Refusal and resistance to care by people living with dementia being cared for within acute hospital wards: an ethnographic study

Article history

The research reported in this issue of the journal was funded by the HS&DR programme or one of its preceding programmes as project number 13/10/80. The contractual start date was in January 2018. The final report began editorial review in February 2018 and was accepted for publication in October 2018. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HS&DR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

none

Disclaimer

This report contains transcripts of interviews conducted in the course of the research, or similar, and contains language that may offend some readers.

Copyright statement

© Queen’s Printer and Controller of HMSO 2019. This work was produced by Featherstone et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2019 Queen’s Printer and Controller of HMSO

Acute hospital care

There is an increasing recognition that a key aspect of care for people living with dementia that needs immediate action is improving their experiences and outcomes following an admission to hospital for a condition unrelated to their dementia. 7,12,30,31 The National Audit Office similarly advocates increased investment to improve care for people living with dementia within general hospitals to enable long-term savings and shorter hospital admissions. 6 The Alzheimer’s Society’s most recent report32 has found that poor care for people living with dementia is still widespread and that the quality of care varies widely between hospitals, with the admission for a person living with dementia being up to five to seven times longer than for other patients over the age of 65 years in the worst-performing hospitals. 33

UK health policy emphasises the importance of keeping people living with dementia out of hospital, minimising the duration of any necessary hospital admission and supporting people living with dementia in the community. 31 However, although health and well-being outcomes for people living with dementia are better when they are supported within the community, there is also an acknowledgement that, in reality, many people living with dementia also need to be cared for in hospitals and other institutions.

Hospitals are not designed to care for this group of patients. 34 Evidence suggests that the structural conditions and standardised care plans within the acute setting often do not fit the needs of people living with dementia and their families,35 with one ethnographic study identifying systemic and organisational factors within the hospital setting that compromised the ability of clinical staff to provide dignified care. 24 Hospital systems are designed to care for patients with one clinical problem but are settings ‘which [are] chiefly subscribed by people who have many things wrong’; these people are thus deemed to be ‘inappropriate’36 patients. This leads to the treatment of the acute clinical problem for patient admission being prioritised, with the care of their additional dementia not being recognised by clinical staff as a priority. 34 However, older patients commonly have more than one chronic condition in addition to their acute illness,24 and, for people living with dementia, there is a need for specific care practices that bring together care of dementia and their acute clinical problem with the acute care setting and its practices. 37

Within the acute setting, there is often the perception that the person living with dementia is at fault rather than the environment. 38 Clinical staff often hold negative stereotypical attitudes towards older patients in their care,39 with clinical features of dementia such as ‘wandering’ perceived by clinical staff as deviant behaviours. 40,41 This lack of understanding of the needs of people living with dementia often results in this group being labelled ‘difficult’. 42

The care of patients living with dementia within hospitals is not only a welfare issue but also a human rights issue. The House of Lords/House of Commons Joint Committee on Human Rights highlights concerns about poor treatment, neglect, abuse, discrimination3 and the lack of dignity, especially with regard to personal care needs. 3 The Counting the Cost of Care report43 identified persons with dementia being treated with a lack of dignity and respect as a key area of concern. An Alzheimer’s Society survey43 found that more than one-third (36%) of carers said that the person living with dementia was never treated with respect and dignity. A more recent Alzheimer’s Society report stated that 60% of carers reported experiencing a lack of dignity or understanding. 44

The undignified care that people living with dementia experience in hospital does not happen in a vacuum; it is rooted in the wider social discrimination that older people experience within our society. 45 The House of Lords/House of Commons Joint Committee on Human Rights identified the powerful impacts of historic and embedded ageism that have contributed to the systematic failures to respect and protect the human rights of older people within the health-care system. 3 This seems particularly notable in the acute setting, where discrimination in the provision of services and quality of care has been acknowledged as ‘an ageist policy’ (p. 5). 18 The Department of Health and Social Care implementation plan for the older people’s National Service Framework46 singled out the deep-rooted negative attitudes and behaviours towards older people in acute care.

This stigmatisation of older people in hospitals is particularly acute for people living with dementia. 47 The poor recognition and systematic undertreatment of post-operative pain among people living with dementia:48,49

illustrates the discriminatory care provided for people with dementia.

p. 62. 12

© Crown copyright. Contains public sector information licensed under the Open Government Licence v3.0

In addition, people living with dementia from black, Asian and minority ethnic communities are more likely to experience stigma and poor care,50 and women with dementia are exposed to a ‘triple jeopardy’51,52 of discrimination because of their age, sex and condition.

There is evidence to suggest that clinical staff routinely believe that people living with dementia do not belong within the acute setting, which is seen as an inappropriate place for older people more generally to receive care. 24 However, there was little exploration by staff of where a more appropriate place for older people would be. Tadd et al. 24 suggest that this may indicate an underlying ageism, but it can also be seen as an example of ‘specialism hegemony’: acute staff tend to dismiss conditions they see as belonging to the specialism of mental health.

A report by The King’s Fund25 highlights that ‘the work of caring for older patients with complex needs is hard and testing – physically, psychologically and emotionally’. 25 This work, carried out by nurses and health-care assistants (HCAs), is often described as ‘basic’ rather than ‘essential’ care, implying that it is unskilled work, and the report acknowledges the physical and emotional impact of this work:

In hospital, patients with dementia and delirium may disturb other patients, or may be challenging and difficult to keep calm and safe . . . changing a doubly incontinent patient and remaking the whole bed requires the effort of two people. Sometimes patients are unresponsive, and beds need to be changed again almost immediately.

Cornwell J. The Care of Frail Older People with Complex Needs: Time for a Revolution. London: The King’s Fund; 2012 (p. 4). 25 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 licence, which permits others to share the work, provided the original work is properly cited. See https://creativecommons.org/licenses/by-nc-nd/4.0/

This presents a significant challenge to the NHS,20 and new approaches are needed to improve what happens to older people, and, more specifically, people living with dementia, when they are admitted to hospital. Consumer groups have lobbied to improve the experiences of elderly patients,9,22 a need also recognised by several government enquiries. 53,54 However, the research agenda has lagged behind20 and there is an evidence vacuum in understanding the experience of elderly patients with dementia and how their care can be improved within the acute setting.

Dementia and hospitalisation

The acute hospital setting has become a key site of care for people living with dementia. The Department of Health and Social Care recognises that as many as one in four acute hospital beds in the UK will be occupied by a person living with dementia at any given time,33,43 the equivalent of 3.2 million bed-days per year. In some areas, these figures may be underestimates, with some hospitals reporting that up to 50% of their acute admissions may also have dementia. 43 A systematic review has identified that, internationally, the prevalence estimates for people living with dementia within acute hospitals range between 12.9% and 63.0% across studies. 55 Dementia as a condition is often thought of as something cared for first in the community, then later in specialist settings and long-term care, but the prominence of people living with dementia in the acute hospital setting and the impact that admissions have on these patients cannot be ignored.

A diagnosis of dementia is associated with an increased risk of hospitalisation. 56 Russ et al. 57 estimate that approximately 6% of all people living with dementia in the UK are inpatients in acute hospitals at any one time, in comparison to only 0.6% of people aged > 65 years without a diagnosis of dementia. A hip fracture,35,58 urinary tract infection,59 pneumonia59 and nutritional disorders35 are among the common causes of admission among people living with dementia. A national review of case notes (n = 7987) for people living with dementia found that the majority were admitted within care of the elderly (40%), general medical (25%) and orthopaedic (11%) acute care wards. 20

Although prevalence rates differ by hospital and are dependent on their specific population, current estimates of prevalence rates are believed to be low owing to under-reporting or late diagnosis of this population. 6 Estimates suggest that, within the acute setting, approximately 50% of those affected by dementia do not yet have a formal diagnosis in their medical records. 57,59,60 There are a range of potential reasons for this potential underdiagnosis61 or delayed diagnosis of dementia, with much of this attributable to clinical teams not having the appropriate expertise. 61 A significant number of people living with dementia may have their first assessment when they are admitted with an acute condition. 58 For example, among older people admitted to hospital following a hip fracture, of those living with dementia (40%), just over one-quarter (27%) received their diagnosis during their admission. 58

However, there are also likely to be many older people within the acute setting who may appear to have features of dementia, but who may have different underlying causes of cognitive decline, including delirium or subsyndromal delirium. One screening study of a large cohort of older patients following an unplanned admission to an acute hospital setting such as a medical assessment unit (MAU) found a high prevalence of delirium (15.5%), and a high rate of undiagnosed (72%) delirium among this population. 62 Other studies have identified similarly high levels of comorbid mental health in this group. 60 Medication and comorbid chronic conditions such as diabetes mellitus can also have an impact on cognitive function57 within the acute setting.

