Perspectives on decision making amongst older people with end‐stage renal disease and caregivers in Singapore: A qualitative study

Abstract Background End‐stage renal disease (ESRD) is increasing both globally and in Asia. Singapore has the fifth highest incidence of ESRD worldwide, a trend that is predicted to rise. Older patients with ESRD are faced with a choice of haemodialysis, peritoneal dialysis or conservative management, all of which have their risks and benefits. Objective This study seeks to explore perspectives on decision making amongst older (≥70) Singaporean ESRD patients and their caregivers to undergo (or not to undergo) dialysis. Design Qualitative study design using semi‐structured interviews. Setting and participants Twenty‐three participants were recruited from the largest tertiary hospital in Singapore: seven peritoneal dialysis patients, five haemodialysis patients, four patients on conservative management and seven caregivers. Results While some patients believed that they had made an independent treatment decision, others reported feeling like they had no choice in the matter or that they were strongly persuaded by their doctors and/or family members to undergo dialysis. Patients reported decision‐making factors including loss of autonomy in daily life, financial burden (on themselves or on their families), caregiving burden, alternative medicine, symptoms and disease progression. Caregivers also reported concerns about financial and caregiving burden. Discussion and conclusion This study has identified several factors that should be considered in the design and implementation of decision aids to help older ESRD patients in Singapore make informed treatment decisions, including patients' and caregivers' decision‐making factors as well as the relational dynamics between patients, caregivers and doctors.


| INTRODUC TI ON
With trends showing increasing incidence, end-stage renal disease (ESRD) is a growing public health concern globally. 1 In Singapore, a small, densely populated and rapidly ageing city-state, ESRD incidence has almost doubled between 1999 and 2015, 2 making it the fifth highest incidence worldwide. 3 These figures are expected to continue rising due to population ageing and a high local prevalence of diabetes. 4 Haemodialysis (HD) and peritoneal dialysis (PD) are primary therapies for ESRD, and evidence from Singapore suggests that for younger and healthier ESRD patients, PD is the most cost-effective treatment choice. 5 However, there is increasing evidence of the lack of survival advantage of dialysis over conservative management (CM), which focuses on symptom management, for ESRD patients over the age of 75 and with multiple comorbidities. 6,7 Given that there is no clear optimal treatment strategy for older ESRD patients, many factors may influence their decision on which treatment to pursue. These include average life expectancy, quality of life, caregiver burden, availability of transport and cost. 8,9 As ESRD patients and their caregivers navigate these choices, there is a need for supportive, contextually appropriate decisionmaking aids to enable informed decision making. Evidence from Singapore has shown that lack of knowledge of CM and deference to physician recommendations may lead older ESRD patients to seek dialysis 10 and that a small proportion of patients later regret their decision to start dialysis. 11 Work done elsewhere has also shown that most older patients have unrealistic expectations of dialysis upon commencing treatment, 12 and has suggested the need for age-specific information tailored to older ESRD patients that clearly communicate the risks, benefits and burdens associated with dialysis. 13 Presently, there is limited research exploring the treatment decision-making process of older ESRD patients and caregivers, and the majority of current literature is centred on Western populations.
As rates of ESRD and associated risk factors rise in Asian populations, it is important to better understand the decision-making process of patients and caregivers within Asian contexts. Singapore's culture and policy structures are uniquely and deeply rooted in the principles of self-reliance and family as the first line of support. 14,15 Individual and family responsibility are emphasized in health-care policies such as Medisave, a mandatory saving scheme that requires citizens to set aside a portion of their income to pay for their own or immediate family members' medical expenses. 16 This paper thus seeks to explore perspectives on decision making amongst older Singaporean ESRD patients and their caregivers to inform the development and implementation of contextualized and culturally appropriate decision-making aids.

| MATERIAL S AND ME THODS
This qualitative study followed the guidelines of the Consolidated Criteria for Reporting Qualitative Studies (Appendix S2).

| Setting
The study was conducted at Singapore General Hospital, the largest tertiary hospital in Singapore.

