Health literacy experiences of multi‐ethnic patients and their health‐care providers in the management of type 2 diabetes in Malaysia: A qualitative study

Abstract Background Patients with type 2 diabetes mellitus (T2DM) require adequate health literacy to understand the disease and learn self‐management skills to optimize their health. However, the prevalence of limited health literacy is high in patients with T2DM, especially in Asian countries. Objective This study aimed to explore experiences related to health literacy in Asian patients with T2DM. Design This is a qualitative study using in‐depth interviews and focus group discussions. A framework analysis was used to analyse the data. Setting and participants articipants (n = 24) were multi‐ethnic patients with T2DM (n = 18) and their primary health‐care providers (n = 6). This study was conducted in four primary health‐care clinics in Malaysia. Results Nine subthemes were identified within the four dimensions of health literacy: accessing, understanding, appraising and applying information. Discussion Motivated patients actively sought information, while others passively received information shared by family members, friends or even strangers. Language and communication skills played important roles in helping patients understand this information. Information appraisal was lacking, with patients just proceeding to apply the information obtained. Patients' use of information was influenced by their self‐efficacy, and internal and external barriers. Conclusion In conclusion, the experiences of multi‐ethnic patients with T2DM regarding health literacy were varied and heavily influenced by their cultures.


| INTRODUC TI ON
Health literacy is an important ability for patients when negotiating complex and long-term health conditions such as type 2 diabetes mellitus (T2DM). 1 It is defined as the degree to which individuals can access, understand, appraise and apply health information to make informed health decisions to maintain quality of life across their life course. 2 It is closely related to literacy and entails patients' knowledge, motivation and competences. 2 Health literacy is of particular importance to patients with T2DM to enable the use of information and services to make appropriate lifestyle and treatment decisions. 3 Studies have shown that limited health literacy in patients with T2DM was associated with adverse health outcomes. These patients had difficulty in reading printed instructions and understanding health recommendations or warnings. 4 They were also reported to have less disease knowledge, 5,6 poorer medication adherence 7 and expended more money on health care. 8 These patients also experienced poorer patient-doctor communications and participated less in shared decision making. 9 However, evidence linking limited health literacy and glycaemic control is more mixed. Some studies have demonstrated links between higher levels of health literacy and lower HbA1c, while others have failed to show an association. 5,[10][11][12] Despite the mixed findings, diabetes self-management interventions that considered health literary aspects showed positive effects on glycaemic control. In 2016, a systematic review and meta-analysis of health literacy-sensitive diabetes self-management interventions found a significant reduction in HbA1c for health literacy-sensitive interventions when compared to usual care; the effect was more pronounced in patients with limited health literacy. 13 Studies that employed interventions focusing on self-management in patients with limited health literacy also found a significant reduction in emergency department visits and hospitalizations. 14 Many of these interventions were educational, employing the use of written communication, oral communication, patient empowerment and tailored communication to patients' language and cultural practices. 13 Limited health literacy in patients with T2DM is prevalent, and the rate is even higher in Asian countries with multi-ethnic populations. A recent systematic review found the pooled prevalence of limited health literacy in the United States of America to be 30% and the prevalence in European countries to range from 7.3% in Switzerland to 9.7% in the Netherlands. 15 In Asian countries with multi-ethnic populations, the prevalence ranged from 71.7% in South Korea to 82% in Taiwan. 4,16 In Malaysia, recent studies have shown the prevalence of limited health literacy in patients with T2DM to be as high as 85.8%. 17 Recent qualitative studies exploring health literacy in patients from Asian cultures found that culture shaped patients' understanding and experiences of health literacy. In a study exploring health literacy experiences of Chinese patients with diabetes living in America, Leung et al 18 found that cultural issues influenced patients' access, understanding and application of information. Similar findings were noted in Thai patients with diabetes, 19 and in Samoa, the health literacy of patients with non-communicable diseases is heavily influenced by culture. 20 Malaysia is an ideal location to study the health literacy experiences of multi-ethnic Asian patients. Malaysia is a multi-ethnic, multi-cultural and multi-lingual country where the major Asian ethnic groups Malay, Chinese and Indian still maintain somewhat separate cultural identities. Malaysian society had been described as 'Asia in miniature'. 21 Furthermore, T2DM is a major health problem in Malaysia. The prevalence of diabetes in Malaysia was 17.5% in 2015. 22 Only 70% of these patients achieved target glycaemic control, thus leading to a high prevalence of microvascular and macrovascular complications. 23,24 Therefore, this study aimed to explore health literacy experiences in the management of T2DM in multi-ethnic patients and their health-care providers (HCPs) in Malaysia.

| Design
This study used a qualitative approach where in-depth interviews and focus group discussions were conducted to explore the experiences of multi-ethnic patients with T2DM and the primary HCPs who supported them.

| Setting
This study was conducted in four primary health-care clinics in the state of Perak, Malaysia, from January to April 2019. Participants were patients with T2DM previously involved in a study measuring the prevalence of limited health literacy at the same setting. They were purposively selected to represent the experiences of patients with adequate and limited health literacy. Health-care providers (HCPs) who participated in the study worked at the study clinics and were involved in delivering care to the patients with T2DM involved in this study. Most patients approached agreed to be interviewed, three patients declined, two had relocated and one was too ill. All HCPs approached agreed to be interviewed.

