Social Interaction

1. Introduction

Several policies and initiatives in Western countries aim to move death and dying from the realm of the unspoken to the sphere of open debates and interpersonal communications (Hospice UK, 2023; Impermanence, n.d.; National Palliative and End of Life Care Partnership, 2021). The assumption is that open discussions about the end of life (EOL) can benefit people, especially those for whom the prospect of dying is nearer, including older and frailer people and people who live with advanced and incurable conditions. There is evidence that timely discussions about disease progression and EOL can afford people opportunities to make their wishes for future care known, reduce futile treatments and hospital admissions, and improve the bereavement experience of those who are important to them (Brighton & Bristowe, 2016; Jimenez et al., 2018; Keeley, 2017). A common theme in the literature is that these discussions do not happen enough (Hancock et al., 2007). It is frequently reported that those who interact with older people and people with advanced illnesses discourage talk about mortality (e.g., Alftberg et al., 2018). This includes people who can be characterised as companions: informal caregivers, family members, and friends, who play an important part in a person's transition towards the EOL. Some previous studies have provided empirical support for these claims.

Some older people participating in a US focus group study reported that their relatives discouraged them from talking about EOL issues (Egbert et al., 2017). In a US focus group study on the experiences of bereaved family caregivers, Kellas et al. (2017) suggested that some caregivers "avoided" conversations about mortality in their interactions with cancer patients and that they performed "fake hope" instead. According to the authors, this happened especially when the caregivers, alongside their family and social networks, maintained a focus on curative treatment and did not openly acknowledge the possibility that the patient might die from cancer. By contrast, caregivers who seemed to accept that possibility appeared more open to discussing mortality with patients. Those caregivers nevertheless seemed to balance honesty with hope in their interactions with patients (Kellas et al., 2017). In another US study, hospice volunteers reported observing that family members sometimes refused to discuss death with patients (Planalp & Trost, 2008). The volunteers attributed this to a state of "denial" that the family members were arguably in (Planalp & Trost, 2008).

Other authors have suggested that the presence of companions (family members and other informal caregivers) in interactions between patients and healthcare practitioners (HCPs) makes discussions about EOL issues less likely. For example, in a discussion piece about palliative care in India, Chaturvedi et al. (2009) suggested that relatives often "collude" with HCPs to withhold prognostic information from patients. A quantitative observational Australian study of recorded palliative care consultations documented that patients report more emotional contents when seeing an HCP on their own, without a companion (Swetenham et al., 2015).

The literature reviewed so far suggests that companions sometimes discourage discussion about EOL matters when interacting with patients. Three considerations are important here. First, previous studies have largely relied on patients' and companions' reports (e.g., in interviews and focus groups) of their own prior experiences. Although these provide indirect evidence about whether and how companions engage or disengage with discussions about patients' mortality, they do not clarify how this happens in practice. Direct observation of real-life interactions can contribute to addressing this gap.

Our second point is about several authors' observation that companions sometimes "avoid" discussions about mortality (Egbert et al., 2017; Kang, 2021; Kellas et al., 2017; Planalp & Trost, 2008). According to existing analyses, companions can "deny" the possibility of the patient's death (Kellas et al., 2017; Planalp & Trost, 2008) or "refuse" to discuss mortality (Kellas et al., 2017), sometimes by "colluding" with HCPs (Chaturvedi et al., 2009). "Avoidance" characterises companions' actions in negative terms, for what they do not do. We will show that several companions' actions are best characterised as pursuing their own locally relevant interactional projects. In turn, these actions can have, as one of their outcomes, that they do not promote elaboration of matters pertaining to disease progression or EOL when this is a relevant possibility. We will demonstrate that these actions cannot be straightforwardly labelled as "avoiding" engagement with EOL issues.

Our third consideration is about cultural and socio-structural factors that have been proposed to affect the extent and ways in which people engage with discussions about mortality. On the one end, public campaigns, such as Dying Matters in the UK (Hospice UK, 2023), promote a culture of openness regarding mortality. Some indirect evidence that these initiatives may be having an impact is that, when asked about their preferences for communication, several patients and companions say that they want at least some discussion about future disease progression and care options (Parker et al., 2007). On the other hand, disease progression and EOL remain fraught subjects in Western contexts. Some authors have proposed that death is a reminder of human vulnerability, which clashes with capitalist and patriarchal values of self-determination, competition, and acquisition in Western societies (Harris, 2009; Kang, 2021; Keeley, 2017). This would lead people to avoid discussions of EOL issues because they are reminders of their own fragility (Harris, 2009). Others factors that may discourage people from openly discussing mortality include an imperative to maintain a positive attitude in everyday social life (Pollock et al., 2022) and a curative ideology that approaches death as an outcome that should be staved off through life-extending treatments—rather than a natural phase of the life cycle that everyone should prepare for (Alftberg et al., 2018). In the context of these contradictions around the desirability and practicality of initiating and sustaining discussions about disease progression and EOL, we can anticipate that people face situated tensions and dilemmas at the ground level of their everyday interactions. Along these lines, Egbert et al. (2017) proposed that family members face competing goals in interactions with older people. Kellas et al. (2017) similarly noted that family caregivers face conflicting communication demands in interactions with cancer patients. Missing in the literature is nevertheless an examination of how these tensions play out in naturally occurring interactions.

