Decision‐making in cancer care for people living with dementia

Increasing numbers of people are expected to live with comorbid cancer and dementia. Cancer treatment decision‐making for these individuals is complex, particularly for those lacking capacity, requiring support across the cancer care pathway. There is little research to inform practice in this area. This ethnographic study reports on the cancer decision‐making experiences of people with cancer and dementia, their families, and healthcare staff.

rates. 1 They receive less treatment than those without dementia. [1][2][3] Individuals with dementia and their relatives may not disclose memory problems, and oncology clinicians may not ask about dementia, 4 meaning the implications of dementia for cancer treatment may be poorly understood.
Decision-making in cancer care and treatment is complex, requiring factors including age and comorbidities to be balanced with patient preferences, and treatment and prognosis expectations. 5 Many people with cancer experience decision-making uncertainty regarding treatment options, not fully understanding treatment intent or prognosis. 5 However, comprehension and decision-making abilities may be further decreased in people with dementia. 6 Few studies have specifically considered the role of decision-making in cancer care for people with dementia 7 and their limitations include single site studies 4,6,8 small-scale samples, 6,8 inclusion of people with Mild Cognitive Impairment as well as dementia 4 and only including staff perspectives. 6 The studies to date identify the central role of families in decision-making, 9 with people with dementia often depending on them to speak and make decisions on their behalf. 5,8 Clinicians who support decision-making may have different priorities to the person with CCD, and their beliefs and understanding about dementia can impact advice given and the treatments offered, meaning people with CCD can experience a power imbalance and lack of autonomy around treatment decisions. 6 Treatment decision-making for people with CCD is complex, particularly for those lacking capacity. 8 Relatives are integral in supporting decision-making, but often have unmet information needs and feel excluded from the process. 8 Reviews conclude decision-making in CCD remains under-researched, 8 particularly regarding treatment goals in palliative and end-of-life decision-making. 2 This paper explores cancer treatment decision-making in CCD as one theme identified within a larger UK study exploring cancer care experiences of people with CCD, their families, and healthcare staff. 9 2 | METHODS

| Design
An ethnographic approach was taken, consisting of participant observations, informal conversations, semi-structured interviews, and medical note review.

| Setting
Oncology and associated departments in two National Health Service (NHS) Trusts in one UK region and their local communities.

| Procedure
Ethical approval was gained from the Bradford Leeds Research Ethics In NHS Trusts, staff members (ie, nurse, consultant) identified and approached participants at post-diagnostic clinic appointment to establish interest in speaking to a researcher about the study. Study adverts were also shared with local community support groups and via social media.
Written informed consent was obtained for interviews and indepth observations from all participants with capacity, with advice provided by a consultee for people lacking capacity. 10 Ongoing consent was established prior to each observation.

| Participants
Participants were people with a diagnosis or suspected dementia (assessed via completion of the FAST 11 ) of any type, (mild to moderate severity), who were receiving/had recently received cancer treatment; relatives of people with CCD who had currently/previously received cancer treatment; and oncology staff with experience of supporting people with CCD. Purposive sampling was used to recruit a range of staff roles.

| Data collection
Researchers observed oncology appointments including consultations, treatment, and follow-up. Participants were typically met in reception areas to observe their entire hospital visit, during which informal conversations were also held, and documented within field notes. Semi-structured interviews focused on experiences of cancer care for people with CCD and were conducted in their home or a quiet hospital room. Interview topic guides were developed with the Lay Advisory Group. People with CCD and their relatives were interviewed individually, or as a dyad/ group, depending on preference. Interviews were audio recorded and transcribed verbatim. Participants with CCD currently being treated at participating hospitals could opt to participate in an interview and/or observations. Medical notes were reviewed for any mention of dementia or associated issues. (See Appendix S1, 1 for more information on participants and procedure).

| Data analysis
Analysis was an iterative process exploring content and patterns in the data via triangulation across sources. Initial coding of interviews and observations was conducted independently by A.G., R.K., C.S. and F.C.
Ethnographically informed thematic analysis 12 was used to develop a coding framework. This was discussed and refined with the wider research team, including lay members. Data collection and analysis ran concurrently, with early analysis informing subsequent data collection.

| RESULTS
Participant characteristics are summarized in Table 1. Observations and informal conversations (totaling 46 hours) were conducted with 12 people with CCD, 8 of whom also participated in interviews, and their families. We conducted 37 interviews (13 people with CCD, 18 relatives and 19 staff) lasting between 9 and 122 minutes due to varying communication abilities and participant preference.
Decision-making challenges were a major theme identified in the data. Not all study participants contributed data to each identified theme. Within decision-making three sub-themes were developed (please see Table 2 for an overview); 1. Ethical dilemmas and challenges 2. Whose decision?

