AIM: To investigate whether continuous veno-venous hemofiltration (CVVH) in different filtration rate to eliminate cytokines would result in different efficiency in acute pancreatitis, whether the saturation time of filter membrane was related to different filtration rate, and whether the onset time of CVVH could influence the survival of acute pancreatitis. METHODS: Thirty-seven patients were classified into four groups randomly. Group 1 underwent low-volume CVVH within 48 h of the onset of abdominal pain (early CVVH, n = 9). Group 2 received low-volume CVVH after 96 h of the onset of abdominal pain (late CVVH, n = 10). Group 3 underwent high-volume CVVH within 48 h of the onset of abdominal pain (early CVVH, n = 9). Group 4 received high-volume CVVH after 96 h of the onset of abdominal pain (late CVVH, n = 9). CWH was sustained for at least 72 h. Blood was taken before hemofiltration, and ultrafiltrate was collected at the start of CVVH and every 12 h during CVVH period for the purpose of measuring the concentrations of TNPα, IL-1β and IL-6. The concentrations of TNPα, IL-1β and IL-6 were measured by swine-specific ELISA. The Solartron 1 255 B frequency response analyzer (British) was used to observe the resistance of filter membrane. RESULTS: The survival rate had a significant difference (94.44% vs 68.42%, P<0.01) high-volume and low-volume CVVH patients. The survival rate had also a significant difference (88.89% vs 73.68%,P<0.05) between early and late CVVH patients. The hemodynamic deterioration (MAP, HR, CVP) was less severe in groups 4 and 1 than that in group 2, and in group 3 than in group 4. The adsorptive saturation time of filters membranes was 120-180 min if the filtration rate was 1 000-4 000 mL/h. After the first, second and third new hemofilters were changed, serum TNF-α concentrations had a negative correlation with resistance (r: -0.91, -0.89, and -0.86, respectively in group 1; -0.89, -0.85, and -0.76, respectively in group 2; -0.88, -0.92, and -0.82, respectively in group 3; -0.84, -0.87, and -0.79, respectively in group 4). The decreasing extent of TNF-α, IL-1β and IL-6 was significantly different between group 3 and group 1 (TNF-α P<0.05, IL-1β P<0.05, IL-6 P<0.01), between group 4 and group 2 (TNF-α P<0.05, IL-1β P<0.05, IL-6 P<0.01), between group 1 and group 2 (TNP-α P<0.05, IL-1β P<0.05, IL-6 P<0.05), and between group 3 and group 4 (TNF-α P<0.01, IL-1β P<0.01, IL-6 P<0.05), respectively during CVVH period. The decreasing extent of TNF-α and IL-1β was also significantly different between survival patients and dead patients (TNF-α P<0.05, IL-1β P<0.05). In survival patients, serum concentration of TNF-α and IL-1β decreased more significantly than that in dead patients. CONCLUSION: High-volume and early CVVH improve hemodynamic deterioration and survival in acute pancreatitis patients. High-volume CVVH can eliminate cytokines more efficiently than low-volume CVVH. The survival rate is related to the decrease extent of TNF-α and IL-1β. The adsorptive saturation time of filter membranes are different under different filtration rate condition. The filter should be changed timely once filter membrane adsorption is saturated. A: To investigate whether continuous veno-venous hemofiltration (CVVH) in result of different efficiency in acute pancreatitis, whether the saturation time of filter membrane was related to different filtration rate, and whether the onset time of CVVH could influence the survival of acute pancreatitis. METHODS: Thirty-seven patients were classified into four groups randomly. Group 1 underwent low-volume CVVH within 48 h of the onset of abdominal pain (early CVVH, n = 9). Group 2 received low Group 3 underwent high-volume CVVH within 48 h of the onset of abdominal pain (early CVVH, n = 9). Group 4 received high-volume CVVH after 96 h of the onset of abdominal pain (late CVVH, n = -volume CVVH after 96 h of the onset of abdominal pain (late CVVH, n = 9). CWH was sustained for at least 72 h. Blood was taken before hemofiltration, and ultrafiltrate was collected at the start of CVVH and every 12 h during CVVH period for the purpo The concentrations of TNPα, IL-1β and IL-6 were measured by swine-specific ELISA. The Solartron 1 255 B frequency response analyzer (British) was used to RESULTS: The survival rate had a significant difference (94.44% vs 68.42%, P <0.01) high-volume and low-volume CVVH patients. The survival rate had also a significant difference (88.89% vs 73.68 %, P <0.05) between early and late CVVH patients. The hemodynamic deterioration (MAP, HR, CVP) was less severe in groups 4 and 1 than that in group 2, and in group 3 than in group 4. The adsorptive saturation time after the first, second and third new hemofilters were changed, serum TNF-a concentrations had a negative correlation with resistance (r: -0.91 , -0.89, and -0.86, respectively in group 1; -0.89, -0.85, and -0.76, respectively in group 2; -0.88, -0.92, and-0.84, -0.87, and -0.79, respectively in group 4). The decreasing extent of TNF-α, IL-1β and IL-6 was significantly different between group 3 and group 1 IL-1 P <0.05, IL-6 P <0.01), between group 4 and group 2 (P <0.05) between group 1 and group 2 (TNP-α P <0.05, IL-1β P <0.05, IL-6 P <0.05) and between group 3 and group 4 , The decreasing extent of TNF-α and IL-1β was also significantly different between survival patients and dead patients (TNF-α P <0.05, IL-1β P <0.05) CONCLUSION: High-volume and early CVVH improve hemodynamic deterioration and survival in patients with acute pancreatitis. High-volume CVVH can eliminate cytokines more efficiently than low-volume CVVH. The survival rate is related to the decre ase extent of TNF-α and IL-1β. The adsorptive saturation time of filter membranes are different under different filtration rate conditions. The filter should be changed timely once filter membrane adsorption is saturated.