Impact of Advanced Therapy Centers on Characteristics and Outcomes of Heart Failure Admissions

Abstract

BACKGROUND: Although much attention has been paid to admission and transfer patterns for cardiogenic shock, contemporary data are lacking on decompensated heart failure (HF) admissions and transfers and the impact of advanced therapy centers (ATCs) on outcomes. METHODS: HF hospitalizations were obtained from the Nationwide Readmissions Database 2016 to 2019. Centers performing at least 1 heart transplant or left ventricular assist device were classified as ATCs. Patient characteristics, outcomes, and procedural volume were compared among 3 cohorts: admissions to non-ATCs, admissions to ATCs, and transfers to ATCs. A secondary analysis evaluated outcomes for severe HF hospitalizations (cardiogenic shock, cardiac arrest, and mechanical ventilation). Multivariable logistic regression was performed to adjust for the presence of HF decompensations and significant clinical variables during univariate analysis. RESULTS: A total of 2 331 690 hospitalizations (81.2%) were admissions to non-ATCs (94.5% of centers), 525 037 (18.3%) were admissions to ATCs (5.5% of centers), and 15 541 (0.5%) were transferred to ATCs. Patients treated at ATCs (especially those transferred) had higher rates of HF decompensations, procedural frequency, lengths of stay, and costs. Unadjusted mortality was 2.6% at non-ATCs and was higher at ATCs, both for directly admitted (2.9%, P <0.001) and transferred (11.2%, P <0.001) patients. However, multivariable-adjusted mortality was significantly lower at ATCs, both for directly admitted (odds ratio, 0.82 [95% CI, 0.78–0.87]; P <0.001) and transferred (odds ratio, 0.66 [95% CI, 0.57–0.78]; P <0.001) patients. For severe HF admissions, unadjusted mortality was 37.2% at non-ATCs and was lower at ATCs, both for directly admitted (25.3%, P <0.001) and transferred (25.2%, P <0.001) patients, with similarly lower multivariable-adjusted mortality. CONCLUSIONS: Patients with HF treated at ATCs were sicker but associated with higher procedural volume and lower adjusted mortality.