Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice J Adv Nurs . 2025 Aug 6;82(5):5430–5443. doi: 10.1111/jan.70129 Search in PMC Search in PubMed View in NLM Catalog Add to search Development and Validation of the End‐of‐Life Assessment Tool for Advanced Dementia: A Multi Method Study Carolyn Bourke Carolyn Bourke 1 Health/UARC, University of Technology Sydney, Ultimo, New South Wales, Australia Find articles by Carolyn Bourke 1 , Lynn Chenoweth Lynn Chenoweth 2 Centre for Healthy Brain Ageing (CHeBA), Discipline of Psychiatry, UNSW Medicine and Health, Sydney, New South Wales, Australia Find articles by Lynn Chenoweth 2, ✉ , Ekavi Georgousopoulou Ekavi Georgousopoulou 3 The University of Notre Dame Australia, Darlinghurst, New South Wales, Australia 4 ACT Health, Phillip, Australian Capital Territory, Australia Find articles by Ekavi Georgousopoulou 3, 4 , Anna Williams Anna Williams 5 School of Nursing and Midwifery, Faculty of Health, University of Technology Sydney, Sydney, New South Wales, Australia 6 School of Nursing and Midwifery & Translational Health Research Institute (THRI), Western Sydney University, New South Wales, Australia Find articles by Anna Williams 5, 6 Author information Article notes Copyright and License information 1 Health/UARC, University of Technology Sydney, Ultimo, New South Wales, Australia 2 Centre for Healthy Brain Ageing (CHeBA), Discipline of Psychiatry, UNSW Medicine and Health, Sydney, New South Wales, Australia 3 The University of Notre Dame Australia, Darlinghurst, New South Wales, Australia 4 ACT Health, Phillip, Australian Capital Territory, Australia 5 School of Nursing and Midwifery, Faculty of Health, University of Technology Sydney, Sydney, New South Wales, Australia 6 School of Nursing and Midwifery & Translational Health Research Institute (THRI), Western Sydney University, New South Wales, Australia * Correspondence: Lynn Chenoweth ( [email protected] ) ✉ Corresponding author. Revised 2025 Jun 2; Received 2024 Nov 12; Accepted 2025 Jul 14; Issue date 2026 May. © 2025 The Author(s). Journal of Advanced Nursing published by John Wiley & Sons Ltd. This is an open access article under the terms of the http://creativecommons.org/licenses/by-nc-nd/4.0/ License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non‐commercial and no modifications or adaptations are made. PMC Copyright notice PMCID: PMC13069191 PMID: 40767121 ABSTRACT Aims To develop and validate the End‐of‐Life Care Assessment Tool for Dementia (EoLC‐ATD). Design A methodological study with multiple phases. Methods Five sub‐studies comprising: a review of 90 validated dementia measures to compile an item bank of advanced dementia symptoms; focus groups with registered nurses on advanced dementia symptom identification and relevance of item bank inclusions; Delphi surveys with dementia experts seeking consensus on the EoLC‐ATD constructs and items; pilot testing of the EoLC‐ATD; and field testing of the EoLC‐ATD in persons with dementia. Results The item‐bank included 180 symptoms, most of which focus group nurses ( n = 17) identified as occurring in advanced dementia. Delphi surveys with dementia experts ( n = 31) achieved 70% consensus for 25 of 26 EoLC‐ATD items. Pilot testing of the EoLC‐ATD by two nurses in eight persons with dementia showed good agreement for six constructs (Cohen's Kappa 0.856–0.927) and 26 items (Cronbach's alpha 77.0). An ‘other sympto’ item was included following RN recommendation. The 27‐item EoLC‐ATD field tested by 17 nurses in persons with dementia ( n = 113) accurately identified advanced dementia symptoms (Cronbach's alpha 77.0, p < 0.001). Mortality at 180 days after baseline EoLC‐ATD was significant ( p < 0.001, area under the ROC curve p = 0.769). Conclusion The EoLC‐ATD accurately and reliably identified symptoms of advanced dementia. Implications for the Profession and/or Patient Care The EoLC‐ATD provides registered nurses with a single measure of advanced dementia symptoms that will help in identifying symptom‐responsive palliative care requirements. Impact The EoLC‐ATD will address the current lack of a validated dementia symptom measure for use by aged care home registered nurses to identify unique palliative and end‐of‐life care needs according to presenting symptoms in persons living with advanced dementia. Reporting Method STROBE Statement for cohort and mixed methods studies. Patient or Public Contribution An eight‐member Expert Advisory Group, which provided guidance and advice throughout the study, was composed of three carers of persons living with dementia, two dementia care clinicians, a dementia care clinical educator, and two dementia clinician researchers. Keywords: advanced dementia, end of life, palliative care, person‐centred care, registered nurses, symptom measurement 1. Introduction More than 55 million people worldwide live with dementia (World Health Organization 2023 ). Globally, dementia is currently the seventh leading cause of death and one of the major causes of disability and dependency among older people (World Health Organization 2023 ). Dementia results from a variety of diseases and injuries that affect the brain, with Alzheimer's disease being the most common form (Alzheimer's Society, UK 2022 ). Every aspect of the person's life is impacted by dementia, and in the advanced stage, long‐term residential aged care may be necessary. As dementia is a terminal illness, aged care staff need the knowledge and skills to provide the person with palliative care that is responsive to the person's unique bio‐psychosocial symptoms (Alzheimer's Society, UK 2022 ). Although progress is occurring in understanding palliative care requirements for these persons, unless aged care staff have a clear understanding of the unique symptomology occurring in advanced dementia, they may misinterpret or inadvertently ignore some symptoms and fail to provide the care required to achieve quality of life. Consequentially, the person's palliative care needs may not be identified until the final or terminal stage of life (Sampson et al. 2017 ). Residential aged care nurses would benefit from gaining access to a measure that can build essential knowledge and skills in identifying advanced dementia symptoms as a first step in ensuring that palliative care is responsive to symptoms unique to the individual (Brennan et al. 2023 ). 2. Background In Australia, dementia is the second leading cause of death overall and the leading cause of death for women, the median age at death being 89.2 years (Australian Institute of Health and Welfare (AIHW) 2022 ). More than half of the 188,000 Australians over 65 years of age who live in residential aged care have a moderate or advanced level of dementia (Australian Institute of Health and Welfare (AIHW) 2022 ). In the advanced stage of dementia, the person is often frail, has significantly impaired memory, comprehension and language ability, reduced physical functioning and self‐care ability, and has difficulty eating and swallowing. They may have co‐existing cardiac and/or vascular illness, dyspnoea and pain, and will be susceptible to urinary and respiratory infections and delirium (Alzheimer's Society, UK 2022 ). The person may also experience neuropsychiatric symptoms, physical and verbal agitation and/or apathy (Sampson et al. 2017 ). Person‐centred palliative care is essential for the relief of these symptoms and maintenance of quality of life (Brennan et al. 2023 ). There are often substantial gaps in the provision of person‐centred palliative care for persons living with dementia in Australian aged care homes, due to the low numbers of registered nurses (RN) with dementia‐specific palliative care training and limited access to medical and allied health staff with expertise in dementia‐specific palliative care (Australian Institute of Health and Welfare (AIHW) 2022 ). Suboptimal palliative care of persons with dementia in the Australian aged care sector was identified by the Royal Commission into Aged Care Quality and Safety (RCACQS) (Royal Commission into Aged Care Quality and Safety 2021 ). This investigation identified an urgent need for RNs to receive additional education on palliative care of persons living with advanced dementia, in supporting the person's “… right to fair, equitable and non‐discriminatory access to palliative and end‐of‐life care” (Royal Commission into Aged Care Quality and Safety 2021 ). As highlighted in the many public submissions to the RCACQS, the uncertainty surrounding advanced dementia symptoms makes it difficult for nurses to recognise the unique and often subtle functional changes occurring and to convey this information to family members and colleagues with some confidence (Goodman et al. 2015 ). The RCACQS recommended that the situation may benefit by having RNs administer a validated dementia symptom measure (Royal Commission into Aged Care Quality and Safety 2021 ) such as the Supportive and Palliative Care Indicators Tool (SPICT) (Boyd and Murray 2010 ), the Gold Standard Framework (GSF) (Royal College of General Practitioners 2011 ) or the Australia modified Karnofsky Performance Status measure (AKPS) (Abernethy et al. 2005 ). The GSF identifies some of the more pronounced symptoms occurring in advanced dementia (Royal College of General Practitioners 2011 ), while the SPICT (Boyd and Murray 2010 ) and the AKPS (Abernethy et al. 2005 ) focus on symptoms occurring in other life‐limiting illnesses such as cancer and chronic obstructive airways disease. None of these measures clearly discriminate between the bio‐psychosocial symptoms that occur in advanced dementia and symptoms occurring in other end‐stage illnesses such as cancer and chronic airways disease (Laver et al. 2016 ). Lack of access to a single, validated advanced dementia symptom measure which is suitable for use by RNs in the Australian aged care sector is, therefore, problematic. The study has addressed this issue by producing the validated End‐of‐Life Care Assessment Tool for Dementia (EoLC‐ATD), which when used by aged care RNs, can inform the provision of symptom‐specific palliative care for persons living with advanced dementia. It is important that registered nurses working in the aged care sector have access to a dementia severity measure which captures sufficient information on this stage of dementia, so that they can anticipate likely symptoms and allocate care and support accordingly (Brennan et al. 2023 ; Laver et al. 2016 ). This paper describes the development and validity testing of the EoLC‐ATD. 3. Study Aims The study aimed to develop and validate the End‐of‐Life Care Assessment Tool for Dementia (EoLC‐ATD) to improve assessment of advanced dementia symptoms by Registered Nurses (RN) in the aged care sector. 