Background
Estimates show that 47 million people worldwide were living with dementia in 2016 and the number is steadily increasing (Prince et al. Reference Prince, Comas-Herrera, Knapp, Guerchet and Karagiannidou2016). Due to government policy drives to increase dementia diagnosis rates, the number of people receiving a diagnosis of dementia has risen significantly in the United Kingdom (UK) (Hodge and Hailey Reference Hodge and Hailey2015) and is increasing worldwide (Alzheimer Europe 2016). Receiving a diagnosis of dementia is a life-changing event, which can cause strong emotional reactions (Aminzadeh 2007; Brooker Reference Brooker2003) which are partly influenced by the stigma attached to this diagnosis (Speechly, Bridges-Webb and Passmore Reference Speechly, Bridges-Webb and Passmore2008). The process from a person or a family member realising that they might have a problem to the dementia diagnosis itself can take up to three years (Chrisp et al. Reference Chrisp, Thomas, Goddard and Owens2011). Thus, when people finally hear the diagnosis, while there may be some level of shock and denial, there are likely to be many concerns and information needs surrounding the condition and its management (Bunn et al. Reference Bunn, Sworn, Brayne, Iliffe, Robinson and Goodman2015).
The way health-care professionals (HCPs) elicit and respond to concerns expressed by patients is an important part of medical communication. Concerns are defined as patient expressions that convey worries, fears or unpleasant emotions (Zimmermann, Del Piccolo and Finset Reference Zimmermann, Del Piccolo and Finset2007). Although eliciting patient concerns might prolong the diagnostic appointment (McLean and Armstrong Reference McLean and Armstrong2004), encouraging their expression can lead to improved outcomes, e.g. by improving engagement of patients in health services who might otherwise not engage (McCabe et al. Reference McCabe, Skelton, Heath, Burns and Priebe2002; Zimmermann, Del Piccolo and Finset Reference Zimmermann, Del Piccolo and Finset2007). This is particularly important in dementia, where involvement in support and activities can enhance quality of life and reduce behavioural symptoms (Trahan et al. Reference Trahan, Kuo, Carlson and Gitlin2014). Additionally, research on cancer consultations has shown that failure to address patient concerns can have a negative impact on patients’ subsequent psychological wellbeing (Brown et al. Reference Brown, Butow, Dunn and Tattersall2001). However, a review of the literature shows that HCPs do not always detect patient concerns and may sometimes unwittingly discourage their disclosure (Zimmermann, Del Piccolo and Finset Reference Zimmermann, Del Piccolo and Finset2007). These studies use various frameworks to explore doctor–patient communication, such as the Verona Coding Scheme for Emotional Sequences (Del Piccolo, Goss and Zimmermann Reference Del Piccolo, Goss and Zimmermann2005; Zimmermann et al. Reference Zimmermann, Del Piccolo, Bensing, Bergvik, De Haes, Eide, Fletcher, Goss, Heaven, Humphris and Kim2011), and focus predominantly on dyadic interactions between HCPs and patients.
In dementia there are specific challenges to HCP–patient communication. Memory clinic consultations usually involve the person with dementia, their companion (usually spouse/family member) and the HCP, and this can create additional communication difficulties when the needs of both the patient and companion have to be considered (Karnieli-Miller et al. Reference Karnieli-Miller, Werner, Aharon-Peretz, Sinoff and Eidelman2012; Robinson et al. Reference Robinson, Bamford, Briel, Spencer and Whitty2010). These triadic interactions can be difficult for the HCP as they try to attend both to patients and carers without marginalising one or the other. Research in memory clinics (e.g. Karnieli-Miller et al. Reference Karnieli-Miller, Werner, Aharon-Peretz, Sinoff and Eidelman2012) has shown that the patient can be marginalised. A key factor influencing these triadic interactions is the impact of dementia symptoms on communication. Dementia typically involves language decline and comprehension problems (Blair et al. Reference Blair, Marczinski, Davis-Faroque and Kertesz2007), which can have practical implications for how HCPs communicate, e.g. using long sentences with multiple clauses can be more difficult for people with memory difficulties to follow than people who have no cognitive impairment (Jones et al. Reference Jones, Drew, Elsey, Blackburn, Wakefield, Harkness and Reuber2016). These factors can lead to the marginalisation of the person with dementia in discussions about their health and treatment (Karnieli-Miller et al. Reference Karnieli-Miller, Werner, Aharon-Peretz, Sinoff and Eidelman2012; Sabat Reference Sabat2005).
