Reliability and validity of the Turkish version of the Rapid Estimate of Adult Literacy in Dentistry (TREALD-30)

• Published in: Acta odontologica Scandinavica Vol. 75; no. 3; p. 198
• Main Authors: Peker, Kadriye, Köse, Taha Emre, Güray, Beliz, Uysal, Ömer, Erdem, Tamer Lütfi
• Format: Journal Article
• Language: English
• Published: England 01.04.2017
• Subjects: Adult; Dentistry; Female; Health Education, Dental; Health Knowledge, Attitudes, Practice; Health Literacy; Humans; Male; Middle Aged; Oral Health; Psychometrics; Quality of Life; Reproducibility of Results; Surveys and Questionnaires - standards; Turkey; REALD-30; Turkish dental patients; psychometric evaluation; Oral health literacy; literacy in dentistry
• ISSN: 1502-3850
• DOI: 10.1080/00016357.2016.1278079

To culturally adapt the Turkish version of Rapid Estimate of Adult Literacy in Dentistry (TREALD-30) for Turkish-speaking adult dental patients and to evaluate its psychometric properties. After translation and cross-cultural adaptation, TREALD-30 was tested in a sample of 127 adult patients who attended a dental school clinic in Istanbul. Data were collected through clinical examinations and self-completed questionnaires, including TREALD-30, the Oral Health Impact Profile (OHIP), the Rapid Estimate of Adult Literacy in Medicine (REALM), two health literacy screening questions, and socio-behavioral characteristics. Psychometric properties were examined using Classical Test Theory (CTT) and Rasch analysis. Internal consistency (Cronbach's Alpha = 0.91) and test-retest reliability (Intraclass correlation coefficient = 0.99) were satisfactory for TREALD-30. It exhibited good convergent and predictive validity. Monthly family income, years of education, dental flossing, health literacy, and health literacy skills were found as stronger predictors of patients'oral health literacy (OHL). Confirmatory factor analysis (CFA) confirmed a two-factor model. The Rasch model explained 37.9% of the total variance in this dataset. In addition, TREALD-30 had eleven misfitting items, which indicated evidence of multidimensionality. The reliability indeces provided in Rasch analysis (person separation reliability = 0.91 and expected-a-posteriori/plausible reliability = 0.94) indicated that TREALD-30 had acceptable reliability. TREALD-30 showed satisfactory psychometric properties. It may be used to identify patients with low OHL. Socio-demographic factors, oral health behaviors and health literacy skills should be taken into account when planning future studies to assess the OHL in both clinical and community settings.