Cultural adaptation of the serious illness conversation guide for caregivers of persons with dementia in Singapore Authors: Alston Ng, Vignesh Sivasamy, Shirlyn Hui Shan Neo Published: 2025 ABSTRACT AbstractIntroductionGoals-of-care (GOC) discussions are often delayed or uninitiated for people with dementia, potentially contributing to inappropriate end-of-life treatment. The Serious Illness Conversation Guide (SICG) facilitates structured GOC discussions about patient values and care preferences, but a dementia-specific Guide adapted to the Asian cultural context has yet to be developed. This study aimed to adapt and iteratively revise the SICG-C, and assess its acceptability among caregivers of people with dementia.MethodsWe adapted the SICG for caregivers of people with dementia in Singapore using a three-phase, approach. Phase 1 involved adapting the patient-facing SICG to caregiver-facing language (version 1). In Phase 2, we iteratively tested the guide with 6 and 13 caregivers in 2 rounds of interviews to derive the final version of the Guide. In Phase 3, 50 caregivers rated each SICG-C item from Versions 1 and 3 using a structured questionnaire, with presentation order alternated to minimise order effects.ResultsSixty-nine caregivers were recruited in this study. Qualitative interviews revealed that caregivers valued professionalism and reassurance from clinicians, while also emphasising their own experiential knowledge as distinct from clinical expertise. Many struggled to articulate their loved ones’ priorities due to cognitive decline and the absence of prior discussions, often expressing uncertainty about whether their assumptions reflected what their loved ones would have wanted. Caregivers were sensitive to language implying poor prognosis and preferred phrasing that conveyed continuity of care and partnership. None of the caregivers had been informed by their loved ones of their goals and priorities, reaffirming the value of the SICG-C. In the final survey (n = 50), caregivers showed a strong preference for Version 3 over Version 1, with 21 of 22 items rated significantly higher (all p<0.001). The final SICG-C was rated highly acceptable.ConclusionThe SICG-C enables clinicians to initiate timely, values-based GOC discussions with caregivers of people with dementia. Its structured, sensitive language addresses caregiver concerns and may be used to support proactive care planning in dementia. Further research is needed to evaluate implementation in routine clinical practice.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12877-025-06761-4.Keywords: Serious illness care, Dementia, Values and preferences, Goals of care, Communication Introduction Dementia is a progressive and life-limiting disease that affects more than 55 million people globally, a number set to reach 139 million by 2050 [1]. A known challenge in dementia care is the tendency for persons with dementia to receive overtreatment at the end of life, including burdensome life-sustaining interventions that offer limited benefit to survival and quality of life [2–6]. Early initiation of goals-of-care discussions are thus important in ensuring that the care people with dementia receive aligns with their values, priorities, and preferences. But these conversations rarely take place or occur too late in the illness trajectory, with only about 3–39% of people with dementia having undertaken some form of advance care planning (ACP) globally [7]. Initiation of goals-of-care discussions with people with dementia and/or their surrogate decision-makers is impeded by multiple factors [8, 9]. Dementia, for instance, may be assumed to be a natural part of ageing by people with dementia and their loved ones [9–12], a misconception that contributes to the perceived irrelevance of goals-of-care discussions and differences in end-of-life care preferences between people with dementia and their loved ones [13]. The Serious Illness Conversation Guide (SICG) is a structured communication tool developed by Ariadne Labs and tested with many patient populations across various health systems. The available evidence suggests that clinicians who use the Guide have better, earlier, and more conversations with their patients [14, 15]. Other beneficial outcomes include enhanced patient satisfaction with care, reduced symptoms of anxiety and depression in patients [16–18], and lower rates of acute care utilisation [19, 20]. However, existing trials primarily focus on patient-clinician conversations; limited research has been conducted on serious illness conversations (SIC) that involve surrogate decision-makers, a gap especially salient to the delivery of quality dementia care. The only dementia-specific adaptation of the SICG currently available is the version developed by Berry and colleagues for use in the American context [21]. However, a review by the study team raised concerns about its acceptability and appropriateness in an Asian context. This study thus builds on a locally adapted version of the SICG [22] to iteratively test and refine a caregiver-oriented SICG (SICG-C), with the longer-term goal of supporting clinicians in eliciting the priorities and care preferences of people with dementia from their caregivers and delivering patient-centred, values-aligned dementia care. Study aims This study aimed to. (i)Adapt and iteratively revise the SICG-C; and. Adapt and iteratively revise the SICG-C; and. (ii)Assess the acceptability of the finalised SICG-C among caregivers of people with dementia. Assess the acceptability of the finalised SICG-C among caregivers of people with dementia. Methods Study design This study employed a three-phase, mixed-methods design combining qualitative and quantitative methods to adapt and assess the acceptability of the SICG-C. Participants and setting Participants were adult caregivers of people with dementia, recruited from the outpatient Geriatric Medicine clinic at Singapore General Hospital (SGH). Eligible participants aware of the patient’s diagnosis, primary caregivers or spokespersons of the patient, able to speak English or Mandarin, and ≥ 21 years in accordance with Singapore’s Human Biomedical Research Act. People with dementia and domestic helpers were excluded. Purposive sampling was used to ensure diversity in caregiver demographics and dementia severity among their care recipients. Written informed consent was obtained from all participants prior to initiation of study activities. Instrument The SICG-C was adapted from the locally validated English-language Serious Illness Conversation Guide [22] and informed by Berry et al.’s SICG-Dementia [21] and various clinician-developed communication tools used in local practice. In addition, the Functional Assessment Staging Tool (FAST) was employed to characterise the stage