Long-term clinical outcomes of delirium after hospital discharge: a systematic review and meta-analysis
Yonas Tesfaye, Courtney R. Davis, Melissa J. Hull, Danielle Greaves, James du Preez, Sally Johns, Alice Bourke, Hannah A. D. Keage
Abstract
BACKGROUND: Delirium has been linked to adverse health outcomes. There has not been a comprehensive attempt to synthesise these outcomes. OBJECTIVE: To synthesise evidence comparing post-discharge clinical outcomes in individuals who experienced delirium in hospital compared to those who did not. METHODS: A systematic electronic search was conducted in Medline, Embase, CINAHL, PsycINFO and Cochrane databases. Random-effects models were used to assess effect size differences between those who experienced delirium and those who did not: odds ratios (OR) for categorical outcomes and Hedges' g for continuous outcomes. Analyses were conducted for each outcome relative to ≤6 months, >6-12 months, 12+ months and collapsed across time post-discharge. RESULTS: Data were synthesised from 253 studies representing 29 814 participants who experienced delirium and 107 583 participants who did not experience delirium. The mean (SD) age of participants was 76.0 (9.3) years. Collapsed over follow-up period, results included, those who experienced delirium in hospital showed higher objective cognitive decline (OR = 1.58, P < .001), greater subjective cognitive impairment (OR = 2.11, P = .041), greater functional decline (g = -0.43, P = .001), lower quality of life (g = -0.44, P < .001), higher burden of poor mental health (OR = 1.69, P < .001), increased risk of dementia (OR = 5.37, P < .001), higher likelihood of institutionalisation (OR = 2.80, P < .001), greater rates of hospital readmission (OR = 1.70, P < .001) and increased mortality (OR = 2.55, P < .001) post-hospital discharge compared to those who did not experience delirium in hospital. Time-specific analyses did not reveal any consistent patterns of effects. CONCLUSIONS: Older adults who experience delirium in hospital demonstrate significantly worse long-term clinical outcomes compared to those who do not.