Effectiveness of Alcohol-Based Hand Hygiene Gels in Reducing Nosocomial Infection Rates

Effectiveness of Alcohol-Based Hand Hygiene Gels in Reducing Nosocomial Infection RatesAuthor(s): By Andreas F. Widmer,, MD, MS; Manfred Rotter,, MDSource: Infection Control and Hospital Epidemiology, Vol. 29, No. 6 (June 2008), p. 576Published by: The University of Chicago Press on behalf of The Society for Healthcare Epidemiologyof AmericaStable URL: http://www.jstor.org/stable/10.1086/587808 .

Accessed: 16/05/2014 19:42

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

The University of Chicago Press and The Society for Healthcare Epidemiology of America are collaboratingwith JSTOR to digitize, preserve and extend access to Infection Control and Hospital Epidemiology.

http://www.jstor.org

This content downloaded from 193.105.154.119 on Fri, 16 May 2014 19:42:29 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=ucpress

http://www.jstor.org/action/showPublisher?publisherCode=shea

http://www.jstor.org/action/showPublisher?publisherCode=shea

http://www.jstor.org/stable/10.1086/587808?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


l e t t e r s t o t h e e d i t o r

infection control and hospital epidemiology june 2008, vol. 29, no. 6

Effectiveness of Alcohol-Based HandHygiene Gels in Reducing Nosocomial

From the Division of Infectious Diseases & Hospital Epidemiology, Univer-sity of Basel Hospitals, Basel, Switzerland (A.F.W.); and the Institute of Hy-giene and Medical Microbiology, Medical University Vienna, Vienna, Austria(M.R.).
Infection Rates
To the Editor—We have read with great interest the articleby Rupp et al.1 The results challenge current infection con-trol policies.2 The study has been carefully conducted andthe results appear to be valid. Surprisingly, the improvedcompliance with hand hygiene recommendations associ-ated with the use of alcohol-based hand hygiene productsdid not result in lower nosocomial infection rates. As theauthors pointed out, the results may be interpreted in sev-eral ways. We want to add points to be discussed that may beassociated with or even responsible for these negativeresults.

First, the level of antimicrobial efficacy of 62% ethanolmay not suffice to interrupt transmission of nosocomialpathogens. In fact, the product does not meet the require-ment of European standard EN 1500,3 which is needed toclear the product for the European market.4 Incidentally,the exact concentration of the ethanol is not described byRupp et al1: it may be 62% by volume (equivalent to 49 g/dL)or 62% by weight. Second, we have shown that training inthe application of alcohol-based hand hygiene products iscrucial to optimize antimicrobial killing.5,6 No formal train-ing is described by Rupp et al1; an absence of training mayhave reduced the effect of the gel. Third, coagulase-negativestaphylococci were cultured mainly from the hands ofhealthcare workers. However, data regarding detection ofclinically important pathogens such as methicillin-resistantStaphylococcus aureus are not given. Finally, a formal samplesize calculation was not mentioned, and lack of this calcu-lation potentially limits the impact of the negative results ofthe trial. The low baseline rate of nosocomial infection mayhave jeopardized the possibility of detecting the clinical ef-fect of the introduction of the gel. In addition, under con-ditions of high patient occupancy or understaffing, handhygiene alone is unlikely to prevent nosocomial infection.7

We congratulate the authors for conducting this importanttrial. It may be the first hint that the antimicrobial activity ofsuch gels is not sufficient to reduce the incidence of nosocomialinfections.

acknowledgmentsPotential conflicts of interest. All authors report no conflicts of interest rele-vant to this article.

Andreas F. Widmer, MD, MS; Manfred Rotter, MD

This content downloaded from 193.105.154.All use subject to JSTOR T

Address reprint requests to Andreas F. Widmer, MD, MS, University ofBasel, Basel, BS 4031 Switzerland ([email protected]).

Infect Control Hosp Epidemiol 2008; 29:576© 2008 by The Society for Healthcare Epidemiology of America. All rightsreserved. 0899-823X/2008/2906-0019$15.00. DOI: 10.1086/587808

references

1. Rupp ME, Fitzgerald T, Puumala S, et al. Prospective, controlled, cross-over trial of alcohol-based hand gel in critical care units. Infect Control HospEpidemiol 2008;29:8 –15.