of dementia in people with dementia. The FAST is a validated, seven-stage global measure of functional decline in dementia and is widely used in both clinical and research settings [23]. English was the sole language used for Guide adaptation and all study procedures; no translation was required as all participants were fluent English speakers. Procedures Phase 1: initial adaptation A first draft was developed by AN, a non-clinician researcher, and reviewed by VS and SN, who drew on their clinical experience to anticipate and address potential concerns raised by caregivers. Phase 2: iterative testing and refinement Two rounds of interviews were conducted. In Round 1 (October 2024), we recruited 6 caregivers for structured interviews to rapidly identify key areas for improvement and generate Version 2 of the Guide. In Round 2 (November 2024), we recruited 13 caregivers to reach thematic saturation. During these interviews, participants were read each item from the SICG-C and asked to rate its acceptability using a three-point scale: 1–Unacceptable; 2–Acceptable with changes; 3–Acceptable as is. They were encouraged to elaborate on their ratings, with attention to clarity, tone, and associations evoked. In Round 2, participants were shown both versions of each item (Version 1 followed by Version 2) and asked to evaluate them independently to reduce the risk of contrast bias. Interviews were audio-recorded, transcribed, and thematically analysed to guide revisions after each round. Where participant feedback diverged, the interviewer clarified the reasoning behind each preference and tested alternative phrasings in real time. The study team reviewed these discussions collectively, considering the totality of feedback across participants. Phase 3: final acceptability assessment A structured questionnaire comparing Versions 1 and 3 using a 4-point Likert scale (1= “Very unacceptable” to 4= “Very acceptable”) was administered. The scale was expanded to enhance response sensitivity and remove a neutral midpoint upon finalisation of Guide wording. Due to the lack of standardised benchmarks for acceptability, we defined an item as acceptable if (i) ≥ 75% of participants rated it 4 (“Very acceptable”) and (ii) ≥ 95% of participants rated it either 3 (“Acceptable”) or 4. A target sample size of 50 was selected to provide adequate power for Wilcoxon signed-rank tests, assuming moderate effect size (P(X > Y) = 0.67), α = 0.05, and power = 0.80. To minimise order effects, two questionnaire versions were developed: in one, Version 1 was presented first for each item tested; in the other, Version 3 appeared first (see Supplementary Appendix 1). These were alternately administered to each successively recruited participant. Questionnaires were self-administered or completed with assistance from the recruiter upon request. Recruitment and questionnaire administration took place between March and May 2025. Data analysis Interview data from Phase 2 were analysed using a hybrid thematic analysis combining deductive and inductive coding. As interviews were structured around the wording of Guide items, initial coding followed a deductive framework based on item sequence and acceptability, with inductively derived codes added to reflect emergent caregiver perspectives. Themes were developed through iterative review of participant quotations, focusing on expectations and experiences related to caregiving and healthcare communication. The first author (AN) conducted primary coding and theme generation across both rounds in Phase 2, with co-authors reviewing codes and themes to ensure analytical coherence and interpretive accuracy. Revisions to the Guide were finalised through consensus, guided by the principles of accessibility, emotional sensitivity, and alignment with the overarching goal of facilitating meaningful conversations about the values and goals of the people with dementia. Quantitative survey data from Phase 3 were analysed using descriptive statistics on Excel and Wilcoxon signed-rank tests in R (version 4.2.0). Results A total of 69 participants were purposively recruited (Table 1). Two participants from Phase 2 Round 1 and two from Phase 2 Round 2 were incidentally re-approached for survey in Phase 3 during routine clinic visits, yielding a sample of 65 unique participants. Table 1Participant characteristicsRound 1 (n = 6)Round 2 (n = 13)Round 3 (n = 50)Females: n (%)5 (83.3%)7 (53.8%)37 (74.0%)Age: Mean (sd)59.5 (10.4)57.2 (14.2)53.3 (9.5)Ethnicity: n (%) Chinese5 (83.3%)10 (76.9%)40 (80.0%) Indian1 (16.7%)2 (15.4%)4 (8.0%) Malay0 (0%)1 (7.7%)6 (12.0%)Religion: n (%) Buddhist1 (16.7%)2 (15.4%)14 (28.0%) Christian3 (50.0%)5 (38.5%)8 (16.0%) Free thinker1 (16.7%)3 (23.1%)13 (26.0%) Hindu1 (16.7%)1 (7.7%)3 (6.0%) Muslim0 (0%)2 (15.4%)7 (14.0%) Others0 (0%)0 (0%)5 (10.0%)Highest educational qualification: n (%) Primary school0 (0%)0 (0%)1 (2.0%) Secondary school1 (16.7%)1 (7.7%)10 (20.0%) Post-secondary5 (83.3%)12 (92.3%)39 (78.0%)Occupational status: n (%) Working full-time4 (66.7%)5 (38.5%)31 (62.0%) Working part-time0 (0%)1 (7.7%)4 (8.0%) Unemployed0 (0%)2 (15.4%)8 (16.0%) Retired2 (33.3%)5 (38.5%)7 (14.0%)Relation to patient/person with dementia: n (%) Child3 (50.0%)10 (76.9%)42 (84.0%) Sibling1 (16.7%)0 (0%)2 (4.0%) Spouse0 (0%)0 (0%)2 (4.0%) Others1 (16.7%)3 (23.1%)4 (8.0%)Patient’s dementia subtype: n (%) Alzheimer’s5 (83.3%)9 (69.2%)26 (52.0%) Frontotemporal0 (0%)0 (0%)1 (2.0%) Mixed1 (16.7%)3 (23.1%)16 (32.0%) Vascular0 (0%)1 (7.7%)7 (14.0%)Patient’s dementia staging (FAST): n (%) Mild1 (16.7%)5 (38.5%)22 (44.0%) Mild-to-moderate0 (0%)1 (7.7%)7 (14.0%) Moderate3 (50.0%)5 (38.5%)18 (36.0%) Moderate-to-severe1 (16.7%)2 (15.4%)2 (4.0%) Severe1 (16.7%)0 (0%)1 (2.0%) Participant characteristics Across all rounds, participants were mostly female (ranging from 53.8% in Round 2 to 83.3% in Round 1), with mean ages varying between 53.3 years in Phase 3 and 59.5 years in Phase 2 Round 1. Most participants were of Chinese ethnicity and had completed post-secondary education. The majority were employed full-time and were adult children of the person with dementia. Alzheimer’s disease was the most commonly reported diagnosis, followed by mixed dementia and vascular dementia. The majority of people with dementia were in the mild or moderate stages of dementia based on clinically documented FAST staging. Rating tabulation for phase 2 round 1 A total of 21 items were tested with 6 caregiver participants. Only 10 items were rated “Acceptable as is” (rating = 3) by at least 5 participants (> 75%), of which 4 items were related to summative or concluding remarks found at the end of the Guide. Key areas for improvement included exploration of acceptable tradeoffs (Serial Number (S/N) 1–16), critical abilities (S/N 1–14, 1–15), understanding of patient’s condition (S/N 1–4), among others. Table 2 shows the 12 items which were rated “Acceptable as is” (rating = 3) by fewer than 75% of participants (n < 5), of which seven were rated “Unacceptable” (rating = 1) from at least one participant (S/N 1–3, 1–4, 1–5, 1–12, 1–14, 1–15, 1–16). (See Supplementary Table 2.1 in Supplementary Appendix 2 for detailed results.) Table 2Results from phase 2 round 1 – items requiring revision (n = 6)S/NaQuestion/prompt topicNo. of responses with rating of 3 (n = 6)1–3“I will be using a guide and taking notes during our talk in case I forget something.”21–16Tradeoffs: “If your ____’s condition gets worse, how much treatment/interventions would you think he/she is willing to go through for the possibility of living longer?” (provide examples: ICU/dialysis/surgery/feeding tubes/more tests/others)21–4“What have you been told about your ____’s condition?”31–14Critical abilities (early in trajectory): “What activities brought joy and meaning to your ___’s life?”