Understanding and integrating the needs and preferences of people living with dementia in the inpatient setting: a qualitative study

Abstract

Background

People living with dementia (PLWD) have poorer outcomes than cognitively normal people when admitted to hospital. One reason for this difference is related to the challenges in learning and integrating the needs and preferences of PLWD into clinical care. We aimed to obtain a range of perspectives on the challenges in supporting PLWD in hospital and explore opportunities for improvement.

Methods

Using an exploratory qualitative study design, we conducted interviews with nine people supporting PLWD (current / ex-spouses or children of PLWD) and 11 nursing, medical and allied health staff members at a single Australian hospital. Data were thematically analysed using a framework approach.

Results

Participants described how best practice in supporting PLWD included understanding and integrating patient needs and highlighted the importance of family and the multidisciplinary team working in partnership. A number of factors inhibiting quality care provision were described, including uncertainty around responsibility for communicating with families to understand needs of PLWD; unsuitable tools; lack of opportunities for families to communicate with staff; and, resource and environmental constraints. Participants discussed potential for improvement, emphasising the need for a pre-emptive, rather than reactive solution. They expressed support for the idea of a ‘hospital admission kit’, containing both information about PLWD and their familiar items. Implementation considerations were also noted, with various perspectives on the timing of initiation, updating, responsible person(s), format, content and how it should be incorporated into clinical workflows.

Conclusions

We found that hospital staff and those supporting PLWD felt that integrating the needs and preferences of PLWD into hospital care was important. The concept of a pre-prepared ‘toolkit’ that was ready in the case of a need to attend hospital was felt to be valid and potentially helpful. More work is required to design aspects such as format, content and the workflows needed to generate accountability and reliability in creating, updating and incorporating it into hospital care.

Clinical trial number

Not applicable.

Background

There is a large body of evidence reporting that people living with dementia (PLWD) who are admitted to hospital have poorer outcomes than inpatients without cognitive impairment [1, 2]. These poorer outcomes include patient-centred outcomes such as greater complications, greater loss of functional independence and increased mortality, as well as process measures such as longer hospital stays and greater healthcare system costs [1]. Although involving patients in their own care is well-established to improve outcomes [3] and is a core part of modern clinical care [4,5,6,7,8,9,10,11], it is often done informally, inconsistently and is poorly communicated to others [4, 6, 7, 12,13,14,15,16].

There are few tools to support health services staff to learn the needs and preferences of PLWD [6, 13, 14, 17] and existing tools are often limited in routine use by a lack of clear guidance about how they should be operationalised, with previous work highlighting that there is often poor understanding of how relevant information should be generated, updated and incorporated into clinical care [18, 19]. One reason for this lack of understanding is that many existing tools have been designed with a focus on either the perspective of those supporting a PLWD at home or hospital staff using such tools [18, 19]. We are unaware of a tool that has been designed by both groups in partnership.

To help develop a tool to facilitate a process for identifying the needs and preferences of PLWD when admitted to hospital, we aimed to explore various perspectives on understanding and integrating the needs of PLWD when they are admitted to hospital, as well as capture opportunities for improvement and potential solutions. The broad research questions for each group were: (1) what information is important for hospital staff to know when providing hospital care for a PLWD? and (2) what processes would help staff identify and use this information to meet the needs and preferences of PLWD?

Methods

Study design

This was an exploratory qualitative study designed to understand various perspectives on understanding and integrating the needs and preferences of PLWD when they are admitted to hospital and was underpinned by a phenomenological approach [20].

Study population and setting

This project is part of the Living Labs Research Program of the National Centre for Healthy Ageing (ncha.org.au), a partnership between Monash University and Peninsula Health, a major metropolitan health service in Victoria, Australia. The study was conducted at the Acute Care of the Elderly (ACE) Unit at Frankston Hospital, an acute hospital with 454 beds within Peninsula Health. The ACE unit is co-located on an oncology ward and can provide acute medical care for a maximum of 30 older people at any time. Frankston Hospital is an acute university-affiliated tertiary metropolitan teaching hospital with 454 beds. Guided by the concept of ‘information power’, i.e. the more information a sample holds that is relevant to the study aims, the lower number of participants is required [21], we included a small number of staff across a range of relevant disciplines with varying levels of seniority. We aimed to recruit 10 people admitted to the Acute Care of Elderly (ACE) Unit at Peninsula Health with a known history of dementia, 10 people who are supporting someone living with dementia admitted to the ACE Unit and 10 hospital clinicians caring for PLWD on the ACE Unit. Hospital clinicians were purposively selected to ensure a variety of perspectives based on professional role (nursing, medical and allied health) and levels of seniority (junior to senior positions). This selection process aimed to capture a broad range of experiences in caring for PLWD. Hospital clinicians were eligible to participate if they had worked in the ACE Unit for more than 7.5 h in the 2 months preceding the interview.

