide survey, JMIR Public Health Surveill. 7 (11) [66] M.K. Nakphong, M.E. De Trinidad Young, B. Morales, I.Y.
(2021) e32951. Guzman-Ruiz, L. Chen, K.G. Kietzman, Social exclusion at
[57] M.R. Vicente, G. Madden, Assessing eHealth skills across the intersections of immigration, employment, and
Europeans, Health Policy Technol. 6 (2) (2017) 161–168. healthcare policy: A qualitative study of Mexican and
[58] Z. Guo, S.Z. Zhao, N. Guo, Y. Wu, X. Weng, J.Y.H. Wong, et Chinese immigrants in California, Soc. Sci. Med. 1 (298)
al., Socioeconomic disparities in ehealth literacy and (2022), 114833.
preventive behaviors during the COVID-19 pandemic in [67] L. Jiyeon, L. Eun-Hyun, C. Duckhee, eHealth literacy
Hong Kong: Cross-sectional study, J. Med. Internet Res. 23 instruments: systematic review of measurement
(4) (2021) e24577. properties, J. Med. Internet Res. 23 (11) (2021).
[59] E. Yang, S.J. Chang, H. Ryu, H.J. Kim, S.J. Jang, Comparing [68] J. Burzynska, M. ´ Rękas, P. Januszewicz, Evaluating the
factors associated with ehealth literacy between young psychometric properties of the eHealth literacy scale
and older adults, J. Gerontol. Nurs. 46 (8) (2020) (eHEALS) among Polish social media users, Int. J.
46–56. Environ. Res. Public Health 19 (7) (2022) 4067.
[60] N.G. Choi, D.M. Dinitto, The digital divide among low- [69] R. van der Vaart, C. Drossaert, Development of the digital
income homebound older adults: Internet use patterns, health literacy instrument: Measuring a broad spectrum of
health 1.0 and health 2.0 Skills, J. Med. Internet. Res. 19
(1) (2017), e27.
[70] M. Lowry, N. Trivedi, P. Boyd, A. Julian, M. Trevino, Y. Lama, et al., Making˜ [72] B.D. Weiss, M.Z. Mays, W. Martz, K.M. Castro, D.A.
DeWalt, M.P. Pignone, et al., decisions about health information on social media: A mouse-tracking study, Quick assessment of
literacy in primary care: the newest vital sign, Ann. Fam. Med.
Cognit. Res.: Principles Implicat. 7 (1) (2022) 68. 3 (6) (2005) 514–522.
[71] B.E. Canady, M. Larzo, Overconfidence in managing health concerns: The Dunning- Kruger effect and health literacy, J. Clin. Psychol.
Med. Set. (2022).