Exploring Patients’ Perceptions of an Advance Care Planning Intervention in the Emergency Department: A Qualitative Study

Abstract

Objectives:

Emergency department (ED) visits offer opportunities for seriously ill patients to formulate future medical care goals, yet ED clinicians lack practical strategies for these conversations. ED GOAL, a behavioral intervention, engages seriously ill yet clinically stable older adults in the ED to address advance care planning (ACP) with their outpatient clinicians. In a randomized trial, goals-of-care documentation was significantly higher in the intervention group compared to controls after three (24.3% vs. 9.9%, p = 0.03) and 6 months (31.4% vs. 12.7%, p < 0.01). This study is a sub-analysis to learn about intervention arm participants’ perceived benefits and obstacles of the intervention.

Methods:

We conducted semi-structured interviews between October 2022 and August 2024 (N = 52) with intervention-arm patients aged 50+ years at three hospitals in Boston, Massachusetts. Using rapid qualitative analyses, we identified themes in intervention-arm participants’ comments to open-ended questions about the intervention’s benefits and obstacles to continue ACP outside the ED.

Results:

Of 70 intervention-arm participants, 52 completed interviews, of which two were surrogates. ED GOAL motivated most patients to initiate ACP with outpatient clinicians and loved ones and improved the quality of conversations by clarifying patients’ wishes and improving patient-clinician relations. Barriers to continuing ACP were the lack of clinician availability and patient/surrogate readiness. Those with clear care goals found the intervention less useful yet harmless.

Conclusions:

The intervention provided participants with insights into actionable ACP steps. To address the lack of clinician availability, these conversations may be completed by non-physician clinicians or through non-personnel resources. Better tailored ACP interventions may improve patients’ readiness.

Trial Registration:

ClinicalTrials.gov identifier: NCT05209880

1 |. Introduction

Emergency departments (ED) may be the prime setting to initiate serious illness conversations (SICs), which are essential goals-of-care discussions between clinicians and patients with advanced illnesses who may face poor prognoses. Approximately 75% of older adults visit the ED during the last 6 months of life, but 56%–99% of these patients do not have advance directives documented at the time of their visit [1, 2]. Thus, the ED provides a unique opportunity to intervene in patient care near the end of life.

SICs allow patients to communicate their medical care values, ensuring their care aligns with their goals, particularly near the end of life [3]. Patients who had SICs were more than twice as likely to receive goal-concordant care [4]. SICs are associated with reduced deaths in the intensive care unit, earlier hospice referrals, increased peacefulness, and less aggressive end-of-life care [5–7]. However, less than 40% of seriously ill older adults have end-of-life conversations [7]. On average, these conversations are initiated 71 days before death, potentially contributing to less goal-concordant care due to the delayed timing [8]. There is a dire need to facilitate SICs among seriously ill older adults to better align their care with their wishes.

We developed ED GOAL, a behavioral intervention to help patients in the ED formulate and communicate their care goals. It is a multi-modal intervention performed by trained research nurses with patients/surrogates during or after an ED visit. The intervention includes: (1) a motivational interview, (2) a structured SIC, and (3) communication to reinitiate SICs with outpatient clinicians. The structured SIC used patient-tested language, demonstrated in previous studies [8, 9]. The research nurses documented patients’ preferences in their medical records, communicated them to the patients’ outpatient clinicians (with the participants’ permission), and provided a handout to encourage additional SICs with clinicians/loved ones.

Our intervention aimed to facilitate SICs after leaving the ED. In a randomized controlled trial (RCT) of English and Spanish-speaking patients/surrogates, we found that ED GOAL led to increased medical record documentation of end-of-life preferences [10]. However, little is known about patient/surrogates’ perceived benefits of ED GOAL and obstacles in reinitiating SICs with outpatient clinicians. We conducted this qualitative study to understand the experience of the intervention-arm patients/surrogates in continuing SICs with their outpatient clinicians after an ED visit.

2 |. Methods

2.1 |. Study Design and Setting

This qualitative analysis is a sub-analysis of a pilot RCT of 141 patients/surrogates seen in the ED of two academic medical centers and one community hospital in Boston, Massachusetts. The RCT is described elsewhere [10]. In brief, English- and/or Spanish-speaking ED patients aged 50 years or older with at least one serious illness were randomized to ED GOAL or usual care [10]. The RCT assessed changes in patient-reported engagement in advance care planning (ACP) at three timepoints after baseline. ACP is a range of activities that support adults in understanding and sharing their values, goals, and preferences regarding future medical care. In our study, ACP included discussions about code status, identification of a health care proxy, documentation of advance directives, and other values and goals for end-of-life care between patients and their clinicians/loved ones.

