Exploring the Acceptability and Relevance of Tool-supported Advance Care Planning (ACP) for a Long-term Care (LTC) Home Environment

Objectives: Despite known benets, advance care planning (ACP) is rarely a component of usual practice in long-term care (LTC). A series of tools and workbooks have been developed to support ACP uptake. Yet, their acceptability and relevance for LTC has yet to be examined. This study explored the extent to which available ACP materials hold promise in improving ACP engagement in LTC by (1) exploring LTC home staff’s reactions to tool supported ACP and (2) examining if available ACP tools include content of relevance to LTC. Methods: A combination of focus group deliberations with LTC home staff (N=32) and content analysis of publicly available ACP workbooks (N=32) were used to meet the study aims. Results: Focus group deliberations suggested that tool-directed ACP is a promising approach for LTC, provided tools include psychosocial elements and are paper-based. Content analysis of available tools revealed that only a handful of paper-based ACP tools (32/611, 5%) include, psychosocial content, with most encouraging psychosocially-oriented reections (30/32, 84%), and far fewer providing direction around other elements of ACP such as communicating psychosocial preferences (14/32, 44%) or transforming preferences into a documented plan (7/32, 22%). Conclusions: Tool-supported ACP appears acceptable to LTC staff. To improve ACP uptake in LTC selected tools should include psychosocial content that can be supported by a range of clinical and non-clinical staff. Available ACP tools may require infusion of scenarios pertinent to frail older persons, and a better balance between psychosocial content that elicits reections and psychosocial content that supports communication.

Advance care planning (ACP) is an important component of a comprehensive palliative approach program because it empowers individuals with non-reversible health conditions to re ect on, communicate, and sometimes document their values, beliefs, and preferences for future end-of-life care (Cornally et al., 2015;Rietjens et al., 2017). As such, ACP provides a mechanism to ensure that persons with life-limiting conditions remain at the centre of their own care by providing avenues for communicating care preferences to family/close friends, legally appointed decision-makers and health providers, when capacity for re ection, communication, and decision-making is consistently present (Fleuren, Depla, Janssen, Huisman & Hertogh, 2020;Howard et al., 2018;Sudore & Friend, 2010).
When ACP is introduced in LTC, it is associated with reductions in hospital admissions, increased concordance between preferred and received care, and decreases in stress, depression, and anxiety for residents and their families (Brinkman-Stoppelenburg, Rietjens., & van der Heide, 2014;Cornally et al., 2015;Robinson et al., 2012;Shanley et al., 2011).
Despite the known bene ts, ACP is rarely a component of usual practice in LTC (Ampe, Sevenants, Smets, Declercq, & Van Audenhove, 2017;Jeong, Higgins, & McMillan, 2011). Key barriers to ACP engagement include lack of resident and family preparedness (Jimenez et al., 2018;Lund, Richardson, & May, 2015), limited staff time, and discomfort and uncertainties from all parties regarding what to discuss beyond single medical decisions and funeral planning (McGlade et al., 2017;Reinhardt, Downes, Cimarolli, & Bomba, 2017;Sussman et al., 2017a;van Soest-Poortvliet et al., 2015). ACP programs that include multiple-steps have shown some successful outcomes in increasing ACP engagement in LTC (Ampe et al., 2017). However such programs have proven di cult to sustain into daily practice in LTC where access to clinical staff is exceptionally limited (Ampe et al., 2017). Hence within these settings ACP materials designed for self-use could be of particular bene t (Fryer, Bellamy, Morgan, Gott, 2016;Kontos, Miller & Mitchell, 2009).
Interactive tools like workbooks and card games that require limited professional support have recently been developed to encourage holistic conversations about EOL issues with one's family and healthcare staff throughout the general population (Butler, Ratner, McCreedy, Shippee, & Kane, 2016;Van Scoy, Reading, Scott, Green and Levi, 2016). Some such tools could speci cally support ACP uptake in LTC, by (a) increasing resident and family awareness of the importance of ACP conversations and (b) providing guidance about topics to re ect on and discuss (Bridges et al., 2018). Yet, the extent to which these available materials hold promise in improving ACP uptake in LTC has not been explored.
Redressing this gap in the literature this study combined focus groups discussions with LTC staff and a content analysis of available ACP tools to explore (1) what, if anything, would make tool-supported ACP an acceptable and feasible method of improving ACP in LTC, and (2) does the content and format of available ACP tools align with the needs and preferences of clinical and support staff in LTC?

