To Örebro University

BACKGROUND: Health Literacy is a crucial factor for health. In Europe, many people have limited health literacy (i.e. difficulties with accessing, understanding, appraising and using health information). This study aimed to evaluate the psychometrics of the Swedish versions of the HLS-EU-Q16 and HLS-EU-Q6, instruments that aims to assess health literacy.

METHODS: In this prospective psychometric study convenience sampling was used, which gave a study population of 347 Swedish-speaking adults. The psychometric evaluation included item distributional statistics, construct validity testing, and principal component analysis to assess structural validity. Internal consistency and test-retest reliability was also investigated.

RESULTS: For the Swedish version of HLS-EU-Q16, no floor effects were detected but a ceiling effect was noted among 28% of the respondents. Construct validity was supported as four out of five expected correlations was confirmed (educational level, self-perceived health, electronic health literacy and HLS-EU-Q6). In terms of structural validity, the principal component analysis yielded a four-factor structure with most items loading significantly only to one factor. The Swedish version of HLS-EU-Q16 had acceptable internal consistency (Cronbach's α = 0.89, split-half reliability = 0.93) and test-retest reliability showed stability over time (Cohen's κ = 0.822). For the Swedish version of HLS-EU-Q6, neither floor nor ceiling effects were observed. Construct validity was supported as HLS-EU-Q6 correlated as our a priori stated hypothesis. The principal component analysis did not support the unidimensionality of the scale as a two-factor structure was identified. The Swedish version of HLS-EU-Q6 had acceptable internal consistency (Cronbach's α = 0.77, split-half reliability = 0.80) and test-retest reliability showed stability over time (Cohen's κ = 0.812). According to the Swedish version of the HLS-EU-Q16, 71% of the participants were classified as having sufficient comprehensive health knowledge (CHL), while only 33% were classified as having this when the HLS-EU-Q6 was used.

CONCLUSIONS: The Swedish versions of the HLS-EU-Q16 and HLS-EU-Q6 have acceptable psychometric properties, and based on the results we recommend its use to measure CHL. However, we are hesitant to use Sw-HLS-EU-Q6 in estimating different CHL levels and further studies need to be conducted to establish validity and accuracy of the thresholds of HLS-EU-Q6.