Journal of Hospital Infection 79 (2011) 115e118

Available online at www.sciencedirect.com

Journal of Hospital Infection

journal homepage: www.elsevierhealth.com/journals/jhin

Testing of the World Health Organization-recommended formulations

for surgical hand preparation and proposals for increased efficacy
M. Suchomel a, *, M. Kundi b, B. Allegranzi c, D. Pittet c, d, e, M.L. Rotter a
a
Institute of Hygiene and Applied Immunology, Medical University of Vienna, Vienna, Austria
b
Department of Public Health, Medical University of Vienna, Vienna, Austria
c
First Global Patient Safety Challenge, World Health Organization, Geneva, Switzerland
d
Infection Control Programme, University of Geneva Hospitals and Faculty of Medicine, Geneva, Switzerland
e
World Health Organization Collaborating Centre on Patient Safety, University of Geneva Hospitals and Faculty of Medicine, Geneva, Switzerland

a r t i c l e i n f o s u m m a r y

Article history: The 2009 World Health Organization (WHO) Guidelines on hand hygiene in health care
Received 7 October 2010
recommend alcohol-based hand rubs for both hygienic and pre-surgical hand treatment. Two
Accepted 9 May 2011
by J.A. Child formulations based on ethanol 80% v/v and 2-propanol 75% v/v are proposed for local prep-
Available online 7 July 2011 aration in healthcare settings where commercial products are not available or too expensive.
Both formulations and our suggested modifications (using mass rather than volume percent
Keywords: concentrations) were evaluated for their conformity with the efficacy requirements of the
Alcohol-based hand rub forthcoming amendment of the European Norm (EN) 12791, i.e. non-inferiority of a product
Hand hygiene
when compared with a reference procedure (1-propanol 60% v/v for 3 min) immediately
Surgical hand preparation
WHO-recommended formulations and 3 h after antisepsis. In this study, the WHO-recommended formulations were tested for
3 min and 5 min. Neither formulation met the efficacy requirements of EN 12791 with
3 min application. Increasing the respective concentrations to 80 w/w (85% v/v) and 75 w/w
(80% v/v), together with a prolonged application of 5 min, rendered the immediate effect of
both formulations non-inferior to the reference antisepsis procedure. This was not the case
with the 3 h effect, which remained significantly inferior to the reference. Although the
original formulations do not meet the efficacy requirements of EN 12791, the clinical signifi-
cance of this finding deserves further clinical trials. To comply with the requirement of EN
12791, an amendment to the formulations is possible by increasing the alcohol concentrations
through changing volume into mass percent and prolonging the duration of application from
3 min to 5 min.
Ó 2011 World Health Organization. Published by Elsevier Ltd on behalf of the Healthcare
Infection Society. All rights reserved.

Introduction
In 2009, the World Health Organization (WHO) issued the Guide-
lines on hand hygiene in health care with a strong recommendation for
the use of alcohol-based hand rubs as the preferred means of routine
hand antisepsis for both hygienic hand disinfection during daily routine
practice and for preoperative hand treatment of the surgical team if
hands are not visibly soiled.1 These guidelines include two hand-rub
formulations for local preparation in healthcare settings where
* Corresponding author. Address: Institute of Hygiene and Applied Immunology,
Medical University Vienna, Kinderspitalgasse 15, 1095 Vienna, Austria. Tel.: þ43 1
4277 79420; fax: þ43 1 4277 9794.
E-mail address: miranda.suchomel@meduniwien.ac.at (M. Suchomel).
commercial products are either not available or too expensive. One
formulation is based on ethanol 80 volume percent (% v/v) (WHO1) and
the other (WHO2) on 2-propanol 75% v/v. Alcohols are the active
components in these formulations, and a low concentration of
hydrogen peroxide is incorporated to help prevent contamination with
bacterial spores during storage. Glycerol is also added as a humectant to
improve dermal acceptability of the solutions.
We undertook to evaluate the bactericidal efficacy of these
formulations in in vivo laboratory studies with volunteers for
conformity with the efficacy requirements of both European
Norms, EN 1500 for hygienic hand disinfection (separate study),
and, in this investigation, the forthcoming amendment of EN 12791,
which was designed to test surgical hand-rub products.2 This norm
requires non-inferiority of a product within maximum application

