Background:
Readmissions are frequent and costly outcomes in patients hospitalized for heart failure (HF) and acute myocardial infarction (AMI). Knowledge of the exact timing of 30-day readmissions after hospitalization for HF and AMI can help identify time periods during which patients are at the highest readmission risk and guide the development of interventions designed to prevent early readmissions.
Methods:
Using Medicare Standard Analytic and Denominator files, we identified all HF and AMI hospitalizations in 2007-2009. We excluded hospitalizations for patients aged<65, transferred out, discharged against medical advice, or with an inpatient death. For both HF and AMI cohorts, we identified all readmissions to short-stay acute care hospitals due to any cause occurring within 30 days of hospital discharge except for planned coronary revascularization. Our primary outcome was the number of observed readmissions occurring during each day (0-30) after discharge. We also calculated the cumulative number of observed readmissions occurring during the 1
st
3 days, 1
st
week, and 1
st
15 days after discharge. We used a one-tailed two-proportion z test to evaluate if the proportion of readmissions during the 1
st
3 days, 1
st
week, and 1
st
15 days was higher than what would be expected had readmissions occurred at an equal rate during the 30 days (alpha=0.05).
Results:
We identified 329,308 readmissions within 30 days after 1,330,157 hospitalizations for HF (4,633 hospitals) and 108,992 readmissions within 30 days after 548,834 hospitalizations for AMI (3,895 hospitals). Readmission frequency by day is described for both HF and AMI in the accompanying figure. Following hospitalization for HF, 13.4% of 30-day readmissions occur during the 1
st
3 days after discharge, 31.7% occur during the 1
st
week, and 61.0% occur during the 1
st
15 days. Following hospitalization for AMI, 19.1% of 30-day readmissions occur during the 1
st
3 days after discharge, 40.1% occur during the 1
st
week, and 67.6% occur during the 1
st
15 days. For both HF and AMI cohorts, readmissions after 3, 7, and 15 days were higher than what would be predicted had readmission rates remained constant (p<0.0001 for all).
Conclusion:
For patients hospitalized with HF and AMI, a disproportionately high percentage of 30-day readmissions occur soon after discharge. Interventions designed to reduce hospital readmissions may therefore generate substantive benefits when applied to the time period shortly after hospitalization.