2. Widmer AF. Replace hand washing with use of a waterless alcohol handrub? Clin Infect Dis 2000;31:136 –143.

3. Rotter ML. European norms in hand hygiene. J Hosp Infect 2004;56(sup-pl2):S6 –S9.

4. Kramer A, Rudolph P, Kampf G, Pittet D. Limited efficacy of alcohol-basedhand gels. Lancet 2002;359:1489 –1490.

5. Widmer AF, Conzelmann M, Tomic M, Frei R, Stranden AM. Introducingalcohol-based hand rub for hand hygiene: the critical need for training.Infect Control Hosp Epidemiol 2007;28:50 –54.

6. Widmer AE, Dangel M. Alcohol-based handrub: evaluation of techniqueand microbiological efficacy with international infection control profes-sionals. Infect Control Hosp Epidemiol 2004;25:207–209.

7. Beggs CB, Noakes CJ, Shepherd SJ, Kerr KG, Sleigh PA, Banfield K. Theinfluence of nurse cohorting on hand hygiene effectiveness. Am J InfectControl. 2006;34:621– 626.

“Cannot Detect a Change” Is Not the Sameas “There Is Not a Change”

To the Editor—The recent article by Rupp et al.1 has generateda great deal of media attention. Unfortunately, the value of thearticle in terms of increasing hand hygiene compliance, use ofalcohol-based hand gel, and useful adherence data from obser-vations was lost. However, we believe the following remarksprovide evidence that the conclusion of “no detectable change”in nosocomial infection rates may not be supported by theiranalysis.

The authors had a null hypothesis of “no change” and analternative of “change.” They have concluded that theirfindings support the null hypothesis. What this means isthat there is not enough evidence to overturn the null hy-pothesis— but that is not the same as saying that the nullhypothesis is true. If you collect very few data or assemble avery uninformative data set, then it is unlikely that you willhave enough evidence to overturn the null hypothesis– evenif it should be overturned.

There are so few infections over the time period in the study1

that the data sets are likely to be uninformative with respect tothe question of infection rates. These units had 12 beds, and the

119 on Fri, 16 May 2014 19:42:29 PMerms and Conditions


patients in each unit were observed for 1 year in each branch ofthe crossover study. Twelve beds multiplied by 365 days is4,380 bed-days; so they had 4,380 bed-days as a maximum (we

Trial of Alcohol-Based Hand Gel in CriticalCare Units

letters to the editor 577

do not know if the units were consistently fully occupied).Their reported infection rates are approximately 1 to 4 infec-tions per 1,000 bed-days. This means that they observed ap-proximately 4 to 16 infections over the entire year of the inter-vention for each arm of the study. This range represents a verysmall number of infections, and without getting into the detailsof the underlying Poisson regression model, the inherent vari-ability on these numbers will be relatively high.

So, what does this mean? It means that the data are verynoisy, and the study is unlikely to be able to demonstrate aneffect even if it is there. Rough calculations suggest that even ifimprovements in hand hygiene adherence were able to de-crease infection rates by 50%, then this study would have onlyroughly a 20% chance of demonstrating the effect.

Another issue important to this data set is whether the in-fections themselves are independent or whether they occurredin clusters (clumped in time). If they were clustered (whichwould mean that they were not statistically independent), thenthis analysis would be weakened even more, because ignoringthe clustering would give a false sense of the amount of infor-mation contained in the data. If the infections are clustered intime, then the analysis is inappropriate.

To the authors’ credit, they do acknowledge that the study is“underpowered to detect small differences in rates of infection,”1

but it may be underpowered to demonstrate larger differences,too. The reviewers of this article should have noted to the authorsthat this is a good article but requested that they leave out thecomments on “detectable changes in the incidence of healthcare-associated infection,” because there does not appear to be enoughinformation to generate a reliable conclusion.

acknowledgmentsPotential conflicts of interest. Both authors report no potential conflicts ofinterest relevant to this study.

Maryanne McGuckin, ScEdD; Richard Waterman, PhD

From MMI and Health Policy Jefferson Medical College (M.M.), and Ana-Bus and the Wharton School, University of Pennsylvania (R.W.), Philadelphia,Pennsylvania.

Address reprint requests to Maryanne McGuckin, ScEdD, 115 E. AthensAve., Ardmore, PA 19003 ([email protected]).

Infect Control Hosp Epidemiol 2008; 29:576 –577© 2008 by The Society for Healthcare Epidemiology of America. All rightsreserved. 0899-823X/2008/2906-0020$15.00. DOI: 10.1086/587087

reference

1. Rupp M, Fitzgerald T, Puumala S, et al. Prospective, controlled, cross-overtrial of alcohol-based hand gel in critical care units. Infect Control HospEpidemiol 2008;29:8 –15.


