ABSTRACT
Importance: Preoperative skin antisepsis is an established procedure to prevent surgical site infections (SSIs). The choice of antiseptic agent, povidone iodine or chlorhexidine gluconate, remains debated. Objective: To determine whether povidone iodine in alcohol is noninferior to chlorhexidine gluconate in alcohol to prevent SSIs after cardiac or abdominal surgery. Design, Setting, and Participants: Multicenter, cluster-randomized, investigator-masked, crossover, noninferiority trial; 4403 patients undergoing cardiac or abdominal surgery in 3 tertiary care hospitals in Switzerland between September 2018 and March 2020 were assessed and 3360 patients were enrolled (cardiac, n = 2187 [65%]; abdominal, n = 1173 [35%]). The last follow-up was on July 1, 2020. Interventions: Over 18 consecutive months, study sites were randomly assigned each month to either use povidone iodine or chlorhexidine gluconate, each formulated in alcohol. Disinfectants and skin application processes were standardized and followed published protocols. Main Outcomes and Measures: Primary outcome was SSI within 30 days after abdominal surgery and within 1 year after cardiac surgery, using definitions from the US Centers for Disease Control and Prevention's National Healthcare Safety Network. A noninferiority margin of 2.5% was used. Secondary outcomes included SSIs stratified by depth of infection and type of surgery. Results: A total of 1598 patients (26 cluster periods) were randomly assigned to receive povidone iodine vs 1762 patients (26 cluster periods) to chlorhexidine gluconate. Mean (SD) age of patients was 65.0 years (39.0-79.0) in the povidone iodine group and 65.0 years (41.0-78.0) in the chlorhexidine gluconate group. Patients were 32.7% and 33.9% female in the povidone iodine and chlorhexidine gluconate groups, respectively. SSIs were identified in 80 patients (5.1%) in the povidone iodine group vs 97 (5.5%) in the chlorhexidine gluconate group, a difference of 0.4% (95% CI, -1.1% to 2.0%) with the lower limit of the CI not exceeding the predefined noninferiority margin of -2.5%; results were similar when corrected for clustering. The unadjusted relative risk for povidone iodine vs chlorhexidine gluconate was 0.92 (95% CI, 0.69-1.23). Nonsignificant differences were observed following stratification by type of surgical procedure. In cardiac surgery, SSIs were present in 4.2% of patients with povidone iodine vs 3.3% with chlorhexidine gluconate (relative risk, 1.26 [95% CI, 0.82-1.94]); in abdominal surgery, SSIs were present in 6.8% with povidone iodine vs 9.9% with chlorhexidine gluconate (relative risk, 0.69 [95% CI, 0.46-1.02]). Conclusions and Relevance: Povidone iodine in alcohol as preoperative skin antisepsis was noninferior to chlorhexidine gluconate in alcohol in preventing SSIs after cardiac or abdominal surgery. Trial Registration: ClinicalTrials.gov Identifier: NCT03685604.
ABSTRACT
BACKGROUND: Preoperative skin antisepsis is an essential component of safe surgery. However, it is unclear how many antiseptic paints are needed to eliminate bacteria prior to incision. This study compared microbial skin counts after two and three antiseptic paints. METHODS: We conducted a prospective cohort study in non-emergency patients receiving a cardiac/abdominal surgery with standardized, preoperative skin antisepsis consisting of an alcoholic compound and either povidone iodine (PI) or chlorhexidine (CHX). We obtained three skin swabs from the participant's thorax/abdomen using a sterile template with a 25 cm2 window: After collection of the first swab prior to skin antisepsis, and once the second and third application of PI/CHX had dried out, we obtained a second and third swab, respectively. Our primary outcome was the reduction in microbial skin counts after two and three paints of PI/CHX. RESULTS: Among the 239 enrolled patients, there was no significant difference in the reduction of mean square root-transformed microbial skin counts with three versus two paints (P = 0.2). But distributions of colony forming units (CFUs) decreased from paint 2 to 3 in a predefined analysis (P = 0.002). There was strong evidence of an increased proportion of patients with zero CFU after paint 3 versus paint 2 (P = 0.003). We did not identify risk factors for insufficient reduction of microbial skin counts after two paints, defined as the detection of > 5 CFUs and/or ≥ 1 pathogens. CONCLUSIONS: In non-emergency surgical patients, three antiseptic paints may be superior to two paints in reducing microbial skin colonization prior to surgery.
