Abstract
Background
Dementia is a leading cause of death among residents in care home settings internationally. Proactive identification of those approaching end-of-life can support future care planning, respecting preferences. We aimed to synthesise existing literature on prognostic tools designed to predict 6-month mortality in individuals with dementia residing in care homes including their content and prognostic performance.
Methods
A systematic review was undertaken, searching Embase, CINAHL, PsycINFO and MEDLINE in December 2024. Studies evaluating prognostic tools for predicting 6-month mortality in individuals with dementia living in care homes were included. Risk of bias was assessed using the PROBAST tool. Data were extracted and narratively synthesised.
Results
Thirteen studies involving 503 501 participants were included from a total of 5438 records screened. Eleven prognostic tools were identified. Only three studies were assessed at low risk of bias. Predictive factors per tool ranged from 4 to 12. Age, change in cognitive function, functional decline/dependency and concerns around nutritional intake/weight were the commonest domains of predictors across included tools. Reporting around the discriminative performance of tools was limited and only two tools (Mortality Risk Index Score and Advanced Dementia Prognostic Tool had acceptable discrimination (area under receiver operating curve >0.70), which was not replicated in validation studies.
Conclusions
While identified predictors have clinical congruence, there is significant variability in how these are assessed and operationalised and in the application of prognostic tools. Specific prediction of mortality remains challenging and would benefit from further research to adapt and validate dynamic prognostic tools for use in practice.
Keywords: dementia, long-term care, prognosis, palliative and end of life care, care home, systematic review, older people
Key Points
Our review identified significant heterogeneity in prognostic tools for people living with dementia in care homes
Two tools using data from the minimum data set were identified with acceptable discrimination, which has not been externally replicated
There has been little new research in dementia prognostication over the past decade
Further research is needed to adapt and externally validate a dynamic prognostic tool which could be used in this population
Introduction
Dementia is a progressive neurodegenerative disease affecting ~57 million people worldwide, with estimates of a growing global population of people living the condition [1]. As the disease advances, individuals often require additional support with activities of daily living (ADL), including continence, personal care and management of neuropsychiatric symptoms [2]. The majority of people living with dementia are also living with multiple long-term conditions and this is associated with poorer outcomes [3].
While there is a lack of consensus around the optimal time for moving-in to a care home setting for individuals living with dementia, [4] the condition itself is strongly associated with the need for care placement [5, 6]. A recent systematic review estimated that individuals diagnosed with dementia can spend up to a third of their remaining life expectancy in a nursing home setting [7]. Thus, dementia is common in people living in care home settings [8, 9]. As the number of people living with dementia grows, so the numbers dying with it are estimated to increase significantly, [10, 11] commonly in care home settings [12]. To-date, dementia of all causes accounts for over one-third of all deaths among care home residents in Scotland [13].
It is well-documented that those with dementia experience barriers to accessing high quality palliative care. Reasons include a widespread failure to recognise dementia as a terminal illness, [14, 15] fragmented access to healthcare services [16, 17] and lack of access to specialist palliative care support [18]. Timely identification of individuals in the last months of life is essential to avoid unnecessary hospital admissions and medical interventions that may be burdensome [19, 20]. Recent data from London supports that where palliative care needs are prospectively identified, there is a lower risk of multiple unscheduled hospitalisations and greater engagement of primary and community care support [21]. However, prognostication in dementia remains an area of challenge in clinical practice due to the clinical trajectory, impact of other long-term conditions and organisation of palliative care services which remain focused around other conditions [15].
Prognostication among adults living with chronic disease is an area of active interest for clinicians and researchers [22, 23]. A 2012 systematic review considered 6-month prognosis in people living with advanced dementia across all settings, concluding there was a lack of prognostic concordance across the literature [24]. A 2020 scoping review of prognostic variables in dementia identified 239 factors across six thematic areas which had been examined in the published literature and variation in predictive factors based on the setting of care [25]. This review grouped hospices and care home settings together in their analysis of individual factors [25]. Despite most dementia-related deaths taking place in care homes, no previous review has evaluated the performance of 6-month prognostic tools for dementia in long-term care settings.
The aim of this systematic review was to establish what tools exist to identify 6-month prognosis among people living with dementia in care homes and to assess their content and performance.
Methods
This review was reported following Preferred Reporting of Items in Systematic Reviews and Meta-Analyses (PRISMA) guidance [26]. The protocol was registered on PROSPERO in 2024 (CRD: 42024623940: https://www.crd.york.ac.uk/PROSPERO/view/CRD42024623940).
Eligibility criteria
Studies reporting on prognostic models were included. These were based on observational cohort studies both prospective and retrospective in design.
We placed no restrictions on date or language of publication. If abstracts were identified, we searched for subsequent full-text publications using author names and abstract titles. If no full text publication was identified, the abstract was included, with limited data extraction and without risk of bias assessment.
Population
People living in care homes who have a diagnosis of dementia of any subtype. We included dementia however this was defined, by study authors, but did not include those diagnosed with mild cognitive impairment. Where studies included people living with and without dementia, we assessed eligibility for inclusion in this review based on whether results data was presented about those living with dementia separately.
Outcomes
Identification of prognostic models/tools for 6-month mortality among people living with dementia, including a summary of their accuracy and performance. The review was interested in the performance of the model/tool including the number and range of predictors, discrimination, calibration and any other relevant statistical summaries.
Other pre-specified outcomes of interest were the applicability of models to specific subgroups: based on dementia sub-type and the applicability of evidence for individuals with other comorbidities alongside dementia—e.g. cancer, vascular disease and other neurodegenerative disease.
Timing
Six month prognosis has been previously identified as clinically significant due to the association with funding for hospice care in some jurisdictions [27, 28] and is a meaningful period for people and their families. We also explored the timing of when a prognostic model was applied for a person living in a care home setting with dementia. We were interested in identifying tools which could be used in practice to identify a resident was approaching their last months of life.
Setting
Our study focused on people living in care homes. Recognising heterogeneity in the terminology, this was inclusive of synonymous long term care settings internationally providing 24-hour residential care and support for adults. [29, 30] Where studies included a wider population of people living with dementia, beyond care homes, we included data from the study where results data relating to the care home population was available separately.