Studies examining the needs of people with dementia and their families highlight issues that arise in the post-diagnostic stage regarding lack of ongoing support from services (e.g. Black et al. Reference Black, Johnston, Rabins, Morrison, Lyketsos and Samus2013; McCabe, You and Tatangelo Reference McCabe, You and Tatangelo2016; van der Roest et al. Reference van der Roest, Meiland, Comijs, Derksen, Jansen, van Hout and Dröes2009). Additionally, studies of how people experience communication of dementia diagnoses demonstrate that they do not feel they receive adequate diagnostic information (Abley et al. Reference Abley, Manthorpe, Bond, Keady, Samsi, Campbell and Robinson2013; Bamford, Lamont and Eccles Reference Bamford, Lamont and Eccles2004). People with dementia and their companions report feeling that they did not know what to ask the HCPs in the diagnostic feedback meeting and that they felt that the concerns they raised were sometimes disregarded (Manthorpe et al. Reference Manthorpe, Samsi, Campbell, Abley, Keady, Bond and Iliffe2013). However, there have been no studies directly observing the communication in dementia diagnosis meetings that have focused on the nature of the concerns that people with dementia and companions are raising in these meetings, and how HCPs are addressing these concerns.
The context within which a diagnosis is delivered is also important as different pathways and practices can have an impact on the experience of people receiving a diagnosis of dementia. In the UK, the diagnosis of dementia typically takes place in specialist memory services. This involves the primary care practitioner screening people who present with dementia symptoms and referring them to specialist memory services, often referred to as memory clinics, where further testing is conducted and the diagnosis is disclosed (Alzheimer's Society 2014). However, people can be referred to memory clinics or neurology and have variable access to pre-diagnostic counselling, psychology and post-diagnostic psycho-social interventions (Guss et al. Reference Guss, Middleton, Beanland, Slade, Moniz-Cook, Watts and Bone2014). There is also some variation in the way memory clinics operate, e.g. in terms of waiting times and information provision (Abley et al. Reference Abley, Manthorpe, Bond, Keady, Samsi, Campbell and Robinson2013).
This study explored concerns expressed in HCP–patient–companion communication in specialist memory clinics during the dementia diagnostic feedback meeting. The aims were to identify (a) the topics of the concerns expressed by patients and companions, (b) whether HCPs elicit patient and companion concerns or whether they are volunteered, and (c) how HCPs respond to these concerns.
Methods
This study was part of a cohort observation study: Shared Decision Making in Mild to Moderate Dementia (ShareD: 13/114/93). Data were collected in the form of video- and audio-recordings from nine specialist memory clinics over four National Health Service (NHS) trusts (one in Devon – site A; three in London – site B), from May 2014 to February 2016.
Setting and data collection
The participating memory clinics were all based in secondary care, so patients had been referred by their general practitioner for assessment and diagnosis. Patients were asked to attend the appointment with a family member or someone who knows them well. The memory clinic process consists of history taking, cognitive assessment and neuro-imaging stages, which culminate in a diagnosis feedback meeting. In site B, these stages took place in separate appointments over a period of months. In site A, these appointments occurred on the same day in one four-hour appointment and there was no contact with the patients prior to the day of diagnostic feedback other than the appointment letter. In site B, some clinics offered pre-diagnostic counselling, i.e. to prepare the patient for a possible diagnosis and discuss their preferences for treatment and support should they have dementia. As in site A, the appointment letter did not mention the possibility of a diagnosis of dementia. The types of professionals present and the meeting length were similar across settings, suggesting no major resource issues that varied across the settings.
The HCPs delivering diagnoses in the memory clinics were identified in the participating trusts and contacted first by email and then by telephone. The consent rate was 88 per cent. Eligible patients who were due to attend a diagnostic feedback appointment with the participating HCPs were identified by administrative staff before their appointment by review of medical records. The only exclusion criterion was that the patient did not require an interpreter to talk to the HCP. The patients received information about the study with their appointment letters. The letter outlined what would happen on the day (e.g. scan followed by a meeting with memory clinic staff). When the patient and their companion arrived at the clinic, a researcher approached them to discuss the study further and obtain written, informed consent. For patients without capacity to provide informed consent to participate (one patient in our sample), we followed the ‘guidance on nominating a consultee for research involving adults who lack capacity to consent’ (Department of Health 2008).
