A 3-Item Measure of Digital Health Care Literacy: Development and Validation Study

Abstract

Background With increased reliance on digital health care, including telehealth, efficient and effective ways are needed to assess patients’ comfort and confidence with using these services. Objective The goal of this study was to develop and validate a brief scale that assesses digital health care literacy. Methods We first developed an item pool using existing literature and expert review. We then administered the items to participants as part of a larger study. Participants were caregivers of children receiving care at a pediatric clinic who completed a survey either on the web or over the telephone. We randomized participants into development and confirmatory samples, stratifying by language so that exploratory factor analysis and confirmatory factor analysis could be performed with separate samples of participants. We assessed the scale’s validity by examining its associations with participants’ demographics, digital access, and prior digital health care use. Results Participants (N=508) were, on average, aged 34.7 (SD 7.7) years, and 89.4% (454/508) were women. Of the 508 participants, 280 (55.1%) preferred English as their primary language, 157 (30.9%) preferred Spanish, and 71 (14%) preferred Arabic; 228 (45%) had a high school degree or less; and 230 (45.3%) had an annual household income of <US $35,000. Using exploratory factor analysis, 3 items were retained in a reduced scale with excellent reliability (Cronbach α=.90) and a high variance explained (78%). The reduced scale had excellent fit, with factor loadings between 0.82 and 0.94. All fit statistics exceeded the criteria for good fit between the proposed factor structure and the data. We refer to this scale as the Digital Health Care Literacy Scale. The scale was positively associated with education (ρ=0.139; P=.005) and income (ρ=0.379; P<.001). Arabic speakers had lower scores than English (P<.001) and Spanish speakers (P=.02), and Spanish speakers had lower scores than English speakers (P<.001). Participants who did not own a smartphone (P=.13) or laptop computer (P<.001) had lower scores than those who owned these devices. Finally, participants who had not used digital tools, including health apps (P<.001) and video telehealth (P<.001), had lower scores than those who had used these tools. Conclusions Despite the potential for digital health care to improve quality of life and clinical outcomes, many individuals may not have the skills to engage with and benefit from it. Moreover, these individuals may be those who already experience worse outcomes. A screening tool such as the Digital Health Care Literacy Scale could be a useful resource to identify patients who require additional assistance to use digital health services and help ensure health equity.

The role of digital literacy in achieving health equity in the third millennium society: A literature review

“…In fact, not always those with high incomes (and therefore likely to have access to high-level technological tools) possess a sufficient degree of digital literacy to make beneficial use of e-Health services ( 42 ). It is therefore essential to develop screening tools that can accurately identify the population groups in need of digital literacy interventions ( 24 , 41 ). These tools already exist, but, as the literature consulted shows, they need to be considerably implemented.…”

Nurse-to-family telehealth for pediatric transfers: protocol for a feasibility and pilot cluster randomized controlled trial

“…Electronic health record data will be used to abstract patient demographics (age, race, ethnicity, sex, insurance, California Healthy Places Index [ 35 ]) and utilization variables (admitting service, length of stay, disposition). Demographic characteristics of family caregivers (age, race, ethnicity, gender, education, smart phone access, digital literacy score [ 36 ]) will be collected in the survey packets.…”

Virtual Family-Centered Rounds in the Neonatal Intensive Care Unit: Protocol for a Cluster Randomized Controlled Trial

“…We will use generalized linear models to estimate intervention effect sizes and con dence intervals and to test hypotheses for outcomes. Independent variables will include a binary indicator for intervention assignment along with a parsimonious set of subject characteristics as covariates (parent age, race, ethnicity, gender, education, transportation security, marital status, other children dependents, digital literacy score 55 ). We will specify the generalized linear mixed models as linear, logistic, or Poisson regressions, according to the outcome type.…”

“…Demographic characteristics of parents (age, race, ethnicity, gender, relationship to the infant, education, employment, transportation security, marital status, housing, other children dependents, computer/smart device access, internet access, and digital literacy score 55 ) will be collected in the survey packets.…”