Markers of assimilation of problematic experiences in dementia within the LivDem project

This study aimed to determine whether the Markers of Assimilation of Problematic Experiences in Dementia scale (MAPED) can be used to identify whether the way in which participants talk about dementia changed during the group. All eight sessions of a LivDem group, which were attended by participants were recorded and transcribed. An initial analysis identified 160 extracts, which were then rated using the MAPED system. Inter-rater reliability was 61% and following a resolution meeting, 35 extracts were discarded, leaving 125 extracts with an agreed marker code. All of the participants were identified as producing a speech marker relating to dementia, and these varied between 0 (warding off) to 6 (problem solution). Examples of these markers are provided. The proportion of emergence markers (indicating the initial stages of assimilation) compared to later markers changed significantly between the first four sessions and the final sessions. This difference was still significant even when the markers produced by the most verbal participant, Graham, were excluded. The use of process measures within psychotherapy complements more conventional outcome measures and has both theoretical and clinical implications.

Emergence of thoughts associated with dementia or memory loss. Further discussion may be avoided and dementia is likely to be talked around rather than named directly. Unfocused strong emotions (e.g., anxiety, fear, anger, sadness) are more salient than the content a. Changing the conversation b. Telling a contradictory story c. Fear-of-loss-of-control (e.g., being mad or abnormal) d. dementia is located elsewhere or referred to indirectly (e.g., as ''it'') e. the significance of dementia is otherwise minimized 2. Vague awareness The problematic experience of dementia is acknowledged and uncomfortable associated thoughts are described. The person is in distress, and this seems to come from some internal conflict relating to dementia, so that in talking they seem to be caught up in the moment of the emotion. Affect focused on acute psychological pain or panic Expressions of a. Feelings of sadness, depression, worry, or anxiety b. Anger or irritation about some aspect of dementia c. Puzzlement or confusion about what is happening to them d. Feeling overwhelmed or that things are getting worse e. Stories that point to the problem but are not clearly described f. Use of metaphors to talk about the problem 3. Problem statement or clarification The person acknowledges the existence of dementia as a problem and attributes this to an illness such as dementia. Affect is negative but manageable a. Describing their reaction to dementia, which may include mixed feelings (e.g., ''yes. . .. But'') b. The person appears ''stuck'' and sees no way out c. Developing a clearer understanding of dementia (e.g., asking questions or being curious) d. Describing contradictions in the illness e. Describing that previous difficulties in recognizing the problem 4. Understanding/ insight The person acknowledges the existence of dementia and is able to describe how this makes them feel, or how they react to this. They are able to stand back from their feelings. Curiosity of affect, with mixed pleasant and unpleasant recognitions a. Describing situations where their feelings differ in intensity b. Emotional distance from the dementia (e.g., through use of humor) c. Making links between dementia and others areas of their life d. Making links between the past and the present (continued) 2002). Each of the eight levels is characterized by a number of specific speech markers (Honos-Webb, Surko, Stiles, & Greenberg, 1998, Honos-Webb, Surko, et al., 1999, which are indicative of the form of dialogue that occurs at the different levels between the Dominant and the Problematic Voice. This process has been formalized through the development of the Markers of Assimilation of Problematic Voices Scale or MAPVS (Honos-Webb, Stiles, & Greenberg, 2003). The MAPVS is primarily a research tool, but it can also be used to help to formulate clinical problems, and as such has strong clinical relevance (Stiles, 2001).

