Methods: 75 CKD 5D patients on high-efficiency hemodialysis (blood flow rate = 350 ml/min) at the Integrated Centre of Nephrology (Guarulhos, Brazil) will be randomized 1:1:1 to receive M-EDTA, C or H locks for 15 weeks. Lock solutions concentrations are M-EDTA 30 mg/ml/3 mg/ml, C 30% (C) and H 1,000 U/ml and both investigators and patients are blinded to treatment allocation. The primary end-point is a 10% reduction in HD blood flow rates (35ml). The frequency of CRBSI will be recorded. Poor Flow Rate (event/1,000 catheter-day) and CRBSI Rate (event/1,000 catheter-day) will be calculated in each arm. Bleeding and lock solution-related adverse events are also the primary safety end points and their rates (events/1,000 catheter-day) will be calculated. Logistic Regression will be performed to evaluate differences in PF, CRBSI, Bleeding and lock solution-related adverse events rates among the treatments (SPSS version 13.0, IBM, USA).
