Why do we need a feasibility study of improving care for end-of-life agitation in nursing homes? Because quality dying with dementia in nursing homes remains elusive, and agitation, among other neuropsychiatric symptoms, is a cause of distress for people with dementia, family members, and professional carers alike. While several intervention studies have targeted agitation per se (Ballard et al., Reference Ballard2018; Chenoweth et al., Reference Chenoweth2014; Lichtwarck et al., Reference Lichtwarck2018; Rapp et al., Reference Rapp2013), and others specifically at the end of life (Van den Block et al., Reference Van den Block2020), we neither know what works nor what is scalable in this area. Multicomponent controlled trials that have demonstrated clinically significant reductions in agitation have combined various elements of staff training, pain assessment, increased social interaction, antipsychotic review, physician review, and case conferencing with individualized treatment plans (Ballard et al., Reference Ballard2018; Lichtwarck et al., Reference Lichtwarck2018; Rapp et al., Reference Rapp2013). However, the cost-effectiveness and scalability of these time- and resource-intensive interventions are unknown.
Notably, the less intensive, and less costly, Managing Agitation and Raising Quality of Life (MARQUE) intervention, a cluster-randomized controlled trial of a manualized, staff training intervention delivered over 8 months in 20 English care homes, was neither efficacious in managing agitation, staff burnout, nor lowering antipsychotic use when compared with usual treatment (Livingston et al., Reference Livingston2019). However, MARQUE was feasible and cost effective in terms of quality of life, based on quality-adjusted life-years and the European quality of life five dimensions questionnaire (EQ-5D-5L). This unpacking of quality of life outcomes from agitation outcomes – previously associated (Livingston et al., Reference Livingston2017) – was surprising and worthy of future note. Interestingly, staff abuse tactics also decreased by about a third in the MARQUE group. Although recognizing limitations of the study, including inadvertent exclusion of those most likely to benefit, the authors recommended a more intensive intervention, implementing whole-home management and cultural change, resident social and activity times, and a longer time to implement change. The Sampson et al. (Reference Sampson2020) study is part of the MARQUE program.
The PACE Steps to Success Program (Van den Block et al., Reference Van den Block2020) was a multicomponent intervention integrating basic non-specialist palliative care in nursing homes, a focus similar to that of the Sampson et al. (Reference Sampson2020) study. The PACE intervention comprised a cluster randomized clinical trial implemented in 78 nursing homes in seven countries over 1 year, using a train-the-trainer approach to support staff in a 6-step program. The steps were (1) advance care planning with residents and family, (2) assessment, care planning, review of needs, and problems, (3) coordination of care via multidisciplinary review meetings, (4) delivery of high-quality care focusing on pain and depression, (5) care in the last days of life, and (6) care after death. Once again, outcomes were negative. Specifically, residents’ comfort in the last week of life did not differ between intervention and control group; although the quality of care in the last month of life did significantly improve, the change being most apparent in the preparatory tasks subscale (including advance care planning). Notably, while staff knowledge of palliative care improved, this was not deemed clinically important. The PACE program also had no effect on other key outcome measures, namely, hospital use in the last month of life and place of death (Honinx et al., Reference Honinx2020). The lack of effect of the PACE program was attributed to suboptimal implementation of an overly complex intervention with a mismatch between intervention elements and primary outcomes (Van den Block et al., Reference Van den Block2020). Other obstacles included challenges in recruitment and collecting data in nursing homes, vividly demonstrated in England, where 75% of eligible nursing homes refused to participate (Moore et al., Reference Moore, Payne, Van den Block, Ten Koppel, Szczerbińska and Froggatt2019). It is well known that study fidelity and implementation are notoriously difficult in nursing homes (Hunt et al., Reference Hunt, Stephens and Smith2020). For example, a randomized controlled trial of palliative care teams (PCTeams) in 25 New York State nursing homes found no statistically significant effect of the intervention, although only 6 of the 14 facilities had continuously working PC Teams throughout the study period. However, in contrast to the PACE program, there was a reduction of in-hospital deaths associated with this intervention (Temkin-Greener et al., Reference Temkin-Greener2018).
