Abstract

AIMS: Inpatients with heart failure and renal impairment have poor outcomes and variable quality of care. We investigate treatment practice and outcomes in an unselected real-world cohort using historical creatinine measurements. METHODS AND RESULTS: Admissions between 1/4/2013 and 30/4/2015 diagnosed at discharge with heart failure were retrospectively analysed. Stages of chronic kidney disease (CKD) and acute kidney injury (AKI) were calculated from creatinine at discharge and 3-12 months before admission. We identified 1056 admissions of 851 patients (mean age 76 years, 56% Caucasian, 36% with diabetes mellitus, 54% with ischaemic heart disease, and 57% with valvular heart disease). CKD was common; 36%-Stage 3a/b, 11%-Stage 4/5; patients were older, more often diabetic, with higher potassium, lower haemoglobin, and more oedema but similar prevalence of left ventricular systolic dysfunction (LVSD) compared patients with Stages 0-2. AKI was present in 17.0% (10.4%-Stage 1, 3.7%-Stage 2, and 2.9%-Stage 3); these had higher potassium and lower haemoglobin than patients with no AKI. Length of stay was longer in Stage 4/5 CKD [11 days; P = 0.008] and AKI [13 days; P = 0.006]. Mortality was higher with Stage 4/5 CKD (13.8% compared with 7.7% for Stages 0-2 CKD (P = 0.036)] and increased with AKI (5%-no AKI, 20.9%-Stage 1, 35.9%-Stage 2, and 48.4%-Stage 3; P < 0.001). Adjusted for age, diabetes, and LVSD, both AKI and Stage 4/5 CKD were independent predictors of in-hospital mortality. In survivors with LVSD, the discharge prescription of angiotensin-converting enzyme inhibitors/angiotensin receptor blockers decreased with progressive CKD, [84%-no-mild, 59%-moderate, and 36%-severe CKD; P < 0.001]; this was not purely explained by hyperkalaemia. CONCLUSIONS: Inpatients with heart failure and renal impairment, acute and chronic, failed to receive recommended therapy and had poor outcomes.