4. Study Methods 4.1. Study Design This was a methodological study with multiple phases following COSMIN methodology (Mokkink et al. 2020 ): defining the conceptual model and measurement constructs; identifying an appropriate type of measure; selecting and formulating measurement items and the scoring system; performing a pilot test in a small representative population to test the relevance, comprehensiveness, comprehensibility, acceptability, and feasibility of the measure; and field testing the measure in a larger representative population. These procedures occurred in five sub‐studies using multiple methods, as shown in Figure 1 . FIGURE 1. Open in a new tab Sub‐studies involved in the EoLC‐ATD development and validation. 4.2. Ethical Considerations Ethical approval for the study was provided by the University of Notre Dame Australia Human Research Ethics Committee (HREC) (Reference HREC 18/12/2017:017189S). In compliance with infection control regulations required by the Australian Government, Department of Health and Aged Care (Australian Government, Department of Health and Aged Care 2020 ) during the Covid‐19 pandemic, HREC approval for amended data collection procedures was obtained in the third year of the study (HREC 22/08/2019:017189S). Convenience sampling was used to recruit potentially eligible study participants in accordance with the ethical standards of the Helsinki Declaration of 1975 (World Medical Association 2013 ). Participant recruitment occurred through the issue of study advertisements and participant information statements and consent forms (PISCF) unique to the study populations involved in sub‐study 2 (Registered Nurses), sub‐study 3 (Delphi panel members), and sub‐studies 4 and 5 (persons living with advanced dementia). A participant information statement was provided to individuals who directly contacted author CB expressing an interest in joining the relevant sub‐study. Author xx provided verbal study information and a consent form to potentially eligible participants and to the legal guardian (proxy) of the person living with dementia. Study participants, including proxies for persons living with dementia, gave written consent for participation. A unique code number was allocated to each participant upon the issue of the signed consent form, which was subsequently recorded on the participant's data collection forms. Hard copies of the participants' consent forms and code numbers were stored separately in the locked office of author CB, and scanned copies of these documents were stored in electronic format on the dedicated study computer. Only author CB had access to confidential study documents. Although an expert dementia care group comprised of dementia carers, advocates, clinicians, an educator, and researchers provided guidance and advice to the researchers on each of the sub‐studies, these individuals were blinded to the identity of study participants. 4.3. Sub‐Studies The aims, settings, populations, methodology, and results of each of the five sub‐studies are presented in accordance with the COSMIN methodology (Mokkink et al. 2020 ), as follows. 4.3.1. Defining the EoLC‐ATD'S Conceptual Framework The conceptual framework which guided the development of the EoLC‐ATD constructs and items was a synthesis of the shared tenets of the palliative (Sawatzky et al. 2016 ) and person‐centred care (Kitwood and Bredin 1992 ) models. These models of care acknowledge that the onset of advanced dementia affects every aspect of the person's life—physical, psychological, familial, social and spiritual. Accordingly, the many symptoms experienced in advanced dementia involve these aspects of life, including those associated with a loss of independence and anticipation of suffering relative to mind, body and spirit, which necessarily impact on the individual's personhood (self‐hood) (Kitwood 1999 ). In general, advanced dementia symptom measures do not include items measuring personhood (Bentvelzen et al. 2017 ). The authors considered this to be an omission, since assessment of personhood in advanced dementia can assist in the prevention and relief of suffering according to the unique emotional, spiritual and practical needs of the person and their family (National Coalition for Hospice and Palliative Care 2018 ). Recognition of personhood was central to the study's conceptual framework, since it acknowledges the relational nature of care, regards family members as equal partners in decisions relating to symptom management, and seeks to sustain the person in a personalised care environment (Kitwood 1999 ). Thus, synthesis of the palliative and person‐centred models provided a useful conceptual framework for the study in recognition of the role that human interaction and relationships play in identifying and developing mutual understanding of advanced dementia symptoms (Kitwood 1999 ; National Coalition for Hospice and Palliative Care 2018 ). 4.3.2. Defining the EoLC‐ATD Constructs and Items To identify potentially suitable constructs and items for inclusion in the EoLC‐ATD items, two sub‐studies were undertaken: Sub‐Study 1 reviewed validated instruments that measured a range of advanced dementia symptoms; and Sub‐Study 2 undertook focus groups with Registered Nurses (RN) who routinely cared for persons living with advanced dementia in the aged care sector. 4.3.2.1. Sub‐Study 1: Compiling an Item Bank of Advanced Dementia Symptoms The study aimed to identify a comprehensive range of advanced dementia symptoms included in validated dementia symptom measures, to define potential EoLC‐ATD constructs and items. 4.3.2.1.1. Methodology An item bank of advanced dementia symptom constructs (domains) and related items was collated following an extensive review of existing dementia symptom measures by two of the authors (CB, LC). The review was guided by dementia symptom terminology, as described in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.) (DSM‐5) (American Psychiatric Association and DSM‐5 Task Force 2013 ) and the International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD 10) (World Health Organization 2010 ). Personhood characteristics were identified with reference to the Socio‐Psychological Theory of Personhood in Dementia (SPTPD) (Kitwood and Bredin 1992 ). The Dementia Outcomes Measurement Suite (DOMS) (Bentvelzen et al. 2017 ) measurement appraisal criteria were employed in reviewing validated dementia symptom measures (Box 1 ). BOX 1. Measurement appraisal criteria. Content validity: The content of the measure must be reported as being found appropriate for assessing dementia stage and/or symptom severity and the measure must not be designed for another specific purpose (e.g., psychosis in persons with schizophrenia), except if it has been shown to have clinical validity in dementia. Rating: Excellent—domain comprehensively sampled; Adequate—domain reasonably well‐sampled; Low—important aspects of domain not sampled. Psychometric validity (as reported in at least 1 peer‐reviewed publication): Evidence of the measure's inter‐rater reliability (Intra Class Correlation, Cohen's Kappa tests). Rating: Excellent (ICC/k 0.90); Adequate (ICC/k 0.70–0.89); Low (ICC/k < 0.70), or no data. Evidence of the measure's test–retest reliability in persons living with dementia (Intra Class Correlation, Cohen's Kappa tests). Rating: Excellent (ICC/k 0.90); Adequate (ICC/k 0.70 to 0.89); Low (ICC/k < 0.70), or no data. Evidence of expected correlations (concurrent validity) with similar validated measures (Pearson correlation coefficient, Cohen's Kappa tests). Rating: Excellent (jr/kj 0.70); Adequate (jr/kj from 0.40 to 0.69); Low (jr/kj < 0.30), or no data. Clinical validity: Quantitative evidence in support of the measure's discriminant validity in a clinical context relevant to dementia (e.g., in distinguishing persons with dementia from those without dementia) when assessed concurrently (e.g., sensitivity and specificity) and/or longitudinally (responsiveness) is reported. Rating: Excellent (can distinguish between > 2 clinically important categories); Adequate (can distinguish between 2 clinically important categories); Low (no evidence). Availability: The measure is publicly available for use by non‐specialist clinical staff. Rating: Excellent (no charge to acquire, use tool and/or for training in use of tool); Adequate (small one‐time costs to acquire, use tool and/or or for training in use of tool); Low (costs charged each time tool is acquired, used and/or for training in use of tool). Feasibility: The measure can be completed in < 35 min of administration time, and does not entail complicated scoring, training or administration procedures that would contraindicate use in a (typical) clinical context. Rating: Excellent (< 5 min); Adequate (6–15 min); Low (> 15 min). Open in a new tab 4.3.2.1.2. Results The initial literature search yielded 90 validated measures of advanced dementia symptoms, according to the categories (constructs) allocated within the DSM‐5, ICD‐10, and SPTPD, which included global cognition, behaviour/neuropsychiatric, activities of daily living, somatic/suffering, personhood, psychological, and physiological. After removal of duplicated symptoms within these seven constructs, an item bank of 108 advanced dementia symptoms was derived from 37 high‐scoring dementia symptom measures (Bentvelzen et al. 2017 ). These 108 advanced dementia symptoms aligned with the following constructs: global cognition (6 symptoms including significantly impaired memory, recognition, comprehension and language); behaviour/neuropsychiatric (28 symptoms including agitation, hallucinations and paranoia); activities of daily living (27 symptoms including all aspects of self‐care); personhood (6 symptoms including failure to express unique needs); somatic/suffering (5 symptoms including withdrawing and crying during care); psychological (13 symptoms including apathy, depression, contentment and happiness); and physiological (23 symptoms including significant weight loss and infections). 4.3.2.1.3. Conclusion Based on a review of the psychometric properties and clinical relevance (Bentvelzen et al. 2017 ) of the reviewed dementia symptom measures, seven constructs and 108 associated symptoms were identified and were suitable for inclusion in the EoLC‐ATD item bank. 4.3.2.2. Sub‐Study 2: Registered Nurse Focus Groups Sub‐Study 2 aimed to ascertain the perceptions of Registered nurses (RNs) on the identification and measurement of advanced dementia symptoms in practice. 4.3.2.2.1. Study Setting and Participants The settings for the focus groups were three dementia specific units (DSU's) that were housed in three government‐accredited residential aged care homes belonging to one Australian not‐for‐profit aged care provider. The aged care provider was subsidised by the Australian Government to provide a range of residential care services for older persons with/without dementia. Two of the three DSUs were located in a major city, and one was located in a country town. The 35‐bed DSUs were designed to provide a home‐like, accessible, and safe environment for people living with dementia. Focus groups were held in private meeting rooms, which were adjacent to the DSUs. Study participants were registered nurses (RN) who were caring for persons living with advanced dementia in these three DSUs. Convenience sampling was used to recruit 14 out of a possible 20 eligible RNs working in these DSUs, a sample size that was considered adequate to reach data saturation on the RNs' perceptions of advanced dementia symptoms and their measurement (Guest et al. 2017 ). Of the 14 consented RNs, 12 of them identified as female and two as male, 10 worked full‐time and four worked part‐time. Nine RNs were clinicians, three were care managers, one was a clinical nurse educator, and one was a residential care manager. All 14 RNs had tertiary qualifications in nursing, more than five years of experience in caring for persons with dementia in a DSU, and routinely used available instruments to assess and support symptoms in people living with advanced dementia. 4.3.2.2.2. Methodology Using a structured interview guide, authors CB, LC and AW held three separate focus groups with the RN participants (two groups of five each, and one group of four). The focus group questions included: (1) the advanced dementia symptoms identified by RNs when caring for persons with dementia; (2) the measures and procedures used by RNs to identify when a person has reached an advanced stage of dementia; and (3) the suitability of the seven constructs and 108 items included in the EoLC‐ATD item bank. Qualitative focus group data were audio recorded with participant permission, transcribed in Microsoft WORD, and thematically analysed by authors CB, LC and AW according to Braun & Clarke (Braun and Clarke 2006 ). Analysis included independent review of focus group transcriptions by these three authors to look for patterns and meanings in the data, tabulating participant responses for each of the focus group questions, independently allocating data codes to the data by analysing the participant quotes, and deriving themes and sub‐themes from allocated data codes. Through discussion, authors CB, LC and AW agreed on the final theme names. 4.3.2.2.3. Results Three themes and eight associated sub‐themes were derived from the RNs' responses to the three focus group questions (Table 1 ). TABLE 1. Registered nurse perceptions of advanced dementia symptoms. Question. ‘What do registered nurses perceive to be the signs and symptoms of advanced dementia’? Question. ‘How do registered nurses identify when a person living with dementia is in the advanced stage’? Question. ‘When registered nurses recognise that the person is in the advanced stage, how do they use this knowledge to plan care’? Theme Theme Theme Observed changes occurring in the person Use a collaborative process to identify changes in the person Engage in a continuous process of assessment, consultation with stakeholders, review and decision‐making Subthemes Decline in independent function related to self‐care ability Inability to communicate needs Changes from usual behaviour Deterioration in physical health Subthemes Everyone is alert to changes occurring in the person Discussion among members of staff, the General Medical Practitioner (GP) and family members/carers Subthemes Use the Residential Aged Care End of Life Care Pathway (RAC‐EoLCP) Use a range of assessment tools to identify care requirements Collaborate to make joint decisions on symptom support Open in a new tab RN participants emphasised that changes occurring in the person with advanced dementia were unique to the individual, since “… deterioration is very individualized and presentation is varied” [P0207]. Most participants expressed an ability to identify functional changes indicative of advanced dementia: “…you know the changes” [P0301], especially “if there is a drastic change” [P0303]. In the main, observed changes related to a decline in the person's self‐care ability, communication skills, behavioural responses and physical health, which signified to the RNs that the person had reached an advanced stage of dementia. A collaborative process was reportedly used to assess any deterioration in the usual abilities, function and health in persons with dementia. It was emphasised that changes experienced by the person “may be subtle” [P0204] and are often identified by care staff “during day‐to‐day care provision for the person” [P0102]. At times, the person's family/carers alerted the staff to changes in symptoms “unique to the individual” [P0207] that were identified when visiting the person. Participants indicated that all those involved in the person's treatment and care collaborated in identifying symptom changes, for example, in the person's behaviour or sleep patterns. Changes occurring in the person such as their physical health, weight, self‐care abilities, pain levels and emotions were regularly monitored in delivering end‐of‐life care. Symptom assessment was undertaken using a variety of measures that were available to RNs, including the “Residential Aged Care‐End of Life Care Pathway (RAC‐EoLCP) … the Comfort Chart …” [P0303] “pain scales” [P0204] and the “Stop and Watch” [P0202] observation form. RNs and care staff were required to chart and “… monitor symptoms over time, such as over a week” [P0304], which were discussed by all team members in planning care. The RNs unanimously confirmed the relevance of the constructs and most of the associated items included in the EoLC‐ATD item bank, which were frequently observed in persons living with advanced dementia. Participants reinforced the importance of educating care staff and RN colleagues in identifying the interplay between neurological impairment, physical health, psychosocial, and spiritual factors occurring in advanced dementia, including “existential thought…” (P0204). 4.3.2.2.4. Conclusion The focus group responses confirmed the content validity of the seven EoLC‐ATD databank constructs and 108 associated symptoms, in being considered relevant, comprehensive and comprehensible to RNs who routinely cared for and assessed symptoms in persons with advanced dementia (Streiner and Kottner 2014 ). Focus group findings also confirmed the face validity of the EoLC‐ATD item bank constructs and associated symptoms, as the RNs considered that most of the items represented the constellation of advanced dementia symptoms observed in their practice (Streiner and Kottner 2014 ). 