To the Editor—We congratulate Rupp et al.1 for their well-designed and well-performed study. However, we have some con-cerns. In contrast to the findings of several studies,2-7 this study didnot find an association between increased hand hygiene adher-ence and a reduction in nosocomial infections in intensive careunits. The authors reported the incidence of 3 types of medicaldevice–related infections (central venous catheter–related bacter-emia, urinary catheter–associated urinary tract infection, andventilator-associated pneumonia) and 3 types of infections asso-ciated with multidrug-resistant pathogens (methicillin-resistantStaphylococcus aureus [MRSA], vancomycin-resistant enterococci[VRE], and Clostridium difficile). Our major concern is that activesurveillance cultures were not performed to identify patients col-onized with MRSA or VRE. This is an important shortcoming,because the rate of importation of MRSA or VRE into intensivecare units and the proportion of ICU patients colonized with suchorganisms (“colonization pressure”) are factors shown to affectthe rate of transmission and, most likely, the incidence of infec-tion.8 Because no surveillance cultures were performed, thepresent study was not able to assess the impact of hand hygiene onnosocomial transmission of these organisms.

The study was statistically underpowered to show a differ-ence in the measured outcomes, and in fact no formal poweranalysis was conducted. With detection of such low rates ofnosocomial infections, the findings can be explained by chancevariability, regression to the mean, and, because nosocomialinfections tend to cluster, overdispersion of infection rates rel-ative to chance variation. The authors might consider poolingthe data on the incidence of infections due to MRSA, VRE, C.difficile, and Pseudomonas aeruginosa to ascertain if there was adifference in the total number of infections caused by thesepathogens during the periods under study, but it is unclear ifthis would overcome the above-mentioned problems. In addi-tion, the study compares infection rates aggregated by timeperiod. As stated in gold standard guidelines for the reportingof intervention studies of nosocomial infections,9 measure-ment at regular intervals (weekly or monthly) would have bet-ter demonstrated trends.

Two of the device-related infections selected as out-comes—namely, catheter-related bloodstream infectionsand ventilator-associated pneumonia— often necessitate acombination (“bundle”) of preventive measures, not justhand hygiene alone, to achieve substantial reductions in in-cidence. However, the authors did not mention if bundleswere used during any of the study periods and, if they were,the degree of compliance with the bundles or other inter-ventions that may have confounded the results.

The authors noted that their inability to demonstrate anassociation between hand hygiene adherence levels and rates ofnosocomial infections may have been due to a failure to



achieve a sufficiently high adherence level. The highest level ofhand hygiene adherence in their study was 69%. A study of therelation between MRSA prevalence and hand hygiene compli-

the World Health Organization World Alliance for PatientSafety has designated hand hygiene promotion as the corner-stone of the First Global Patient Safety Challenge, which is

578 infection control and hospital epidemiology june 2008, vol. 29, no. 6

ance conducted in a rehabilitation hospital found that wardswith compliance greater than 70% had a lower prevalence ofMRSA than wards with less compliance.5 Additionally, there isongoing debate about the efficacy of alcohol-based hand rubformulations with an ethyl alcohol content lower than 80%, inparticular with gels and foam formulations.7 The liquid formu-lations have achieved greater log reductions in the concentra-tion of pathogens in in vivo laboratory-based studies of handantisepsis. To our knowledge, however, to date no randomizedclinical trials or epidemiologic data have demonstrated that theliquid formulations reduce transmission of pathogens to agreater degree than gel formulations.

Moreover, we would like to emphasize that what the au-thors have clearly designed and conducted, and what theirresults support, is a successful multimodal hand hygienepromotion campaign modeled on various experiences.2,3,6,7

Their intervention included most key components of such astrategy7: education of healthcare staff by using varioustools, face-to-face meetings with nursing staff, reminders inthe workplace, monitoring of compliance, monitoring ofthe rate of nosocomial infections, surveillance feedback,and the introduction of alcohol-based hand rub at the pointof care in 2 units at different time periods. The last, referredto as “system” change,7 is the prerequisite for successfulhand hygiene promotion (as clearly apparent from the studyresults1), but it is not sufficient in itself when introduced asa unique component of promotion.7 In this regard, we be-lieve that the abstract is somewhat misleading, as the au-thors and hospital healthcare staff efforts are insufficientlyrecognized.