Involving people with dementia as participants in research is challenging as it raises issues of informed consent: the voluntary agreement by an individual with capacity to understand the research protocol and decide whether to participate in the research, including the risks and benefits involved and voluntary withdrawal (Black et al. Reference Black, Brandt, Rabins, Samus, Steele, Lykestsos and Rosenblatt2008). In this context, capacity is important, as people with cognitive impairment may have reduced ability to understand and make decisions (Karlawish Reference Karlawish2008). However, people with cognitive impairment may have some ability to reason and make choices (Karlawish Reference Karlawish2008) and it is important to ensure that people who might want to take part in research projects are not excluded (British Psychological Society 2008). If people have reduced capacity, guidelines have been developed (British Psychological Society 2008) to enable people to participate by obtaining proxy consent provided by a spouse, family member or a carer, who can help to confirm that the person with dementia agrees to participate.
Of 315 patients approached to take part in the study, 215 took part (62%), of whom 101 were diagnosed with dementia. The first 60 consultations involving people receiving a diagnosis of dementia were selected consecutively from the larger sample for analysis. The diagnosis feedback meetings were video-taped using two GoPro HERO3 cameras, and audio-recorded to ensure maximum chance of quality audio capture. The researcher was not present to minimise observer effects. Recordings were transcribed verbatim, including additional features such as pauses, laughter and acknowledgement tokens (e.g. hm). The video-recordings were used during coding when it was not clear from the transcript if the HCP was speaking to the patient or the companion.
Data analysis
Concerns were operationalised as patient/companion expressions that convey worries, fears or unpleasant emotions, identified inductively from the transcripts (Silverman Reference Silverman2006), and defined as issues topicalised by the patient or their companion including: physical health problems; mental health concerns (feeling worried, anxious, afraid, upset, frustrated, embarrassed); negative consequences of symptoms of dementia (i.e. irritating people, getting confused, feeling like an idiot); associated dependency arising from symptoms; stigma and response to diagnosis (shocked, not prepared).
Patient and companion concerns were coded according to: (a) topic, (b) whether they were elicited or volunteered and (c) HCP initial response to the concern.
Topic
Each concern was coded using content analysis (Schreier Reference Schreier2012). This was done by creating a list of first-level exhaustive codes from the data describing the nature of the concern raised. Second-level codes (categories) were developed to group together similar first-level codes. The categories were then counted to obtain frequencies.
Whether they were elicited or volunteered
Concerns were elicited when the HCP solicited or facilitated their expression, usually through direct questioning (Zimmermann et al. Reference Zimmermann, Del Piccolo, Bensing, Bergvik, De Haes, Eide, Fletcher, Goss, Heaven, Humphris and Kim2011). Concerns were volunteered when patients or companions spoke about the concern uninvited. Elicitations could be topic-specific, e.g. prompting the patient or their companion to talk about their reaction to the diagnosis ‘what do you think about the diagnosis’ or eliciting perspectives on dementia ‘all of that can for some people can be something that they don't want to hear, how do you feel about that?’ These could then lead to concerns being discussed on the same topic as the question, or create space for concerns to be raised on a different topic. Elicitations could also not contain topic-specific content, e.g. ‘any questions?’ This was coded and taken into account in the analysis.
HCP initial response to the concern
The Verona Coding Scheme was used to code HCP initial response as either encouraging or discouraging elaboration of the concern (Zimmermann et al. Reference Zimmermann, Del Piccolo, Bensing, Bergvik, De Haes, Eide, Fletcher, Goss, Heaven, Humphris and Kim2011). Responses encouraging elaboration included acknowledgement or agreement tokens (yes, hmm, okay) alongside pauses, inviting the patient or companion to continue:
Companion: I think [patient name] doesn't want to admit that, there is a problem as such you know you.
HCP: Yeah (pause).
HCPs could also encourage elaboration with a more substantive verbal response, which often aimed to advise or reassure:
Patient: I don't want to be a burden to my children.
HCP: No I mean, I think it seems that they are, they're doing, little bits and bobs for you.
Patient: Yeah that's it.
HCP: As I said before if it ever gets to be a problem, there's other support and other help.
Patient: Okay.
HCP: So, I wouldn't feel like, I wouldn't worry too much about that at this stage.
Responses discouraging elaboration were characterised by HCPs not providing space for the patient/companion to elaborate further on their concern, most often by providing a minimal acknowledgement followed by an immediate change of topic:
Patient: You know what I mean, I grew up, if I see my mother and father I would go into a fit and stuff like that.
HCP: Okay, alright, so I think that the most important thing you can do to help your memory really is to stay off the alcohol.
Reliability
The coding was developed inductively in four analytic meetings between RM, PX and IM. A further five meetings took place to finalise the coding scheme, where there was discussion of 28 cases that were uncertain. For example, it could be unclear whether there was an expression of concern when patients, prompted by the HCP, were explaining symptoms: ‘[memory] ain't like it was but it's not terrible’. Additionally, there were cases where utterances across a number of turns related to the same concern. These were coded as one concern.
