Network analysis and psychometric properties of the Brazilian version of the eHealth Literacy Scale in a dental clinic setting

Abstract

Background: The eHealth Literacy Scale (eHEALS) has become a standard tool for the evaluation of digital health literacy.

Objectives: This study aims to examine the validity and reliability of the Brazilian version of the eHEALS in a sample of patients from a dental clinic setting.

Material and methods: A cross-sectional survey was conducted in a dental setting. The sample included patients from the Dental Clinic of the Dentistry School of the University of São Paulo between September 2019 and March 2020. The underlying structure of the data was exploratorily investigated using Exploratory Graph Analysis, a network analysis method. The internal reliability was assessed using the McDonald's Omega coefficient. Confirmatory Factor Analysis (CFA) was used to assess the fit of the structure identified by the network analysis. Convergent validity was assessed using the Average Variance Extracted, and measurement invariance was examined using nested models in multigroup CFA. Criterion-related validity was examined calculating the latent mean differences between subgroups (genders, age groups, and educational levels).

Results: The sample included 132 adults aged 18 to 82 years (mean 44.7 years). The eHEALS network indicated that items form a single-factor structure. The 1-factor model presented adequate fit (χ2 (18) = 29.873, p < 0.039; R-CFI = 0.997; R-TLI = 0.996; R-RMSEA = 0.032, 90% CI [0.000, 0.052]), good internal reliability, and convergent validity. Configural invariance was found for genders, educational levels, and age groups. Scalar invariance was observed for genders and age, whereas partial scalar invariance was confirmed for education. Participants aged 18 to 45 and those with higher education presented greater latent means for eHEALS subscales. There were no differences between genders.

Conclusion: The BR-eHEALS presented good internal reliability, convergent validity, measurement invariance, and was able to discriminate the levels of eHealth literacy among groups with different ages and educational levels. These findings demonstrate that the tool is valid and reliable for use in a dental setting with the Brazilian population.

Keywords: Consumer health informatics; Digital literacy; Health information; Health literacy; Measurement invariance; Network analysis; eHealth literacy.