Understandings of disease among Pacific peoples with diabetes and end‐stage renal disease in New Zealand

Abstract Background Compared with New Zealand Europeans, Pacific peoples in New Zealand develop type 2 diabetes at a higher rate and a younger age, and have 3.8 times higher incidence of end‐stage renal disease (ESRD). Objective To investigate contextual factors that shape understandings of disease for Pacific peoples with diabetes and ESRD. Methods Focussed ethnography. In‐depth interviews were conducted with 16 Pacific people on haemodialysis for diabetic ESRD, in Auckland, New Zealand. Study participants aged between 30 and 69 years old were of Samoan, Cook Islander, Tongan, Niuean or Tokelauan ethnicity. Thematic analysis was used to code and identify themes. Results Participants were embedded in a multigenerational legacy of diabetes. The limited diabetes‐related education of earlier generations influenced how future generations behaved and understood diabetes. Perceptions were compounded by additional factors including the invisibility of early‐stage diabetes; misunderstandings of health risks during communication with health providers; and misunderstandings of multiple conditions’ symptoms and management. Participants had limited engagement with health services until their diagnosis of ESRD acted as a trigger to change this behaviour. However, this trigger was not effective in itself—rather, it was in combination with relevant education delivered in a way that made sense to participants, given their current understandings. Conclusions Illness representations drive choices and behaviours with respect to self‐management of diabetes and engagement with health services. Diabetes is often present in multiple generations of Pacific people; therefore, illness representations are developed and shared within a family. Changing illness representations requires engagement with the individual within a family context.


Compared with New Zealand Europeans, Pacific peoples in New
Zealand develop type 2 diabetes mellitus (diabetes) at a higher rate 1 and a younger age, and have a 3.8 times higher incidence of endstage renal disease (ESRD). 2 More than two thirds of Pacific ESRD is due to diabetes compared with just more than one third in New Zealand Europeans. 2 Furthermore, diabetes and ESRD are increasingly being seen across multiple successive generations of Pacific families. Our findings demonstrate how understandings of diabetes are shaped by multigenerational family beliefs and behaviours and are further compounded by misunderstandings and limitations to engagement with health providers.
Control of type 2 diabetes, sufficient to reduce complications, requires constant self-management which becomes more complex in the presence of complications and comorbidities. Effective selfmanagement requires health literacy and self-management skills supported by patient-provider communication, which, to be effective, must respond to culture-specific understandings of diabetes. 3 As a key determinant of health, 4,5 health literacy is critical to personal empowerment. In its broadest sense, health literacy is a 'composite' term that can include 'knowledge' and 'understanding' as targets of health promotion activities. 4 Health care is complex and health literacy is best defined in terms of what it enables us to do, 6 noting that people with adequate general literacy may not necessarily have adequate health literacy. 7 Furthermore, health literacy is constructed within a social context-in environments that include families, communities and social networks. 4,8,9 Family context is pivotal to an individual's understanding, beliefs and expectations (illness representations) of diabetes. 10,11 Because diabetes is epidemic, and commonly present across multiple family generations, judgements, (biased) perceptions and myths associated with disease can develop within families. 11 What is observed within the family becomes the 'appropriate' or normalized way to care for, or selfmanage, diabetes; thus, the platform is established for an individual's foundational knowledge and illness representations 12 about diabetes.
Diabetes is largely asymptomatic in the years preceding a diagnosis or even to the onset of complications. This invisibility can lessen an individual's motivation to engage in diabetes self-management. The difference between an individual's illness representations [10][11][12] and a clinical assessment can strongly influence adherence or commitment to recommended treatments. [13][14][15] Leventhal and colleagues propose a model of beliefs in five domains to conceptualize how illness representations determine judgement of illness and health behaviour. 12 These beliefs address the label or diagnosis of diabetes and its associated symptoms (identity); possible causes of diabetes (causes); duration of diabetes (timelines); perceived physical, psychological or financial consequences of diabetes (consequences); and feelings of self-efficacy for managing illness (controllability). Perceptions of consequences and controllability 14 are prominent within the literature of diabetes outcomes where increasing evidence associates a family history of diabetes with a sense of inevitability and lack of controllability. 10,14 The traditional Pacific Island worldview of health differs from a Western or European perspective. Expressed as a notion of maintaining social order and harmony, health is viewed as a holistic concept of well-being (a balance of the body, spirit and mind), incorporating spirituality, and values and obligations centred around family kinship and communalism. 16 Although the importance of family among Pacific has been maintained since Western contact, there have been noticeable shifts in cultural orientation and health perceptions. 17

