RESUMO
PURPOSE: To compare peristomal infection rates following percutaneous gastrostomy (PG) after a single dose of prophylactic antibiotics versus placebo and evaluate rates of peristomal infection in patients receiving concurrent antibiotics. MATERIALS AND METHODS: This single-center, randomized trial (2012-2016) enrolled 122 patients referred for image-guided PG; all enrolled patients completed the study. Of enrolled patients, 68 were randomly assigned to receive either antibiotics (n = 34) or placebo (n = 34) before PG placement. The remaining 54 patients were taking pre-existing antibiotics and were assigned to an observation arm. Stoma sites were assessed for signs of infection by a blinded evaluator at early (between 3-5 d and 7-10 d) and late (between 14-17 d and 28-30 d) time points after the procedure. The primary outcome was peristomal infection. RESULTS: Under intention-to-treat analysis, early infection rate was 11.8% (4/34 patients; 95% CI, 0.0%-9.4%) in the placebo arm and 0.0% (0/34 patients; 95% CI, 0.0%-8.4%) in the antibiotic arm (P = .057 for comparison of infections in the 2 arms). Under per-protocol analysis, early infection rate was 13.3% (4/30 patients; 95% CI, 4.4%-29.1%) in the placebo arm and 0.0% (0/32 patients; 95% CI, 0.0%-8.9%) in the antibiotic arm (P = .049). The number needed to treat to prevent 1 early infection was 8.5 and 7.5 from the 2 analyses, respectively. CONCLUSIONS: There is a trend toward reduction in rate of peristomal infection after PG when prophylactic antibiotics are administered.