After the coding scheme was finalised, transcripts were coded by one author (IM) and a second rater (PX) coded 15 per cent (randomly selected) of the total 60 transcripts in order to establish inter-rater reliability. Cohen's kappa (McHugh Reference McHugh2012) was calculated using Stata (StataCorp 2015). Overall agreement was 89 per cent (mean of the agreement of individual weighted kappas): patient concern kappa = 0.73 (second coder identified five new concerns); companion concern kappa = 1; response encouraging elaboration (patient) kappa = 0.83; response discouraging elaboration (patient) kappa = 0.83; response discouraging elaboration (companion) kappa = 1; and response encouraging elaboration (companion) kappa = 1.
Results
Participant information
Nineteen lead HCPs, 60 patients and 59 companions were observed. In 90 per cent of the consultations there was one companion present and in 8 per cent there were two companions present and in one consultation (2%) there was no companion present. The HCPs who delivered the diagnosis to the patient/companion were all either psychiatrists (71%), geriatricians (15%) or specialist doctors (14%). One to six patient consultations were recorded per HCP (four HCPs had one consultation recorded). In 12 appointments other HCPs (dementia advisors, nurses and medical students) attended the meeting and participated in discussions. The majority of the elicitations and responses were attributed to the HCP (80%). The HCP number of years working in dementia ranged from one to 25 years (mean = 12 years).
Thirty-five patients were recruited from site B and 25 from site A. Patients were aged from 52 to 92 years, with an average age of 81 years (for patient characteristics, see Table 1). The companions were either a (a) spouse/partner (30%), (b) child/child-in-law (42%), (c) sibling (2%), (d) friend (5%), or (e) other, e.g. nephew/residential home warden (10%).
Table 1. Participant (patient) characteristics

A total of 249 concerns were identified (see Figure 1). An average of four patient/companion concerns per consultation were identified, ranging from zero to 17 concerns per consultation. In 54 of the 60 consultations at least one patient concern was coded, and in 36 of the 60 consultations at least one companion concern was coded. In five consultations there were no patient or companion concerns raised, however, there were HCP invitations to elicit concerns in all these consultations.

Figure 1. Concern and response totals.
Topics of concern
Table 2 presents the concerns expressed by patients and their companions. The most common concerns were in relation to the dementia diagnosis (N = 31). This involved personal negative representations of dementia, e.g. ‘I had a friend, they had a dementia and they ended up being very violent’ (Companion (C) 23). These also included responses to the word dementia, e.g. ‘I hope I haven't got the thing you talked about, dementia’ (Patient (P) 34), and concerns raised as a result of the disclosure of diagnosis, e.g. ‘I didn't think you were going to tell me this’ (P19). The next most common concern was about dementia symptoms – primarily regarding memory loss. This included worries about remembering names and things that happened recently, e.g. ‘say I watch Coronation Street you say what was it about half an hour after, I can't even tell you’ (P33), and companion anxiety about patients not accepting their symptoms, e.g. ‘I think [patient name] doesn't want to admit that there is a problem as such’ (C27). Patients discussed concerns about the dementia diagnosis and dementia symptoms almost twice as often as companions.
Table 2. Patient and companion concerns topics

Notes: P: patient. C: companion.
Patients and companions discussed concerns about their family, including traumatic events that are relevant to their current situation, e.g. ‘sometimes I feel really awful about [my brother's death]’ (P35). Concerns about the patient's mental and physical health, were the only concern topics to be raised more commonly by companions than patients.
A number of concerns related to medication. These included concerns about new tablets, side-effects and how to manage treatment, e.g. ‘I live on my own so I don't want any drastic treatment’ (P57). Additionally, both patients and companions expressed concerns about the role of the companion, the patient often with anxiety about being cared for, e.g. ‘I hate being nursed’ (P82), and the companion expressing concern about their default position as carer in the relationship, e.g. ‘I do feel a bit inhibited on sort of going ahead with making plans for a long-term carer’ (C83).
Some concerns were only expressed by patients. Frustration with dementia symptoms was common (N = 25). This included concerns about losing competencies, e.g. ‘they exasperate me and frustrate me, looking for a word’ (P24), and worry about the impact of their memory problems, e.g. ‘when something happens or anything, I sort of afterwards think oh I've lost my cool’ (P79). Also, only patients expressed concerns about prognosis and how dementia would affect them in the future, e.g. ‘what does the future hold for me?’ (P117), as well as concerns surrounding lack of social connection, e.g. ‘you can always go out for a walk but you see nobody at that time ‘cause all families are indoors you know’ (P84).