Based on the limited information from studies constrained by these implementation issues, we are currently left with a dichotomy between overly complex versus overly simplistic interventions, both of which have their limitations. A broader focused systematic review of changed staff care practices that improved resident outcomes found that no single or combination of interventions were associated with greater likelihood of positive outcomes. Moreover, changing staff behavior did not necessarily improve resident outcomes. However, studies targeting specific care tasks such as oral care and physical restraints were more likely to produce positive outcomes than those relying on broader practice changes such as care philosophy (Low et al., Reference Low2015). This argues for single component and against multicomponent interventions (Van den Block et al., Reference Van den Block2020). Alternatively, the broader, behavior/culture change interventions may take longer to see benefit, be measurable or embedded in the facility long term, particularly in the light of the dynamic nature of nursing home staffing and staff turnover, especially among senior nursing management. These realities do not fit easily within research funding period constraints.
Our failure to identify the ingredients of optimal end-of-life care for people with dementia in nursing homes is clearly not for lack of trying. Moreover, the refreshing publication of negative studies in this area is informative. Sampson’s current study looks at the feasibility and acceptability of a modified version of the MARQUE intervention, specifically targeting agitation in dementia at the end of life, underpinned by a conceptual framework linking agitation to unmet needs and in turn quality of life. Despite narrowing the scope, there are many complex themes that the training intervention touches on. Four 2-h training modules have been developed to train care staff and provide information to families on a range of topics from “What are end of life and palliative care?” to “How to communicate effectively with families, GPs and nurses.” Of the 53 staff that undertook the training (74% of eligible staff), 84% attended at least three of the four sessions and the training was found to be acceptable. Feedback on the training was positive from both staff and family carers. Of interest was the role of the trainers and how they engaged with the staff during the intervention. Staff reported feeling listened to, understood, and appreciated, and results from the Sense of Competence in Dementia Care (staff competency) scale showed positive change between pre- and post-intervention phases. This engagement and validation aspect of clinical training interventions may be significant and need to be considered for inclusion in future studies, as a means of augmenting resident outcomes. As suggested by Sampson et al. (Reference Sampson2020), whilst improving knowledge is evidently important, it may be that attention given to staff well-being and personhood are actually the drivers of change.
The changes that we are interested in are sustainable, long-term cultural changes. It is one thing to know and remember how to provide care for older residents with complex needs, but to sustain effort, empathy, and compassion in doing so is predicated on feeling acknowledged and appreciated. This is what Sampson et al. pens as “investing in staff.” What unfolded from the qualitative data was the importance of shared learning, communication, and most of all, “reflective practice.” Providing quality care for residents at the end of life can only be manualized to a certain extent. Much relies on care staff having the impetus and awareness to note possible distress or suffering, reflecting on why this is happening and then having a safe and supportive environment in which to discuss and explore possible solutions, with junior and senior colleagues and the resident’s family. These solutions should ideally be based on the resident’s will and preferences; however, they are expressed. In this way, we rely on nursing staff to be “needs interpreters” for residents who may be “silent and suffering” (Aminoff and Adunsky, Reference Aminoff and Adunsky2006; Peisah et al., Reference Peisah, Weaver, Wong and Strukovski2014). We need an eyes wide open approach to unmet needs and suffering for people with dementia at the end of life. What remains elusive is how to detect and measure such suffering – whether it be through agitation or other “signs” identified in the Mini Suffering State Examination (i.e. not calm, screams, pain, decubitus ulcers, malnutrition, eating disorders, invasive action, unstable medical condition, and suffering perceived by medical staff or family) (Aminoff, Reference Aminoff2008).
Quality end of life is a complex construct grounded on the assumption that quality of living is the same as quality of dying, when that may not be true, or that quality of life can be validly rated by a proxy. Moreover, quality care provision for people with dementia at the end of life is a complex, multifaceted process, nursing homes are complex multifaceted environments, and nursing home research implementation is complex and scalability of findings even more so (Hunt et al., Reference Hunt, Stephens and Smith2020; Van den Block et al., Reference Van den Block2020). We need a feasibility and acceptability study such as that of Sampson et al. (Reference Sampson2020) because we are like Goldilocks and the Three Bears, we need interventions for quality care at the end of life that are not too complex, or not complex enough, but just right.
