Portal de Periódicos da CAPES

Prokinetics in acute upper GI bleeding: a meta-analysis

2010; Elsevier BV; Volume: 72; Issue: 6 Linguagem: Inglês

10.1016/j.gie.2010.08.011

1097-6779

Alan Barkun, Marc Bardou, Myriam Martel, Ian M. Gralnek, Joseph J.�Y. Sung,

Gastric Cancer Management and Outcomes

Background Recent data suggest that administration of prokinetics before gastroscopy may be useful in patients with acute upper GI bleeding (UGIB). Published studies are limited in the number of subjects evaluated, and the conclusions are disparate. Objective To assess the evidence of administering prokinetic agents before EGD in acute UGIB. Design and Setting Comprehensive literature searches from 1990 to January 2010 were performed. We selected for meta-analysis randomized trials assessing prokinetic agents in acute UGIB. The primary outcome was the need for a repeat EGD. Secondary outcomes included endoscopic visualization, blood transfusions, duration of hospitalization, and surgery. Results were reported as odds ratios (ORs) or weighted mean differences (WMDs). Results From 487 citations identified, we selected 3 fully published articles and 2 abstracts assessing a total of 316 patients. Erythromycin (3 studies) and metoclopramide (2 studies) were compared with either placebo (2 studies) or no treatment (3 studies). A prokinetic agent significantly reduced the need for repeat EGD (OR 0.55; 95% CI, 0.32-0.94). The number of units of blood was not significantly altered (WMD, −0.40; 95% CI, −0.86 to 0.06) nor was hospital stay (WMD, −1.04; 95% CI, −2.83 to 0.76) or the need for surgery (OR 1.11; 95% CI, 0.27-4.67). Endoscopic visualization was not analyzed because the disparate definitions across studies did not allow for meaningful clinical inferences. Limitations The results are limited by the small number of subjects. Conclusions Intravenous erythromycin or metoclopramide immediately before EGD in acute UGIB patients decreases the need for a repeat EGD, but does not improve other clinically relevant measurable outcomes.