Sustained low-efficiency extended dialysis (SLED) with single-pass batch system in critically-ill patients with acute kidney injury (AKI)

Abstract

Background: Single-pass batch dialysis (SBD) is a well-established system for treatment of end-stage renal disease. However, little evidence is available on sustained low-efficiency extended dialysis (SLED) performed with SBD in patients with acute kidney injury (AKI) in the intensive care unit (ICU). Methods: All SLED-SBD sessions conducted on AKI patients in nine ICUs between March and June 2010 were retrospectively analyzed regarding the achieved metabolic and fluid control. Logistic regression was performed to identify the risk factors associated with hypotension and clotting during the sessions. Results: Data from 106 patients and 421 sessions were analyzed. Patients were 54.2 ± 17.0 years old, 51 % males, and the main AKI cause was sepsis (68 %); 80 % of patients needed mechanical ventilation and 55 % vasoactive drugs. Hospital mortality was 62 %. The median session time was 360 min [interquartile range (IQR) 300–360] and prescribed ultrafiltration was 1500 ml (IQR 800–2000). In 272 sessions (65 %) no complications were recorded. No heparin was used in 269/421 procedures (64 %) and system clotting occurred in 63 sessions (15 %). Risk factors for clotting were sepsis [odds ratio (OR) 2.32 (1.31–4.11), p = 0.004], no anticoagulation [OR 2.94 (1.47–5.91), p = 0.002] and the prescribed time (hours) [OR 1.14 (1.05–1.24), p = 0.001]. Hypotension occurred in 25 % of procedures and no independent risk factors were identified by logistic regression. Adequate metabolic and fluid balance was achieved during SLED sessions. Median blood urea decreased from 107 to 63 mg/dl (p < 0.001), potassium from 4.1 to 3.9 mEq/l (p < 0.001), and increased bicarbonate (from 21.4 to 23.5 mEq/l, p < 0.001). Median fluid balance during session days ranged from +1300 to −20 ml/24 h (p < 0.001). Conclusions: SLED-SBD was associated with a low incidence of clotting despite frequent use of saline flush, and achieved a satisfactory hemodynamic stability and reasonable metabolic and fluid control in critically-ill AKI patients.