Finally, other concerns involved patient and companion concerns that occurred infrequently and were thus grouped together. These included concerns about the patient's difficult life, not liking visiting hospitals, a concern about the amount of information given to the patient, e.g. ‘I think she won't you know take all this in’ (C101), smoking, and a patient concern about the disability determination services (DSS) examination, e.g. ‘I had to go through an examination through that bloody DSS doctors, well my God, the questions on there’ (P33).
Elicited and volunteered concerns
There were a total of 177 HCP elicitations of patient/companion concerns, three elicitations on average per consultation. There was one consultation where there was no elicitation of concerns by the HCP, but six patient and companion concerns were expressed.
Forty-three per cent of the concerns in the consultations were elicited by the HCP. In 67 per cent of these cases (29% of total concerns), the HCP's topic of elicitation was on the same topic as the concern raised, e.g. in this discussion about the patient's (P31) hallucinations:
HCP: Tell me about them.
Patient: Yeah, well, I can't really tell you about them because um nobody else ever believes you.
The remaining 33 per cent of elicited concerns (14% of total concerns) were a result of HCP elicitations that were non-specific in topic (e.g. ‘how are things?’) or in a small minority (7% of elicited, 3% of all concerns) concerns were a response to elicitations on a different topic, e.g. (P105):
HCP: Do you have any questions about diagnosis or treatment?
Patient: Not really. I've got real problems when, just recently when I lost my diary, financial details.
Some elicitations were addressed to the companion, e.g. eliciting concerns about being a carer, e.g. ‘I guess that you're the one Mr [name] who actually has to bear the brunt of all of this and all these changes’; and managing on a daily basis, e.g. ‘do you have any worries about do you know managing from day to day?’
There were four instances where the companion elicited a patient's concern, e.g. their reaction to the diagnosis, e.g. ‘does it bother you … that the doctor's telling you have dementia I mean, does it frighten you?’
HCPs elicited some topics of concern more often than others (see Table 3). For concerns about the dementia diagnosis, symptoms of dementia and patient's mental health, the number of concerns that were elicited by the HCP was roughly proportionate to the number of overall concerns elicited (around 40%). However, some of the concerns were more often volunteered by patients than elicited by HCPs. For example, out of the 28 concerns raised about patient physical health, two of these were directly elicited (with an elicitation on the same topic) by the HCP. Similarly, none of the concerns about prognosis or lack of social connection were directly elicited by the HCP. Another topic of concern that was mainly volunteered by the patient was about frustration with memory problems, where five of the 25 concerns were elicited directly. Additionally, while the role of the companion was raised in five of nine cases as a result of the elicitation, only two of these were results of direct elicitations about the companion role.
Table 3. Health-care professional elicitation and response by concern type

Seventy invitations to elicit concern were not followed up by a concern. These cases usually involved patients reporting that they were not concerned about the topic discussed – such as the following two examples (P31 and P117):
HCP: Now is there anything that you're, any problems that you have that we haven't talked about?
Patient: I can't think of any.
HCP: What about driving is that a concern for you?
Patient: It's not a concern for me at all but I haven't driven for ages.
HCP response to concerns
All 249 concerns were given a response by the HCPs. The majority (62%) were given a response that encouraged elaboration, while in more than a third (38%) elaboration was discouraged by HCPs (see Figure 1).
When concerns were elicited by HCPs, they were more likely to respond by encouraging elaboration of the concern (71%) than not (29%). On the other hand, when concerns were volunteered by the patient or their companion, HCPs were (comparatively) more likely to discourage elaboration of the concern (55%) than not (45%).
HCPs encouraged elaboration of concerns more often with some topics of concern than others (see Table 3). Concerns about family, patient's mental health and the role of the companion were encouraged over 80 per cent of the time they were raised. Concerns about frustration with memory problems were encouraged in nearly three-quarters of cases, and concerns surrounding medication were encouraged in two-thirds of cases.
Concerns about the dementia diagnosis, dementia symptoms and lack of social connection were encouraged in over half of the cases in which they arose as topics of concern, but still discouraged in over 40 per cent of cases. Elaboration of concerns about prognosis were discouraged in 77 per cent of cases in which they arose.
Discussion
There were 249 concerns across the 60 consultations. The majority of consultations contained at least one patient concern and over half contained at least one companion concern. Patient and companion areas of concern included: the symptoms of dementia, receiving a diagnosis, dementia symptoms, patients’ mental and physical health, and prognosis. HCPs elicited more patient than companion concerns, and 43 per cent of all concerns raised. Concerns elicited by the HCPs were mostly linked with the agenda of diagnosis feedback: dementia symptoms, reaction to diagnosis and mental health. Concerns about physical health or prognosis were mostly volunteered. HCPs’ responses to concerns were more likely to encourage elaboration when they elicited the concern. Concerns about family and mental health were encouraged most often. Nearly 40 per cent of concerns were followed by HCPs discouraging elaboration of the concern by changing topic, and concerns about prognosis were most commonly discouraged.