Eligible serious illnesses included New York Heart Association (NYHA) Class III/IV congestive heart failure, chronic obstructive lung disease on home oxygen, chronic kidney disease on dialysis, or metastatic solid tumor cancer. Patients with non-metastatic cancer, NYHA Class I/II heart failure, chronic obstructive lung disease not on home oxygen, and chronic kidney disease not on dialysis were included if patients had a hospitalization in the 12 months prior to recruitment. Our inclusion of patients with recent hospitalizations was guided by prior literature identifying recent hospitalization as a marker for high risk of serious illness progression and mortality. While some of these conditions may be curable or stable, we aimed to capture patients for whom a SIC could still be appropriate due to their recent acute healthcare utilization, which may reflect evolving or poorly controlled disease. Multiple studies have shown that recent or recurrent hospitalizations are associated with increased mortality across a range of chronic diseases—including “low-risk” heart failure—even after adjusting for multiple comorbidities [11, 12]. Surrogates were co-enrolled for patients with mild cognitive impairment or mild dementia. For patients with moderate or severe dementia, only surrogates were enrolled. Mini-Cog and the Quick Dementia Rating System were administered to determine cognitive impairment [13, 14]. Surrogates were patients’ healthcare proxies or next of kin [10].

This study was approved by the Mass General Brigham IRB (Protocol 2021P003093). Verbal consent was obtained from all participants. We followed the Consolidated Criteria for Reporting Qualitative Research in reporting this study (Data S1) [15].

2.2 |. Selection of Participants

All intervention patients/surrogates were approached for a semi-structured interview, unless they had expired, refused, or could not be reached.

2.3 |. Interview Guide Development

We created a semi-structured interview guide with clinical (KO: MD/MPH, emergency physician, male) and qualitative research experts (ACR: PhD, sociologist, female; ANC: MD/PhD, medical anthropologist, female) in English and Spanish. The qualitative interview focused on the intervention’s benefits and obstacles that participants faced when re-initiating SICs with their clinicians/loved ones. Clinicians include outpatient healthcare professionals directly involved in patient care (e.g., physicians, physician assistants, nurses, and social workers). Loved ones refer to family members, caregivers, or friends who can make decisions for the patient if they become too sick to speak for themselves.

We included six open-ended questions about patients’/surrogates’ (1) attitudes about ACP, (2) actions taken regarding ACP, (3) intervention effects on conversations with clinicians, (4) effects on conversations with loved ones, (5) effects on attitudes towards medical crises (i.e., if patients are unable to speak for themselves), and (6) effects on end-of-life decision making. These six domains were pre-specified by our expert clinician panel, who identified key aspects of the intervention and anticipated areas of impact or barriers to ACP engagement. These domains guided the development of our interview questions and were used as the framework for qualitative analysis. The Spanish guide was translated by a certified translator and reviewed by a Spanish-speaking research assistant outside of the research team.

2.4 |. Data Collection

Interviews were conducted by phone from October 14, 2022, to August 9, 2024, by three research assistants (YS: MPH, research assistant, female; HO: BA, research assistant, female; LK: BA, research assistant, female), trained by the principal investigator (KO). The participants were contacted at least 6 months after baseline enrollment in the RCT. The interviews were audio/visual recorded unless participants declined, in which case the interviewers took detailed notes. The interviewers had provided participants with a study information sheet prior to the intervention. At the beginning of the interview, participants were reminded of ED GOAL through an explanation of the study and a definition of ACP. Some participants were not asked questions if interviewers noticed signs of distress caused by the interview, such as emotional reactions or reluctance to answer due to discomfort. Participants were encouraged to skip questions or stop the interview entirely if they felt uncomfortable. Participants were not excluded for declining to respond or if the interviews were cut short.

2.5 |. Qualitative Analysis

We utilized a rapid analysis approach, a validated method that allows for timely dissemination of results with fewer resources, producing findings consistent with more in-depth qualitative analyses [16]. Rapid analyses of the interviews identified recurring themes to illustrate qualitative benefits (e.g., how ED GOAL changed their care) and obstacles to complete SICs (e.g., why attempted SICs were unsuccessful). The recordings were analyzed without full transcription. Field notes were created after the interviews.