In summary, previous studies have not directly observed how companions' actions contribute to engaging or disengaging with explicit and focused discussion about disease progression and EOL. We suggest that this has contributed to a limited understanding of companions' actions in the context of discussions about mortality. The prevailing take in the literature is that companions can "avoid" or "refuse" to engage with EOL issues. We propose that an observational analysis of recorded real-life interactions is needed to gain a more rounded and nuanced understanding of how companions' actions can facilitate or hamper discussions about mortality. As we will show, companion actions that are not primarily designed to "avoid" discussion of EOL issues but, rather, to pursue other interactional projects, can have as one of their outcomes that they do not promote explicit and focused discussion about EOL matters when such discussion is a relevant possibility.

In this article, we examine a corpus of audio-visually recorded palliative care consultations involving HCPs, patients with advanced conditions, and their companions. The latter are family members and friends who accompanied the patients at their appointments at a UK hospice, and who were involved in supporting them. The patients had been diagnosed with life-limiting, incurable conditions and referred to the hospice for review and management of physical and psychological symptoms and sometimes to discuss and plan for palliative and EOL care. This makes this setting conspicuous for an investigation of talk about disease progression and EOL.

Our aim is to examine companions' actions following patients' possible allusions to disease progression and EOL. We identify the interactional projects (Schegloff, 2007) that the companions' actions observably advance, and their implications for the possibility of engaging or disengaging with explicit and focused discussion about disease progression and EOL. Our study thus advances understandings about how companions' actions affect the possibility of engaging in discussion about patients' mortality. We will discuss the theoretical and practical implications of our findings for understandings of companion participation and of discussions about disease progression and EOL in healthcare interactions.

2. Methods

Data for the study consist of 72 audio-visual and 13 audio recordings of outpatient and inpatient consultations collected in a large UK hospice. Participants included 85 patients, 38 companions, six palliative medicine doctors, three physiotherapists, and five occupational therapists. Companions included family members (most of whom were partners), friends, and one professional caregiver. Ethical approvals were obtained from UK NRES Committees Coventry & Warwickshire (Ref: 14/WM/0128) in 2014, and Nottingham 2 (Ref: 17/EM/0037) in 2017. All participants gave written informed consent to have their interactions recorded and analysed, and for pseudonymised transcripts to be included in published reports. The interactions were in British English.

The analyses we report are part of a broader programme of research on interactions in palliative and EOL care, one strand of which focuses on ways in which discussions about disease progression and EOL are initiated. Using the method of conversation analysis (CA; Sidnell & Stivers, 2013), we previously identified actions that HCPs use to promote discussion about disease progression and EOL (Pino & Jenkins, 2023; Pino et al., 2016). We found that, recurrently, those actions are responsive to or occasioned by something that a patient or companion has raised either in the previous turn or earlier in the interaction, which can be understood (and appears to be treated by the HCPs) as possibly related to disease progression or EOL. We subsequently examined those patient and companion turns, and we found that they rarely raise disease progression or EOL explicitly. Rather, they present themselves as possible allusions to those matters. Following Schegloff (1996a), we defined allusions broadly as actions where something "can be seen to have been 'planted' (so to speak), or conveyed, in the preceding talk, but to have not been said 'in so many words,' or in any words 'explicitly'" (Schegloff, 1996a, p. 181; emphasis in the original). Central to Schegloff's and our own analyses (and in line with CA methodology) is agnosticism regarding whether an allusion is intentional (see Schegloff, 1996a, pp. 183-184). Whilst it is possible that a participant intended to "plant" a certain meaning or action, this is ultimately unknowable. Importantly, asking about participants' intentions is unnecessary to identify the outcomes of their actions for the possibility of engaging or disengaging with explicit and focused discussion about disease progression and EOL.

We assembled a collection of possible allusions¹ about disease progression and EOL produced by patients and companions.² We first built a core collection of instances where an allusion is demonstrably the starting point of subsequent explicit and focused discussion about disease progression or EOL; in these cases, there is evidence—in terms of the next-turn proof procedure (Sidnell, 2013)—that a participant treats it as having possibly alluded to disease progression or EOL. Following Schegloff (1996a, p. 189), we located some of these allusions "in retrospect" by first identifying an action that responded to or built on them and then "seeking [them] out in the earlier talk".