Impact on patient outcomes

People living with dementia are a highly vulnerable group within the hospital setting, and, following an acute admission, the functional abilities of this population can deteriorate significantly. 60 A systematic review has identified that people living with dementia in the acute hospital are older, require more hours of nursing care, have longer admissions, are at higher risk of delayed discharge and are likely to experience functional decline during their admission. 55 Similarly, a screening study of emergency admissions of patients aged > 70 years with cognitive decline showed that the majority were at risk of malnutrition (80%) and just under half (47%) were classified as incontinent and as needing help at mealtimes (49%). 56 An acute hospital admission for people living with dementia is associated with an increased risk of functional decline63,64 and with experiencing adverse events. 63,65 A longitudinal cohort study of acute emergency admissions found that patients aged > 70 years with cognitive impairment had markedly higher short-term mortality, with 24% dying during admission,59 and this group of patients being more likely to be readmitted and also more likely to die post admission than patients without cognitive impairment admitted for the same reason. 5 Similarly, a prospective cohort study found that people living with dementia who have an unplanned acute hospital admission had half the survival time of similar inpatients without a diagnosis of dementia. 66

The poor recognition and systematic undertreatment of pain among people living with dementia48,49 illustrates the inequity in care experienced by people living with dementia in the acute setting. 12 People living with dementia receive poor end-of-life care, fewer palliative medications67 and only one-third the level of painkilling opioid medication48 compared with patients without a diagnosis of dementia, with one study concluding that the majority of people living with dementia were in severe pain post operatively. 48,68 People living with dementia find it difficult to articulate their pain,49 and pain is poorly identified and undertreated in people living with dementia. 68

People living with dementia in the acute setting are also at risk of ‘cascade iatrogenesis’, whereby the treatment or intervention used to treat the initial acute admitting condition leads to an unintended sequence of multiple medical complications and a cascade of decline in the person,69 which can result in further dependency, institutionalisation and, potentially, death during their acute admission. 70 Adverse events are common for people living with dementia during an acute admission and, importantly, these are typically associated with identifiable risk factors, with staff failing to understand and recognise the impact of a diagnosis of dementia on patient outcomes. 65

George et al. 70 in their review of challenges and solutions to the care of people living with dementia in the acute setting concluded that there is a need to increase awareness and understandings of the way health-care-related harm manifests in patients admitted to hospital with a diagnosis of dementia, given that they can often present as common geriatric syndromes. The subtlety of adverse events in patients with dementia is said to be overlooked and compounded by a culture of low expectation running counter to a dominant safety culture. 70

Examining culture and organisation

The House of Lords concluded that ‘an entire culture change is needed’3 to improve the care people living with dementia receive in hospital, stating:

We believe that this question of the culture in a hospital is absolutely crucial.

p. 44. 3

© Parliamentary Copyright. Contains Parliamentary information licensed under the Open Parliament Licence v3.0

Research highlights the importance of the culture of care in influencing front-line delivery. Meta-ethnography findings71 reflect that, in spite of acute care nurses’ aspirations to deliver a high standard of psychosocial care, standards of care are largely dependent on ward-level social and organisational conditions. National Institute for Health Research (NIHR) studies of acute settings report a strong association between ward climate and quality of care. Patterson et al. 72 found that more positive patient and carer assessments of acute care for older patients was correlated with higher staff ratings of team climate in terms of ‘supporting each other’, highlighting the vital role of the ward manager in shaping a positive climate for care. Such findings were mirrored in a second Service Delivery Organisation (SDO) study73 highlighting that the experiences of working in wards directly affects patient experiences of care and that complex or demanding patients received less personalised care from staff. 73

Ward nurses play a key part in facilitating the flexibility needed for hospital systems and processes to function effectively. 74 SDO studies have examined the role and impact of allied health professionals (AHPs)75 and Advanced Nursing Practitioners76 in ward leadership. Studies highlight the importance of focusing on the relationships between nurses and less privileged groups, which influences how care is organised, supervised and delivered,77 given the increasing delegation of ‘hands-on’ care to HCAs. 77

An ethnography of HCAs identified that they are typically a close-knit, marginalised group with a strong identity, which could negatively affect the organisation of wards; this has implications for multidisciplinary teamworking and the provision of care. 78 The impact of different shifts on work goals and priorities,79 a culture of reactivity that responds to immediate problems,80 ward environments characterised as austere and as emphasising routines, plus the lack of communication opportunities, combined to restrict patients and staff, thus contributing to moral distress and burnout among ward nurses. 81

Ethnography has also examined the ways in which information about patients and the delivery of care is defined and communicated within wards, with nurses often using ‘scraps’ as a form of invisible documentation that they had ownership of. 82 In contrast, the formal nursing record had limited relevance to informing practice and supporting the delivery of care,74 with ward structures further hindering the development of nurses’ practice. 83

A systematic review of qualitative studies highlighted the importance of the relational work in the quality of care that patients receive within acute wards84 and the significance of the nursing role in identifying and promoting dignity for older people living with dementia. 85 Meta-ethnography concludes that service improvements need to focus on how to optimise the organisational conditions that support nurses in their relational work with patients71 and that take into account the challenges of implementing innovations and embedding change. 86

National Institute for Health Research ethnographic studies have examined the complexity of hospital discharge87 and transitions in care,88 with a Research for Patient Benefit ethnographic study of hospital discharge identifying the complexity of judgements about a person’s capacity, best interests and decision-making for people living with dementia,89 emphasising the importance of communication and collaboration during transitions in care. 88

Ethnographic studies

There are a small number of ethnographic studies exploring the experiences and care of older people in acute settings,90–96 which have a focus on dignity in older patients24 with and without97 a diagnosis of dementia. Ethnographies have explored dementia in the context of long-term care settings,98–101 with only a small number of studies examining people living with dementia in acute settings;97,102,103 importantly, both sets of studies highlight the failure of acute settings to respond to the needs of this patient population.

Norman’s observation of wards within a general hospital found that people living with dementia were viewed by the health-care professionals caring for them as belonging to one of two groups, either ‘positive and acceptable patients’ or ‘negative and unacceptable patients’. 102 Similarly, the Tadd et al. 24 ethnography of four acute care settings identified ageist attitudes among ward staff. In examining the occupational, organisational and cultural factors that affect care, the authors found a failure to provide dignified care and assessed the impact of this on both the quality of care and patient outcomes. 24 Both Tadd et al. 24 and The et al. 96 found that staff believed that the acute hospital was not an appropriate setting for older people to receive care and also found that this patient population was perpetually moved both between and within hospital wards.

Do we know anything about best practice in acute settings?

Few studies have explored what constitutes ‘good care’ for people living with dementia within the acute setting, with studies developing models of care for this patient group typically assuming that the principles of good care that have been integrated into long-term facilities can transfer unproblematically into other settings. 42 Only a small number of studies have evaluated proposed models of care to improve patient outcomes within acute settings for dementia; however, the focus of these interventions has been to address ‘chronic confusion’104,105 rather than dementia. There is little evidence that training interventions to enhance health-care workers’ expertise and capacity in working specifically with patients with dementia are effective. Recent reviews42,106 suggest that, prior to developing interventions, further detailed research to understand the role and needs of health-care workers with this patient population106 and an exploration of what constitutes ‘good care’ within the acute setting42 are needed.

In their review identifying best practice for caring for this patient population within the acute setting, Moyle et al. 42 identified a number of key features: the early identification of dementia and assessment of the patient, the knowledge and attitudes of nursing staff, a multidisciplinary approach to care, providing focused communication, a reduction in stressors, and carer and family involvement. Staff within an Australian hospital reported that the main constraints to being able to provide best practice were environmental, sociocultural and economic, concluding that the acute setting was not appropriate for this patient group. 107

Conceptualising refusal and resistance to care

Refusal and resistance to care by people living with dementia is not a clearly defined concept within the wider research literature. The terms ‘resistance’ and ‘refusal’ are used interchangeably, both clinically and in the literature, to cover both the range and extent of this behaviour or phenomenon. 108 Other terms used in the literature include ‘non-compliant behaviour’,109 ‘challenging behaviours’,110 ‘behaviours that challenge’111 and ‘agitated’112 and ‘aggressive’113 behaviours. At their core, refusal and resistance to care remain something of umbrella terms and cover behaviour(s) characterised as non-compliant in response to health care. 109

Within the research, examining what Ishii et al. 108 conceptualise in their review under the umbrella term ‘rejection of care’, there is a significant focus on long-term settings, including community care, home care, care home and nursing home settings, psychiatric wards, and specialist mental health settings. Within these settings there has been sufficient research to support systematic reviews that examine how dementia and associated behaviours, including rejection of care, affect friendship,114 marriage,115 caregivers,116–119 the person as patient,119,120 self and identity,121 quality of life,122 costs of care,123–125 access to care,55,126 care staff resilience,106 care staff stress,127 care staff competence,128 care staff communication skills129,130 and care staff retention. 131 The impact of behaviours relating to a dementia diagnosis, including rejection of care, has been explored extensively within long-term and specialist settings.

Locating refusal and resistance to care in the acute setting

Thus, although there is considerable evidence exploring the long-term settings in which one would expect to find people living with dementia being cared for, research examining the impact of other settings in which people living with dementia are also found is scarce. A notable setting in which there is little research is the acute care setting, in which people living with dementia form a significant population. Acute care nurses in the UK have reported that they are ‘always’ responsible for caring for at least one patient with a diagnosis of dementia132 and almost all find such care a ‘challenging’ part of their work. 22 ‘Rejection of care’ is a recognised feature of these admissions,24,42,132,133 yet, unlike in other long-term and specialist settings, few studies have examined the prevalence, causes, experiences and impacts of this phenomenon within the acute ward.

This disparity within the research literature is reflected at the organisational level of care delivery. Despite the number of admissions of people living with dementia to the acute setting, only 36% of hospitals have a fully developed dementia care pathway in place. 20 A large number of reports highlight the need for specialist dementia training for all clinical staff within the acute setting,18,19,31,58,134 but health-care professionals working on acute wards continue to lack the necessary skills, competencies and knowledge to care for this group of patients. 12,23,24

Within the hospital setting, refusal of care typically refers to non-compliance towards staff-led provision of care, which ranges from essential treatments to the more everyday, mundane but still vital elements of care, including, but not limited to, food, hydration, medication, personal care and toileting. 40 This characterisation and conceptualisation of refusal and resistance as non-compliance means that it is commonly perceived by clinical staff as deviant behaviour, leading to patients admitted with a diagnosis of dementia being labelled, and conceived of by acute care staff, as a ‘difficult’ patient group to care for. 42 Although the literature examining the refusal of care of people living with dementia in the acute setting is scarcer than for other settings, there is still a small body of work that raises crucial issues for consideration.

