Factors Associated with Health Literacy Competencies: Analysis of Thai Health Literacy Survey 2019

The consumption of health-related products has been increasing continuously. Information on health-related products can make it dicult for some people with limited health literacy to use. This study investigated the extent to which health literacy competencies in the aspect of consumer protection related to demographic and socio-economic factors, using data from the Thai Health Literacy Survey (THL-S) among Thai citizens aged 15 years and above (2019). The THL-S used a stratied three-stage-sampling to draw a sample of Thais aged 15 years and above. Participants were interviewed with a questionnaire of 34 items measuring health literacy and 8 items measuring behavioural practices. Proportions of responses (the 6-Likert scales) in accessing, understanding, communicating and making decisions related to the consumer protection aspect were performed. Logistic regression models were used to explore the association between health literacy competencies and participant’s socioeconomic, demographic, health and social characteristics.


Background
In 2019, Thailand had its rst survey on health literacy among a population above 15 years of age. A strati ed-three-stage sampling was used together with a weighing method to represent both entire and regional populations. This survey was one of the largest national surveys, with the sample size of 17,530 people. The results showed approximately 19 %of Thais who had insu cient health literacy in accessing, understanding, communicating and making decisions, based on four domains of health-related information including health care, disease prevention, health promotion and health and medical products. Furthermore, Thais found that accessing health-related information in all aforementioned domains as the most di cult, followed by communicating about health (1). The survey has become a crucial starting point for shaping national health care reform strategies. Health literacy has also been included in national legislative initiatives. The 20-year National Strategic Plan: 2018 -2037 has included health literacy as one of the eleven health care reform areas. The ve-year National Health Development Plan: 2018 -2022 also indicates that health literacy among Thais should be improved by 25 % by 2020.
From the literature, health literacy might play a signi cant role in maintaining or improving health across one's lifespan. It is also thought to be a potential predictor of health inequity (2)(3)(4). A cross-sectional study on health literacy among Dutch adults (2013) (4) investigated on how health literacy competencies related to demographic and socio-economic characteristics, using some data from the European Health Literacy Survey (HLS-EU) in the Netherlands. The study showed interesting ndings, arguably leading to another point of view on analysing and interpreting health literacy competencies at a national level, rather than focusing on general health literacy. The previous study categorised health literacy competencies into different health domains and analysed their relationships with education, income, social status, age and sex, as well as individual competency within each domain. The study showed inconsistent relationships between both demographic and socio-economic characteristics and all health literacy competencies from each domain as well as general health literacy. For example, education and income were signi cantly associated with accessing and understanding health information, but to a lesser extent with appraising and applying. With regard to accessing and understanding, the group with the lowest income had lower health literacy scores compare to the group with the highest income, yet this difference was found only in the health care domain. These ndings allow policy makers to better identify potential characteristics of the target group when designing strategies for improving access to and understanding of health information in the health care domain.
However, within the context of the general Thai population, little is known on how different health literacy competencies relate to each other, as well as what factors in uence these competencies and in what way. Understanding these relationships would not only help us to better understand health literacy but would also contribute to creating more effective ways of designing, planning, implementing and evaluating health services in all relevant domains. The ndings could empower Thai citizens and reduce health inequity in Thailand, pointing out the importance of addressing health literacy in the national legislative initiatives.
According to the Thai National Health Examination Survey 2014, the percentage of Thais aged 15 years and above who consumed health-related products and supplements within 30 days increased by 2.25 folds within ve years (14.8 % in 2010 versus 33.3 % in 2014) (5,6). Moreover, the out of pocket expenditure of households in Bangkok due to the consumption of supplements accounted for one-third of the total amount of the out of pocket expenditure on health care (7). In a highly competitive market of health-related products and supplements, individuals and companies use marketing strategies, social media and easy-to-understand formats to draw customer's attention. There is plenty of information about health-related products and supplements but such information lacks characteristics of trustworthiness (8). Some people pay attention to pictures, colours, and celebrities on the product's labels and the advertisements. This makes it di cult and often misinforming to those with limited health literacy skills to distinguish reliable information, and access reliable sources. Therefore, there is need to ensure reliable, accessible, and understandable health information on health related products as it helps protect the consumer against misinformation on health related products. This study investigated the extent to which health literacy competencies, which are accessing, understanding, communicating about health, and making health-related decisions in the aspect of consumer protection, relate to demographic and socio-economic factors, using data from the Thai Health Literacy Survey among Thais aged 15 years and above (2019).