| Topic guide
The conceptual model of health literacy by Sørensen 2 guided the development of the topic guide. 2 Questions were kept as openended as possible with the aim of exploring participants' views and experiences within the four dimensions of health literacy. Separate topic guides were created for patients' interviews and for HCPs' interviews. The topic guide for patients was developed in English and then translated to Bahasa Malaysia and Mandarin, whereas that for HCPs was in English. The topic guide was revised after each interview to include new issues raised by the participants. There were no new revisions of the topic guide after three patient interviews and two HCP interviews. The topic guides for patient and HCP interviews were included as Appendix S1 and S2.

| Sampling and data collection
Patients were recruited through phone calls, and the HCPs were recruited by SMS messages (WhatsApp) and email. All participants were provided with the participant information sheet and given time to read through it before the interview. Those who agreed to participate were then asked to sign a consent form. All interviews were conducted in a private room at the study sites. Interviews in English and Bahasa Malaysia were conducted by the first author (AA). A trained research assistant (CHY) conducted the Mandarin interviews. CHY was present at three of the interviews conducted by AA and took notes for both focus group discussions. The interviews were audio-recorded. There were 18 patients' in-depth interviews, one HCP in-depth interview and two focus group discussions involving five HCPs.

| Data analysis
The interviews were transcribed verbatim, and the data were managed using ATLAS.ti 9 software. Transcripts in other languages were translated and analysed in English. Interviewers' reflective notes were written after the in-depth interviews, and field notes were taken during the focus group discussions. Each in-depth interview lasted for 30-45 minutes, and the focus group discussions lasted for 45-60 minutes each. Data reached saturation after the 16th patient interview. The HCP interviews did not reach data saturation, but these interviews were performed to triangulate patients' data.
Transcriptions were done by professional transcribers and checked by the first author by listening back to the audio recordings of the in-depth interviews (IDIs) and focus group discussions (FGDs).
The data were analysed using the five stages of the thematic analysis framework: (a) familiarization; (b) identifying a thematic framework; (c) indexing; (d) charting; and (e) mapping and interpretation. 25 Familiarization involved gaining an overview of the literature, research objectives and data. Then, a framework was identified. In this study, in order to explore all the capacities associated with health literacy, the thematic framework was informed by Sørensen's 2 conceptual model of health literacy. The domains of health literacy made up the categories for data analysis.
All interview transcripts were then systematically coded using the thematic framework created. All the coding was completed by AA, and LSM and NCJ double-coded two transcripts. The researchers involved in the analysis were primary care physicians (AA, LSM and NCJ) who were conscious of their personal views and biases concerning health literacy in patients with T2DM. Quotations that best reflected the themes that emerged from the transcripts are presented in results.
The rigour of this research was ensured by using the criteria of credibility and confirmability. In order to achieve credibility, the topic guides were developed using an established conceptual framework and data triangulation was performed by using two data collection techniques, that is IDI and FGD, as well as collecting data from patients and their HCPs to capture patients' experiences related to health literacy from two different perspectives. Transcripts of the interviews were also shared and checked with the participants in a process called member checking. Researcher bias was addressed by the main researcher completing a reflective diary at the end of every interview, which captured the researcher's perceptions and opinions, which were then bracketed during data analyses. Dependability was achieved by having two other researchers read through the transcripts and agree on the codes and themes generated.

| RE SULTS
The interviews yielded rich data pertaining to the way patients access, understand, appraise and apply health information in the management of their T2DM. The participants' characteristics and study sites are summarized in Table 1. Categories, themes and subthemes are set out in Table 2. Quotations are labelled with the prefix P for participant followed by the participant's number, for example P1 denoting participant 1.

| The health literacy experiences of multi-ethnic patients with T2DM
Patients with T2DM and their HCPs shared their health literacy experiences, and these were then coded into the four health literacy domains of accessing/finding, understanding, appraising and applying information. For each domain, subthemes were identified to represent factors that influence participants' experiences ( Figure 1).
Direct quotations from participants' experiences were used to illustrate the findings.

| Accessing/finding information
We found three subthemes relating to participants' experiences in accessing/finding information: formal and informal sources of information, push and pull information, and motivation.

| Formal and informal sources
Patients sourced their information from informal and formal sources.
The main formal source was their HCPs, particularly doctors.
Participants felt that it was the doctors' responsibility to provide them with information and that 'if the doctor did not tell, I won't know': HCPs understood their responsibility to share information with patients but felt they should also empower patients to look for information themselves. HCPs admitted that they only gave advice to patients with poor control, as a lack of consultation time limited the number of patients and the depth of discussion that they could engage in: So, we play a very important role to educate. And I think we should also actively provide sources of information, ya.
We cannot be telling everything because of time again. If anything is abnormal then we will bring it up.
Otherwise, if everything is okay, we don't discuss it anymore. (P20)

| Push and pull information
As opposed to accessing information by themselves (pull), participants told us that information was also 'pushed' to them. When patients were actively searching for information to assist in their effort to manage their diabetes, information was being pulled by patients from various sources as discussed above. However, there were instances where information was shared with patients whether they solicited it or not: Anybody where… anyone wherever we go, sometimes we go (to) temple will talk… (they will say) this one got diabetes, they said okay, you all take this take that.