Importantly, this literature emphasises the importance of recognising both that people living with dementia are to be found and belong within acute care wards and that, in their current form, acute care wards are an inappropriate place for people living with dementia to be. 42,132 Although an admission to an acute ward is associated with poor care outcomes for people living with dementia,48,49,56,59,60,135 the reviews by Moyle et al. 42 and Dewing and Dijk132 suggest a major factor for this could be the acute ward itself. The built environment of acute care is unfamiliar, frustrating and threatening for people living with dementia. Acute wards are busy, with a constantly changing number of unfamiliar people doing unfamiliar things against a backdrop of unfamiliar and irregular noises from alarms and machines, with artificial bright lighting, in a space that is disorientating. All of these factors can be disorientating, distressing or agitating for the admitted patient with a diagnosis of dementia and may lead to behaviours that Ishii et al. 108 conceptualise as ‘rejection of care’. 136

This literature confirms that people living with dementia can be found in the acute setting, are likely to refuse care while accommodated within it, and that this both poses challenges for ward staff and has negative outcomes for patients. However, what is unusual is for such well-documented phenomena to remain so under-researched. Within the acute setting, refusal of care is often considered in detail within the literature only when it is associated with another well-researched phenomenon. The refusal of food, for instance, is discussed in the literature but is typically examined as one feature of other phenomena, such as the ethics of artificial feeding137–139 or the efficacy of food supplements for patients presenting with reduced nutritional intake. 139 Refusal of treatments and medications may also seem, at a first glance, to be well examined, but too often terms such as rejection, refusal and resistance are used as synonyms for non-compliance or non-adherence with a long-term or new drug regime. Rather than exploring the behavioural aspects of the rejection of medication, this literature is concerned with an examination of the potential biomedical and technical solutions, such as the assessment of the efficacy of non-oral administration. 140,141 At present, although the gaps in our understanding of the behavioural features of refusal and resistance to essential everyday care in the acute setting have been remarked on,138,142 little has been done to address them.

With little research evidence or best practice that can be implemented in the acute care ward, key responses to refusal of care are its management through restraint, sedation or Deprivation of Liberty Safeguard (DoLS) applications. 136,143,144 A retrospective review of DoLS applications (in the UK, this is an application of ‘deprivation of liberty’ to protect a vulnerable person who is judged to lack the capacity to consent to the care or treatment they need) identified that 22.5% of all DoLS applications were for patients with a dementia diagnosis and that 23.6% of all applications were directly related to refusal of care, with the person attempting to leave their ward a key trigger for a DoLS application for a further 21.3%. 143 However, the overall evidence base for why restraint is used, and in what circumstances, is limited, emerging briefly as a discussion point within a small number of quantitative studies piloting therapeutic interventions,136,144 with few studies examining the use of restraint in response to resistance to care.

A small body of research has identified that these techniques and approaches in response to resistance continue and are embedded within hospital nursing culture,111,136 despite recognition that they are undesirable and contribute to missed diagnosis, poor care experiences, poor outcomes and functional decline in the person. 111,133,136,144–146 However, although the evidence around the embedded use of restraint, and the lack of alternative methods of recognising and managing the refusal of everyday care, is consistent, it is limited. It includes a single case study,111 limited case studies (n = 4) derived from larger randomised controlled trials,145 small-n quantitative studies (n = 36)133 or studies focused on topics tangential to rejection of care (barriers to post-operative pain-management). 146

Although this evidence is consistent with studies that suggest that people living with dementia respond more positively to autonomous cultures of care than to restrictive ones,85 there remains a significant gap in the literature that provides systematic evidence of the extent of refusal and resistance to care, how it manifests, the ways in which ward staff respond, and whether or not cultures of restraint continue to exist in acute settings in response to rejection of care behaviours, and, if so, why. The limited body of literature examining the rejection of care in the acute hospital setting acknowledges this, highlighting the scarcity of high-quality studies examining this phenomenon within the acute environment. 133,136,144,146,147

Refusal and resistance towards food

What research we do have is useful for examining the ways in which resistance to everyday care is currently conceptualised and managed. Archibald148 suggests that resistance towards or refusal of food is one of the most common manifestations of rejection of care across all settings. Archibald148 argues that although it is known that as many as half of people living with dementia in long-term settings regularly refuse food, there are few data to illustrate how common a phenomenon this is in a short-term acute hospital setting, although it is likely to be significantly higher. The lack of data is arguably symptomatic of how a behaviour or response can become a mundane and even accepted behaviour during an admission, as demonstrated by the scarcity of research around it. 147 Importantly, it means that many of the recommendations from the literature to improve or manage the rejection of food are drawn from data and research within long-term settings.

The recommendations of this literature in managing the refusal of food are relatively straightforward. Many of them could be described as ‘common-sense’. Examples include serving only easily manipulated and easily swallowed food to people living with dementia,148,149 flexibility in mealtimes (both in terms of when they are served and their duration) and flexibility around sleeping routines and daytime rhythms,150 serving meals in preferred or recognisable dining settings and using recognisable plates and cutlery,151 and spending time familiarising patients with both their meal and the associated technologies. 142 The key theme running through this literature is of making meals as normal as possible, making them social and triggering memories of an archetypal ‘normal’ meal, sitting around a dining table at home with family. However, there is little within these recommendations to address the extraordinary environment of the acute ward. In addition, these recommendations also remove the agency of the person living with dementia, suggesting that if the person responsible for care serves the meal in the correct way, then refusal of care can be managed. That the person living with dementia may have agency or rationality in their refusal, beyond unfamiliarity with the meal or their surroundings, is completely overlooked.

This literature providing recommendations that focus on the presentation of the meal rather than the needs of the person is reflected within acute wards. The potential underlying causes of resistance to food are rarely investigated by nursing staff. 147 Byron et al. 138 criticise a lack of any primary research examining nurse decision-making around mealtimes and food-related issues in the hospital setting. However, it has been shown that ward staff are rarely trained to recognise or investigate factors underlying the rejection of food, instead viewing rejection as a feature of the dementia diagnosis. 137 As a result, the consistent or consecutive refusal of meals is likely to be responded to by altering the meal or method of intake, for example by giving pureed meals, thickened fluids, nutritional drinks or intravenous (i.v.) drips or, in extreme cases, by undertaking artificial feeding. 137,138,149,151 The person refusing food may be labelled as difficult,42 be infantilised by their rejection of the meal or be seen as ungrateful by staff. 148 The literature suggests that it is rarer for the refusal behaviour to be perceived as a responsive behaviour expressing something about the person and their feelings in that moment to their condition, their comfort or the environment and surroundings, which would be in line with current recommendations. 32

Refusal and resistance towards medication

The literature examining medication is clearer in identifying that the underlying causes of resistance, refusal and rejection need to be investigated rather than simply responded and reacted to. While acknowledging the difficulties in doing so, the literature suggests that staff must take on the role of a detective to identify the underlying cause at the root of a patient’s behaviour,111 which, it is suggested, may often be pain or discomfort that the patient is unaware of or is unable to communicate. 146 However, nurses working in the acute setting may lack the training to do this kind of detective work111 or believe that such a role is for specialist pain management staff146 rather than for general acute nurses. Rantala et al. 146 argue that the current lack of investigation of the potential underlying causes of refusal and resistance can be illustrated by the significantly lower prescription rates for post-operative pain-killing medication among people living with dementia than among the general acute patient population. Such findings are repeated in other studies152,153 and are associated with poor patient outcomes and significantly higher post-operative mortality rates.

At the same time, research focusing on improving how staff care for patients refusing medication in the acute setting is limited, particularly in comparison to studies that focus on food refusal. There is a small body of translatable behavioural work from long-term care settings,149–151,154 and work examining the use of alternative methods of medication delivery. However, this is often concerned with longer-term adherence to a drug regime, rather than specific interventions,140,155 or is instead concerned with the covert delivery of medication and the ethics thereof. 156,157

In contrast, the limited research carried out within the acute setting identifies more challenges than it answers. Rantala et al. 146 describe the ways in which aggression can escalate from poor pain management, leading to cycles of aggression that, in turn, create further barriers to future treatment and pain management itself once recognised. Such cycles are often treated with sedation, which calms the observable aggression but not the less visible pain. Similar issues have been observed when clinical investigation fails to differentiate the features of dementia and related symptoms of agitation from a potential diagnosis of delirium and/or confusion superimposed on to the patient’s dementia. 158

There is some research addressing resistance to medication that is not associated with underlying pain. Abetz et al. 141 discuss how factors including suspicion of new medication, suspicion of unfamiliar people bringing medications and fear of unpleasant side effects can contribute to or trigger resistive behaviour towards medication. However, these findings come from a body of pharmaceutical research promoting the use of transdermal medication delivery in place of oral medication for people living with dementia. Reviews of such delivery methods suggest some reduction in medications refusal,140 although these findings do not factor in the suitability or accessibility of these methods within an acute ward environment.

Implications for our research strategy

This small body of literature highlights that there is a clear need for systematic research exploring the ways in which refusal and resistance manifests in the acute care setting, and what can be done to improve care in response to this to enhance both care experiences and outcomes.