Assimilation and dementia
The threat of a dementia diagnosis can be understood in terms of the Assimilation Model of Problematic Voices. Within this framework, a diagnosis is likely to represent such a powerful threat to the person's psychological equilibrium that it may resist being easily assimilated into the self. This way of representing the threat of dementia, and thus the nature of the psychological change that is required for people to ''come to terms with'' dementia, has been used in a series of studies about changes in descriptions of dementia (e.g., Betts & Cheston, 2011;Cheston, 2013;Cheston, Jones, & Gilliard, 2004;Snow, Cheston, & Smart, 2014;Sims & McCrum, 2012;Watkins et al., 2006). The MAPVS was adapted so that it is appropriate for use with people with dementia (Lishman, Cheston, & Smithson, 2014). The revised scale (known as the Markers of Problematic Experiences of Dementia or MAPED scale 1 ) follows the MAPVS by coding accounts of dementia in terms of a series of markers, or speech events, which can be grouped into one of the eight possible levels (see Table 1). Ratings of these markers take into account both the use of language to frame dementia (and dementia-related experiences) and the emotional intensity that surrounds this discourse. Adapting the MAPED scale from the original MAPVS involved identifying passages from either psychotherapy groups or qualitative interviews with people affected by dementia in which descriptions of dementia were initially rated as indicative of a specific marker and then used as illustrative templates to guide future analysis. For instance, a passage would be coded as falling into level 1 (unwanted thoughts) if the account includes an aspect of dementia such as memory loss and the person's level of affect indicates feelings such as anxiety, anger, or sadness. In addition, there needs to be a strong sense that exploration of the wider implications of the dementia is being avoided because this is experienced as being frightening or destabilizing to the person's emotional equilibrium. Level 1 markers include the person: changing the conversation; locating the existence of dementia elsewhere by talking about other people and not themselves as having dementia; limiting the problems the person experiences to some areas without this being explicitly or implicitly associated with dementia; references to fears of being mad or abnormal; asserting that other areas of functioning are unimpaired or that they can be easily overcome; or other speech acts that minimize the significance of the dementia experience. One marker of a level 1 rating that was identified in the original MAPVS rating process is referred to as the ''fear-of-loss-of-control'' marker (Honos-Webb, Lani, & Stiles, 1999). This marker can be identified as occurring when a person describes a choice between continuing to explore the problematic experience (and risking destabilizing their internal, emotional equilibrium), and avoiding further discussion. The clinical importance of this marker is that it signifies the emergence of unwanted thoughts into the person's awareness, an early indication of the person beginning to assimilate emotionally charged or problematic experiences. The MAPED coding frame translates this into the context of dementia by identifying instances in which a person avoids discussing dementia, for fear of how this might make them feel. Thus, in the psychotherapy group subsequently reported by Watkins, Cheston, Jones, and Gilliard (2006), Mr E states ''I can't think about coming here, it just brings it all back.'' One indication of a possible ''fear-of-loss-of-control'' marker can be found in what has been described by Cheston (2015) as the Voldemort phenomenon. This refers to the way in which some people affected by dementia may refer to aspects of their experiences indirectly, and without using the terms ''dementia'' or ''Alzheimer's disease.'' Just as in the Harry Potter series of novels, the evil Voldemort is for many characters too frightening to be named directly, so too some people affected by dementia find their illness too powerful to be referred to by name. To name dementia, then, is to acknowledge its existence and thus to risk a loss of emotional equilibrium. Thus, Sue's description of her dementia as ''like being given a ticking time bomb'' (Snow et al., 2014) and Henry's analogy of fighting an unnamed foe because he did not want to ''wave the white flag'' (Lishman et al., 2014) were both rated being markers of a ''fear-of-loss-of-internal-control''-and thus acquired a Level 1 rating. Finding alternative ways to talk about dementia, for instance by using euphemisms (such as memory loss) or figurative language is therefore part of a process of emotional regulation.

Living well with dementia therapy groups
The Living well with Dementia (LivDem) group therapy project was a pilot-randomized control trial, which compared the impact on people who have recently received a diagnosis of dementia of attending a therapy group compared to a waiting-list control group. Seven groups were established across the south of England, each of which lasted for 10 weeks and involved memory clinic staff delivering a manualized intervention (see Table 2 for details of the course content). Sixty participants were randomized to receive either the intervention or usual care in the form of a waiting-list control arm. The analysis of data arising from the study involved both quantitative and qualitative components. The collection of outcome measures at baseline, the end of the study, and follow-up were complemented by the assimilation analysis of changes in how participants talked about their problems. These two strands of analysis were included in the study protocol, with the recording, transcribing, and analysis of material from the sessions being specified in the consent process. For the primary outcome measure of quality of life and a secondary outcome, self-esteem, there were strong, but nonsignificant trends toward improvement in the intervention group compared to the control group (Marshall et al., 2014). This is consistent with previous research that has indicated that short-term group therapy or support intervention can be effective in lowering levels of depression (Cheston & Jones, 2009;Cheston, Jones, & Gilliard, 2003;Logsdon et al., 2010). However, as a pilot study, the research was not powered to produce significant findings, and these results thus need to be treated with caution.
The aim of the LivDem intervention is to support people who have recently received a diagnosis of dementia to plan ahead and to take control over their illness. For many people, this may present a significant emotional challenge in which there is a need, in effect, to balance the practical benefits that might accrue from greater understanding of dementia on the one hand, against the existential pain that might arise from this increase in awareness (Cheston, 2013). The aim of the qualitative part of the study was to establish whether it was possible to use the MAPED technique to identify whether there were changes in the type of markers that are found over the course of one of the LivDem groups.