Unsurprisingly given the setting of the dementia diagnosis feedback meeting, the majority of concerns were about perceptions of or reactions to the diagnosis, or concerns surrounding the symptoms. Previous studies have shown that the stigma of dementia can delay seeking a diagnosis (Speechly, Bridges-Webb and Passmore Reference Speechly, Bridges-Webb and Passmore2008), which is demonstrated by the negative perspectives about dementia that were raised by patients and companions in the diagnostic meeting. The negative reactions to the diagnosis itself have also been found in other studies of dementia diagnosis delivery (e.g. Aminzadeh et al. Reference Aminzadeh, Byszewski, Molnar and Eisner2007; Brooker Reference Brooker2003; Derksen et al. Reference Derksen, Vernooij-Dassen, Gillissen, Olde Rikkert and Scheltens2006). The concerns raised around patients’ mental health, primarily around anxiety and low mood, also reflect studies of patient experiences of receiving a dementia diagnosis, with many patients experiencing stress when noticing the early stages of dementia symptoms (Cahill et al. Reference Cahill, Gibb, Bruce, Headon and Drury2008; Perry-Young et al. Reference Perry-Young, Owen, Kelly and Owens2018; Samsi et al. Reference Samsi, Abley, Campbell, Keady, Manthorpe, Robinson, Watts and Bond2014).
Concerns about the diagnosis and the patients’ mental health were most often elicited by HCPs. This again reflects the context of the dementia diagnostic feedback meeting, where the HCP's agenda will include the need to discuss the diagnosis and the patient's symptoms (including in test result feedback). This is also demonstrated by concerns surrounding medication also being the most often elicited by HCPs, as prescribing cholinesterase inhibitors is indicated on receiving a diagnosis of Alzheimer's disease. That HCPs are eliciting patient concerns about taking medication demonstrates an aim to involve patients in medication decisions (Stevenson et al. Reference Stevenson, Barry, Britten, Barber and Bradley2000).
That some of the most commonly volunteered concerns were around family issues and the patient's physical health reflects the impact of dementia on the wider family system and the presence of comorbid illnesses. The symptoms of dementia often occur alongside other difficulties, both physical and psychological (Clague et al. Reference Clague, Mercer, McLean, Reynish and Guthrie2017), with most frequent comorbid illnesses in dementia patients identified as hypertension, diabetes, coronary artery disease, chronic heart failure and stroke (Schubert et al. Reference Schubert, Boustani, Callahan, Perkins, Carney, Fox, Unverzagt, Hui and Hendrie2006). Having comorbid illnesses can negatively impact patient–HCP communication as research shows that patients with multiple chronic conditions report worse experiences in health care more frequently than those with one condition (Paddison et al. Reference Paddison, Saunders, Abel, Payne, Campbell and Roland2015).
Additionally, studies have shown that a dementia diagnosis can have an effect on a person's social network: changing roles and responsibilities of the people surrounding the person with dementia, as well as impacting on the person with dementia's identity within that network (Bunn et al. Reference Bunn, Goodman, Sworn, Rait, Brayne, Robinson, McNeilly and Iliffe2012). While the role of the companion was not one of the most common topics of concern, this also reflects the wider impact of the diagnosis. It may be argued that patients topicalising concerns about physical health or family problems may demonstrate a lack of awareness of the significance of their memory problems (Karnieli-Miller et al. Reference Karnieli-Miller, Werner, Aharon-Peretz, Sinoff and Eidelman2012). However, these topics were actually raised more often by companions, demonstrating that they are not just being raised to avoid discussions of memory problems, but are occurring alongside and are of equal importance as concerns about memory.