Study design and data collection
The Thai Health Literacy Survey used a strati ed-three-stage sampling to draw a sample of Thais aged 15 years and above (1). The sample was strati ed based on health regions, provinces, enumeration areas, and households. In the rst stage, three provinces in each health region (13 health regions) were systematically identi ed by ranking the number of inhabitants aged 15 years and above from smallest to largest. Thirty-seven provinces were selected in total. In the second stage, 492 enumeration areas within the selected provinces were de ned based on their locations. In the third stage, a total of 7,380 households were systematically selected from each enumeration area. In each household, all members who were 15 years old or older were interviewed with a questionnaire. Data were collected at the participant's home between March to August 2019. A total of 7,295 households participated in the survey in which a total of 18,832 people met the survey criteria. The enumerators were able to interview 17,530 people (response rate of 93%). Other 867 people were unavailable after a three-time follow-up and 375 people refused to participate in the interview.

Health Literacy
The questionnaire contained 42 items of which 34 items measure health literacy competencies in four domains, while 8 items measure behavioural practices. The questionnaire was developed by the Department of Health, Ministry of Public Health with nancial support from the Health System Research Institute. The items measuring health literacy competencies in the consumer protection aspect are shown in Table 1. The questionnaire was pre-tested for understandability and relevance with a sample of 722 people from six provinces across all regions (n = 120 in each province). Focus groups were conducted with 10 -12 respondents who had di culties understanding the questionnaires. The overall internal reliability of the questionnaire, as indicated by Cronbach's alpha coe cient of 0.94, was good. The internal reliability for the health literacy competencies was also good (Cronbach's alpha coe cient of 0.88 for 'accessing', 0.86 for 'understanding', 0.90 for 'communicating' and 0.88 for 'making health-related decisions'). For accessing, understanding, and making health-related decisions, the respondents were asked to choose from a 6-point Likert scale ranging from,1 =very easy, 2 =fairly easy , 3 = fairly di cult, 4 = very di cult 5 = unable to perform, and 6 = con dent in performing but never had a chance to perform. For communicating, the 6-point Likert scale ranged from, 1 =all the time, 2 = sometimes, 3 = never do , 4 = not dare to do , 5= don't want to do, and 6 = having someone do it,. Table 1 Health Literacy items for the consumer protection aspect 6. = con dent in performing but never had a chance to perform how easy is it for you to…

Accessing
Find reliable information about medicine, cosmetic products, herbal products and food supplements.

Understanding
Understand information on labels of medicine products, cosmetic products, herbal products, medical devices, and hazardous chemical products.

Communicating
Inquire healthcare providers about health-related products

Making decisions
Make decisions about what food supplements or herbal products are suitable for you.

Demographic and socio-economic characteristics
The demographic and socio-economic characteristics analysed in this study are sex, age group, marital status, the highest level of education, level of reading di culty, level of writing di culty, holding leading roles in the community, income su ciency, occupation, insurance scheme, presence of chronic disease, level of hearing di culty and use of eyeglasses or contact lens. All of the demographic and socio-economic data were analysed as categorical variables.

Statistical analysis
Proportions of responses (the 6-Likert scales) in four competency domains in the consumer protection aspect were performed. Logistic regression models were used to explore the association between health literacy competencies and socioeconomic, demographic, health and social characteristics. The responses of each domain were grouped into a dichotomous variable representing whether respondents have di culties in accessing, understanding, communicating or making decisions. For each competency, a binomial logistic regression was performed using R 3.1.0 (9, 10).