To improve care we must first address the tension that lies in the clear articulation of what we mean and understand by refusal and resistance when observing it. In other words, how do we know it when we see it? Thus, a key aspect of our examination and analysis of resistance, refusal and rejection of care focuses on the ways in which ward staff (nurses and HCAs) interpret, recognise, assess, classify and subsequently respond to these behaviours. Ishii et al. 108 give a very helpful definition of ‘rejection of care’, namely to ‘reject evaluation or care [e.g. bloodwork, taking medications, ADL (activities of daily living) assistance] that is necessary to achieve the patient’s goals for health and well-being’. 108 Importantly, the authors go on to state that this definition requires a patient’s intent to refuse or reject care and does not constitute a rejection if the patient’s somnolence renders them unable to swallow medication. This definition also excludes aggressive behaviours. However, as this report shows, a person being non-responsive or aggressive in response to care delivery was a key feature of how ward staff practically recognised and understood dementia, with ‘refusal of care’ interpreted within the ward as an expected feature or outcome of the behavioural or psychological symptoms of dementia during an admission.

In the course of this analysis, the concept of resistance is treated as emergent and ‘in the making’, which surfaced in the ways staff interpreted and responded to behaviours during the mundane encounters, interactions and the everyday routine work of the ward. The actual act of refusal, resistance or rejection is always context bound and involves many factors (which we have tried to capture in our analysis) and should not be seen as an isolated element within a wider repertoire of dementia-related behaviours. Thus, rather than classifying these behaviours as distinct entities, the term ‘resistance’ is used and the ways in which this manifests as part of a continuum of responses that people living with dementia have to the ways in which their care is delivered at the bedside and their wider experiences of their admission to an acute hospital ward are assessed. In turn, these responses are viewed by staff as problematic and difficult, are believed to signify a lack of capacity and, because they do not ‘fit’ the organisation and timetabled routines of the ward, are viewed as something that must be overcome or managed.

Data collection and analysis

Data collection (observations and interviews) and analysis has been informed by the analytic tradition of grounded theory,175 a practical and flexible approach for ethnographic research. 176 It uses the constant comparative method and theoretical sampling whereby data collection (observation and interview data) and analysis are inter-related175,177 and carried out concurrently. 178,179 The flexible nature of this approach is important because it has allowed us to increase the ‘analytic incisiveness’176 of the study; as data were collected at one site, preliminary analysis of the data was able to proceed in parallel, with this preliminary analysis informing the focus of later stages of data collection and analysis.

Although these traditions have developed independently, they are complementary. Grounded theory strengthens the ethnographic aims of achieving a theoretical interpretation of the data, and the ethnographic approach prevents grounded theory from being applied in a mechanistic and rigid way. 176 A common concern with an ethnographic approach is that it can treat everything within a setting as data, which can lead to the ethnographer collecting large numbers of unconnected data and producing a heavily descriptive analysis. 180 This approach provides a middle ground in which the ethnographer, often seen as a passive observer of the social world, can use grounded theory to provide a systematic approach to data collection that can be used to develop theory to address the interpretive realities of the range of actors within this setting. 176

Ethnographic fieldwork

This ethnography was carried out in 10 wards within five hospitals across England and Wales, which were purposefully selected to represent a range of hospital types, geographies and socioeconomic catchments. Across these sites, 155 days of observational ethnographic fieldwork were carried out in areas of acute hospitals known to admit large numbers of people with dementia for acute conditions: T&O wards and MAUs or variants thereof. Approximately 600,000 words of observational fieldnotes were collected, written up, transcribed, cleaned and anonymised by the ethnographers (KF and AN). To provide a detailed contextual analysis of the events observed, the expertise involved and the wider conditions of patient care, we also carried out ethnographic interviews (during observation) with ward staff (n = 414). Ethnographic interviews (n = 71) were also conducted with case-study participants (n = 10) and their family members (n = 37), with care being observed at the bedside throughout a patient’s admission. Given the scope of our data set, in this report we focus on presenting our analysis of the observational fieldwork.

Multisite ethnography defines the object of study via a number of techniques or tracking strategies, and within the fieldwork we recognise the importance of focusing on the ‘busy intersections’182 and of seeking out sites of tension at which a large number of interests and identities are expressed. It is argued that it is at these points that identity and culture become articulated, enacted and constructed. The aim of this study was to provide a detailed understanding of the clinical and interactional work and processes that influence nursing, HCA and other clinical staffs’ [this included mealtime assistants, specialist registrars (SPRs), consultants, AHPs and staff with managerial responsibilities] responses to ‘refusal of care’ with a focus on everyday care, medication rounds and mealtimes. We studied actions and accounts within their natural everyday settings to explore how individuals, teams, wards and hospitals respond to and manage refusal and what influences these approaches. Our focused observational strategy within each setting did the following:

• Concentrated on the work of nurses and HCAs and other clinical staff from a range of disciplines and roles when they are involved in the care of people living with dementia, focusing on medication rounds and mealtimes. We mapped the organisation of care, responses to refusal and management and communication of refusal of care with wards.

• Followed nurses and HCAs within each ward setting to explore their everyday work, and what informs this work. The processes of decision-making, the management of uncertainty and treatment procedures in response to refusal of care were identified by examining the everyday routine behaviours of individuals and teams within wards.

• Focused on observing handovers, admissions and conversations with carers, all of which present opportunities for sharing information about refusal behaviours and how these might best be managed.

• Where possible, collected routine data (from ward managers and within patient records) about ward staffing levels, overall work allocation, bed occupancy, patient acuity, turnover and recorded levels of refusal and resistance at the time of fieldwork to provide context and an understanding of the workload of refusal and resistance within this setting.

This has provided a detailed understanding of organisational and care processes that have an impact on the responses to and the management of refusal and resistance within this patient group. We examined the everyday work of staff, their practices and the interactions among staff, and between staff and patients/carers within these wards.

Ethnographic interviews with patients who refuse care and their carers

We carried out ethnographic interviews to explore the impact of refusal of care on patient and carer experience and the factors that could lead to improved care and support. However, ward staff and the NHS Research Ethics Committee (REC) were concerned that the consent process for interviews was too demanding for patients living with dementia within wards and, in addition, patients who refused within the wards could not be identified as a distinct group (as we had assumed from the available literature) within the wider bay or ward populations. In response, recruitment was informed by the importance of recognising issues of capacity to consent and the impact of an individual’s acute admitting condition on their ability to participate in an interview. To comply with our NHS REC approvals, we obtained written consent for interviews and observation at the bedside only when approved and supported by ward staff and patients’ families. When the patient instigated a conversation with us, we obtained verbal but not written consent and recorded only the broad themes of the conversation within fieldnotes rather than verbatim quotations.

Where possible, we carried out ethnographic (during observation) interviews with patients and their carers within the ward setting to explore the experiences and needs of this patient population and issues of refusal from their perspectives. Although we originally proposed to interview patients resisting care, this was ultimately not possible as we underestimated the emotional duress and agitation linked to such experiences. Where possible, we spoke to people living with dementia following a period of resistance, but the prior resistance was not discussed with them, in order to minimise further agitation.

Thus, in response to these limitations, in consultation with ward staff and the NHS REC, we used our case studies to obtain the perspectives of people living with dementia and their families and carers via ethnographic interviews. Ethnographic interviews were carried out with the case study participants throughout their admission, as well as with their family members. Ethnographic interviews (n = 71) were conducted with case study participants (n = 10) and their family members (n = 37) to explore:

• experiences of admission and care, the impact of the physical environment and wider hospital structures

• processes of being listened to, communication and decision-making

• fears and concerns about treatment and management, particularly around refusal and resistive behaviours

• the identification of refusal of care and involvement and recognition of carer expertise.

Ethnographic interviews with nurses and health-care assistants

The aim of ethnographic interviews was to provide a detailed understanding of the influences on health-care professionals’ responses to ‘refusal of care’. Ethnographic (during observation) interviews were carried out with nursing staff, HCAs and clinical staff (n = 436) from a range of disciplines (including mealtime assistants, SPRs, consultants, AHPs and staff with managerial responsibilities) as they cared for this patient group within each ward and focus on medication rounds and mealtimes. This allowed us to question what staff were doing and why. The following key questions were asked:

• What is the articulation work within those settings? How do staff account for and make sense of their actions?

• What is the experience of and training for staff working with people living with dementia and refusal of care and what informs their practice?

• What aspects of caring are defined as ‘difficult’, demanding or rewarding and what is the level of staff confidence in their competence of working with this patient group? What are the barriers to and enablers of supporting this patient group?

• What is the recognition and reward for providing care to this patient group from patients, relatives, colleagues and managers?

Case studies

Across the hospital sites we carried out detailed case studies (n = 10) of patients living with dementia who were identified as refusing or resisting food or medicines (n = 5). This allowed us to extend our fieldwork to provide a detailed contextual analysis of the events, the clinical staff and expertise involved in the events, and the wider conditions of care and their impact over time. This provides an understanding of the broader care systems within the acute setting that affect care and enable a multiperspective analyses. Purposive sampling was used, which was informed by our early analysis of observational data within each setting. Each case study involved:

• A detailed systematic observation of patient care during their admission (114 days of observation), interviews with carers and family members (n = 37) and, where possible, patients (n = 71), to explore the needs of this patient population and issues of refusal (target = 20–40).

• Interviews with nurses, HCAs and other clinical staff (including feeding assistants, SPRs, consultants, AHPs and staff with managerial responsibilities) involved in the care of these patients (see Ethnographic interviews with patients who refuse and their carers) to explore their response to refusal or resistive behaviour with a focus on medication rounds and mealtimes.

This allowed us to follow the impact of the everyday routine care carried out by nurses, HCAs and other clinical staff and the consequences of their response to and management of refusal and resistive behaviour for this group of patients and their carers.

Field notes of observation and near verbatim text have been written up as Microsoft Word 2016 (Microsoft Corporation, Redmond, WA, USA) files183,184 and all audio recordings of observations and interviews (ethnographic and in-depth) have been written up as Word files or transcribed verbatim by a professional transcription service. All site and individual data have been anonymised and organised in accordance with the Data Protection Act 1998185 and the NHS England Data Protection Policy 2014. 186 Storage of the data is managed by the Cardiff University Information Security Framework Program.

