Abstract: Objectives To investigate the effects on blood coagulation, circulation status and filter life in critically ill patients with high bleeding risk requiring continuous renal replacement therapy (CRRT) without anticoagulation. Methods Patients with high risk of bleeding requiring CRRT without anticoagulation in the ICU were randomly divided into saline flushes group and no-rinse group. Levels of prothrombin fragment 1+2 (PF1+2) and β-thromboglobulin (β-TG) were measured, and the transmembrane pressure (TMP), pressure before the filter (PBE) and filter pressure decrease value (ΔP) were observed. Changes of heart rate and blood
pressure were monitored during CRRT. Results There was no significant difference in the levels of PF1+2 and β-TG between saline flushes group and no-rinse group. In no-rinse group the TMP at 3h (t=3.813, P＜0.001), 4h (t=4.230, P＜0.001) and at the end of CRRT (t=7.014, P＜0.001) were significantly higher than
that at the start of CRRT. In saline flushes group, the TMP at 4h (t=3.296, P=0.002) and at the end of CRRT (t=3.930, P＜0.001) were significantly higher than that at the start of CRRT. Filter pressure in the two groups showed significant differences after 1h (u=3.056, P=0.002), 2h (u=2.788, P=0.005), 3h (u=2.009, P=0.045) and at the end of CRRT (u=2.201, P=0.043). In no- rinse group the filter pressure decrease value at 1h (t=2.738, P=0.009), 2h (t=3.590, P＜0.001), 3h (t=4.771, P＜0.001), 4h (t=4.754, P＜0.001) and at the end of CRRT (t=5.144, P＜0.001) were significantly higher than that at the start of CRRT. In saline flushes group, the filter pressure decrease value at 3h (t=3.013, P=0.005), 4h (t=3.020, P=0.005) and at the end of CRRT (t=3.814, P=0.001) were significantly higher than that at the start of CRRT. Filter pressure decrease value in the two groups showed significant differences at the end of CRRT (u=2.155, P=0.031). The CRRT treatment time were more than 5 hours in all patients in the two groups. CRRT treatment was terminated in 5 patients (one case in saline flushes group and 4 cases in no-rinse group) due to high pressure limit alarm. The rest of the patients completed continuous veno-venous hemofiltration (CVVH) for 6～10h following doctor's advice. Compared between the two groups, there were significant differences in systolic pressure changes at 1h (t=2.845, P=0.007), 2h (t=3.353, P=0.002), 3h (t=3.367, P=0.002), 4h (t=3.745, P=0.001), 5h (t=3.355, P=0.002) and at the end of CRRT (t=2.711, P=0.010), significant differences in mean arterial blood pressure changes at 2h (t=2.508, P=0.016), 3h (t=3.078, P=0.004), 4h (t=3.023, P=0.004), 5h (t=2.412, P=0.021), significant differences in pulse pressure values at 2h (t=2.635, P=0.012), 3h (t=2.805, P=0.008), 4h (t=3.070, P=0.004), 5h (t=2.893, P=0.006) and at the end of CRRT (t=2.254, P= 0.030), and significant differences in pulse pressure changes at 1h (t=2.769, P=0.008), 2h (t=3.154, P=0.003), 3h (t=2.614, P=0.013), 4h (t=2.973, P=0.005), 5h (t=3.063, P=0.004) and at the end of CRRT (t=2.672, P=0.011). Conclusion Physiological saline irrigation is conducive to slow down the rise of pressure value in the machine and to prevent blood coagulation in extracorporeal circuit in CRRT for more than 5h. Patients with acute or chronic renal failure with high risk of bleeding associated with cardiovascular disease should be treated with saline irrigation without anticoagulation during CRRT.

Key words: Continuous renal replacement therapy, Saline flushes, Anticoagulation, Clotting, Nursing