Abstract BACKGROUND AND AIMS Haemodialysis catheters are frequently dysfunctional due to catheter-related thrombosis, which potentially compromises dialysis adequacy and may limit catheter survival. Strategies to prevent haemodialysis catheter dysfunction include catheter locks, that are generally either heparin or citrate. Our haemodialysis unit converted to locking all central venous haemodialysis catheters with sodium citrate 4% instead of heparin 10 000 U/mL. A retrospective analysis was conducted to evaluate whether replacing heparin with sodium citrate 4% would ensure cost-effective satisfactory catheter function. METHOD Our haemodialysis unit converted to locking all central venous haemodialysis catheters with sodium citrate 4% instead of heparin 10 000 U/mL as a unit policy change. In the heparin period, all catheters were locked with concentrated heparin 10 000 U/mL, in the citrate period, all catheters were locked with sodium citrate 4%. Partially or fully clotted catheters were treated with rt-PA. A retrospective analysis compared the outcomes of the 6 months prior and after the conversion. We included all patients in whom a central venous haemodialysis catheter was used during at least part of each period. Exclusion criteria included patients with a contraindication to systemic anticoagulation, documented or suspected heparin-induced thrombocytopaenia, or known allergy to citrate. Age, gender, dialysis vintage and comorbidities were analysed. The compared outcomes were incidence of catheter thrombosis, catheter-related infection, flow-related catheter exchange rate, dialysis efficiency measured by Kt/V and catheter dysfunction. RESULTS A total of 23 patients were included, with an age of 68 ± 16 years. Dialysis vintage was 2.8 ± 2 years. The patients had multiple comorbidities, with a mean Charlson comorbidity index of 7 ± 2. Regarding catheter characteristics, 19 (83%) had a right jugular vein tunnelled catheter, the remaining 4 (17%) had a left jugular vein tunnelled catheter. These 23 patients were dialyzed using a central venous haemodialysis catheter during both the heparin and citrate lock period, so the patient characteristics are the same in both groups. Between the two audit periods, the rate of catheter thrombosis measured by a number of treatments with rt-PA was 0 (0/1000 catheter days) and 7 (2.78/1000 catheter days) during the heparin and citrate period respectively (P = 0.08). The number of catheter exchanges due to catheter dysfunction was 0 (0/1000 catheter days) during the heparin period and just 1 (0.39/1000 catheter days) during the citrate period (P = 0.317). There were no catheter-related infections in either group. Dialysis adequacy measured by Kt/V was 1.55 ± 0.34 in the heparin group and 1.59 ± 0.34 in the citrate group (P = 0.725). Regarding catheter dysfunction (inability to achieve and maintain a blood flow of more than 350 mL/min), it was recorded in 146 sessions (12.3%) in the heparin period and 59 sessions (14.7%) in the citrate period (P =0.729). CONCLUSION When comparing catheter lock with heparin versus citrate, this study did not find any statistically significant differences regarding major catheter-related outcomes. However, there was a trend favouring heparin locking. In our institution, the direct cost of citrate locking is considerably less than the cost of heparin locking. Given the lack of major differences in outcomes, the total cost-benefit is likely to be greater with citrate locking (although cost related to thrombolytic use was not assessed). Besides potential financial advantages in switching to citrate locking, there are other potential advantages, namely the minimization of heparin-related side effects or limitations. In conclusion, citrate locking is an effective and less expensive alternative to heparin locking of haemodialysis catheters.