Interviews were analyzed by two trained researchers who did not participate in the interviews (SM: BS, BA, research assistant, female, NS: MS, research assistant, female), ensuring interpretive integrity and consistency. Based on the prefigured codes from the domains, YS, SM, and NS developed a codebook, which was iteratively refined through discussions. All interviews were independently coded (SM, NS) using a framework based on the prefigured codes. Ten percent of the interviews were double-coded to ensure interrater reliability of over 90%, which assessed the consistency of coding, crucial in minimizing bias [15]. SM and NS created a matrix to identify common themes across domains. The matrix was reviewed by YS, who added emerging codes by expanding on the predefined codes, and these were discussed collaboratively with SM and NS to ensure consistent integration into the framework. Data was saturated when no new themes were identified and confirmed with the team. Disagreements were discussed to establish a consensus.

3 |. Results

3.1 |. Participant Characteristics

Of the 70 intervention arm participants in the RCT, 52 were interviewed for this qualitative study. Of those not interviewed, 6 had expired, 2 refused, and 10 could not be reached. Of the 52 interviewees, two were surrogates. The patients’ mean age was 65.7 years (SD 8.66); 26 (50.0%) were female, 36 (69.2%) were White, and 11 (21.2%) were Black/African American. Thirty patients (57.7%) had metastatic solid tumor cancer, and 2 (3.85%) had non-metastatic solid tumor cancer with recent hospitalization (Table 1). Two interviews were conducted in Spanish by Spanish-speaking staff (LK).

TABLE 1 |.

Participant characteristics.

	No. (%)a
Characteristics	Total (N = 52)
Age, years
Mean (SD)	65.7 (8.66)
Median [Min, Max]	66.0 [50.0, 85.0]
Sex
Female	26 (50.0)
Male	26 (50.0)
Race
White	36 (69.2)
Black/African American	11 (21.2)
Missing	5 (9.62)
Ethnicity
Hispanic/Latino	3 (5.77)
Non-Hispanic/Latino	49 (94.2)
Serious illness
Solid tumor cancer with metastases	30 (57.7)
Non-metastatic solid tumor cancer with recent hospitalization	2 (3.85)
COPD on home oxygen	1 (1.92)
COPD not on home oxygen with recent hospitalization	1 (1.92)
CHF (NYHA Class III/IV)	5 (9.62)
CHF (NYHA Class I/II) with recent hospitalization	5 (9.62)
CKD on dialysis	7 (13.5)
ED Clinician would not be surprised if patient died in the next 12 months	1 (1.92)

Abbreviations: CHF, congestive heart failure; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; ED, emergency department; NYHA, New York Heart Association Functional Classification.

^aValues are reported as No. (%) unless otherwise indicated.

3.2 |. Qualitative Themes

We found five cross-cutting themes across domains: (1) Most patients found ED GOAL motivating to initiate SICs with outpatient clinicians/loved ones. (2) ED GOAL improved the quality of SICs by clarifying patients’ wishes, improving clinician-patient relationships, and reducing the burden on loved ones. (3) Some participants were frustrated that they could not access their clinicians for these discussions. (4) Patient/surrogate readiness for ACP was a barrier in continuing ACP with outpatient clinicians. (5) Patients who had already communicated their goals of care to clinicians/loved ones found ED GOAL less useful; regardless, they did not find it harmful, and some liked the review (Table 2).

TABLE 2 |.

Themes and representative quotes.