Sampling

Sampling in ethnography requires a flexible, pragmatic approach, using a range of variables that may influence any phenomena, as well as what is known based on the available literature. Probability sampling is not appropriate; instead, non-probability sampling, which is not representative of the wider population, was used to provide analytically rather than statistically generalisable findings. 187,188 This is the most appropriate means by which to study organisations or a clearly defined group, and the size of the sample required for this approach is determined by the nature and scope of the study aims. Using this approach, the number of sites and participants in the sample is judged appropriate not on the basis of size but on the quality and appropriateness of the sample and when saturation of data has been achieved. 188

Setting and access of hospitals

This approach emphasises the importance of comparisons across sites,189 allowing for and optimising the generalisability of findings190 and enhancing the ability for the findings to affect policy and practice. 160 Hospital settings are well suited to an ethnographic approach. At first glance, hospitals may appear to operate in similar ways; however, they often have their own unique culture, which is informed by local dominant cultures and belief systems, which in turn means that care and decision-making can vary widely within institutions. 191,192 Thus, we identified a range of variables that may influence the phenomenon of refusal using purposive and maximum variation sampling to include five hospitals that represent different hospital types, geographical location, expertise, interventions and quality. The five acute hospital settings were identified from across the UK to represent the:

• types of acute hospital [two large university teaching hospitals (Sites B and D), two medium sized general hospitals (Sites A and C) and one smaller general hospital (Site E)]

• geographical locations, to include urban (Site A), inner-city (Sites C and D) and a large rural and urban catchment area (Sites B and E), situated across England and Wales

• range of specialist and non-specialist clinical and non-clinical staff, from staff with no formal expertise to dementia specialist workers (Site B).

Sampling of wards for observation

We observed episodes of care involving patients within MAU and T&O wards, which receive a large population of patients who have dementia and who require acute medical attention:

• MAUs – these are where unscheduled admissions arrive for assessment from accident and emergency (A&E), an outpatient department clinic or a general practitioner. Following assessment, patients are discharged, transferred to a specialist centre or admitted to an inpatient bed (see Collins et al. 62). These are high-turnover settings, which aim to discharge or transfer patients within 24 hours. There is no routine within this setting, with staff geared to deal with acute admissions with fast turnaround, with no continuity or personalised care and a chaotic atmosphere. Thus, they are not areas designed for or conducive to patients who have any cognitive deficit or dementia at a critical time where escalation of symptoms may occur. Members of the Carer Steering Group all had poor experiences of this setting and found it a frightening time, when they felt that they were not listened to and were often separated from their partners; we observed 267 patients in this type of setting.

• T&O wards – these settings allowed us to observe routine ward care and those patients who have an unscheduled admission following an accident where there may be no opportunity to provide additional support. The patients within this setting have had an accident or fall that has typically resulted in a fracture; we observed 65 patients in this type of setting.

Sampling and recruitment of staff for observation and interviews

We followed the work of nurses and HCAs. We purposively sampled to ensure that we included the range of clinical grades (clinical support worker nursing, nurse associate, entry-level nurse, nurse specialist, nurse team leader, advanced nurse, nurse team manager, modern matron, nurse consultant) across the ward settings as well as other clinical staffs’ (mealtime assistants, SPRs, consultants, AHPs and staff with managerial responsibilities) responses to ‘refusal of care’. Within each acute setting we worked with our key contact, who was typically the senior nurse responsible for care of the elderly and people living with dementia within each trust, who facilitated the process of identifying and introducing the team to key informants at hospital and ward levels. These key informants also provided introductions to our wards.

Sampling and recruitment of patients and carers for interview and observation

Within wards we focused on capturing the ‘incidents, events, and happenings that denote the work that they do, the conditions that facilitate, interrupt, or prevent their work, the action/interaction by which it is expressed and the consequences that result’. 177 Importantly, the focus of observation was on the daily practice of nurses and HCAs and other clinical staff during mealtimes and medication. It was not possible to predict the numbers of patients and carers within each hospital ward during the fieldwork period; however, we found that a high number of patients in the wards we observed had a diagnosis of dementia (as identified in ward records).

Case studies

We carried out a focused series of individual case studies (n = 10), identified using purposive sampling, which was informed by our early analysis of our initial observational data. We included a range of subjects (maximum variation sampling) who have had particular types of experiences within the setting (critical case sampling). Thus, sampling included patients who represented a range of presenting, diagnostic and prognostic, and, where possible, sociodemographic factors:

• We expected to identify refusal and resistance via the ward team and ward records; however, although we found high rates of refusal in the wards, we found low rates of identification, recognition and recording.

• Men (n = 5) and women (n = 5), aged > 65 years (range 70–99 years) with an unplanned admission (fractured hip, n = 5; fractured wrist, n = 1; high number of falls, n = 2; head injury, n = 1; gout, n = 1) and an accompanying comorbid diagnosis of dementia (dementia, n = 6; ‘unspecified dementia’, n = 1; Alzheimer’s disease, n = 1; vascular dementia, n = 2) formally recorded in their medical records were identified.

• Expected length of stay can be highly variable for this group, ranging from days to weeks and months; thus, we followed these patients for up to 6 weeks within each hospital setting.

• Where possible, we carried out follow-up interviews with individuals and families after discharge. Of these patients, two died during their admission, six were medically fit and waiting for a placement in a nursing home and two were medically fit and waiting for a placement in a care home.

Ethics approvals

The REC approval for the study was granted by the NHS Research Ethics Service via the Wales REC 3 on 24 June 2015 (reference number: 15/WA/0191). Substantial amendments to the study protocol were approved at a meeting of the Wales REC 3 committee on 10 December 2015. The committee has approved this research project for the purposes of the Mental Capacity Act 2005193 and confirmed that it meets the requirements of Section 31 of the Act in relation to research carried out as part of this project on, or in relation to, a person who lacks capacity to consent to taking part in the project. The study was accepted by NHS Research Permissions Wales on 16 July 2015, with permission received from the NIHR Coordinated System for gaining NHS Permission and West Midlands Clinical Research Network on 11 March 2016 and from the Health Research Authority on 27 May 2016. Recruitment for the study was managed and recorded through the Central Portfolio Management System and closed on 31 January 2017.

The safety of all participants was a key priority of the research team at each stage of the study. Before undertaking this study, the ethics of observing care, and the ethics of reporting where necessary what was observed, was frequently discussed with staff in the hospital sites and our carers group. In meetings with the NHS REC that approved this study it was clarified that although neither of the researchers technically had a clinical duty of care (both researchers were academics without clinical qualifications or professional affiliations), they would still be bound to safeguard any patient participants observed over the course of the project.

Prior to granting research passports, the host site ensured that both researchers had completed Good Clinical Practice and the Protection of Vulnerable Adults (POVA) training to level 1. The researchers were made aware of safeguarding and whistleblowing procedures at each site and had a named member of staff (the site principal investigator or research nurse on shift) to contact if malpractice or behaviour that put vulnerable patients at risk was observed.

Over the course of the observations, the researchers saw many aspects of everyday practice that would not be considered best practice or in the interests of the individual patient at that time. However, the examples presented within this report were not isolated and formed part of systemic and established everyday routine practice within every ward at each hospital site. Over the course of the observations we did not observe individual malicious behaviour or isolated incidents of deviance that placed a vulnerable adult at risk. Instead, we observed how the everyday organisation and delivery of bedside care itself often places the vulnerable adult at risk, but that this is part of the routine and established culture of the hospitals and the wards within them. At no point did the researchers feel that any individual or ward team was acting in a way that required escalation or whistleblowing. Had the researchers observed behaviour that they felt breached POVA guidelines they would have immediately ceased observations and reported the incident in line with site procedures.

The researchers did, however, frequently intervene to support people living with dementia and their families and carers where they felt it was necessary to protect the comfort of the patient. People living with dementia would frequently tell the researchers that they wanted to go to the bathroom or that they were in pain or share their concerns (about their home, their family or their pets, or how to pay for their care). In response to these disclosures, the researcher (with permission from the patient) would inform ward staff and ensure that this was not forgotten and was attended to by the ward team.

In addition, the researcher was sometimes the only person on a hospital bay with patients and would regularly ask patients if they needed anything. Sometimes when staff were not present or able to be called quickly to a bay, the research team provided immediate support and help, for example if patients were trying to reach for a drink or a newspaper. However, if a patient was trying to get out of bed unaided and if they were in some physical danger or at risk, the researcher would call staff. Although the researchers accept that this may have, on occasion, contaminated the purity of the data, the welfare of those under observation was always their priority.

Between sites, our research was regularly presented to the research team, including nurses, clinicians and trust leads, and although it was agreed that the care observed could be detrimental or distressing to a person living with dementia, it was also routine and recognisable as the everyday practice of acute ward staff. Since collecting the data and completing the analysis, we have carried out a programme of dissemination to present these findings at conferences and symposia to clinicians, advocates and nurses. We have also carried out a series of consultation events with people living with dementia and their carers, where our findings have resonated with participants. This work represents a form of member checking or respondent validation and has reinforced how recognisable the everyday nature of the care observed by the researchers, and the reliability and validity of the data collection and analysis, is.

The data presented here report on patients left to shout in beds or being restricted, restrained and coerced into actions that they had verbally or physically refused. Although in isolation these actions may appear to breach patients’ rights, the researchers hope that they have demonstrated that these are not isolated incidents but rather the everyday reality of care delivery that each person living with dementia will experience during their admission within an acute hospital. We also show that nurses and HCAs lack the support to respond to the needs of people living with dementia in ways other than those outlined in this report, and, because of the fixed timetables and culture of the ward, routinely prioritise nutrition, medication and routine over the comfort or preferences of the patient. The actions taken by nurses and HCAs presented in this report were taken in good faith and attempted to protect the patient and the ward and to respond to the policies and perceived expectations of the wider institution. The researchers hope that the evidence presented in this report highlights both the challenges faced by ward staff as they deliver care in the acute environment and the need to better support both staff and patients living with dementia within this setting.