4.3.3. Confirming the Suitability of EoLC‐ATD Constructs, Items and Scoring System Confirmation of the suitability of the EoLC‐ATD constructs, items, response options and instructions, and scoring decisions, occurred through a Delphi survey study with experts in dementia and/or palliative care. 4.3.3.1. Sub‐Study 3: Delphi Survey Study The Delphi Survey study aimed to develop consensus among an international group of dementia experts on the relevant constructs, items, and scoring of the EoLC‐ATD. 4.3.3.1.1. Setting and Participants The Delphi Survey study was conducted online with a convenient sample of Australian and international dementia and/or palliative care experts. The literature on Delphi panel numbers recommends a sample size of 25–30 panel members for all Delphi survey rounds to achieve consensus on survey items and account for any loss to follow up (Ogbeifun et al. 2016 ). The study statistician (author EG) calculated that a target sample size of 52 was required to account for loss to follow‐up, to achieve completion of all survey rounds by a minimum of 30 participants, and to reach consensus on EoLC‐ATD items. Convenience sampling was used to recruit Delphi survey panellists, who were identified through a review of publicly available profiles of dementia and/or palliative experts, referrals provided by these experts, and through inquiries to dementia advocacy and carer support services including Dementia Australia. Of the 52 dementia experts (family carers, medical and nursing clinicians, aged care managers and researchers) invited to participate, 31 individuals consented to completing all Delphi survey study rounds. 4.3.3.1.2. Methodology Three online Delphi surveys were undertaken to obtain panel member consensus on which of the EoLC‐ATD item bank's seven constructs and 108 associated symptoms would be suitable for inclusion in the EoLC‐ATD. The online surveys were developed and completed using Survey Monkey software (Survey Monkey 2018 ) according to the Conducting and REporting DElphi Studies (CREDES) approach (Jünger et al. 2017 ). In all Delphi survey rounds, panel members were asked to rate each of the EoLC‐ATD item bank constructs and symptoms using a 7‐point ordinal Likert scale—from Strongly Disagree = 1, Disagree = 2, Somewhat Disagree = 3, Neither = 4, Somewhat Agree = 5, Agree = 6, to Strongly Agree = 7. The percentage of agreement was quantified with each Delphi survey round. Only items that achieved ‘strongly agree/agree’ ratings were included in Delphi surveys 2 and 3. It was decided that panel member consensus was achieved when the predetermined threshold of ≥ 70% ‘agree/strongly agree’ was met for EoLC‐ATD constructs and items in each survey (Schifano and Niederberger 2025 ). Delphi panel members were invited to provide free text comments via an open‐ended section at the end of each survey, including suggestions on item wording and inclusion of commonly observed symptoms that were missing from the Delphi surveys. Quantitative data were entered into Microsoft Excel then transferred into IBM Statistical Package for the Social Sciences Statistics for Windows, Version 26 (IBM Corp 2019 ) for analysis. Participant demographic and clinical data were tabulated and analysed descriptively. Cronbach's alpha was applied to the Delphi panel member rankings of the eight EoLC‐ATD databank constructs and 108 symptoms to confirm internal consistency, that is, that similar items received similar agreement rates. Principal Components analysis was used to confirm data adequacy. Data adequacy was further confirmed using the Kaiser–Meyer–Olkin test and varimax rotation to interpret components. Scree plot and eigenvalues > 1 were used to decide which components were prominent. These analyses identified which of the constructs and associated symptoms were sufficiently inter‐correlated and to determine if the grouped items measured the underlying construct, to assist with organising items according to the relevant construct/s (Sullivan and Artino 2013 ). With the third Delphi survey, Cronbach's alpha was calculated to determine the internal consistency of the EoLC‐ATD constructs and items, and whether expected trends of ranking were observed and thus, to confirm that the constructs and items were correctly organised. The free text Delphi survey data provided by panel members were inspected and discussed by all four authors, to identify common recommendations for item inclusion or exclusion and item wording. Author agreement on non/acceptance of Delphi panel member item recommendations considered whether suggested rewording of an item made the meaning clearer, whether the item was included in high‐scoring dementia symptom measures, and whether the symptom (item) was frequently observed in practice, or not, as reported in the RN focus groups (Streiner and Kottner 2014 ). 4.3.3.1.3. Results Twenty‐four panel members identified as female and seven as male, having an average age of 55.9 years (range 30–76, SD 13.4). There were 28 Australians, one from the USA and two from the UK. Most ( n = 28) had more than 20 years of experience in dementia and/or palliative care as a clinician, manager, or clinical researcher, and three were family carers and advocates for persons living with dementia. All members held tertiary education qualifications, including PhD ( n = 8), Masters ( n = 9), Post Graduate Diploma ( n = 8), Bachelors ( n = 3) and Diploma ( n = 1). The combined scores of ≥ 70% for Strongly Agree and Agree ratings in the first Delphi survey resulted in panel member agreement on 66 of 108 symptoms included in the EoLC‐ATD item bank (range 70.9%–93.5%, mean 77.7%, SD 4.37). The highest combined ratings were allocated to symptoms relating to global cognition and personhood (ideation, decision‐making, memory) (range 23–29, 83%–93%), function in activities of daily living (continence, self‐care, eating) (range 20–29, 65%–94%), psychological state (low mood, upset, lost interest) (range 23–25, 29%–77%) and behaviour/neuropsychiatric (verbal and physical agitation, resistance to care) (range 9–24, 29%–77%). The second Delphi survey resulted in ≥ 70% Strongly Agree and Agree ratings for 39 of 66 items (Cronbach's alpha 0.876) in the constructs of global cognition (5 symptoms), personhood (4 symptoms), behaviour/neuropsychiatric (4 symptoms), function in activities of daily living (9 symptoms), physiological (11 symptoms), somatic/suffering (3 symptoms) and psychological (3 symptoms). The third Delphi survey achieved ≥ 70% Strongly Agree and Agree ratings for 24 of 26 items (Cronbach's alpha 0.881) in global cognition (3 symptoms), personhood (4 symptoms), behaviour/neuropsychiatric (2 symptoms), activities of daily living (7 symptoms), physiological (6 symptoms) and psychological (2 symptoms). Items rated ≤ 70% Strongly Agree/Agree included 1_1_1a ‘Unable to express personhood’ and 3_1_3a ‘Resistance to care’ (Table 2 ). TABLE 2. Delphi survey round 3 EoLC‐ATD items achieving ≥ 70% in combined ‘strongly agree and agree’ ratings. Construct Items Strongly agree and Agree ≥ 70%, n (%) Global cognition ( conscious mental processes that occur in the process of thinking, including identification, understanding and perception of knowledge ) 1_1_1a Unable to express personhood (personal identity, including life meaning, personality, spirit, character, history, interests) 15 (68.2) 1_1_2a Unable to focus/pay attention 19 (86.4) 1_1_3a Unable to retain new information 18 (81.8) 1_1_4a Confused (to place and person) 19 (86.4) Personhood ( autonomy / self‐determination in making one's own choices and decisions ) 2_1_1a Does not indicate their preferences for care 18 (81.8) 2_1_2a Does not engage in decisions regarding whether, or not, to take medications 16 (72.7) 2_1_3a Does not engage in decisions about treatment for a medical condition 18 (81.8) 2_1_4a Does not indicate the need to eat or drink 18 (81.8) Behaviour/neuropsychiatric ( way in which one acts or conducts oneself, especially towards others ) 3_1_1a *Apathetic (appears not engaged most of the time) 16 (76.2) 3_1_2a *Agitated (e.g., frequently requests attention or calling out, pacing, repetitive actions such as fidgeting or plucking) 15 (71.4) 3_1_3a Resistance to care (e.g., physically or psychologically resisting, pulling away, tightening limbs, stiffening, waving arms and legs and verbally objecting to care with words and/or sounds) 21 (66.7) Function ( ability to perform basic and instrumental activities of daily living ) 4_1_1a Unable to self‐feed 20 (95.3) 4_1_2a Unable to attend to personal hygiene 19 (90.5) 4_1_3a Unable to dress completely in the right way 18 (85.7) 4_1_4a Unable to self‐toilet without full assistance, unrelated to physical disability 18 (85.7) 4_1_5a Incontinent of urine 18 (85.7) 4_1_6a Incontinent of faeces 17 (80.9) 4_1_7a Bedfast, chairfast 22 (100.0) Physiological ( mechanical, physical, and biochemical function ) 5_1_1a Significant weight loss unrelated to physical illness (BMI less than 18.5 kg/m2) 20 (95.2) 5_1_2a Altered appetite and/or decreased food intake 20 (95.2) 5_1_3a Difficulty chewing 20 (95.2) 5_1_4a Choking on food and fluids 20 (95.2) 5_1_5a Altered thirst and decreased fluid intake 19 (90.6) 5_1_6a Pressure injury/or high risk 19 (95.0) Psychological ( state of mind including emotions/mood/suffering ) 6_1_1a Anguish (e.g., from pain, grief or fear) 15 (71.4) 6_1_2a Anxiety 16 (76.2) Open in a new tab 4.3.3.1.4. Conclusion The Delphi survey study confirmed the content and face validity of the EoLC‐ATD and resulted in panel member consensus of ≥ 70% for 24 of the 26 items included in EoLC‐ATD. Two items, 1_1_1a Unable to express personhood (personal identity, including life meaning, personality, spirit, character, history, interests) and 3_1_3a Resistance to care (e.g., physically or psychologically resisting, pulling away, tightening limbs, stiffening, waving arms and legs and verbally objecting to care with words and/or sounds), scored just below the ≥ 70% cut‐off rating. Further review of the literature and inspection of the RN focus group data confirmed that these symptoms frequently occur in advanced dementia (Sampson et al. 2017 ; Brennan et al. 2023 ; Royal Commission into Aged Care Quality and Safety 2021 ; Laver et al. 2016 ). Therefore, it was decided to retain these two EoLC‐ATD items for pilot testing (Streiner and Kottner 2014 ). 