Aims: Approximately 1% of people with dementia experience severe behavioural and psychological symptoms of dementia (BPSD) that require management in hospital. This review aimed to evaluate the effectiveness of hospital-based specialised behavioural management units in improving outcomes for older patients with severe BPSD.
Methods: This systematic review was registered with PROSPERO (CRD42024585674) and conducted in accordance with PRISMA guidelines. MEDLINE, EMBASE, PsycINFO, CINAHL, Scopus and Web of Science were searched on 1 September 2024 and updated on 30 January 2026. Eligible studies reported quantifiable outcomes in older patients admitted to specialised hospital units for BPSD. Two reviewers undertook screening, full-text review, critical appraisal using JBI tools, and data extraction. Narrative synthesis was performed, with random-effects meta-analysis conducted where appropriate.
Results: Of 3,774 records identified, 17 studies met inclusion criteria, comprising one randomised controlled trial, 14 quasi-experimental studies, and two cohort studies. Mean participant age ranged from 77.4 to 85.0 years. Considerable heterogeneity existed across specialised units. Three studies compared specialised units with standard care; two reported greater improvement in BPSD symptoms, while one found no significant difference. Twelve studies assessed pre-post outcomes following admission, with 11 reporting improvement in BPSD. Meta-analysis of five studies (582 admissions) demonstrated a mean reduction of 18.15 points on the Neuropsychiatric Inventory (95% CI 7.4-28.9) following admission. Evidence regarding medication use, functional outcomes, and hospital-acquired complications were inconsistent.
Conclusions: Admission to specialised behavioural management units is associated with improvement in severe BPSD. However, evidence comparing these units with standard ward care and evaluating broader clinical outcomes remains limited.