Design and participants
A cross-sectional survey was conducted in Liaoning Province, China, from August to December 2021. The participants were hypertensive patients from Fuxin, Chaoyang, Panjin and Jinzhou. All patients provided informed consent before participating in the study. The research procedures complied with the ethical standards of the Ethics Committee of Jinzhou Medical College, the 1964 Helsinki Declaration, and its later amendments. A total of 1577 hypertensive patients were recruited for the survey. During the survey, the authors and investigators explained the study’s purpose and methods to patients. The questionnaires were individually delivered to each participant and completed in the presence of the authors and the investigator. The participants were encouraged to give truthful answers. Subjects who had not fully completed the scale and provided questionnaires with obvious logical errors were excluded. The remaining 1472 patients (93.34%) were retained. The survey was anonymous, except that 37 patients in Jinzhou were required to write their names as the test-retest participants. One month later, 37 patients who participated in the first test were recruited to evaluate the test-retest reliability.
The first step
We obtained permission from Dr. PH Chau to revise and verify the Chinese Health Literacy Scale For Low Salt Consumption. The Chinese Health Literacy Scale For Low Salt Consumption was formulated in 2015 .
The second step
We then compared the questions with the corresponding culture of the researchers, companies (Hanjing Zhang, Song Li, et al.)and professors (Hongyu Li, PH Chau, et al.). We revised some items to conform to the diet. Considering the different food preferences between Hong Kong and mainland China, we replaced some food examples with this scale. Based on the literature review, the team and experts removed some items based on evidence, considering that some items were outdated and unsuitable for future applications in various countries. When filling in the questionnaire, the patients felt that their functional literacy and knowledge of international standards were similar. After experts and a literature review, the two dimensions were combined. Regarding statistics, experts said a scale should have no more than seven dimensions. Finally, a pilot study was conducted among 37 patients with hypertension. They were invited to complete the scale and then were asked about their understanding of the scale’s introduction section, items, and options. We communicated with the survey respondents, who reported that they had no difficulty in understanding the content of each scale item, and the revised scale was obtained.
The third step
The revised scale was investigated in 469 older people referred 60 years old and older.. Through the statistical analysis, the revised scale indicated well reliability and validity. So, we ended up with the Chinese Health Literacy Scale For Low Salt Consumption of 22 items. The scale was tested on 300 non-hypertensive subjects to determine its explanatory degree and stability. ROC curve analysis was conducted with hypertensive group as experimental group and non-hypertensive group as control group. The objective of the ROC curve is to understand the trend and critical value of hypertensive low-salt health literacy.
The fourth step
We applied the CHLSalt-22 scale to 1003 hypertensive participants The purpose of this step was to use the revised scale in different age groups, further verify the applicability of the scale in inland Chinese population, and reveal the low-salt health literacy of inland Chinese population.
All participants completed the CHLSalt-22, the measuring change in restriction of salt (sodium) in the diet in hypertensives (MCRSDH-SUST) , the Brief Illness Perception Questionnaire (BIPQ) , and the Benefit Finding Scales (BFS) . Participants were also asked to complete a checklist assessing sociodemographic variables (e.g., sex, age, and income). Height and weight were also measured to calculate each participant’s body mass index (BMI) Participants were categorized as underweight (< 18.5 kg/m2), normal weight (18.5–23.9 kg/m2), overweight (24–27.9 kg/m2), and obese (≥28 kg/m2) based on Chinese criteria of weight for adults .
The Chinese health literacy scale for low salt consumption (CHLSalt-22)
The CHLSalt-22 consists of 22 items with three items assessing Functional literacy, four items assessing Salty food knowledge, three items assessing Disease knowledge, three items assessing Myths about salt intake, three items assessing Salt intake attitudes, three items assessing Salty food consumption, and three items assessing Nutrition label practices. The questions’ responses were in the form of either a 5-point Likert scale or four multiple-choice options. For the Likert-scale questions, the most favorable option scored two points, the following profitable option scored one point, and the remaining three scored 0 points. The correct answer scored 2 points for multiple-choice questions, a score of 2 was assigned to the correct answer, and the remaining options scored 0 points. The total score was calculated by summing up the scores for each item .
