Real-World Adherence to a Delirium Screening Test Administered by Nurses and Medical Staff during Routine Patient Care

Abstract

Delirium is often the first symptom of incipient acute illness or complications and must therefore be detected promptly. Nevertheless, routine screening for delirium in acute care hospital wards is often inadequate. We recently implemented a simple, user-friendly delirium screening test (RMA) that can be administered during ward rounds and routine nursing care. The test was found to be non-inferior to 4AT in terms of sensitivity and specificity. However, the dominant factors to take into account when assessing the performance of a test added to the routine work of busy acute care hospital wards are ease of administration, real-life amenability and the ability of the staff to adhere to testing requirements. In this study, we evaluated the prevalence of daily RMA tests that were not administered as scheduled and the impact of these omissions on the overall real-world performance of RMA. Using point-in-time assessments of 4AT by an external rater, we found that complete RMA was administered in 88.8% of the days. Physicians omitted significantly more tests than nurses, but their results were more specific for delirium. Omissions reduced the sensitivity and specificity of RMA for delirium (compared to 4AT) from 90.7% to 81.7%, and from 99.2% to 87.8%, respectively. Ideally, the number of omitted RMA tests should be minimized. However, if over 85% of the daily quota of complete tests are administered, the sensitivity and specificity of RMA for diagnosing delirium as soon as it appears remain at acceptable levels.

Keywords

• delirium
• screening test
• adherence
• cognition
• RADAR