Our next step was to expand the collection to instances where possible allusions are not the starting point of subsequent talk about disease progression or EOL; in these cases, there is no next-turn evidence that a participant treats it as having possibly alluded to disease progression or EOL. To include these, we relied on the notion of possible action, that is, "a describable practice of talk-in-interaction which is usable to do recognizable [actions] [...] (a claim which can be documented by exemplars of exchanges in which such utterances were so recognized by their recipients)" (Schegloff, 1996b, p. 116). For example, having identified several instances where a mention of panic during episodes of breathlessness is recognised by participants as possibly alluding to EOL matters (Pino & Land, 2022), we included comparable instances where references to panic are not followed by these forms of recognition. Our wider collection thus comprises 90 possible allusions. The examples in this article have been selected from the core collection to provide internal evidence that the participants treat a patient's action as possibly alluding to disease progression or EOL.

Using analytic procedures in CA, we identified the type of action sequence in which each allusion is embedded, and the action accomplished by the turn containing it. We examined turn compositional and other features that make an allusion recognisable as such. These include participant orientations to shared knowledge about the patient's condition. For example, some reports of difficulties (such as experiencing depression or panic; Pino & Land, 2022) can be heard as possibly related to disease progression or EOL in the context of shared knowledge about the patient's advanced and incurable condition. Additionally, aspects of the institutional setting can explain why the HCPs sometimes actively follow up on and explore those possibilities. Part of their professional remit is to support patients and companions in discussing and preparing for the future (including deterioration and EOL). In this context, some of the HCPs in our data (especially the doctors) appear to treat aspects of patients' and companions' talk that can be heard as possibly related to disease progression and EOL as opportunities to initiate discussion about these matters. Modifying Sacks' hearer's maxim (Sacks, 1972), we could say that in some of the interactions at the hospice, the HCPs orient to a rule whereby 'If a patient or companion says something that can be related to disease progression and EOL, then hear it that way' (we might add, regardless of whether the patient or companion intended it in that way). It is thus important to note that the hearability of an allusion results from a complex and reflexive relationship between its design and the context in which it is produced (including the broader institutional context and the interactional projects that can be pursued within it).

In this article, our aim is not to explicate how allusions are designed or the different activities and sequences in which they are used. Rather, our focus is on companions' actions that are positioned after patients' possible allusions to disease progression and EOL. To this end, we identified all instances in the wider collection where: (a) a patient produces a possible allusion to disease progression or EOL, and (b) a companion is the participant who takes a full turn next (sometimes after an HCP has passed the opportunity to take a full turn, for example through the use of continuers). This search resulted in nine instances occurring in seven consultations. These involved seven patients (five women, two men), seven companions (two wives, three husbands, one daughter, and one granddaughter), and seven HCPs (three palliative medicine doctors, two physiotherapists, and two occupational therapists; one of the doctors was a man, the other HCPs were women). The patients were aged 66-79. Their diagnoses included motor neurone disease, chronic obstructive pulmonary disease, bronchiectasis, pulmonary fibrosis, and cancer. One recorded interaction involved an inpatient; the other interactions were outpatient appointments. We transcribed relevant data segments using CA conventions for verbal/vocal conduct (Jefferson, 2004) and embodied/visible conduct (Mondada, 2018)—see the Appendix for details. The instances included in this article are the most concise to illustrate patterns across the collection.

3. Results

In all nine cases in our collection, the companions' actions have as one of their outcomes that they do not promote elaboration of matters pertaining to disease progression or EOL following a patient's possible allusion (in circumstances where such elaboration is a relevant possibility). The ways in which they do so vary depending on whether the patient has introduced possible considerations about disease progression and EOL explicitly or allusively. It is thus instructive to first consider the only case in our collection where a patient almost immediately moves from a possible allusion to an explicit consideration about the EOL. This case shows that a patient's explicit mention of EOL matters can be met with a companion's equally explicit contradiction.

Extract 1 is from a consultation involving Jason (P-Jas), a patient with chronic obstructive pulmonary disease, who is attending an outpatient appointment at the hospice with a physiotherapist, Sharon (PT), and is accompanied by his wife Julie (C-Jul). A physiotherapist trainee is also present but does not speak in this extract. Earlier in the consultation, Jason complained that he lacks energy for activities of daily living and gets out of breath without doing much. The physiotherapist recommended that Jason redistribute his morning activities to help conserve his energy. She used the metaphor of a battery to illustrate that point (data not shown). In extract 1, she reiterates that recommendation (lines 1-2 and 5). She supports it by describing the negative outcome that her recommendation would help prevent (running out of energy), and, to do so, she again uses the battery metaphor (8-9 and 11). She appends laughter at the end of this turn (line 11), possibly to mitigate the force of the recommendation (Shaw et al., 2013). Another possibility is that the laughter is designed to head off more troubling understandings of "you've got nothing left in your battery", such as the possibility that it alludes to further deterioration.³ Jason reciprocates the laughter (line 12). Possibly pursuing acceptance, the physiotherapist reiterates the point ("You've got nothing left in your battery", line 13). We focus on how Jason's response (line 15) works as a possible allusion to EOL matters, which he subsequently makes explicit (line 17), and on Julie's response (line 18). In the extracts, we have highlighted the patient's allusion in grey and the companion's subsequent action with borders.

