Pain is prevalent among people with dementia, often resulting from advanced age and multiple comorbidities [1–4]. Pain recognition and assessment in this patient population is widely recognised to be challenging as extensive cognitive deterioration in advanced stages often significantly impairs or removes the possibility of patient self-report, exposing patients to the risks of under-assessment and under-treatment of pain [1, 5–8]. Where patient testimony is unavailable or extensively impaired, health professionals are advised to observe for behavioural and nonverbal cues (such as grimacing, guarding, frowning, moaning, agitation, and aggression) which may indicate pain in nonverbal, cognitively impaired adults [9].

Standardised pain assessment tools provide an organised format by which professionals can attribute an estimated pain severity score to each behavioural indicator and aggregate them on completion to provide an overall score of estimated pain in the patient [10, 11]. However, difficulties associated with the use of these tools have been reported due to the overlap of behavioural indicators with indicators of other, non-pain related conditions such as hunger, distress and boredom [12, 13]. It is therefore recommended that scores from these assessments be interpreted in the context of other collateral knowledge and information about patients [14, 15].

As nurses’ and physicians’ administrative workloads continue to increase and, as a consequence, time spent with patients is reduced, much of the direct care lies with healthcare assistants (HCAs) [16–18]. HCAs (also known as nurse auxiliaries, healthcare support workers, and personal/clinical support workers) work in health and social care settings providing care to patients across a range of physical and psychosocial domains usually under the supervision of Registered Nurses (RNs). Their typical duties include: providing personal care, maintaining patient hygiene, assisting patients with eating and toileting, providing social interaction and psychological support, and basic housekeeping (e.g. making beds, laying tables for meals) [16–21]. These staff spend more time with patients than any other healthcare professional and often develop detailed knowledge of patients’ preferences, routines, and normative patterns of behaviour, mood, appetite and disposition [19–22]. As a result, they are often the first to recognise and alert professional staff to changes in patients’ physical and cognitive functioning [22].

Previous work has explored the impact of HCAs on patient care and outcomes in the context of the care of older adults, palliative care and dementia [19–23]. No previous work has examined HCAs’ experiences and perspectives of and contributions to pain assessment and management in nonverbal patients with dementia approaching the end of life. This study aimed to explore HCAs’ perspectives and experiences of pain assessment and management in people with advanced dementia approaching the end of life in hospice, acute care and nursing home settings to address this gap in an area of clinical practice widely acknowledged to be challenging and critically important. It forms part of a wider programme of research into assessing and managing pain in this complex patient group, from which a number of articles have been published [24–27].

The complexity and challenges of pain assessment and management for people with advanced dementia are well-recognised and documented with much of this previous work focused on the experiences of nurses and physicians on whom the responsibility of these assessments usually falls [12, 32, 33]. To our knowledge, this is the first study which explores HCAs’ perspectives and experiences of, and their role in, pain assessment and management for people with advanced dementia approaching the end of life.

HCAs in this study were not required to conduct formal pain assessments for their residents and most reported being unfamiliar with the contents and application of standardised pain assessment tools. However, participants’ narratives revealed that most regularly performed informal relationship-centred pain assessments as an inherent part of care provision. Such assessments occur when knowledge and understanding of patients’ normative patterns of behaviour, physical and cognitive functioning and past reactions to pain are used to inform recognition and interpretation of behavioural and nonverbal pain cues [34]. Daily care provision brought HCAs in this study into close physical and social contact with residents, allowing them to develop a comprehensive knowledge and understanding of their care-recipients over time [19–22]. In most cases, changes in residents’ norms prompted HCAs to observe for behavioural and nonverbal indictors of pain and interpret them within contextual knowledge of the resident. Closer analysis revealed that most were performing many of the observations required by standardised assessment tools recommended by healthcare policies for use in this patient population [10, 11, 35–38]. Participants’ knowledge of residents also allowed idiosyncratic expressions of pain to be recognised as such and facilitated distinction of behavioural indicators of pain from non-pain related cues, overcoming a commonly reported difficulty associated with pain assessment in nonverbal patient populations [12, 32]. Understanding the residents for whom they cared was perceived to be critical to recognising and reporting pain, illness or distress in these patients, particularly for those unable to self-report. Many participants in this study reported difficulties and limitations in recognising pain in residents with whom they were unfamiliar. Previous research reported similar findings for nurses and certified nursing assistants (CNAs) [34]. Interestingly, in the present study, most HCAs sought to gain insight into residents’ past pain experiences and behaviours through liaison with patients’ families, friends and other key health and social care staff. The involvement of families and other key social contacts in pain assessment for patients with severe cognitive impairment and/or dementia for whom self-report is unavailable, has been widely recommended [39–43].