Resistance: a routine part of everyday care

We treat ‘resistance’ as an emergent concept and we found that it surfaced in the ways staff interpreted and responded to and assessed people’s behaviours within the ward as legitimate or not. Staff were aware that there were patterns of resistance throughout their shifts; however, they were typically not able to recognise the potential triggers or patterns of resistance for individuals within their bay. Resistance was typically seen as a feature of a dementia diagnosis and a behaviour to be expected from this patient group within every shift, as evidenced in the following fieldnote:

HCA discusses the ward: ‘It goes in waves, it’s not bad today, I am watching [bed number]’. As we are talking a nurse comes over and says she will take over the 1-2-1 work and sits in the chair and fills in the patient chart and takes a new ‘side room checks’ document for him.

Site B, day 3

However, despite these routinely observable patterns or ‘waves’ across shifts, staff interventions in response to perceived resistance could be highly variable.

Overall, we identified extremely high levels of resistance to care among people living with dementia within acute hospital wards. It was a common feature of ward life; every patient living with dementia who we observed within the acute hospital ward resisted care during our period of observations. In total, we identified 1052 (T&O, n = 523; MAU, n = 529) incidents or episodes of resistance to care. This could manifest in a number of ways:

• Attempting to get out (and getting out) of bed, standing and walking around. Episodes included the patient wanting to go home; trying to leave the bay, ward or hospital; going to other rooms or areas of the ward or unit; pulling at the bed rail; approaching the nursing station.

• Verbal and physical cues. These included shouting; becoming angry or sounding agitated; crying, screaming and sounding distressed; asking to go home, see family or be taken somewhere; biting, pushing, pulling, spitting at or holding on to staff; refusing to leave the ward, be transferred or be discharged.

• Not submitting to the timetabled rounds of the ward. This included problems in relation to mealtimes, medication, personal care, observation rounds and examinations.

• Removal of equipment. This included pulling or removing i.v. lines, gastric tubes, dressings, catheters, nebulisers, cannulas, oxygen masks and other medical equipment crucial to care; pulling sheets off the bed or removing clothing.

Resistance, in whatever form it manifested, was always identified and interpreted by staff as a feature of the person’s dementia diagnosis that signified their lack of capacity, and, as such, was something to be overcome or managed. Similarly, it was often interpreted by family carers and visitors as ‘bad behaviour’, something to be embarrassed by and to apologise for. However, we identified that resistance was typically a response by people living with dementia to the everyday organisation of their care within the wards (fixed timetabled routines) and the ways in which ward staff delivered care at the bedside. Importantly, these findings were represented in our data within all hospital sites, regardless of ward, time of day, day of week, etc. Our ethnographic approach allowed the researchers (KF and AN) to spend extended time within and across shifts observing specific ward bays and the individual staff and patients within them. These extended periods of observation almost always revealed the potential underlying reasons or triggers for a person’s resistance to aspects of their care or their admission. These were typically rational to that patient’s present ontology and perceptions.

Ward staff responses to resistance were either (1) to ignore resistance, particularly if the behaviour was judged not to be an immediate risk for the patient or to interfere with the organisation of the ward, or (2) to intervene and manage, particularly if staff believed that the provision of care was essential, that there was an issue of patient safety or that resistance was interfering directly with or delaying the organisation, timetables and wider work of the ward. The subsequent patterns of responses and interventions typically involved nurses and HCAs using multiple interactional approaches that combined language (orient to reality, instructions to be obeyed, the necessity of the institution, negotiation and bargaining) and the performative (a focus on completing essential care and containment of the patient at the bedside). These approaches created damaging cycles of stress for patients, families and ward staff.

Physical resistance: getting out of bed, standing and walking

People attempting to get out of their bed or bedside chair and, if that was achieved, to stand and walk, was interpreted by ward staff as a key form of resistance. This was the most overt and commonly observed resistive activity and was consistently high across all wards and sites (although slightly lower in Sites C and D). This response to care was expressed by the most immobile patients, such as those patients who were able only to clutch at or pull and ‘rattle’ the raised rails at the sides of their bed (for this patient group, if they were in bed, the side rails on the bed were typically raised); people sitting in their chairs, who were able to push down on the arms of the chair in an attempt to stand up; and people who were able (often unsteadily) to walk away from the bedside. However, this behaviour was an immediate concern for staff and was identified as particularly problematic only if the person was able to stand, which was likely to lead to the person walking from the bedside and walking unsupervised within, or from, the ward.

As one example, an 86-year-old woman who is living with vascular dementia was admitted following a ‘fall’ and over one afternoon made multiple attempts to try to get out of bed. Every time she tried to leave the chair or the bed the HCA in the bay immediately responded to and repeatedly helped her to sit back in the chair (‘let’s sit you up for lunch’) or led her back to her bed. The team (particularly the HCAs) encouraged her to sleep, arranging the sheets and pillows and holding her hand to try to calm her. However, every time the patient appeared settled, she would immediately try to reach for something, to sit up and get out of bed or stand. Her need to get up can be related to the immediate concerns and anxieties she expressed for her home, where her house keys were and who was picking her up and taking her home. Throughout the course of this afternoon, staff responses to her ranged from distraction (lunch) to repeatedly questioning her ‘what are you doing?’, ‘where do you want to go?’ and ‘who?’, before escalating to giving her clear directives, ‘you need to rest’, and referencing the wider institution by suggesting that other people had the power to decide whether or not she could leave, saying ‘We need to talk to the doctors first’ and ‘your sons will sort it out for you’:

She is wearing her dark red cashmere jumper over her hospital gown and she tries to get out of bed. She has lifted the sheets off her legs and swings round to a sitting up position with her feet on the floor. She has been looking in her handbag and getting out of bed a number of times all morning and I think she was looking for her key and expecting to go home. The HCA immediately goes over to her.

‘Ooh I am out of breath’ – she breathes heavily.

‘It’s lunchtime, let’s sit you up for lunch’. She helps her to move slowly and gently from the bed to the bedside chair and puts lots of pillow behind her and tidies up the bed.

Thank you.

The HCA tucks her into the chair with her mohair blanket over her legs with the trolley long-ways right in front of her, which keeps her in place. The trolley wheels are on the mohair blanket that trails on the floor and she stoops to pull it off – she is perilously close to the edge of the chair because there are lots of pillows behind her that also are pushing her forward. The HCA goes over from her chair at the other end of the bay to help her, puts the blanket on the bed and helps her into bed . . .

‘It’s a bit worrying they are supposed to come for me I only live over the way’ – she points out of the window next to her bed. She looks uncomfortable in bed and looks around for something on the bed and then gets her legs out of bed.

The HCA goes over to her: ‘What are you doing?’ She lifts her legs back up into bed and tucks her back in and moves the trolley next to the bed within reach. As soon as the HCA leaves, she leans over to the end of the bed with some difficulty and gets her mohair blanket and goes to get out of the bed and lifts her legs over the bed and her feet on the floor. The HCA returns: ‘Where do you want to go? Where do you want to go?’ She is kind and gentle in tone, but also sounds slightly exasperated.

She’s picking me up in a minute so she said.

Who?

The patient sits on the bed and shakes her head, she is trying to remember and the HCA leaves and as she turns this patient tries to get up from the bed so the HCA returns and goes to help her to pull her hospital gown down, it has ridden up, and helps her to move from the bed to the chair and sits her down and strokes her face – ‘I am just over there’ – she points to the chair in the corner.

Where are you staying tonight?

I am going home.

‘Someone is going to pick me up, I only live across the way’ – she points outside the window. The HCA holds her hand and strokes her gently and soothingly. ‘I am waiting for someone to take me home.’

A young male HCA arrives and has joined them: ‘Take your medicine, eat well and you can go home!’ The HCA sits with her and a nurse comes over and relieves her while she goes on her break and helps this patient into the bed: ‘You need rest’. She puts the bed flat and helps her to settle down in the bed – she is curled up on her side and she places one pillow behind her back and one in front of her shins and covers her up. She tells her, ‘This is your blanket’, and she makes sure the mohair blanket is over her and she can see. The nurse sits in the chair and holds her hand and strokes it gently.

‘I wonder who has the key?’ The nurse strokes her head gently. The patient sits up: ‘What do you suggest?’

The nurse is next to her: ‘Nothing, you need rest, it’s best to sleep you need rest’ . . . The nurse leaves and the patient then pulls the blanket aside and moves the pillows and swings her legs out of the bed and sits on the side of the bed. The HCA watches her and goes over: ‘Where are you going?’

Someone is going to take me home I hope.

We need to talk to the doctors first.

OK . . . where are the keys for my home?

I am sure your son or daughter has them, you must stop worrying about it, try not to worry, your sons will sort it out for you.

Site E, day 11

Although the bay team talked to the patient throughout the course of this afternoon, as they did this they also continued to subtly restrict her movement to the bed and the bedside. They started with tucking the blanket over her legs and placing the trolley in front of her in the chair and then used pillows around her body, covered her up with the sheets and sat next to her holding her hand.

