THE LANCET32 Jamestown Road London NW1 7BY UK

Axel KramerPeter RudolphGonter Kampf

Didier Pittet

Limited efficacy ofalcohol-based hand gels

Reprinted from THE LA CETSaturday 27th April 2002 Vol. 359No. 9316 Pages 1489-1490

Limited efficacy of alcohol-basedhand gelsAxel Kramer, Peter Rudolph, Gonter Kampf, Didier Pittet

Alcohol-based gels have been Introduced recently In manyhospitals worldwide for hand antisepsis. We investigated theantimicrobial efficacy of ten gels and four rinses according toEuropean standards (EN 1500). No gel met the EN 1500requirements within 30 s of application, whereas all handrinses did. The tested hand gels should be considered aretrograde step for hand hygiene because the application timeIn clinical practice Is often shorter than 30 s; they should notreplace alcohol-based liquid hand disinfectants currently usedIn hospitals or be Implemented as first choice agents.

Lancet 2002; 359: 1489-90

Control of hospital-acquired infections represents a majorchallenge to modern medicine. osocomial pathogens aremostly transmined via the hands of health-care workers andhand hygiene is considered the leading preventive measure toreduce cross-transmission in health-care senings.'

Two different types of hand hygiene procedures can bedistingui hed worldwide.' Antinllcrobial or plain soap andwater is mainly used for hand hygiene in the USA, andwaterless alcohol-based hand rub in many parts of Europe ifhands are not visibly dirty. Arguments for the preference ofwaterless hand antisepsis are that it acts faster, irritates handsIe soften, ha higher efficacy than handwashing with plain orantimicrobial soap, and can be immediately available at thebedside.'" Recently, we verified the relation betweensustained improvement in hand-hygiene compliancefacilitated by the routine use of hand rub and reducednosocomial infection rates.'

Irritant dermatitis resulting from frequent application ofsoaps and detergents is commonly experienced by caregivers. Gel formulations have been proposed to reduce thedrying effect of alcohols and potentially enhance compliancewith hand hygiene, which remains almost universally 10w.'.2

Hand-hygiene agents should have antimicrobial efficacyagainst nosocomial pathogens prior to introduction. InEurope, the state-of-the-art protocols to test their efficacyare referred to as the European norms (E .' EN 1500 is thestandard by which the efficacy of waterless product such ashand rinses or gels are tested under practical conditions bycomparison with the reference disinfectant (2-propanol, 60%volume per volume lv/v]) tested on Eschen'chia coli KI2(NTCC 1053).' The tested product should not beignificantly less effective than the reference alcohol.

The antimicrobial efficacy of each product was comparedwith 2-propanol 60% (v/v) on artificially contaminatedhands using a crossover design with 15 volunteers.' Eightvolunteers rubbed their hands with the tested product first,and the other seven with the reference alcohol first. Handswere washed for I min with soft soap, dried with papertowels, immersed in the contamination fluid up to themidmetacarpals for 5 s with fingers spread, and then allowedto dry for 3 min. Fingenips were rubbed for I min in a petridish containing a liquid broth (prevalues). 3 mL of the testedproduct were applied to the hands. Postvalues weredetermined immediately after the rub-in period using petridishes containing liquid broth with neutralisers' For bothreference and test procedure the log countS of bacteriafrom the left and right hands of each person were averagedseparately for prevalues and postvalues. The arithmeticmeans of all individual log reduction factors were calculated.Wilcoxon's matched-pairs signed rank test (one-sided) wasused for comparison (p=O'O I).'

THE lANCET· Vol 359· April 27, 2002· www.thelancel.com

RE EARCH LETTERS

Tested hand gels and rinses contained mainly ethanol orI-propanol or 2-propanol as active ingredients (table). Thetotal alcohol content of the gels varied from 53% to 70%(v/v). The mean microbial reduction factors of the referencealcohol varied from 3·7 to 5·07 and those of the gels from2'13 to 4·09. The mean reduction factor of each gel wasabout I log-step lower than that of the reference alcohol(table). The mean reduction factors of the four hand rinsestested varied from 4·26 to 4· 8 and did not differsignificantly from that of the reference alcohol (table).

Our data show that a 30 s handrub with a gel containing atotal amount of up to 70% (v/v) alcohol is significantly lesseffective than a hand rub with 2-propanol 60% (v/v). Mostalcohol-based hand rinses meet the E 1500 requirementwithin 30 s of application,'" a highly desirable prerequisitefor their use in the health-care setting. The tested gels didnot fulfil this criterion, whereas the rinses did.

Thus, in hospitals where most health-care workers usealcohol-based solutions that already meet the E 1500requirements, the introduction of any of the tested gelswould be a backward step and unnecessarily lower thehygiene standard. An increased risk of cross-transmissionwould certainly result because the application time in dailypractice averages 8-15 s and is unlikely to exceed 30 s.'·' Themain reason for the use of gel formulation is to reduce skinirritation and dryne s, but this can be prevented by theaddition of emollients in hand rinses or by the use of aprotective skin cream.'