Methods
This paper reports ndings from two initiatives: (1) a series of focus groups held with LTC home staff to explore the acceptability of tool-supported ACP in LTC, including content and format that would support feasible implementation, and (2) a content analysis of publicly available ACP tools, including material identi ed by focus group participants as relevant for a LTC home context.
The research was conducted in accordance with the standards of the Tri-Council Policy Statement for Ethical Conduct for Research Involving Humans 1998(with 2000, 2002. Procedures were approved by the O ce of Research Ethics Board at X (Blinded for Review) University and Y (Blinded for review) University.
Step One: Focus Groups:

Focus Group Recruitment and Data Collection:
In order to rst explore the acceptability and feasibility of tool-supported ACP in a LTC environment, staff in four LTC homes in southern Ontario were invited to participate in focus group deliberations. Staff were known by the research team to be palliative leaders in their respective LTC homes, as this was a component of a larger study in all four LTC homes aimed at strengthening a palliative approach in LTC (Blinded for review).
Recruitment took place one month prior to each focus group and was limited to staff who had been active members of palliative care champion teams (n = 55). All potential participants were contacted directly by the research team via email or verbal invitation. 32 of 55 eligible staff participated in four focus groups, including 13 nurses, seven allied health professionals, four support staff, three clinical leaders, and ve trainees/volunteers. Focus groups lasted 60 minutes and were facilitated by two members of the research team. Participants were rst provided with a range of ACP tools including one card game, two workbooks (one of which culminated in the development of advance directives and another which did not) and one worksheet with an accompanying video. Participants were then invited to share any reactions to the tools and discuss their thoughts on implementing them in a LTC environment. Written informed consent was attained on the day of focus group deliberations.

Data Analysis Focus Groups:
All focus group deliberations were audio-recorded and transcribed, and eld notes were taken throughout. Transcripts and eld notes were analyzed by two members of the research team in two steps guided by the principled of content analysis (Neuendorf, 2019). In step one, researchers independently and then together coded comments about each tool as either strengths or limitations. In step two, commonalities and discrepancies between tools and across groups were examined to identify if particular tools were consistently viewed either positively or negatively and to illuminate overall themes related to the strengths and limitations of all tools to support ACP discussions in LTC. The following four themes emerged as a result of this two-step process as together they were thought to capture a comprehensive narrative of staffs' reactions to the materials ( : (1) Tools with su cient information and direction are key; (2) Psychosocially-focused tools seen as preferable; (3) Workbooks are a preferred medium for ACP tools; and (4) No one tool will t all.
Step Two: Content Analysis Search Strategy: Informed by ndings from our focus group deliberations, we aimed to identify and analyze publicly available paper-based ACP workbooks with psychosocial content that could be easily accessed and utilized by residents, families, and staff in LTC. We sought to identify workbooks through a grey literature search using Google as our search engine. The key words "Advance Care Planning Tools" were used to locate paper-based materials. We also solicited Key Informants for additional tool identi cation.
Guided by the themes emerging from our focus group deliberations, tools were excluded from review if they (1) focused exclusively on recording medically-oriented advance directives or decisions including donot-resuscitate orders or the cessation of particular treatments such as kidney dialysis for advanced renal failure; (2) were not available in printable workbook form; and (3) did not provide su cient information and direction for use by non-clinically trained staff. The research team also decided to exclude tools that were (4) developed for substitute decision-makers only because we were interested in materials that would support the inclusion of residents in their own future care planning (Sussman et al., 2017b;2019).
Two researchers were involved in determining tool retention/exclusion. The rst researcher conducted an initial review of identi ed tools once exclusion criteria were established. The second was consulted with a random selection of tools slated for exclusion and purposefully when uncertainties arose.
A total of 611 tools were identi ed through our key word Google search, of which 489 (80% of identi ed tools) were excluded because of their sole medical focus. An additional 49 tools (8%) were excluded because they were not paper-based and/or required internet access to be utilized and 32 tools (5%) could not be accessed because the website housing them had expired or the tool was otherwise unavailable. A further 12 tools (2%) were excluded because they were designed solely for clinicians or decision-makers.
Hence, we retained 29 ACP workbooks for further review. To ensure we had located all viable paper-based ACP materials, we emailed the list of 29 tools to our broad network of national and international collaborators and researchers engaged in the areas of palliative care, long-term care, and aging (N = 75). This process yielded an additional 3 tools. A total of 32 ACP workbooks were retained for full content review.

