Aims: Delirium may hasten nursing home admission but inconsistent evidence has left its real-world impact unclear. We evaluated the independent associations of hospital-acquired delirium with the risk and timing of nursing home admission among older adults with and without dementia.
Methods: Matched cohort design in UK Biobank participants with linked hospital records. Participants with hospital-acquired delirium, with or without dementia, were matched 1:1 to controls without delirium on age, sex, Hospital Frailty Risk Score, admission diagnosis, length of stay and intensive care. Delirium from ICD-10 codes and incident nursing home admission from hospital discharge destination were analysed using competing-risk and accelerated failure time models.
Results: In participants without dementia (N = 26,008), delirium was associated with 23% higher risk of nursing home admission (sHR 1.23, 95% CI 1.14, 1.34) and hastened admission by 1.6 years (95% CI 1.1, 1.9) at a survival probability of 0.90; each additional delirium episode was associated with 20% higher risk (sHR 1.20, 95% CI 1.13, 1.27) and hastened admission 20% (95% CI 14, 26). In participants with dementia (N = 3,580), delirium was associated with 22% higher risk of nursing home admission (sHR 1.22, 95% CI 1.07, 1.38) and hastened admission by 1.3 years (95% CI 0.6, 1.8) at a survival probability of 0.70; additional episodes were not significantly associated with risk.
Conclusions: Hospital-acquired delirium is a consistent, independent predictor of nursing home admission, both for people with and without dementia. Strategies aimed at mitigating hospital-acquired delirium might delay the need for nursing home care.