DETERMINANTS OF LENGTH OF HOSPITAL
STAY IN PATIENTS ADMITTED WITH HEART FAILURE S P Wright*, D Verouhis, H Walsh,
G Gamble, N Sharpe, R N Doughty Dept of Medicine, University of
Auckland, Auckland, New Zealand. Background: Over the last 10 years
the length of stay (LOS) for patients admitted with heart failure (HF) in New
Zealand has halved from 14 to 7 days. The reasons for this decrease in LOS over
the decade is unclear. In addition, the determinants of LOS for contemporary
admissions with HF are uncertain. This study examined the determinants of LOS
in 197 prospectively selected HF patients. Methods: The index admissions of
the 197 patients prospectively identified for enrolment into the Auckland Heart
Failure Management Study (a post-discharge intervention study) were studied.
Data was obtained from the prospective recruitment process and also from
retrospective chart review. Determinants of LOS for the index admission of each
patient were determined using univariate and multiple regression techniques. Results: 197 patients were
included: mean age 73 (SD 10.5), 60% male, 68% NYHA class 4; 55% had 1 or more
prior admissions for heart failure, 79% were caucasian. The median length of
stay was 6 days (IQR 4, 9). Many patients had a complication or concomitant
acute medical condition prolonging their stay: 11.7% had renal failure; 33.5%
angina, MI, or arrhythmia; 20% pneumonia or an exacerbation of COPD; and 12%
social problems requiring intervention. On univariate analysis, LOS > 6 days
was associated with the presence of oedema, chest pain, S3 or hepatomegaly at
admission; change in weight; number of medications at admission; frusemide dose
at admission; serum albumin, serum sodium, peak creatinine; the development of
complications or other acute problems during stay; and number of days treated
with iv frusemide. On multivariate analysis, the strongest predictor of LOS was
the number of days treated with iv frusemide. Other predictors included social
complications during stay, renal impairment, concurrent respiratory infection,
symptoms of peripheral congestion at admission, and LVEDV. Conclusion: LOS in HF is determined by the duration of treatment of iv
frusemide and the development of complications during hospital stay. A subset
of patients with severe peripheral congestion stayed longer. Based on this
data, LOS for patients with HF may be at a minimal level and may be difficult
to modify further.