Extract 1. VERDIS AHP_12 27,05 VT968 EL12.1 "There'll be nothing left before long"

Jason elaborates on the physiotherapist's statement (line 13) with "There'll be nothing left before long" (line 15), notably uttered in overlap with a similar elaboration by the physiotherapist herself (line 14). However, Jason's version uses the future tense ("There'll be") and can thus be heard as referring to an inevitable outcome (and an imminent one, as "before long" suggests). In the context of shared knowledge about Jason's progressive condition, this can be heard as doing more than confirming the physiotherapist's statement about energy levels; it can be heard as alluding to future disease progression and perhaps EOL. By building the allusion as an elaboration of the physiotherapist's statement (line 13), Jason further shows that the physiotherapist's statement could be understood as already alluding to future deterioration. Julie produces minimal laughter (line 16), which may be responsive to Jason's own laughter at the end of line 15. Jason then makes the possible allusion explicit ("I'll be dead in- I'll be dead soon", line 17). He does so whilst turning towards Julie, which appears to prompt her response: the contradiction "No you won't" (line 18).

Jason treats Julie's contradiction as obvious and unnecessary ("I know", line 19), perhaps also conveying that his complaint in line 17 should not be taken literally. Consistent with this interpretation, he goes on to share a broader feeling of "giving up" (lines 19-21 and 24). Whilst the physiotherapist aligns as a recipient by gazing at Jason (line 20), Julie observably disengages by announcing taking her cardigan off (lines 22-23) and accounting for it (line 26), which may embody disinclination to engage with discourse about mortality.

Jason further elaborates in a way that retrospectively treats his own reference to "giving up" (line 24) as another allusion. He does so by elaborating it (Schegloff, 1996a) with "completely" (line 28) and "stopping everything" (line 29). Jason does not specify what exactly he would be stopping. However, in the context of having discussed activities of daily living (see lines 1-2), Jason can be heard as voicing an inclination to discontinue efforts to engage in such activities (with further implications for not engaging with the physiotherapist's recommendations). The extreme case formulations (Pomerantz, 1986) "completely" and "everything" can further be heard as implying resignation to inevitable deterioration and perhaps transition to the EOL.

Following a continuer by the physiotherapist and a substantive silence (lines 30-31), Jason formulates an upshot with the reported thought "Sod it. Why should I bother" (line 32). Its idiomatic nature can be heard as closing off the complaint whilst providing an opportunity for recipient affiliation (Drew & Holt, 1988). Indeed, the physiotherapist immediately responds by showing recognition of Jason's predicament (lines 33-34). This response links Jason's reported troubles to his experience of being "tired all the time", and thus embodies the physiotherapist's understanding that his complaint is about his physical deterioration. Jason confirms (line 35) and elaborates with an example (line 38). Julie intervenes at this point with another contradiction (line 39). This is done by recycling and opposing an aspect of Jason's complaint (Goodwin, 2018): his reference to not wanting to "bother" (back in line 32). She exhorts Jason to "bother", which can be heard as an invitation to engage with activities of daily living (and by implication, the physiotherapist's recommendations) for Julie's sake (line 39). Julie can thus be heard as pursuing the interactional project of motivating Jason to maintain engagement with activities of daily living. One of its outcomes is nevertheless to create an environment that is unfavourable to further discussion about EOL issues. Once again, Jason treats the point as obvious (line 40). Julie further claims that Jason is not going to die (albeit euphemistically, line 41). This embodies her understanding that Jason's complaints have been alluding to the inevitability of deterioration and death whilst contemporaneously disagreeing with such contemplations. The physiotherapist semi-seriously cooperates with Julie's exhortation (lines 45-46) and uses this to accomplish a transition back to a central activity within the consultation: assessing Jason's activities of daily living (lines 48-50).

