Effectiveness of non-pharmacological and pharmacological interventions in preventing delirium in older adults: a systematic review and meta-analysis of randomised controlled trials

BACKGROUND: Delirium is a serious neuropsychiatric condition common among older adults, associated with prolonged hospital stays, increased morbidity and mortality. Although guidelines emphasise prevention, identifying most effective measures is crucial. OBJECTIVE: To evaluate the efficacy of non-pharmacological and pharmacological interventions in preventing delirium in older adults (≥65 years). METHODS: We conducted a systematic review and meta-analysis of randomised controlled trials across settings, excluding ICU-only studies. MEDLINE, Cochrane, Web of Science and PsycINFO were searched through October 2024. Comparators included standard care, placebo, or other drugs. Random-effects models estimated pooled risk ratios. Risk of bias was assessed using RoB 2.0. Subgroup analyses were performed by setting and drug type. PROSPERO: CRD42024500387. RESULTS: Eighty-seven trials (19,289 randomised patients) were included. Non-pharmacological multicomponent interventions (k = 17) significantly reduced delirium occurrence (risk ratio [RR] = 0.56, 95% CI 0.45-0.72; I2 = 63%) versus usual care, although certainty of evidence was very low. Effects were significant in non-cardiovascular surgical and medical patients. Single-component non-pharmacological interventions (k = 9) showed mixed or inconclusive findings. Among pharmacological trials (k = 53), dexmedetomidine reduced delirium risk (RR = 0.49; 95% CI 0.43-0.57; low-certainty) in non-cardiovascular surgical patients. Corticosteroids and intranasal insulin showed significant effects (high-certainty), based on a small number of trials. Mixed strategies, such as haemodynamic or cerebral oxygen-guided approaches, suggested potential benefit but were heterogeneous. CONCLUSIONS: Non-pharmacological multicomponent interventions reduce delirium occurrence and should be prioritised. Dexmedetomidine showed benefit in non-cardiac surgical populations, while evidence for other drugs remains inconclusive. Overall, certainty was low to very low, highlighting the need for further high-quality randomised trials to ensure generalisability.