Objective: To study the effectiveness of local antiseptic spray with or without a three-dose antibiotic regimen in the prevention of stomal infection following percutaneous endoscopic gastrostomy.
Methods: Ninety-six patients were randomized into three groups: A, intravenous cefuroxime 750 mg just before the procedure followed by two further doses every 8 h; B, single application of povidone-iodine (Betadine) antiseptic spray; C, combination of A and B. The stomal site was examined at midweek and at the end of week 1 for evidence of infection using a validated scoring system. Fisher's exact test was used for analysis of primary end point, namely, stomal infection at midweek and at the end of week 1. Logistic regression models were used for secondary analysis to consider the effects of diabetes, acid suppressants, steroids, age and sex on outcome.
Results: Group A had 34 patients, group B had 28 and group C had 34. Age, sex and indications in groups A, B and C are broadly comparable. Stomal infection was 32% in group B vs. groups A (6%) and C (9%) (P = 0.0114) at midweek, and 3% in group C vs. 32% each in groups A and B (P = 0.0013) at the end of week 1. Cumulative infections (n) at the end of week 1 were lower in group C (3) (9%) than in groups A (11) (32%) and B (12) (43%) (P = 0.003). No significant difference was observed between the three groups in terms of the number of patients who were given antibiotics for other indications (P = 0.363). By logistic regression only diabetes, but not other covariates, seems to have a significant effect on stomal infection (odds ratio, 33.34; 95% CI, 4.33-256.7).
Conclusion: A combination of cefuroxime and Betadine spray significantly reduces stomal infection following percutaneous endoscopic gastrostomy at the end of week 1. Both the antibiotic group (A) and the combined group (C) did well compared with the Betadine only group (B) at midweek.