Abstract 13648: Hospital Admission Patterns and Outcomes of Heart Failure (HF) Patients Discharged from the Emergency Department (ED)
BACKGROUND Understanding variations in HF hospitalizations requires knowledge of admission-discharge decisions in the ED. Hospital-level variation in admission rates and outcomes after discharge are unknown. OBJECTIVES To examine the association of hospital admission rates among HF patients presenting to the ED, with institution and patient characteristics, and outcomes of 30-day hospitalization, repeat ED visits, and death.
METHODS Using the National Ambulatory Care Reporting System, we identified patients presenting to an ED in Ontario, Canada, with HF (Apr 2004-Mar 2008). Hospitalizations were identified using the Canadian Institute for Health Information Database. Based on admission rates for HF patient visits, we classified EDs into low (L), medium (M), or high (H) admission rate tertiles, and examined outcomes of those discharged from the ED.
RESULTS Among 78,259 HF patients, hospitalization rates were 80%, 71%, and 55% at H (n=25,946), M (n=26,330), and L (n=25,983) admission rate institutions, respectively (p<0.001). Mean age (76-77 [SD 12] yrs), %male (49-50%), and comorbidities (Charlson score 1.6-1.7 [SD 2.0-2.1]) were similar across ED tertiles. Compared to L-admission institutions, H-admission EDs were more often teaching or large (94% [H] vs 53% [L]), urban (94% [H] vs 38% [L]), and had greater total bed capacity (210±108 [H] vs 61±92 [L]). Multivariable-adjusted hazards ratios [aHR] for 30-day repeat ED visits for HF were 1.32 (95%CI 1.19-1.47, p<0.001) for L and 1.16 (95%CI 1.04-1.30, p=0.008) for M vs. H-admission EDs. At L-admission EDs, aHR for HF-related ED visits or hospitalizations was 1.19 (95%CI 1.08-1.31, p<0.001). Patients returning to the ED for HF within 7 days after initial ED discharge exhibited increased 30-day mortality risk with an adjusted odds ratio 2.79 (95%CI 2.22-3.51, p<0.001) compared to those who did not return to the ED. Cubic regression splines accounting for ED-level clustering showed an inverse relationship between HF admission rate and log odds of repeat ED visits for HF.
CONCLUSIONS Without system-wide use of acute HF risk-stratification, institutions with L-admission rates exhibited worse outcomes after ED discharge, including higher rates of repeat ED visits, hospitalizations and mortality.
- © 2012 by American Heart Association, Inc.