Tradução para o português e análise das propriedades psicométricas do instrumento DISCERN para avaliação da qualidade de texto sobre saúde para leigos Tese
Data
2018-07-26
Autores
Logullo, Patricia Julia 
Orientadores
Riera, Rachel
Tipo
Tese de doutorado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Objetivo: Produzir a tradução, para a língua portuguesa, do DISCERN, adaptá-la
culturalmente e avaliar suas propriedades psicométricas como instrumento para
avaliação da qualidade de textos em português sobre saúde; avaliar o alfabetismo em
saúde das pessoas que usam o DISCERN traduzido para o português brasileiro,
utilizando o Short Assessment of Health Literacy for Portuguese-Speaking Adults
(SAHL-PA), e estudar a correlação entre os escores produzidos pelos dois
instrumentos. Métodos: Três tradutores, trabalhando de forma independente,
traduziram as 16 perguntas e módulos de texto adicionais de orientação ao usuário
do questionário DISCERN. As traduções foram submetidas a um comitê de
especialistas, que, por meio de votação em ambiente eletrônico, chegou a uma versão
em português. Esta foi ajustada e traduzida de volta para o inglês por outros dois
tradutores. As versões foram comparadas quanto à semântica e a versão em
português foi novamente ajustada. O DISCERN traduzido foi então aplicado a
estudantes de jornalismo de primeiro ano, que o utilizaram para avaliar um texto sobre
tratamentos para cessação do tabagismo. Esses estudantes também foram
submetidos a avaliação de seu alfabetismo em saúde por meio do SAHL-PA. Quatro
semanas após a primeira aplicação, os participantes responderam ao DISCERN uma
segunda vez, avaliando o mesmo texto. A consistência interna do instrumento foi
avaliada utilizando-se o alfa de Cronbach e o coeficiente de correlação intraclasse
(CCI) foi utilizado para verificar sua reprodutibilidade. A correlação entre os escores
do DISCERN e do SAHL-PA foi investigada utilizando-se o teste de Spearman. A
associação entre o escore do DISCERN e o sexo dos participantes foi investigada por
meio do teste t de Student, e a correlação entre o escore e a idade, pelo de Pearson.
Resultados: A distribuição dos votos dos membros do comitê foi heterogênea e exigiu
mais de uma rodada de votação e ajustes para se atingir uma tradução adequada do
DISCERN. A versão final foi aplicada para três turmas de faculdades de jornalismo (n
= 126). Não houve dificuldade que impedisse as respostas do questionário. A versão
traduzida do DISCERN teve boa consistência interna, com alfa de Cronbach alto, de
0,865 (intervalo de confiança de 95%, IC, de 0,826 a 0,898), e o CCI entre as duas
avaliações foi de 0,845 (IC de 0,717 a 0,912). O alfabetismo em saúde dos 98
participantes que responderam ao questionário SAHL-PA foi considerado adequado.
Não houve correlação entre os escores do SAHL-PA e do DISCERN (p > 0,05), ou
seja, a avaliação do texto não se mostrou estatisticamente correlacionada com o alfabetismo em saúde. A idade ou o sexo não teve qualquer relação com o escore
médio do DISCERN. Conclusões: A tradução do questionário DISCERN para a língua
portuguesa foi culturalmente adaptada para uso no Brasil. O DISCERN em português
é capaz de medir o fenômeno “qualidade de textos na área de saúde” de maneira
adequada, com boa convergência entre as questões, boa consistência interna e boa
reprodutibilidade. O instrumento está à disposição de qualquer cidadão, profissional
de saúde ou instituição de saúde no Brasil que deseje, na prática, obter uma avaliação
da qualidade de textos sobre saúde disponíveis em qualquer meio.