Extract 1 has shown that a patient's explicit mention of EOL matters can be met with an equally explicit contradiction by a companion. Following Julie's first contradiction (line 18), Jason persists in sharing thoughts and feelings related to deterioration. However, Julie's second contradiction (lines 39, 41, and 44) results in an abandonment of the topic. Other aspects of Julie's comportment (her dealing with the cardigan from line 22) may be taken to embody disinclination to engage with discourse about mortality. At the same time, Julie's actions can be seen as advancing the interactional project of reassuring and motivating Jason to maintain engagement with activities of daily living. Julie's opposition to a focus on the EOL thus emerges as "supportive disagreement" (Weiste, 2015). One of its outcomes is nevertheless to create an environment that is unfavourable to the articulation of EOL issues.

This instance is arguably akin to what previous studies have described as a "refusal" to engage with matters related to the EOL (Planalp & Trost, 2008). Julie's exhortations might also be similar to what has been described as "performing hope" (Kellas et al., 2017) in response to a patient's reflections about the inevitability of deterioration and the resulting feeling of wanting to "give up". We will show that in our remaining cases, companions' actions cannot be straightforwardly seen as opposing a focus on EOL matters.

3.1 A companion introduces a related but distinct matter

Extract 1 has shown that a relatively explicit mention of disease progression and death can be followed by an equally explicit contradiction. But what happens in our eight remaining cases where the patients' actions only allude to disease progression and EOL? The companions' actions positioned after those allusions are not ostensibly concerned with opposing a focus on disease progression and EOL. Rather, they pursue interactional projects of their own, which are coherent with a sequence of action underway and tied to a wider activity in progress. However, one of their outcomes is to facilitate a move away from the possibility of elaborating the allusion and of making matters pertaining to disease progression or EOL explicit (despite these being relevant possibilities). The companions' actions shape the subsequent interaction, which observably moves away from the possibility of immediately elaborating on the allusion. Elaborating on the allusion and engaging in more explicit and focused talk about disease progression or EOL subsequently requires some additional interactional work to retrieve the allusion itself.

The companions' actions achieve these outcomes in two ways: by introducing a related but distinct matter after a patient's possible allusion (this section), and by introducing positive considerations (next section).

In Extract 2, Avery (P-Ave), a patient diagnosed with chronic obstructive pulmonary disease, and her husband Ray (also her main caregiver; C-Ray) are attending an outpatient appointment at the hospice with a doctor, Michelle (Doc), whom they have met previously. Before this extract, Ray semi-seriously complained about Avery sometimes being difficult to be with ("a right pain in the bum"). Avery confirmed, saying that she sometimes gets "agitated" and reacts negatively ("I get up in the air [...] then he gets it"). The doctor asked what this might be about; Avery responded that she does not know, and Ray suggested that this is probably because Avery cannot do what she used to do. Subsequent discussion featured examples of those situations, such as Avery worrying that the neighbours might see Ray when he puts the washing out in the garden (which Avery used to do; data not shown). This leads to the doctor's question at lines 1-4, which invites Avery to elaborate on the experience of feeling agitated (in terms of what is "most difficult about that"). Our focus is on how Avery's answer can be heard as alluding to disease progression (lines 6-7 and 10), and on Ray's subsequent turn (lines 14-16).

Extract 2. VERDIS35 32,30 VT1115 EL35.2 MP

Avery's answer establishes witnessing Ray's engagement with household work as difficult on the basis that she "used to do that" (lines 6-7). In the context of shared knowledge about Avery's advanced condition, this can be heard as alluding to a sense of loss associated with diminished capacity for activities of daily living caused by progressive health deterioration. This is reinforced by the finality of the expression "That's gone" (line 10), produced in overlap with the doctor's continuer.

Part of the remit of consultations at the hospice is to promote discussions about disease progression, and Avery's allusion can be heard as presenting an opportunity to do so. The doctor may orient to this possibility as she lets a silence emerge (line 11) and foregoes taking a full turn with the continuer "Right" (line 12), embodying an expectation that Avery might say more. Avery appears to treat this as an opportunity to say more as she draws an inbreath, apparently preparing to speak (line 13). She nevertheless relinquishes the floor and turns to Ray as he takes a turn (line 14). Ray intervenes unsolicited (neither Avery nor the doctor gaze at him before he starts to speak). Ray's turn does not take up, develop, or respond to Avery's answer. Rather, Ray produces his own answer to the doctor's question by proposing that Avery also worries about his wellbeing (he refers to suffering backpain when "hoovering", a British term for cleaning the floor with a vacuum cleaner).