Method
The assimilation analysis of recordings collected during the LivDem study falls into two parts: first of all, the analysis presented here of all eight of the sessions attended by one of the seven groups; and secondly, the analysis of a randomly selected sample of sessions from the remaining six groups, which will be reported separately. The group from which this analysis is presented was the first of the seven groups to take place, although otherwise it does not differ significantly from the other groups that were part of the project.
All names used here are pseudonyms. The group was initially made up of five participants, one of whom (Ellen) was only able to attend the preliminary session with her husband, as she subsequently slipped at home and broke her arm. All of the four other participants attended every session (see Table 3 for details). Although each group consisted of 10 sessions, the first and final sessions had a different format: they involved not just group participants, but a relative or friend also attended. As the dynamics within these two sessions were different, they have not been included in this analysis. All eight sessions from one of the seven groups that were part of the project and which involved only participants were recorded and transcribed (although only 15 minutes of session 1 were recorded due to technical difficulties).
The procedure for identifying markers of assimilation is adapted from psychotherapy process research Honos-Webb, Surko, et al., 1999;Knobloch, Endres, Stiles, & Silberschatz, 2001). This consists of four steps: . Transcribing and familiarization. All participants in the LivDem study gave their consent for each session to be audio-recorded. The initial step of the analysis involves intensive exposure to the transcripts, making systematic notes to locate passages that relate to change, cognitive loss, or expressions of heightened affect. . Extracting passages. Speech markers are identifiable events in discourse that recur throughout the transcripts and that indicate important clinical phenomena . In this analysis, LG identified 160 extracts in which one or more of the group participants either referred directly to an aspect of dementia or were provided with a clear opportunity to do so (e.g., through a question by a facilitator) but avoided doing so. . Rating passages.
LG was trained using extracts that had initially been rated in Lishman, Cheston, and Smithson (2014).
LG and RC then independently coded the extracts. RC was blind to the session in which the extracts occurred, although LG was not blind as she had extracted the data from the transcripts. There was an initial agreement on 98 of the 160 extracts, giving an inter-rater agreement level of 61%. . Clarifying disagreements. The 62 ratings where the raters disagreed were discussed in a resolution meeting. This enabled agreement to be reached for 27 extracts, with the remaining 35 extracts being discarded as they were deemed either not to have enough content to be rated or as being too difficult to be rated reliably. This left a combined total of 125 extracts with an assigned marker code.