| Participants and recruitment
This study used two methods to sample key informants: purposive sampling to recruit patients with ESRD and snowball sampling to recruit their caregivers. The inclusion criteria were as follows: patients had incident chronic kidney disease Stage 5 (glomerular filtration rate < 10 mL/min), were aged 70 years or older and were currently receiving HD, PD or CM. Caregivers were primary informal Participants were approached and recruited face-to-face by researchers (EH, JK, VH, RQ) during appointments at Singapore General Hospital with the assistance of renal coordinators, nurses and doctors. In total, 62 participants were approached; 38 declined or did not respond, and one dropped out due to hospitalization. Reasons for refusal included being too tired and frail to have a prolonged conversation or not wanting to talk about their experience. Twentythree participants from four participant categories were included in the study: seven PD patients, five HD patients, four patients on CM and seven caregivers. Of these participants, 14 were female and nine were male. All but three participants were Chinese. Patients had a mean average age of 75 (range 71-85). Figure 1 displays a flow chart of the participant recruitment process, and Table 1 shows the basic demographic characteristics of each included participant.

| Data collection
Semi-structured, in-depth interviews were conducted with all 23 participants by researchers (JK, EH, RQ, VH, HLQ) trained in the study protocol, interview guide and qualitative research methods. Interviews were conducted in English (n = 8), Chinese (n = 9) or Malay (n = 1) by researchers fluent in that language. All caregivers were interviewed jointly with patients except two who were interviewed alone as the patient declined to participate in the study. Efforts were made to ensure that both the caregiver and patient had the opportunity to contribute equally in these dyadic interviews, 17 such as probing for the other participant's opinions if only one of them responded initially.
Interviewers followed an interview guide (Appendix S1) while conducting interviews. The guide included topics such as sociodemographics and caregiver support, medical history and medications, doctor-patient relationships, decision-making processes and lived experiences. Interviews primarily took place in participants' homes or waiting areas within the hospital (at the request of participants). Each interview lasted approximately 30-60 minutes. Audio recordings and field notes were taken during all the interviews. Data collection ceased when the researchers were confident that data saturation had been reached, which is defined by Saunders et al 18 as the point at which no new information is being generated by additional interviews.

| Ethics
Informed consent for participation and recording was obtained before the interviews started, and participants signed a Participant All interview materials were also stored securely to ensure data confidentiality. Each excerpt in this paper includes the number of the interview and code letters (F for female, M for male) so that extracts from the same individual can be linked. To maintain anonymity, all names reported are pseudonyms and identifying data have been excluded.

| Reliability and validity
In establishing the rigour of our enquiry, we employ the four trustworthiness criteria for qualitative research developed by Lincoln and Guba: credibility, dependability, confirmability and transferability. 19 To increase the credibility and dependability of our research, investigator triangulation and stepwise replication were implemented. This involved multiple investigators analysing the same data set separately, comparing their interpretations of the data and resolving inconsistencies. Where interpretations differed, the investigators discussed until consensus was reached on the interpretation that was most consistent with the original meaning of the data. Confirmability of the findings was also secured by requiring all investigators to keep reflexive notes and participate in regular meetings to ensure that interpretations of the findings were not influenced by personal opinions and biases but were grounded in the data. Finally, the transferability of our study has been facilitated by the use of purposive sampling and detailed descriptions of the study methodology and context.

| Patient and public involvement (PPI)
Patient and public involvement in research is not a common practice in Singapore for a variety of reasons, including lack of funding.
Consequently, patients were not involved in the design and conduct of the study, interpretation of results or the writing of this paper.

