Non-pharmacological interventions to reduce neuropsychiatric symptoms in hospitalised patients for behavioural crises with cognitive impairment: A systematic review

• This systematic review summarises the current effectiveness of non-pharmacological interventions to reduce NPS in older patients with cognitive impairment hospitalised for behavioural crises. • Five studies testing physical exercise (n=1), listening to individualised music (n=1), implementation of physical exercise and music (n=1) and meaningful activities (n=2) in geriatric psychiatry were included. • Three of the five studies showed significant effects on NPS. These studies tested physical exercise, (behavioural activation for) meaningful activity and listening to individualised music. • More research is needed and may include interventions that allow for flexible adjustments to patient characteristics, such as their needs, contextual factors and existing care practices. Non-pharmacological interventions are pivotal to reducing neuropsychiatric symptoms (NPS) in patients with cognitive impairment. This systematic review assessed the effectiveness of non-pharmacological interventions applicable by nursing staff to reduce NPS in older patients with cognitive impairment hospitalised for behavioural crises. Six databases were searched for randomised or non-randomised controlled trials. Two authors screened full-texts and assessed the quality of the studies using the Modified Downs and Black Checklist. Results were presented narratively using the PRISMA guideline. Only five studies could be included, all were conducted in geriatric psychiatry. Three studies showed significant effects on NPS. They tested physical exercise, behavioural activation for meaningful activity and listening to individualised music and were of good, moderate and low quality, respectively. The best evidence was found for physical exercise. Due to the small number of studies with heterogeneous quality, the results must be interpreted with caution, limiting the scope of conclusions. While more research is needed, we recommend that non-pharmacological interventions in clinical practice be adjusted to patient characteristics, contextual factors and existing care practices.