Information sources
We searched MEDLINE (Ovid); EMBASE (Ovid), CINAHL (EBSCOhost) and PsycInfo (EBSCOhost) from 1980 to December 2024. The search was developed involving three concepts—people living with dementia (‘population’), prognostic tools for mortality (‘concept’) in care home settings (‘context’). The full strategy is reported in Appendix S1, it was designed drawing on other published review search strategies [6, 31–34]. We checked reference lists of identified systematic or scoping reviews to identify any additional studies.
Study selection and data collection
We used Covidence software [35] for deduplication and study selection. Title/abstract screening was undertaken by independent pairs of reviewers (EW, LM, PP and JKB). Conflicts were resolved by team discussion, involving senior authors (JKB and LM). Full text screening followed the same approach.
Data extraction and risk of bias assessment
We developed an electronic data extraction form and piloted this on a sample of four papers, making adaptations for consistency and completeness. Independent data extraction by two authors (EW and LM) was undertaken for all full-text papers, and a single author (JKB) undertook data extraction for those where only the abstract data was available.
Data were extracted on: year of publication; country/countries; study design; number of care homes; description of care home setting; number of participants; age; sex; race/ethnicity; diagnostic criteria for dementia; dementia subtype; severity score to define advanced dementia; other comorbidities; causes of death; follow-up duration; 6-month survival; prediction model; data source; statistical modelling; number of predictors; list of predictors; timing of prediction; discrimination; calibration; other statistics reported and any comparison to other scores or tools.
A narrative synthesis of results is presented due to the significant heterogeneity between included studies. We grouped predictors into common domains to look at their distribution across the included studies. We found two types of predictive tools: tools specifically designed for predicting mortality in people living with dementia in care homes and a second group of tools designed for broader use in predicting mortality outcomes applied to a care home population. We reported where studies had evaluated discrimination and calibration and used recognised cut-off points to describe the level of discrimination based on area under receiver-operating-characteristics curves [23].
Risk of bias was evaluated by a single author (JKB) using the PROBAST tool which is specific for use in prediction model studies [36]. Four assessments of risk of bias are made individually (across the domains of participants, predictors, outcome and analysis) then three assessments of applicability are made about the included participants, predictors and outcomes. These are then considered together to present an overall judgement for each study. We used the risk of bias Excel template developed by Fernandez-Felix et al. to enable generation of a graphical summary of findings [37]. Risk of bias assessment was undertaken for full-text studies only, due to the lack of detail available from conference abstracts alone.
Results
We screened 5438 records by title and abstract after automated de-duplication. Five additional studies were identified from the reference lists of similar review articles. We assessed 65 studies in full and included 13 publications, three in abstract form only. We excluded 52 studies with main reasons for exclusion including n = 11 no prediction model reported; n = 9 wrong condition (results not available for people with dementia separately); n = 8 wrong outcome (not 6-month mortality); n = 8 not original research. All reasons for exclusion are reported in Figure 1. Of the eight studies which looked at prediction of a different mortality time-point, these included 14 days, [38] 30 days, [39] 1 year, [40, 41] 19 months, [42] 2 years [43, 44] and 5 years [45].
Figure 1.
PRISMA diagram [46].
Included study characteristics
The review includes ten original articles [47–56] and three abstracts [57–59], included study characteristics are summarised in Table 1. Included studies were mainly of cohort design, with six being retrospective [52, 53, 55, 57–59]. Nine studies were conducted in North America, [48–54, 56, 58] with one in both North America and the Netherlands [55]. All were situated in high-income countries.
Table 1.
Included study characteristics: published papers and conference abstracts.
| Study ID | Country | Sample size | Number of care homes | Age in years (mean +/− SD unless stated) | Female sex (%) | How dementia diagnosed and included subtype(s) | Severity inclusion criteria | 6-month survival (%) | Risk of bias and applicability |
|---|---|---|---|---|---|---|---|---|---|
| Published papers | |||||||||
| Esteban-Burgos 2023 [47] | Spain | 149 146 | 8 | 84.7 +/− 9.3 | 70.6 | Not reported | Not reported | 74.8 | R: high A: high |
| Hicks 2010 [48] | USA | 123 | 3 | 81.5 +/− 7.1 | 55.3 | Diagnosis of dementia in medical record All subtypes: 57.7% Alzheimer’s, 42.3% othersa |
Met hospice criteria for dementia or were receiving hospice or palliative care | 68.0 | R: high A: high |
| Hirdes 2014 [49] | Canada | 118 429 | Not reported | Age distribution: 2.6% <65, 7.4% 65–74, 32.7% 75–84, 57.4% >84 |
68.3 | Resident Assessment Instrument (RAI 2.0) diagnosis of Alzheimer’s and related dementias | Not reported | Not reported | R: high A: high |
| Marsh 2000 [50] | USA | 112 | 5 | 82.0 +/− 7.7 | 75.0 | Diagnosis of Alzheimer’s dementia | GDS >6 | 80.4 | R: high A: high |
| McCusker 2014 [51] | Canada | 180 | 7 | >65 | 56.0 | Diagnosed by a physician on resident chart Subtypes not reported |
Not reported | 89.0 | R: high A: high |
| Mitchell 2004 [52] | USA | 11 430 (6799 derivation; 4631 validation) |
1082 (642 derivation; 440 validation) |
Median 83 | 66.8 | Diagnosis of Alzheimer’s and other dementias in MDS data | Cognitive performance score 5–6 | 68.96% (71.7% derivation; 64.9% validation) |
R: low A: low |
| Mitchell 2010(a) [53] | USA | 22 405 | Not reported | 84.5 +/− 7.5 | 77.0 | Diagnosis of dementia in MDS data | Cognitive performance score 5–6 (from MDS) | Not reported | R: low A: low |
| Mitchell 2010(b) [54] | USA | 606 | 21 | 85.0 +/− 7.0 | 81.9 | Dementia diagnosed in medical record | Cognitive performance score 5–6 (assessed prospectively) | 81.7 | R: high A: high |
| Van der Steen 2007 [55] | Netherlands and USA | 557 (288 Dutch; 269 USA) | 41 (6 Dutch; 35 USA) | Proportion > 83 Dutch 56.6%; USA 62.8% |
74.7 | Diagnosis of Alzheimer’s and other dementias in MDS data | Cognitive performance score 5–6 | 69.7 (Dutch 75.7; USA 63.2) |
R: low A: low |
| Volicer 1993 [56] | USA | 139 (development 68; testing 71) |
3 | 67.5 +/− 0.8 | 3 | Neurologist clinical evaluation of Alzheimer’s dementia using DSM 3 criteria | Not reported | 66.2 | R: high A: high |
| Conference abstracts | |||||||||
| Almeida 2012 [57] | Not reported | 24 | 1 | Not reported | Not reported | Dementia as cause of death Subtype not specified |
Not reported | Not reported | Not assessed |
| Voldberding 2013 [58] | USA | 236 | 1 | Not reported | Not reported | Not reported | Not reported | 36.9 | Not assessed |
| Wong 2018 [59] | Hong Kong | 114 | Not reported | Not reported | Not reported | Clinical management system data Subtype not specified |
Not reported | 50.0 | Not assessed |
Abbreviations: DSM—Diagnostic and Statistical Manual; SD—standard deviation.
