Differences in the prognoses of patients referred to an advanced heart failure center from hospitals with different bed volumes

Abstract

Few reports have discussed appropriate strategies for patient referrals to advanced heart failure (HF) centers with available left ventricular assist devices (LVADs). We examined the association between the characteristics and prognoses of referred patients with advanced HF and the bed volume of the referring hospitals. This retrospective analysis evaluated 186 patients with advanced HF referred to our center for consultation about the indication of LVAD between January 1, 2015, and August 31, 2018. We divided the patients into two groups according to the bed volume of their referring hospital (high bed volume hospitals (HBHs): ≥ 500 beds in the hospital; low bed volume hospitals (LBHs): < 500 beds). We compared the primary outcome measure, a composite of LVAD implantation and all-cause death, between the patients referred from HBHs and patients referred from LBHs. The 186 patients with advanced HF referred to our hospital, who were referred from 130 hospitals (87 from LBHs and 99 from HBHs), had a mean age of 43.0 ± 12.6 years and a median left ventricular ejection fraction of 22% [15-33%]. The median follow-up duration of the patients was 583 days (119-965 days), and the primary outcome occurred during follow-up in 42 patients (43%) in the HBH group and 20 patients (23%) in the LBH group. Patients referred from HBHs tended to require catecholamine infusion on transfer more often than those referred from LBLs (36.5% (HBH), 20.2% (LBL), P = 0.021). Kaplan-Meier analysis indicates that the occurrence of the primary outcome was significantly higher in the HBH patients than in the LBH patients (log-rank P = 0.0022). Multivariate Cox proportional hazards analysis revealed that catecholamine support on transfer and long disease duration were statistically significant predictors of the primary outcome. Patients from HBHs had a greater risk of the primary outcome. However, the multivariate analysis did not indicate an association between referral from an HBH and the primary outcome. In contrast, catecholamine support on transfer, long duration of disease, and low blood pressure were independent predictors of the primary outcome. Therefore, these should be considered when determining the timing of a referral to an advanced HF center, irrespective of the bed volume of the referring hospital.

Figure 1

Flowchart of the study depicting the referral of advanced HF patients from other hospitals to our advanced HF center. Certain exclusion criteria were applied. HF, heart failure; LVAD, left ventricular assist device.

Figure 2

Numbers of HF patients referred to our advanced HF center from HBHs and LBHs. HF, heart failure; LBHs, low bed volume hospitals; HBHs, high bed volume hospitals.

Figure 3

(a) Differences in event-free survival curves of the primary outcome measures, including LVAD implantation and death, between patients referred from HBHs and patients referred from LBHs. (b) Differences in event-free survival curves of the secondary outcome measures between patients referred from HBHs and patients referred from LBHs. LVAD, left ventricular assist device; LBHs, low bed volume hospitals; HBHs, high bed volume hospitals.

Figure 4

(a) Differences in event-free survival curves of the primary outcome measures, including LVAD implantation and death, between patients with and patients without catecholamine infusion during transfer. (b) Differences in event-free survival curves of the secondary outcome measures between patients with and patients without catecholamine infusion during transfer. LVAD, left ventricular assist device.