Themes	Representative quotes
Theme 1: ED GOAL was useful in initiating SICs with outpatient clinicians/loved ones	“[ED GOAL] spurred a family conversation with [loved ones] about how important it is. It was a good reminder for me and my wife.” (ID 63)
	“I think that it added specificity to the conversation. Instead of just talking in general … we’ve been able to talk more specifically about end-of-life questions.” (ID 84)
	“I had not yet had that conversation before. It was good to start having that conversation, talking with my wife … It was the boost I needed to get going. I had always thought about it but had never done anything.” (ID 93)
Theme 2: ED GOAL improved the quality of SICs by clarifying patients’ wishes, improving clinician-patient relationships, and reducing the burden on loved ones	“[ED GOAL] made me think a little [deeper] about being more specific … how important it is to have that laid out early so that someone’s not in a really difficult position without having this conversation with me.” (ID 44)
	“It made it a little easier to talk to [my family] about [ACP] and to say what I want and don’t want.” (ID 79)
	“Talking to my doctor, I found out … he’s really rooting for me.” (ID 118)
Theme 3: Frustration with lack of access to clinicians to have SICs	“Well, I’ve never taken the opportunity. We just go in there for a physical and go through what I have to do to maintain a healthy lifestyle.” (ID 20)
	“We haven’t had an appointment with our PCP … an extensive appointment.” (ID 91)
	“I have no one to talk to. so it has been very difficult for me. My HCP … isn’t someone I can talk to.” (ID 127)
Theme 4: Patient/surrogate readiness is a major barrier in continuing ACP	“It’s discouraging and worrying because I don’t know what happens if … my health worsens.” (ID 21)
	“Every time I bring up anything, even planning my funeral … They get upset at the thought of it. My kids aren’t ready for it.” (ID 37)
	“It’s so hard for me to determine what I want or don’t want.” (ID 58)
Theme 5: ED GOAL was not as useful for patients who already did ACP	“I don’t think it affected it at all because I had already thought about all this.” (ID 12)
	“We spent time with [clinician] and went over very carefully my choices for future care, which we previously didn’t fully understand … I have since changed some of my answers.” (ID 25)
	“I’m comfortable that my family knows what my wishes are.” (ID 91)

3.3 |. Attitudes and Actions Towards ACP

Most participants thought ED GOAL raised self-awareness and encouraged ACP (Appendix S1). They felt ED GOAL was educational and helped jumpstart conversations that are challenging to initiate when acutely ill: “It’s nice to have this all taken care of. [ED GOAL] was a nudge to get me going” (ID 93). ED GOAL increased self-awareness about ACP documentation: seven patients completed a Medical Order for Life Sustaining Treatment form, five completed a healthcare proxy form, and nine completed a living will. Even patients who had already taken ACP actions used ED GOAL as an opportunity to reassess decisions.

Participants who noted ED GOAL did not affect their attitudes felt that ACP was not relevant yet or had anxieties surrounding ACP. Others were still unsure what their preferences were: “I don’t know what actions to take. I don’t know what the situation will be” (ID 21).

3.4 |. Intervention Effects on Conversations With Clinicians and Loved Ones

Most participants spoke with clinicians/loved ones about ACP after ED GOAL. Some participants commented on plans to speak with their clinician, while others reported no plans.

Participants who spoke with their clinicians and/or loved ones were asked if ED GOAL had any effect on these conversations. Many noted ED GOAL helped initiate this conversation by clarifying wishes and discussing scenarios: “It made it a little easier to talk … about [ACP] and to say what I want and don’t want” (ID 79). Another noted that “carefully [going over their] choices for future care, which [they] previously didn’t fully understand,” allowed him to reassess his previous answers with his clinician (ID 25).

One participant noted ED GOAL encouraged more explicit discussions with his clinician regarding his prognosis: “My sense was that I had a lot of years left … when [the cancer meant] I don’t, that got my attention” (ID 93). Another participant reflected, “[It] made me think a little [deeper] about being more specific … about what I would like to have happen” (ID 44).

Often, participants described improved relationships with their clinician, realizing that “[my clinician is] really rooting for me” (ID 118) and “[my clinician] does understand my feelings about my health” (ID 48). Conversely, some participants did not express improved SICs stating, “Doctors can only talk to me about treatment, but they can’t do more than that” (ID 21). Some preferred speaking with loved ones over their clinicians out of comfort: “What felt more comfortable was to have a healthcare proxy … knowing that she’ll make the right choices for me” (ID 44). Others reported no intervention effect on conversations because they had previously documented or expressed wishes to their clinicians/loved ones.

Barriers for SICs with clinicians included lack of opportunities and not prioritizing ACP. These challenges stemmed from their current health, which had not deteriorated enough to warrant SICs. Participants unable to speak with their loved ones reported having nobody to talk with or a lack of family readiness. Similarly, many participants noted that they were unable to speak with their outpatient clinicians due to short visit lengths or clinicians leaving the practice. Some participants commented that they still did not understand ACP or felt no need to discuss it with their clinicians as their healthcare proxy was aware of their wishes.

Most participants who spoke to their clinicians/loved ones about ACP recommended SICs to other seriously ill patients. One participant reflected, “You might assume that your family understands what you want or not want; in reality, they might not” (ID 43). Another noted, “It’s generally a conversation that comes up during crisis … I think people should talk about this earlier” (ID 44).

