Study setting and participants
A cross-sectional survey was conducted and reported, following the statement Strengthening the Reporting of Observational Studies in Epidemiology23. A convenience sample was recruited for the study from the Health Control Center and Department of Endocrinology of the Second Affiliated Hospital of Zhejiang University School of Medicine in Hangzhou, Zhejiang Province, China. A paper survey and an online survey platform powered by WJX (www.wjx.com) between April and May 2022 were conducted. Inclusion criteria were a citizen of Zhejiang, over the age of 18, not undergoing treatment for a psychiatric disorder and no communication barriers. Before the evaluation process, the participants were informed about the subject of the study and gave their informed consent. Sample size was based on subject to recommended item of (5-10):1, for which 130-260 participants were needed. Assuming a missing data rate of 10% and an adequate sample size for confirmatory factor analysis (CFA)24318 participants were included.
Chinese version of the ABCD Risk Perception Questionnaire
Permission has been obtained from the author to translate the English version into Chinese. We conducted this procedure following the Intercultural Adaptation Guidelines of Self-Assessment Measures25.26 and consensus standards for the selection of health measurement instruments27.
First, two native Mandarin Chinese translators who were fluent in English independently translated the ABCD English Risk Perception Questionnaire into Mandarin Chinese. Then the two Chinese versions were compared to the original scale by a third translator. The research team discussed any discrepancies and formed a consensus version in Mandarin. This consensus Mandarin version was given to two bilingual translators who did not know the original English version and who re-translated the Mandarin version into English. The back-translated versions of the ABCD Risk Perception Questionnaire were compared to the original version, and the translators clarified some ambiguities and inconsistencies.
Intercultural adaptation process
A panel of experts, including a cardiologist, two specialist nurses from a cardiology department, a health management specialist, a methodologist and a language professional, reviewed all translation reports with translators in terms of semantic equivalence, idiomatic equivalence, experiential equivalence and conceptual equivalence. until a consensus is reached. The pre-final version proposed by these experts was sufficiently close to the original version. Finally, all the experts evaluated the relevance of each item of the ABCD questionnaire on risk perception in the Chinese context on a 4-point Likert scale (1 = very little relevant; 4 = very relevant). In addition, completeness and comprehensibility, according to the COSMIN rating system, were determined by experts, using a 3-point scale (1 = not clear, 2 = not sufficient, 3 = sufficient )27.
Pilot test of the pre-final version
The pre-final ABCD risk perception questionnaire was tested with 40 Chinese adults in accordance with the above cross-cultural adaptation guidance.25. Participants took approximately 5 minutes to complete the questionnaire. In addition, each participant was invited to comment on their understanding of the wording, particularly with regard to confusing or illegible statements. After minor revisions, the final ABCD Risk Perception Questionnaire was generated for psychometric assessment.
Final Chinese Version of the ABCD Risk Perception Questionnaire (ABCD-C)
This questionnaire is made up of four subscales. The assessment of knowledge of cardiovascular risk and prevention consists of eight statements about cardiovascular risk, indicating whether the respondent agrees or disagrees with the statements with three options (true/false/don’t know). For each item, the correct answer was scored 1, and an incorrect or “don’t know” answer was scored 0. The values are added together to create a summary score that can range from 0 to 8, for which higher values indicate superior CVD knowledge. Other dimensions include perceived risk of heart attack/stroke (8 items), perceived benefits and intention to change behavior (7 items), and healthy eating intentions (3 items). Response options are presented on a 4-point scale and range from 1 = strongly disagree to 4 = strongly agree. Items 15, 21 and 26 were reverse coded.
Demographic characteristics included age, sex, marital status, education level, religion, employment status, smoking and alcoholism, family background and related variables. Subjective health status was assessed using the question “In general, how would you rate your health status?” (1 = very poor, 2 = poor, 3 = fair, 4 = good, 5 = excellent). Smoking and drinking status was determined by answering the question, “What is your current smoking/drinking status?” (1 = never smoked/drank, 2 = ever smoked/drank, 3 = currently smokes/drinks). Also, one item estimated absolute cardiovascular risk: “What do you think is your risk of getting any type of cardiovascular disease in the next 10 years?” selecting a number from 1 to 10, with 0 = no risk and 10 = very high risk, a question commonly used in studies related to CVD risk perception16.28.