In earlier talk before the extract, references to Avery being "agitated" had been used as a euphemism for her having angry reactions at Ray. It is thus possible that Ray's action (lines 14-16) implements the interactional project of offsetting this focus (which the doctor's question may be heard as insisting on) with a more altruistic characterisation of Avery (see the turn-initial contrastive "but"). Additionally, Ray's action may be face-saving because it steers the focus away from a potentially embarrassing aspect of Avery's experience: that she feels shame at the idea that their neighbours might judge her for not discharging household duties that she treats as pertaining to her (something that had also been discussed before this extract; data not shown).⁴ Ray's move is thus affiliative. Furthermore, it can be seen as fitted to the finality of Avery's "That's gone" (line 10), which can be heard as projecting 'nothing more to say about this' (Drew & Holt, 1998). Nevertheless, one of its outcomes is that it steers the talk away from the possibility of further elaborating on the matter that Avery has possibly alluded to. It does so by shifting from a focus on Avery's self-experience to her concerns about another person (Ray himself). We will now examine evidence that Ray's action shapes the immediately subsequent interaction, moving the focus away from the allusion. We will show that the doctor's subsequent conduct embodies an orientation to Avery's allusion as an opportunity to discuss matters related to disease progression, and that doing so requires the doctor to do some additional interactional work due to Ray's actions in between.

Avery agrees with Ray through an upgraded assessment of the problem with his back (line 17). After the doctor's continuer done whilst gazing at Avery (line 19), Ray produces an increment to his own answer (the idiomatic "If you know what I'm trying to say", with appended laughter, uttered whilst gazing at the doctor, lines 20-21), possibly designed to retain the floor and solicit a more affiliative response from the doctor.⁵ Avery nevertheless expands in overlap with Ray (line 22) and then again in overlap with a doctor's continuer (line 24) by shifting the focus towards her own preferences (that Ray does not do housework). Importantly, at this point, the talk has moved away from Avery's possible allusion to disease progression (lines 6-7 and 10).

The doctor responds to Avery with a formulation, which focuses on two aspects of her experience. The doctor first formulates the most proximal aspect of Avery worrying about Ray, which she renders as a concern about Ray doing "extra work" (line 28), which Avery confirms (line 29). The doctor then goes on to retrieve Avery's earlier answer, which featured the possible allusion. She formulates Avery's experience of witnessing Ray engage with housework as "hard" on the basis of it being "stuff you used to do" (line 32), which ostensibly links back to Avery's answer in lines 6-7. The doctor further proposes that for Avery this is also a reminder of her deteriorating health (lines 34-35). In this way, the doctor proposes that Avery's answer had conveyed more than it had stated "in so many words" (Schegloff, 1996a, p. 181), specifically, by alluding to matters related to disease progression, which Avery confirms (lines 36 and 38).

Extract 2 has shown that a companion can intervene after a patient's possible allusion to disease progression by developing a related but different matter. This becomes consequential for subsequent talk, moving it away from the possibility of immediately elaborating on the patient's allusion (despite this being a relevant possibility). Unlike extract 1, there is no evidence in extract 2 to suggest that the companion's intervention is designed to oppose a focus on disease progression. Ray's turn is ostensibly occupied with answering the doctor's initial question, and it can further be seen as implementing other interactional projects (including providing a positive characterisation of Avery). A lack of engagement with a possible allusion to disease progression is thus best seen as one of the outcomes of Ray's action and the interactional projects it implements.

3.2 A companion introduces positive considerations

We now examine the second way in which a companion's action can move the interaction away from the possibility of elaborating a possible allusion: responding to the patient's turn carrying the allusion by introducing positive or reassuring considerations. Jefferson (1988) documented that "optimistic projections" are recurrently used in ordinary conversation to transition out of troubles-talk. In an examination of interactions involving palliative care specialists and families of imminently dying patients, Anderson et al. (2021) found that the participants moved out of prognostic talk by transitioning towards more controllable matters, such as comfort care. In our cases, a companion responds to a patient's allusion to disease progression or EOL with a shift to reassuring matters, such as what can be done in the future to ameliorate the patient's quality of life. This move treats the patient's action as having alluded to disease progression or EOL (Ekberg et al., 2019) but does not promote its further elaboration (when this is a relevant possibility). Rather, it promotes a shift to other matters.

Extract 3 is from an interaction between Jan (P-Jan), a patient with lung cancer and brain metastases admitted to the hospice in-patient ward, her grand-daughter Melanie (C-Mel), and an occupational therapist (OT). The purpose of the interaction is to assess the kinds of home adaptations and support Jan will need when discharged back home. The participants have been discussing the type of bed Jan has at home and alterations it might need to enable her to get in and out of it (data not shown). Building on that, the OT announces that part of the agenda for this interaction is to assess Jan's mobility (lines 1-2). Jan takes a turn in recognitional overlap (Jefferson, 1984) to announce having a worry. She frames it as something that has been on her mind and has now been touched off by the OT's agenda setting (lines 3 and 5), and this suggests that it might be related to coping with her reduced mobility at home. Jan's announcement projects more talk designed to detail the nature of her worry. She starts a storytelling from line 7. Our analysis focuses on how this alludes to EOL matters and on Melanie's response (lines 42-44 and 48).