| Data analysis
Interviews were recorded and then translated into English (if necessary) and transcribed in full. To ensure reliability, professional bilingual transcriptionists who were familiar with local culture and colloquialisms were hired to translate and transcribe the interviews simultaneously. The respective interviewers (EH, JK, RQ) then double-checked all transcripts against the audio recordings and rectified any missing or misinterpreted information to ensure semantic and conceptual accuracy. This approach was preferred to back and forward translation, which is usually insufficient to obtain equivalence in meaning. 20 Three researchers (EH, JK and VH) coded interviews using QSR Nvivo 11 Software utilizing a combination of inductive and deductive thematic analysis. A priori themes were first derived from the research aims, interview questions and previous literature.
Next, a posteriori themes were generated from the data and refined using techniques from grounded theory, including the constant comparative method and line-by-line analysis to test preliminary assumptions. 21

| Conceptual framework
We used a conceptual framework to guide our analysis of patient and caregiver treatment decision making ( Figure 2). We first considered the role of socio-economic factors, as well as doctor-patient and caregiver-patient relationships, in the decision-making processes of older ESRD patients and their families. [22][23][24] We then used emergent themes to better inform the structure and content of our framework. Table 2 presents a summary of the key themes, subthemes and examples of evidence. Firstly, we will describe the roles of patients, families and doctors in the decision-making process. Secondly, we will explore the factors that informed patients' and families' initial decision to accept or decline dialysis, as well as factors that caused patients who initially declined dialysis to start it. All quotes by patients are presented with the accompanying ID number, pseudonym, gender, age range, ethnicity and treatment type (HD/PD/CM).

| RE SULTS
Quotes by caregivers are presented with their ID number (matched with the patient's), pseudonym, gender, relationship to the patient and the patient's treatment type.

| Roles of individuals involved in the decisionmaking process
Some patients reported that they had made an independent treatment decision, while others reported feeling like they had no choice in the matter or that they were strongly persuaded by their doctors and/or family members to undergo dialysis.

| Role of patient
Many patients stressed that the final decision to accept or reject dialysis was theirs to make and that they made it on their own, even if they had consulted with health-care professionals and family members during the decision-making process: This insistence on making an independent decision largely stemmed from a notion of self-reliance and taking personal responsibility for one's own problems. Patients expressed a reluctance to implicate others in the consequences of their treatment decision:

TA B L E 2 Key themes and examples of evidence
Interviewer: Back then, when you were asked to undergo dialysis, did you discuss it with your children?
Patient: Yes, but I was told to make the decision myself.
This decision cannot be made by others. Interviewer: Please explain more. Patient: If they make the decision for me to undergo dialysis, it means they will have to take care of me.

| Role of family
These narratives reflected a process of negotiation between patients and families and showed the powerful influence that families had over patients' decisions. On the other hand, some patients described accepting the advice of their doctors to undergo dialysis within a positive doctor-patient relationship as doctors made an effort to encourage, reassure and support them during the decision-making process:

| Role of doctor
He would use positive words to help us and advise us.
It's because of him that I accepted dialysis. Otherwise, I wouldn't have accepted it.
[PC42_Chia_M_71-75_Chinese_PD]  The quotes reveal that for these patients, it was quality rather than quantity of life that featured as a leading consideration when making their treatment decision.

| Financial burden
The financial burden of dialysis was raised as a significant decision-making factor by many patients, who perceived that treatment expenses would be extremely burdensome for themselves or their families.
For example, Irene, who is currently still deciding whether to start dialysis, shared that the cost of dialysis is her primary consid- In short, while many patients were hesitant to start dialysis because of the perceived financial burden, the family caregivers in our study did not regard the cost of dialysis as a reason to reject it but rather as a challenge they should and could find ways to cope with.

| Caregiving burden
Patients recognized that their decision to undergo dialysis would give rise not just to financial but also caregiving obligations for their family, and this was a major factor in their decision making. In short, while the caregiving burden associated with dialysis was a prominent reason for patients to reject it, the families in our study showed a willingness to shoulder this burden and even persuaded patients to undergo dialysis.