4.3.4. Testing the Relevance, Comprehensiveness, Comprehensibility, Acceptability and Feasibility of the EoLC‐ATD 4.3.4.1. Sub‐Study 4: EoLC‐ATD Pilot Test Sub‐study 4 aimed to test the 26‐item EoLC‐ATD's functionality and reliability in identifying advanced dementia symptoms, compared with validated measures that assessed similar constructs and items. 4.3.4.1.1. Study Setting and Participants The EoLC‐ATD was piloted in two of 11 Dementia Specific Units (DSU) that were housed in two aged care homes in a major city (as described in Section 4.3.2.2 ). As the pilot study took place during the Covid‐19 pandemic when government health regulations severely restricted researcher access to Australian aged care homes, convenience sampling was used. Permission was granted by the health authority to recruit all potentially eligible persons living in these two DSUs with a diagnosis of dementia, as recorded in their medical files. Eligibility assessment was undertaken by author CB and included having a Functional Assessment Staging Tool (FAST) (Reisberg et al. 1984 ) score of 6 or 7 (advanced dementia), not being on an end‐of‐life pathway, health status indicators suggesting that the person would likely remain alive in the next three months, and consent. Self‐consent and/or proxy consent from their legal guardian was obtained in writing for the person's participation, as outlined in Section 4.3.2.2 . 4.3.4.1.2. Methodology To test the relevance, comprehensiveness, comprehensibility, acceptability and feasibility of the 26‐item EoLC‐ATD, it was pilot tested alongside two validated measures which included many items that were similar to those included in the EoLC‐ATD. These measures were the Global Deterioration Scale (GDS) (Reisberg et al. 1982 ) which included items on cognition, behaviour/neuropsychiatric symptoms and function in activities of living, and the Mini Suffering State Examination (MSSE) (Aminoff 2004 ) which included items on personhood, psychological state and mood. Two Registered Nurses (RN) were pre‐trained by author CB in the administration of the EoLC‐ATD, GDS and MSSE, supported by a measurement instruction and scoring manual applicable to case studies identified by the RNs. The EoLC‐ATD, GDS and MSSE were then independently administered by the two RNs and author CB on the same day (day 1) to eight participants living with advanced dementia in the two DSUs. To test the repeatability of these measures, these data were independently collected 1 week later (day 8) for the same eight participants and on the same day by author CB and the two RAs. Participant demographic and clinical data were obtained from their electronic and hard‐copy medical records according to the HREC‐approved protocol. All data were recorded on hard copy data collection forms. Field notes were hand‐recorded by author CB to record the RNs' perceptions on the comprehensiveness, comprehensibility and feasibility of the EoLC‐ATD. 4.3.4.1.3. Statistical Analyses Demographic and clinical data, and GDS (Reisberg et al. 1982 ), MSSE (Aminoff 2004 ) and EoLC‐ATD scores, were analysed in IBM Statistical Package for the Social Sciences Statistics for Windows, Version 26 (IBM Corp 2019 ). Participant demographic and clinical data were analysed descriptively. Categorical variables were presented as frequencies (relative frequencies). For continuous variables, normality was confirmed with a combination of graphical representation and Shapiro–Wilk tests, due to the small sample size. Continuous variables were presented as means (standard deviation) if normally distributed, or median (1st, 3rd quartile) when normality was not met. Cohen's Kappa test was used to calculate inter‐rater reliability (IRR) of the EoLC‐ATD, GDS and MSSE scores. Test–retest reliability and repeatability were calculated with Chi‐square and Cronbach's alpha to assess EoLC‐ATD accuracy, and to determine the level of agreement between the EoLC‐ATD, GDS and MSSE scores allocated at the first and second data collection time points, with cut‐offs set at Cohen's kappa = 0.70 (Boateng et al. 2018 ). McNemar's test was used on paired nominal data to determine if the proportions of categories in the two sets of allocated scores significantly differed from each other. 4.3.4.1.4. Results Participant Demographics and Clinical Profile Eight of 11 potentially eligible persons with dementia consented to the pilot study. The mean age of the eight participants was 84.87 years (SD 8.965), two identified as male and six as female, and three were born overseas. Seven participants had a diagnosis of Alzheimer's disease and one had mixed dementia. Six had a FAST (Jünger et al. 2017 ) score of seven (7), two had a FAST score of six (6) and three were prescribed psychotropic medicines for neuropsychiatric symptoms. EoLC‐ATD Reliability and Validity At the first and second data collection time points, there was a substantial level of agreement between author CB and the RNs for GDS (Reisberg et al. 1982 ) scores (Cohen's kappa = 0.75) and for MSSE (Aminoff 2004 ) scores (Cohen's kappa = 0.75). There was a substantial or good level of agreement for 20 of the 26 EoLC‐ATD items (Cohen's kappa = 0.77), slight or fair agreement for four items and no agreement for two items (Table 3 ). There was good test–retest reliability of the EoLC‐ATD as a whole (Pearson's r = 0.82, Cronbach's alpha > 0.85). TABLE 3. EoLC‐ATD inter‐rater reliability (IRR) scores. Construct Item Cohen's Kappa Interpretation Cognition 1a. Unable to express personhood 1.00 Perfect agreement 1b. Unable to focus/pay attention ≤ 0 No agreement 1c. Unable to retain new information 1.00 Perfect agreement 1d. Confused (to place and person) 0.21 Slight agreement Personhood 2a. Does not indicate their preferences for care 1.00 Perfect agreement 2b. Does not engage in decisions whether, or not, to take medicines 1.00 Perfect agreement 2c. Does not engage in decisions about treatment 1.00 Perfect agreement 2d. Does not indicate the need to eat or drink 0.59 Moderate agreement Behaviour 3a. Apathetic 0.21 Slight agreement 3b. Agitated 0.09 No agreement beyond chance 3c. Resistance to care 0.59 Moderate agreement Function 4a. Unable to self‐feed 0.70 Substantial agreement 4b. Unable to attend to personal hygiene 0.70 Substantial agreement 4c. Unable to dress completely and correctly 1.00 Perfect agreement 4d. Unable to self‐toilet without full assistance, unrelated to physical disability 1.00 Perfect agreement 4e. Incontinent of urine 1.00 Perfect agreement 4f. Incontinent of faeces 1.00 Perfect agreement 4g. Bedfast, chair‐fast 0.70 Substantial agreement Physiological 5a. Significant weight loss 1.00 Perfect agreement 5b. Altered appetite/decreased food intake 0.37 Fair agreement 5c. Difficulty chewing 0.70 Substantial agreement 5d. Choking on food and fluids 1.00 Perfect agreement 5e. Altered thirst and decreased fluid intake 1.00 Perfect agreement 5f. Pressure injury/or high risk 0.72 Substantial agreement Psychological 6a. Anguish 1.00 Perfect agreement 6b. Anxiety 0.24 Fair agreement Open in a new tab The validity of the EoLC‐ATD constructs and associated items was established with Cronbach's alpha for the constructs ranging from 0.856 to 0.968. There was internal consistency between the EoLC‐ATD items (Cronbach's alpha 0.965), and excellent reliability of the EoLC‐ATD as a whole (Cronbach's alpha > 0.85). Registered Nurse Perceptions The RNs advised that use of the EoLC‐ATD had relevance in identifying advanced dementia symptoms, and that it was comprehensive and comprehensible and likely to be acceptable to RNs in the aged care sector. The RNs did not indicate that there were too many items; rather they recommended that an additional item be included at the end of the measure ‘7. Other symptoms (please describe)’, so that nurses could record unique symptoms not included in the EoLC‐ATD. Through discussion, all four authors agreed to this recommendation, since the aim of developing the EoLC‐ATD was for nurses to more accurately identify presenting symptoms for future planning of symptom‐responsive palliative care. 4.3.4.1.5. Conclusion The 26‐item EoLC‐ATD had excellent reliability and repeatability, and was identified by the RNs to be relevant, comprehensive, comprehensible, and acceptable in measuring advanced dementia symptoms. In response to the RNs' recommendation, an additional item for recording ‘other symptom’ was included in the EoLC‐ATD after piloting, so that aged care RNs could record unique symptom/s that were not included in the existing 26 items. 4.3.5. Field Testing the EoLC‐ATD in a Larger Representative Population 4.3.5.1. Sub‐Study 5: EoLC‐ATD Validity Testing The aim of field testing the 27‐item EoLC‐ATD was to determine its accuracy in identifying advanced dementia symptoms in a representative sample of persons living with advanced dementia. 