Content Analysis Acp Tools:
The content of the workbooks was analyzed in three stages by the two researchers responsible for tool selection (Hsieh & Shannon, 2005). In the rst stage we used conventional content analysis to categorize psychosocially-oriented excerpts from each guide into the broad ACP areas of re ection (components designed to encourage thinking about personal values, goals, and preferences), communication (components designed to encourage conversations between patients, decision-makers, and others, including steps to take to begin the conversation), and documentation (components left for patients to record any wishes or preferences that could be referred to by others). These aspects are commonly denoted in the literature as critical and unique steps in the ACP process (Sudore et al., 2013;Sudore & Fried, 2010;You et al., 2014).
In the second stage, through discussion and a collective review of coded excerpts, the team further re ned the categories for re ection, communication, and documentation into what we considered vague psychosocial content (e.g. re ecting on what one considered quality of life) and speci c psychosocial content (e.g. re ecting on music preferences at end-of-life). We also speci ed whether psychosocial re ections, communication, and documentation were geared towards future EOL care or on designating a decision-maker, as both are considered unique aspects of ACP (Sudore et al., 2013;Sudore & Fried, 2010;You et al., 2014). Noting the extent to which substitute decision-makers were encouraged to participate in re ections, communication, and documentation seemed important both because these distinctions were emerging from the content and because recent models of ACP recognize that decision-maker involvement in all stages of the ACP process can encourage ACP activation for older adults living with frailty (Piers, et al., 2018;Wendrich-van Dael., Bunn, Lynch, Pivodic, Van den Block, & Goodman, 2020).
In the third and nal stage, the frequency with which each tool had psychosocial content falling within each category was recorded to illustrate the most and least common applications of psychosocial content. At this stage two researchers independently coded content. Any identi ed discrepancies were discussed with a third researcher who helped make a nal determination based on consensus. We also created formatting-related categories noted in the literature to support readability for older persons, including tool length, use of pictures/images, and inclusion of scenarios applicable to older persons (National Institute on Aging, 2007).

Focus Group Results:
Deliberations in all focus groups suggested that staff consider tool-supported ACP to be a promising approach to trigger more consistent ACP discussions between residents, families, and LTC staff. While staff felt such tools could not direct when discussions should happen, many expressed that with the proper content and format, ACP tools could provide important direction for open, honest, and comprehensive discussions with residents and families. The following sections highlight the content and format considered most acceptable and feasible for a LTC environment.
Tools with su cient information and direction are key: ACP tools with su cient information and direction were seen as particularly helpful to guide ACP discussions between staff, residents, and families. As one participant stated, "this [tool] can be used as a great guide to staff when talking with families" (Focus Group 1). Others re-iterated that tools that offered "good ways to start a conversation" (Focus Group 2) and "good direction about what to ask" (Focus Group 4) were useful as they would provide staff with the support they need to activate such conversations.
Psychosocially-focused tools seen as preferable: Focus group deliberations further suggested that staff were apprehensive about tools with a strong medical focus. This theme was particularly evident when discussing one selected tool which had an associated advance medical directive. This tool was consistently viewed as "too medical" (Focus Group 2), "too harsh" (Focus Group 4), and "not the best t for a LTC setting"( Focus Group 3). Overall, participants within and between groups suggested that "having psychosocial aspects covered in a tool is good" (Focus Group 4) and would enhance uptake in LTC. Psychosocial topics considered of high relevance in a LTC environment included views on family involvement, beliefs about quality of life/death, environmental conditions thought to enhance care, and religious/spiritual preferences.
Workbooks are a preferred medium for ACP tools: While all tools reviewed had strengths and weaknesses, participants across groups expressed preferences for paper-based workbooks over videos, card games, and web-based materials. This medium was favoured because it could include education/information alongside interactive elements, and did not depend on computer accessibility.

No one tool will t all:
Because ACP is about exploring individual wishes, beliefs, and preferences, most attendees suggested that no one tool would ever be appropriate for all LTC residents. Rather, participants expressed the desire for a list of potential tools they could refer to. As one participant stated about the tools, "We need different ways of delivering [ACP] as not every resident will like any one option." (Focus Group 2). Content Analysis Results: Table 1 provides a brief overview of all 32 tools retained for analysis including where they can be located and whether they have been evaluated in the scholarly literature. It is noteworthy that of the 32 tools located only three appear to have been formally evaluated as evidenced by a search in the published literature and email-outreach to all organizations wherein tools were housed. Table 2 reports results of our content review. These ndings are also reported below.