Findings in this study illustrated three cases of disparities between HCAs’ and nurses’ knowledge of residents which may have critical implications for the interpretation of pain scores. The finding reported here suggests that both staff should be present during assessment to allow outcomes to be considered alongside other collateral patient information [14, 15, 44]. Participants with task-oriented approaches to care were a minority in this study. These staff reported being under-skilled in recognising pain and relied largely on nurses’ reports of existing pain to approach care routines with additional caution, rather than attempting to observe for signs of new pain. Previous studies have reported poor patient and staff outcomes which result when healthcare staff lack understanding of the pain experience in people with dementia and are inadequately trained and supported to recognise, assess and manage pain for these patients [45–48].

Pain reporting was heavily impacted by the nature of relationships with other healthcare staff. Participants who were openly encouraged and supported to raise and discuss all concerns, including pain, and who felt valued for their contribution to patient care provided detailed reports which expounded their interpretations of observed behaviours [49]. These HCAs monitored for and reported back on patient response to analgesia and, without nurse direction, planned care tasks to allow the benefits of pain relief to manifest before attempting to move, lift or turn patients. Positive work-related identities and relationships with other staff have resulted in improved staff morale, increased confidence and improvements to care quality and patient outcomes [19–21, 49]. Negative relationships were reported in cases where professional hierarchies dominated and participants felt ignored or undervalued. Strong in-group identities among HCAs can propagate ‘us and them’ attitudes leading to negative perceptions of and noticeable professional distance from nursing and medical staff; a finding reported in another study of HCAs in working in dementia care [21]. Pain reporting was severely impacted by negative work-related identities and relationships; in such cases it was at best perfunctory and detail-poor, and at worst uninformative, with participants referring to general ill health rather than pain. Work-related identity construction is recognised to impact significantly on patient care and outcomes and has been studied extensively using social identity theory [50–56]. Negative work-related identities, strong in-group membership and dysfunctional team dynamics are associated with deterioration of collaborative approaches to care and withholding or poor exchange of information, exacerbating the challenges of complex care and resulting in negative outcomes for patients and staff [21, 57–60]. The impact of the quality of communication between healthcare professionals on patient outcomes has been studied extensively across multiple health conditions including dementia; much of this previous work has focused on nurse-physician interactions [60–62]. Although this study provides insight into HCA reporting of pain, the findings are limited to their perspectives alone. Future studies may wish to examine nurses’ and physicians’ experiences of and perspectives on communicating with HCAs, or explore communication dyads and triads among these staff in order to elucidate how information provided by HCAs is received, understood and processed by qualified health professionals. This is particularly important given that tensions in relationships between CNAs and RNs have been previously reported [34].

Participants expressed frustration and disappointment at being excluded from MDT meetings with many believing this resulted from stigma regarding their status as unqualified workers. Most believed the benefits of HCA participation in the MDT would be reciprocal with all members of the team benefitting from having access to patient information that could inform care practice. Lloyd and colleagues (2011) warn that exclusion may result in HCAs feeling isolated from the wider care team, strengthening in-group identities and lead to difficulties as described above [21]. The importance of inclusivity and teamwork among health professionals is emphasised in a number of health policies and is a required competency of the HCA role [63–68]. Future research could evaluate the impact of the inclusion of these staff in the multidisciplinary team on patient care and team dynamics, and should include the perspectives and experiences of HCAs as well as other members of the MDT.