Perhaps because of restrictions in the length of the article,the results were not compared with those of studies thatshowed a reduction in the transmission of nosocomial in-fections following promotion of alcohol-based handhygiene.2-7 In an intervention conducted in a neonatal unit,investigators monitored hand hygiene compliance, alcohol-based hand rub consumption, and nosocomial infections atthe individual patient level.8 Improved compliance was in-dependently associated with a decreased risk of nosocomialinfections and reduced cross-transmission of genotypicallyrelated bloodstream pathogens. Other researchers usingquasi-experimental designs reported reduced MRSA infec-tion acquisition following implementation of hand hygienecampaigns that included promotion of alcohol-based handhygiene.3,5,7

We believe that it is important to evaluate further the impactof hand hygiene and other infection control interventions onthe incidence of nosocomial infections. We are surprised bythe tremendous attention that this article has drawn in the laypress with the take-home message that hand hygiene has noimpact on nosocomial infections, a message that we considerharmful to the international patient safety movement. Of note,


dedicated to tackling nosocomial infections as a worldwidepriority.7,10 To contribute to this field of endeavor in ameaningful way, future investigations must be carried outwith appropriate statistical power and scientific rigor. Inap-propriate interpretation of the study results by nonscientificexperts is a disservice to the authors and the healthcare com-munity, as it is of seminal importance for such a study to befully understood.

acknowledgmentsThe authors thank all members of the Infection Control Program at Universityof Geneva Hospitals and members of the World Health Organization GlobalPatient Safety Challenge “Clean Care is Safer Care” core group: John Boyce,Barry Cookson, Nizam Damani, Don Goldmann, Lindsay Grayson, ElaineLarson, Geeta Mehta, Ziad Memish, Hervé Richet, Manfred Rotter, Syed Sat-tar, Hugo Sax, Wing Ho Seto, Andreas Voss, and Andreas Widmer.

Potential conflicts of interest. J.M.B. has consultantships with Gojo Indus-tries, Advanced Sterilization Products, Clorox Corporation, Soap and Deter-gent Association, and 3M Corporation. All other authors report no potentialconflicts of interest relevant to this study.

Leonard A. Mermel, DO ScM; John M. Boyce, MD;Andreas Voss, MD, PhD; Benedetta Allegranzi, MD;

Didier Pittet, MD, MS

From the Warren Alpert Medical School of Brown University and RhodeIsland Hospital, Providence, Rhode Island (L.A.M.); the Hospital of St.Raphael, New Haven, Connecticut (J.M.B.); the Department of Medical Mi-crobiology, Radboud University Medical Centre, Nijmegen, the Netherlands(A.V.); the World Health Organization (WHO) World Alliance for PatientSafety First Global Patient Safety Challenge, WHO Headquarters (J.M.B., A.V.,B.A., D.P.), and the Infection Control Program, University of Geneva Hospi-tals and Faculty of Medicine (D.P.), Geneva, Switzerland.

Address reprint requests to Didier Pittet, MD, MS, Director, Infection Con-trol Program, University of Geneva Hospitals and Faculty of Medicine, 24 RueMicheli-du-Crest, 1211 Geneva 14 / Switzerland ([email protected]).

Infect Control Hosp Epidemiol 2008; 29:577–579© 2008 by The Society for Healthcare Epidemiology of America. All rightsreserved. 0899-823X/2008/2906-0021$15.00. DOI: 10.1086/587812

references

1. Rupp ME, Fitzgerald T, Puumala S, et al. Prospective, controlled, cross-over trial of alcohol-based hand gel in critical care units. Infect ControlHosp Epidemiol 2008;29:8 –15.

2. Pittet D, Hugonnet S, Harbarth S, et al. Effectiveness of a hospital-wideprogramme to improve compliance with hand hygiene. Lancet 2000;356:1307–1312.

3. Johnson PD, Martin R, Burrell LJ, et al. Efficacy of an alcohol/chlorhexi-dine hand hygiene program in a hospital with high rates of nosocomialmethicillin-resistant Staphylococcus aureus (MRSA) infection. Med J Aust2005;183:509 –514.

4. Gordin FM, Schultz ME, Huber RA, Gill JA. Reduction in nosocomialtransmission of drug-resistant bacteria after introduction of an alcohol-based handrub. Infect Control Hosp Epidemiol 2005;26:650 – 653.

5. Girou E, Lagrand P, Soing-Altrach S, et al. Association between hand hy-giene compliance and methicillin-resistant Staphylococcus aureus preva-



lence in a French rehabilitation hospital. Infect Control Hosp Epidemiol2006;27:1128 –1130.

