Objective   To evaluate the curative effect of gastroendoscopic titanium clips

injection with titanium clips

and injection with argon plasma coagulation (APC) for acute non-variceal upper gastrointestinal bleeding (ANVUGIB) (ForrestⅠ and Ⅱ

a

) and the effects of different Rockall scoring of rehemorrhage on endoscopic hemostasis.     Methods   A total of 182 cases of ANVUGIB of Forrest Ⅰ and Ⅱ

a

confirmed gastroendoscopically were randomly divided into group of titanium clips (78 cases)

group of injection combined with titanium clips (49 cases)

and injection with APC (55 cases) for treatment of bleeding. Rockall score was conducted to compare the hemostatic rate by the three methods and the effects of different Rockall scores on the curative effect in the patients were investigated.     Results   The hemostatic rates in the group of injection with titanium clips and the group of injection with APC were 95.92% and 98.18%

respectively

higher than that in the group of titanium clips (84.62%) (P0.05). The hemostatic rate in the group of titanium clips with Rockall risk score lower than 2 was 100%

but no difference was found in the hemostatic rate between the groups of injection with titanium clips and injection with APC (both of them were 100%) and the group of titanium clips (92.31%) when Rockall risk score was 3-4. When Rockall risk score was higher than 5

there were no difference in hemostatic rate between the group of injection with APC (95.45%) and the group of injection with titanium clips (96.30%)

but it was higher than that in the group of titanium clips (75%) (P0.05).     Conclusion   The curative effect of injection with titanium clips and injection with APC are reliable

and are superior to the treatment with titanium clips alone. The hemostatic rate is correlated with Rockall risk score.