RESUMO
To compare the complications and mortality related to gastrostomy procedures performed using surgical and percutaneous endoscopic gastrostomy techniques, this review covered seven studies. Five of these were retrospective and two were randomized prospective studies. In total, 406 patients were involved, 232 of whom had undergone percutaneous endoscopic gastrostomy and 174 of whom had undergone surgical gastrostomy. The analysis was performed using Review Manager. Risk differences were computed using a fixed-effects model and forest and funnel plots. Data on risk differences and 95% confidence intervals were obtained using the Mantel-Haenszel test. There was no difference in major complications in retrospective (95% CI (-0.11 to 0.10)) or randomized (95% CI (-0.07 to 0.05)) studies. Regarding minor complications, no difference was found in retrospective studies (95% CI (-00.17 to 0.09)), whereas a difference was observed in randomized studies (95% CI (-0.25 to -0.02)). Separate analyses of retrospective and randomized studies revealed no differences between the methods in relation to mortality and major complications. Moreover, low levels of minor complications were observed among endoscopic procedures in randomized studies, with no difference observed compared with retrospective studies.