Pages, J., Hazera, P., Mégarbane, B., du Cheyron, D., Thuong, M., Dutheil, J.J., . . . Daubin, C. (2016). Comparison of alcoholic chlorhexidine and povidone–iodine cutaneous antiseptics for the prevention of central venous catheter-related infection: A cohort and quasi-experimental multicenter study. Intensive Care Medicine, 42, 1418–1426. 

To compare the effectiveness of different skin antiseptics in the prevention of catheter-related infection (CRI)

The type of antiseptic used for skin disinfection for catheter care was chosen by the intensive care units of participating hospitals. Maximal sterile precautions for catheter insertion according to guidelines were used. All catheters were nontunneled, and none were used for routine blood sampling. Decisions to remove catheters were at the physicians' discretion and, after removal, catheter tip and peripheral blood cultures were conducted. A one-step procedure of skin cleansing was conducted with 2% chlorhexidine, and a four-step protocol of scrub, rinse, dry, and disinfect was used with other antiseptics. A propensity score was calculated from analysis of covariance to determine the propensity for CRI and was controlled in analysis techniques. In four ICUs, staff switched from a povidone iodine antisepsis to chlorhexidine, and separate analysis of differences in outcomes were analyzed individually (1,368 patients). All suspected cases of CRI were reviewed and determined by a blinded committee.

• N = 3,027
• MEAN AGE = 62.3 years
• MALES: 50.3%, FEMALES: 49.7%
• CURRENT TREATMENT: Not applicable
• KEY DISEASE CHARACTERISTICS: Patient with multiple diseases, including cancer. All were in intensive care units. Over half were on antibiotics, 78% were on ventilators, 10%–14% had an immunodeficiency, and, overall, less than 1%  had neutropenia.

• SITE: Multi-site
• SETTING TYPE: Inpatient
• LOCATION: France

• Prospective cohort comparison

CRI defined as catheter-related bloodstream infection (CRBSI) or the combination of a catheter tip colonization and clinical signs of sepsis with no other cause identified.

The use of chlorhexidine was associated with a decreased risk of CRI (2 per 1,000 catheter days, p = 0.037). The unadjusted incidence of CRI was higher in the povidone iodine group compared to the chlorhexidine group (2.8 versus 2 per 1,000 catheter days, p = 0.001). Overall, CRI risk in the units that switched from povidone iodine to chlorhexidine was lower with chlorhexidine use (hazard ratio [HR] = 0.31, p = 0.005). However, no significant differences in CRBSI existed between groups.

The use of a skin antisepsis with a 2% chlorhexidine alcohol preparation for catheter care may be associated with a lower incidence of CRI.

• Risk of bias (no random assignment) 
• Key sample group differences that could influence results
• No data or control regarding other aspects of catheter care and use existed, which can also affect the development of CRI.
• No information is provided regarding the frequency of dressing changes, etc.
• The various solutions used for skin antisepsis were unclear—both a 2% and 1% chlorhexidine and 5% and 10% povidone iodine solutions were mentioned, but the study does not indicate which patients got which solution, or which solution was associated with which reported result.
• Differentiation of CRI versus CRBSI were unclear, and the association of these with the skin antisepsis used differed.

The use of chlorhexidine skin antisepsis may be associated with a lower incidence of CRI. The evidence has several limitations; however, it is consistent with the body of evidence showing the efficacy of chlorhexidine skin preparation as part of central venous catheter care.