| Ethical dilemmas and challenges
Decision-making raised complex ethical dilemmas, for example the potentially negative impact of people receiving their cancer diagnosis multiple times due to short-term memory loss. Staff and relatives experienced uncertainty and sometimes conflict over whether this upsetting process was the right thing to do. Thus, extended decisionmaking processes and support were required. Ethical concerns arose when people became distressed receiving cancer treatment. Staff questioned whether they were doing the right thing, even though the person with dementia had consented and/or it was decided to be in their best interests.
It was very, very uncomfortable for all the staff because it's, "we shouldn't be doing this, because he's agitated.
He's not liking it." So, it's supporting the staff, it's like, "well, if you want me to treat him, I will, because I've had the long conversation with him and he definitely

| Evaluating treatment options
Interviewees reported a range of ways in which the additional needs of people with CCD were considered within treatment decisionmaking including the range of options available, their ability to cope with these and any resulting side-effects or longer-term impacts, and likely prognosis. "Fitness for treatment" was frequently discussed, with concerns here including how well people with dementia might tolerate anaesthetic, hormone treatment or chemotherapy. Decisions were more complex when cumulative treatments were possible. In other cases, decisions were determined not on physical treatment experiences or consequences, but by potential after effects on memory problems or the potentially traumatic treatment impacts.
If they've had for instance, a breast removed, they may have forgotten why that's been done. They may feel that they've been mutilated against their will or wishes. In contrast, when the "bigger picture" was considered, clinicians could recognize dementia-related concerns, such as the potential impacts of treatment on the person due to their dementia. to ensure decisions made are fully informed, but does not routinely happen in practice. 19 We found relatives were integral to treatment decision-making, supporting reiteration and explanation of key information and discussion of treatment options. Previous research has highlighted their important role in facilitating inclusion of the person with CCD in decision-making 6 but has also indicated there may be deference to relatives in this process. 4,8 Our study confirmed these findings and identified the considerable burden caused by reliance on families for decision-making. While family burden associated with caring for an older person with cancer is well-documented, 20-22 the impacts of decision-making have not been highlighted, and our work indicates this burden may be greater when someone has CCD.
Evaluating cancer treatment options was also made more complex by comorbid dementia. Whilst cognitive impairment is known to reduce treatment adherence, 23 in our study fitness for treatment, rather than cognitive impairment specifically, informed decisions. Where this "bigger picture" was considered -including the impact of a person's dementia on prognosis, treatment suitability, understanding of potentially cumulative treatment options and ability to identify and report side-effects-shared treatment decisions and plans were able to be made. Again, while cancer treatment decisions, particularly in older people, always require consideration of an array of complex issues, 24 dementia brought greater complexity and prognostic uncertainty. Geriatric evaluation can prove beneficial in supporting decision-making in older populations, 25 however, as these did not arise within the data we collected, it was unclear whether these routinely occurred in oncology settings.

| Study limitations
This study was conducted in one area of the UK, across two NHS Trusts. The experiences of people with CCD may vary between hospitals, with further research required to validate our results. Triangulation of data sources was not possible for all participants and we sometimes relied on retrospective reflections on decision-making experiences. Although subjectivity is inevitable within ethnographic research, we mitigated this by acknowledging our preconceptions before data collection began, data being collected by two authors, reflexive journal keeping, group data analysis, and synthesis of multiple data sources. 26

| Clinical implications
Our study has indicated the following implications for oncology practice for people with dementia: • Treatment decision-making requires additional time and repeated re-evaluation; longer and/or additional appointments may be required to support this; • Oncology clinicians may benefit from additional training and support in communication with, and assessment of, decision-making capacity in people with dementia; • Families play an important role in treatment decision-making and may experience additional distress and burden when supporting a person with dementia through this. Consideration should be given to how their information and support needs can be met within oncology services; • Oncology clinicians need to ensure they apply an individualized ability focused assessment, considering the "bigger picture" of how dementia may impact cancer treatment options and side-effects, alongside potential longer-term impacts.

| CONCLUSIONS
Cancer treatment decision-making for people with dementia is complex, challenging and fraught with uncertainties, and should be considered as an ongoing process with emotional impacts for the individual, relatives, and staff. Oncology clinicians are currently working with limited evidence-based guidance on how dementia impacts on treatment prognosis, adherence and efficacy. More research is required to provide a stronger evidence base for treating people with dementia in oncology services.