Semantics bias in cross-national comparative analyses : Is it good or bad to have "fair" health?

The Health Behavior in School-aged Children is a cross-national study collecting data on social and health indicators on adolescents in 43 countries. The study provides comparable data on health behaviors and health outcomes through the use of a common protocol, which have been a back bone of the study sine its initiation in 1983. Recent years, researchers within the study have noticed a questionable comparability on the widely used item on self-rated health. One of the four response categories to the item "Would you say your health is….?" showed particular variation, as the response category "Fair" varied from 20 % in Latvia and Moldova to 3-4 % in Bulgaria and Macedonia. A qualitative mini-survey of the back-translations showed that the response category "Fair" had a negative slant in 25 countries, a positive slant in 10 countries and was considered neutral in 9 countries. This finding indicates that there are what may be called semantic issues affecting comparability in international studies, since the same original word (in an English original) is interpreted differently across countries and cultures. The paper test and discuss a few possible explanations to this, however, only leaving to future studies to hold a cautious approach to international comparisons if working with the self-rated health item with four response categories.

Impact of oral health on health-related quality of life : A cross-sectional study

Background: Despite the consensus regarding the existence of a relationship between “impacts on oral health” and “health-related quality of life”, this relationship, considering the latent nature of these variables, is still poorly investigated. Thus, we performed this study in order to determine the magnitude of the impacts of oral health, demographic and symptom/clinical variables on the health-related quality of life in a Brazilian sample of dental patients. Methods: A total of 1,007 adult subjects enrolled in the School of Dentistry of São Paulo State University (UNESP) - Araraquara Campus for dentistry care between September/2012 and April/2013, participated. 72.4 % were female. The mean age was 45.7 (SD = 12.5) years. The Oral Health Impact Profile (OHIP-14) and the Short Form Health Survey (SF-36) were used. The demographic and symptom/clinical variables collected were gender, age, economic status, presence of pain and chronic disease. The impact of studied variables on health-related quality of life were evaluated with a structural equation model, considering the factor “Health” as the central construct. The fit of the model was first analyzed by the evaluation of the goodness of fit indices (χ 2 /df ≤ 2.0, CFI and TLI ≥ 0.90 and RMSEA < 0.10) and the evaluation of the variables’ impact over health-related quality of life was based on the statistical significance of causal paths (β), evaluated by z tests, for a significance level of 5 %. Results: We observed adequate fit of the model to the data (χ 2 /df = 3.55; CFI = 0.95; TLI = 0.94; RMSEA = 0.05). The impacts on oral health explained 28.0 % of the variability of the health-related quality of life construct, while the total variance explained of the model was 39.0 %. For the demographic and symptom/clinical variables, only age, presence of pain and chronic disease showed significant impacts (p < 0.05). Conclusion: The oral health, age, presence of pain and chronic disease of individuals had significant influence on health-related quality of life.