6. Pessoa-Silva CL, Hugonnet S, Pfister R, et al. Reduction of health careassociated infection risk in neonates by successful hand hygiene promo-

(about 10-fold; ie, 1 log less) than do liquid alcohol hand rubs.5

This has 2 implications: the antimicrobial activity is very low tostart with, and it is further compromised by the gel formula-


tion. Pediatrics 2007;120:e382– e390. 10.1542/peds.2006 –3712. PublishedJuly 30, 2007. Accessed May 5, 2008.

7. World Health Organization. WHO guidelines on hand hygiene inhealth-care (advanced draft). Geneva: World Health Organization,2006. Available at: http://www.who.int/patientsafety/information_centre/Last_April_versionHH_Guidelines%5b3%5d.pdf. Accessed February 8,2008.

8. Merrer J, Santoli F, Appéré de Vecchi C, Tran B, De Jonghe B, Outin H.“Colonization pressure” and risk of acquisition of methicillin-resistantStaphylococcus aureus in a medical intensive care unit. Infect Control HospEpidemiol 2000;21:718 –723.

9. Stone SP, Cooper BS, Kibbler CC, et al. The ORION statement: guidelinesfor transparent reporting of outbreak reports and intervention studies ofnosocomial infection. Lancet Infect Dis 2007;7:282–288.

10. Pittet D, Donaldson L. Clean Care is Safer Care: a worldwide priority.Lancet 2005;366:1246 –1247.

Alcohol-Based Hand Hygiene andNosocomial Infection Rates

To the Editor—I read with great interest the study by Rupp etal., describing a crossover trial of alcohol hand gel use in criticalcare units.1 I was surprised that the significant increase in com-pliance observed in this study did not appear to be associatedwith a decrease in nosocomial infection rates. I am concernedthat superficial readers may conclude that alcohol-based handhygiene does not provide a benefit in the healthcare environ-ment compared with hand washing.

The causal role of microorganisms on hands in the patho-genesis of nosocomial infections is extremely well established.2

However, the interplay between various factors involved inclinical practice (eg, availability of appropriate hand hygieneagents, correctness of their use, compliance with hand hygienerecommendations) and the outcome in terms of nosocomialinfection rates is highly complex and multifactorial. Apartfrom the question of whether the study by Rupp et al.1 hadpatient numbers sufficient in size and observation periods suf-ficient in length to demonstrate a difference, the authors ap-parently have not considered one factor that I think is impor-tant: the antimicrobial activity of a product used for handhygiene.

The hand gel chosen by the authors has an ethanol contentof only 62%. To determine the implications of this, it is neces-sary to look at some facts about alcohol-based hand hygiene.First, the published useful range of antimicrobial activity ofalcohols is about 60%– 80% for most microorganisms, withethanol the least potent, followed by isopropanol andn-propanol.2,3 The triclosan component (0.3%) of the gel usedin the study has very negligible immediate antimicrobial activ-ity.4 With an ethanol content of 62%, this gel is at the very lowend of the published range of activity. In addition, gel formu-lations often have considerably less antimicrobial activity


tion. The consequences are that there is no safety marginagainst handborne microbial contamination and that minoramounts of other liquids on the hands (eg, sweat, water) willrender the agent inactive by dilution. Such issues have beenaddressed by the European EN testing standards. Hand rubsthat pass EN 1500 typically produce a reduction in microbialcontamination of about 4 log (about 10,000-fold) on handswithin 30 seconds.3,5 Very few gels pass EN 1500, and the onesthat do typically contain 80% or more ethanol.6 The WorldHealth Organization’s standardized hand hygiene solutionscontain either 75% isopropanol or 80% ethanol, and each ofthese formulations pass EN 1500.2

Why is the antimicrobial activity of a hand hygiene agentimportant? First, it is beyond doubt that microorganisms onhands are responsible for nosocomial infections and that itis the killing or elimination of microorganisms on handsthat prevents these infections2; it is not the act of performinghand hygiene per se. Second, although the relationship isnot a formal mathematical one, there is a quantitative dose-response relationship between microorganisms eliminatedfrom hands and infections prevented.7 Third, there is noestablished “threshold” of microbial elimination beyondwhich hands can be considered “safe” from the risk of trans-mitting infections, such that lesser microbial reduction maybe considered equally good. Fourth, with regard to user ac-ceptability and compliance, it is important to bear in mindthat antimicrobial activity per se has no negative impact oneither; instead, user acceptability and compliance are influ-enced by overall hand rub composition and emollient addi-tives.8 As a consequence, it is necessary to choose hand hy-giene products that have both significant antimicrobialactivity and optimized composition for the users.