Recruitment

Potential participants were invited to participate in a semi-structured interview by Peninsula Health clinical researchers, who had the expertise to determine whether PLWD had the capacity to consent to participate in an interview. These clinicians (senior nurse and geriatrician) had extensive training and expertise in working with PLWD. Capacity to participate in an interview was assessed via a clinical interview if the PLWD was alert enough to participate. Staff members were invited to participate by either the nurse lead of the project (FL) or the medical lead (CM). PLWD and those supporting them were approached to participate either face to face or over the telephone by the cognition nurse consultant (RN) or the medical lead (CM), not their usual treating clinician. Research aims and procedures were outlined in an Explanatory Statement given to all participants prior to the interview and all participants provided voluntary informed consent.

Data collection

A trained and experienced female researcher (AL) conducted all interviews via telephone or Zoom between July and December 2022. The interviewer was a Senior Research Officer with a background in psychology (BPsych (Hons)), who had no lived experience of working with or caring for someone with dementia, nor did they have any prior relationships with the participants. Participants were informed of the interviewer’s professional role and role in the study. No interviewer characteristics were reported. The interview guide (See Appendix 1) was developed to address the aims of the study. While this was not pilot tested, it was reviewed by the broader research team, including those working with PLWD and a consumer with lived experience of caring for a PLWD. Questions for support people explored PLWD’s hospital admission experiences; existing home routines and translation to the hospital environment; support provided by hospital staff; preparation for hospital admission; and communication about needs and preferences. Questions for staff explored tools to understand PLWD’s needs and preferences; integration of needs and preferences into care; and training and support. Both groups of participants were also asked to consider the usefulness of a ‘hospital admission kit’, similar to a ‘go bag’ used in maternal health settings, as a potential solution. This was defined broadly with the intention to gather initial concept ideas before progressing to a separate co-design phase. Therefore, it was not designed to constitute a comprehensive exploration of this concept. Probes were used throughout the interviews to encourage participants to explain or expand on their responses.

All interviews were conducted in English and the mean interview time was 36 min (range: 23–59). No non-participants were present during the interviews. Interviews were digitally audio-recorded and transcribed verbatim using an online transcription service. The first author reviewed the transcripts for accuracy. Participants were able to request a copy of their transcript for review, however no one did.

Data analysis

Interview transcripts were loaded into a computer-assisted qualitative data analysis software program (NVivo 20, QSR International Pty Ltd 2022, Doncaster). Thematic analysis was conducted by the first author using a framework approach [22]. This process commenced with repeatedly reading the transcripts to ensure familiarisation. Repetitions in text and meaning were identified and coded. Through an inductive and iterative process, key themes and subthemes were formed based on patterns and connections within and between transcripts. Similar themes were clustered. The resulting framework was applied to the original transcripts to check for meaningful links and connections and adjustments were made to ensure accurate representation. Direct quotations were used to illustrate key themes. Participants did not provide feedback on the findings.

Results

A total of 20 interviews were conducted. This included 11 healthcare professionals, including nurses (n = 9), allied health staff (n = 1) and medical staff (n = 1). A total of 13 people supporting a PLWD initially agreed to participate in the study with nine ultimately participating. Most PLWD who were admitted to the ward during the study period were too confused (many with superimposed delirium) to be able to concentrate for long enough to understand the purpose of the study and consent. Two PLWD were deemed potentially appropriate to participate in the study but did not consent to do so due to becoming frequently unwell. All staff members who were approached, agreed to participate. For staff, years of experience in their current role ranged from one month to 10 years. Support people included spouses (n = 4), ex-partners (n = 1) and children (n = 4) of those living with dementia. Among the PLWD that they were supporting, experience of supporting symptoms of dementia ranged from 8 weeks to 8 years. The majority of participants reported that the person that they were supporting was living at home (n = 6) prior to admission, with the remainder in a nursing home, respite or a serviced retirement village. However, one participant noted that the PLWD they were supporting would be moving into a nursing home following the recent hospital admission.