“What activities do you enjoy doing with your _____?”31–17Family/social support: “How are you and your family coping during this difficult time?”31–5“What would you like to know about your ____’s condition?”41–8Time (for doctors only): “I wish this was not the case. It is also possible that time may be as short as ___ (express as a range, e.g. days to weeks, weeks to months, months to a year).”41–10Priorities (now): “What do you think are your ___’s most important priorities now?”“Are these your priorities for him/her too?”41–11Priorities (future): “If your ___’s health gets worse, what do you think would be his/her most important priorities?”“Are these your priorities for him/her too?”41–12Worries: “What would you say are your____’s biggest worries?”Caregiver’s worries: “What are your biggest worries?”41–15Critical abilities (late in trajectory): “Given your ____’s condition, what activities do you think he/she would like to continue doing?”41–19“Keeping that in mind, and what we know about your ____’s condition, let’s make the best of this situation. I recommend that we ________.”4aSerial Number Results from phase 2 round 1 – items requiring revision (n = 6) Critical abilities (early in trajectory): “What activities brought joy and meaning to your ___’s life?” “What activities do you enjoy doing with your _____?” Priorities (now): “What do you think are your ___’s most important priorities now?” “Are these your priorities for him/her too?” Priorities (future): “If your ___’s health gets worse, what do you think would be his/her most important priorities?” “Are these your priorities for him/her too?” Worries: “What would you say are your____’s biggest worries?” Caregiver’s worries: “What are your biggest worries?” aSerial Number Qualitative findings and rationale for key revisions to version 1 of SICG-C Interviews with caregivers yielded three key themes that directly informed revisions to the SICG-C (additional representative quotes in Supplementary Appendix 3). Table 3 provides a side-by-side comparison of Versions 1 and 2 of the Guide to illustrate the revisions introduced and discussed below. Guide items in which no changes were made are shaded grey. Table 3Side-by-side view of versions 1 and 2 of SICG-CVersion 1Version 2S/NGuide itemS/NGuide item1–1“How is your ____ doing today?”2 − 1“How is your ____ doing today?”1–2“Is it okay if we talk together about your ____’s condition?”2–2“Can we talk about your ___’s condition?”1–3“I will be using a guide and taking notes during our talk in case I forget something.” (Use if appropriate)2–3“I will be using a guide and taking notes during our discussion.” (Use if appropriate)1–4“What have you been told about your ____’s condition?”2–4“What do you know about your ____’s condition?”1–5“What would you like to know about your ____’s condition?”2–5“Is there anything that you’d like to find out about your ____’s condition?”1–6“I would like to tell/inform/share with you my understanding of where things are right now with your ____’s condition – is that okay?”2–6“Can I tell/inform/share with you where things currently stand with your ___’s condition?”1–7Uncertain: “I hope that your ____ will stay as well as possible for a long time and we will work towards that. It is also possible that his/her health could change quickly. I think it is important that we prepare for that possibility.”2–7Uncertain: “I hope that your ____ will stay as well as possible for a long time and we will work towards that. It is also possible that his/her health could change quickly. I think it is important that we prepare for that possibility.”1–8Time (for doctors only): “I wish this was not the case. It is also possible that time may be as short as ___ (express as a range, e.g. days to weeks, weeks to months, months to a year).”2–8Time (for doctors only): “I would like to know if understanding the prognosis/lifespan for your ___ is important to you.”(If yes) “We cannot predict exactly how much time your ___ will have, but based on what we know about his/her health, we estimate that he/she has about ____ (express as a range, e.g. days to weeks, weeks to months, months to a year).”1–9Function: “I hope ___ will feel as well as possible for a long time, and we will work towards that. It is also possible that his/her ____’s (e.g., strength, energy level, alertness, ability to do daily activities) will be affected.”2–9Function: “I hope your ___ will stay well. It’s possible that your ____’s (e.g. strength, energy level, alertness, ability to do daily activities) will be affected, depending on the medical issues and problems that arise over time.”1–10Priorities (now): “What do you think are your ___’s most important priorities now?”“Are these your priorities for him/her too?”2–10Priorities: “Has your ___ shared with you what is important to him/her?”(If yes) “What did he/she say is important to him/her?”(If no) “Tell me more about your ___. What was he/she like and what do you think is important to him/her?”1–11Priorities (future): “If your ___’s health gets worse, what do you think would be his/her most important priorities?”“Are these your priorities for him/her too?”2–11Priorities (future): (If answer to preceding question is “yes”) “If your ___’s health gets worse, do you think there’d be any changes to what’s important to him/her?”“From your point of view, are there other things that you think are important for your ___’s care that we have not discussed or planned for?”1–12Worries: “What would you say are your____’s biggest worries?”Caregiver’s worries: “What are your biggest worries?”2–12Caregiver’s worries: “What worries you about your ___?”1–13Sources of strength: “What do you think gives your ____ strength living with his/her condition?”Caregiver’s sources of strength: “What gives you strength as you care for ____?”2–13Sources of strength: “What brings your ___ comfort or peace in his/her daily life?”Caregiver’s sources of strength: “What gives you/your family strength as you care for ____?”1–14Critical abilities (early in trajectory): “What activities brought joy and meaning to your ___’s life?”