3.5 |. Intervention Effects on Attitude Towards a Crisis

Most participants described a positive intervention effect, reporting that ED GOAL was “a decent check-in … if things got into crisis mode and how we would handle it” (ID 63). Participants who described no intervention effects chose to defer decision-making to their loved ones during a crisis: “I have no idea what my feelings would be … I might rely on the judgement of my family” (ID 03). Others noted increased anxieties regarding their health: “This study makes me think that this could happen to me at any time … I would get sick worrying about it” (ID 35).

3.6 |. Intervention Effects on End-of-Life Decision Making

Most participants reported a positive intervention effect resulting in improved autonomy over end-of-life decision making: “I don’t have to rely on someone guessing … I’ve already said what I want” (ID 93). Others noted increased preparedness for medical crises: “You’re having the conversation while you’re [healthy] … you’re likely to know your own mind if you’re not faced with a crisis” (ID 15). Participants who reported no intervention effect on decision making cited previously completed ACP.

4 |. Discussion

In this qualitative study of seriously ill older adults, we found that ED GOAL motivated patients to initiate ACP with their clinicians/loved ones after an ED visit. ED GOAL enhanced the quality of SICs by clarifying medical care wishes, improving clinician-patient relationships, reducing burdens, and preparing loved ones for end-of-life situations. However, many participants faced obstacles, including limited opportunities to revisit ACP with clinicians and a lack of patient/surrogate readiness to initiate SICs. Finally, those who had previously communicated their goals of care to clinicians/loved ones found ED GOAL to be less useful. These findings suggest that SIC interventions in the ED are perceived as helpful in facilitating ACP, particularly for patients who have not yet addressed their care preferences.

While SICs have been shown to reduce deaths in the hospital and lead to less aggressive end-of-life care, their use in the ED is limited. We focused on the experiences of ED patients, building on prior studies that have examined SIC effects in other settings. A study investigating the acceptability of ACP interventions in outpatient clinics found SICs clarified end-of-life care preferences, a parallel to our study [17]. While some participants found that ED GOAL improved the clinician-patient relationships, other studies found no change [17]. In our study, clinicians’ lack of availability was an obstacle in reinitiating SICs, a common theme in a qualitative study on nursing home staff [18]. Furthermore, patient/surrogate readiness is a critical theme in SICs as highlighted in another qualitative study on SIC barriers [19]. We found that patients who already expressed their preferences found ED GOAL less useful, similar to a finding that patients with documented advance directives were less interested in discussing their preferences with ED clinicians [20].

Our study’s implications to clinicians are ED visits can be teachable moments for patients to engage in SICs; educating patients on initiating SICs may facilitate more effective decision-making toward the end of life. ED GOAL encouraged many participants to re-initiate SICs with their clinicians by providing guidance on different approaches to ACP, while improving the SIC’s quality. ED GOAL taught patients about actionable steps, such as completing advance directives, by prospectively capturing seriously ill older adults in these teachable moments. Similarly, an RCT evaluating decision aids for ACP in primary care settings found that the intervention improved patient knowledge of medical decisions [21]. Our findings showed that SIC interventions can provide educational benefits in the ED, motivating patients to be proactive in their ACP. Increasing patients’ ACP engagement could reduce burdens on patients/surrogates if they return to the ED or have another medical crisis.

Another implication for clinicians is the need to address clinician availability by developing systems for patients to explore their goals-of-care without direct physician involvement. This can be achieved by (1) training nurses/social workers/laypersons for SICs, (2) having remote conversations, or (3) utilizing technology and non-personnel resources. While many patients felt motivated to complete ACP, we identified several non-patient challenges to reinitiating ACP, including (1) clinicians leaving practice, (2) clinicians lacking sufficient time, and (3) uncertainty in prognosis consistent with clinician-level barriers found in other studies [22–25]. To address clinician accessibility, training nurses or social workers for SICs in outpatient settings may be effective, as supported by other studies [26]. A qualitative analysis found that clinicians often perceive SICs as the “physician’s job”; however, using a structured conversation guide empowered nurses/social workers to initiate SICs [27]. Another study showed that patients preferred nurses or physician assistants for their bedside manner and convenience compared to physicians [28]. Additionally, training laypersons and nonclinical personnel to conduct SICs by building upon the Patient Care Connect Program, which utilizes the Respecting Choices curriculum, may overcome clinician-accessibility barriers [29–31]. Furthermore, remote conversations, technology (e.g., artificial intelligence), or non-personnel resources, including goals-of-care decision aid videos, may enable patients to more readily address their ACP, demonstrated in other studies [32–35].

