William G. Ward Sr，Robert M. Corey

There has been heightened interest in the prevention and minimization of perioperative infections. These efforts have included, but are not limited to, the NSQIP (National Surgical Quality Improvement Program) and SCIP (Surgical Care Improvement Program). Notwithstanding this heightened interest, the substantial morbidity and increased costs of perioperative infections are best avoided.

A few recent clinical studies have indicated decreased infection rates in surgical procedures in which the surgeons employed intraoperative irrigation with an antiseptic solution. Brown et al. reported a reduction in the ninety-day infection rate in hip and knee arthroplasties from 0.97% (in wounds irrigated with 1 L of normal saline solution prior to closure) to 0.15% with a three-minute irrigation with dilute (0.35%) Betadine and a 10% povidone-iodine skin application prior to final closure. Cheng et al. performed a prospective randomized spine surgery trial and reported no infections in 208 patients following irrigation with dilute (3.5%) Betadine compared with one superficial and six deep infections in 206 patients treated without irrigation (p < 0.01). However, there have been few such studies, and the antiseptic agents and their concentration have varied among the studies, presumably on the basis of which antiseptic solutions were readily available and on empirical practices. To our knowledge, there have been no in vivo studies comparing the efficacy or toxicity of one solution with that of another in a surgical wound model, and there have been few studies evaluating the effect of the concentration of these solutions on their efficacy and/or tissue toxicity during surgery. Howells et al. did evaluate the effect of irrigation solutions on the strength of the cement-bone interface in a cadaveric animal model and reported inferior cement fixation associated with irrigation with normal saline solution or povidone-iodine compared with hydrogen peroxide. They theorized that the effervescent effect of the hydrogen peroxide better cleansed the bone, allowing for improved cement penetration and interdigitation, thereby improving the strength of the cement-bone interface. Several studies have indicated potential toxic effects of antiseptic irrigation solutions, especially where articular cartilage remains. Von Keudell et al. reported chrondrotoxicity of povidone-iodine even at 0.35% strength if applied for longer than one minute to bovine cartilage explants. Hirsch et al. found that povidone-iodine exhibited lower antibacterial efficacy and higher toxicity to human skin cells compared with a number of other antiseptic solutions in a cell culture environment. Therefore, it is natural that questions have arisen regarding which antiseptic agent (if any) is the best for irrigation of an open wound and what concentration it should be used in. These are the core issues that van Meurs et al. have addressed.

Their basic-science laboratory study attempts to provide some scientific basis to guide the clinical selection of an antiseptic solution and its concentration to be used for irrigation of a surgical wound. In the model systems utilized, povidone-iodine antiseptic solution emerged as the best of the five agents that were tested. It possessed the best balance of providing bactericidal capability while minimizing cytotoxicity at clinically relevant concentrations. As explained in the article, the other agents possessed unacceptable cytotoxicity at the concentrations required to achieve sufficient bactericidal effect. Unfortunately, even povidone-iodine exhibited some cytotoxicity at the more effective bactericidal concentrations. It should be noted, however, that this was a study based on cytotoxicity in a cell culture environment, which may not precisely reflect cellular toxicity, tissue toxicity, or wound-healing interference in an in vivo environment. The toxicity tolerance of cells in a monolayer on culture medium may well be substantially different from the toxicity tolerance of exposed tissue in an in vivo surgical environment that is well vascularized. Similarly, the bactericidal effect of antiseptic irrigation in an in vivo environment, in which blood and other tissue fluids are present, is unlikely to precisely reflect the bactericidal effect of such a solution on bacteria suspended in the solution. These concerns were acknowledged by the authors in their discussion. Because of such differences involving the local environments and the actual tissue configurations tested, the validity of their conclusions must be considered in light of the experimental environment. Be that as it may, this study does contribute to the body of knowledge and provides scientific data supporting recent clinical studies indicating that wound irrigation with povidone-iodine solution decreases the risk of subsequent wound infections.

On the basis of the results of this study and the limited existing current body of information, wound irrigation with dilute povidone-iodine solution appears reasonable as a potential method to minimize the risk of perioperative infection in cases without retained articular cartilage. Before endorsing the use of any of these agents at any concentration for any specified period of time as a new standard of care for any particular type of procedure, however, we need further clinical studies; these should include animal studies investigating toxicity to various tissues as well as studies documenting the in vivo bactericidal efficacy of these various agents at various concentrations and under specific surgical conditions of use. Long-term studies are required to confirm the absence of any unsuspected or currently undetected deleterious effects of such a practice. To date, we are unaware of any clinically relevant complications of dilute povidone-iodine wound irrigation. If the simple, low-cost practice of dilute povidone-iodine wound irrigation is confirmed to decrease infection rates appreciably, and if substantial complications are not observed, then such a practice seems prudent to us, given the severe morbidity associated with perioperative infections. Until such additional data are available, the study by van Meurs et al. does contain additional basic-science data that provide a scientific rationale to guide clinicians in their practice of antiseptic wound irrigation. We intend to continue to utilize dilute povidone-iodine wound irrigation in arthroplasty procedures without retained cartilage, pending further investigations of this and other infection prevention practices.