Finally, we can learn from history. It is now 160 years sinceSemmelweis made his seminal observations.9 He showedclearly that soap-based handwashing—which is now known tocause only a minimal reduction in the number of microbialpathogens on hands— did not have the same beneficial effectin preventing puerperal sepsis as did hand treatment with chlo-rinated lime, which is now known to kill microorganisms veryeffectively. In essence, this study by Rupp et al.1 appears tounderline the observation by Semmelweis that very potent an-timicrobial agents are most beneficial in reducing the inci-dence of nosocomial infections. Even high compliance withproducts that have limited activity may not sufficiently de-crease the rate of nosocomial infections.

acknowledgmentsPotential conflicts of interest. The author reports no conflicts of interest rele-vant to this study.

Matthias Maiwald, MD, PhD



From the Department of Microbiology and Infectious Diseases, FlindersUniversity and Flinders Medical Centre, Adelaide, Australia.

Address reprint requests to Matthias Maiwald, MD, PhD, Department

onstrate a statistically significant association between hand hy-giene and nosocomial infections. We also clearly noted thatactive surveillance cultures for methicillin-resistant Staphylo-


of Microbiology and Infectious Diseases, Flinders University and Flin-ders Medical Centre, Adelaide (Bedford Park) SA 5042, Australia([email protected]).


references1. Rupp ME, Fitzgerald T, Puumala S, et al. Prospective, controlled, cross-

over trial of alcohol-based hand gel in critical care units. Infect Control HospEpidemiol 2008;29:8 –15.

2. World Health Organization. WHO guidelines on hand hygiene in healthcare (advanced draft). Geneva: World Health Organization; 2006.

3. Rotter ML. Hand washing and hand disinfection. In: Mayhall CG, ed. Hos-pital Epidemiology and Infection Control. 3rd ed. Philadelphia. LippincottWilliams & Wilkins; 2004:1727–1746.

4. Ayliffe GA, Babb JR, Davies JG, Lilly HA. Hand disinfection: a comparisonof various agents in laboratory and ward studies. J Hosp Infect 1988;11:226 –243.

5. Kramer A, Rudolph P, Kampf G, Pittet D. Limited efficacy of alcohol-basedhand gels. Lancet 2002;359:1489 –1490.

6. Kampf G, Muscatiello M, Hantschel D, Rudolf M. Dermal tolerance andeffect on skin hydration of a new ethanol-based hand gel. J Hosp Infect2002;52:297–301.

7. Larson E. A causal link between handwashing and risk of infection? exam-ination of the evidence. Infect Control 1988;9:28 –36.

8. Kramer A, Bernig T, Kampf G. Clinical double-blind trial on the dermaltolerance and user acceptability of six alcohol-based hand disinfectants forhygienic hand disinfection. J Hosp Infect 2002;51:114 –120.

9. Harbarth S. Handwashing—the Semmelweis lesson misunderstood? ClinInfect Dis 2000;30:990 –991.

Not Perfect–Just Among the Best Available:Reply

To the Editor—We thank the authors of the letters for theirinterest in our study regarding hand hygiene1, and we sharetheir concern regarding the media attention and potentialmisinterpretation of the results.2-5 From an optimistic view-point, the widespread coverage of a study concerning handhygiene, which would have been unfathomable a few yearsago, points to the increasing recognition of the importanceof nosocomial infections and infection control. We hope theprofession can harness this new interest for the bettermentof the field. We regret that the value of the study may havebeen diminished by the widespread misrepresentation ofour conclusions. We have previously released notices6,7 torepudiate the perception that this study somehow “contra-dicts” Centers for Disease Control and Prevention or WorldHealth Organization recommendations.

Many of the specific points raised in the letters to the editorsimply reemphasize points we made in the article. We ac-knowledged that, despite the more than 2-year duration of ourstudy, the low infection rate rendered it underpowered to dem-


coccus aureus (MRSA) or vancomycin-resistant enterococci(VRE) were not performed, which may have precluded detec-tion of a statistically significant effect on the acquisition ofthese organisms. Similarly, we noted that the pathogenesis ofnosocomial infections is complex, and prevention requires amultifaceted or “bundle” approach. Indeed, we have longsupported a multifaceted approach to prevention of noso-comial infections. However, major changes to this approachdid not occur during the hand hygiene study, and monitor-ing of compliance with the bundles did not occur until morerecent years.

