[Epidemiology of delirium in elderly inpatients]

Our objective was to determine delirium incidence and risk factors in a cohort of elderly inpatients. We randomly selected 149 patients, aged 65 years or older, from admission to general wards, without evidence of delirium. They were evaluated daily with the Confusion Assessment Method, an instrument validated for the diagnosis of delirium. We obtained relative risks for delirium and those independently associated were included in a logistic regression model. We used the chi-square test with Yate's corrections for univariate analysis, and t-test for comparisons of means. We observed that 51 patients (20.5%) developed delirium during their hospital stay. Severity of disease (RR 1.28, 1.14-1.43), having chronic diseases (RR 3.45, 2.4-4.96), and having fever at admission (RR 1.84, 1.33-2.56) were found independently associated with delirium. Patients who developed delirium had longer hospital stay (9.87 days +/- 3.48 vs 6.95 days +/- 2.45, p < 0.05) and higher mortality (RR 2.19, CI 1.26-3.79). We conclude that delirium in our setting is very frequent and has negative effects on resource utilization and mortality in elderly inpatients. Its association with the severity of the disease seems interesting. Appropriate prospective identification of patients at risk for delirium may allow the implementation of preventive strategies in order to minimize the impact of this complication.