Concerns about general Pacific health and well-being in New
Zealand are not new. [21][22][23][24] National prevention strategies have identified disparities in access, provision and coordination of health services, 25,26 with government and sector reports proposing interventions to improve health outcomes for Pacific peoples. [27][28][29] A low degree of engagement with health services by Pacific peoples has been identified as a central reason underpinning their much poorer health status compared with the general New Zealand population. 27 Reasons for low engagement include issues relating to communication, lack of trust and experiences of discrimination influencing decisions to engage with services and low health literacy. 15,27,29,30 For these reasons, commentators suggest that there is a need to determine baseline levels of knowledge and understanding of diabetes within Pacific Island communities so that effective strategies for patient engagement can be developed. 27 However, research relating to low health literacy often centres upon 'what is understood' rather than 'what are the underlying contexts and constructs explaining why things are understood in that way'. 9 The research reported here forms part of a larger focused ethnographic study that sought to address the question 'what is the lived reality of illness for Pacific peoples with diabetes?' In this article, we investigated the contextual factors that shape understandings of disease for Pacific peoples with diabetes and ESRD.

| Design, setting and sampling
This qualitative enquiry employed a focused ethnographic approach, which, unlike traditional ethnography, examines a single issue or experience of a subgroup of people within a context-specific setting. 31 The larger study aimed to understand the contexts and meanings underpinning self-management behaviour-through participant accounts of living with diabetes. As an applied research method, focused ethnographic approaches facilitate better understandings of complexities surrounding an issue or phenomenon from the perspective of the participant. 32 In this context, the researcher can examine personal and societal factors that influence, for example, decision making. Cruz and were selected because the researchers believed the original meanings within participant responses would be retained and not be lost during translation between languages. Study information packs were made accessible in general spaces within the units for prospective patients to read and self-nominate if they wished to take part in the study.

| Data collection
Data were collected by the first author (JS-B) who made field observations, conducted face-to-face interviews and recorded field notes.
Interviews were between 30 and 80 minutes, in English, and undertaken while participants were undergoing haemodialysis and were either lying on their bed or sitting semi-recumbent in a chair. The interviewer stopped the shortest interview after noticing the participant was becoming increasingly breathless. Three interviews were re-scheduled due to general unwellness, despite participants in each case wanting to tell their story. Those three interviews were conducted at their next dialysis treatment. Another three participants wanted to be interviewed during the initial recruitment meeting 'because I might not be here next week' (Sa.2). A relaxed conversational approach easily accommodated interruptions from nurses who attended to alarms or checked monitor readings. The interview aimed to explore: understandings of diabetes and kidney disease; health information and communication; self-management; and reflections on diabetes (Table 1). Written informed consent was gained prior to interview, and all participants consented to their interview being audio-recorded. Interviews were transcribed, and interview data were managed using Nvivo12-Pro and Microsoft Excel.

| Data analysis
A thematic analysis approach informed by Saldaña 35 was used to identify contextual factors underpinning participants' understandings of diabetes and ESRD. All identifying information on transcripts were anonymized prior to being read by those other than the interviewer. Transcripts were read several times by members of the wider research team. Data were categorized using first cycle coding (descriptive or basic labelling) prior to being reorganized into groups  The study was guided by our obligation to conduct trustworthy research throughout the enquiry. 36 As an important component of trustworthiness, credibility was supported by prolonged engagement, peer debriefing, insider knowledge, crystallization and reflexivity. Field observations were also recorded along with reflective notes and memos to minimize distortion of participant information. 37 Furthermore, to enable participants experiences to become meaningful and resonant to the reader, 38 rich descriptive accounts of experiences and the context within which they occurred are provided.