Results
Across the eight sessions as a whole, markers from every level except Level 7 (Mastery) were identified. Over half of the extracts (66 of the 125) came from one participant: Graham. Previous research (e.g., Cheston, 2013) suggests that the markers can be subdivided into three groups: emergence, in which the therapeutic task is to enable dementia to emerge as the cause of problems without the person being overwhelmed (levels 0-2); naming dementia as the problem and finding distance from it (levels 3 and 4); and working through (levels 5-7). For an extract to be identified as a Level 0 marker, the person needs to actively avoid discussing dementia, or any subject related to dementia, or to reject the possibility that dementia plays a role in their life. Often this is by attributing any changes in life to the effects of age or to a physical or sensory impairment. Thus, in this group when Mary was asked to engage personally with issues around memory loss, which arose in the initial sessions, she attributed her difficulties in life to her physical health problems (she used a wheel chair). Her affect was minimal suggesting that she had successfully avoided the problem and her replies tended to close down further discussion. Level 1: unwanted thoughts. For a speech extract to be rated as occurring in this level, there needs to be evidence both that the person has thought about dementia and also that these thoughts are actively avoided. Thus in the fifth session, the group watched a recording of people talking about how the diagnosis of dementia had affected them personally. Mary who had been quiet, was asked about the impact of the dementia diagnosis on her: Thus, Mary acknowledged that she had been diagnosed (with vascular dementia), but then denied it had affected her (2.1) and minimized the impact on her ''if I was dying or something like that'' (2.2). Two features of this extract are typical of a Level 1 marker: first, the expression of the need to ignore the Problematic Voice, as Mary repeats the phrase ''carry on'' (2.3 and 2.4); and secondly, the Problematic Voice is referred to only indirectly as ''no use sitting there thinking about it is there'' (2.5). At this point in the therapeutic process, clients are faced with a psychological dilemma: to continue to explore the emerging, threatening areas of their life (and thus risk the escalation of distress), or to continue to push this difficult material away. As has been detailed above, the therapeutic significance of the fear-of-loss-of-control marker (a level one marker) within therapy has been explored both within general psychotherapy (Honos-Webb, Lani, & Stiles, 1999) and within therapy for people with dementia (Cheston, 2015). Emerging affect can involve stronger negative feelings such as anxiety, fear, anger, and sadness, but these are typically controlled and spoken of as being pushed away.
Level 2: vague awareness. In this level, the person is in distress that comes from an internal conflict relating to an aspect of their dementia. In articulating this distress, the person seems to be caught up in the moment of the emotion. They express material with a strong level of affect, and while the nature of this affect varies, nevertheless there is a clear intensity to the way in which they express themselves. Thus, in this series of extracts, the participants express their distress which we might characterize as, respectively, sadness, resentment, and anxiety: In all three extracts. Gina and Graham are caught up in the emotion of their distress. In this sense, they are talking from within their emotion, rather than being able to stand outside it and talking about their distress, both of which would be characteristic of markers at later levels of assimilation.

Naming dementia as the problem
Level 3: clarifying dementia as the problem. At this level, the person both acknowledges the existence of a problem and attributes this to or recognizes that this is caused by their dementia. Thus, in session 5, Gina and Graham comment on a recording of people talking about a diagnosis of dementia that they had just watched (this was part of a DVD used within sessions): [6] Facilitator 1: You were watching the television just now people were talking about how they felt when they first had their diagnosis. Gina: It does it makes you feel you know that you're going downhill very fast. Graham: That's right it makes you feel that it's far more serious than you imagined it to be. Session 5 At this level, people describe or talk about their reaction to the knowledge that they have dementia (e.g., feeling upset, silly, angry, or trying to cover up for mistakes) but without being caught up in the emotion of this as was the case with the Level 2 markers. Although at times the person may feel ''stuck'' and see no way out of their problem, they are beginning to be able to stand back from the problem and talk about it, rather than being overwhelmed by the distress, as is the case in Level 2 markers. For example, they may ask questions about the problem or express curiosity about the problem in order to aid their understanding. Thus in the final session (week 8), Graham asks about the issue of sundowning: [7] Graham: You, fortunately for me brought up an area which was snowboarding. I1: Oh, sun downing? Graham: Sun downing sorry sun downing I happened to say that that's something I get now what I then went on to say is is that something that's going to get worse? is it going to get better is it going to stay as it is? it's the broadening of more than just memory.
Level 4: understanding and gaining perspective. In this level, the person acknowledges the existence of a dementia or a central aspect of dementia such as a memory problem, but is also able to describe how this makes them feel, or how they react to this. However, the central feature of this level is the way in which the person is able to stand back from their feelings and from the diagnosis-they have achieved some emotional distance from the dementia even though they recognize that it affects their life at times profoundly. Consequently, the person's affect is mixed-for instance there may be a tone of relief in addition to distress, or the person may show that they have achieved some emotional distance from the dementia, rather than being overwhelmed by it, by making a joke or laughing. Thus in the final session, Graham both describes his concern at being a burden on his wife, but makes light of this: [8] Facilitator 1: . . . is that a concern as well the fact that maybe other people will see changes happening and you won't see those changes? Graham: my concern if you want to know the real truth is at what stage am I going to be a real burden on my wife. F1: Right yeah. Graham: That's my real concern. F1: Yes have you shared that at all with your wife have you talked to her? Graham: She wouldn't accept it if I did. F1: Really. Graham: She would tell you I am already a burden (laughs) Gina: In a joking way probably? Graham: (laughs) No seriously. Session 8