That these types of concern, which are less explicitly related to the dementia diagnosis, were not elicited by the HCP as often again reflects the HCPs’ agenda in the diagnosis feedback meeting. Research demonstrates that HCPs are under institutional pressure to cover certain topics within the diagnosis feedback (Bailey, Dooley and McCabe Reference Bailey, Dooley and McCabein press) which may account for the fact that they are not asking about wider social issues that may be concerning the patient. This is in line with previous research that shows that a lack of social connection and daily activities are the most common unmet needs after a dementia diagnosis (Black et al. Reference Black, Johnston, Rabins, Morrison, Lyketsos and Samus2013; van der Roest et al. Reference van der Roest, Meiland, Comijs, Derksen, Jansen, van Hout and Dröes2009). The fact that patient frustration about their memory symptoms were also volunteered rather than elicited in the majority of cases also reflects previous research showing that HCPs avoid exploring emotional aspects of people's dementia symptoms (Zaleta and Carpenter Reference Zaleta and Carpenter2010). Given the call for patient-centred care in dementia (Brooker Reference Brooker2003; Robinson, Clare and Evans Reference Robinson, Bamford, Briel, Spencer and Whitty2010), it may be that services need to take a more holistic approach in discussing dementia with patients and their companions at this early diagnostic stage, focusing not only on the condition itself but also on wider aspects of the patient's current situation.
When considering how HCPs respond to the concerns, concerns about psycho-social issues were encouraged in over 80 per cent of instances, whereas concerns surrounding the diagnosis, symptoms and medication were encouraged in 60–70 per cent cases. Additionally, while the lack of social connection of the patient had not been elicited directly in any instances, it was encouraged in nearly 60 per cent of cases where it was discussed. Therefore, once concerns have been raised, HCPs are digressing from the relevant topics of diagnosis feedback to discuss other matters concerning patients and companions.
Although the majority of concerns were followed by HCPs encouraging elaboration of the concern, nearly 40 per cent were followed by HCPs discouraging further discussion of the concern. Additionally, HCPs were far more likely to encourage elaboration when they had initially elicited the concern. This HCP control of the communication has been described in other dementia health-care settings (Karnieli-Miller et al. Reference Karnieli-Miller, Werner, Aharon-Peretz, Sinoff and Eidelman2012; Sakai and Carpenter Reference Sakai and Carpenter2011), and is common across all HCP–patient interactions, which by their nature will be driven by the HCP in their role as a questioner and the expert in the interaction (Heritage Reference Heritage, Duchan and Kovarsky2005). The fact the HCP directs the discussion of concerns in the diagnostic feedback is thus to be expected. In fact, given that studies of cancer consultations show HCPs do not explicitly invite patients to express concerns at all (Brandes et al. Reference Brandes, Linn, Smit and van Weert2015) and often actively discourage the disclosure of concerns (Zimmermann, Del Piccolo and Finset Reference Zimmermann, Del Piccolo and Finset2007), the fact that HCPs are engaging patients and companions in discussions of their concerns may demonstrate enhanced patient involvement compared to other settings. However, research shows that patients and companions feel their concerns around dementia diagnosis are disregarded (Manthorpe et al. Reference Manthorpe, Samsi, Campbell, Abley, Keady, Bond and Iliffe2013), and thus there may need to be more effort to address concerns more often, despite other institutional pressures.
A key topic of concern that HCPs appeared to be avoiding was that of prognosis. While there were only 12 cases where patients raised concerns about prognosis, this was the only topic of concern where none of these cases were a result of direct elicitation from the HCP. Additionally, this was the only concern around which HCPs discouraged further discussion. HCPs have highlighted anxiety about discussing prognosis, specifically in causing a negative emotional impact on the patient at the early diagnosis stage, as well as uncertainty in predicting an individual patient's prognosis given the wide variability of the progression of dementia between individuals (Bailey, Dooley and McCabe Reference Bailey, Dooley and McCabe2018). This is not just an issue in dementia, but has also been discussed in oncology (The et al. Reference The, Hak, Koëter and van der Wal2000). However, given that patients are raising these concerns without invitation by HCPs, this is an area that may need to be addressed despite the challenges it presents. One of the key reasons for increasing dementia diagnosis rates is to encourage advance care planning in dementia (Brown Reference Brown2015). Given that a recent survey of people with dementia found that nearly half see planning for the future as an unmet need (Black et al. Reference Black, Johnston, Rabins, Morrison, Lyketsos and Samus2013), concerns about prognosis will need to be discussed explicitly in order to aid these discussions. Avoiding details about prognosis might lead to reduced opportunities for patients and their families to adjust their lives accordingly and aim towards attainable goals for the future (Fallowfield, Jenkins and Beveridge Reference Fallowfield, Jenkins and Beveridge2002; Hancock et al. Reference Hancock, Clayton, Parker, Wal der, Butow, Carrick, Currow, Ghersi, Glare, Hagerty and Tattersall2007).