| Alternative medicine
Furthermore, some patients stated that they preferred to try alternative medicine, in particular traditional Chinese medicine (TCM), before accepting dialysis. Only when TCM was perceived to no longer be an effective mode of treatment because of the severity of their conditions did these patients resort to dialysis.
In particular, Vivian understood that TCM would not provide a cure for her kidney disease, but she still sought TCM to delay the initiation of dialysis: The [TCM] doctor there said that, if there's need for dialysis, at last, you also must go. Not to say that the Chinese herbs will cure you, will help you to improve anything… just to control it, that's all…not enough… doctor does say that. [PB36_Vivian_F_76-80_Chinese_HD] On the other hand, Chiang Tee claims that he has been cured by a folk remedy that was recommended to him by a friend and available on the Internet, which is why he believes that he does not need to undergo dialysis anymore: After that, someone introduced to me to a concoction of lychee seeds and pig kidney. 7

| Symptoms and disease progression
Finally, many patients explained how their decision to initiate dialysis was strongly influenced by increasingly severe symptoms that occurred alongside disease progression. This was a determining factor for several patients who initially declined but subsequently accepted dialysis.
For example, Siew Leng described how she originally rejected dialysis because she was young and symptom-free, but extreme swelling in her body eventually drove her to initiate dialysis because she could no longer walk: [CPC42_Ying_F_PD_Spouse] In other words, many patients only decided to start dialysis when their symptoms were beginning to affect their everyday functioning and quality of life, which is consistent with their concerns regarding loss of autonomy in their daily life.

| D ISCUSS I ON
This qualitative study explored perspectives and experiences of the decision to undergo dialysis amongst older ESRD patients and their caregivers in Singapore ahead of a study to design ESRD patient decision-making aids. We organized our findings based on a framework to understand the patient decision-making journey and the various factors that influence this decision. Based on our findings, we make some recommendations regarding the creation and implementation of decision-making guides.

| Key findings and recommendations
Patients' decision-making process involved weighing the per- the perceived pain and suffering caused by dialysis. [25][26][27][28][29] Our study adds that many older ESRD patients in Singapore seek alternative medicine such as traditional Chinese medicine (TCM) as a means to delay dialysis, and the main reason why many eventually decide to start dialysis despite initial rejection is to relieve intensifying symptom burden (rather than to extend their lives). Our study also

| Strengths and limitations
This study is the first study to focus on the perspectives and experiences of ESRD patients above the age of 70 in an Asian population.
It is also the first study to include the views of family caregivers to capture the complex inter-relational dynamics in decision making for this population. A further strength of our study was the inclusion of participants from multiple ethnic backgrounds (Chinese, Malay and Indian), with interviews conducted in the language of participants' preference, which adds to the diversity of experiences and factors reported.
Nonetheless, our sample was still largely dominated by Chinese Singaporeans, who make up more than 75% of Singapore's total population. 45 Another limitation of the study is the underrepresentation of certain patient groups including those taking conservative management for ESRD from community clinics. Our patient population was also comprised of older patients, who have been found to report higher levels of satisfaction with care than younger groups. 46 Furthermore, many patients approached had declined to be interviewed, due to either being too frail to hold a long conversation or not wanting to talk about their treatment choice. This could have skewed the views reported in this study, which reflects the perspectives of patients who were more open to talking about their decision-making experience. Lastly, there was a lack of PPI in this study, as the practice is not commonly adopted in Singapore. Given that our study sought to understand patient decision making and inform patient decision-making aids, PPI could have helped to improve the relevance of our research to patients, strengthen the validity of our findings and recommendations, as well as increase the usefulness of our research outputs. The potential of introducing patient and public involvement to research in Singapore thus merits serious consideration.

| CON CLUS ION
This study has identified several factors from patient and caregiver perspectives to consider in the creation and implementation of decision-making aids for older ESRD patients in Singapore. These include

ACK N OWLED G EM ENTS
We sincerely thank Ms Suan Ee Ong for helping to conduct a Malay interview and all the participants for sharing their time and insights.

CO N FLI C T O F I NTE R E S T
None declared. Jafar conceived and designed the study and revised the manuscript.

AUTH O R S ' CO NTR I B UTI O N S
Farah Shiraz revised the manuscript. Helena Legido-Quigley conceived and designed the study, collected the data and revised the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
Participant consent and ethical approval have not been obtained for data sharing. Due to the confidential and sensitive nature of our data, there are no data that can be disclosed beyond that contained within this paper.