Working through
Level 5: accepting dementia and developing strategies. Here, the person acknowledges the existence of a dementia or a central aspect of dementia such as a memory problem and recognizes that the dementia cannot be cured, but can be managed. They can point out exactly what needs to change or to be worked on and show that they are weighing up attempts at specific or partial solutions to the problem (e.g., considering taking medication, supporting the Alzheimer's society or challenging stigma). The affect (mood) is positive, business-like, and optimistic. Thus Graham comments in the penultimate session: [9] Graham: I was told I had dementia and the best thing I could do was tell everybody I've done exactly that I've told everybody that If I could tell you the truth I don't think they've taken one blind bit of notice. Session 7 Level 6: problem solution. At this level, the person is able to acknowledge that they have dementia and describe having achieved a change in their life in living with dementia that they feel positive about. This change may be in their understanding of what is happening to them or they may talk about a change in their relationships with others, or comment on how others have noticed that they have changed. Thus in session 6, Graham's affect is positive, satisfied, and proud of his accomplishment: [10] Graham: Well I try and be jolly all the time and I keep thinking to myself I should have made a change by now because when I was told I got dementia it was as if you know the end of the world had come where as far as I'm concerned I'm not on so many tablets I'm probably jollier not necessarily happy but jollier that I've been for years Does that make sense? Session 6

Shifting use of markers across sessions
This analysis shows that the levels of markers used in sessions altered as the group progressed. As Figure 1 indicates, there was a shift toward an increase in the proportion of markers from more assimilated levels in the later sessions. Alongside this was a decrease in markers signifying emergence of awareness (i.e., levels 0-2) as there are no level 0 markers after week 2, no level 1 markers after week 5, and only one level 2 marker in each of sessions 6, 7, and 8. Similarly, there was an increase in higher level markers over successive sessions: from week 4 onward, there was a significant increase in the number of markers from levels 3-6. A 2 Â 2 contingency, table was constructed to examine whether there were differences between the proportion of markers indicating an emerging awareness (levels 0-2) and more assimilated markers (levels 3-6) between the first four and the final four sessions. Fisher's exact test gave a highly significant two-tailed value of less than 0.0001. Over half of the extracts (66 of the 125) came from one participant: Graham. However, although Gina, Mark, and Mary provided fewer references to dementia that could reliably be identified as markers, nevertheless Figure 2 suggests that when they did contribute, then their use of markers too altered. Even when Graham's markers were excluded from the contingency table, Fisher's exact test gives a P of 0.0016.

Discussion
This study set out to test whether the markers that were used within a therapy group changed across the eight participant-only sessions of a group. We have argued that the Markers of Assimilation of Problematic Experiences of Dementia is a useful way of tracking change, and also that in the second half of the group, participants talked about their dementia in a different way to that in the first four sessions. Change is still present even when Graham, the most verbal and assimilated participant is removed from the analysis.
There are two possible reasons for this apparent change in levels of markers found in the group over the initial and latter stages of the group: either it may represent a therapeutic shift in which participants gradually assimilated the problematic elements of their diagnosis; or alternatively, it may reflect the way in which the topics that the facilitators raised altered across these sessions. In many ways, this change in the type of markers used probably reflects both of these factors-thus although the facilitators talked more directly about dementia in later sessions, participants were able to assimilate this understanding in a way that may well have not have been possible at an earlier point. Indeed, it is this combination of participant and therapeutic factors that typifies the psychotherapeutic process of change.
In addition, talking about dementia, was rarely characterized as existing at a single level of assimilation. Instead, individuals moved between different levels within the same session, and sometimes within the same extract. This is a process that has been described elsewhere, for instance within qualitative analysis of interviews with people affected by dementia. Typically, these studies represent awareness as a fluid process in which both denial and acknowledgement of dementia may be present. Thus, Robinson, Clare, and Evans (2005) described couples' shared constructions of the diagnosis of dementia as a process that cycled through ''denial, minimisation and gradual realisation as couples gradually began to accept the changes in the person with dementia were likely to be permanent, linked to an oscillating process of acknowledging what had been lost, as well as carrying on as a couple by focusing on what remained for each person and the couple'' (Robinson, Clare, & Evans, 2005, p. 344).
Importantly, both the growth in awareness for some participants and the oscillating ambivalence described by both Robinson, Clare, and Evans and by Lishman et al. (2014) occurred outside a formal psychotherapeutic process. This suggests that for some people affected by dementia, it is possible to gain increased awareness of their illness without attending support groups such as LivDem. However, the personal and social characteristics that lie behind such shifts in awareness are not clear.

