876

Abstract

Introduction: Timely recognition of delirium in the intensive care unit (ICU) using validated tools such as The Confusion Assessment Method for the ICU (CAM-ICU) is essential to appropriate management. The purpose of this study was to describe the state of delirium assessment prior to the implementation of an ICU delirium algorithm. Methods: Patients 18-89 years old newly admitted to the surgical ICU with an expected ICU length of stay > 48 hours were eligible for evaluation. A pharmacist assessed patients daily using the CAM-ICU over a 2 week period. The pharmacist evaluation was compared to the nursing assessment for correlation. Nursing and physician assessments were also compared. The nursing CAM-ICU was considered inappropriate if: 1) CAM-ICU was documented while the patient had a RASS score of -4 or -5 or 2) CAM-ICU was charted as 'unable to assess' with a RASS score of -3 or greater. Additional risk factors for delirium were also collected. Descriptive statistics were used to summarize demographic variables and clinical outcomes. Results: A total 89 patient-days for the 20 patients were included. The median patient age was 50 years (range 23-89) and 60% of were male. No patients had hepatic encephalopathy, 3 were uremic, 1 experienced illicit drug withdrawal, and 3 experienced alcohol withdrawal. Only one patient had documented dementia prior to ICU admission. Pharmacy and nursing CAM-ICU assessments correlated in 59.6% of cases and physician and nursing correlation occurred in 60.6% of cases. The CAM-ICU was characterized as inappropriate for 46 of 178 nursing documented evaluations (25.9%). The most common error involved documentation as 'unable to assess' when the assessment should have occurred based on the documented level of sedation. There was only one occurrence where a CAM-ICU was inappropriately performed when the RASS score was -4 or -5. Conclusions: Assessment of ICU delirium is inconsistent in our surgical ICU. A low degree of correlation was seen between practitioners from different professions. Additionally, the CAM-ICU assessment was frequently documented as 'unable to assess' based on level of sedation. Future directions to improve assessment of ICU delirium within our surgical ICU include practitioner education and algorithm implementation.