LAUSR

Background Ultrafiltration (UF) is a novel non-pharmacological therapy for decompensated chronic heart failure (CHF) patients, although its precise clinical role remains unclear. Current research suggests that decompensated CHF patients refractory to intravenous diuretics may benefit from UF therapy. We report our experience of using UF and evaluate its safety and impact on renal function and length of hospital stay in a select group of patients who did not respond to intravenous diuretics. Method UF treatment was delivered by the Aquadex Flexflow device via a central venous catheter. A retrospective analysis was performed of patients who underwent UF from January 2014 to June 2017. We analysed pre-UF diuretic dose, fluid volume removed, weight change, renal function, BNP and admission duration. Results A total of 13 decompensated CHF patients underwent UF (mean age 72 years (55–86 years); 10 males, 3 females; all with LV systolic impairment, median Ejection Fraction 27.5% (12.3%–49%)). All patients received optimal heart failure medication prior to hospital admission, including an angiotensin-converting enzyme inhibitor/angiotensin receptor blocker, beta blocker and mineralocorticoid receptor antagonist. Median intravenous furosemide dose prior to UF was 180 mg (120–240 mg) daily, given as a continuous infusion or as bolus injections. Nine patients received metolazone and 2 patients were prescribed dobutamine infusions. Mean daily weight loss while receiving intravenous diuretic treatment was 0.1 kg/day over a median duration of 6 days (4–30 days), indicating diuretic resistance. During UF, mean daily weight loss was substantially higher at 2 kg/day, with a mean total weight loss of 6.8 kg, over a median treatment duration of 60 hours (36–72 hours). Mean fluid volume removed was 6.5 l. Patients remained in hospital for a median duration of 3 days post UF (mean: 6.9 days; 1–21 days). Seven patients had BNP levels measured. The mean BNP level fell from 834 pg/ml to 709 pg/ml post UF. Patients had varying degrees of renal impairment prior to UF, with a mean serum creatinine level of 152 ümmol/l (97–233ümmol/l). Renal function remained stable post UF, with only a mild rise in creatinine level of 19ümmol/l. In 2 patients, serum creatinine rose more than 50 ümmol/l post UF, but subsequent testing showed that renal function returned to baseline within 2 weeks. We did not encounter any adverse events including thrombosis or bleeding, with strict APTT monitoring. There were no cases of infection. Conclusion UF appears to be an effective treatment in decompensated CHF patients with diuretic resistance. In our experience, its use was safe and was not associated with a deleterious effect on renal function. It resulted in an improved BNP level and for most patients enabled an early hospital discharge following its use. Our data suggests that UF is feasible and beneficial in select decompensated CHF patients with little response to intravenous diuretic therapy.