Results
The demographic and socio-economic characteristics of the sample are shown in Table 2. Female samples accounted for 61% of all samples, which were overrepresented comparing to the distribution of the Thai population (11). In terms of age group, most of the samples were in 60 years and above and 46-59 years groups. These gures were not in line with the distribution of the Thai population, as the samples aged 60 years and above and aged 46-59 years were overrepresented while the samples aged 25-45 years and 15-24 years were underrepresented. Majority of the samples were married or living together. Half of the samples completed primary education while 4% was illiterate. Approximately 10% had insu cient income to support their family. One-third of the samples worked in the agricultural section. Besides, the majority of the samples (79%) were registered under the universal health care coverage scheme (UCS). Approximately one-third of Thais (38%) were unable to or experience di culties in accessing reliable information about medicine, cosmetic products, herbal products and food supplements. The least di cult competency rated by this population was understanding information on labels of medicines, cosmetic products, herbal products, medical devices, and hazardous chemical products. Socioeconomic, demographic, health and social characteristics associated with all health literacy competencies in the aspect of consumer protection include the highest level of education, sex, level of reading di culty, not receiving health screenings, and holding leading roles in the community as shown in Table 4. However, the extent of association varies among competencies. These characteristics associated mostly with accessing, followed by communicating, understanding and making decisions. Overall, it was found that people with a lower level of education, a higher level of reading di culty, no previous health screening, no leading roles in the community, or were male, experienced more di culties in practising health literacy competencies. People who reported inadequacy of income were more likely to experience more di culties in accessing, understanding or communicating than people who had savings. Government o cers tended to experience fewer di culties in accessing, understanding or communicating than unemployed people. However, no association between occupation and ability to choose suitable food supplements or herbal products was found. People under private insurance or social security scheme were less likely to experience di culties in accessing information compared to those registered with the UCS. Regarding understanding, people with private insurance or Civil Servant Medical Bene t Scheme (CSMBS) tended to experience fewer di culties than people with UCS. The older the people were, the more di cult they experienced in accessing, understanding and communicating. The marital status associated only with accessing. Level of reading di culty showed a stronger association with health literacy competencies compare to the level of writing di culty. People with no chronic diseases or were diagnosed with chronic diseases were less likely to experience di culties in health literacy competencies than those who never received screening. People with hearing impairment were more likely to experience di culties compared to those without, while people who wear eyeglasses tended to experience less extent of di culties than people who do not.