The large majority of participants believed that ongoing, needs-driven training which actively enhances knowledge and practical skills is required for all professionals and frontline staff involved in the care of people with dementia. Interestingly, the minority of participants who provided sparse answers for other interview questions responded comprehensively when asked whether further training was required. Appropriately educating staff to understand the complexities of dementia care results in better management of the emotional and physical demands of the work, increased engagement with the role and adoption of person-centred approaches to care [69]. Healthcare policy emphasises a need for all staff (including HCAs) working in dementia to be appropriately educated, trained and equipped to competently provide high quality care but does not suggest appropriate platforms for delivery of this training [70, 71]. Most participants reported that e-learning was unengaging, generic and resulted in little, or no, skills and knowledge development. These findings are reiterations of those in many other studies indicating little progress in this area and suggesting the need for a continued effort towards developing and trialling engaging and appropriate educational programmes for these staff [22, 23, 72]. Few studies have been completed in this regard; however, positive outcomes for HCAs’ confidence, motivation to engage in care and reporting to professionals have been reported following practical skills training in palliative care [73]. Many participants believed that the HCA role could be expanded in dementia care; a sentiment echoed in recent work [74]. In the current study, most participants (including task-oriented HCAs) expressed significant interest in learning how to monitor for and report on treatment response, side and adverse effects, and how to use and report basic assessments such as the Abbey Pain Scale to standardise and improve the quality of reporting among HCAs. Tools such as the Abbey Pain Scale are simple to use and do not require extensive training; previous work reports no significant difference in the ability of qualified health professionals and unqualified staff working outside the healthcare setting to detect pain in facial expressions [75]. However, given the challenges experienced by nursing and other health professionals in the use of these tools, and that assessment scores are often interpreted in conjunction with clinical judgement and collateral information from multiple sources, a number of factors must be considered when exploring an expanded role for HCAs in pain assessment. Firstly there is the question of how training should be delivered, and by whom. Secondly, given the variation in knowledge, skills and competence of HCAs, the process of selecting staff for this expanded role, the selection criteria used and how competence is determined must be considered. The manner in which this might be implemented in clinical practice without duplicating or complicating current use of pain tools and approaches to pain assessment, and without creating and/or exacerbating interdisciplinary tensions, requires careful consideration. Finally, there must be formal, robust evaluation of the use of pain tools by HCAs and the impact on patient care and outcomes, the culture of multidisciplinary working and approaches to end of life care across healthcare settings.

There are some limitations to this study. Although we aimed to recruit for maximum variation among participants, recruitment relied on contacting participants through networks established by members of the PMG. Acute and hospice care settings in this study are linked to teaching hospitals and a university; therefore it is likely that staff are better aware of pain in people with dementia than those in settings without connections to academic teaching and research. The majority of HCAs were recruited from nursing home settings; the shortfall of HCAs from acute care reflects the difficulties in interviewing staff working in this setting, whilst limited numbers recruited from hospice resulted from low take up of participation among these staff despite attempts to publicise the study in these organisations. Although data analysis did not identify significant difference between perspectives and experiences across care settings for most findings, negative relationships were only reported in the nursing home setting. Future studies may wish to expand on this study using a broader selection of HCAs across settings. The self-selecting nature of the sample means that the views of engaged, motivated participants with an aptitude for providing person-centred care may be overrepresented. However, whilst findings related to specific elements of pain assessment and management may reflect more instances of best practice than general practice among HCAs, many of the general principles regarding the way these elements are impacted by social and group identities, dynamics and relationships are supported by other studies of staff working in palliative and dementia care and other patient populations.

Assessing and managing pain in dementia is a priority in provision of end of life care for people with dementia but is one of the most commonly reported challenges among nursing and medical staff. HCAs provide the majority of direct care to these patients, attending to a broad spectrum of physical and emotional needs, and as a result are often well placed to recognise problems as they arise. However, this study indicates that whether this critical information is shared with professional staff and the quality of that information exchange are highly dependent on individual and group identities which form in response to relationships with professional staff. This study highlights the potential for the large workforce of non-qualified healthcare staff to contribute meaningfully to a critical and complex area of dementia care. However, with the prevalence of people living and dying with dementia ever increasing, it is essential that serious consideration be given to, and progress made in, supporting, educating, training and upskilling these staff to improve the quality of care and care experience for people with dementia and their families. Critical to this is changing perceptions and the stigma that the HCA role is unqualified, poorly remunerated manual labour, reconceptualising it as a role which requires significant dedication, commitment and the resilience to respond to and meet the physical and emotional demands of caring for people with dementia with compassion and respect.