Our study’s implications for policymakers and researchers are to consider tailoring SIC interventions to patients’ readiness, using screening tools to identify those who may benefit most, and addressing barriers such as patient reluctance. For individuals who are not ready, repeated motivational interviewing may be beneficial, suggesting that readiness should guide the approach to intervention. In our study, patient readiness was a barrier in continuing ACP with clinicians, and those with pre-established goals found ED GOAL less useful. This aligns with other studies that showed patients who are less ready may have less SICs [19, 25, 36]. The lack of readiness may stem from several factors, such as: (1) patients’ loved ones not being ready, or (2) patients not feeling “sick enough” to engage in such discussions [35]. Notably, some patients may have conditions with a potentially curable trajectory—such as non-metastatic cancer—where the tone and content of SICs may differ from those in patients with terminal illnesses. The timing and framing of these conversations, guided by both prognosis and patient preference, may be critical factors influencing readiness. In a qualitative study assessing perspectives on SICs among patients with advanced cancer, some patients preferred having these conversations earlier in their illness trajectory, while others preferred to wait until experiencing a “downhill” progression in their condition [37]. Another qualitative study investigating the perceptions of the timing and SIC content among patients with advanced cancer emphasized the importance of tailoring discussions to individual needs, addressing different topics at various points in the care trajectory [38]. These obstacles imply that SIC interventions in the ED may be tailored to individuals’ readiness and prognosis. By using the Geriatric End-of-life Screening Tool, an automatable prognostic model, it may be possible to better target those with a high mortality risk [39]. For those who are not ready, repeated sessions of ED GOAL and its motivational interview may be effective [40].

4.1 |. Strengths

The strength of our study lies in its novel focus on exploring the experiences of those who initiated SICs in the ED, which has critical implications for informing ED practices for seriously ill patients. To our knowledge, no qualitative follow-up study of an RCT on SIC interventions in the ED has been conducted before. Furthermore, our study incorporated diverse perspectives, including older adults with various serious illnesses, further enhancing the generalizability of how patients in different settings approach ACP. By offering the interview guide in both English and Spanish, we ensured that our findings are inclusive of Spanish-speaking populations.

4.2 |. Limitations

This study has several limitations. First, 12 patient interviews were not audio recorded due to patient preference. While efforts were made to capture responses through notetaking, the absence of recordings may have resulted in the loss of some quotes or nuances. However, their content was consistent with findings from recorded interviews. Recall bias is another potential limitation, as participants were contacted at least 6 months after enrolling in the study. The interviewers made efforts to refresh participants’ memories before starting the interviews. Additionally, since some of the study staff were closely involved in the implementation and evaluation of the ED GOAL intervention, there is a risk of confirmation bias in interpreting the qualitative data in favor of the intervention. Participants may also have been predisposed to report favorably due to social desirability bias. These factors may have influenced how the findings were framed and should be considered when interpreting the results.

5 |. Conclusion

In this qualitative study of older adults who received an SIC intervention in the ED, we found that ED GOAL motivated participants to engage in ACP with both their outpatient clinicians and loved ones. Major barriers to reinitiating SICs included limited opportunities with clinicians and a lack of readiness. These findings imply that future studies in outpatient clinics may explore how nurses and social workers could initiate conversations even through telehealth or by training laypersons and utilizing non-personnel resources. ED GOAL may be better targeted to patients with low readiness and high mortality, potentially enhancing educational benefits by focusing on actionable ACP steps that align with individual needs.

Supplementary Material

Additional supporting information can be found online in the Supporting Information section.

APPENDIX 1: Interview Guide

Refresh the participants’ memory about the study:

“I am a research assistant, calling about the advance care planning study that you joined in ____. Thank you for completing our phone surveys so far. I am calling to see if now would be a good time to answer some open-ended questions about your experiences in our study. There are about 10 questions and this usually takes 15 minutes. I usually take notes or make an audio recording while I talk. Would it be okay if I record our call?

Introduction:

My name is ____. Thank you for helping us today. I would like to ask you a few questions about your thoughts and experience after enrolling in our study.