R—risk of bias; A—applicability (full reporting of risk of bias assessment included in Appendix S2 and Appendix S3).
aVascular dementia, mixed dementia, Lewy body dementia, frontotemporal dementia, dementia type not specified, dementia due to HIV, dementia due to Parkinson disease.
Eight studies were undertaken in nursing homes, four were in settings described as long-term care facilities, institutions or homes [49–51, 56] and one in a dementia special care centre [58]. Across ten studies, residents in 1172 care homes were represented, three studies did not report the total number of care homes they included [49, 53, 59].
Included participant characteristics
The total review population includes 503 501 people. Individuals study sample sizes varied from 24 to 222 405 participants (median = 236, mean = 38 730). Average sample size is significantly skewed by three large studies of >100 000 participants [47, 49, 52].
Demographics
The average (typically reported as a mean) age of participants was between 67.5 and 85 years. Three studies did not report average age, with one providing distribution in age bands, [49] one specifying all had to be >65 years [51] and one reporting the proportion aged >83 years [55]. Women accounted for the largest proportion of participants in all studies apart from one, where they accounted for just 3% [56]. The authors of the latter study did not provide further explanation on their imbalanced population. Only six studies reported on participant ethnicity and those from white ethnic groups accounted for the majority of all six studies [48, 50, 52–54, 56].
Dementia
Dementia severity was assessed in six studies. This included meeting hospice criteria for dementia, [48] a Global Deterioration Scale (GDS) greater than six [50] and four studies who defined severity based on Cognitive Performance Score of five or six [52–55].
Two studies included only patients with Alzheimer’s type dementia, [50, 56] five studies enrolled patients with Alzheimer’s and other dementia types [48, 49, 52–54]. The remaining six studies did not define the subtypes of dementia included in their study. Only one study presented quantitative data on the proportion of their sample living with Alzheimer’s dementia versus other dementia subtypes [48].
Comorbidity
Three studies used formal tools to quantify comorbidity—one used the weighted index of comorbidity [50] and two others used the Charlson Index [51, 59]. Two studies did not measure comorbidity among included participants [57, 58]. Of the remaining studies, approaches to comorbidity reporting included five studies which reported the prevalence of a range of conditions, [47, 48, 52, 53, 56] while three other studies took a specific focus on conditions of interest including lower respiratory tract infection, [55] congestive heart failure, [54] or eleven neurological conditions [49].
Timing of prediction tool application
Five studies used time of enrolment into a research project as the timing of prognostication [47, 48, 50, 54, 59]. Three of these focused on enrolling those thought to be approaching end of life, based on meeting hospice criteria [47, 48] or clinician judgement [59]. Three studies applied mortality prediction at the time of nursing home admission [51–53]. Two studies used retrospective review of cases who died and the timing of tool application is unclear from the abstract data available [57, 58]. One study evaluated prognostic factors for survival following an episode of fever [56]. One study used the most recent assessment data recorded using the Resident Assessment Instrument, the study authors anticipated this assessment would have been made within the last 3 months [49]. The final study applied different approaches to tool application, with one cohort meeting a clinical definition of advanced dementia and the other in residents with advanced dementia who had experienced lower respiratory tract infection [55].
Follow-up and mortality
Participants were followed up for a duration between six and 71 months (average 23.6 months).
None of the studies reported causes of death. Survival at 6 months ranged from 36.9%–89.0% (average 68.5%). Three studies did not report 6-month survival [49, 53, 57].
Risk of bias within studies
Three studies were considered overall to have low risk of bias and low concerns around applicability [52, 53, 55]. Whereas the remaining seven were considered at overall high risk of bias and high concern around applicability [47–51, 54, 56]. These are summarised in Appendices S2 and S3.
Issues in the participant domain largely related to retrospective study designs and selection bias around the involvement of people living with dementia in care home settings. This was commonly due to recruitment concerns where individual/proxy consent to participate was needed. Predictor assessment was at low risk of bias across all studies, largely as predictors were evaluated using consistent national datasets or standardised researcher assessments. Outcome assessment methods varied, but this domain was also considered at low risk of bias across all studies due to the nature of the outcome being assessed as binary and concrete. The most significant concerns were around the analysis approach adopted and included selection of predictors based on univariate performance, the lack of evaluation of model performance including accounting for overfitting and optimism in performance.
Prognostic tools
Six studies report on the development of a mortality risk prediction tool [48, 50–53, 56]. Two studies report on validations of existing mortality risk prediction tools [49, 55] and two studies compare the performance of existing mortality risk prediction tools in the care home population [47, 54]. Of the three publications in abstract version only, two examined the performance of existing tools [57, 58] and one described developing a new tool [59].
The prognostic tools examined in the included studies, and the predictor variables included are summarised in Table 2. The degree of overlap between tools in core domains are presented in Figure 2. This is split to display domains included in the seven tools specifically designed for predicting mortality in people living with dementia in care home settings, compared to the four tools with broader use. No tool included all domains. Age and change in cognitive function were the most common domains included in eight tools followed by functional decline or dependency and oral intake/nutrition/weight loss, both examined in seven tools.
Table 2.
