Favorable bed utilization and readmission rates for emergency department observation unit heart failure patients.

OBJECTIVES

The objective was to compare readmission rates and hospital bed-days between acute decompensated heart failure (AHF) patients admitted or discharged following accelerated treatment protocol (ATP)-driven care in an emergency department observation unit (OU).

METHODS

This was a retrospective cohort study conducted at two urban university-affiliated hospitals. A total of 358 selected AHF patients received treatment on an ATP in the OU between October 1, 2007, and June 30, 2011. The comparison of interest was admission or discharge following OU treatment. The outcome of interest was readmission within 30 and 90 days of hospital discharge following care in the OU. We also examined resource use (inpatient, inpatient plus outpatient-days) between the admitted and discharged groups. Time to readmission analysis was performed with Cox proportional hazards regression.

RESULTS

Discharged and admitted patients were similar with respect to age, race, sex, ED length of stay (LOS), and OU LOS. Patients admitted from the OU had a higher median B-type natriuretic peptide (BNP; 1,063 pg/mL [interquartile range {IQR} = 552 to 2,067 pg/mL] vs. 708 pg/mL [IQR = 254 to 1,683 pg/mL]; p = 0.002) and blood urea nitrogen (BUN; 19 mg/dL [IQR = 14 to 26 mg/dL] vs. 17 mg/dL [IQR = 13 to 23 mg/dL]) than those discharged (p = 0.04) and a lower median ejection fraction (EF; 22.5% [15% to 43%] vs. 35% [IQR 20% to 55%]; p = 0.002). In models controlling for age, race, sex, clinical site, BNP, BUN, creatinine, and EF, the 30-day readmission rate (13.8% in the study population as a whole) was not significantly different between the patients discharged or admitted following OU care (hazard ratio [HR] = 0.99; 95% confidence interval [CI] = 0.47 to 2.10). The readmission rates were also not significantly different at 90 days (HR = 1.07; 95% CI = 0.65 to 1.77). Within 30 days of discharge from the OU, patients spent a median of 1.7 days (IQR = 0.0 to 5.1 days) as inpatients, compared to 3.5 days (IQR = 2.3 to 5.8 days) among patients admitted from the OU (p < 0.0001). Among readmitted patients, the total median inpatient time was not significantly different between the comparison groups at both 30 and 90 days of follow-up.

CONCLUSIONS

Selected acute heart failure (HF) patients managed by a rapid treatment protocol in the OU demonstrated favorable hospital use, with discharged patients using fewer bed-days and demonstrating readmission rates that were not higher than admitted patients.