Readability And Formatting:
Retained tools ranged in length from 3-102 pages. Although a small minority were less than 12 pages (6/32, 19%), one third (11/32, 34%) exceeded 25 pages. Most tools in our review (21/32, 67%) included photos or graphics and offered vignettes or case situations to encourage ACP re ection (18/32, 56%). Workbooks often featured photos that appeared to be of people having discussions with loved ones, or graphics or symbols depicting certain healthcare scenarios such as hospitalization. Half of the tools reviewed included vignettes about the care of an older person (16/32, 50%), such as imagining oneself having Alzheimer's disease and being unable to recognize loved ones or communicate with them, or of someone contracting pneumonia and moving from a nursing home to a hospital.
Psychosocial Content In Tools: Psychosocial Re ections: The vast majority of tools included in the review encouraged some form of psychosocial re ection (30/32, 94%).

Communication about Psychosocial Issues:
All tools (32/32, 100%) mentioned the importance of communicating and sharing re ections with families, friends, healthcare professionals, and substitute decision-makers, and many (27/32, 84%) encouraged users to think about who they may select as a substitute decision-maker in light of their re ections and preferences. Yet far fewer tools (14/32, 44%) provided users with tips on how to initiate conversations with families, friends, and decision-makers, how to speak with their decision-makers about their comfort acting in the role in light of their wishes (13/32, 41%), or provided guidance directly to substitute decision-makers regarding what they may ask or think about (7/32, 22%).
When communication tips were provided they tended to be general in nature, such as encouraging people to nd the right time to have a conversation or providing them with an opening prompt such as "I need to think about the future. Will you help me?" (Conversation Starter Kit). Only a handful of workbooks (5/32, 17%) speci cally encouraged the user to speak about their psychosocial preferences for end-of-life care with those close to them, via prompts including "…ACP Conversations are a process so you do not have to think about this until you are ready. But if you have thought about it, tell your SDM(s), family and friends: what is important to you at the end of your life… music you want to listen to, books you want to read or have read to you…" (Acclaim Health) or "What is important [in the following exercise] is that you understand what each person involved in your conversation wants for himself or herself… Would you like someone to be with you when you die? Who?" (Critical Conditions Planning Guide).

Documentation of Psychosocial Directives:
While documentation is a common element of medically-oriented ACP workbooks, only a fth (7/32, 22%) of the workbooks we reviewed included a formal opportunity to document psychosocial preferences for EOL. Those that included a documentation component encouraged users to turn their re ections into a formalized plan that identi ed speci c psychosocial directives such as "I wish to have religious readings and well-loved poems read aloud when I am near death" (Five Wishes); "I want X at my bedside [at end of life]" : … [I] want someone to hold [my] hand" (Begin the Conversation). While not necessarily legally binding, this process ensures the translation of re ections and discussions into actionable conditions that users hoped would be applied in their nal moments of life.

Discussion
This study used the combination of focus group deliberations and content analysis of existing ACP tools to explore the extent to which existing ACP materials hold promise for improving ACP uptake in LTC. Our focus group deliberations suggested that tool-supported ACP seems acceptable in LTC, especially in its potential for directing conversations about end-of-life topics. More speci cally, LTC staff considered paper-based workbooks to offer viable direction for ACP uptake if such material included information about the importance of ACP and interactive exercises and prompting questions stimulating re ection and communication. Focus group deliberations further uncovered that tools that were too medicallyfocused were considered ill-suited for LTC. Yet our content review identi ed that a striking 80% (489/611) of existing ACP tools focused solely on supporting discussions and documentation of preferred medical care at EOL including non-resuscitation and non-intubation orders (i.e. advance directives). Current ACP research in LTC a rms the high prevalence of medically-focused ACP materials, as most studies exploring ACP impacts in LTC actually look more speci cally at the implementation of advance medical directives (Capps, Gillen, Haley & Mason, 2018). While outcomes from this work have certainly shown some promise, our ndings suggest such interventions are unlikely to be successfully adopted into usual practice in LTC if psychosocial issues are excluded, such as preferences for family involvement/noninvolvement in EOL care, views about dying alone, spiritual beliefs that may provide comfort at EOL, and values related to quality of life/quality of care (Brinkman-Stoppelenburg et al., 2014;Sussman et al., 2017b).
Quite possibly the recurrent distribution and testing of medically-oriented ACP materials also perpetuates the limited understanding of ACP amongst residents and families in LTC reported in the literature (Mignani, Ingravallo, Mariani & Chattat, 2017). Until such time as ACP is understood more broadly to include re ections on values, beliefs, and preferences that can be used to inform in-the-moment decisions rather than solely on the identi cation of pre-speci ed medical decisions that may not be easily implemented within unforeseen contexts, it will not be consistently implemented in LTC or elsewhere (Howard et al., 2018;You et al., 2014).
Our initial screening of available ACP workbooks resulted in the retention of 32 tools found to include some psychosocially-oriented content. The vast majority of these tools encouraged re ections about psychosocial issues pertinent for EOL decision-making in LTC such as beliefs about quality of life/death.
Many also went a step further to help users unpack such re ections by orienting them towards smaller, more speci ed re ections such as preferred location of death, rituals of importance, and speci c actions that may provide comfort. Encouraging speci c re ections that can be more easily communicated has been found to be a useful approach to ACP activation (Sudore & Fried, 2010).
However, most of the tools we reviewed included less content and direction around how to communicate psychosocial preferences to decision-makers, family/friends, and health providers or how to select decision-makers and speak to them about their roles and responsibilities. Further, and almost none of the tools offered formalized directions on how to transform wishes and preferences into a documented plan that included psychosocial elements. While some have argued that pre-prescribed plans are of limited use because they fail to account for the speci c and contextualized decisions that emerge at EOL (Tulsky, 2005), selection of and communication with substitute-decisions makers and health providers about roles, responsibilities, and preferences has been consistently recognized as a critical component of ACP (Bridges et al., 2018;Howard et al., 2018;Sudore & Fried, 2010;Sudore et al., 2008;You at el., 2014). Although workbooks alone may not be su cient to support the movement from re ection to oral or written communication, more orientation towards this challenging aspect of ACP could hold promise for a LTC home environment by providing residents, families, and staff at all levels with tips and directions that may encourage communication on this emotionally complex and di cult topic.
In terms of applicability to a LTC home environment, two thirds of the tools reviewed were a moderate length (12-25 pages), most included graphics and images supporting readability, and half infused scenarios and examples relatable to circumstances/situations older persons living with frailty may face. Hence, some of the psychosocially-oriented workbooks and materials available may still require adaptations for a LTC environment.