“What activities do you enjoy doing with your _____?”2–14Critical abilities (earlier in trajectory): “What activities does your ___ enjoy doing and would like to continue?”1–15Critical abilities (late in trajectory): “Given your ____’s condition, what activities do you think he/she would like to continue doing?”2–15Critical abilities (later in trajectory): “To help me make better plans for your ____’s care, I’d like to find out what abilities/daily functions are important to him/her.” (Prompt using Activities of Daily Living)1–16Tradeoffs: “If your ____’s condition gets worse, how much treatment/interventions would you think he/she is willing to go through for the possibility of living longer?” (provide examples: ICU/dialysis/surgery/feeding tubes/more tests/others)2–16Tradeoffs: “Every treatment has its pros and cons, and they can affect your ___’s quality of life in different ways. I’d like to find out what additional measures your ___ can accept if his/her health gets worse.”(Provide examples of interventions that may be relevant to patient, e.g. ICU, dialysis, surgery, feeding tubes, more tests, etc. Describe potential benefits and drawbacks.)1–17Family/social support: “How are you and your family coping during this difficult time?”2–17Family/social support: “How are you/your family doing in managing ___’s care?”1–18“I’ve heard you say that ____ is important to your ____.”2–18“I’ve heard you say that ____ is important to your ____.”1–19“Keeping that in mind, and what we know about your ____’s condition, let’s make the best of this situation. I recommend that we ________.”2–19“Keeping that in mind, and what we know about ___’s condition, I recommend/suggest _______.”1–20“This will help us make sure that your ____’s treatment/care plans reflect what’s important to him/her.”2–20“This will help us make sure that your ____’s treatment/care plans reflect what’s important to him/her.”1–21“Is this plan ok to you? If you think of anything else later, we can revisit this conversation another time.”2–21“Is this plan ok? If you think of anything else, we can talk about this again at your next visit.” Side-by-side view of versions 1 and 2 of SICG-C Time (for doctors only): “I would like to know if understanding the prognosis/lifespan for your ___ is important to you.” (If yes) “We cannot predict exactly how much time your ___ will have, but based on what we know about his/her health, we estimate that he/she has about ____ (express as a range, e.g. days to weeks, weeks to months, months to a year).” Priorities (now): “What do you think are your ___’s most important priorities now?” “Are these your priorities for him/her too?” Priorities: “Has your ___ shared with you what is important to him/her?” (If yes) “What did he/she say is important to him/her?” (If no) “Tell me more about your ___. What was he/she like and what do you think is important to him/her?” Priorities (future): “If your ___’s health gets worse, what do you think would be his/her most important priorities?” “Are these your priorities for him/her too?” Priorities (future): (If answer to preceding question is “yes”) “If your ___’s health gets worse, do you think there’d be any changes to what’s important to him/her?” “From your point of view, are there other things that you think are important for your ___’s care that we have not discussed or planned for?” Worries: “What would you say are your____’s biggest worries?” Caregiver’s worries: “What are your biggest worries?” Sources of strength: “What do you think gives your ____ strength living with his/her condition?” Caregiver’s sources of strength: “What gives you strength as you care for ____?” Sources of strength: “What brings your ___ comfort or peace in his/her daily life?” Caregiver’s sources of strength: “What gives you/your family strength as you care for ____?” Critical abilities (early in trajectory): “What activities brought joy and meaning to your ___’s life?” “What activities do you enjoy doing with your _____?” Tradeoffs: “Every treatment has its pros and cons, and they can affect your ___’s quality of life in different ways. I’d like to find out what additional measures your ___ can accept if his/her health gets worse.” (Provide examples of interventions that may be relevant to patient, e.g. ICU, dialysis, surgery, feeding tubes, more tests, etc. Describe potential benefits and drawbacks.) Theme 1: expectations of clinical authority This theme captures caregivers’ expectations of their clinicians in projecting confidence and authority in goals-of-care discussions. The phrase “in case I forget something” (S/N 1–3), for instance, received negative feedback as it was interpreted as a sign of diminished professionalism and competence. “Don’t say ‘In case I forget something,’ that’s very scary… (You) don’t have to remind me that there’s a case that the doctor forgets something.” (Cg06, 52, F, daughter of care recipient diagnosed with moderate Alzheimer’s dementia). For many, confidence in the clinician’s expertise and diligence was foundational to their trust in the clinician and their recommendations. Accordingly, S/N 1–3 was revised to “I will be using a guide and taking notes during our discussion” (S/N 2–3). Theme 2: divergent forms of expertise This theme highlights how participants differentiated their lived experiences in caregiving from clinicians’ clinical expertise. While they valued the clinician’s insights and recommendations, they emphasised the distinctiveness of their own experiential and relational understanding of the needs, behaviours, and lifestyle preferences of their loved ones, as exemplified in the following quote: “For me, what I know is not the same as what the doctors know, because I’m the one that’s taking care of her.” (Cg09, 54, F, daughter of care recipient diagnosed with moderate Alzheimer’s dementia). In response to S/N 1–4, caregivers noted that they possessed a more holistic, person-centred familiarity that is “more than what the doctor said.” In recognition of the person-centred knowledge of the caregiver, S/N 1–4 was revised to “What do you know about your ___’s condition?” (S/N 2–4). Theme 3: challenges in articulating the priorities and critical abilities of people with dementia This theme reflects the participants’ difficulties in discerning or expressing what mattered most to their loved ones. It captures how uncertainty, emotional discomfort, and projection inflected caregiver attempts to represent the values and priorities of care recipients. Caregivers reported varying