Three main themes and multiple sub-themes were identified and are described below. Additional detail and quotations supporting the themes and sub-themes are presented in Table 1.

Table 1 Themes, sub-themes and supporting quotations

Key principles of quality care for PLWD in hospital

Understanding and integrating needs improves patient care

All staff acknowledged the importance of understanding and integrating the needs of PLWD, highlighting that it helps to humanise patients, bring a sense of routine and familiarity to the hospital environment, engage them in their care, facilitate discharge planning, provide appropriate reassurance, de-escalate situations and be more proactive in terms of avoiding escalation of behaviours associated with dementia, thereby reducing the need for medical intervention and preventing or reducing patient harm.

Both staff and support people recounted several examples in which understanding and integrating patients’ needs worked well. These generally involved accommodating patients’ usual routines, encouraging independence, providing reassurance and allowing them to engage in activities of interest to them or tasks associated with their previous occupations, often utilising resources available on the ward.

Both staff and support people highlighted that there is no ‘one size fits all’ approach to caring for PLWD and that care needs to be individualised in order to respond to their behaviour as effectively as possible. However, it was also acknowledged that there were occasions where certain behaviours could not be prevented.

Partnering with families enables person-centred care

Staff members reported that the most efficient way to get to know patients and understand how to care for them in a resource-limited environment is often by engaging with family members rather than only talking to PLWD themselves. This is particularly the case if the cognitive presentation is different from their baseline (e.g., delirium) or in the setting of advanced dementia. Staff also acknowledged that while families generally know PLWD best, nursing home staff and primary care physicians could also provide useful background information, particularly if families were not involved in caring for PLWD at home. Staff recalled receiving very comprehensive information from nursing homes, particularly if the home had developed a behaviour care plan for the PLWD.

Staff emphasised the importance of speaking with families (or obtaining information from other sources) as soon as possible in order to incorporate strategies identified through past experience into current care plans, as well as to assist with forward planning / discharge planning. Limited availability of family members and the PLWD not speaking English made it difficult to obtain information, but one staff member reported that the family of PLWD usually made a significant effort to be available. Staff noted that COVID-19 acted as a barrier to obtaining information from families. Prior to COVID-19, staff could easily ask family members questions while they were visiting, whereas it often required multiple phone calls in the presence of visitation restrictions.

Directly involving family members in the care of PLWD was also seen as important as a de-escalation strategy, with staff acknowledging that family members can provide a sense of familiarity and often know how best to reassure them. Staff were also grateful for the extra physical assistance (e.g., repositioning in bed).

Support people also emphasised the importance of their physical presence in hospital, describing helping their family members with showering, feeding and dressing, as well as providing reassurance, noting that it is often difficult for PLWD to be left alone. Being physically present also enabled them to provide real-time information to staff. Support people recounted a number of situations in which their presence helped to ensure that appropriate care was provided.

Multidisciplinary teamwork and communication between staff facilitate holistic care

Staff emphasised that multidisciplinary teamwork and good communication was key to ensuring that PLWD receive the best care possible. They noted the holistic approach and multidisciplinary nature of good care plan development. Staff also acknowledged that COVID-19 further enabled multidisciplinary teamwork as meetings took place online and were easier for people to attend.

While many staff members favoured verbal communication, the importance of written communication (e.g., in the medical record or shift notes) was also highlighted to reduce duplication of effort. However, staff noted that, given the time taken to read all potentially relevant information as well as the time required to locate it, verbal communication can often be the most efficient as it enables staff to prioritise the most important information. Staff considered handover to be particularly important, arming staff with up-to-date information and strategies prior to seeing PLWD and also acknowledged that the more people communicate, the more knowledge staff accumulate, thereby improving care over time.

Learning to care for PLWD comes with experience

Staff highlighted the lack of formal training they had undertaken around caring for PLWD, instead learning from others and through experience. While some staff members specifically mentioned dementia modules offered by Dementia Australia and the University of Tasmania, others were not aware of any formal training opportunities. Some staff described conducting their own self-directed learning based on areas of particular interest.

Staff also reported attending workplace-based education and study days prior to COVID-19 and suggested that these types of educational activities were particularly useful for newer staff, but also served as refreshers for others.

Some staff described modelling specific strategies to other staff in order to build their confidence in caring for PLWD.

While most staff reported that they felt equipped with a toolkit of strategies to care for PLWD, some did not, suggesting the need for more education, particularly for new staff, whilst also acknowledging the value of refreshers for older staff. They suggested that education should focus on behaviour management, as well as building capacity for nursing staff to take on specialist roles.

