Double-blind prospective randomized study comparing topical mupirocin and placebo for the prevention of infection associated with central venous catheters.

OBJECTIVE

To compare the incidence of exit site colonization, local catheter-related infection and catheter-related bacteremia in patients randomized to receive either topical 2% mupirocin or placebo at the catheter exit site.

PATIENTS AND METHODS

Patients requiring central venous catheters for more than three days were randomized to receive in a double-blind fashion either topical mupirocin or an identical placebo at the exit site three times weekly at the time of dressing change. Insertion, site care and removal of catheters were standardized. Serial semiquantitataive cultures of the skin at the catheter insertion site were performed using a sterile 25 cm(2) template. The distal and proximal catheter segments were cultured using a standardized semiquantitative technique, and any suspect catheter-related bacteremia was investigated with two sets of peripheral blood cultures, a 10 mL sample of infusate and clinical assessment. Both univariate and multivariate analyses were conducted on individual risk factors to determine factors that might influence the outcomes of local or systemic catheter-related infection.

RESULTS

Local catheter-related infection (defined as more than 15 colony forming units [cfu] on culture of the proximal or distal catheter segment) occurred in six of 57 (10.5%) in the mupirocin group versus 18 of 69 (26%) in the placebo group (P<0.05) for the distal catheter segments and in one of 40 (2.5%) versus 13 of 47 (27.6%) for the proximal segments in the mupirocin and placebo groups (P<0.006), respectively. Catheter-related bacteremia occurred in one of 57 (1.8%) of the mupirocin group but in five of 69 (7.2%) of the placebo group (P=0.15). Stepwise logistic regression revealed that cutaneous colonization at the insertion site of at least 10(3) cfu/mL/25 cm(2) (OR 2.6; CI 1.0 to 6.9) and the use of placebo (OR 3.3; CI 1.2 to 9.0) were significant factors predicting local catheter-related infection; whereas mupirocin was associated with reciprocal protective effect (OR 0.3; 95% CI 0.1 to 0.8).

CONCLUSIONS

These findings suggest that patients receiving topical mupirocin at the exit site for long term central venous catheters have significantly less local catheter-related infection, and there is a trend towards less catheter-related bacteremia.