The measuring change in restriction of salt (sodium) in the diet in hypertensives (MCRSDH-SUST)
The Continuous Behavior Change Sub-scale (McRsdh-sust) includes nine items, including emotional change (three items), behavior change practice (three items) and social environment change (three items). Likert 5 rating method was used for this scale, 0 = “very uncertain”, 1 = “uncertain”, 2 = “not certain”, 3 = “certain”, 4 = “very certain”. The higher the total score, the more likely people with hypertension will change their salt-restricted diet .
The brief illness perception questionnaire (BIPQ)
The Brief Illness Perception Questionnaire (BIPQ) was compiled by Broadben et al. It includes two dimensions of cognitive disease symptoms and the degree of understanding for patients with emotional disorders, with eight items as a self-assessment questionnaire. The cognitive disease representations include disease influence course, symptom recognition, individual control, and treatment control. The emotional disease representations include disease worry and mood, with each item rated from 0 to 10. 0 = “degree of tiny “,10 = “degree of extremely strong”,with a total score of 80 points . The higher the score, the stronger negative perception.
The benefit finding scale (BFS)
The Chinese scale includes 19 items, all scored on a 4-point Likert scale, assessing 1 to 4 points from none to very much. The total score was 19–76, and the higher the score was, the higher the perceived benefit level of the participant. The Cronbach’s α coefficient of the scale in this study was 0.910 .
Statistical analysis of data
Data analysis was performed using SPSS 26.0 and Mplus 8.0. Given that all the items were dichotomous, Kuder-Richardson’s α (KR-20) was used to assess the internal consistency of the CHLSalt-22. The test-retest correlation coefficient (intraclass correlation coefficient, ICC) was used to calculate the scale’s stability. Values of ICC were interpreted as follows: > 0.75 was excellent, between 0.40 and 0.75 was fair to good, and < 0.40 was poor . The content validity index (CVI) and Pearson’s correlation coefficients between items and total scores were used to evaluate the scale’s content validity. The CVI includes item-level content validity index (I-CVI) and average S-CVI (S-CVI/Ave) . Each expert chose the relevance of each item to the corresponding dimension. A 4-point rating scale was used to calculate CVI (1 = no relevance, 2 = low relevance, 3 = strong relevance, 4 = very strong relevance).EFA and CFA were used to examine the construct validity of the CHLSalt-22. Data were divided into two groups. Sample 1 consisted of 469 hypertensive patients (53.3% women, mean BMI = 26.61, SD = 3.12), while sample 2 consisted of 1003 hypertensive patients (32.2% women, mean BMI = 26.89, SD = 3.61). The factor ability of the correlation matrix was assessed with the Kaiser–Meyer–Olkin (KMO) statistic and Bartlett’s test for sphericity , and EFA was conducted on sample 1. A scree plot was used to constructed to determine the number of factors. CFA was performed on Samples 1 and 2, and the test level was α = 0.05. To assess the quality of the factor model, we estimated the following indices: minimum function chi-square (χ2), comparative fit index (CFI), Tucker-Lewis index (TLI), standardized root mean residual (SRMR), and the root mean square of approximation (RMSEA). An acceptable model should have a χ2/df < 3, an RMSEA and an SRMR< 0.08 , and a CFI and a TLI > 0.9 . To assess convergent and discriminant validity, we used the path coefficient of CFA to analyze composite reliability (CR) and average variance extract (AVE). An acceptable model should have a CR>0.7 and an AVE>0.45 . Through SPSS 26.0 to analyze the predictive validity of the CHLSalt-22 scale. The area under the ROC curve was>0.70, and the best cut-off points for the CHLSalt-22 scale were analyzed. An AUC of 0.5 represents a test with no discriminating ability, while an AUC of 1.0 represents a test with perfect discrimination . Independent sample t-tests or single-factor ANOVA of the difference in the total score of symptom counts between sociodemographic classifications and Bonferroni’s test were used to calibrate the inspection level for pairwise comparisons. The correlation between the CHLSal-22 t count score and MCRSDH-SUST, BIPQ, and BFS was evaluated by calculating Pearson’s correlation coefficient. The CHLSalt-22 count score was taken as the dependent variable, and the classified and continuous variables were used as independent variables for multivariate linear regression analysis. The multi-classified disordered variables were set as dummy variables according to the requirements of multivariate linear regression for independent variables. The significance level was set at P < 0.05.