Factors inhibiting quality care provision for PLWD

Responsibility for Understanding the needs and preferences of PLWD is unclear

While the various disciplines and team members involved in developing behaviour care plans were clear, responsibility for gathering additional information from families was not. Support people reported that they did not know to whom they should communicate their knowledge of the PLWD they were supporting to ensure that information would be passed on and implemented. Some staff suggested that responsibility for understanding and integrating the needs of PLWD and completing relevant tools should be clear, but should not fall to one person or discipline. One staff member suggested that responsibility for gathering information should fall to whoever contacted the family first, whether that be medical, nursing or allied health staff.

Current tools to understand and integrate the needs of PLWD are surface-level or reactiv

Staff described two main tools for understanding and integrating the needs of patients with dementia– namely, the Sunflower tool and Behaviour Care Plans. The Sunflower tool was developed by the Agency for Clinical Innovation (New South Wales) and visually displays a person’s preferred name and up to 9 other areas including important people, past occupation and hobbies [23]. Behaviour Care Plans are locally developed clinical documents that are completed for those identified as potentially benefitting from them that focusses on assessment regarding changed behaviours, potential behaviour triggers, and successful and unsuccessful strategies [24]. Staff had mixed opinions about the Sunflower Tool, but largely reported that it was only useful at a surface level. While the tool provided some conversation topics that could help to build rapport or potentially de-escalate a situation, staff felt that it did not contain meaningful strategies for supporting PLWD. Staff appreciated the prominence of the tool at the bedside, highlighting that while the information may or may not help to de-escalate a situation, it was easily visible at the point of care provision. Some staff acknowledged that the tool was often not completed or still displayed information about a previous patient and they were not stored for potential future admissions. They also reported that families often noticed the Sunflower tools and offered to help the nursing staff complete them but this was not possible during COVID-19.

Conversely, Behaviour Care Plans were seen as very useful due to their comprehensive nature and the practical strategies they contained. However, their completion was seen as very reactive. The significant time and resources required to develop them were also highlighted, with staff noting that this limited their use to only the most challenging patients. Staff reiterated the need for a tool that could be completed prior to hospital admission to ensure that all patients could benefit from tailored strategies that work to integrate their needs. Difficulties accessing the Behaviour Care Plans and poor awareness of their existence limited their use, which staff identified as a missed opportunity.

Neither Behaviour Care Plans nor the Sunflower tool was mentioned by any participating support people. Support people reported that the extent to which staff asked about the needs of PLWD varied. They reported providing information about their family members to hospital staff on an ad-hoc basis, either because a staff member asked about a particular need or support people noticed that a particular need was not being met.

Formal communication channels are lacking between hospital staff and support people

While support people did not report any difficulties voicing the needs of their family member once they had the opportunity to do so, opportunities to communicate with staff were often limited. Support people expressed a need for more formalised communication processes in order to stay informed about their family member’s care and assuage any uncertainty, particularly when they were unable to visit.

Support people emphasised the importance of closed-loop communication to ensure that requests have been actioned. Suggestions for improvement included scheduling regular updates (either via phone or email), having one point of contact (i.e. a patient liaison) with a direct number to call and implementing communication books.

Lack of staff and time impact the quality of care received by PLWD

Staff described how competing demands within the hospital environment limited the time they had to understand and integrate the needs of patients with dementia, including spending adequate time with patients and completing and updating available tools (e.g. Sunflower tool and behaviour care plan). It was noted that the responsibility for completion of the Sunflower tool frequently fell to nursing staff. Staff also mentioned that if the behaviour care plan had been completed, they often lacked time to read the information that had been documented and subsequently incorporate any strategies into care.

However, staff acknowledged the importance of spending time with patients and the benefits when they were able to take the time to do so.

While staff and support people reported the benefits of having additional staff available, including allowing one-on-one time with patients, completion of tools and identification of patients who may require specialised input, staff highlighted that they do still require adequate knowledge around working with patients with dementia in order to add value to the team.

Considerations for future improvements to understanding and integrating the needs and preferences of PLWD

Importance of a pre-emptive, rather than reactive solution

Staff expressed a need for some kind of solution, like the proposed hospital admission kit, in order to make care less reactive, spend less time trying to source information from other staff and families and allow more time for patient care and streamline discharge planning.