Summary of prognostic tools and their included predictors.
| Study ID | Prognostic tool evaluated | Data source to derive predictors | Number of predictors | Predictors included in the tool | |
|---|---|---|---|---|---|
| Almeida 2012 [57] Mitchell 2004 [52] van der Steen 2007 [55] Volberding 2013 [58] |
Mortality risk index score | MDS | 12 | ADL scale Age > 83 years Bedbound Bowel incontinence Cancer Congestive heart failure |
Male sex Not awake most of the day Shortness of breath Unstable medical condition Use of supplementary oxygen <25% of food eaten at most meals |
| Esteban-Burgos 2023 [47] | NECPAL 4.0 | Original data collection | 7 | Age Disease specific criteria of severity (For dementia global deterioration score > 6) Functional decline |
Multimorbidity Nutritional decline Palliative care needs identified by healthcare professional Use of resources |
| Palliative prognostic index (PPI) | Original data collection | 5 | Delirium Dyspnoea at rest Oedema |
Oral intake Palliative performance score |
|
| PROFOUND Index | Original data collection | 9 | Age > 85 Active neoplasia Barthel index <60 Caregiver other than spouse Dementia present Delirium in last hospitalisation |
Grade III–IV functional class on NYHA and/or MRC classifications Haemoglobin <100 g/l ≥4 hospitalisations in last 12 months |
|
| Hicks 2010 [48] | Model not named | Care of Nursing Home Residents with Advanced Dementia (CareAD) study | 6 | Age Dementia subtype Gender Pneumonia in prior 6 months |
Severe Impairment Rating Scale (SIRS score) 7+ years with dementia |
| Hirdes 2014 [49] | Changes in Health, End-Stage Disease and Signs and Symptoms (CHESS) Scale | inteRAI assessment data from home care and continuing care | 4 | Clinician determined prognosis <6-months Decline in ADLs |
Decline in cognition Weight loss, shortness of breath or oedema present |
| Marsh 2000 [50] | Alzheimer’s Hospice Placement Evaluation (AHOPE) scale | Original data collection | 9 | Ambulation Eye contact Food intake Fluid intake Level of consciousness |
Muscle flexibility Speech Swallowing Weight history |
| McCusker 2014 [51] | Model not named | Original data collection | 7 | Age Charlson Comorbidity Index Delirium Dementia |
Depression Sex Time since admission |
| Mitchell 2010(a) [53] Mitchell 2010(b) [54] |
Advanced Dementia Prognostic Tool (ADEPT) | MDS | 12 | Admitted <90 days previously Age Bed bound Body mass index Bowel incontinence Congestive heart failure |
Dyspnoea Insufficient intake Male sex Pressure ulcers Total functional dependence Weight loss |
| Volicer 1993 [56] | Model not named | Original data collection | 4 | Age Dementia severity (Bedford Alzheimer Nursing Scale) |
Management strategy (Level of care e.g. antibiotics vs palliative) Time from admission |
| Wong 2018 [59] | Model not named | CMS (abbreviation not defined) | 7 | Age Charlson Comorbidity Index Estimated Glomerular Filtration Rate Number of hospitalisations in last 6 months |
Serum albumin Sex Tube feeding |
Abbreviations: interRAI—international Resident Assessment Instrument; MRC—Medical Research Council; NYHA—New York Heart Association.
Figure 2.
Mapping predictor variable domains used in included tools.
Included studies poorly reported the performance of their models in terms of discrimination and calibration, with less than half providing area under curve (AUC) statistics and only four reporting on calibration (Table 3, Appendix S4). Only two models [Mortality Risk Index (also known as the MDS Risk Index) and ADEPT] had AUC values in the acceptable range (>0.70) for 6-month mortality [52, 53]. The Mortality Risk Index has been applied by others, including a joint US and Dutch validation study, in collaboration with the index authors [55]. However, the discrimination of the model in these cohorts was poor and lower than the original development study [55]. We did not identify other external validations of the ADEPT and note that the prospective validation study undertaken by the original authors demonstrated lower discrimination both when a cut-off score or continuous variable were used [54]. In the original study, using a cut-off score for ADEPT resulted in area under receiver operating curve (AUROC) 0.65 in the whole cohort and 0.59 among residents living in the home for at least 12 months [53].
Table 3.
Reporting of discrimination and calibration across included studies.
| Study ID | Prognostic tool | Discrimination of model | Evaluation of calibration |
|---|---|---|---|
| Marsh 2000 [50] | Alzheimer’s Hospice Placement Evaluation (AHOPE) scale | Not reported | Not reported |
| Mitchell 2010(a) [53] | Advanced Dementia Prognostic Tool (ADEPT) | AUROC 0.73 | Calibration plots provided |
| Mitchell 2010(b) [54] | Advanced Dementia Prognostic Tool (ADEPT) | AUROC 0.67 (95% CI 0.62–0.72) | Hosmer-Lemeshow test result (P 0.69) |
| Hirdes 2014 [49] | Changes in Health, End-Stage Disease and Signs and Symptoms (CHESS) Scale | Not reported | Not reported |
| Almeida 2012 [57] | Mortality Risk Index Score | Not reported | Not reported |
| Mitchell 2004 [52] | Mortality Risk Index Score | Derivation: AUROC 0.74 Validation: AUROC 0.70 Sensitivity and specificity data present for each cut-off point of the score |
Not reported |
| Van der Steen 2007 [55] | Mortality Risk Index Score | Dutch cohort AUROC 0.65 (95% CI 0.58–0.72) US cohort AUROC 0.64 (95% CI 0.58–0.71) |
Assessed by comparing observed mortality over risk strata in the development cohort with observed mortality in the study cohorts |
| Volberding 2013 [58] | Mortality Risk Index Score | Not reported | Not reported |
| Esteban-Burgos 2023 [47] | NECPAL (NECesidades PALiativas) 4.0 | AUROC 0.41 | Not reported |
| Palliative Prognostic Index (PPI) | AUROC 0.48 | Not reported | |
| PROFOUND Index | AUROC 0.66 | Not reported | |
| Hicks 2010 [48] | Model not named | Not reported | Not reported |
| McCusker 2014 [51] | Model not named | Not reported | Not reported |
| Volicer 1993 [56] | Model not named | Not reported | Not reported |
| Wong 2018 [59] | Model not named | AUROC 0.63 | Hosmer-Lemeshow test result (>0.05) |
Abbreviation: CI—confidence interval.