You have
Access
Why do we need a feasibility study of improving care for end-of-life agitation in nursing homes? Because quality dying with dementia in nursing homes remains elusive, and agitation, among other neuropsychiatric symptoms, is a cause of distress for people with dementia, family members, and professional carers alike. While several intervention studies have targeted agitation per se (Ballard et al., Reference Ballard2018; Chenoweth et al., Reference Chenoweth2014; Lichtwarck et al., Reference Lichtwarck2018; Rapp et al., Reference Rapp2013), and others specifically at the end of life (Van den Block et al., Reference Van den Block2020), we neither know what works nor what is scalable in this area. Multicomponent controlled trials that have demonstrated clinically significant reductions in agitation have combined various elements of staff training, pain assessment, increased social interaction, antipsychotic review, physician review, and case conferencing with individualized treatment plans (Ballard et al., Reference Ballard2018; Lichtwarck et al., Reference Lichtwarck2018; Rapp et al., Reference Rapp2013). However, the cost-effectiveness and scalability of these time- and resource-intensive interventions are unknown.
Notably, the less intensive, and less costly, Managing Agitation and Raising Quality of Life (MARQUE) intervention, a cluster-randomized controlled trial of a manualized, staff training intervention delivered over 8 months in 20 English care homes, was neither efficacious in managing agitation, staff burnout, nor lowering antipsychotic use when compared with usual treatment (Livingston et al., Reference Livingston2019). However, MARQUE was feasible and cost effective in terms of quality of life, based on quality-adjusted life-years and the European quality of life five dimensions questionnaire (EQ-5D-5L). This unpacking of quality of life outcomes from agitation outcomes – previously associated (Livingston et al., Reference Livingston2017) – was surprising and worthy of future note. Interestingly, staff abuse tactics also decreased by about a third in the MARQUE group. Although recognizing limitations of the study, including inadvertent exclusion of those most likely to benefit, the authors recommended a more intensive intervention, implementing whole-home management and cultural change, resident social and activity times, and a longer time to implement change. The Sampson et al. (Reference Sampson2020) study is part of the MARQUE program.
The PACE Steps to Success Program (Van den Block et al., Reference Van den Block2020) was a multicomponent intervention integrating basic non-specialist palliative care in nursing homes, a focus similar to that of the Sampson et al. (Reference Sampson2020) study. The PACE intervention comprised a cluster randomized clinical trial implemented in 78 nursing homes in seven countries over 1 year, using a train-the-trainer approach to support staff in a 6-step program. The steps were (1) advance care planning with residents and family, (2) assessment, care planning, review of needs, and problems, (3) coordination of care via multidisciplinary review meetings, (4) delivery of high-quality care focusing on pain and depression, (5) care in the last days of life, and (6) care after death. Once again, outcomes were negative. Specifically, residents’ comfort in the last week of life did not differ between intervention and control group; although the quality of care in the last month of life did significantly improve, the change being most apparent in the preparatory tasks subscale (including advance care planning). Notably, while staff knowledge of palliative care improved, this was not deemed clinically important. The PACE program also had no effect on other key outcome measures, namely, hospital use in the last month of life and place of death (Honinx et al., Reference Honinx2020). The lack of effect of the PACE program was attributed to suboptimal implementation of an overly complex intervention with a mismatch between intervention elements and primary outcomes (Van den Block et al., Reference Van den Block2020). Other obstacles included challenges in recruitment and collecting data in nursing homes, vividly demonstrated in England, where 75% of eligible nursing homes refused to participate (Moore et al., Reference Moore, Payne, Van den Block, Ten Koppel, Szczerbińska and Froggatt2019). It is well known that study fidelity and implementation are notoriously difficult in nursing homes (Hunt et al., Reference Hunt, Stephens and Smith2020). For example, a randomized controlled trial of palliative care teams (PCTeams) in 25 New York State nursing homes found no statistically significant effect of the intervention, although only 6 of the 14 facilities had continuously working PC Teams throughout the study period. However, in contrast to the PACE program, there was a reduction of in-hospital deaths associated with this intervention (Temkin-Greener et al., Reference Temkin-Greener2018).