We hope to understand how, if at all, our study influenced your advance care planning process with your doctor and loved ones. Our goal is to describe your thought process based on your experience. There is no right answer to these questions, and your responses will be completely confidential. This interview will in no way affect your care.

Please allow me to explain what I will be asking about:

ED GOAL interview is the conversation you had with our study clinician, (insert name).

ACP conversation is the conversation you may or may not have had with your doctor or family/loved ones about what healthcare you want if you get too sick to speak for yourself.

As you recall, you gave us permission to record this conversation. You can skip any questions you don’t wish to answer, and We can stop the interview at any point if you feel uncomfortable or do not wish to continue.

1. Since enrolled in the study, what thoughts, if any, have you had about ACP? For example, some people have told us that they thought XYZ (advance care planning may help my family know that they are honoring my wishes if I cannot speak for myself).

   • - Can you give me an example?
   • - Tell me more.
2. Since enrolled in the study, what actions, if any, have you taken around ACP? For example, some people have told us that they did XYZ (shared their study sheet with their health care proxy).

   • - Can you give me an example?
   • - Tell me more.

3. Have you spoken to your doctor about ACP?

   If yes: Please tell me how this study affected the ACP conversation with your doctor.

   • - Did you use the handoff sheet that was provided? If so, how was the experience?
   • - In the study, our study clinician (insert name), asked you about things that are important to you if your health were to worsen, life states that are worse than dying itself, hopes for the future, etc. In what ways, if any, has talking about this affected your ACP conversation with your doctor?
   • - What would you tell someone else who was faced with a serious illness like you about the study? Would you recommend this conversation to them?
     If no: what are some reasons that you haven’t had these conversations with your doctor?

4. Have you spoken to your family/loved ones about ACP?

   If yes: Please tell me how this study affected the ACP conversation with your family/loved ones.

   • - In the study, our study clinician (insert name), asked you about things that are important to you if your health were to worsen, life states that are worse than dying itself, hopes for the future, etc. In what ways, if any, has talking about these affected your ACP conversation with your family/loved ones?
   • - What would you tell someone else who was faced with a serious illness like you about the study? Would you recommend this conversation to them?
     If no: what are some reasons that you haven’t had these conversations with your family/loved ones?
     Can tell the participant: I have two more questions that are a bit heavy and may bring up some emotions. You may skip them if you do not feel like answering. Would it be okay if I continue with these last two questions?
5. If heaven-forbid you become sick and cannot speak for yourself, in what ways if any, did our study affect how you think about this situation?

   • - How did the study affect your own decision making for end-of-life care?
     To wrap up: Is there anything else that we should know about your experience in our study?

APPENDIX 2: Codes

Domain	Subdomain (n)	Codes (n)
Attitudes towards ACP	-	Raised awareness/clarity about wishes (16)
		Learned about ACP (7)
		Encouraged to take action (17)
		Making wishes known (3)
		No effect (15)
Actions towards ACP	-	Spoke with clinician (17)
		Spoke with family/loved ones (21)
		No action (16)
		MOLST (7)
		Living will (9)
		HCP/Power of attorney (5)
Conversation with clinician	Effects of ED GOAL on conversation (24)	Helped conversation (11)
		Motivated to initiate (6)
		Clarified specific wishes (8)
		No effect (7)
	Recommendations for others to have ACP conversation with clinician (24)	Would recommend (24)
	Reasons for no ACP conversation with clinician (28)	No opportunity (13)
		HCP aware, no need to tell clinician (2)
		Not a priority (6)
		Not sick enough (5)
		Don’t understand what ACP is (3)
		Uncomfortable (1)
Conversation with loved ones	Effect of ED GOAL on conversation (40)	Lessened the burden (2)
		Family prepared (4)
		Helped conversation (12)
		Clarified specific wishes (14)
		Motivated to initiate (16)
		No effect (9)
	Recommendations for others to have ACP conversation with loved ones (40)	Would recommend (36)
		Would not recommend/maybe (4)
	Reasons why patients did not have ACP conversation with loved ones (12)	No opportunity (1)
		Don’t understand what ACP is (2)
		Not a priority (2)
		No one to talk with (2)
		Family not ready (3)
		Already done (3)
Effects on attitudes towards crisis (1 not asked)	Positive effect (33)	N/A
	Negative effect (0)	N/A
	No effect (18)	N/A
Effects on end-of-life decision making (4 not asked)	Positive effect (27)	N/A
	Negative effect (1)	N/A
	No effect (20)	N/A

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.