When taking a closer look at the antimicrobial efficacy ofdifferent alcohols, it is not surprising that an ethanol contentof up to 70% (v/v) is not as effective as 2-propanol 60%(v/v). In terms of bactericidal efficacy, I-propanol can beregarded as the most effective alcohol followed by2-propanol and ethanol.' Comparison of 2-propanol withethanol showed that the efficacy of 2-propanol 60% (v/v) isalmost equivalent to ethanol 80% (v/v). Therefore,ethanol-based hand formulations should contain at least

0% ethanol (v/v). Based on our efficacy data, we do notconsider any of the tested alcohol-based hand gels to besuitable for hand antisepsis in the health-care sening becausetheir antimicrobial efficacy may be insufficient to prevent thespread of pathogens. Future ethanol-based hand gels used inhospitals should contain at least 80% (v/v) ethanol as theactive ingredient and should be as effective as the E 1500reference alcohol within 30 s.

ContributorsA Kramer and D Pinet designed the srudy. D Pinet was responsible forstudy coordination, dala colltction, analysis, and writing of the paper.A Kramer and P Rudolph contributed to the proiect idea, laboratoryanalyses, and wrinng of the paper. G Kampf participated in datacollection and writing of the paper.

Conflict of interest statementG Kampf is a paid employec of Bodc Chemie GmbH.

AcknowledgmentsBode Chemie GmbH assisted with data collection and analysis of handrinses.

Rotter ML. Hand washing and hand disinfection. In: Mayhall CG, ed.Hospital epidemiology and infection control, 2nd edn. Philadelphia:UppinCOtt Williams and Wilkins, 1999: 1339-55.

2 Pinet D, Hugonner ,Hamanh S, et al. Effecth'eness of a hospital­wide programme to improve compliance with hand hygiene. LanUl2000; 354: 1307-12.

3 Deutsches Instirut fUr ormung, Europcan orm 1500. ChemischeDesinfektionsminel und Antiseptika. Hygienische Hiindedcsinfektion.Priifverfahren und Anforderungen (Phase 21Srufc 2), 1997.

4 CE (European Comminee for Standardization) EN 1500. Chemicaldisinfectants and antiseptics-hygcnic handrub: tcst method andrequirements (phase 2, step 2). Brussels: CE ,1997.

5 Societe fran~aisc d'hygiene hospitaliere. Uste positive des

14 9

RESEARCH LETTERS

Product Active Ingredients Mean (median) Mean (median) Difference preduction factor reduction factorof product alcohol of reference alcohol

Hand gelsAssanis pro Ethanol (53%) 3·31 (3·28) 4·28 (4·28) 0·97 <0·01Endure 300 Ethanol (70%) 2·13 (2'19) 4·12 (4·10) 1·99 <0·01Gel-Hydro-alcoollque Ethanol (60%) 4·09 (4·11) 5·07 (5·03) 0·98 <0·01Levermed Alcohol Gel 1-propanol and 2-propanol (total: 70%) 3·87 (3·98) 4·58 (4·43) 0·71 <0·01Manugel 2-propanol (60%) plus other antiseptic ingredients 4·07 (3·99) 4·96 (4·66) 0·89 <0·01Microsan Ethanol (70%) 3·36 (3·43) 4·26 (3·98) 0·89 <0·01Prevacare Ethanol (60%) 3·07 (3·10) 4·12 (4·10) 1·05 <0·01Purell Ethanol (62%) 3·07 (3·05) 4·10 (4·10) 1·03 <0·01Splrigel Industrial methylated spirits (70%) 3·58 (3·57) 4·68 (4·64) 1·10 <0·01Stokosept Ethanol (57%) 2·68 (2·59) 3·78 (3·79) 1·10 <0·01

Hand rinsesAHD 2000 Ethanol (75%) 4·78 (4·62) 4·78 (4·89) 0 NSMonorapid Synergy Ethanol (54%) and 1-propanol (10%) 4·32 (4'59) 4·45 (4·51) 0·13 NSSoftaman CH 2-propanol (45%) and 1-propanol (30%) 4·88 (5·24) 4·23 (4·07) 0·55 NSSterillium 2-propanol (45%) and 1-propanol (30%) plus 4·26 (4'19) 4·10 (4·03) 0·16 NS

Mecetromium etllsulfate (0·2%)

NS=not significant. Compounds are listed in alphabetical order. Mean (median) reduction factors are given. p values were derived using Wllcoxon's matched·pairsslgned·ranks test.

Comparative efficacy of alcohol-based hand antisepsis agents with the EN 1500 reference alcohol

desinfectanlS 2000. Hygiene 2000; 8: 73-89.

Institute for Hygiene and Environmental Medicine, University ofGrelfswald, Grelfswald, Germany (A Kramer MO. P Rudolph MO);

Bode Chemle GmbH, Scientific Affairs, Hamburg, Germany(G Kampf MO); and Infection Control Programme, Department ofInternal Medicine, University of Geneva Hospitals,1211 Geneva 14, Switzerland (Prof D Pittet MO)

Correspondence to: Prof Didier Pittet(e-mail: didier.pittet@hcuge.ch)

1490 THE lANCET· Vol 359 • April 27, 2002 • www.thelanceLcom

The Lancet is a weekly subscription journal. For further information on how to subscribe please contact our Subscription DepartmentTel: +44 (0) 20 76114102 Fax: +44 (0) 2076114479

North America Tel: +12126333807 Fax: +12126333850

Printed in UK © 2002 The Lancetby Mailing and Printing Company Limited Unit B5, Haslemere Industrial Estate, Pig Lane, Bishop's Stortford, Herts, CM23 3HG, UK