Based on the limited information from studies constrained by these implementation issues, we are currently left with a dichotomy between overly complex versus overly simplistic interventions, both of which have their limitations. A broader focused systematic review of changed staff care practices that improved resident outcomes found that no single or combination of interventions were associated with greater likelihood of positive outcomes. Moreover, changing staff behavior did not necessarily improve resident outcomes. However, studies targeting specific care tasks such as oral care and physical restraints were more likely to produce positive outcomes than those relying on broader practice changes such as care philosophy (Low et al., Reference Low2015). This argues for single component and against multicomponent interventions (Van den Block et al., Reference Van den Block2020). Alternatively, the broader, behavior/culture change interventions may take longer to see benefit, be measurable or embedded in the facility long term, particularly in the light of the dynamic nature of nursing home staffing and staff turnover, especially among senior nursing management. These realities do not fit easily within research funding period constraints.
Our failure to identify the ingredients of optimal end-of-life care for people with dementia in nursing homes is clearly not for lack of trying. Moreover, the refreshing publication of negative studies in this area is informative. Sampson’s current study looks at the feasibility and acceptability of a modified version of the MARQUE intervention, specifically targeting agitation in dementia at the end of life, underpinned by a conceptual framework linking agitation to unmet needs and in turn quality of life. Despite narrowing the scope, there are many complex themes that the training intervention touches on. Four 2-h training modules have been developed to train care staff and provide information to families on a range of topics from “What are end of life and palliative care?” to “How to communicate effectively with families, GPs and nurses.” Of the 53 staff that undertook the training (74% of eligible staff), 84% attended at least three of the four sessions and the training was found to be acceptable. Feedback on the training was positive from both staff and family carers. Of interest was the role of the trainers and how they engaged with the staff during the intervention. Staff reported feeling listened to, understood, and appreciated, and results from the Sense of Competence in Dementia Care (staff competency) scale showed positive change between pre- and post-intervention phases. This engagement and validation aspect of clinical training interventions may be significant and need to be considered for inclusion in future studies, as a means of augmenting resident outcomes. As suggested by Sampson et al. (Reference Sampson2020), whilst improving knowledge is evidently important, it may be that attention given to staff well-being and personhood are actually the drivers of change.
The changes that we are interested in are sustainable, long-term cultural changes. It is one thing to know and remember how to provide care for older residents with complex needs, but to sustain effort, empathy, and compassion in doing so is predicated on feeling acknowledged and appreciated. This is what Sampson et al. pens as “investing in staff.” What unfolded from the qualitative data was the importance of shared learning, communication, and most of all, “reflective practice.” Providing quality care for residents at the end of life can only be manualized to a certain extent. Much relies on care staff having the impetus and awareness to note possible distress or suffering, reflecting on why this is happening and then having a safe and supportive environment in which to discuss and explore possible solutions, with junior and senior colleagues and the resident’s family. These solutions should ideally be based on the resident’s will and preferences; however, they are expressed. In this way, we rely on nursing staff to be “needs interpreters” for residents who may be “silent and suffering” (Aminoff and Adunsky, Reference Aminoff and Adunsky2006; Peisah et al., Reference Peisah, Weaver, Wong and Strukovski2014). We need an eyes wide open approach to unmet needs and suffering for people with dementia at the end of life. What remains elusive is how to detect and measure such suffering – whether it be through agitation or other “signs” identified in the Mini Suffering State Examination (i.e. not calm, screams, pain, decubitus ulcers, malnutrition, eating disorders, invasive action, unstable medical condition, and suffering perceived by medical staff or family) (Aminoff, Reference Aminoff2008).
Quality end of life is a complex construct grounded on the assumption that quality of living is the same as quality of dying, when that may not be true, or that quality of life can be validly rated by a proxy. Moreover, quality care provision for people with dementia at the end of life is a complex, multifaceted process, nursing homes are complex multifaceted environments, and nursing home research implementation is complex and scalability of findings even more so (Hunt et al., Reference Hunt, Stephens and Smith2020; Van den Block et al., Reference Van den Block2020). We need a feasibility and acceptability study such as that of Sampson et al. (Reference Sampson2020) because we are like Goldilocks and the Three Bears, we need interventions for quality care at the end of life that are not too complex, or not complex enough, but just right.
Description of authors’ roles
CP conducted the primary review of the paper, literature review for discussion, planned the commentary structure and wrote the paper. TJ provided secondary reviews and intellectual input into discussion of literature and contributed to writing the paper.