Prospective randomized trial of 10% povidone-iodine versus 0.5% tincture of chlorhexidine as cutaneous antisepsis for prevention of central venous catheter infection.

A multicenter prospective, randomized, controlled trial, with 0.5% tincture of chlorhexidene versus 10% povidone-iodine as cutaneous antisepsis for central venous catheter (CVC) insertion, was conducted for patients in intensive care units. Of 374 patients, 242 had a CVC inserted for >3 days and were used for the primary analysis. Outcomes included catheter-related bacteremia, significant catheter colonization (> or = 15 colony-forming units [cfu]), exit-site infection, serial quantitative exit-site culture (every 72 h), and molecular subtyping of all isolates. Patients in both study groups were comparable with respect to age, sex, underlying disease, length of hospitalization, reason for line insertion, and baseline APACHE II score. Documented catheter-related bacteremia rates were 4.6 cases per 1000 catheter-days in the chlorhexidine group (n=125) and 4.1 cases per 1000 catheter-days in the povidone-iodine group (n=117; not significant [NS]). Significant catheter-tip colonization occurred in 24 (27%) of 88 patients in the povidone-iodine group and in 31 (34%) of 92 patients in the chlorhexidine group (NS). A mean exit-site colony count of 5.9 x 10(5) cfu/mL per 25 cm(2) of the surface area of skin in the povidone-iodine group versus 3.1 x 10(5) cfu/mL per 25 cm(2) in the chlorhexidine group (NS) was found. There was a trend toward fewer exit-site infections in the chlorhexidine group (0 of 125 patients) versus those in the povidone-iodine group (4 of 117 patients; P=.053). Results of an intention-to-treat analysis were unchanged from the primary analysis. No difference was demonstrable between 0.5% tincture of chlorhexidine and 10% povidone-iodine when used for cutaneous antisepsis for CVC insertion in patients in the intensive care unit.