Lastly, the role of the companion's concerns is interesting, especially related to previous literature around the triadic nature of medical consultations in dementia (Dooley, Bailey and McCabe Reference Dooley, Bailey and McCabe2015; Karnieli-Miller et al. Reference Karnieli-Miller, Werner, Aharon-Peretz, Sinoff and Eidelman2012). That patient concerns occurred twice as often, and were elicited over twice as often, than companion concerns potentially demonstrates that HCPs are oriented towards discussing the diagnosis with the patient rather than the companion as previous literature has suggested (Hasselkus Reference Hasselkus1994; Karnieli-Miller et al. Reference Karnieli-Miller, Werner, Aharon-Peretz, Sinoff and Eidelman2012). That HCPs are eliciting patient concerns could be particularly beneficial for this population as people with dementia show reduced conversational initiation (Blair et al. Reference Blair, Marczinski, Davis-Faroque and Kertesz2007).
In many cases, companion concerns were similar to those of patients, particularly surrounding the dementia diagnosis. However, concerns that were not discussed by companions support previous findings that companions may withhold from entering discussions about certain topics with HCPs in order to save the ‘face’ of the person with dementia (Dooley, Bailey and McCabe Reference Dooley, Bailey and McCabe2015). For example, the literature suggests that people caring for people with dementia find the lack of information about prognosis frustrating (Bamford, Lamont and Eccles Reference Bamford, Lamont and Eccles2004; McCabe, You and Tatangelo Reference McCabe, You and Tatangelo2016). However, they did not raise these concerns in the current data. Furthermore, in many previous studies companions report a burden of care-giving and not having enough psychological support (e.g. McCabe, You and Tatangelo Reference McCabe, You and Tatangelo2016), but concerns about the companion role were rare. Companions also did not raise concerns about patient frustration with symptoms or lack of social connection. The fact that companions are avoiding these topics in the diagnosis feedback meetings, but commonly reporting them as major areas of concern after the meeting, may be a form of ‘protective care-giving’ – trying not to upset the person with dementia by talking about difficult topics with the HCP (Hasselkus Reference Hasselkus1994). HCPs highlight these subtle aspects of the triadic interaction as a key challenge in dementia diagnostic feedback (Bailey, Dooley and McCabe Reference Bailey, Dooley and McCabe2018). However, research demonstrates the importance of considering the person with dementia and their companion as a whole unit in order to support the adjustments surrounding a dementia diagnosis (Robinson, Clare and Evans Reference Robinson, Clare and Evans2005), and thus strategies to meet the needs of both patients and their companions are an important area for future research.
Strengths
The analysis was based on real-time video-recordings of people receiving a diagnosis of dementia rather than interviews and included a range of HCPs from multiple memory clinic services across urban and rural areas. Analysing transcripts in detail (turn by turn) facilitated comprehensive exploration of how concerns were elicited/volunteered and further encouraged/discouraged by patients, companions and HCPs.
Limitations
There were some positive emotions expressed in the consultations, however, the low frequency did not lend itself to systematic coding. Video-recording the consultations could have had an impact on the HCP–patient interactions. There could also be selection bias for the HCPs and patients who agreed to participate. The HCPs who agreed to participate may have been more interested in communication while the patients may have been more engaged in the memory clinic process. Reliability coding was conducted on 15 per cent of the sample and, although there was high agreement across all categories, there was some variation across coders in identifying a patient concern.
Conclusions
Patients and companions expressed concerns about the symptoms of dementia and the accompanying frustration along with feeling shocked about receiving a diagnosis, alluding to negative portrayals of dementia. They were also concerned about non-dementia issues such as physical health and family, reflecting patients’ holistic approach to medical consultations. Patient concerns were more often elicited than companion concerns, which is encouraging given the potential for patient marginalisation in the context of dementia. Avoidance of prognosis demonstrated delicacy in discussing future planning at dementia diagnosis.
Practice implications
For doctors and other health-care professionals involved in the dementia diagnostic process, it may be helpful to be aware of the types of concern patients and companions express when receiving a diagnosis of dementia. This may help to attend to these concerns proactively and support long-term planning, thereby reducing anxiety for people.
Acknowledgements
We thank all the patients, companions and health-care professionals who allowed us to film the consultations, Dr Maya Soni, Professor Gill Livingston and the ShareD project team. This paper presents independent research funded by the National Institute for Health Research (NIHR) under its Research for Patient Benefit (RfPB) Programme (Grant Reference Number PB-PG-1111-26063). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health. RM, NB and JD contributed to the conception and design of the study. RM, PX and IM analysed and interpreted the data and drafted the article. JD was involved with data collection, interpretation of the data and drafting of the article. All authors contributed to revisions of the article and approval of the final version. This study was approved by the NRES Committee London, North East REC Office (REC Reference 13/LO/1309). There are no conflicts of interest.