| RE SULTS
Of the 42 potential participants who indicated an interest in the study, 9 did not meet the inclusion criteria and 17 withdrew due to hospitalization (2), feeling unwell (4) and a change of mind (11).
Thirteen were Island-born. All were aged between 37 and 68 years, and half were female. Nine had three or more comorbidities in addition to ESRD (   Yeah they told me about it but it was more just giving information…all about diabetes, not 'me with the diabetes'…But when I started this I've had so much help, like dieticians, the nurses telling us, educating us. If we can prevent that even before we come on this, if we had the same help before like around the diabetes and stuff, I think I wouldn't even be on this.

| Symptoms to diagnoses
(Sa.5) When asked about what might prevent future generations developing diabetes and ESRD, the majority advocated towards including more education about the interactions and management of multiplecomorbidities, and using different modes of delivery: To learn about the different diseases eh, how diabetes affects the heart and the heart blood pressure the diabetes and my legs. These kids don't know those things. (Ck.6) Sometimes people need to experience the sickness, like we do when the people died, they body is dead, but it's real eh? Yeah, the same thing to sickness. I come here and you know the smell, sometimes it's worse for the leg cut off but not healing, aawww I know it!. (Ni.1)

| D ISCUSS I ON
These findings explore the experiences of living with diabetes, and its progression to ESRD, for 16 Pacific Island adults in New Zealand.
Irrespective of age, gender or ethnicity, participants recounted similar experiences in relation to their knowledge of diabetes, diabetes self-management and illness representations within their families.
The diagnosis of ESRD created a disruption that impacted them emotionally, physically and socially. Although they accepted this reality, participants blamed themselves for making bad decisions about health care and were angry that they had missed earlier opportunities to manage their diabetes.
The literature repeatedly confirms that health-care decision making is influenced by an individual's perception(s) of their control over diabetes and the severity of consequences related to becoming ill. 10 Relevant knowledge shifted participants from responding to acute symptoms to managing diabetes in the longer-term. This shift was similarly seen in a New Zealand study of Pacific adults with chronic obstructive pulmonary disease (COPD) who also initially focused on present symptoms rather than longer-term strategies. 41 However, unlike the early stages of diabetes, the early symptoms of COPD such as shortness of breath and production of sputum are more obvious and persistent. For diabetes, the invisibility of symptoms can often lead to unpredictable or no response from the individual and therefore contribute to progressive complications and disabilities. 42 A growing literature supports the significant potential contribution of health literacy to reducing the burden of disease. 4,5,43,44 Robust evidence highlights the need for individuals to gain contextually relevant knowledge, skills and abilities to self-manage health conditions, especially for migrant and refugee populations where linguistic and cultural barriers often exist. 45 Health literacy has recently been established as 'an identifiable and manageable risk in clinical care ', 9 p.3 specifically within the management of chronic and complex conditions that depend upon successful patient engagement. While, previously, health literacy education centred on the individual's skills and capabilities, the focus has shifted towards the context within which health information is obtained, understood and acted upon. 9 The WHO argue relevant health communications must help individuals 'understand that there is a health risk for themselves… [and] that the risk could be severe and that they reduce that risk by undertaking recommended actions'0. 46 However, diabetes education programmes are often pre-packaged to convey a general message, commonly focussing on the dominant population or group. In the current study, participants perceived these sessions as too generalized, whereas education tailored towards their own health needs was considered effective and empowering.
Messages tailored to a specific person or audience have been shown to be effective in a wide range of health outcomes including diabetes 47 and ESRD. 48 As a common element within health literacy models, 8

| Limitations and practice implications
Participant self-reports often covered a long period of time. Risk of recall bias was reduced through not asking for specific dates but rather asking about events in relation to one another, for example experiences before or after diagnoses of diabetes or ESRD. We do not claim participant responses are representative of all Pacific peoples. However, given the increasing global burden of non-communicable diseases on Pacific (and migrant) populations, 53 this study may contribute to on-going development and risk management associated with patient engagement and disease management.
To consider family influences, we purposefully explored partic-

CO N FLI C T O F I NTE R E S T
None.

E TH I C A L A PPROVA L
Institutional ethical approval was obtained from the University of Auckland Human Participants Ethics Committee: Reference 020658.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets generated and analysed during the current study are not publicly available as participants did not consent to transcripts of interviews being shared.