Getting up, standing, trying to walk, walking or walking unaccompanied within the ward was always discouraged and problematised as a form of resistance for all patients. However, this could be subject to a variety of responses depending on the assessment of the immediate risks to the individual patient. Patients who evidently did not have the strength to follow through and achieve their aim of getting out of bed, standing or walking were typically not assessed by staff as an immediate concern. In the following fieldnotes, a 94-year-old man with a fractured hip and pneumonia continues to rattle and pull at the raised bars on the sides of his bed. However, even though he does this forcefully and over an extended period during the shift, an indication that he appears to be distressed, the team continue their focus on the routines that must be completed and recorded around him – the observation and medication rounds of the people in the other beds within the bay. They do not check on him directly; he is contained safely within the bed:

The HCA returns to the bay with the mobile blood pressure unit from taking the measurements from the patients in the cubicles, the nurse finds the yellow charts for him have returned and are open on the trolley at the end of the bed and updates them. He is a very thin man wearing a hospital gown lying in bed – the sides of the bed are up and he is holding on tight to the left side bar and shaking it and rattling it as if trying to get out. He moans and sounds very distressed. The nurse continues and moves on with the medication round to other patients in the bay and as she is doing this, he holds tightly onto the raised side bar on his bed, continuing to rattle and pull at it.

Site A, day 4

The organisation of care and the routines within the ward meant that, often, staff were not able to see this behaviour as a priority for care, and, if they did, the perceived priorities of the ward meant that they often appeared unable to respond. However, walking or walking unaccompanied within the ward was almost always discouraged and problematised as a form of resistance for all older patients. Staff typically questioned older patients who were walking and the language discussing walking referred to it as a danger or a form of ‘escape’, often using humour to reduce this behaviour:

The HCA wakes a patient up for her observation and medication. As she does this a younger man in his 20s from the bay at the end of the ward walks past, he is just wearing his underpants and is holding a wash bag and pushing his mobile drip heading to the bathroom.

At the same time an 87-year-old woman with a diagnosis of dementia who has been admitted with a fracture is trying to get up from her chair. She is wearing a full-length quilted dressing gown that is buttoned all the way up to her neck (it is the middle of summer and very hot in the ward). It is purple with tiny sprigs of flowers like heather all over it. As she does this, the HCA calls to her: ‘Where are you off to?’. She responds that she is heading to the toilet and the HCA is relieved: ‘Oh I thought you were going to escape!’. She helps her to reach her walking frame and stands behind her as she mutters ‘I know I know I know . . .’ all the way to the bathroom and back. Afterwards she turns to the HCA and smiles and says, ‘Thank you so much’. The HCA gives me a look as the young man from the bay at the end of the ward walks back from the bathroom in just his underpants.

Site A, day 1

Throughout, judgements were also made about the person, their condition and whether or not their particular behaviour was legitimate and able to be permitted within the ward. A man in his twenties was able to walk up and down the ward in his underwear without comment; however, this woman with a diagnosis of dementia was met with close control over her movements.

Verbal resistance: shouting, anger and agitation

It was common for people who were shouting, agitated or angry in the ward or bay to receive the most attention from staff. This was a key response across four of the sites and was low only at Site C. This behaviour often resulted in the individual being assigned one-to-one agency care within the shift (‘specialing’ a patient means that their needs have been assessed as being so great that they require care at all times and they are assigned a private agency HCA to provide one-to-one care). The key responses from ward staff to this behaviour were to (in turn) orient people to where they were, direct them to calm down and rationalise with them. Throughout these processes there were typically multiple attempts to contain the person at the bedside, which could lead to them being supervised more closely, which in turn could cause the patient further anxiety and trigger the ward procedures of ‘specialing’, a DoLS order and one-to-one HCA care. In the following example, a 77-year-old man with a diagnosis of dementia has been in this ward for > 5 months and, although he is ‘medically fit to leave’, there is no package of care organised for him. He is in an individual room and is ‘specialed’ with one-to-one HCA care and, although he has difficulty communicating, he is clearly waiting for his lunch. Shouting ‘HEY’ is his main form of verbal communication during the day and, as he waits, he gets louder. The one-to-one HCA and the team passing join in and they repeatedly switch between encouraging his excitement about the arrival of lunch and using a number of techniques in an attempt to calm him down when lunch does not arrive on schedule. They orient him to where he is, ‘you are in your room’, direct him to calm down and then rationalise with him, ‘what do you want?’:

I am walking through the corridor with two HCAs and the person in the side room shouts for us. The door to his single room is open and he is sitting in the chair looking out wearing hospital pyjamas and red hospital socks and he has a large knitted hat on his head. The one-to-one HCA is sitting very close and opposite him filling in his bedside chart. We go over to see him and he holds up three of his fingers to indicate that there are three of us. The HCA with me asks him what he is having (pasty) and asks him what sauce he would like to go with it and she gets him sachets of sauce to have with his chips: red, brown and vinegar. He is now very excited about the chips. I introduce myself again and we talk about the hospital food and he tells me the food is terrible and then covers his mouth like he shouldn’t say this and we laugh together and sympathise. He gets more and more impatient for lunch – ‘Hey HEY’ and asks for the time when the chips should arrive and the team count down the time – 15 . . . 10 . . . 6 . . .

HEY HEY Give it, give it, give it me.

He is now getting very angry and agitated and repeatedly shouts ‘HEY HEY’ as people pass his door.

What’s the matter now? Listen you need to calm down.

Hey hey.

The HCA continues to talk to him and repeatedly tries to calm him down . . .

HEY HEY HEY.

The HCA jokes and laughs with him and is very positive, ‘Your pasty is coming and ice cream’.

HEY

Are you alright?

NO I AM NOT

‘What do you want?’ Lunch fails to arrive and the team call catering. When it finally arrives the pasty is cold, the chips are soggy and his agitation, unhappiness and frustration increases.

Site D, day 5a

This behaviour suggests that the person was attempting to communicate with ward staff. Importantly, it was also a response to the ways in which care was organised and was being delivered at the bedside, especially (as in this case) if this does not meet the patient’s expectations. In the case above, staff encouraged the patient’s excitement about his lunch only for it to fail to arrive on time and be unappetising and inedible when it did.

Resistance towards the timetabled rounds of the ward

Ward staff identified patients failing to submit to the timetables of the ward as a common way in which resistance to care was manifested. This resistance occurred in relation to mealtimes, medication, personal care, observation rounds and examinations. This study has a specific focus on examining mealtimes and medication rounds and resistance to food, drink and medications. Thus, we examine these aspects of routine care in more detail within the wider context of the timetabled rounds of the ward, which includes mealtimes, food and drink, medication, personal care, observation rounds and examinations.

Removal of equipment

Language and performance: communicating the ‘rules’ of the ward in response to resistance

The ways in which ward staff talked to patients living with dementia in terms of the provision of care at the bedside and in response to perceived resistance were highly repetitive. Staff typically addressed the person by locating them very clearly in relation to the reality of where they were, what had happened to them and what was going on around them. This talk focused on reorienting and locating the person to the rules of the ward and the accepted behaviour within it, and on overcoming perceived resistance to care or managing and containing the person within the context of the routines of the ward. However, people living with dementia did not learn the rules and fit into the timetables of the ward. Importantly, this talk was rhetorical and did not require or expect the patient to respond or assent. The content of this talk did not necessarily reflect what staff actually did, with care and work on the body continuing during these encounters.

The most common approach staff used in response to resistance was to locate the person living with dementia within the institution, ‘you are in hospital’, and within the reality of what had happened to them, ‘you have broken your hip’. Staff also gave very clear instructions to be followed and obeyed, often emphasising the potential imminent danger caused by a patient’s actions; these instructions typically contained a powerful sense of urgency that often displayed staffs’ own underlying anxiety and fears. Staff also negotiated and bargained with patients, offering to leave them alone and stop disturbing them if they co-operated with the request. Appeals to the necessity and expectations of the institution were commonly made, and these appeals emphasised that there was no choice for either the person or the ward team caring for them, ‘we have to change you’. This repetitive talk was directed at reminding the person living with dementia of their place in the world and of the status of ward staff; they must all fit in with the expectations and timetables of the institution.

These exchanges provide ways of uncovering aspects of the loss of identity and social standing of people living with dementia. It is important to note how much of this talk is directed at trying to remind the patient of their place in the hierarchy of the ward and of the rules to be followed. Although staff often gave the appearance of seeking permission and negotiating with the person, the delivery of care typically continued as staff carried out this ‘talk’, with a tacit assumption of assent, focused on obtaining the correct response from the person to allow care, which was already being carried out, to continue.

The moral classifications of resistance

We found that people living with dementia who resisted care, particularly when this was regularly disruptive to the ward routines and timetables, were vulnerable to moral assessment. Their resistance was often interpreted by staff as another feature of who the person was and this could become a key part of their identity in the context of the ward. In addition, some patients were also identified by the ward team as more likely to resist care based on their assessment of the patient’s personality and background.

The emotional and somatic impacts of the organisation and timetables of the ward

Our observations identified that resistance appeared to be a response and reaction to the impact of an admission on individuals. These responses were both emotional and somatic and included difficulties in communicating needs and high levels of anxiety as a result of the unfamiliar environment of the ward, which could lead to disorientation. A key impact of this was the person being viewed within the ward as having increased dependency, which in turn resulted in them losing skills and independence.

Resistance was typically triggered by unfamiliarity with the ward, or the fixed routines, timetables and organisation and delivery of care within an acute ward itself. Loud noises, bright lights and unfamiliar people were distressing for many of the people living with dementia. Because care in these settings is delivered by shift and rota, there was little continuity of care, exacerbating the unfamiliarity of the environment and the wider distress of other patients within it.

As such, resistance can be framed as a rational response to the organisation and delivery of care, with routinised, timetabled and compartmentalised care producing fear and anxiety within a patient but never being acknowledged in the assignment of staff at each shift handover. Our analysis was drawn from observations that allowed the researchers to spend extended time within and across shifts, monitoring specific ward bays and the individual staff and patients within them. This almost always revealed the potential underlying reasons or triggers for patients’ resistance to aspects of care or their admission. These were typically rational to that patient’s present ontology and perceptions.

