Based on exploratory qualitative analysis of data from both non-profit and for-profit RCH and ALF services about staff experiences of using the DöBra cards to discuss future EOL care, we were able to clarify the perceived outcomes, impacts and usability of a novel EOL conversation tool. This is particularly relevant in contexts, such as Swedish elder care, where such discussions are infrequent. We found that the cards can be directly useful, by providing a framework to structure EOL conversations, helping to elicit valuable information about residents’ EOL values and preferences, and enabling an interpersonal connection that strengthens rapport. In addition, our findings suggest that the cards may have more long-term impact by encouraging continued communication among stakeholders, with or without staff facilitation, and aligning care goals by supporting discussions to clarify what matters at the EOL. Factors that influenced use and usability related to both characteristics of staff members, and contextual features, based on the needs and health status of the resident involved, as well as organizational features.
Given the novelty of facilitating proactive EOL conversations in elder care, and using a conversation tool to do so, a qualitative study design was crucial to comprehensively explore staffs’ experiences, and reflections regarding use and usability. Our prolonged engagement with participants may have contributed to a sense of trust for participants to openly share their insights and experiences and engagement in the study was high, suggesting that the study focus was perceived as relevant. Furthermore, using a participatory approach during data collection enabled preliminary findings to be discussed with participants throughout. This was particularly the case in the working group in Database 2, where sharing and discussing data and initial interpretations was part of the process of co-developing the practical guidance. While member-checking does not verify results, general agreement and recognition of findings strengthens their trustworthiness [38]. Nevertheless, it should be remembered that these data reflect only participating staff’s views of the EOL conversations and subsequent occurrences. It is worth noting that only staff who contacted the authors after having used the DöBra cards were interviewed for Database 1. It is possible that other staff utilized the tool without notifying us. Staff with more negative experiences might have been less likely to make contact, leading to under-representation of such experiences. Likewise, in Database 2, it is possible that participants who remained in the working group were more positive to testing an EOL conversation tool than others. Thus, self-selection may constitute a potential bias that should be remembered when considering transferability of the findings.
Unlike other, often script-based, EOL conversation tools, the cards provided examples that could be freely explored, serving as a framework stimulating in-depth reflection about EOL values and preferences. Our study adds more detailed understanding about how using a tool in EOL conversations can affect several important aspects of interaction. In addition to helping staff address and ask about residents’ EOL values and preferences, the cards stimulated a reflective process and provided interactive sharing of information, stories, and emotions. These findings highlight the interpersonal aspects of EOL conversations, e.g. strengthening relationships and developing shared narratives and goals, which have been shown to act as major contributors to ACP benefit [39], particularly with older populations [40, 41]. We found, as did Sussman et al. [42], that interactive tools, such as card games, help target reflection and can cover a variety of aspects of EOL care. However, since most EOL conversation tools focus primarily on medical aspects of care, other important dimensions risk being overlooked [42]. Prior research has suggested that non-medical issues may even be more imperative to discuss as advanced age affects care preferences [43]. As the DöBra cards cover physical, practical, existential, and social matters, they offer a more comprehensive perspective on EOL values and preferences beyond medical treatment options alone.
The DöBra cards were found to be useful to strengthen person-centered care provision and rapport both during and after EOL conversations. This is in contrast to a recent study by Groebe et al. [44], in which some care staff considered EOL conversation tools to be counterintuitive to an individualized care approach. It may be that the physical format of the DöBra cards better allowed residents and/or relatives to be actively involved in directing the discussion than other conversation tools, as the cards they choose served as route markers for mapping the discussion, with as much – or little – commentary as residents and/or relatives wished. However, this required considerable flexibility and attentiveness from staff in facilitation. Our findings thus highlight the delicate balance between seeing the cards as a tool to complete a task with a set goal on the one hand or as a trigger for an unfolding conversation in which involved parties together determine issues important to discuss on the other. This raises questions about how underlying goals of EOL conversations should be negotiated and determined. Using the cards might create forums for sharing stories or thoughts that residents and/or relatives need to express and allow discussion of topics that otherwise would not be addressed, making them meaningful experiences, even if they do not directly contribute to, or even risk hindering, completion of the card exercise as planned. We propose that staff instructions for card use need to be carefully considered as to not create perceptions of failure if the card exercise cannot be completed as suggested, which is important to consider when designing competence-building initiatives and in future research on tools for EOL conversations.
The results of this study also strongly suggest that there is no optimal ‘one size fits all’ procedure for using the cards and that each EOL conversation needs to be adjusted to fit the needs and constraints those participating, in line with suggestions in prior research [45, 46]. The flexibility of DöBra card use therefore appears to be a key feature for their potential usability in RCHs and is particularly important if the tool is to be used with residents with cognitive decline. In general, EOL conversations are rarely conducted with residents with dementia [47] and participants’ experiences highlighted cognitive function as a challenge, even though this study primarily involved residents who staff considered cognitively competent. The expected increase in dementia prevalence further emphasizes the need to find feasible ways to support EOL conversations also with this population [48, 49]. In these cases, while the ranking exercise appeared too complex, the card statements were still useful as probes to clarify what matters at the EOL, as also noted by Eneslätt et al. [26]. Nevertheless, the usefulness of the DöBra cards to discuss EOL values and preferences with residents with moderate and severe cognitive impairment need to be studied further [50].
This study adds to the extant research that indicates that communication skills are a precondition for EOL conversations [51–53]. The direction and depth of EOL conversations seemed largely influenced by staff’s interest and curiosity about what matters to other people. Additionally, the practice of ‘holding space’, i.e. actively listening, being mentally and emotionally present, and setting aside one’s own agenda to allow the other person to lead [54] appeared to constitute a key skill for facilitating EOL conversations described as richer and more memorable. This highlights that EOL conversations differ from other conversations, by requiring staff to respond to existential needs and shift their focus from ‘doing’ towards ‘being’ and listening [55]. Building such skills may require a more experience-based approach to training, as suggested by Sand et al. [56].
The influence of contextual factors on implementation of EOL conversations is well known and this study corroborates several previously identified barriers, such as unclear mandates, under-staffing, and negative staff attitudes [57–59], as well as new prerequisites, e.g., related to systems for shared written documentation. In addition, we extend understanding about how time constraints influence not only prevalence of EOL conversations, but also their depth and salience, as stress when EOL conversations took longer than expected was a source of frustration or impatience among participants. These findings point to fundamental challenges in introducing new processes in care systems already pressed for time and resources, and as argued by Lund et al. [51], it may be that until these are dealt with, the benefits of using tools to support EOL conversations will be limited. Impact of wide-scale implementation of EOL conversations using the cards into routine RCH practice thus remains a critical question for future research to explore.