degrees of difficulty in identifying what their loved ones prioritised, valued, or feared, largely due to the cognitive decline faced by care recipients and the absence of prior goals-of-care conversations. In trying to speak on behalf of their loved ones, most inadvertently projected their own views, as demonstrated in the following quote. “Actually, what does ‘priority’ mean? Does that mean that the doctor is asking me what my mum wants now or what is more important for her now? For example, what is more important for her now is to eat more [caregiver’s priority for care recipient]” (Cg06). This tendency risks undermining the very purpose of serious illness conversations, which aim to elicit what the care recipient would have wanted if they retained decisional capacity. In response, S/N 1–10 was revised to explicitly ask whether the care recipient had previously expressed what was important to them. If they had not, the Guide prompts caregivers to reflect on the patient’s character and preferences prior to cognitive decline, before considering what their goals and priorities might be now. Similarly, we revised the question on critical abilities to ask which activities of daily living (ADLs) are important to the care recipient (S/N 2–15), as opposed to asking them to reflect on activities that they enjoy or would like to continue to do, reflecting a shift from aspirational to realistic concerns that better align with the personal experience of dementia. While participants were largely accepting of the frailty of their loved ones and took well to prognostic cues indicated in S/N 1–7, 1–8, and 1–9, one participant suggested that the question about remaining time (S/N 1–8) may be made less abrupt and jarring with the addition of a preference-based prompt. In S/N 2–8, therefore, the Guide asked if the caregiver would like to know how much time their loved one might have. Rating tabulation for phase 2 round 2 Thirteen newly recruited participants were asked to compare original (Version 1) and revised (Version 2) versions of the SICG-C using the same 3-point acceptability scale from Round 1 and elaborate on their ratings. 21 items were presented, of which 17 were paired comparisons. Revisions led to improved acceptability ratings across all 17 items except S/N 2–2, which received mixed reviews with some finding S/N 2–2 “less serious” and “formal”, while others claiming that it stated the obvious. The greatest improvements in acceptability, defined as an increase in the number of participants giving a rating of 3, were observed for S/N 2–16, which gained 10 additional ratings of 3, and S/N 2–9, 2–13, 2–15, which gained 9 additional ratings of 3. The smallest improvements were seen in S/N 2–3, 2–11, and 2–19, for which an additional 2 participants gave a rating of 3. No improvement was recorded for S/N 2–2, likely because the revision was slight. (See Supplementary Table 3.2 for detailed results.) Four items (S/N 1–1, 1–7, 1–18, and 1–20) were retained in Version 2 without revision. Three of these were rated “Acceptable as is” (rating = 3) from at least 11 of 13 participants (> 75%). S/N 2–7 (Version 2), however, was deemed acceptable by only five participants, suggesting the need for further refinement and exploration of participant feedback. Qualitative findings and rationale for further revisions to version 2 of SICG-C Most participants in Round 2 interviews found Version 2 of the SICG-C acceptable and affirmed the revisions made. Nevertheless, several items merited closer attention, particularly S/N 2–7 and 2–11. Beyond semantic refinements, two key themes were identified and used to guide further iterations of the Guide. Table 4 contrasts Versions 2 with Version 3 of the SICG-C to foreground the changes made and indicate how items in one version correspond to those in the other. Table 4Side-by-side view of versions 2 and 3 of SICG-CVersion 2Version 3S/NGuide itemS/NGuide item2 − 1“How is your ____ doing today?”3 − 1“How are you doing?”2–2“Can we talk about your ___’s condition?”3 − 2“Is it okay if we discuss your ____’s condition?”2–3“I will be using a guide and taking notes during our discussion.” (Use if appropriate)3–3“I will be using a guide to structure our discussion and to take notes for future reference.” (Use if appropriate)2–4“What do you know about your ____’s condition?”3–4“What do you know about your ____’s condition?”2–5“Is there anything that you’d like to find out about your ____’s condition?”3–5“Is there anything that you’d like to find out about your ____’s condition?”2–6“Can I tell/inform/share with you where things currently stand with your ___’s condition?”3–6“May I tell/inform/share with you my assessment of your ___’s current condition?”2–7Uncertain: “I hope that your ____ will stay as well as possible for a long time and we will work towards that. It is also possible that his/her health could change quickly. I think it is important that we prepare for that possibility.”3–7 Uncertain trajectory: “While we continue to do our best for your ___’s care, it is possible that your ___’s condition could change, resulting in differences in his/her health/strength/energy level/alertness/ability to do daily activities. I think it is important we work together to prepare for what may lie ahead.”(Where appropriate, provide reassurance that you wish to work with the caregiver to plan ahead while the patient’s condition is stable/manageable.)2–8Time (for doctors only): “I would like to know if understanding the prognosis/lifespan for your ___ is important to you.”(If yes) “We cannot predict exactly how much time your ___ will have, but based on what we know about his/her health, we estimate that he/she has about ____ (express as a range, e.g. days to weeks, weeks to months, months to a year).”3–8Time (for doctors only): “I would like to know if you would like to discuss the prognosis/how much time your ___ could have”.(If yes) “We cannot predict exactly how much time your ___ has, but based on what we know, we estimate that he/she has about ____ (express as a range, e.g. days to weeks, weeks to months, months to a year).”(Where appropriate, provide assurance of support.)2–9Function: “I hope your ___ will stay well. It’s possible that your ____’s (e.g. strength, energy level, alertness, ability to do daily activities) will be affected, depending on the medical issues and problems that arise over time.” Note: S/N 2–7 and 2–9 have been merged to form S/N 3–7. 2–10 Priorities: “Has your ___ shared with you what is important to him/her?”(If yes) “What did he/she say is important to him/her?”(If no) “Tell me more about your ___. What was he/she like and what do you think is important to him/her?”3–9 Patient’s priorities: “Has your ___ shared with you what is important to him/her?”(If yes) “What did he/she say is important to him/her?”