Ward strategies of care in response to resistance

Overall, the key response to refusal and resistance to care by people living with dementia was one of containment and restraint. Although there was some variation in specific techniques between wards, the overall strategy was always to keep the person living with dementia within their bed or sitting at the bedside. Across all sites, staff expressed high levels of concern and anxiety about people living with dementia attempting to leave or leaving the bed or bedside, and this increased exponentially if they were standing or walking in the bay, the wider ward and corridor or close to the ward entrance. Multiple techniques of containment and restraint by ward teams were observed. Importantly, these approaches to patient care and containment at the bedside were both a response to resistance and also frequently the trigger of resistance or the cause of patient anxiety.

In addition, a key response to ongoing regular patterns of resistance by individuals living with dementia within wards was ‘specialing’ and the assignment of one-to-one agency HCA staff to care for them. This was used particularly when individuals were disrupting the ward routines and was described as a way to support both people living with dementia and ward staff.

Interventions to improve care

We have provided a detailed analysis of the experiences of care and of the impact of the culture and organisation of wards and interactional work of delivering care at the bedside from the perspectives of people living with dementia, their family carers and ward staff. We have identified the ways in which resistance to care manifests, is recognised by staff and is responded to routinely. We have also examined the impact of these responses over time. A number of studies have highlighted the difficulties of caring for people living with dementia within the acute hospital setting;24,132 however, there is an evidence vacuum in understanding how the care of these patients can be improved within the acute setting,20 and few studies have demonstrated how to practically address these challenges. In response, within this chapter, it is demonstrated how the findings of this study can be used to develop training and test the feasibility of interventions to improve the quality and humanity of care for this key population and to support the ward staff caring for them.

Within this chapter we focus on our approaches and strategies to build on and utilise the study findings to (1) consult with, raise awareness of and inform wider user communities and the public about our findings and the experiences of care in acute wards for people living with dementia, their family carers and ward staff, (2) extend the dissemination and delivery of training to nurses and HCAs working in acute wards via open-access online training and (3) develop and test interventions at ward level to improve the quality and humanity of care that people living with dementia receive during a hospital admission. To do this, we have:

• involved people living with dementia, their carers, their families and the wider public in the research process and in the discussion of initial findings via an ongoing dissemination and consultation process in collaboration with the arts through the development of a programme of dementia-friendly film screenings and festivals

• presented the initial findings of our research to nurses, students, care staff, patients, advocacy groups and experts from a variety of disciplines in the UK and Europe

• extended and enhanced our community of interest by developing social media programmes to extend the reach of our research via a co-ordinated cross-platform campaign of films, training and research updates on open-access social media platforms [dedicated website, Twitter (Twitter, Inc., San Francisco, CA, USA), Facebook (Facebook, Inc., Menlo Park, CA, USA), Instagram (Instagram, Menlo Park, CA, USA) and YouTube (YouTube, LLC, San Bruno, CA, USA)]

• established a website (www.storiesofdementia.com) via accessible logos and illustrations to create an inclusive, distinct, memorable and consistent identity that reflects the overall objectives and values at the heart of our research and to ensure that our outputs are instantly recognisable as belonging to our project

• collaborated with Dementia UK and Admiral Nurses on developing –

  • no-cost interventions or ‘hacks’ for acute wards

  • short training (n = 17) and ‘trigger’ (n = 12) films for nurses and HCAs working in acute wards available via open-access online training

• conducted a feasibility study within one ‘laboratory’ ward to introduce three interventions or ‘hacks’ and to establish the feasibility of introducing them in acute wards.

Our detailed analysis has identified features of the social and organisational context of front-line care and key moments within the hospital system and ward routines and timetables that we believe are pivotal opportunities for change. In response, we have developed small changes or ‘hacks’, which do not require ‘permission’, cost money or demand involvement of the hospital management or wider governance structures. We believe that these changes can improve both the quality and humanity of care for people living with dementia and also improve the working lives of ward staff. An individual nurse or HCA may believe that they lack the power or influence to make large transformations to their hospital, but small changes can have a significant impact on the local culture of the ward, the ward team and patient care.

We are extending our training and interventions to involve other hospital trusts and applying our findings to other areas of care (delirium) that affect people living with dementia during an acute hospital stay. The findings have also informed another Health Services and Delivery Research study that is currently ongoing [Understanding how to facilitate continence for people living with dementia in acute hospital settings: raising awareness and improving care (project number: 15/136/67)].

Recommendations

Reflections and limitations

Key limitations identified from the study included the potential for the Hawthorne effect to influence data collection and establishing the generalisability of the findings.

We are aware of the potential for the Hawthorne or researcher effect to have an impact on research. Importantly, ethnography does not aim to achieve distance and detachment, with one of the strengths of this approach being the development of close ties with social actors (those being observed) in the field. It has been argued (e.g. by Monahan and Fisher198) that any performances observed, however staged or influenced by the presence of researchers, often reveal critical insights by displaying how people see themselves and how they want to be seen. We acknowledge that behaviour may have been changed by the researchers’ presence in the wards and believe that it is important to explore this potential effect and to learn from it.

In acknowledging this limitation, we enrolled a number of strategies intended to minimise this potential effect on our data:

• we believe that our extended periods of observation mean that the participants became used to the researchers’ presence and supported the development of trust with ward staff

• we believe that our extended periods of observation made it difficult for participants to maintain performed or exaggerated behaviours over time

• interviews with ward staff, patients and carers during the periods of observation also explored this effect and participants were explicitly asked if they felt a need to change or adapt practice (staff) or if they experienced any changes in the care they received (patients and families) during the research and when care was being observed.

Our approach emphasises the importance of comparisons across sites,189 which we believed allowed us to optimise the generalisability of our findings. 190 We identified a range of variables that we believed may have influenced the phenomenon by using purposive and maximum variation sampling to include five hospitals that represent hospital types, geographical location, expertise, interventions and quality. 199 However, there may be other important variables that we did not consider that could have affected the generalisability of our findings. Although at first glance hospitals may appear to operate in similar ways, they often have their own unique culture, which is informed by local dominant cultures and belief systems, and which in turn means that care and decision-making can vary widely within institutions. 191,192

Practical limitations presented themselves throughout the research. As a result of the permissions granted to the researchers, they were unable to follow or track what happened to patients once they were transferred to other parts of the hospital. Thus, we are unable to provide evidence of whether or not the delivery of care, and a patient’s response to it, changed after a patient was moved to another ward, or of whether or not the patient’s resistance or behaviour changed in a new setting. As such, the research recounts the interactions around resistance to care only while individual patients were admitted to MAU or T&O wards.

Similarly, the researchers could observe and make notes for only a limited number of hours and days at each site. It was not uncommon to arrive on a ward and be told that we had missed a patient behaving in a certain way or had missed the team identifying a patient need. Nor was it always possible to stay on the ward long enough to see how a patient’s behaviour may or may not have changed over the entire period of their admission, if nurses adapted their strategies or if a shift change could prompt any differences to care delivery and/or the management of resistive behaviours. However, the everyday and routine nature of the interactions observed leave the researchers confident that they did not miss significant changes to care (or response to care) once observations ceased.

A final limitation is that acute wards are difficult to observe. By their nature, they are large and busy places. It is simply not possible to observe everything going on at any given time on a ward. The researchers acknowledge that a lone researcher can focus on only a single area of the ward or unit and cannot be aware of all interactions taking place around them.

Contributions of authors

Katie Featherstone (Reader in Sociology and Medicine) was the chief investigator for the study. She led the initial conception and design of the project, managed the day-to-day running of the project, led the ethnographic fieldwork across all sites, conducted ethnographic fieldwork at five T&O wards and transcribed fieldnotes. She led the analysis and theory development, prepared the results for publication and was a primary author of this report.

Andy Northcott (VC2020 Lecturer, Sociology of Medicine) was the lead researcher for the study. He conducted ethnographic fieldwork at five MAUs, transcribed fieldnotes, conducted analysis of data, prepared the results for publication and was a primary author of this report.

Jane Harden (Senior Lecturer, Adult Nursing) contributed to all stages of the study. She contributed to the initial conception and design of the project and the recruitment of hospital trusts, and prepared the results for publication. She has led the engagement with curriculum development, continuing education outputs, the theoretical development and implementation of interventions.

Karen Harrison Denning (Head of Research & Publications, Dementia UK) has made significant contributions to the review of literature, the project’s engagement with curriculum development, continuing education outputs and the development of interventions. She has also made significant contributions to the project engagement with Admiral Nurses, Dementia UK and UK health-care policy.

Rosie Tope (Carers UK Trustee) contributed at all stages of the study. She contributed to the initial conception and design of the project and the consultations with carers and people living with dementia and public collaborations, contributed to the review of the literature and the analysis, and ensured that the perspectives of carers were represented throughout the study.

Sue Bale (Director of Research and Development and Professor of Wound Management) contributed to the initial conception and design of the project, site recruitment, research governance, participant identification and recruitment and the preparation of results, with a focus on dissemination, interventions and follow-up studies.

Jackie Bridges (Professor of Older People’s Care) contributed to the review of the literature, site recruitment and analysis, and to the preparation of results for publication.

Publication

Boddington P, Featherstone K. The canary in the coal mine: continence care for people with dementia in acute hospital wards as a crisis of dehumanization. Bioethics 2018;32:251–60.

Data-sharing statement

This is a qualitative study and, therefore, the data generated are not suitable for sharing beyond those that are contained within the report. Further information can be obtained from the corresponding author.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health and Social Care.