Implications And Recommendations:
Our ndings provide some direction to improve ACP engagement and uptake in LTC. First, our review identi ed a series of tools that include psychosocial components. We encourage LTC home administrators and directors of care to use the list provided as a starting point to select material for possible distribution. We also encourage them to reconsider any tendencies to use materials with a strictly medical focus. Second, prior to implementation within a selected LTC home environment, we suggest preferred material be altered (if required) to include typical LTC home scenarios that may encourage ACP re ections and to ensure a relatively equal balance between prompting questions and exercises that encourage re ections and those that encourage communication and documentation. With the advent of the COVID-19 pandemic, additional adaptations or companion resources may be crucial to ensure the relevance of these tools. A number of the tools we identi ed have already evidenced this, including the Speak Up tool (https://www.speakupontario.ca/wp-content/uploads/2020/04/Engaging-in-Advance-Care-Planning-for-COVID-19.pdf). Third, given the dearth of evidence located on the effectiveness or implementation of the identi ed tools we strongly encourage more research on ACP in LTC that evaluates the identi ed psychosocially-oriented ACP materials. At the time of writing, we were only able to locate published studies that empirically evaluated three of the tools we reviewed (see Table 1). While our review suggests these materials are of high relevance to a LTC environment, studies evaluating the distribution and use of such materials in LTC are warranted to maximize impact and uptake.

Limitations
At the time of writing electronic materials were considered inaccessible for many residents in LTC and hence our review was limited to paper-based resources. In response to the COVID 19 pandemic and the need to ensure connections between residents and families, electronic communication has become more common in some LTC environments. It is possible that more psychosocially-oriented ACP tools are available in these digital formats and it may be timely to revisit the feasibility of electronically-based ACP materials in LTC.
We did not perform a full systematic review of published work that has evaluated our identi ed tools, and therefore may have omitted some relevant studies. However, our results made clear that the vast majority of the tools on our list have not been evaluated.
Our review failed to distinguish between workbooks that had some content applicable to a frail older population and those that were predominantly focused on issues related to aging and frailty. Hence our estimate of available psychosocially-oriented workbooks applicable to older persons in their current form may be in ated. We consider this a minor limitation as we expect all identi ed tools to require some adaptations to ensure applicability in local LTC environments.

Conclusions
A handful of ACP tools currently exist that hold promise for improving uptake of ACP re ection in LTC.
However, these psychosocially-oriented tools may require infusion of scenarios pertinent to frail older persons, re-formatting, and a better balance between prompts and exercises encouraging re ections and those geared towards supporting communication to ensure applicability for LTC and to support communication between residents, decision-makers, and staff. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. A list and location of all ACP tools included in the content analysis (Phase II) is provided in the paper (see table 1).