ABC | Volume 112, Nº1, January 2019

Original Article Santos et al Validation of CADE-Q II in portuguese Arq Bras Cardiol. 2019; 112(1):78-84 changed over the years, 11,16 and a second version of the CADE (CADE-QII) was developed aiming at updating the instrument and including additional educational components, such as psychosocial health. 17 The aim of the present study was to translate, cross‑culturally adapt and psychometrically validate a Brazilian Portuguese version of the CADE-QII. Methods Study design and procedures This study was approved by the human research ethics committee of the University Universidade Federal de Minas Gerais (approval number 1.350.973), according to the 466/12 resolution of the Brazilian National Health Council. This was an observational, cross-sectional, multicenter study, involving research centers in the cities of Belo Horizonte and Florianopolis in Brazil. Data were collected from January to May 2016. First, the processes of translation and cross-cultural adaptation of the instrument were performed according to precise criteria previously approved by the authors, based on the protocol proposed by Guillemin et al. 18 – (1) initial translation, (2) back translation, (3) review of the questionnaire versions built by specialist committee members, (4) pre-testing for equivalence with bilingual individuals, and (5) revision of the weighting of scores. The translated, cross-culturally adapted version of the instrument was then tested for clarity in coronary patients. The results were used to refine the CADE‑QII version in Portuguese. Second, psychometric validation was conducted. The refined instrument was administered to a larger sample of patients, participants of three CR programs in the metropolitan area of Florianopolis, Brazil (one private and two public programs) and of one public program in Belo Horizonte, Brazil. The instrument was administered by interview. The interviewers maintained a neutral position, answering questions about the study and encouraging respondents to answer all the questions. The questionnaire was readministered two weeks following the first administration in patients selected by convenience for test-retest analysis. Participants For psychometric validation, we recruited participants of CR programs developed in four participating institutions, where a total of 500 patients attended every month. A convenience sample was recruited from these patients, based on availability of recruiters and patients that accepted to participate in the study. The sample was composed of 307 patients, corresponding to 61.4% of the patients attending the participating institutions. The programs lasted for at least three month and none of them had a structured educational component. Inclusion criteria were a confirmed diagnosis of CAD or the presence of cardiovascular risk factors and signing of the consent form. Exclusion criteria were age younger than 18 years and any significant visual, cognitive or mental impairments that could limit patients’ ability to answer the questionnaire. CADE-QII CADE-QII was constructed to verify the level of knowledge of coronary patients, participants of CR programs about CAD. CADE-QII evaluates the knowledge of patients for five domains consisting of medical condition, risk factors, exercise, nutrition and psychosocial risk, based on official documents and guidelines in the area. 17 The instrument has 31 items, each item with four possible answers about one of these domains of knowledge. One of the answers is the most “correct” one, i.e. , the answer states complete and true information about the domain, and gets a score of 3; one of the answers is “partially correct” and gets a score of 1, and the two other options are “incorrect”, one describing incorrect information and “I don’t know”, which should be chosen when the patient is not sure about any of the previous options. Both incorrect and “I don’t’ know” options get a “0” score. The total sum is calculated, and it represents the level of patients’ knowledge about the domains. 17 Thus, CADE-QII not only quantifies the level of knowledge about cardiovascular disease but also identify the domains where patients get the lowest scores, i.e ., in which knowledge is deficient. Variables The following characteristics of the patients participating in the psychometric validation were collected for analysis – sex, age, educational level, monthly family income, comorbidities, cardiac risk factors and previous diseases. All these characteristics were self-reported. The Brazilian version of the CADE-QII were tested for the following psychometric properties – clarity, content validity, test-retest reliability, internal consistency and criterion validity. A descriptive analysis of total score was also performed, both per question and per domain. Statistical analysis The Statistical Package for Social Sciences (SPSS), version 20.0 was used for data storage, classification and analysis. The level of significance was set at 5%. Data were excluded from analysis when more than 20% of the CADE-QII items were incomplete. To test clarity and validity of the content, a pretest was conducted with the patients, in order to get a feedback on the items and to verify the time required to complete the questionnaire. In addition, five specialists evaluated the clarity of the CADE-QII version in Brazilian Portuguese. The intraclass correlation coefficient (ICC) was used in the test retest reliability analysis. Values with ICC lower than 0.7 were excluded from analysis. 19 The internal consistency of the instrument was assessed by Cronbach alpha coefficient; values above 0.70 were considered acceptable. 20 Criterion validity was assessed by comparing CADE-QII scores with family income and educational level. 17 Data normality was tested using the Kolmogorov-Smirnov test. Descriptive analysis of the Brazilian version of the CADE- QII was also performed; continuous variables with normal distribution were described as mean and standard deviation, 79