4.3.5.1.1. Study Setting and Participants Sub‐study 5 was conducted in 11 Dementia Specific Units (DSU), including the two pilot study DSUs. Each of the DSUs was located in one of 11 care homes belonging to the Australian aged care provider, as described in Sections 4.3.2.2 and 4.3.4 . Six of the DSUs were located in geographically diverse suburbs of a major city, three in regional country areas and two in small rural towns. Convenience sampling was used to recruit all potentially eligible persons living with a diagnosis of dementia in the 11 study DSUs, as recorded in their medical files. Eligibility included having a FAST (Reisberg et al. 1984 ) score of 6 or 7, not being on an end‐of‐life care pathway, health status indicators suggesting that the person would likely remain alive for the next 3 months, and consent. The sample size calculation was informed by the literature on measurement testing in vulnerable health populations (Boateng et al. 2018 ). The literature suggested that while a sample of 10 participants per measurement item is an ideal number, frequently resource constraints and limited population availability in clinical settings make this unattainable (Boateng et al. 2018 ). The study statistician (author EG) calculated that a minimum sample of 100 persons with advanced dementia (with complete EoLC‐ATD scores) would be required to determine whether the EoLC‐ATD was predictive of death at 180 days after baseline assessment, to achieve statistical power (80%) and an effect size of 10% in the primary study outcome (advanced dementia), and to achieve the mean score and standard deviation of the EoLC‐ATD that were previously assessed in pilot study participants with similar FAST scores (Swan et al. 2023 ). 4.3.5.1.2. Methodology As occurred with the pilot study, author CB provided training to 17 of the DSU RNs in the administration and scoring of the FAST (Reisberg et al. 1984 ) and the EoLC‐ATD. The RNs screened 136 potentially eligible participants with the FAST, using clinical assessment and observation, electronic and hard‐copy medical file review, and held discussions with direct care staff on the person's current symptoms and remaining abilities. Demographic and clinical data, including weight, body mass index (BMI), pressure injury risk, and Psychogeriatric Assessment Scales (PAS) (Jorm et al. 1997 ) score were obtained by the RNs for consented participants by reviewing their electronic and hard‐copy medical files. The EoLC‐ATD was administered to study participants by the RNs at baseline (on recruitment to the study). Mortality data were recorded at 90 days post‐baseline and 180 days post‐baseline. 4.3.5.1.3. Statistical Analysis The same statistical procedures used with the EoLC‐ATD pilot study data were applied to analyse participant demographic and clinical data (Section 4.3.4 ). Receiver Operator Characteristic (ROC) curve analysis was used to explore the diagnostic ability of the EoLC‐ATD against death at 90 days post‐baseline and 180 days post‐baseline. ROC curve analysis determined the most appropriate cut‐off point of the EoLC‐ATD score that was able to identify individuals at high risk of significant decline (i.e., death within 90 days and at 180 days post‐baseline), by using the ROC curve coordinates of Sensitivity and 1‐Specificity for all possible scores. Binary logistic regression was used to explore the role of the EoLC‐ATD score to independently predict mortality status at 90 days post‐baseline and 180 days post‐baseline, as a proxy of disease progression. Analysis was adjusted for participant age and sex (Australian Institute of Health and Welfare (AIHW) 2022 ). On the basis of their clinical significance in advanced dementia, the confounders PAS (Jorm et al. 1997 ), BMI, and English language skills were also included in the analysis (Sampson et al. 2017 ). Statistical significance was set at alpha = 0.05 and analysis was performed on valid data. As these data are routinely recorded in Australian residential care homes for funding purposes, there were very few missing data. The analysis was performed on complete sets of data and no imputation was made for missing data. 4.3.5.1.4. Results Of 113 participants, the mean age was 86.7 years (range 67–107 years), 76 (67%) identified as female and 37 (33%) as male, 28 (25%) were born overseas, 85 (75%) were Australian‐born and 107 (95%) spoke English as their primary language. Alzheimer's disease was the most common dementia type ( n = 88, 78%), 64 (57%) had a FAST (Reisberg et al. 1984 ) score of level 6 (moderately severe dementia), 47 (43%) had a FAST score of level 7 (severe dementia) and 60 (53%) were prescribed psychotropic medicines for neuropsychiatric symptoms. Forty‐three (38%) participants were either chair or bed fast and 60 (53.1%) had a very high risk of pressure injury. The 27‐item EoLC‐ATD performed well in identifying symptoms of advanced dementia in these participants. The EoLC‐ATD total mean score was 15.3 out of a possible score of 27 (range 8–26), while based on ROC curve analysis, the cut‐off with the best coordinate was equal to 14.50, sensitivity was 0.833 and specificity was 0.386. A score of 14.50 or greater out of 27 accurately identified advanced dementia symptoms in the study sample (Area Under the ROC curve [AUC] = 0.769, 95% CI: (0.670, 0.868, p < 0.001). This result confirms that the EoLC‐ATD performed well in identifying advanced dementia symptoms and high‐risk of death at 180 days (6 months) post‐baseline) (Table 4 ). TABLE 4. Area under the ROC Curve coordinates. Area under the curve (AUC) Test result variable(s): EoL_ATD_Score_Final Area Std. error a Asymptotic sig. b Asymptotic 95% confidence interval Lower bound Upper bound 0.769 0.051 0.000 0.670 0.868 Open in a new tab Note: The test result variable(s): EOL_Score_Final has at least one tie between the positive actual state group and the negative actual state group. Statistics may be biased. a Under the nonparametric assumption. b Null hypothesis: true area = 0.5. Of 113 participants, 18 (15.9%) died within 90 days post‐baseline and 30 (26.5%) died within 180 days post‐baseline. Mortality was significant at 180 days post‐baseline ( p < 0.001); AUC = 0.769. The best performing cut‐off point for sensitivity/specificity was found at the'ye' score of 14.5 (out of 27). Of the 30 (26.5%) participants who died within 180 days post‐baseline, the FAST (Sullivan and Artino 2013 ) score was 7 (advanced dementia) and EoLC‐ATD score was high in the areas of function (unable to perform activities of daily living, 93%), global cognition (unable to focus and engage in decision‐making, 76%), physiological (high risk of pressure injury, 76%) personhood (no expressed needs, 73%) and behaviour/neuropsychiatric (resistance to care, 70%). Multivariable binary regression analysis confirmed that the EoLC‐ATD score was significant in predicting death at 180 days post‐baseline (per unit increase =1.213; 95% CI: 1.044, 1.408, p = 0.011), after adjusting for participant age and sex (Table 5 ). TABLE 5. Multivariable logistic regression model. Step 1 a B SE Wald df Sig 95% CI for XP(B) Exp(B) Lower Upper Age_final 0.049 0.035 1.995 1 0.158 1.051 0.981 1.125 EoLC_score_final 0.311 0.071 19.439 1 < 0.001 1.365 1.189 1.567 Sex 0.879 0.535 2.701 1 0.100 2.409 0.844 6.877 Constant −11.555 3.718 9.658 1 0.002 0.000 Open in a new tab a Variable(s) entered on step 1: Age_final, EoLC_score_final, Sex. 4.3.5.1.5. Conclusion Field testing of the EoLC‐ATD showed it to be a valid and reliable measure of advanced dementia symptoms. The 30 (26.5%) participants who died within 180 days of baseline EoLC‐ATD measurement had higher EoLC‐ATD scores for all areas of function, global cognition, personhood, physiological, and behaviour. These higher EoLC‐ATD scores aligned with the FAST scores at level 7, which indicates that the higher the EoLC‐ATD score, the more severe the dementia symptoms and the higher likelihood of death within 6 months. 5. Discussion As a measure of advanced dementia symptoms that is suitable for use by Registered Nurses (RN) in persons living with dementia, the EoLC‐ATD plays a crucial role in progressing evidence‐based palliative care in dementia. RNs are faced with the issue of multidimensionality in being able to readily identify when a person is living with advanced dementia symptoms, since these symptoms encompass an array of bio‐psychosocial factors which are frequently unique to the person (Brennan et al. 2023 ; Goodman et al. 2015 ). Although several ‘gold standar’ dementia symptom measures exist (Bentvelzen et al. 2017 ), including the FAST (Reisberg et al. 1984 ), the review of available measures revealed the difficulty of comprehensively capturing the different symptoms occurring in advanced dementia in one unidimensional measure. This was particularly apparent when searching for measures which included the important concept of personhood (Kitwood and Bredin 1992 ), as well as the psychosocial and physiological features of advanced dementia (Sampson et al. 2017 ; Brennan et al. 2023 ). The review identified a bias, or incomplete representation, of these constructs in most of the single measures reviewed. By contrast, the EoLC‐ATD was developed as a single measure to comprehensively assess commonly occurring symptoms through qualitative and quantitative methods, including with RNs for whom the measure was designed. Following the COSMIN (Mokkink et al. 2020 ) taxonomy, field testing the EoLC‐ATD by RNs who routinely worked in the dementia care setting showed this methodological approach to be helpful. The EoLC‐ATD was shown to be reliable, valid and responsive to each of the bio‐psychosocial constructs which were included; it achieved inter‐rater and test–retest reliability, face and content validity, and internal consistency and specificity. The EoLC‐ATD was highly responsive to the construct of interest, that is, advanced dementia, in the study sample (Boateng et al. 2018 ). These test results recommend the EoLC‐ATD as a measure of choice for the aged care sector, since inadequate measurement of advanced dementia