While most support people had not previously considered the idea of a hospital admission kit, they could see the value of doing so, feeling it could assist staff with rapport building and facilitate the provision of patient-centred care.

Hospital admission kits should contain both information about PLWD and their familiar items

Staff suggested that a hospital admission kit could include information about a person living with dementia and familiar items and ideally have this ready in case of a hospital attendance. Staff suggested including information to assist in getting to know patients and build rapport, as well as how to help support them in hospital. Specific suggestions from both staff and support people included: routines (e.g. what a normal day looks like and how and when they take their medication); religion; languages spoken; usual roles within the home; personality traits; stage of dementia; common behaviours (e.g. wandering); concerning behaviours (e.g. violent or aggressive tendencies); triggers for escalation; level of assistance required for various activities (e.g. dressing); level of continence; meaningful strategies to support them that have been effective in the past (as well as those strategies that were not effective); preferences for their general environment; main contacts and carers (including names of surrogate decision makers); important people in their life; likes and dislikes (e.g. food); hobbies / interests; pets; and previous occupations.

Staff also mentioned a number of physical items that could be useful to include, highlighting that as long as items are safe for the patient and not too loud, they should bring as much of the home environment into hospital as possible to foster a sense of familiarity. Suggestions included: photographs (e.g. of travels or family); items that represent their passions (e.g. football scarf or trophies); things that they use on a daily basis (e.g. electric shavers, hearing aids, glasses and iPads or other electronic devices); things to keep them busy (e.g. notebook and pen, magazines, newspapers, playing cards, fidget toys, puzzles, music and colouring books); favourite foods and drinks; religious items of significance (e.g. rosary beads); orientation-related items (e.g. clocks or calendars); and things to help them feel calmer and more comfortable (e.g. their own clothes, pillows, blankets and weighted toys). Two staff members expressed concern that patient items could be lost or damaged during their stay and some support people who had previously brought in personal items had also taken this into consideration when choosing what to bring. One support person was also worried about making the environment look too permanent, which influenced the items they chose to bring for their family member.

Although they had not put a ‘kit’ together prior to admission, some support people had brought in personal items for their family members and seen an immediate benefit. However, the integration of these items was largely by support people rather than staff (i.e. family members looking through photos with PLWD).

Information about PLWD should be readily accessible

Staff emphasised the need for the informational component of the hospital admission kit to be easily accessible in multiple settings (i.e. in the community setting for completion and in the hospital setting for utilisation and updating), but were unsure about how this could be coordinated. Support people suggested developing a form that was easily accessible online and via hard-copy to suit individual preferences. One participant also suggested that it could be completed via an app. They suggested that the form should include tick-boxes, space for photos and should be customisable (i.e., only sections relevant to the PLWD need be completed). While some staff stated a preference for a hard copy document, they acknowledged that it could easily get lost amongst other paperwork. Other staff suggested that it would need to be electronic so that it could be easily updated and travel with the patient across settings. However, they recognised the complexity of this format as not all staff can access the electronic medical record (EMR) and health services often use different systems, thereby limiting transferability. They also mentioned that if the hospital admission kit was to form part of the EMR, an alert would need to appear so that staff are aware that this additional information is available. Other suggestions included using the Australian government-developed, My Health Record, to enable access across settings, with a notification of this being visible within the local EMR [25]; use of a medical alert bracelet; or personal card using QR codes.

Variation in views regarding when hospital admission kits should be created

Some staff members suggested that the creation of a hospital admission kit should commence as early as possible (i.e. when someone is first diagnosed with dementia) as they are more likely to be relatively well and stable and can contribute to personalising the information included. Staff suggested that this process could be supported by a primary care physician, staff members working in an aged care facility or another support person who is familiar with dementia (either through personal or professional experience e.g. peers via Dementia Australia). They also noted that this information could then be updated as things change over time. Alternatively, it could be conducted at key moments of change (e.g. when someone is undergoing an assessment for formal social supports).

Conversely, some support people suggested that it would be good to have a conversation with a staff member and complete the informational component of a hospital admission kit when a PLWD is being admitted to the ward. They also noted the potential time available to complete this task when waiting to be seen in the Emergency Department.

The utility of a hospital admission kit is likely to be limited by the advancement of dementia

Both staff and support people acknowledged that the informational component of a hospital admission kit may not be representative of how a PLWD presents in hospital if they have deteriorated either prior to or following admission, such that family members may not know how to support them anymore.

