Abstract 14895: Predictors of Six-Month Mortality for Older Medicare Beneficiaries Hospitalized for Heart Failure: Plans for the Development and Validation of University of Alabama at Birmingham Heart Failure Model to Optimize Discharge Hospice Referral
Introduction: Reduction of hospital readmission to reduce the cost of Medicare is an Affordable Care Act priority and over the next 10 years hospitals are projected to lose over $7 billion in Medicare payments due to above-average 30-day all-cause readmission, for which heart failure (HF) is the leading cause. Findings from the Alabama HF Project demonstrated that among those who did not receive a discharge hospice referral, 20% died within 6-month of discharge and compared with these patients, those receiving a discharge hospice referral had a significant lower risk of 30-day all-cause readmissions. The objective of the current study is to develop a model to identify HF patients at risk for 6-month mortality so that the discharge hospice referral process may be optimized.
Methods: In the Alabama HF Project, of the 8032 Medicare beneficiaries hospitalized for decompensated HF in 103 U.S. hospitals between 1999 and 2001, 7042 were aged ≥65 years. Of these, 170 (2.4%) received discharge hospice referral and were excluded. Of the remaining 6872 patients, 1492 (22%) died within six months and were considered hospice-eligible. Multivariable-adjusted Cox regression model was used to identify baseline characteristics significantly associated with 6-month post-discharge mortality.
Results: Patients had a mean age of 78 (±8) years, 60% were women, 22% African American. 30-day all cause hospital readmission occurred in 37% of patients who died and 17% who did not die within 6 months post-discharge (p <0.001). Multivariable-adjusted predictors for 6-month post-discharge mortality are depicted in Table below. Race, Left ventricular ejection function, serum potassium, intensive care stay or cardiology care had no significant association with 6-month mortality. The model had a c-statistic of 0.73 (95% CI, 0.71-0.74).
Conclusions: Several routinely available clinical and laboratory characteristics can reliably predict 6-month mortality in patients with HF.
Author Disclosures: S. Mazimba: None. R. Kheirbek: None. N.M. Shara: None. C. Morgan: None. M. Bakitas: None. O. Intrator: None. M. Safford: None. P. Deedwania: None. G. Fonarow: None. W.S. Aronow: None. R.D. Fletcher: None. R.M. Allman: None. A. Ahmed: None.
- © 2014 by American Heart Association, Inc.