Objective: To translate the DISCERN tool into Portuguese, to adapt it culturally and to evaluate its psychometric properties as an instrument to evaluate the quality of texts in Portuguese about health issues; to assess the health literacy of people using DISCERN translated into Brazilian Portuguese using the Short Assessment of Health Literacy for Portuguese Speaking Adults (SAHL-PA) and to study the correlation between the scores produced by the two instruments. Methods: Three translators, working independently, translated the 16 questions and additional text modules for user guidance from the DISCERN questionnaire. The translations were submitted to a committee of experts, who, through electronic voting, reached a Portuguese version. This version was adjusted and translated back to English by two other translators. The translations were compared regarding semantics, and the Portuguese version was adjusted again. The final version of the DISCERN in Portuguese was then tested among first-year journalism students who used it to evaluate a text on smoking cessation treatments. These students also underwent assessment of their health literacy through SAHL-PA. Four weeks later, participants used DISCERN a second time, evaluating the same text. The responses of both questionnaires were tabulated and analyzed using the Cronbach’s alpha (to assess the internal consistency) and the intraclass correlation coefficient (ICC, to compare the two occasions of DISCERN's response). The correlation between the DISCERN and SAHL-PA scores was examined using the Spearman test. The association between the DISCERN score and the gender of the participants was investigated using Student's t-test, and the correlation between score and age, by Pearson’s. Results: The distribution of committee members’ votes was heterogeneous and required more than one round of voting and adjustments to achieve a proper translation of DISCERN. The final version was applied to three groups of journalism students (n = 126). The participants had no difficulty in using the questionnaire. The internal consistency analysis revealed a Cronbach’s alpha of 0.865 (95% confidence interval, CI, of 0,826 to 0,898), and the ICC between the two evaluations was 0.845 (CI: 0,717 to 0,912). The health literacy of the 98 participants who answered the SAHL-PA questionnaire was adequate. There was no correlation between the SAHL-PA and DISCERN scores (p > 0.05), that is, the text evaluation was not statistically correlated with health literacy. Age or gender had no relation to the DISCERN mean score. Conclusions: The translation of the DISCERN questionnaire into the Portuguese language was culturally adapted for the use in Brazil. The DISCERN version in Portuguese is capable of measuring the phenomenon of “quality of texts in the health area” adequately, with good convergence between the questions, good internal consistency and good reproducibility. The instrument is available to any citizen, health professional or health institution in Brazil that aims to evaluate the quality of texts about health available in any medium.
Objective: To translate the DISCERN tool into Portuguese, to adapt it culturally and to evaluate its psychometric properties as an instrument to evaluate the quality of texts in Portuguese about health issues; to assess the health literacy of people using DISCERN translated into Brazilian Portuguese using the Short Assessment of Health Literacy for Portuguese Speaking Adults (SAHL-PA) and to study the correlation between the scores produced by the two instruments. Methods: Three translators, working independently, translated the 16 questions and additional text modules for user guidance from the DISCERN questionnaire. The translations were submitted to a committee of experts, who, through electronic voting, reached a Portuguese version. This version was adjusted and translated back to English by two other translators. The translations were compared regarding semantics, and the Portuguese version was adjusted again. The final version of the DISCERN in Portuguese was then tested among first-year journalism students who used it to evaluate a text on smoking cessation treatments. These students also underwent assessment of their health literacy through SAHL-PA. Four weeks later, participants used DISCERN a second time, evaluating the same text. The responses of both questionnaires were tabulated and analyzed using the Cronbach’s alpha (to assess the internal consistency) and the intraclass correlation coefficient (ICC, to compare the two occasions of DISCERN's response). The correlation between the DISCERN and SAHL-PA scores was examined using the Spearman test. The association between the DISCERN score and the gender of the participants was investigated using Student's t-test, and the correlation between score and age, by Pearson’s. Results: The distribution of committee members’ votes was heterogeneous and required more than one round of voting and adjustments to achieve a proper translation of DISCERN. The final version was applied to three groups of journalism students (n = 126). The participants had no difficulty in using the questionnaire. The internal consistency analysis revealed a Cronbach’s alpha of 0.865 (95% confidence interval, CI, of 0,826 to 0,898), and the ICC between the two evaluations was 0.845 (CI: 0,717 to 0,912). The health literacy of the 98 participants who answered the SAHL-PA questionnaire was adequate. There was no correlation between the SAHL-PA and DISCERN scores (p > 0.05), that is, the text evaluation was not statistically correlated with health literacy. Age or gender had no relation to the DISCERN mean score. Conclusions: The translation of the DISCERN questionnaire into the Portuguese language was culturally adapted for the use in Brazil. The DISCERN version in Portuguese is capable of measuring the phenomenon of “quality of texts in the health area” adequately, with good convergence between the questions, good internal consistency and good reproducibility. The instrument is available to any citizen, health professional or health institution in Brazil that aims to evaluate the quality of texts about health available in any medium.