Length of stay and risk of very early readmission in acute heart failure

Abstract

Introduction and objectives: In patients admitted for acute heart failure (AHF), optimal length of stay (LOS) remains controversial. Longer hospitalizations are associated with worse prognosis, but little is known about short hospitalizations. The aim of this work was to evaluate the relationship between LOS and the risk of short-term readmission in patients discharged after a hospitalization for AHF.

Methods: We included 2110 consecutive patients. The independent associations between LOS and unplanned 10, 15 and 30-day readmissions were evaluated by Cox regression analysis adjusted for competing events. LOS was categorized as LOS1: ≤4days, LOS2: 5-7days, LOS3: 8-10days, and LOS4: >10days.

Results: The mean age was 73±11years and 52.6% exhibited left ventricle ejection fraction≥50%. The median (IQR) LOS was 7 (5-11) days. At 10, 15 and 30-day follow-up, 130 (6.2%), 181 (8.6%), and 282 (13.4%) unplanned readmissions were registered. Rates of 10 and 15-day readmission among LOS categories showed a J-shaped pattern with lower rates for those in LOS2 and higher at the both extremes (p=0.001). At 30-day, only longer stays showed higher rates of readmission (p=0.002). In the multivariate analysis, the U-shaped curve remained significant for 10 and 15-day readmissions (p<0.05). Compared to LOS2, LOS1, LOS3 and LOS4 showed about two-fold increased risk. At 30-day only longer stays showed a borderline and modest increase of risk.

Conclusions: Shorter and longer stays are associated with the risk of very early readmissions after an episode of AHF. These associations are marginal for 30-day readmissions.

Keywords: Acute heart failure; Early readmissions; Length of stay.