Implications for clinical development
Drawing on the assimilation model, Cheston (2013) has suggested that there are three steps within this process of adjusting to the illness. The first task is for people to be able to name their problems as being dementia without being emotionally overwhelmed (levels 0-2). Then, once people are able to talk about the dementia, they have to find some emotional distance from the illness (levels 3 and 4). Finally, they are able to see dementia as only one part of their lives, and make appropriate plans (levels 5-7).
Within this model, the transition between levels 2 and 3 (i.e., from warding off awareness of the illness, to identifying dementia as the problem) is of critical importance. This process involves the person managing to name their problems as being caused by dementia without, at the same time, being emotionally overwhelmed. Intrinsic to this change in talking about dementia is an emotional process involving the person articulating the Problematic Voice. This may be around their fears for the future, around loss (such as giving up driving), or around embarrassment and frustration. From a clinical perspective, what seems to facilitate people in approaching their dementia includes other people in the group articulating their fears and concerns (Watkins et al., 2006) and using indirect ways of exploring threat (Cheston et al., 2004). Conversely, barriers to change may be setting too fast a pace for therapy, providing information at too soon a point in the therapeutic process or involving people from outside the group (Cheston & Jones, 2009).

Implications for research
The Assimilation Model has a number of differences compared to other approaches which examine the awareness that people affected by dementia have into their illness. One clear difference is the way it conceptualizes the relationship between affect and awareness as a complex and bilinear process. Thus, the Assimilation model suggests that initially increased levels of awareness from warding off (Level 0) through to Vague Awareness (level 3) are associated with increasing levels of affect. This distress can be dissipated through the use of affect regulation behaviors such as repression and the projecting out of distress, but it may also be associated with behavioral strategies such as avoidance. For some people, increased awareness of the Problematic Voice of dementia may be associated with verbal markers of a fear-of-loss-of-control in which the person expressed concerns that continuing to engage with the problematic material may lead to a loss of psychological equilibrium or well-being (Cheston, 2015). However, if the person affected by dementia is able to continue to articulate their Problematic Voice and this becomes assimilated into the community of selves, then levels of affect decline. Thus, if a person is able to identify their problems as being associated without being emotionally overwhelmed and are both to hold onto their dementia while also identifying partial solutions to some of the problems that it raises, then levels of affect diminish.

Final thoughts
The importance of postdiagnostic support for people affected by dementia is increasingly recognized (Watts et al., 2014). The LivDem groups provide a short-term intervention that can be facilitated by nurses and other memory clinic workers after a two-day training course and receiving supervision from clinical psychologists. Although evidence from the Pilot study was encouraging, further work is needed to clarify its therapeutic impact. This includes identifying whether participants are more able to acknowledge their dementia as a result of attending the groups, and to identify therapist behaviors that facilitate or impede this process. The MAPED method provides an innovative and potentially important way of tracking therapeutic change or growth in awareness, and this is an important tool in both research and clinical practice.

Statement of ethical approval
The protocol was registered online (http://www.controlled-trials.com/ISRCTN25079950/), and received approval from NRES Ethics Committee South Central-Oxford B (REC Number 11/SC/0363, approval dated 18 November 2011, accepted protocol amendments on 28 June 2012 and 23 August 2012).