Table 1:

Patient-level analysis of compliance to guidelines for the management of nonvariceal upper gastrointestinal bleeding

GuidelineBefore intervention, % (no. of patients)After intervention, % (no. of patients)Intracluster correlation coefficientPercentage difference (95% CI)
Experimental groupControl groupExperimental groupControl group
Primary outcome
Endoscopic hemostasis followed by high-dose IV PPI13.2 (20/152)7.1 (11/155)9.8 (14/143)4.8 (6/124)0.04−5.0 (−12.0 to 2.0)
Secondary outcomes
Injection and/or thermal coagulation without clips and bolus IV PPI followed by infusion for 72 h for high-risk ulcers8.6 (13/152)5.2 (8/155)6.3 (9/143)4.8 (6/124)0.07−1.5 (−8.1 to 5.2)
Injection and/or thermal coagulation with clips and bolus IV PPI followed by 8 mg/h infusion for 72 h and no oral PPI. If the patient received a pre-endoscopy infusion, the absence of bolus IV PPI was ignored69.1 (105/152)58.7 (91/155)57.3 (82/143)62.9 (78/124)0.185.6 (−12.7 to 23.8)
Injection and/or thermal coagulation with clips and pre-endoscopy IV bolus PPI plus 80-mg bolus IV PPI followed by 8 mg/h infusion for 72 h after endoscopy and no oral PPI1.3 (2/152)1.9 (3/155)1.4 (2/143)0.8 (1/124)−0.05−0.6 (−2.5 to 1.3)
Injection and/or thermal coagulation upon endoscopy for high-risk ulcers with clips alone81.6 (124/152)74.8 (116/155)72.7 (104/143)72.6 (90/124)0.35−0.1 (−20.4 to 20.1)
Injection and/or thermal coagulation without clips for high-risk ulcers alone49.3 (75/152)56.1 (87/155)50.4 (72/143)46.8 (58/124)0.33−3.6 (−25.4 to 18.2)
80-mg IV bolus PPI followed by 8 mg/h PPI infusion for 72 h after endoscopy and no oral PPI15.3 (27/177)10.0 (17/170)13.6 (21/155)15.9 (24/151)0.272.3 (−12.1 to 16.8)
80-mg IV bolus PPI followed by 8 mg/h infusion for 72 h after endoscopy and no oral PPI. If patient received a pre-endoscopy infusion, the absence of bolus IV PPI afterward was ignored82.5 (146/177)80.0 (136/170)81.3 (126/155)89.4 (135/151)0.068.1 (−1.4 to 17.6)
Pre-endoscopy bolus of PPI plus 80-mg bolus IV PPI followed by 8 mg/h infusion for 72 h after endoscopy and no oral PPI1.1 (2/177)1.8 (3/170)1.9 (3/155)9.3 (14/151)0.357.3 (−2.9 to 17.5)
Use of validated scoring system based on a combination of clinical and endoscopic characteristics to stratify patients into low- and high-risk categories0.0 (0/402)0.2 (1/424)1.9 (7/361)0.0 (0/389)0.19−1.9 (−5.0 to 1.1)
Early endoscopy (in the first 24 h) with risk classification by clinical/endoscopic criteria62.9 (253/402)65.6 (278/424)57.3 (207/361)63.1 (245/388)0.095.6 (−5.6 to 16.9)
No hemostasis for low risk lesion88.9 (200/225)93.7 (223/238)94.2 (196/208)89.0 (218/245)0.13−5.2 (−13.2 to 2.8)
Endoscopic hemostasis for adherent clot87.9 (29/33)88.0 (22/25)80.7 (25/31)80.8 (21/26)0.090.1 (−22.9 to 23.1)
Endoscopic hemostasis for other active bleeding or visible vessel85.9 (152/177)83.3 (155/186)93.5 (143/153)86.1 (124/144)0.06−7.4 (−15.8 to 1.1)
Low-risk patients who received treatment11.1 (25/225)6.3 (15/238)5.8 (12/208)11.0 (27/245)0.135.2 (−2.8 to 13.2)
Low-risk patients not given IV PPI bolus or infusion after endoscopy61.8 (139/225)59.2 (141/238)56.3 (117/208)64.1 (157/245)0.058.0 (−2.9 to 18.9)
  • Note: CI = confidence interval, IV = intravenous, PPI = proton pump inhibitor.