Discussion
This study found that di culties in the consumer protection aspect vary among health competencies, with communicating being the most di cult and understanding being the least di cult. Thai citizens felt that it is relatively di cult to inquire about healthcare providers about medicine, cosmetic products, herbal products and food supplements. They were also unable to or had some di culties in accessing reliable information about medicine, cosmetic products, herbal products and food supplements. Each health literacy competency requires different skill sets and knowledge to practice. For accessing, a person requires pro cient skills and knowledge to use technology to access resources (12). For understanding, a person requires to understand meanings of words and terms as well as having experiences, to interpret information correctly (13). To be able to communicate con dently, a person has to understand the situation, realize the impact of using health products on themselves and their relatives, know how to formulate questions and has enough self-con dence to ask or start a conversation with others (13). For making decisions, a person needs to have enough relevant information about the issue as well as self-determination and to be able to critically analyse possible consequences of different choices (13). On the other hand, the context of a person plays a role in determining the degree of di culties in practising health literacy (13). If a person responds to stimuli very well, the same person might respond differently in another context with different stimuli and different level of di culties. Possible reasons that Thais do not ask or have di culties in asking or communicating with healthcare providers about health-related products relate to norms, beliefs, perception, prior experiences as well as lack of question formulation and communication skills. The paternalistic nature of the patient-doctor relationship in the Thai context could also be another explanation. Health care providers are the authority who responsible for diagnosis and treatment, while patients are viewed as passive and are not expected to actively participate in the process of decision making on their care or to ask any questions (14,15). Doctors normally do not promote health-related products such as herbal supplements or food supplements and asking questions about health-related products or supplements may offend them, leading to arguments and negatively affect the patient-doctor relationship. Some people think that having arguments with their doctors might affect how the doctors treat them or their relatives. This nding is in line with a previous study in Thailand that showed hypertensive patients did not ask for information from medical staffs because they felt obligated to their physicians, thereby missing an opportunity to gain related knowledge to take care of themselves (16).
This study con rms that health literacy levels can re ect responses as a result of interaction between individual skills required to practice health literacy competencies and the complexity of health care contexts. For optimal improvement of health literacy, there should be a match between individual skills and system demands. Skills such as searching for reliable sources and appraising reliability of information as well as asking for clari cation from healthcare providers are needed to be addressed and trained. A system should be designed in a way that reduces barriers for practising these skills. Healthcare providers should encourage patients to ask questions in hospitals and primary care setting to improve patients' understanding of relevant health information on healthrelated products. A good example is 'Ask Me Three' approach, a practice that encourages patients and family members to ask three speci c questions to better understand their health conditions. The practice was found to be effective in improving patient's understanding, communication skills, and compliance with health-related advice (17). In addition, there should be a mass communication to create a new norm that asking is a necessary action to protect one's bene ts. Also, reliable sources of information are important to gain knowledge about health-related products and supplements. Another important recommendation is to promote and build skills and knowledge of the population for evaluating health information on health-related products and supplements. Finally, a monitoring and alert system for consumers about untrustworthy information of health-related products on the Internet and communities should be developed.
Our study found that people with a lower level of education, could not read, did not receive health screening, were living in poverty, did not hold leading roles in the community, were male, or have a hearing impairment or were at older age experienced more di culties when practising health literacy competencies. The ndings are similar to prior small-scale studies in Thai patients (18,19). The extent of the association varies among competencies. These factors associated in a greater extent with accessing, followed by communicating, understanding and making decisions. The level of education in uences development of health literacy competencies in a way that students have the opportunity to acquire and practice sets of skills and knowledge, especially literacy, numeracy, and critical thinking, which are crucial for practising health literacy competencies (13). Under the Thai national health education curricula for primary and secondary schools, there are sets of literacy skills, knowledge, and health-related practices that students are required to have. A lower level of education indicates fewer practices in health literacy competencies in the classroom, potentially leading to experiencing more di culties in practising health literacy competencies in health-related contexts. The level of reading di culty has a stronger association with health literacy competencies than the level of writing di culty. Reading ability is crucial for accessing and understanding information, as most health-related information is presented in written forms. In Thailand, those with higher educations are more likely to have higher incomes and employed in companies with either private insurance or social security scheme or both (20). This could explain the differences that those with more income experienced less degree of di culties in practising health literacy competencies. People who hold leading roles in the community may have more exposure to health-related information and events, and people who received an annual health screening may have more experiences in coping with various demands of the health service systems, which then helps to improve their health literacy competencies at a faster rate compared to those who did not have one.

Strengths and limitations of the study
The Thai Health Literacy Survey 2019 included health literacy measures that were relevant in Thai contexts. It measured the health literacy skills in four health domains; health care, disease prevention, health promotion and consumer protection. As the consumption of health-related products and supplements in Thailand has been increasing (5, 6), the improvement of health literacy in the consumer protection aspect might help Thai citizens make healthy choices during their life course.
Another strength comes from the sampling methods. The survey used a three-stage sampling technique based on health regions, provinces, enumeration areas and households. With the questionnaire administered face to face, the survey results ensured a better representation of Thai citizens in remote areas throughout the nation including some minorities who might have inadequate reading and writing abilities in Thai language.
A limitation of this study is that the sample overrepresented the elderly, which might have affected other factors such as adequacy of income, level of education, occupation, and ability to read and write (4). The questionnaire did not include some variables that might have affected the opportunities to gain and practice health literacy skills such as experiences of taking care of ill people in the family and duration of living with the current disease (21).

Conclusions
In conclusion, characteristics of a person, as indicated by demographic, socioeconomic, health and social factors, can explain differences in how persons develop and practice health literacy skills and experience degrees of di culties in practising health literacy skills in the consumer protection aspect. In particular, vulnerable consumers in the society, such as people with a low level of education, living in deprivation, at old age and hearing impairment, face signi cant barriers in accessing, understanding, communicating and making decisions in health-related products. The context of deprivation and social vulnerability in which these people live, might even worsen their quality of life. Therefore, health literacy programs should be developed to build health literacy competencies and empower vulnerable consumers. Responsible organizations should also promote adult education among elderly people to address basic literacy skills, media literacy, and health communication skills. Furthermore, there is a need to adapt the current health services and information on health-related products to meet health literacy needs among vulnerable consumers. In addition, health information on health-related products or services should be designed in a way that is understandable and accessible to everyone, regardless of their literacy levels.