Incidence, characteristics and risk factors of delirium in the intensive care unit: An observational study.

AIMS AND OBJECTIVE

To investigate the incidence, characteristics and risk factors of delirium in the ICU.

BACKGROUND

Identifying the risk factors of delirium is important for early detection and to prevent adverse consequences.

DESIGN

An observational cohort study conducted according to STROBE Guidelines.

METHOD

The study was conducted with patients who stayed in ICU ≥24 h and were older than 18 years. Patients were assessed twice daily using the RASS and CAM-ICU until either discharge or death. Cumulative incidence was calculated. Demographic/clinical characteristics, length of stay and mortality were compared between patients with and without delirium. A logistic regression model was used to investigate risk factors.

RESULTS

The incidence of delirium was 31.8% and hypoactive type was the most frequent (41.5%). The median onset of delirium was 3 days (IQR = 2) with a mean duration of 5.27 ± 2.32 days. Patients with delirium were significantly older, had higher APACHE-II, SOFA and CPOT scores, higher blood urea levels, higher requirements for mechanical ventilation, sedation and physical restraints, longer stays in the ICU and higher mortality than those without delirium. The logistic regression analysis results revealed that a CPOT score ≥3 points (OR = 4.70, 95% CI: 1.05-20.93; p = .042), physical restraint (OR = 10.40, 95% CI: 2.75-39.27; p = .001) and ICU stay ≥7 days (OR = 7.26, 95% CI: 1.60-32.84; p = .010) were independent risk factors of delirium.

CONCLUSIONS

In this study, the incidence of delirium was high and associated with several factors. It is critical that delirium is considered by all members of the healthcare team, especially nurses, and that protocols are established for improvements.

RELEVANCE TO THE CLINICAL PRACTICE

Based on the results of this study, delirium could be decreased by preventing the presence of pain, prudent use of physical restraints and shortening the ICU stay.