(If no) “Tell me more about your ___. What do you think is important to him/her?”2–11Priorities (future): (If answer to preceding question is “yes”) “If your ___’s health gets worse, do you think there’d be any changes to what’s important to him/her?”“From your point of view, are there other things that you think are important for your ___’s care that we have not discussed or planned for?”3–10 Caregiver’s priorities: (Describe what disease progression may look like for patient, highlight commonly seen symptoms)“From your point of view, what are some important things we should take note of when managing your ___’s condition?”2–12Caregiver’s worries: “What worries you about your ___?”3–11 Worries “Has your ____ ever shared with you his/her concerns or worries? What are they?”“What worries you about your ___?”2–13Sources of strength: “What brings your ___ comfort or peace in his/her daily life?”Caregiver’s sources of strength: “What gives you/your family strength as you care for ____?”3–12Sources of strength: “What brings your ___ comfort or peace in his/her daily life?”3–13Caregiver’s sources of strength: “What gives you/your family strength as you care for ____?”2–14Critical abilities (earlier in trajectory): “What activities does your ___ enjoy doing and would like to continue?”3–14Critical abilities (earlier in trajectory): “What activities does your ___ enjoy doing and would like to continue?”2–15Critical abilities (later in trajectory): “To help me make better plans for your ____’s care, I’d like to find out what abilities/daily functions are important to him/her.” (Prompt using Activities of Daily Living)3–15Critical abilities (later in trajectory): “To help me make better plans for your ____’s care, I’d like to find out what abilities/daily functions are important to him/her.” (Prompt using Activities of Daily Living)2–16Tradeoffs: “Every treatment has its pros and cons, and they can affect your ___’s quality of life in different ways. I’d like to find out what additional measures your ___ can accept if his/her health gets worse.”(Provide examples of interventions that may be relevant to patient, e.g. ICU, dialysis, surgery, feeding tubes, more tests, etc. Describe potential benefits and drawbacks.)3–16Tradeoffs: “Every treatment has its pros and cons, and they can affect your ___’s quality of life in different ways. I’d like to discuss with you what you think are additional measures your ___ can accept if his/her health gets worse.”(Provide examples of interventions that may be relevant to patient, e.g. ICU, dialysis, surgery, feeding tubes, more tests, etc. Describe potential benefits and drawbacks.)2–17Family/social support: “How are you/your family doing in managing ___’s care?”3–17Family/Social support: “How are you/your family doing in managing ___’s care?”2–18“I’ve heard you say that ____ is important to your ____.”3–18“I’ve heard you say that ____ is important to your ____, and that ____ is important to you/your family.”2–19“Keeping that in mind, and what we know about ___’s condition, I recommend/suggest _______.”3–19“To ensure that your ___’s care reflects what’s important to your ___ and to you/your family, may I suggest/propose that ____?”2–20“This will help us make sure that your ____’s treatment/care plans reflect what’s important to him/her.”2–21“Is this plan ok? If you think of anything else, we can talk about this again at your next visit.”3–20“Is this plan ok? If you think of anything else, we can talk about this again at your next visit.” Side-by-side view of versions 2 and 3 of SICG-C Uncertain trajectory: “While we continue to do our best for your ___’s care, it is possible that your ___’s condition could change, resulting in differences in his/her health/strength/energy level/alertness/ability to do daily activities. I think it is important we work together to prepare for what may lie ahead.” (Where appropriate, provide reassurance that you wish to work with the caregiver to plan ahead while the patient’s condition is stable/manageable.) Time (for doctors only): “I would like to know if understanding the prognosis/lifespan for your ___ is important to you.” (If yes) “We cannot predict exactly how much time your ___ will have, but based on what we know about his/her health, we estimate that he/she has about ____ (express as a range, e.g. days to weeks, weeks to months, months to a year).” Time (for doctors only): “I would like to know if you would like to discuss the prognosis/how much time your ___ could have”. (If yes) “We cannot predict exactly how much time your ___ has, but based on what we know, we estimate that he/she has about ____ (express as a range, e.g. days to weeks, weeks to months, months to a year).” (Where appropriate, provide assurance of support.) Priorities: “Has your ___ shared with you what is important to him/her?” (If yes) “What did he/she say is important to him/her?” (If no) “Tell me more about your ___. What was he/she like and what do you think is important to him/her?” Patient’s priorities: “Has your ___ shared with you what is important to him/her?” (If yes) “What did he/she say is important to him/her?” (If no) “Tell me more about your ___. What do you think is important to him/her?” Priorities (future): (If answer to preceding question is “yes”) “If your ___’s health gets worse, do you think there’d be any changes to what’s important to him/her?” “From your point of view, are there other things that you think are important for your ___’s care that we have not discussed or planned for?” Caregiver’s priorities: (Describe what disease progression may look like for patient, highlight commonly seen symptoms) “From your point of view, what are some important things we should take note of when managing your ___’s condition?” Worries “Has your ____ ever shared with you his/her concerns or worries? What are they?” “What worries you about your ___?” Sources of strength: “What brings your ___ comfort or peace in his/her daily life?” Caregiver’s sources of strength: “What gives you/your family strength as you care for ____?” Tradeoffs: “Every treatment has its pros and cons, and they can affect your ___’s quality of life in different ways. I’d like to find out what additional measures your ___ can accept if his/her health gets worse.” (Provide examples of interventions that may be relevant to patient, e.g. ICU, dialysis, surgery, feeding tubes, more tests, etc. Describe potential benefits and drawbacks.) Tradeoffs: “Every treatment has its pros and cons, and they can affect your ___’s quality of life in different ways. I’d like to discuss with you what you think are additional measures your ___ can accept if his/her health gets worse.” (Provide examples of interventions that may be relevant to patient, e.g. ICU, dialysis, surgery, feeding tubes, more