Without belaboring the details, we respond that most of thestudies cited by Mermel and colleagues,3 as well as numerousother reports purported to support the role of hand hygiene inthe prevention of nosocomial infections, are even more meth-odologically flawed than our own study. In general, these stud-ies were not controlled trials and often involved numerous in-terventions, including active surveillance cultures, isolationpractices, environmental disinfection, and patient decoloniza-tion. A cautionary note, tempering somewhat unrealistic ex-pectations of hand hygiene in the intensive care unit, has beenpreviously sounded.8-10

To more specifically address the questions raised, we wish torelate that when the various measures of nosocomial infectionsin our study were combined, a statistically significant associa-tion between hand hygiene compliance and infection was notdetected. In addition, clustering of infections was not ob-served. As we noted, coagulase-negative staphylococci were byfar the most common organisms recovered from the hands ofnurses. S. aureus was recovered only once (when gel was notavailable in the intensive care unit); gram-negative bacilli wererecovered from 7.2% of cultures of hand samples when gel wasavailable and from 11.7% of cultures when gel was not avail-able; yeasts were recovered from 2% of cultures when gel wasavailable and were not observed when gel was unavailable. Al-though not specifically mentioned in our article, the educa-tional program that preceded the introduction of the hand gelinto the critical care units explained when and how to use thehand gel.

Several of the letters2-4 noted the controversy regarding theefficacy of alcohol-based hand hygiene preparations in rela-tionship to alcohol content and formulation (gel or liquid).The hand gel used in our study contained a blend of 88% wt/wtethanol and 4.6% wt/wt isopropanol, and the total alcoholcontent was 68.5% vol/vol or 60.7% wt/wt (written communi-cation, M. Dolan, Gojo Industries, February 2008). The anti-microbial activity of alcohols is derived from their capacity todenature proteins, and they are most potent at concentrationsof 60%– 80%.11,12 At higher concentrations, they are less effec-tive because proteins are not denatured as readily in the



absence of water.11,12 In general, higher-chain alcohols aremore active than shorter-chain alcohols, and tertiary alcoholsare less effective than primary or secondary alcohols. Alcohols

actions speak louder than words, and we have recentlylaunched a hospital-wide campaign to improve adherence tohand hygiene recommendations. In the Discussion section of


have excellent in vitro and in vivo activity against vegetativebacteria, mycobacteria, a variety of fungi, and some envelopedviruses.11,12 The efficacy of hand hygiene preparations is influ-enced by a number of parameters including the type of alcohol,the concentration of alcohol, the volume used, the contact time,whether the hands are wet, and whether the hands are contami-nated with organic debris. A large number of studies have docu-mented the antimicrobial effect of alcohol-based hand rubs.11,12

Contrary to the claim of Widmer and Rotter,4 the hand gelused in our study does meet European standard EN 1500 re-quirements.13 Hand hygiene gel formulations and antiseptic ordisinfectant applications of alcohols are stringently regulatedin the United States by the Food and Drug Administration andthe Environmental Protection Agency, and there is some con-cern that the maximum alcohol content of alcohol-based handrubs in the United States is capped at 70% vol/vol by firecodes.14 As noted by Mermel et al.3 and by Maiwald,2 liquidformulations of alcohol-based hand rubs have been associatedwith improved in vitro performance compared with gels. Toour knowledge, however, to date no clinical data indicate thatthe liquid formulations are more effective. Recently, a prospec-tive trial in a critical care setting indicated considerably im-proved hand hygiene compliance when a gel formulation wasavailable, compared with compliance when a liquid formula-tion was in use.15 Any increase in potency of the liquid formu-lation compared with that of the gel might be mitigated by adecrease in hand hygiene compliance.

We appreciate the comments of the authors of the lettersacknowledging that our study was well designed and well per-formed, because data from other prospective, controlled trialsconcerning the efficacy of hand hygiene are very limited. Asnoted, some of the finer points of the study regarding the suc-cessful conduct of this multimodal performance improvementproject were lost on the media. In addition, the media largelyignored information about hand microbial ecology, drug re-sistance, preservation of hand hygiene compliance with in-creased workload, and the important effects on hand hygieneof nail length and the wearing of rings.

