Occurrence, associated factors, and outcomes of delirium in patients in an adult acute general medicine service in England: a 10-year longitudinal, observational study

Background

Reliable estimates of delirium occurrence and outcomes are necessary to inform hospital services, research, and policy, but inclusive cohorts with long-term follow-up are scarce. We aimed to assess the age-specific occurrence of delirium in acute general (internal) medicine, associated factors, and 10-year outcomes stratified by age and comorbid dementia status.

Methods

This longitudinal, observational study was done at the John Radcliffe Hospital (Oxford, UK). We included consecutive adult patients aged 16 years and older in an acute general (internal) medicine service over six 8-week periods (between Sept 4, 2010, and Nov 15, 2018). Delirium was diagnosed prospectively using the Confusion Assessment Method and Diagnostic and Statistical Manual of Mental Disorders, fourth edition, criteria and subcategorised as prevalent (≤48 h of admission) or incident (>48 h postadmission). Odds ratios adjusted (_adjORs) for demographics, comorbidity, frailty, and illness severity were calculated for binarised outcomes and adjusted hazard ratios (_adjHRs) were calculated for time to death.

Findings

1846 patients were admitted to acute general (internal) medicine (mean age 68·2 years [SD 20·0], age range 16–102 years), 426 (23% [95% CI 21–25]) of whom had delirium (prevalent n=290 [68%], incident n=73 [17%], both prevalent and incident n=63 [15%]), of whom 134 (31·5%) had dementia. 950 (51·5%) patients were female, 895 (48·5%) were male, and sex data were missing for one patient. Delirium increased with age, from six (2% [95% CI 1–4]) of 340 patients younger than 50 years and 31 (9% [6–13]) of 333 patients at age 50–64 years to 57 (20% [16–25]) of 281 at age 65–74 years, 245 (35% [31–38]) of 704 at age 75–89 years, and 87 (46% [39–54]) of 188 at age 90 years and older. Of the 37 patients younger than 65 years who had delirium, 28 (76%) had an underlying neurological or neuropsychiatric disorder. In those aged 65 years or older, delirium was overall associated (all p<0·001, age and sex adjusted) with dementia (_adjOR 3·63 [95% CI 2·65–4·98]), pre-admission dependency (2·63 [2·02–3·43]), comorbidity burden (1·04 [1·02–1·05]), and frailty (moderate vs low risk 3·62 [2·70–4·85] and high vs low risk 11·85 [7·24–19·42]), with stronger associations in patients without comorbid dementia than in those with comorbid dementia. Delirium predicted inpatient stay longer than 7 days (_adjOR 2·48 [1·84–3·35]), discharge care needs (2·41 [1·70–3·40]), and mortality during admission (2·45 [1·52–3·94]). The increased risk of death in the delirium group was highest in the immediate postadmission period and attenuated thereafter, but was maintained for up to 10 years of follow-up (_adjHR 2·03 [95% CI 1·40–2·97] for 30-day mortality vs 1·52 [1·30–1·77] for 10-year mortality). Excess inpatient mortality was highest in younger age groups versus older age groups (_adjOR 4·38 [95% CI 1·18–16·31]; p=0·028 at age 65–74 years vs 1·96 [1·02–3·75]; p=0·043 at age 75–89 years and 2·86 [1·14–7·16]; p=0·025 at age 90 years or older) and in those without versus with comorbid dementia (_adjOR 3·02 [1·73–5·25]; p<0·001 vs 1·47 [0·58–3·75]; p=0·42).

Interpretation

Our findings support current guidelines for routine on-admission delirium screening from age 65 years. Delirium outcomes are relatively more adverse in those aged 65–74 years without comorbid dementia in whom interventions and clinical trials should be prioritised.

Funding

National Institute for Health and Care Research and the Medical Research Council.