Adaptação transcultural e avaliação das propriedades psicométricas de dois instrumentos de letramento digital em saúde para adolescentes brasileiros: eHealth Literacy Scale e Digital Health Literacy Instrument

Abstract

With the expansion of internet access and digitization of health services, adolescent users must have a set of skills for their proper use, called Digital Health Literacy (DHL). The objective of this study was to cross-culturally adapt and evaluate the psychometric properties of two DHL instruments, eHealth Literacy Scale (eHEALS) and Digital Health Literacy Instrument (DHLI), for Brazilian adolescents. The eHEALS instrument is an eight-item measure that assesses DHL self-perception. The DHLI instrument consists of 21 self-report items and seven skill items that assess the individual’s practical performance. The study was approved by the Research Ethics Committee of the Federal University of Minas Gerais. The sample consisted of individuals aged 13 to 19 years, enrolled in public schools in Belo Horizonte, MG, Brazil. The eHEALS and DHLI instruments were previously translated and adapted for Brazilian adults. We used the Brazilian versions of the instruments and the following methodological steps were performed: expert committee; back-translation; synthesis of back-translations; evaluation by original authors; pre-test with 42 adolescents using cognitive interviews with probing questions and production of the final Portuguese version directed at adolescents. A total of 260 adolescents participated in the testing phase and answered DHL instruments, eHEALS and DHLI directed at Brazilian adolescents (eHEALS-BrA and DHLI-BrA), health literacy instruments, Newest Vital Sign (NVS-BR) and Rapid Estimate Adult Literacy in Medicine and Dentistry (REALMD-20), a quality-of-life instrument, World Health Organization Quality of Life-Bref (WHOQOL-bref), and a sociodemographic questionnaire, internet access and use of eHealth. In addition, after 15 days, 67 adolescents responded again to eHEALS-BrA and DHLI-BrA. The instrument’s reliability was tested through internal consistency (Cronbach’s alpha and McDonald’s omega) and test-retest reliability. Confirmatory factor analyses (CFA) evaluated the instrument’s dimensionality and the validity was tested by associations between the eHEALS-BrA and DHLI-BrA scores and variables of interest (p<0.05). eHEALS-BrA demonstrated good internal reliability (α and ω=0.71) and good stability (ICC=0.82, 95% CI:0.70-0.89). Preliminary tests showed that the data were adequate for principal component analysis: KMO=0.78 and Barlett’s sphericity test (p<0.001). The best-fitting model in CFA consisted of a single factor (χ²=49.884 (p=0.0002), CFI=0.934, TLI=0.908, RMSEA=0.076, and RMSR=0.045). The self-report items of DHLI-BrA demonstrated satisfactory internal reliability (α and ω = 0.83) and good stability (ICC=0.906 (IC 95%: 0.75-0.95). In CFA, the best-fitting model consisted of six factors (χ²=229.173 (p=0.0032), CFI=0.944, TLI=0.933, RMSEA=0.035, and RMSR=0.047). Performance-based DHLI-BrA demonstrated satisfactory internal reliability (α=0.57e ω =0.58) and good stability (ICC=0.86; IC 95%: 0.76-0.92). In CFA the best-fitting model consisted of a single factor (χ²=17,901 (p=0,2113), CFI= 0,952; TLI= 0,927; RMSEA=0,033; RMSR=0,038). It is concluded that the versions of eHEALS-BrA and DHLI-BrA instruments produced demonstrated good psychometric properties for measuring DHL in Brazilian adolescents.