Unfortunately, in both the professional and the lay press wenote a trend toward oversimplification of the pathogenesis andprevention of nosocomial infections. It appears that manypeople believe that if we would only improve our hand hygienepractices, all of our nosocomial infection problems would beresolved. In this regard, we believe our study sounds a caution-ary note. We stand behind the major conclusion of our article:the prevention of nosocomial infection is a multifaceted issue,and hand hygiene is but one part of the equation.1 As we noted,the lack of association between hand hygiene compliance andnosocomial infection incidence should not be interpreted tomean that hand hygiene is not important.1 In fact, we believe


our article, we did not neglect to state the inherent limitationsof our study, which were, to a large extent, inherent in a single-center study with limited financial support.1 We hope govern-mental funding agencies will note the need to support ade-quately powered, multicenter, cluster-randomized studies toanswer the important questions that exist in our field. Al-though our study was not perfect, which we pointed out in theDiscussion section of the article1 and which was again empha-sized by the various letters to the editor,2-5 it is among the beststudies available on the subject.

acknowledgmentsFinancial support. The work reported in our study1 was supported in part bya study contract between the University of Nebraska Medical Center, Omaha,Nebraska, and Gojo Industries.

Potential conflicts of interest. M.E.R. reports receiving a stipend fromGojo Industries to cover travel expenses associated with presentation of thestudy1 at the 14th Annual Scientific Meeting of the Society for HealthcareEpidemiology of America in 2004. Gojo Industries approved the study butdid not have a role in study design, data acquisition, data analysis, ormanuscript preparation.

Mark E. Rupp, MD

Department of Internal Medicine, University of Nebraska Medical Center,Omaha, Nebraska.

Address reprint requests to Mark E. Rupp, MD, 984031 Nebraska MedicalCenter, Omaha, NE 68198-4031 ([email protected]).


references

1. Rupp M, Fitzgerald T, Puumala S, et al. Prospective, controlled, crossovertrial of alcohol-based hand gel in critical care units. Infect Control HospEpidemiol 2008;29:8 –15.

2. Maiwald M. Alcohol-based hand hygiene and nosocomial infection rates.Infect Control Hosp Epidemiol 2008;29:579 –580.

3. Mermel LA, Boyce JM, Voss A, Allegranzi B, Pittet D. Trial of alcohol-based hand gel in critical care units. Infect Control Hosp Epidemiol 2008;29:577–579.

4. Widmer AF, Rotter M. Effectiveness of alcohol-based hand hygiene gels inreducing nosocomial infection rates. Infect Control Hosp Epidemiol2008;29:576.

5. McGuckin M, Waterman R. “Cannot detect a change” is not the same as“there is not a change.” Infect Control Hosp Epidemiol 2008;29:576 –577.

6. Study demonstrates sustained infection prevention takes a compre-hensive strategy. Available at: http://www.shea-online.org/Assets/files/SHEA_News_-_February_2008.pdf. Accessed May 9, 2008.

7. Leading expert disputes interpretation of hand hygiene survey at Univer-sity of Nebraska Medical Center. Available at: http://www.shea-online.org/Assets/files/handgelreleaseFinal.pdf. Accessed May 9, 2008.



8. Sebille V, Chevret S, Valleron AJ. Modeling the spread of resistant noso-comial pathogens in an intensive care unit. Infect Control Hosp Epidemiol1997;18:84 –92.

9. Silvestri L, Petros AJ, Sarginson RE, de la Cal MA, Murray AE, vanSaene

12. World Health Organization. WHO guidelines on hand hygiene in healthcare (advanced draft). Geneva: World Health Organization, 2006. Avail-able at: http//www.who.int/patientsafety/information_centre.

13. Test Report B 10988ae, Purell Hygienic Hand Rub Gel, EN 1500-Hygienic


HKF. Handwashing in the intensive care unit: a big measure with modesteffect. J Hosp Infect 2005;59:172–179.

10. Eckmanns T, Schwab F, Bessert J, et al. Hand rub consumption and handhygiene compliance are not indicators of pathogen transmission in inten-sive care units. J Hosp Infect 2006;63:406 – 411.

11. Ali Y, Dolan MJ, Fendler EJ, Larson EL. Alcohols. In: Block SS, ed. Disin-fection, Sterilization, and Preservation. 5th ed. Philadelphia: LippincottWilliams & Wilkins; 2001:229 –253.


Handrub, May 19, 2006, HygCen International, Bischofshofen, Austria.Copies available from Gojo Industries.

14. International Code Council. Chapter 34. Flammable and CombustibleLiquids. In: 2006 International Fire Code. Falls Church, VA: InternationalCode Council; 2006:305–343.

15. Traore O, Hugonnet S, Lubbe J, Griffiths W, Pittet D. Liquid versus gelformulation: a prospective intervention study. Critical Care 2007;11:R52.



Documents

Effectiveness of Alcohol-Based Hand Hygiene Gels in Reducing Nosocomial Infection Rates