Clinical associations of delirium in hospitalized adult patients and the role of on admission presentation.
Academic Article
Overview
abstract
OBJECTIVE: To describe clinical associations of delirium in hospitalized patients and relationships to on admission presentation. DESIGN: Retrospective analysis of an administrative hospitalization database 1998-2007. SETTING: Acute care hospitalizations in the New York State (NYS). MEASUREMENTS: Four categories of diagnosis related group (DRG) hospitalizations were extracted from a NYS administrative database: pneumonia, congestive heart failure, urinary tract/kidney infection (UTI), and lower extremity orthopedic surgery (LEOS) DRGs. These hospitalizations were examined for clinical associations with delirium coding both on and after admission. RESULTS: Delirium was coded in 0.8% of the cohort, of which an on admission diagnosis was present in 59%. On admission delirium was strongly associated with dementia (adjusted odds ratio 0, 95%CI 5.8-6.3) and with adverse drug effects (ADEs) (adjusted odds ratio 4.6, 95%CI 4.3, 5.0). After admission delirium was even more highly associated with ADEs (adjusted odds ratio 22.2, 95%CI 20.7-23.7). The UTI DRG category had the greatest proportion of on admission delirium. However after admission delirium was more common in the LEOS DRG category. Over time, there was a greater increase in delirium proportions in the UTI DRG category, and an overall increase in coding for encephalopathy states (potential alternative delirium descriptors). CONCLUSION: ADEs play an important role in delirium regardless of whether or not it is present on admission. While the finding that most delirium hospitalizations presented on admission suggests that delirium impacts more as a clinical admitting determinant, in-hospital prevention strategies may still have benefit in targeted settings where after admission delirium is more frequent, such as patients with LEOS.
