Admission Delirium and In-Hospital Mortality in Older Adults: A Systematic Review and Meta-analysis

5353

K H Kyaw1; J Soldera1

University of South Wales

Abstract

Introduction

Delirium is common among older adults admitted to hospital and is associated with adverse outcomes. However, the magnitude of its association with in-hospital mortality and whether this differs between ward and intensive care unit (ICU) settings remain uncertain. This systematic review and meta-analysis evaluated the association between delirium present on admission and mortality.

Methods

Observational studies comparing outcomes in hospitalised older adults with and without delirium on admission were included. Meta-analysis was performed in R using the meta and metafor packages. Odds ratios (ORs) with 95% confidence intervals (CIs) were pooled using random-effects models with restricted maximum likelihood estimation. Heterogeneity was quantified using I². Prespecified subgroup analysis and meta-regression compared ward and ICU settings. Longer-term mortality, ICU admission and hospital readmission were analysed when reported by at least two studies.

Results

Fifteen studies involving 33,884 patients were included in the primary analysis; 6,708 had delirium and 27,176 did not. Delirium on admission was associated with substantially higher in-hospital mortality (OR 4.49, 95% CI 3.52–5.74; I²=67.3%). The association remained significant in ward studies (OR 4.88, 95% CI 3.70–6.45; I²=52.7%) and ICU studies (OR 3.78, 95% CI 2.35–6.07; I²=84.7%), with no significant difference between settings. Delirium was also associated with longer-term mortality (OR 3.73, 95% CI 3.08–4.52) and ICU admission (OR 5.05, 95% CI 1.37–18.66), but not hospital readmission (OR 0.98, 95% CI 0.69–1.40). Egger’s test showed no evidence of small-study effects (p=0.589).

Conclusions

Delirium present on hospital admission is a strong prognostic marker for in-hospital and longer-term mortality. Its adverse association was consistent across ward and ICU settings, supporting early recognition and risk stratification in older hospitalised adults.