LETTERS TO THE EDITOR 1095 3. School Board of Nassau County v Arline, 480 US 273 (1987). 4. Centers for Disease Control and Prevention (CDC). Recommendations for preventing transmission of human immunodeficiency virus and hep- atitis B virus to patients during exposure-prone invasive procedures. MMWR Morb Mortal Wkly Rep 1991;40(RR-8):l-9. 5. Turkel S, Henderson DK. Strategies for managing providers infected with bloodborne pathogens. In: Programs and abstracts of the Fifth Decennial International Conference on Healthcare-Associated Infections. Arlington, VA: Society for Healthcare Epidemiology of America, 2010. Abstract 684. 6. UK Department of Health, UK Expert Advisory Group on AIDS. AIDS/ HIV-infected health care workers: guidance on the management of infected health care workers. Heywood Lancashire, England: UK Department of Health, 1994. 7. Reitsma AM, Closen ML, Cunningham M, et al. Infected physicians and invasive procedures: safe practice management. Clin Infect Dis2005;40(ll): 1665-1672. 8. Tuboku-Metzger J, Chiarello L, Sinkowitz-Cochran RL, Casano-Dickerson A, Cardo D. Public attitudes and opinions toward physicians and dentists infected with bloodborne viruses: results of a national survey. Am J Infect Control 2005;33(5):299-303. 9. Sadoh AE, Sadoh WE, Fawole AO, Oladimeji A, Sotiloye O. Attitude of health care workers to patients and colleagues infected with human im- munodeficiency virus. SAHARA J 2009;6(l):17-23. Chlorhexidine Is Not the Main Active Ingredient in Skin Antiseptics That Reduce Blood Culture Contamination Rates To the Editor—We read with great interest the article by Mar- lowe et al 1 describing a significant reduction in rates of pe- diatric blood culture contamination after implementation of skin antisepsis with 70% isopropanol plus 3.15% chlorhex- idine, compared with 10% aqueous povidone-iodine. We would like to congratulate the authors on the achievement of this landmark study with a very large number of partic- ipants and blood cultures investigated. However, we would like to raise the point that interpre- tation of the results by the authors contains a likely error in the attribution of the positive effects observed. In the title and abstract and throughout the text, the authors describe this as a study of chlorhexidine versus povidone-iodine and— by implication—attribute the positive effects to the chlor- hexidine component. However, the solutions used were 70% isopropanol with 3.15% chlorhexidine and 10% aqueous pov- idone-iodine. The first solution has 2 active ingredients, and the second has only 1. The differential antimicrobial activity of these compounds has been a topic of intense research and evaluation since the 1970s and is well described in standard textbooks on antisepsis and infection control. 2 " 4 If chlorhex- idine or povidone-iodine in aqueous solutions is compared with various alcohols, it turns out that the immediate anti- microbial activity (ie, the capacity to act as a disinfectant) of standard alcohol compounds is significantly greater than that of the other agents if they are used in an aqueous solution. The difference between alcoholic and aqueous hand and skin antiseptics is typically approximately 1 log, or a 10-fold dif- ference in favor of the alcohols. Furthermore, overview tables on the differential activity of skin antiseptics have been pub- lished in the Centers for Disease Control guidelines on the prevention of surgical site infections 5 and in the World Health Organization guidelines on hand hygiene in health care. 6 Ac- cording to this information, alcohols are the most rapid acting skin antiseptics, whereas both chlorhexidine and povidone- iodine only have intermediate speed of action. However, al- cohols lack any residual activity, whereas chlorhexidine ap- pears to exert such an effect. This is also well illustrated by a comparative experiment on surgical hand antisepsis as pub- lished in a textbook chapter.4 The results and associated figure indicate that 70% isopropanol generates an immediate re- duction of resident hand flora of approximately 2.5 log, aque- ous 10% povidone-iodine generates a reduction of approx- imately 1.8 log, and 4% aqueous chlorhexidine generates a reduction of approximately 0.8 log. The immediate microbial reduction achieved by both aqueous agents is significantly less than that of 70% isopropanol, and that caused by a 70% isopropanol plus 0.5% chlorhexidine mix is almost the same as of pure isopropanol. However, when a second time point of 180 minutes under surgical gloves is examined, it becomes apparent that there is bacterial regrowth after use of pure isopropanol, whereas continued microbial suppression occurs with use of the mix. These results indicate that there is almost no contribution from chlorhexidine to the immediate kill caused by isopropanol, and the benefit from such a mix is that of immediate plus sustained action. Similar results con- cerning immediate and sustained action of alcohols alone versus alcohol plus chlorhexidine were also obtained in a more recent study of antisepsis at various other skin sites. 7 Mixtures of alcohol and chlorhexidine have benefits if both immediate and sustained action are required; examples in- clude use for surgical skin antisepsis, with sustained action under surgical drapes; for surgical hand antisepsis under the gloves; and at the sites of vascular catheter insertion. 4 ' 8,9 The biological principles underlying blood culture collection are different. Culture samples are obtained immediately after an- tisepsis and after observing the appropriate contact time of the antiseptic on skin. As opposed to the other applications mentioned above, there is no requirement for sustained ac- tion at the site of blood culture collection. Arguably, the positive effect observed in this landmark study by Marlowe et all is more likely to be derived from the action of the isopropanol component than from the chlorhexidine com- ponent. In conclusion, alcohol is likely the key component in the immediate disinfection process, and the impact of chlorhexidine would require further investigation. ACKNOWLEDGMENTS Potential conflicts of interest. All authors report no conflicts of interest rel- evant to this article.