Delirium prevalence and incidence in acutely admitted older patients: an observational cohort study

Delirium is common in acutely ill older adults and is associated with multiple unfavourable outcomes, including an increased risk of dementia and death. As the population ages and more people live with dementia, updated and accurate estimates of delirium prevalence are important. The primary aim of this study was to identify delirium point prevalence in the emergency department and incidence during hospitalisation. Our secondary aim was to compare outcomes: length of stay, need for a higher level of care, and mortality after discharge in patients with and without delirium. In this unselected observational cohort study, all older adults aged 65 years or above acutely admitted to the emergency department of a large Norwegian hospital during a 5-day and 4-night midweek period, were screened for delirium by the 4 “A”s test. A final consensus delirium diagnosis was made based on review of all available information in the patients’ electronic health record to consider if The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition’s criteria (DSM-V) for delirium were fulfilled. Of 240 patients assessed, 14% (n = 33) fulfilled DSM-V criteria for delirium in the emergency department, and 8% (n = 17) of the remaining 207 patients developed delirium later during the hospitalisation. Only 4 of the 50 patients with delirium (8%) had a documented diagnosis of delirium in their discharge summary. For patients with delirium, the current hospital admission was more often a readmission (42% vs. 18%, p < 0.001). Delirium patients also had longer hospitalisations (4 vs. 2 days, p < 0.001), and higher 9-month mortality (52% vs. 14%, p = 0.002), corrected for age, sex and severity of acute illness by The National Early Warning Score 2. Delirium was common and underdiagnosed in our study and associated with unfavourable outcomes for the patients. Delirium prevention, identification and management should be key priorities for the health care system. Our article does not report results of a health care intervention, and the protocol is not registered in a trial registry.