Subgroup analyses
Of our pre-specified outcomes of interest, no studies reported data separately based on dementia sub-type. However, one study included mixed or other dementia (versus Alzheimer’s dementia) as a variable in their predictive model, finding those with Alzheimer’s dementia and pneumonia had shorter survival than for other dementia subtypes [48].
Similarly other comorbidities were included in predictive models with congestive heart failure in two models, [52, 53] cancer in two models, [47, 52] depression in one model [51] and multimorbidity or comorbidity included in three models [47, 51, 59].
Discussion
Findings in context
Our systematic review identifies that there is no optimal prognostic tool to identify 6-month mortality among people living with dementia in care homes. There is significant heterogeneity across the included studies. The two tools (Mortality Risk Index and ADEPT) developed by Mitchell et al. in 2004 [52] and 2010 [53] respectively offer the most acceptable performance and use data from the minimum data set (MDS; structured standardised data-collection approach used internationally). However, external validation has either not replicated original model performance or not been undertaken, meaning further work is required before considering use in routine practice. This includes the use of cut-off scores and how to operationalise the tool, issues present at time of development [53]. The literature thus has not significantly advanced since the 2013 systematic review, across care settings, undertaken by Brown et al. [24]
The predictor variables used within tools have clinical congruence in identifying those with more advanced disease—advanced age, worsening cognition, dependency and concerns around oral intake/weight. They also align with predictor variables identified across non-communicable, non-cancer disease prediction models [22]. However, operationalising these for use in clinical practice is challenging and risks overlooking the holistic needs of an individual dying with dementia [60].
A gradually declining illness trajectory for people with dementia has been described in the literature [61]. The role of acute intercurrent illnesses, such as infections or fractures, to alter prognosis for an individual living with dementia are also increasingly recognised [62]. Pneumonia is a specific prognostic factor which has been identified previously in this population, [63] and for which a prediction model has been developed, albeit for 14-day mortality [38]. Risk prediction in this population requires to be dynamic to changing circumstances in which factors, such as infection, fever and difficulties with eating are common and associated with increased 6-month mortality [64].
Strengths and limitations
This study benefits from structured conduct based on a pre-specified protocol and use of good practice in undertaking systematic review research. It addresses an applied clinical topic and area of uncertainty in an under-researched population. However, we have been unable to perform any quantitative synthesis based on the heterogeneity of tools identified and limited data using the same tool in different populations.
However, by focusing on those living in care homes we have excluded evidence about the population living with advanced dementia cared for elsewhere in the community and in hospital settings. This led to exclusion of a model developed in Taiwan, across the whole population living with dementia, [65] a model derived on those presenting at in acute care [66] and a deep learning algorithm developed on hospital inpatients [67]. Our focus on the care home population was intended both to consider those with complex needs and to identify if sector-specific tools exist making use of data collected in care home settings, which may not be consistently present for those elsewhere, such as the MDS or Inter-RAI (used as the source for predictors in several included studies).
We were interested in those living in care homes with dementia, but this approach excludes those living with other long-term conditions. Focusing primarily on dementia can be justified based on the prevalence of the condition, but note that some studies have looked at predicting care home mortality as a whole, based around tools such as the FRAIL-NH [68] or Mortality Risk Index [69]. These tools would also benefit from exploration in care home settings internationally.
Implications for clinical practice and future research
For clinical audiences the message is clear that there is uncertainty in predicting mortality in this population which needs to be managed in care interactions. A recent scoping review identified that only ~40% of caregivers recognised dementia as a life-limiting condition and that their understanding of the condition is associated with the care experienced by their loved-one during the dying phase, [70] thus ensuring caregivers are informed and supported is critical. Jurisdictions providing palliative care support for those felt to be at risk of mortality demonstrate beneficial impacts on service utilisation measures [71]. However, while the components of palliative care interventions for this population have been identified, [72] delivering this in a consistent way remains a challenge and active area for research [73]. Consistent delivery of palliative care interventions may be improved if a well-performing and usable prognostic model for 6-month mortality was routinely available.
This review highlights the ongoing evidence gap around a high-performing, externally validated prognostic tool for 6-month mortality among people living with dementia in care homes. We note with interest a 2020 protocol to develop a tool for use in Australian Residential Aged Care Facilities [74]. More research is needed to determine the models which perform best in clinical practice, beyond their original settings of development. Future prognostic model studies need to be clear on how variables are defined and measured, when models are applicable and reduce bias through inclusion of more representative cohorts. For countries not using established data capture mechanisms, such as the MDS, using digital care record data collected in care home settings offers the potential to develop tools from routinely-collected data [75]. This could enable modelling of the impact of acutely changing health (as a consequence of dementia or other comorbid long term conditions) on prognosis and survival.
Supplementary Material
Contributor Information
Emily West, Academic Geriatric Medicine, School of Cardiovascular and Metabolic Health, University of Glasgow, Glasgow, Scotland, UK; Department of Medicine for the Elderly, Glasgow Royal Infirmary, Glasgow, Scotland, UK.
Laura Mulligan, Department of Medicine for the Elderly, Glasgow Royal Infirmary, Glasgow, Scotland, UK; School of Medicine, Dentistry & Nursing, University of Glasgow, Glasgow, UK.
Prabin Paudyal, Medicine for Older Adults, University Hospital Hairmyres, East Kilbride, UK.
Terence J Quinn, Academic Geriatric Medicine, School of Cardiovascular and Metabolic Health, University of Glasgow, Glasgow, Scotland, UK; Department of Medicine for the Elderly, Glasgow Royal Infirmary, Glasgow, Scotland, UK.
Jennifer Kirsty Burton, Academic Geriatric Medicine, School of Cardiovascular and Metabolic Health, University of Glasgow, Glasgow, Scotland, UK; Medicine for Older Adults, University Hospital Hairmyres, East Kilbride, UK.
Declaration of Conflicts of Interest:
None.
Declaration of Sources of Funding:
JKB is supported by a Joint NHS Education for Scotland/Chief Scientist Office (NES/CSO) Postdoctoral Clinical Lectureship (PCL/21/01).
Funding from the Vivensa Foundation Academy Excellence Award (EA2402_51) was used to facilitate this review.
The funders played no part in the conduct or writing of this review.
Data Availability:
In addition to the data available in the supplementary materials, the underlying data supporting this review is available on reasonable request through the corresponding author.
