Systematic review of prognostic factors for poor outcome in people living with dementia that can be determined from primary care medical records

Background Dementia has a major impact on individuals, their families and caregivers, and wider society. Some individuals experience a faster decline of their function and health compared to others. The objective of this systematic review was to determine prognostic factors, measurable in primary care, for poor outcome in people living with dementia. Methods Cohort studies set in the community or primary care, and examining prognostic factors for care home admission, cognitive decline, or palliative care were included. Databases were searched from inception to 17th June 2022. Identified papers were screened, the risk of bias assessed using Quality in Prognostic Studies (QUIPS) tool, and data extracted by 2 reviewers, with disagreements resolved by consensus or a 3rd reviewer. A narrative synthesis was undertaken, informed by GRADE, taking into consideration strength of association, risk of bias and precision of evidence. Patient and Public Involvement and Engagement (PPIE) and stakeholder input was obtained to prioritise factors for further investigation. Results Searches identified 24,283 potentially relevant titles. After screening, 46 papers were included, 21 examined care home admission investigating 94 factors, 26 investigated cognitive decline as an outcome examining 60 factors, and 1 researched palliative care assessing 13 factors. 11 prognostic factors (older age, less deprived, living alone, white race, urban residence, worse baseline cognition, taking dementia medication, depression, psychosis, wandering, and caregiver’s desire for admission) were associated with an increased risk of care home admission and 4 prognostic factors (longer duration of dementia, agitation/aggression, psychosis, and hypercholesterolaemia) were associated with an increased risk of cognitive decline. PPIE and other stakeholders recommended further investigation of 22 additional potential prognostic factors. Conclusions Identifying evidence for prognostic factors in dementia is challenging. Whilst several factors highlighted as of relevance by our stakeholder groups need further investigation, inequalities may exist in care home admission and there is evidence that several prognostic factors measurable in primary care could alert clinicians to risk of a faster progression. Registration PROSPERO CRD42019111775.