Extract 3. VERDISAHP_19 14,46 VT443 EL19.2 MP "this is part of what's worrying me"

Jan's storytelling (lines 7-30) focuses on the time when she was caring for her late husband at home as he was nearing the EOL. Jan describes lack of professional support from the palliative care team (lines 9-24), eventually leading to her not being able to care for her husband at home despite having promised him that he could die there (lines 25-30). Specifically, she was not able to lift him anymore (lines 29-30), which resonates with the possibility that her earlier announcement (line 5) alluded to a concern about her own reduced mobility. Therefore, her storytelling can be heard as alluding to concerns about the time when she will be discharged home and might find herself in a similar situation as her husband as she nears the EOL.

Jan goes on to share her guilt for not fulfilling her husband's wishes, with signs of hearable and visible upset (lines 33-34). This is met with empathic and supportive responses by Melanie and the OT (lines 35-40). As Jan accepts this (line 41), Melanie goes on to respond to Jan's storytelling (lines 42-44 and 48). This response is designed to reassure that Jan will not be in the same situation as her husband because the hospice care team will help get "things in place" for her (line 44). Melanie proposes that this will help Jan "be wherever you want to be for as long as you can" (lines 44 and 48). In the context of shared knowledge about Jan's progressive and terminal condition, this can be heard as referring to Jan's wishes for where to die. Melanie's response is thus not ostensibly concerned with rejecting a focus on the EOL. Rather, it displays recognition that Jan's storytelling alluded to concerns about being able to remain (and possibly die) at home because of her reduced mobility. It is affiliative insofar as it is designed to reassure. At the same time, Melanie does not open up those issues for further discussion and rather provides for a move to other matters. Indeed, Melanie's response can be seen as implementing the project of facilitating a transition towards the practical task of assessing Jan's mobility, which the OT announced in lines 1-2, by referring to it again in lines 42-44. The OT collaborates with this move by starting to detail the types of support to which Melanie has referred (from line 50). A move away from a focus on EOL matters is thus best seen as one of the outcomes of Melanie's response and the projects it implements, rather than upfront refusal to engage with such matters.

4. Discussion and Conclusions

We examined companion actions that follow patients' possible allusions to disease progression and EOL in a corpus of UK palliative care consultations. Our findings support the consideration that the presence of companions can add complexities to healthcare interactions. This is because companions' actions can advance interactional projects that diverge from the patients' or the HCPs' in specific moments in the interaction. In the cases examined in this article, companions' actions promote conversational environments that are not favourable to immediately elaborating on patients' allusions and making matters related to disease progression or EOL explicit.⁶ However, our analyses suggest that it is difficult to see these actions as designedly "refusing" or "avoiding" an engagement with disease progression and EOL as suggested by some of the extant literature (see introduction). Our findings suggest that, in several cases, disengagement with explicit and focused discussion about disease progression and EOL is the emergent outcome of companion actions that observably pursue other projects in the interaction, which are affiliative.⁷

In only one case in our collection (extract 1), a patient's explicit EOL considerations are met with an equally explicit contradiction by a companion. By contrast, in the other cases in our collection, where patients' turns allude to (rather than explicitly focus on) disease progression or EOL, companions' subsequent actions are not observably designed to forestall a focus on those matters. Rather, they implement actions that are fitted to a sequence of action in progress (e.g., answering the HCP's question in extract 2, and responding to the patient's storytelling in extract 3). They also advance supportive interactional projects that are tied to a wider activity underway. In extract 2, the companion's action offers a more benign characterisation of the patient, offsetting previous complaints about her. In extract 3, the companion's action promotes the patient's engagement with the occupational assessment that the OT had announced prior to the patient's storytelling. Moving the interaction away from the possibility of focusing on disease progression or EOL is best seen as one of the emergent outcomes of these actions (perhaps even a "collateral effect"; Enfield & Sidnell, 2017). Elaborating on the patient's allusion is nevertheless a relevant possibility in these environments as evidenced in comparable cases within our wider collection.

Our findings have theoretical implications because they show that engaging or disengaging with explicit and focused discussion about disease progression and EOL can be the contingent and emergent outcome of participants' actions designed to pursue various projects in the interaction. Whether disease progression and EOL become a sustained focus in the interaction depends on all participants' actions, including the HCPs'. As an illustration, in extract 2, the HCP pursues the possibility of more explicit and sustained discussion about disease progression and EOL by retrieving the patient's allusion after the companion's actions have steered the talk away from it. By contrast, in extract 3, the OT pursues a move back to the agenda of the occupational assessment, rather than promoting further talk about the EOL considerations that the patient has alluded to.