References
- 1. Nichols E, Steinmetz JD, Vollset SE et al. Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the global burden of disease study 2019. Lancet Public Health 2022;7:e105–25. 10.1016/S2468-2667(21)00249-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Risco E, Cabrera E, Jolley D et al. The association between physical dependency and the presence of neuropsychiatric symptoms, with the admission of people with dementia to a long-term care institution: a prospective observational cohort study. Int J Nurs Stud 2015;52:980–7. 10.1016/j.ijnurstu.2015.02.013. [DOI] [PubMed] [Google Scholar]
- 3. Stirland LE, Choate R, Zanwar PP et al. Multimorbidity in dementia: current perspectives and future challenges. Alzheimers Dement 2025;21:e70546. 10.1002/alz.70546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Cole L, Samsi K, Manthorpe J et al. Is there an “optimal time” to move to a care home for a person with dementia? A systematic review of the literature. Int Psychogeriatr 2018;30:1649–70. 10.1017/S1041610218000364. [DOI] [PubMed] [Google Scholar]
- 5. Luppa M, Luck T, Weyerer S et al. Prediction of institutionalization in the elderly. A systematic review. Age Ageing 2010;39:31–8. 10.1093/ageing/afp202. [DOI] [PubMed] [Google Scholar]
- 6. Harrison JK, Walesby KE, Hamilton L et al. Predicting discharge to institutional long-term care following acute hospitalisation: a systematic review & meta-analysis. Age Ageing 2017;46:547–58. 10.1093/ageing/afx047. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Brück CC, Mooldijk SS, Kuiper LM et al. Time to nursing home admission and death in people with dementia: systematic review and meta-analysis. BMJ 2025;388:e080636. 10.1136/bmj-2024-080636. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Honinx E, van Dop N, Smets T et al. Dying in long-term care facilities in Europe: the PACE epidemiological study of deceased residents in six countries. BMC Public Health 2019;19:1199. 10.1186/s12889-019-7532-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Burton JK, Lynch E, Love S et al. Who lives in Scotland’s care homes? Descriptive analysis using routinely collected social care data 2012–16. J R Coll Physicians Edinb 2019;49:12–22. 10.4997/jrcpe.2019.103. [DOI] [PubMed] [Google Scholar]
- 10. Etkind SN, Bone AE, Gomes B et al. How many people will need palliative care in 2040? Past trends, future projections and implications for services. BMC Med 2017;15:102. 10.1186/s12916-017-0860-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Finucane AM, Bone AE, Etkind S et al. How many people will need palliative care in Scotland by 2040? A mixed-method study of projected palliative care need and recommendations for service delivery. BMJ Open 2021;11:e041317. 10.1136/bmjopen-2020-041317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Teggi D, Dixon J, Dickson M. Demographic, health, and social predictors of place of death in England, 2004–2013: identifying barriers to dying in the community. J Aging Soc Policy 2025;37:972–95. 10.1080/08959420.2025.2482298. [DOI] [PubMed] [Google Scholar]
- 13. MacRae J, Ciminata G, Geue C et al. Mortality among long-term care residents: a retrospective cohort study using national linked social care and health data. BMJ Support Palliat Care 2025;16:188–97. 10.1136/spcare-2024-005163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Mitchell S, Kiely D, Hamel M. Dying with advanced dementia in the nursing home. Arch Intern Med 2004;164:321–6. 10.1001/archinte.164.3.321. [DOI] [PubMed] [Google Scholar]
- 15. Harrison KL, Hunt LJ, Ritchie CS et al. Dying with dementia: underrecognized and stigmatized. J Am Geriatr Soc 2019;67:1548–51. 10.1111/jgs.15895. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Kinley J, Hockley J, Stone L et al. The provision of care for residents dying in U.K. nursing care homes. Age Ageing 2014;43:375–9. 10.1093/ageing/aft158. [DOI] [PubMed] [Google Scholar]
- 17. Sampson EL, Candy B, Davis S et al. Living and dying with advanced dementia: a prospective cohort study of symptoms, service use and care at the end of life. Palliat Med 2017;32:668–81. 10.1177/0269216317726443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Crowther J, Horton S, Wilson K et al. A UK qualitative study of living and dying with dementia in the last year of life. Palliat Care Soc Pract 2022;16:26323524221096691. 10.1177/26323524221096691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Cardona-Morrell M, Kim J, Turner RM et al. Non-beneficial treatments in hospital at the end of life: a systematic review on extent of the problem. International J Qual Health Care 2016;28:456–69. 10.1093/intqhc/mzw060. [DOI] [PubMed] [Google Scholar]
- 20. Robinson J, Boyd M, O’Callaghan A et al. The extent and cost of potentially avoidable admissions in hospital inpatients with palliative care needs: a cross-sectional study. BMJ Support Palliat Care 2015;5:266–72. 10.1136/bmjspcare-2014-000724. [DOI] [PubMed] [Google Scholar]
- 21. Leniz J, Higginson IJ, Yi D et al. Identification of palliative care needs among people with dementia and its association with acute hospital care and community service use at the end-of-life: a retrospective cohort study using linked primary, community and secondary care data. Palliat Med 2021;35:1691–700. 10.1177/02692163211019897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Gebresillassie BM, Attia JR, Mersha AG et al. Prognostic models and factors identifying end-of-life in non-cancer chronic diseases: a systematic review. BMJ Support Palliat Care 2024;14:e2316–29. 10.1136/spcare-2023-004656. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Ho L, Pugh C, Seth S et al. Performance of models for predicting 1-year to 3-year mortality in older adults: a systematic review of externally validated models. Lancet Healthy Longev 2024;5:e227–35. 10.1016/S2666-7568(23)00264-7. [DOI] [PubMed] [Google Scholar]
- 24. Brown M, Sampson E, Jones L et al. Prognostic indicators or 6-month mortality in elderly people with advanced dementia: a systematic review. Palliat Med 2013;27:389–400. 10.1177/0269216312465649. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Kaur P, Tan WS, Gunapal PPG et al. Deaths in dementia: a scoping review of prognostic variables. BMJ Support Palliat Care 2021;11:242–52. 10.1136/bmjspcare-2020-002217. [DOI] [PubMed] [Google Scholar]
