ABC | Volume 111, Nº6, December 2018

Original Article Ghisi et al Validation of the Brazilian-Portuguese CADE-Q SV Arq Bras Cardiol. 2018; 111(6):841-849 Methods Design and Procedures The design of this study consisted of a series of cross- sectional, observational studies. Data was collected between September 2017 and February 2018. First, the translation and cultural adaptation was performed. This process followed strict norms approved by the author and co‑authors and was based on the protocol proposed by Guillemin et al: 22 (1) initial translation, (2) back-translation, (3) committee review of those translations and back‑translations, and (4) pre-testing for equivalence using bilingual individuals. The initial translation was performed by an independent translator, aware of the objectives and concepts underlying the study and sought to detect ambiguities and unexpectedmeanings in the original items. The back‑translation was performed by a second translator, blinded to the initial objectives of the study and the original version. All versions were reviewed by a committee of three bilingual experts. This version was then pre-tested in 20 patients and clarity of items was checked. To assess clarity, patients were asked to rate each item on a Likert-type scale ranging from 1 (not clear) to 10 (very clear). Results were used to refine the Brazilian‑Portuguese version of CADE-Q SV. Second, a psychometric validation was performed. The refined tool was administered to a larger sample of current cardiovascular ambulatory patients from a public hospital in Belo Horizonte, Minas Gerais. The instrument was applied through monitored self-administration (i.e. researchers maintained a neutral stance during the administration, answering questions about the research and encouraging participants to answer all questions). The questionnaire was re-administered one month after the first application in 21 randomly selected participants to assess test-retest reliability. Data were collected between June and November 2017. Participants For the psychometric validation, a convenience sample of 200 ambulatory cardiovascular patients was recruited. The sample size calculation for this analysis was based on Hair &Anderson’s 23 recommendation of 10 subjects per item. Since CADE-Q SV has 20 items, a sample size of 200 is considered valid. The inclusion criteria were the following: confirmed cardiac diagnosis or multiple cardiovascular risk factors. The exclusion criteria were the following: younger than 18 years old, illiterate, any significant visual, cognitive or mental impairment which precludes the participant’s ability to answer the questionnaire. CR participants were characterized according to gender, age, educational level, family income, comorbidities, clinical risk factors, and history and duration of participation in CR. The participant’s clinical characteristics were obtained from the medical chart, and socio-demographic characteristics were self-reported. Measure: The CADE-Q SV scale CADE-Q SV assesses cardiovascular patients’ knowledge about their condition. It was designed to be a true/false/I don’t know questionnaire, with 20 items, four in each domain as follows: medical condition, risk factors, exercise, nutrition, and psychosocial risk. Each correct answer equals to one point; therefore, the maximum score possible is 20 overall, four by domain, and one per item. The tool has been developed in English and psychometrically tested inCanadianCRparticipants. 21 This tool can be used to tailor any type of educational intervention addressed to cardiovascular patients, not only in CR programs. Statistical analysis SPSS Version 24.0 was used. 24 The level of significance for all tests was set at 0.05. Psychometric properties were tested as per the Consensus-based Standards for the selection of health Measurement Instruments (COSMIN) taxonomy. 25 First, internal consistency was assessed by the Kuder-Richardson-20 (KR-20) overall, and by Cronbach's alpha of each factor (based on factor structure, described below). For this analysis, values equal to, or higher than 0.70 were considered acceptable, 23 reflecting the internal correlation between items of the same area. Second, factor structurewas assessed using confirmatory factor analysis. The main component method for factor extraction was used with consideration being given only to those with eigen values > 1.0. After the selection of the factors, a correlation matrix was generated, whereby the associations between items and factors were observed through factor loadings greater than 0.40 on only one factor. 23 The varimax method with Kaiser normalization was used to interpret the matrix. 26 Third, test-retest reliability was assessed using intraclass correlation coefficient (ICC). ICC values lower than 0.70 27 were considered bad items. Finally, criterion validity was also assessed by comparing CADE-Q SV total scores with the participant’s level of education, family monthly income and time of diagnosis, using independent sample t-tests and Pearson’s correlation. Item completion rates were also described. A descriptive analysis of the Portuguese CADE-Q SV was performed. A mean total score was computed to reflect total knowledge. Independent sample t-tests, one‑way analysis of variance, and chi-square tests were used as appropriate to assess differences in total knowledge based on patient’s socio‑demographic and clinical characteristics. Continuous variables were all normally distributed (confirmed by Kolmogorov‑Smirnov test) and were reported with mean and standard deviations. Categorical variables were reported by absolute numbers, percentages and, when applicable, confidence intervals. Results Participants’ characteristics The characteristics of participants from the psychometric validation are described in Table 1. Overall, 200 cardiovascular ambulatory patients completed the Portuguese version of CADE-Q SV, of which 118 (59.0%) were male, and the mean age was 58.4 ± 11.6 years old. Translation, cultural adaptation and pre-testing During the process of translation and cultural adaptation, it was observed that one item needed to be adapted to be 842