symptoms may not only result in policies that support and fund unhelpful treatment options, neglectful and/or inefficient palliative care practices may follow (Goodman et al. 2015 ; Laver et al. 2016 ). Having access to an appropriately designed and statistically robust measurement, such as the EoLC‐ATD, is essential in ensuring timely and accurate assessment of advanced dementia symptoms (Brennan et al. 2023 ; National Coalition for Hospice and Palliative Care 2018 ). Of perhaps even greater importance is that the EoLC‐ATD was tested and validated in the aged care setting by RNs who would be expected to employ this measure as part of the care planning involved in advanced dementia symptom support. The test results indicate that these RNs were able to accurately identify symptoms of advanced dementia with the EoLC‐ATD and make meaning of the scores in their practice (Swan et al. 2023 ). There is currently no other composite dementia measure which includes global cognition, personhood, behaviour/neuropsychiatric, function (activities of daily living), physiological, psychological, and somatic symptoms. The EoLC‐ATD, therefore, will avoid the necessity of administering several symptom‐specific and physiological measures to identify advanced dementia symptoms. Another advantage is that the EoLC‐ATD scores will assist in identifying the person's unique symptom support requirements and associated palliative care needs (Sawatzky et al. 2016 ). As dementia has an unpredictable trajectory with considerable variability in its progression (Alzheimer's Society, UK 2022 ), an individual's advanced dementia symptoms may not be recognised and well supported without regular monitoring with the EoLC‐ATD (Sampson et al. 2017 ; National Coalition for Hospice and Palliative Care 2018 ). Personhood, for example, often receives inadequate attention in advanced dementia care. This occurs when family carers, RNs, and care staff are unsure how to help the person maintain their personhood with the progression of dementia (Kitwood 1999 ). Thus, administration of the EoLC‐ATD at different stages of the advanced dementia trajectory will potentially benefit not only the person living with advanced dementia, but also those providing care, in its ability to identify personhood factors unique to the person and therefore, in knowing which of these require support (Kitwood and Bredin 1992 ). As well, the inclusion of psychological items in the EoLC‐ATD will assist RNs to be better informed about these important aspects of person‐centred palliative care (National Coalition for Hospice and Palliative Care 2018 ), especially when it is difficult to interpret the person's bodily and facial expressions in the absence of verbal language (Laver et al. 2016 ). Where this occurs, the person's unique psychosocial symptoms may be inadvertently neglected (Brennan et al. 2023 ; Kitwood 1999 ), resulting in a tendency to focus on tangible signs of advanced dementia such as verbal and physical indicators of pain, discomfort, dyspnoea, and physiological dysfunction or breakdown (Sampson et al. 2017 ; Goodman et al. 2015 ). This has been recognised as an issue of concern in the aged care sector, particularly when the failure to identify unique advanced dementia symptoms results in neglect of vital care support at end‐of‐life (Alzheimer's Society, UK 2022 ; Royal Commission into Aged Care Quality and Safety 2021 ). Critically, the EoLC‐ATD cut‐off score of 14.5 (out of a possible 27) aligned with the FAST (Reisberg et al. 1984 ) score of level 7, which identifies many of the symptoms occurring in the most advanced stage of dementia. The advantage of the EoLC‐ATD is that it includes items on personhood, and on physiological and psychological status. These items are not included in the FAST, which primarily screens for cognitive and functional decline. Where a person with dementia is assessed to have a score of 14.5 and higher on the EoLC‐ATD, RNs will be in a better position to anticipate likely mortality within 180 days of baseline assessment. By sharing the EoLC‐ATD score with family carers, the care team and medical practitioners, RNs will be able to facilitate discussions on ways to support the person's symptom‐specific end‐of‐life care requirements (Sampson et al. 2017 ; Brennan et al. 2023 ; Boyd and Murray 2010 ). 5.1. Study Strengths and Limitations 5.1.1. Strengths A study strength is the alignment of the EoLC‐ATD with international efforts to improve palliative care of persons living with dementia in aged care homes (Alzheimer's Society, UK 2022 ; Brennan et al. 2023 ; Royal Commission into Aged Care Quality and Safety 2021 ). From the study's inception, an eight‐member Expert Advisory Group contributed to the development of the study methodology and procedures, and they provided confirmation of the EoLC‐ATD's relevance in advanced dementia care. Another study strength was advice on the EoLC‐ATD contributed by RN participants who were intimately associated with care of the person living with advanced dementia (Boateng et al. 2018 ; Jorm et al. 1997 ). The RN contributions included valuable insights on the limitations of currently available advanced dementia measures, the practical requirements of a comprehensive advanced dementia symptom measure for use by RNs, and the value of including global cognition, personhood, behaviour, function, psychological, and physiological symptoms in the EoLC‐ATD. Involving RNs with considerable dementia care experience in the pilot study and the validity testing of the EoLC‐ATD confirmed its clinical relevance (Streiner and Kottner 2014 ). 5.1.2. Limitations A methodological limitation resulting from government regulations during the COVID‐19 pandemic (Australian Government, Department of Health and Aged Care 2020 ) was the small sample size ( n = 8) of persons living with suspected advanced dementia able to be recruited to the EoLC‐ATD pilot study, and government restrictions on the potential number of DSU participants suitable for EoLC‐ATD field testing. Thus, the study DSUs and participants were not necessarily representative of the expansive landscape of Australian and international dementia‐specific long‐term care services, which include for‐profit, not‐for‐profit, and government providers. Another potential methodological limitation was applying the cut‐off of ≥ 70 to achieve Delphi survey consensus for the 26‐item EoLC‐ATD. The authors weighed up the benefit and cost of the cut‐off being over and under ≥ 70 by considering the clinical value of the EoLC‐ATD (Laver et al. 2016 ). It was decided that as the EoLC‐ATD is a single measure, it is more likely to be used to screen and monitor a range of advanced dementia symptoms, rather than using two or more measures for these different symptoms. Thus, a cut‐off of ≥ 70 would not be likely to adversely impact care planning to support identified symptoms (Streiner and Kottner 2014 ; Boateng et al. 2018 ). As some of the EoLC‐ATD constructs and associated items had a greater impact on the total score (e.g., function) than others (e.g., behaviour), overall specificity was low (0.386), which may influence the accuracy of the EoLC‐ATD in predicting death at 90‐ and 180‐days post‐baseline. However, as the measure's main purpose is to identify advanced dementia symptoms, low specificity is not as crucial when scores are used as an aid to care planning (Jorm et al. 1997 ). 5.2. Recommendations for Future Research Further testing of the EoLC‐ATD with a larger sample of persons living with dementia in different contexts and over time is warranted, to confirm whether the EoLC‐ATD consistently identifies advanced dementia symptoms in persons from culturally, socially, and linguistically diverse backgrounds. Given the methodological limitations identified, future studies could undertake further validation of the EoLC‐ATD, including confirmatory factor analysis and responsiveness. Future research could also include alignment of palliative care planning with identified EoLC‐ATD symptoms in the person living with advanced dementia, with the aim of evaluating symptom relief, wellbeing, and quality of life, and evaluating nursing staff's knowledge and skills in identifying, planning, and delivering person‐centred palliative care that is responsive to unique advanced dementia symptoms. 5.3. Implications for Policy and Practice Validation of the 27‐item EoLC‐ATD has addressed the urgent need for an advanced dementia symptom measure that is suitable for use by registered nurses in the aged care sector. Information obtained with the EoLC‐ATD provides clarity on the person's unique advanced dementia symptoms that require support and thus, justification for harnessing suitable resources in the aged care sector to meet those needs. 6. Conclusion This study developed and validated an End‐of‐Life Care Assessment Tool for Dementia (EoLC‐ATD) using a multimethod approach in five sub‐studies, underpinned by a person‐centred palliative care conceptual framework. The EoLC‐ATD will support informed and collaborative decision making on advanced dementia symptom assessment and inform palliative and end‐of‐life care management, and it has potential to promote better care outcomes for people living with advanced dementia. Thus, the EoLC‐ATD represents a promising start to ensuring person‐centred palliative care for the person living with advanced dementia. The study also lays the foundation for future studies to further validate the EoLC‐ATD and to develop a care planning component that is responsive to EoLC‐ATD‐identified symptoms. Author Contributions All authors have agreed on the final version and meet at least one of the following criteria (recommended by the ICMJE*): (1) substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data; (2) drafting the article or revising it critically for important intellectual content. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Data S1: jan70129‐sup‐0001‐DataS1.docx. JAN-82-5430-s001.docx (34.3KB, docx) Acknowledgements The authors acknowledge the contributions of the Expert Advisory Group members, the study participants and the Registered Nurses who participated in piloting and field testing the End‐of‐Life Care Assessment Tool for Dementia (EoLC‐ATD). Open access publishing facilitated by University of New South Wales, as part of the Wiley ‐ University of New South Wales agreement via the Council of Australian University Librarians. Bourke, C. , Chenoweth L., Georgousopoulou E., and Williams A.. 2026. “Development and Validation of the End‐of‐Life Assessment Tool for Advanced Dementia: A Multi Method Study.” Journal of Advanced Nursing 82, no. 5: 5430–5443. 10.1111/jan.70129. Funding: The authors received no specific funding for this work. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. References Abernethy, A. P. , Shelby‐James T., Fazekas B. S., Woods D., and Currow D. C.. 2005. “The Australia‐Modified Karnofsky Performance Status (AKPS) Scale: A Revised Scale for Contemporary Palliative Care Clinical Practice.” BMC Palliative Care 12, no. 4: 1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Alzheimer's Society, UK . 2022. “End of Life Care for a Person With Dementia: Alzheimer's.” https://www.alzheimers.org.uk/get‐support/help‐dementia‐care/end‐life‐care . American Psychiatric Association , and DSM‐5 Task Force . 2013. Diagnostic and Statistical Manual of Mental Disorders: DSM‐5. 5th ed. American Psychiatric Publishing, Inc. 10.1176/appi.books.9780890425596. [ DOI ] [ Google Scholar ] Aminoff, B. Z.
2004. “Mini Suffering State Examination (MSSE) Scale: First 5 Years.” IMAJ 6, no. 10: 645–646. [ PubMed ] [ Google Scholar ] Australian Government, Department of Health and Aged Care . 2020. “National COVID‐19 Aged Care Plan.” 7th ed. https://www.health.gov.au . Australian Institute of Health and Welfare (AIHW) . 2022. “Admissions Into Aged Care: 2021–2022.” https://www.genagedcaredata.gov.au/Topics/Admissions‐into‐aged‐care . Bentvelzen, A. , Aerts L., Seeher K., Wesson J., and Brodaty H. A.. 2017. “Comprehensive Review of the Quality and Feasibility of Dementia Assessment Measures: The Dementia Outcomes Measurement Suite.” JAMA 18, no. 10: 826–837. [ DOI ] [ PubMed ] [ Google Scholar ] Boateng, G. O. , Neilands T. B., Frongillo E. A., Melgar‐Quiñonez H. R., and Young S. L.. 2018. “Best Practices for Developing and Validating Scales for Health, Social, and Behavioral Research: A Primer.” Frontiers in Public Health 6: 149. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Boyd, K. , and Murray S. A.. 2010. “Recognising and Managing Key Transitions in End‐of‐Life Care.” BMJ (Clinical Research Ed.) 341: c4863. [ DOI ] [ PubMed ] [ Google Scholar ] Braun, V. , and Clarke V.. 2006. “Using Thematic Analysis in Psychology.” Qualitative Research in Psychology 3, no. 2: 77–101. [ Google Scholar ] Brennan, F. , Chapman M., Gardiner M. D., Narasimhan M., and Cohen J.. 2023. “Our Dementia Challenge: Arise Palliative Care.” Internal Medicine Journal 53, no. 2: 186–193. [ DOI ] [ PubMed ] [ Google Scholar ] Goodman, C. , Froggatt K., Amador S., Mathie E., and Mayrhofer A.. 2015. “End of Life Care Interventions for People With Dementia in Care Homes: Addressing Uncertainty Within a Framework for Service Delivery and Evaluation.” BMC Palliative Care 17, no. 14: 42. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Guest, G. , Namey E., and McKenna K.. 2017. “How Many Focus Groups Are Enough? Building an Evidence Base for Nonprobability Sample Sizes.” Field Methods 29, no. 1: 3–22. 10.1177/1525822X16639015. [ DOI ] [ Google Scholar ] IBM Corp . 2019. IBM SPSS Statistics for Windows, Version 26.0. IBM Corp. [ Google Scholar ] Jorm, A. F. , Mackinnon A. J., Christensen H., Henderson A. S., Jacomb P. A., and Korten A. E.. 1997. “The Psychogeriatric Assessment Scales (PAS): Further Data on Psychometric Properties and Validity From a Longitudinal Study of the Elderly.” International Journal of Geriatric Psychiatry 12, no. 1: 93–100. 10.1002/(sici)1099-1166(199701)12:1<93:aid-gps468>3.0.co;2-c. [ DOI ] [ PubMed ] [ Google Scholar ] Jünger, S. , Payne S. A., Brine J., Radbruch L., and Brearley S. G.. 2017. “Guidance on Conducting and REporting DElphi Studies (CREDES) in Palliative Care: Recommendations Based on a Methodological Systematic Review.” Palliative Medicine 31, no. 8: 684–706. [ DOI ] [ PubMed ] [ Google Scholar ] Kitwood, T.
1999. “Dementia Reconsidered. The Person Comes First.” BMJ 318: 880. 10.1136/bmj.318.7187.880a. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Kitwood, T. , and Bredin K.. 1992. “Towards a Theory of Dementia Care: Personhood and Well‐Being.” Ageing and Society 12: 269–287. [ DOI ] [ PubMed ] [ Google Scholar ] Laver, K. , Cumming R. G., Dyer S. M., et al. 2016. “Clinical Practice Guidelines for Dementia in Australia.” Medical Journal of Australia 204, no. 5: 191–193. [ DOI ] [ PubMed ] [ Google Scholar ] Mokkink, L. B. , Terwee C. B., Knol D. L., et al. 2020. “The COSMIN Checklist for Evaluating the Methodological Quality of Studies on Measurement Properties: A Clarification of Its Content.” BMC Medical Research Methodology 10: 22 2010. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] National Coalition for Hospice and Palliative Care . 2018. National Consensus Project for Quality Palliative Care. 4th ed. Clinical Practice Guidelines for Quality Palliative Care. https://www.nationalcoalitionhpc . [ Google Scholar ] Ogbeifun, E. , Agwa‐Ejon J., Mbohwa C., and Pretorius J. H. C.. 2016. “The Delphi Technique: A Credible Research Methodology. Proceedings of the International Conference on Industrial Engineering and Operations.” In Proceedings of the 2016 International Conference on Industrial Engineering and Operations Managemen t. Kuala Lumpur, Malaysia, March 8–10, 2016. Reisberg, B. , Ferris S., de Leon M., and Crook T.. 1982. “The Global Deterioration Scale for Assessment of Primary Degenerative Dementia.” American Journal of Psychiatry 139, no. 9: 1136–1139. [ DOI ] [ PubMed ] [ Google Scholar ] Reisberg, B. , Ferris S. H., Anand R., et al. 1984. “Functional Staging of Dementia of the Alzheimer Type.” Annals of the New York Academy of Sciences 435: 481–483. 10.1111/j.1749-6632.1984.tb13859.x. [ DOI ] [ Google Scholar ] Royal College of General Practitioners . 2011. The Gold Standards Framework: NHS End of Life Care Gold Standards Framework (GSF) Prognostic Indicator Guidance: To Aid Identification of Adult Patients With Advanced Disease, in the Last Months/Year of Life, Who Are in Need of Supportive Palliative Care. 4th ed. Gold Standards Framework National Team. http://www.goldstandardsframework.nhs.uk/ . [ Google Scholar ] Royal Commission into Aged Care Quality and Safety . 2021. “Final Report Volume 1.” Commonwealth of Australia. https://www.royalcommission.gov.au/system/files/2021‐03/final‐report . Sampson, E. L. , Candy B., Davis S., et al. 2017. “Living and Dying With Advanced Dementia: A Prospective Cohort Study of Symptoms, Service Use and Care at the End of Life.” Palliative Medicine 32, no. 3: 668–681. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Sawatzky, R. , Porterfield P., Lee J., et al. 2016. “Conceptual Foundations of a Palliative Approach: A Knowledge Synthesis.” BMC Palliative Care 15, no. 15: 5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Schifano, J. , and Niederberger M.. 2025. “The Delphi Technique in the Health Sciences.” In Handbook of Concepts in Health, Health Behavior and Environmental Health, edited by Liamputtong P.. Springer. 10.1007/978-981-97-0821-5_132-1. [ DOI ] [ Google Scholar ] Streiner, D. L. , and Kottner J.. 2014. “Recommendations for Reporting the Results of Studies of Instrument and Scale Development and Testing.” Journal of Advanced Nursing 70, no. 9: 1970–1979. 10.1111/jan.12402. [ DOI ] [ PubMed ] [ Google Scholar ] Sullivan, G. M. , and Artino A. R.. 2013. “Analyzing and Interpreting Data From Likert‐Type Scales.” Journal of Graduate Medical Education 5, no. 4: 541–542. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Survey Monkey . 2018. “Free Surveys.” https://uk.surveymonkey.com/Free/Survey . Swan, K. , Speyer R., Scharitzer M., et al. 2023. “Measuring What Matters in Healthcare: A Practical Guide to Psychometric Principles and Instrument Development.” Frontiers in Psychology 14: 1225850. 10.3389/fpsyg.2023.1225850. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] World Health Organization . 2010. International Statistical Classification of Diseases and Related Health Problems (ICD) 10th Revision. World Health Organization. https://icd.who.int . [ Google Scholar ] World Health Organization . 2023. “Dementia.” https://www.who.int/news‐room/fact‐sheets/detail/dementia . World Medical Association . 2013. “Declaration of Helsinki. Medical Research Involving Human Subjects.” https://www.wma.net/what‐we‐do/medical‐ethics/declaration‐of‐hel… . Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Data S1: jan70129‐sup‐0001‐DataS1.docx. JAN-82-5430-s001.docx (34.3KB, docx) Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. Articles from Journal of Advanced Nursing are provided here courtesy of Wiley ACTIONS View on publisher site PDF (513.1 KB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top