- 26. Liberati A, Altman D, Tetzlaff J et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med 2009;6:e10000100. 10.2427/5768. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Jayes RL, Arnold RM, Fromme EK. Does this dementia patient meet the prognosis eligibility requirements for hospice enrollment? J Pain Symptom Manage 2012;44:750–6. 10.1016/j.jpainsymman.2012.08.004. [DOI] [PubMed] [Google Scholar]
- 28. Quinn KL, Grossman DL. At the crossroads of religion and palliative care in patients with dementia. Isr J Health Pol Res 2020;9:43. 10.1186/s13584-020-00401-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Burton JK, Quinn TJ, Gordon AL et al. Identifying published studies of care home research: an international survey of researchers. J Nurs Home Res 2017;3:99–102. 10.14283/jnhrs.2017.15. [DOI] [Google Scholar]
- 30. Achterberg WP, Everink IH, van der Steen JT et al. We’re all different and we’re the same: the story of the European nursing home resident. Age Ageing 2019;49:3–4. 10.1093/ageing/afz145. [DOI] [PubMed] [Google Scholar]
- 31. Harrison JK, Fearon P, Noel-Storr AH et al. Informant questionnaire on cognitive decline in the elderly (IQCODE) for the diagnosis of dementia within a secondary care setting. Cochrane Database Syst Rev 2015;3. Art. No.: CD010772. 10.1002/14651858.CD010772.pub2. [DOI] [PubMed] [Google Scholar]
- 32. O’Brien JT, Thomas A. Vascular dementia. Lancet 2015;386:1698–706. 10.1016/S0140-6736(15)00463-8. [DOI] [PubMed] [Google Scholar]
- 33. National Institute for Health and Care Excellence . Dementia: assessment, management and support for people living with dementia and their carers NICE guideline [NG97] 2018. [cited 2023 14th December]: Available from: https://www.nice.org.uk/guidance/ng97. [PubMed]
- 34. Ho L, Pugh C, Seth S et al. Predicting short- to medium-term care home admission risk in older adults: a systematic review of externally validated models. Age Ageing 2024;53:afae088. 10.1093/ageing/afae088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Veritas Health Innovation . Covidence Systematic Review Software. Melbourne, Australia, 2023. [Google Scholar]
- 36. Wolff RF, Moons KGM, Riley RD et al. PROBAST: a tool to assess the risk of bias and applicability of prediction model studies. Ann Intern Med 2019;170:51–8. 10.7326/M18-1376. [DOI] [PubMed] [Google Scholar]
- 37. Fernandez-Felix BM, López-Alcalde J, Roqué M et al. CHARMS and PROBAST at your fingertips: a template for data extraction and risk of bias assessment in systematic reviews of predictive models. BMC Med Res Methodol 2023;23:44. 10.1186/s12874-023-01849-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Rauh SP, Heymans MW, van der Maaden T et al. Predicting mortality in nursing home residents with dementia and pneumonia treated with antibiotics: validation of a prediction model in a more recent population. J Gerontol A Biol Sci Med Sci 2019;74:1922–8. 10.1093/gerona/gly260. [DOI] [PubMed] [Google Scholar]
- 39. Kayyali A, Petite SE. Identification of deprescribing opportunities for patients with diabetes and dementia discharged to a nursing home using a mortality prognostic index. Drugs Ther Perspect 2020;36:41–6. 10.1007/s40267-019-00683-z. [DOI] [Google Scholar]
- 40. Gambassi G, Landi F, Lapane KL et al. Predictors of mortality in patients with Alzheimer’s disease living in nursing homes. J Neurol Neurosurg Psychiatry 1999;67:59–65. 10.1136/jnnp.67.1.59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Suh GH, Kil Yeon B, Shah A et al. Mortality in Alzheimer’s disease: a comparative prospective Korean study in the community and nursing homes. Int J Geriatr Psychiatry 2005;20:26–34. 10.1002/gps.1256. [DOI] [PubMed] [Google Scholar]
- 42. Navarro-Gil P, González-Vélez AE, Ayala A et al. Which factors are associated with mortality in institutionalized older adults with dementia? Arch Gerontol Geriatr 2014;59:522–7. 10.1016/j.archger.2014.07.007. [DOI] [PubMed] [Google Scholar]
- 43. Liu J, Li X, Yu W et al. Prediction of survival of persons with advanced dementia using the advanced dementia prognostic tool: a 2-year prospective study. Geriatr Nurs 2024;55:64–70. 10.1016/j.gerinurse.2023.11.005. [DOI] [PubMed] [Google Scholar]
- 44. van Dijk PT, Dippel DW, Habbema JD. A behavioral rating scale as a predictor for survival of demented nursing home patients. Arch Gerontol Geriatr 1994;18:101–13. 10.1016/0167-4943(94)00537-0. [DOI] [PubMed] [Google Scholar]
- 45. Carlson MC, Brandt J, Steele C et al. Predictor index of mortality in dementia patients upon entry into long-term care. J Gerontol A Biol Sci Med Sci 2001;56:M567–70. 10.1093/gerona/56.9.M567. [DOI] [PubMed] [Google Scholar]
- 46. Page MJ, McKenzie JE, Bossuyt PM et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. 10.1136/bmj.n71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Esteban-Burgos AA, Hueso-Montoro C, Mota-Romero E et al. The prognostic value of the NECPAL instrument, palliative prognostic index, and PROFUND index in elderly residents of nursing homes with advanced chronic condition. BMC Geriatr 2023;23:715. 10.1186/s12877-023-04409-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Hicks KL, Rabins PV, Black BS. Predictors of mortality in nursing home residents with advanced dementia. Am J Alzheimers Dis Other Demen 2010;25:439–45. 10.1177/1533317510370955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Hirdes JP, Poss JW, Mitchell L et al. Use of the interRAI CHESS scale to predict mortality among persons with neurological conditions in three care settings. PloS One 2014;9:e99066. 10.1371/journal.pone.0099066. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Marsh GW, Prochoda KP, Pritchett E et al. Predicting hospice appropriateness for patients with dementia of the Alzheimer’s type. Appl Nurs Res 2000;13:187–96. 10.1053/apnr.2000.7654. [DOI] [PubMed] [Google Scholar]
- 51. McCusker J, Cole MG, Voyer P et al. Six-month outcomes of co-occurring delirium, depression, and dementia in long-term care. J Am Geriatr Soc 2014;62:2296–302. 10.1111/jgs.13159. [DOI] [PubMed] [Google Scholar]
