Early Identification of Delirium in Intensive Care Unit Patients: Improving the Quality of Care.
Publication Title
Critical care nurse
Document Type
Article
Publication Date
4-1-2020
Abstract
BACKGROUND: Delirium has long-term consequences for intensive care unit patients. The project site, an urban academic hospital, did not previously use a validated delirium screening tool, and patients commonly received sedative medications to treat agitation.
OBJECTIVE: To minimize the risk of delirium by implementing the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) as the standard-of-care delirium assessment tool in the intensive care unit and by decreasing use of high-risk medications (ie, opioids and benzodiazepines).
METHODS: An observational pretest-posttest design was used to analyze deidentified patient data from electronic health records. The evidence-based practice intervention focused on educating nurses on high-risk medications and CAM-ICU implementation. Control charts, χ2 tests, and mixed regression models were used to evaluate the effectiveness of the intervention in reducing delirium risk by decreasing use of high-risk medications.
RESULTS: High-risk medication use significantly decreased after intervention among patients at low risk for delirium (before intervention, 7.37%; after intervention, 3.92%; P < .001) and at high risk for delirium (before intervention, 4.73%; after intervention, 2.99%; P < .001). Hospital stays were significantly shorter in patients at low risk than at high risk for delirium (P < .001) but increased by a mean of 0.13 days with each additional high-risk medication used (P < .001).
CONCLUSIONS: The variation of high-risk medication use was significantly controlled with the implementation of CAM-ICU and education. Nurses felt that hands-on training with the CAM-ICU increased their comfort in identifying patients at risk for delirium. Future work will focus on assessment accuracy.
Specialty/Research Institute
Nursing
Specialty/Research Institute
Critical Care Medicine