While support people could easily think of topics to include in the informational component of a hospital admission kit, some struggled to think of physical items that would be useful (beyond photos) because their family members had either lost interest in the activities they had previously enjoyed or lacked the attention span to engage with them. They noted that this could make it difficult to put together the physical aspect of a hospital admission kit as the benefits of certain physical items may differ as dementia progresses. However, they could specify items that would have been useful to include prior to the advancement of dementia or onset of other physical limitations (e.g. deterioration in vision or hearing alongside dementia).

Discussion

We found that participants were consistent in highlighting the importance of understanding and integrating the needs of PLWD into clinical care provision and that communication helped people supporting a PLWD to partner with the multidisciplinary healthcare team to provide high quality care. Participants reported that a lack of time, staff and training, environmental constraints and a lack of opportunities for families to communicate with staff were barriers to providing quality care. Participants felt that currently used tools were not fit for purpose, but were supportive of the concept of having materials pre-prepared in case of the need of hospital attendance. Hospital staff and people supporting PLWD had a number of different suggestions for what such a ‘toolkit’ might entail. However, there were substantial variations in the nature of these suggestions, supporting our planned next steps to enter into a formal co-design process.

The high priority that participants placed upon incorporating the needs and preferences of PLWD into hospital care provision is supported by the results of previous studies [4, 6, 7, 13,14,15, 17, 26, 27]. Previous work has also highlighted the absence of effective interventions to facilitate this integration [6, 13, 15]. Participants in our study reported that they felt locally-used tools were not fit for purpose. This is consistent with existing literature examining a broader scope of widely available tools, (e.g. healthcare passports) further highlighting the need to consider both the design and implementation aspects of potential solutions [18, 19].

Conducting this study during the COVID-19 pandemic highlighted that processes involving face-to-face communication with family members to understand needs and preferences, are weak and easily disrupted. As such, ideal solutions will need to be functional in the absence of family members being physically present. As highlighted by participants, the designed solution will likely need to pre-empt rather than react to an acute deterioration necessitating hospital attendance. An important part of the design process will involve ensuring the responsibility to create, update and link this to healthcare workers is clearly assigned. PLWD, those supporting them, primary care providers, residential care homes, advocacy groups (e.g. Dementia Australia) and hospital staff were all identified as potential candidates. The optimal time to complete and update this information also requires further input. Suggestions included at the time of dementia diagnosis, at primary care visits, supported by family at home or on hospital admission/discharge. All of these epochs have specific advantages and disadvantages but do not necessarily need to be mutually exclusive. Similarly, the actions performed by staff to meet the hospital care needs of PLWD once they use the toolkit needs to be considered. Healthcare provider workflows will need to be designed in such a way that material within the toolkit will be sought and consistently incorporated into clinical care with clear understanding of roles and accountability.

Additional design considerations include the need for the hospital admission kit to be somewhat fluid or ‘living’ to accommodate evolving conditions, behaviours, needs and preferences, which presents a different design challenge, than with a ‘set and forget’ tool. Furthermore, staff highlighted that a significant amount of information would be useful to know, whilst also repeatedly reporting that they were time poor. Consequently, the quick and timely translation of key and/or filtered information will need to be considered in the design.

This study has a number of strengths and limitations. We were able to obtain the perspectives of a range of people supporting PLWD, including spouses and children. We also obtained perspectives from hospital staff over a range of clinical disciplines and experience. However, an important limitation was our inability to include PLWD. Although including this important perspective was an aim of this study, those PLWD who were potentially eligible did not remain medically stable enough post hospital discharge to be able to participate. We chose to recruit from the acute hospital setting as we wanted to optimise the ability of a PLWD being able to remember their recent hospital experience. Future work should consider recruiting people in the community setting to examine the feasibility of this approach. However, our experience highlights the clinical need and a range of design challenges associated with developing a toolkit for such a vulnerable and clinically unstable group. Reassuringly, many of the themes we identified in our study were consistent with those reported by a previous review of seven qualitative studies that did include PLWD [14]. A further limitation of our study was our focus on participants’ experience at a single clinical site and in those who speak English. Therefore, the generalisability of our results is unclear.

Conclusions

We found that hospital staff and those supporting PLWD felt that integrating the needs and preferences of PLWD into hospital care was important. The concept of a pre-prepared ‘toolkit’ that was ready in the case of a need to attend hospital was felt to be valid and potentially helpful. More work is required to design aspects such as format, content and the workflows needed to generate accountability and reliability in creating, updating and incorporating it into hospital care.