- 52. Mitchell S, Kiely D, Hamel M et al. Estimating prognosis for nursing home residents with advanced dementia. JAMA 2004;291:2734–40. 10.1001/jama.291.22.2734. [DOI] [PubMed] [Google Scholar]
- 53. Mitchell SL, Miller SC, Teno JM et al. The advanced dementia prognostic tool: a risk score to estimate survival in nursing home residents with advanced dementia. J Pain Symptom Manage 2010;40:639–51. 10.1016/j.jpainsymman.2010.02.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Mitchell S, Miller S, Teno J et al. Prediction of 6-month survival of nursing home residents with advanced dementia using ADEPT vs hospice eligibility guidelines. JAMA 2010;304:1929–35. 10.1001/jama.2010.1572. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. van der Steen JT, Mitchell SL, Frijters DH et al. Prediction of 6-month mortality in nursing home residents with advanced dementia: validity of a risk score. J Am Med Dir Assoc 2007;8:464–8. 10.1016/j.jamda.2007.05.004. [DOI] [PubMed] [Google Scholar]
- 56. Volicer BJ, Hurley A, Fabiszewski KJ et al. Predicting short-term survival for patients with advanced Alzheimer’s disease. J Am Geriatr Soc 1993;41:535–40. 10.1111/j.1532-5415.1993.tb01891.x. [DOI] [PubMed] [Google Scholar]
- 57. Almeida J, Galicia-Castillo M. Predicting mortality in end stage dementia: a retrospective comparison of medicare guidelines and Mitchell criteria for hospice referral. J Am Geriatr Soc 2012;60:S124. [Google Scholar]
- 58. Volberding K, Patki D, Watson N et al. Does the mortality risk index predict 6 month mortality in residents living in a dementia special care center? J Am Geriatr Soc 2013;61:S222. [Google Scholar]
- 59. Wong WM. P2-320: how to do better prognostication in advanced dementia. Alzheimers Dement 2018;14:P805. 10.1016/j.jalz.2018.06.1010. [DOI] [Google Scholar]
- 60. Browne B, Kupeli N, Moore KJ et al. Defining end of life in dementia: a systematic review. Palliat Med 2021;35:1733–46. 10.1177/02692163211025457. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61. Murray SA, Boyd K, Moine S et al. Using illness trajectories to inform person centred, advance care planning. BMJ 2024;384:e067896. 10.1136/bmj-2021-067896. [DOI] [PubMed] [Google Scholar]
- 62. Hunt LJ, Morrison RS, Gan S et al. Mortality and function after hip fracture or pneumonia in people with and without dementia. J Am Geriatr Soc 2025;73:1179–88. 10.1111/jgs.19354. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63. Hsieh PC, Wu SC, Fuh JL et al. The prognostic predictors of six-month mortality for residents with advanced dementia in long-term care facilities in Taiwan: a prospective cohort study. Int J Nurs Stud 2019;96:9–17. 10.1016/j.ijnurstu.2018.12.013. [DOI] [PubMed] [Google Scholar]
- 64. Mitchell S, Teno J, Kiely D et al. The clinical course of advanced dementia. N Engl J Med 2009;361:1529–38. 10.1056/NEJMoa0902234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65. Lee KC, Hsu WH, Chou PH et al. Estimating the survival of elderly patients diagnosed with dementia in Taiwan: a longitudinal study. PloS One 2018;13:e0178997. 10.1371/journal.pone.0178997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 66. Kaur P, Kannapiran P, Ng SHX et al. Predicting mortality in patients diagnosed with advanced dementia presenting at an acute care hospital: the PROgnostic model for advanced DEmentia (PRO-MADE). BMC Geriatr 2023;23:255. 10.1186/s12877-023-03945-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 67. Wang L, Sha L, Lakin JR et al. Development and validation of a deep learning algorithm for mortality prediction in selecting patients with dementia for earlier palliative care interventions. JAMA Netw Open 2019;2:e196972. 10.1001/jamanetworkopen.2019.6972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 68. Kaehr EW, Pape LC, Malmstrom TK et al. FRAIL-NH predicts outcomes in long term care. J Nutr Health Aging 2016;20:192–8. 10.1007/s12603-016-0682-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 69. Porock D, Parker-Oliver D, Petroski GF et al. The MDS mortality risk index: the evolution of a method for predicting 6-month mortality in nursing home residents. BMC Res Notes 2010;3:200. 10.1186/1756-0500-3-200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 70. Balasubramanian I, Andres EB, Poco LC et al. Prognostic understanding among caregivers of persons with dementia: a scoping review. J Am Geriatr Soc 2025;73:947–55. 10.1111/jgs.19245. [DOI] [PubMed] [Google Scholar]
- 71. Chu CP, Huang CY, Kuo CJ et al. Palliative care for nursing home patients with dementia: service evaluation and risk factors of mortality. BMC Palliat Care 2020;19:122. 10.1186/s12904-020-00627-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 72. Kochovska S, Garcia MV, Bunn F et al. Components of palliative care interventions addressing the needs of people with dementia living in long-term care: a systematic review. Palliat Med 2020;34:454–92. 10.1177/0269216319900141. [DOI] [PubMed] [Google Scholar]
- 73. Tropea J, Nestel D, Johnson C et al. Evaluation of IMproving palliative care education and training using simulation in dementia (IMPETUS-D) a staff simulation training intervention to improve palliative care of people with advanced dementia living in nursing homes: a cluster randomised controlled trial. BMC Geriatr 2022;22:127. 10.1186/s12877-022-02809-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 74. Bicknell R, Lim WK, Maier AB et al. A study protocol for the development of a multivariable model predicting 6- and 12-month mortality for people with dementia living in residential aged care facilities (RACFs) in Australia. Diagn Progn Res 2020;4:17. 10.1186/s41512-020-00085-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 75. Gordon AL, Rand S, Crellin E et al. Piloting a minimum data set for older people living in care homes in England: a developmental study. Age Ageing 2025;54:afaf001. 10.1093/ageing/afaf001. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
In addition to the data available in the supplementary materials, the underlying data supporting this review is available on reasonable request through the corresponding author.