Our analyses further augment existing cultural and social-structural explanations for the reported absence of discussions about EOL matters (see introduction) through detailed examination of actual social interactions where EOL matters are raised. Wider analysis of our data set of palliative care consultations (not reported in this article) shows that, in the majority of cases where explicit and sustained discussions about disease progression and EOL happen, their starting point is not an explicit reference to EOL matters but, rather, a possible allusion (like those exemplified in this article). This suggests that participants handle disease progression and EOL as delicate matters that should not be raised explicitly or unilaterally, but rather developed interactionally in a stepwise manner (see Jefferson, 1980). One consequence is that possible trajectories towards EOL talk are intrinsically vulnerable because allusions can be dis-attended (extract 2) or attended to only tangentially (extract 3); in both cases, the matters that the patient has possibly alluded to do not immediately become a sustained focus of subsequent interaction (extract 2 further exemplified how retrieving and focusing on those matters subsequently requires additional interactional work by the HCP). These findings thus start to provide social-interactional grounding for considerations about the absence of discussions about EOL matters in Western contexts: their interactional properties as a delicate activities make them vulnerable to what Jefferson and Lee (1992) called interactional "asynchrony", in which the possibility of exploring troubling matters possibly raised by a patient's allusion is abandoned (at least provisionally) in the service of pursuing alternative interactional projects.

There are some related practical implications. Our findings invite caution when interpreting companions' actions that follow patients' possible allusions to disease progression or EOL. We proposed that these actions cannot be straightforwardly reduced to forms of "refusal" or "avoidance"; rather, they can advance supportive interactional projects. This nevertheless creates particular challenges for HCPs who wish to promote discussions about disease progression and EOL. Although examining HCP responses was beyond the scope of this article, extract 2 is instructive. It shows how an HCP can display recognition of the concerns introduced by a companion's action whilst also working to retrieve the themes raised by a patient's possible allusion. Elsewhere, we examined the careful interactional work that HCPs mobilise to pursue a recommendation in environments where a patient's and a companion's actions embody diverging alignments towards that recommendation (Pino, Doehring, & Parry, 2021). The analyses reported in this article further suggest that promoting discussions about disease progression and EOL also entails careful management of the intersecting interactional projects that are advanced by patients' and companions' actions.

Acknowledgments

We would like to thank the staff, patients, and companions who consented to being recorded. We are grateful to Ruth Parry for her overarching work in establishing and leading the VERDIS project and building and supporting the team that made this article possible. We thank two anonymous reviewers for their detailed and helpful comments on an earlier version of this article. Finally, we thank Deborah Chinn, Niklas Norén, and Joe Webb for bringing together the special issue in which this article is published and for their support throughout the editorial process.

Funding

Creation of the data set for this study and earlier analyses were funded by The Health Foundation [Grant IDs: AIMS 1273316, AIMS 1273743/GIFTS 7210], an independent charity committed to bringing about better health and social care for people in the UK, and by the National Institute for Health Research Academy Career Development Fellowship to Ruth Parry: 'Enhancing staff-patient communication in palliative and end of life care' (Grant ID CDF-2014-07-046). The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care.

Transcription Conventions

Transcription of verbal and vocal conduct is based on Jefferson's (2004) conventions.

Transcription of visible and embodied conduct is based on Mondada (2018) as follows:

^{1 In earlier work, we have referred to these as possible EOL considerations as well as 'cues' (Pino & Land, 2022; Pino et al., 2016).↩}

^{2 We focus on patient allusions in this article. We analysed companion allusions in Pino and Land (2022).↩}

^{3 We thank an anonymous reviewer for suggesting this possibility.↩}

^{4 Another possibility raised in a data session is that Ray may be working to attract attention to the difficulties he is experiencing. There is some internal evidence to support this, as previously in the consultation Ray mentioned some of the struggles involved in working as Avery's family caregiver, and he complained that the doctor had not asked about this in the consultation. This analysis is not incompatible with our main point here: that Ray's action steers the focus away from Avery's experience, and thus also from the possibility of elaborating her possible allusion.↩}

^{5 Another possibility, brought to our attention by an anonymous reviewer, is that this is designed to retain the doctor's gaze (which has just switched to Avery) and works as an appeal to the doctor not to hear Ray's previous turn as an attempt to turn the focus away from Avery's problems to his own problems. Either way, "If you know what I'm trying to say" appears to treat his own prior talk as conveying more than he has said in so many words.↩}

^{6 We find it important that our findings are not taken to suggest that the presence of companions simply makes discussion of EOL more difficult. Elsewhere, we examined companion actions that facilitate the emergence of these matters (Pino & Land, 2022). This happens when it is a companion who alludes to concerns about disease progression or EOL.↩}

^{7 Although it is possible that some of the companions in our data intentionally pursued these emergent outcomes, but did so covertly, this is ultimately unknowable to us (and arguably, also to the other participants in the interaction). ↩}