Data collection procedures
Data was collected onsite and online. Well-educated and eligible research assistants who all held a doctorate in nursing. candidates collected data in face-to-face interviews. All participants were informed of the purpose of the research and that their participation was voluntary and confidential. Questionnaires with more than three empty items in the paper survey were excluded. No items were missing for the online survey because all items were required, but a response time of less than 2 minutes or similar option choices invalidated the questionnaire, which was excluded. Additionally, the ABCD risk perception questionnaire was completed twice by twenty participants who were randomly selected from the total sample at two-week intervals to calculate test-retest reliability.
Psychometric assessments and statistical analyzes
The reliability of the ABCD-C scale was tested by internal consistency and test-retest reliability. The degree of internal consistency is described as the Cronbach degree a and McDonald’s ω. An item‐total corrected correlation coefficient was used to calculate item discrimination. Test-retest reliability was assessed by an intraclass correlation coefficient (ICC), rooted in a two-way analysis of variance in a random-effects model. 95% confidence intervals (CI) of the ICC value have also been reported. The two Cronbachs aMcDonald’s ω and ICC values greater than 0.70 are recommended, indicating that the scale has acceptable reliability29.30.
Construct validity and content validity were assessed to verify the validity of ABCD-C. The content validity of the scale was assessed by the item content validity index (I-CVI) and the medium scale content validity index (S-CVI/Ave) based on the ratings experts on this questionnaire. An S-CVI/Ave value of 0.9 is considered an excellent criterion and a value of 0.8 the lower limit of content validity for acceptance of the entire scale31I-CVI ≥ 0.78 is considered appropriate if the number of experts is ≥ 632. Maximum likelihood exploratory (EFA) and confirmatory (CFA) factor analyzes were performed to further explore item dimensionality. Standardized factor loadings and an estimate of the variance of the measured variable explained by the latent variable (R2), as well as the fitting statistics (χ2, CFI—comparative fit index, IFI—incremental fit index; TLI—Tucker-Lewis index, RMSEA—root mean square error of approximation). A CFA > 0.95, TLI > 0.90, RMSEA < 0.08 were considered acceptable33. The mean variance extracted (AVE) was applied to assess the internal convergent validity of each factor, with a score ≥ 0.5 indicating satisfactory convergent validity34. The square root of the AVE value exceeding each of its correlations with other factors indicates appropriate discriminant validity34. To assess concurrent validity, Spearman rank correlations were analyzed between the ABCD-C and subjective single-item CVD risk perception. The correlation of |r|= 0.10–0.30, |r|= 0.31–0.60, and |r|= 0.61–1.00 were considered low, medium, and high, respectively35.
Item difficulty for the knowledge subscale was determined by descriptive statistics, consistent with previous knowledge studies36. The difficulty level index was calculated as the number of correct answers divided by the total number of answers, a higher index indicates a lower level of difficulty37. The difficulty level of a test should be around 0.5037. Corrected correlations between performance between individual items and the overall test were calculated to determine each item’s ability to discriminate between high and low scoring participants. A correlation value of 0.5 to 0.7 indicates good discrimination power38.
The Statistical Package for the Social Sciences, version 26.0 (SPSS, Chicago, IL, USA), was used for statistical analysis. The data are expressed as the mean (M) and the standard deviation (SD). The mean difference in ABCD-C scores between sociodemographic variables was determined using one-way analysis of variance (ANOVA) and Tukey’s multiple comparison test. The effect size is partial squared (η2) was calculated as the sum of squares of the effect divided by the total sum of squares; η2= 0.01 indicates a small effect, η2= 0.06 with middle and η2= 0.014 with large effect36. FCA was performed using AMOS 22.0 to assess the structural validity of the scale. SPSSAU was used to perform reliability analysis. A p-a value less than or equal to 0.05 was considered statistically significant.
The study protocol had been approved by the Institutional Review Board of the Second Affiliated Hospital of Zhejiang University School of Medicine (ID No.: 2022-0280).
Informed consent was obtained from all subjects involved in the study. The study was consistent with the principles of the Declaration of Helsinki.