| Motivation
Patients who went out to search for information were those who were motivated to do so and could appreciate the value of such pursuit. Some had seen the suffering of other patients with T2DM and believed that with appropriate information they could better control their illness: For us diabetics must be the individual, the one with the illness, the sick, the sick themselves.
Because I see this person suffering from diabetes.
suffer … my friend has diabetes, and heart disease.
In contrast, there were participants who did not see any value in looking for information. They believed in fate and did not have any adverse symptoms from the disease and thus were not motivated to look for information. Some even felt scared to look for information which might tell them of the potential negative outcomes of the illness: (Laughing) Don't know, it is Allah's job to hold our lives. If He wanted to take it, He will take it….

| Understand information
Subsequent to finding information, participants shared their experiences in understanding the information they found. We found two subthemes in this domain: the influence of language, and communication and clarification skills.

| Language
Patients said that English is the language of science, but those who were not fluent in English sought information in the language in which they were fluent. They even looked for doctors who were able to converse in their mother tongue: English………the language of science.

| Communication and clarification skills
When patients gathered information independently, they found that they had unanswered questions that required clarification. Some patients tried to engage the HCPs to assist them:

| Appraising information
Participants moved from accessing/finding and understanding information straight to applying information. They did not engage in evaluation of the information prior to applying it: I believe more in the information regarding medications and diet only. The other information, I take them less seriously. I am also not sure which is right or wrong. Yes, I just accept. If I can follow, I will follow. If I can't then I will just forget it.
HCPs also made the same observation. They felt patients did not reflect on what was given to them and were more likely to just follow the instructions given by HCPs: For the older generation er they tend to just get from the information and they just take in the information, they did not er you know, clarify or, or criticise whether the information that they received is accurate or not. They believe it as a gospel truth.

| Trust and belief
Much of the information was gathered orally by approaching their HCPs, families and friends (principally those with diabetes). These were people who they trusted to give them the right information.
Patients also trusted information found online, because they believed that the government regulated this information and that most of it was written by HCPs. They also believed that if the information was consistent across many websites or repeated by different doctors, then it must be true: (Information) that is on the Internet can be trusted.
Maybe the information is written by doctors, the source of the information is from doctors also…now information is at your fingertips.
If it's on the Internet, then it should be harmless right, they are regulated by the government. (P18) The info, most of the websites, the info is about the same. So, I… I trust that it is true. (P6) But for me…I don't just refer to one source, right?
Maybe two or three. And if they (different sources) state the same information, then it must be right.

| Applying information
In the final stage of their health literacy journey, patients shared their experiences of applying health information. Self-efficacy seemed to be the main influencing factor here, but patients did discuss the personal and environmental barriers faced when applying information.

| Self-efficacy
Patients in our study declared that the decisions to act on the information were their own. They had ownership of their lives and health, as one participant stated, 'diabetes is my life' (P3): Not really, it starts with your own self, doesn't it? Like the proverb 'when there is a will, there is a way'. HCPs found self-efficacy to be the main factor for patients to act on information given to them. They also noted some internal and external barriers faced by patients: Er Ok most patients actually I think they can understand our instructions and all but some they find it difficult to follow, I'm not sure why, maybe because of their work, er some they are working shift hours then it is difficult to take their medicine or sometimes they often forget,…diet control is definitely (chuckles) quite bad. (P21)

| Barriers
Patients also shared the barriers they faced when trying to apply health information. These barriers were related to personal barriers and environmental barriers. Patients cited lack of self-control and self-discipline as internal barriers, and time and family responsibilities as external barriers: Aa that (laughs) people like to give (sweet food) to me.
My heart says no, but I still want them (laughs).

| Strengths
This study interviewed patients with adequate and limited health literacy and used a framework approach to analyse the data.

| Limitations
One criticism is that the number of health-care providers inter-

| Implications for policy and practice
This study identified several points for intervention towards enhancing patients' health literacy levels. Adequate health literacy is essential for patient empowerment and involvement in their health care. 32 Health-care policymakers and providers need to be aware of current barriers faced by patients with T2DM and limited health literacy in order to deliver health care sensitive to patients' health literacy level. Interventions need to address the language barriers that are

| CON CLUS ION
This study found a duality in the health literacy experiences of multi-ethnic patients with T2DM. These experiences were heavily influenced by culture, which explained the absence of this pathway in previous health literacy conceptual frameworks. Therefore, interventions to improve health literacy in these patients need to be targeted not just at patients but their families, friends, the general population and HCPs. Appropriate training must be provided to HCPs in view of the important role they play in improving patients' health literacy.

ACK N OWLED G EM ENTS
We are grateful to the study subjects for participating in this study.
We thank Dr James Gnanasigamani, Mastura bt In, Uma Muthu and Sattieswary A/P Govindasamy Kee for their support in data collection and participant recruitment.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.