tests, etc. Describe potential benefits and drawbacks.) Theme 1: worries and misconceptions over implied poor prognosis This theme captures caregivers’ sensitivity to signals of bad news and their desire for reassurance from clinicians. This underscores the need for careful phrasing and context-setting in goals-of-care discussions. S/N 2–7 (“I hope your ___ will stay as well as possible…”) elicited mixed responses from the participants. While some appreciated its empathetic tone, others found the phrasing disconcerting, interpreting it as suggestive of a poor prognosis. “Is there any other word other than ‘hope’? Because ‘hope’ is like… something consoling. I feel like ‘hope’ means that (things are) going down, then alamak [colloquialism for expresing dismay], hope – there’s only hope left.” (Cg26, M, 37, son of care recipient diagnosed with moderate vascular dementia). Nonetheless, there was broad agreement that reassurance about ongoing clinical oversight and support was helpful. In response, we revised the prompt to begin with “While we continue to do our best to care for your ___,” thereby shifting the emphasis from prognosis to the clinician’s commitment to continued care. The revised prompt nonetheless maintains the possibility “that (the care recipient’s) condition could change” (S/N 3–7). Participants also expressed a need for greater context at the outset of the conversation, namely, clarity on why the clinician was initiating the discussion and what was meant by “his/her health could change quickly.” Several participants did not view disease progression as a clinical concern, but they responded positively nonetheless when interviewers clarified that the conversation was not prompted by worsening health but aimed at preparing for the future. Drawing on these insights, we merged S/N 2–7 and 2–9 into a single prompt (S/N 3–7), included a note to clinicians to provide reassurance to the caregivers, and reinforced a sense of partnership by inviting caregivers to “work together to prepare for what may lie ahead.” Theme 2: uncertainty and interpolation in caregivers’ accounts of patients’ priorities This theme points to the caregivers’ discomfort and uncertainty when speaking on behalf of their loved ones. Despite revisions to reduce this interpretive burden, some caregivers voiced their uneasiness with discussing their loved ones’ future goals, revealing the emotional and ethical complexity of proxy decision-making and representation in dementia care. S/N 2–11 presented considerable challenges for participants. Many felt ill-equipped to speak to the potential changes in their loved one’s priorities, particularly as people with dementia were often perceived to lack the cognitive capacity to hold or express even present-day priorities. None of the participants have had a discussion with their loved ones about their goals of care nor had they been informed on their priorities. “No, no, you know, people of that era, right? They don’t want to talk about that. They don’t want to talk about, yeah… being sick or whatever.” (Cg18, F, 71, daughter of care recipient diagnosed with mild Alzheimer’s dementia). “If her condition gets worse, then she is not in the position to respond, to express (her wishes). Then it’s a question of what the caregiver actually thinks is best for the mother… But the condition is already bad. I don’t know whether my mum wants to… have her life extended or not.” (Cg13, F, 68, daughter of care recipient diagnosed with Alzheimer’s dementia). As in Round 1 interviews, participants frequently substituted their own priorities for those of care recipients. Some caregivers of people with moderate to severe dementia expressed discomfort and moral uncertainty at whether their assumptions aligned with what their loved ones would have wanted, had they retained decision-making capacity. We thus considered a question about the care recipients’ future priorities to be unnecessary, and that any potentially relevant information about their goals of care would be elicited through the preceding prompt. As such, S/N 3–10 (corresponding to 2–11) asks caregivers, “What are some important things we should take note of when managing your ___’s condition?” For added clarity, this question includes a reminder for clinicians to “Describe what disease progression may look like for the patient,” recognising the clinician’s role in contextualising the discussion based on the patient’s dementia subtype and clinical history. Although not directly prompted by participant feedback, we revised S/N 2–12 to ensure consistency in how the Guide distinguishes between patient and caregiver perspectives. The original version asked caregivers only about their own worries regarding the patient. However, cognisant that certain people with dementia—especially those in the earlier stages of disease—may still be able to express their worries and concerns, we applied the same recall-based framing used for S/N 3–9 to S/N 3–11, asking if they have “ever shared anything they are worried or concerned about” with the caregiver. Acceptability assessment in phase 3 (n = 50) Upon finalising Version 3 of the SICG-C by the end of Phase 2, we conducted a final acceptability assessment with 50 caregivers, four of whom had been interviewed in Phase 2. As we separated S/N 1–13 (patient’s sources of strength and caregiver’s sources of strength) into its constituent elements to facilitate comparison with its Version 3 counterparts (Questions 13 and 14 of questionnaire in Supplementary Appendix 1), the questionnaires contained 22 paired items for the participants’ evaluation. Both study-specific acceptability criteria (i) and (ii) were satisfied for every item in Version 3 of the Guide. By contrast, only one item in Version 1 (S/N 1–1) met both criteria, with five other items (S/N 1–5, 1–6, 1–13, 1–18, 1–21) meeting only criterion (ii). (For full results, see Supplementary Table 2.3.) Wilcoxon signed-rank tests were performed for each pairwise comparison between Versions 1 and 3. Of 22 paired comparisons, 21 showed statistically significant improvements favouring Version 3 (all p < 0.001, Table 5). Table 5Results from Wilcoxon signed-rank test in phase 3S/NWilcoxon Vp-valueInterpretationQuestionaVersion 1Version 311–13 − 1115.50.6702No significant difference21–23 − 200.0000000553Revision favoured31–33–3170.0000000403Revision favoured41–43–465.50.0000217000Revision favoured51–53–5129.50.0001530000Revision favoured61–63–6640.0000379000Revision favoured71–73–7820.0000001030Revision favoured81–93–717.50.0000000063Revision favoured91–83–812.50.0000000009Revision favoured101–103–9510.0000001480Revision favoured111–113–10330.0000000896Revision favoured121–123–111080.0000037200Revision favoured131–133–1292.50.0000001920Revision favoured141–133–13660.0000154000Revision favoured151–143–14140.0007280000Revision favoured161–153–1580.0000189000Revision favoured171–173–171600.0018650000Revision favoured181–163–16480.0000001080Revision favoured191–183–1800.0000000154Revision favoured201–193–1982.50.0000262000Revision favoured211–203–19560.0000176000Revision favoured221–213–20850.0000651000Revision favouredaQuestion order as shown in Supplementary Appendix 1 (questionnaire administered during Phase 3 of testing) Results from Wilcoxon signed-rank test in phase 3 aQuestion order as shown in Supplementary Appendix 1 (questionnaire administered during Phase 3 of testing) While participants did not consistently favour S/N 3 − 1 over S/N 1–1, the differences are slight, with the former asking “How are you doing?” and the latter “How is your ___ doing today?”