RESUMO
BACKGROUND: A wide variability in the approach towards delirium prevention and treatment in the critically ill results from the dearth of prospective randomised studies. METHODS: We launched a two-stage prospective observational study to assess delirium epidemiology, risk factors and impact on patient outcome, by enrolling all patients admitted to our Intensive Care Unit (ICU) over a year. The first step - from January to June 2008 was the observational phase, whereas the second one from July to December 2008 was interventional. All the patients admitted to our ICU were recruited but those with pre-existing cognitive disorders, dementia, psychosis and disability after stroke were excluded from the data analysis. Delirium assessment was performed according with Confusion Assessment Method for the ICU twice per day after sedation interruption. During phase 2, patients underwent both a re-orientation strategy and environmental, acoustic and visual stimulation. RESULTS: We admitted a total of respectively 170 (I-ph) and 144 patients (II-ph). The delirium occurrence was significantly lower in (II-ph) 22% vs. 35% in (I-ph) (P=0.020). A Cox's Proportional Hazard model found the applied reorientation strategy as the strongest protective predictors of delirium: (HR 0.504, 95% C.I. 0.313-0.890, P=0.034), whereas age (HR 1.034, 95% CI: 1.013-1.056, P=0.001) and sedation with midazolam plus opiate (HR 2.145, 95% CI: 2.247-4.032, P=0.018) were negative predictors. CONCLUSION: A timely reorientation strategy seems to be correlated with significantly lower occurrence of delirium.