First-line eradication of H pylori infection in Europe: A meta-analysis based on congress abstracts, 1997-2004.

摘要

AIM: To meta-analyse the European abstracts presented between 1997-2004 at the European H pylori Study Group, United European Gastroenterology Week meetings and World Congresses of Gastroenterology. METHODS: The abstracts of randomized/controlled prospective studies were classified into groups based on first-line eradication schedules. The quality of the abstracts was checked by a validated score system. The pooled eradication rates (PER) and combined odds ratios (OR) were calculated and compared with the published meta-analyses. RESULTS: The PER of proton pump inhibitor-based (PPI) one week triple therapies was 81.4% (confidence interval, 95% CI: 78.5-84.5). Ranitidine bismuth citrate-based (RBC) triple regimens have an efficiency rate of 78.5% (95% CI: 70.5%-84.3%) (P = 0.28 vs PPI). The OR for PPI effect vs RBC regimens was 1.1 (95% CI: 0.92-1.30). H(2) receptor antagonist-based triple therapies achieved 64.1% (95% CI: 52.6-75.6) (P = 0.02 < 0.05 vs PPI), the OR vs PPI regimens was 1.55 (95% CI: 0.72-3.78). PPI-based double combinations were less efficient than triple regimens (PER: 55.0%, OR: 4.90, 95% CI: 2.36-9.70). Quadruple regimens were successful in 82.6% (95% CI: 76.0-89.7), the OR vs triple therapies was 0.80 (0.62-1.03). Clarithromycin + amoxicillin or nitroimidazole combinations were efficient in 80.5% (95% CI: 77.2-84.2) and 83.8% (95% CI: 81.7-85.9), respectively. Amoxicillin + nitromidazole therapies eradicated the infection in 73.5% (66.6-78.5) (P = 0.01 < 0.05 vs clarithromycin-based regimens). CONCLUSION: PPI/RBC-based triple therapies achieved comparable results with the meta-analyses. H(2)-receptor antagonists and PPI-based double combinations were less efficient. Triple and quadruple regimens were equally effective. Clarithromycin + either amoxicillin or nitroimidazole containing regimens were more effective than amoxicillin + nitroimidazole combinations. High quality congress abstracts constitutes a valuable pool of data which is suitable for meta-analytical workup.