. The show of concern and support for caregivers in S/N 3 − 1 was appreciated by some participants, though most were accepting of both conversation openers. These results indicate that the final version of the SICG-C is highly acceptable and that the iterative revisions led to an overall improvement in its acceptability. Discussion The participants’ preference for the final version of the adapted caregiver-oriented Serious Illness Conversation Guide (SICG-C) reflects a clear improvement in meeting caregivers’ informational and emotional needs, as well as their expectations for clinician communication. With a median rating of 4 (“Very Acceptable”) for every item, the final Guide achieved high acceptability among caregivers of people with dementia, indicating its suitability for adoption in clinical practice and caregiver engagement. (See Supplementary Appendix 4 for the finalised SICG-C.) Our findings affirm prior research showing that goals of care are rarely discussed within Asian families, often due to generational attitudes and cultural norms around talking about health, illness, and dying [24–27]. In particular, the belief that filial piety entails intuitively knowing and acting on a loved one’s wishes, i.e., without needing explicit instruction, accords value to leaving things unspoken [24, 27, 28]. Without having been informed of their loved ones’ priorities and goals, most of the participants claimed to understand what is important to them, with many considering the idea of having such a discussion with their elders uncomfortable and at odds with how they relate to each other. While involving caregivers as proxies in goals-of-care discussions is suboptimal, persistent challenges in eliciting preferences directly from people with dementia underscore the need for a caregiver-specific guide as an interim measure. Further research is warranted to develop and evaluate interventions that support earlier family communication about goals of care, particularly in the early stages of frailty and cognitive decline for people with dementia. This work complements Berry et al.’s dementia-specific SICG [21] by offering an adaptation for the Singaporean context and potentially for other Asian populations, expanding the repertoire of communication tools that clinicians may draw on, adapt, and integrate into practice to support goals-of-care discussions with caregivers of people with dementia. At the same time, while it is intended for use with caregivers, the Guide may also be considered in scenarios involving undiagnosed cognitive impairment, patients with communication difficulties, or where treatment decision-making responsibilities have been delegated to family caregivers. This study has four limitations. First, there is a potential risk of contrast bias from presenting two Guide versions for assessment. While participants were asked to rate each item independently, there remains a residual risk of contrast bias. Second, caregiver feedback was based on hypothetical reactions, rather than observations of actual clinical use. Third, as all participants were English-speaking caregivers, findings may not represent the expectations and experiences of non-English speakers. Lastly, most participants took part in only one phase of the study, with four incidentally re-approached for the final survey. While feedback was largely convergent, some participants offered contradictory suggestions—for instance, where some caregivers preferred plain, direct language while others recommended gentler phrasing and reassurances. Final revisions were made conservatively, prioritising language that enhanced emotional sensitivity and accessibility for caregivers. Future work may involve adapting the SICG-C for non-English speakers, evaluating the use of the SICG-C in clinical settings, and testing the cultural acceptability of the SICG-C in other Asian societies. Conclusion This study describes the iterative adaptation of the Serious Illness Conversation Guide for Caregivers (SICG-C) to support caregiver-clinician goals-of-care discussions for persons with dementia. Revisions improved its clarity, emotional sensitivity, and relevance. The final version was deemed highly acceptable by caregivers of people with dementia and offers a practical, culturally appropriate communication tool for clinical applications and training in Singapore and potentially in the wider region. Supplementary Information Supplementary Material 1. Appendix 1: Questionnaires used in Phase 3. Supplementary Material 2. Appendix 2: Detailed results. Supplementary Material 3. Appendix 3: Representative quotes from Phase 2. Supplementary Material 4. Appendix 4: Finalised SICG-C. Acknowledgements The study team extends its gratitude to the Lien Centre for Palliative Care (LCPC), the Lien Foundation, and Dr Anupama Roy Chowdhury (Department Head of Geriatric Medicine, SGH) for their generous support for this project. Authors’ contributions AN: Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing – Original draft, Writing – review & editing. VS: Methodology, Resources, Supervision, Writing – review & editing. SN: Conceptualization, Methodology, Investigation, Project administration, Resources, Supervision, Writing – review & editing. Funding statement This work was supported by the Lien Centre for Palliative Care Research Award [Grant number: LCPC-EX22-0001]. Data availability The data that support the findings of this study are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate All participants gave their written informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the SingHealth Centralised Institutional Review Board (CIRB), reference number: 2022/2736. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Appendix 1: Questionnaires used in Phase 3. Supplementary Material 2. Appendix 2: Detailed results. Supplementary Material 3. Appendix 3: Representative quotes from Phase 2. Supplementary Material 4. Appendix 4: Finalised SICG-C. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.