OBJECTIVE: Given the high cost of caring for patients with congestive heart failure, there are strong incentives to decrease hospital costs by shortening length of hospital stay. We sought to identify factors associated with length of stay among patients admitted for the treatment of heart failure resulting from systolic dysfunction. STUDY DESIGN: Retrospective cohort study. METHODS: We examined data from patients with a principal discharge diagnosis of congestive heart failure who had been admitted to 1 of the 49 academic hospitals across the United States that participated in the CHF Benchmark Project, a large collaborative quality improvement project coordinated by the University HealthSystem Consortium. Patients were discharged between January 1 and June 30, 1996. We obtained patient characteristics and hospitalization data by retrospectively reviewing medical records. We used linear regression models to identify major determinants of length of stay. RESULTS: Among the 1046 patients eligible for the study, 59% were women, 55% were white, and 58% were aged 65 years or older. Adjusting for patient demographic and admission clinical characteristics, the mean length of stay was 4.9 +/- 0.9 days. Length of stay varied significantly among hospitals, even after adjusting for differences in patient characteristics. In multivariate regression models, factors that were independently associated with a significantly longer length of stay were prior renal failure, peripheral edema, atrial fibrillation, hyponatremia, urinary catheter on admission, initiation of an antiarrhythmic or warfarin, and major complications. Patient characteristics and hospital events combined explained 16% of the variation in the length of stay. Adjusting for the individual hospitals explained an additional 10% of the variation in the length of stay. CONCLUSIONS: Although a number of patient and hospitalization factors were associated with length of stay in patients with congestive heart failure resulting from systolic dysfunction, much unexplained variation remained. Clinical factors alone explained about 50% more variation than did factors specific to the individual hospitals.
OBJECTIVE: Given the high cost of caring for patients with congestive heart failure, there are strong incentives to decrease hospital costs by shortening length of hospital stay. We sought to identify factors associated with length of stay among patients admitted for the treatment of heart failure resulting from systolic dysfunction. STUDY DESIGN: Retrospective cohort study. METHODS: We examined data from patients with a principal discharge diagnosis of congestive heart failure who had been admitted to 1 of the 49 academic hospitals across the United States that participated in the CHF Benchmark Project, a large collaborative quality improvement project coordinated by the University HealthSystem Consortium. Patients were discharged between January 1 and June 30, 1996. We obtained patient characteristics and hospitalization data by retrospectively reviewing medical records. We used linear regression models to identify major determinants of length of stay. RESULTS: Among the 1046 patients eligible for the study, 59% were women, 55% were white, and 58% were aged 65 years or older. Adjusting for patient demographic and admission clinical characteristics, the mean length of stay was 4.9 +/- 0.9 days. Length of stay varied significantly among hospitals, even after adjusting for differences in patient characteristics. In multivariate regression models, factors that were independently associated with a significantly longer length of stay were prior renal failure, peripheral edema, atrial fibrillation, hyponatremia, urinary catheter on admission, initiation of an antiarrhythmic or warfarin, and major complications. Patient characteristics and hospital events combined explained 16% of the variation in the length of stay. Adjusting for the individual hospitals explained an additional 10% of the variation in the length of stay. CONCLUSIONS: Although a number of patient and hospitalization factors were associated with length of stay in patients with congestive heart failure resulting from systolic dysfunction, much unexplained variation remained. Clinical factors alone explained about 50% more variation than did factors specific to the individual hospitals.