A systematic review of 13 studies involving 503,501 participants identified 11 prognostic tools designed to predict 6-month mortality in people with dementia living in care homes, but only two tools showed acceptable discrimination and this was not replicated in validation studies.
- Researchers reviewed 13 studies covering over 500,000 care home residents with dementia to assess tools predicting 6-month mortality
- Only 2 of 11 identified tools showed acceptable accuracy, and validation studies could not replicate these results
- Common predictors included age, cognitive decline, functional dependency, and nutritional concerns
How this compares to prior research
Dementia is a leading cause of death among care home residents internationally. While clinicians have long sought reliable methods to identify residents approaching end-of-life to support care planning and respect patient preferences, the accuracy and consistency of existing prognostic approaches remained unclear. This review synthesized evidence on available tools to determine which predictive factors and instruments show the most promise for clinical use.
Dementia prevalence trends, global population, 1990–1999
Key findings
- Eleven prognostic tools were identified across 13 studies involving 503,501 participants with dementia in care homes
- Only the Mortality Risk Index Score and Advanced Dementia Prognostic Tool demonstrated acceptable discrimination with area under receiver operating curve greater than 0.70
- The most common predictor domains were age, change in cognitive function, functional decline or dependency, and concerns around nutritional intake or weight loss
What this means in practice
- Ask your care team about advance care planning if a loved one with dementia shows declining function or nutritional concerns
- Recognize that predicting exact timelines remains difficult even with assessment tools
- Consider discussing end-of-life preferences early rather than relying solely on prognostic scores
Frequently asked questions
Can doctors accurately predict 6-month survival for dementia patients in care homes?
Current prognostic tools show limited accuracy. Only two tools demonstrated acceptable discrimination, and validation studies could not replicate these results, indicating prediction remains challenging.
What factors help predict mortality in care home residents with dementia?
The most common predictors across tools include age, declining cognitive function, loss of functional independence, and nutritional concerns such as weight loss or reduced food intake.
How many prediction tools exist for dementia patients in care homes?
Researchers identified 11 different prognostic tools designed to predict 6-month mortality, though these tools varied significantly in how they assessed and applied predictive factors.
Key terms explained
Prognostic tool
A clinical instrument that uses specific factors to predict future health outcomes, such as likelihood of death within a timeframe.
Area under receiver operating curve
A statistical measure of how well a test distinguishes between outcomes; values above 0.70 indicate acceptable accuracy.
PROBAST
Prediction model Risk Of Bias ASsessment Tool, a standardized method for evaluating quality of prognostic studies.
Source: Determining six-month prognosis among people with dementia living in care homes: a systematic review of prognostic tools. · DOI: doi: 10.1093/ageing/afag087