0195-6701/$ e see front matter Ó 2011 World Health Organization. Published by Elsevier Ltd on behalf of the Healthcare Infection Society. All rights reserved.
doi:10.1016/j.jhin.2011.05.005
116 M. Suchomel et al. / Journal of Hospital Infection 79 (2011) 115e118
duration of 5 min when compared with a reference hand antisepsis
procedure (R) using 1-propanol 60% v/v performed for 3 min in
parallel with the study formulation on the hands of the same
volunteers. In this study, the formulations were tested for 3 min
and 5 min.
Methods
The following formulations were prepared at the Institute of
Hygiene and Applied Immunology, Medical University Vienna,
Austria, according to the instructions contained in the WHO
guidelines.1
e WHO1: ethanol (pro analysi, Merck, 1.00983, Darmstadt,
Germany) 80% (v/v); hydrogen peroxide (pro analysi, Merck,
1.07210) 0.125% (v/v); and glycerol (pro analysi, Merck,
1.04092) 1.45% (v/v).
e WHO2: 2-propanol (pro analysi, Merck, 1.09634) 75% (v/v);
hydrogen peroxide 0.125% (v/v); and glycerol 1.45% (v/v).
e WHO1-modified: ethanol 80% [weight/weight (w/w)];
hydrogen peroxide 0.125% (v/v); and glycerol 1.45% (v/v).
e WHO2-modified: 2-propanol 75% (w/w); hydrogen peroxide
0.125% (v/v); and glycerol 1.45% (v/v).
e R: 1-propanol (pro analysi, Merck, 1.00997) 60% (v/v), the
reference alcohol of EN 12791.2
Twenty-four volunteers participated in the study. Exclusion
criteria were: aged <18 years; pregnancy; and skin breaks on
hands, such as cuts, abrasions or other skin disorders. Nails were
short and clean and the volunteers agreed not to take or use any
antibacterial substance or antibacterial soap during the trial start-
ing from one week prior to testing. All volunteers gave written
informed consent. The study protocol was approved by the insti-
tutional ethics committee of the Medical University of Vienna.
Culture media were those described in EN 12791.2 Sampling and
dilution fluids were: tryptic soy broth (TSB) (Caso broth, Merck,
1.05459); and counting plates: tryptic soy agar (TSA) (Caso agar,
Merck, 1.05458). Neutralising agents were not necessary for any of
the tested WHO-recommended or modified formulations because
even dilution with pure broth without supplement was shown in
many previous validation tests to neutralize any antimicrobial
effect.
The efficacy of the WHO-recommended formulations, WHO1
and WHO2, and the modified formulations was compared in three
individual test series with R, the standardized reference surgical
hand preparation procedure of EN 12791, that requires hands to be
kept wet with N  3 mL 1-propanol 60% (v/v) for 3 min. A Latin-
square design was used with three groups of eight randomly
allotted volunteers each, and as many experimental runs were
performed as there were formulations including R. In every run, all
hand antisepsis procedures were applied concurrently. At the end
of the third series, every volunteer had used each procedure once.
Test runs were spaced by one week to allow regrowth of the normal
skin flora.
The test method was as described in EN 12791.2 The efficacy of
the formulations was tested after an initial prewash with soap and
water. Pretreatment values were established by rubbing and
kneading the fingertips, including the thumbs, of both hands for
1 min at the base of a Petri dish containing 10 mL sampling fluid,
one for each hand. After hands had been dried with paper towels,
3 mL portions of the respective formulations were poured into the
cupped hands of volunteers and vigorously rubbed on to both
hands up to the wrists; as many portions were applied as were
necessary to keep hands wet for either 3 min or 5 min. To assess
post-treatment values, fingertips were sampled as described above
Table I
Immediate and 3 h effects after a 3 min application of WHO-recommended
formulations WHO1 and WHO2 compared with the reference (R) of EN 12791
Formulation Application Mean log10 reduction  SD (N ¼ 24)
Immediate effect 3 h effect
WHO1 N  3 mL per 3 min 1.61  0.70a 0.76  0.74a
WHO2 N  3 mL per 3 min 1.43  0.53a 0.58  0.65a
R N  3 mL per 3 min 2.38  0.89 1.66  0.85
WHO1: ethanol 80% (v/v), hydrogen peroxide 0.125% (v/v), and glycerol 1.45% (v/v);
WHO2: 2-propanol 75% (v/v), hydrogen peroxide 0.125% (v/v), and glycerol 1.45%
(v/v); R: 1-propanol 60% (v/v).
a
Significantly inferior to R [HodgeseLehmann upper one-sided 97.5% confidence
limit of individual differences in log10 bacterial reductions between test formulation
and R  0.75 log10 (agreed inferiority margin)].
for determination of pretreatment values, either immediately after
hand antisepsis on one randomly selected hand or, after air-drying
and gloving, 3 h after antisepsis on the other hand. Quantitative
surface cultures from all sampling fluids and their decimal dilutions
were done on TSA, and counting plates were incubated at 36  1  C
for a total of 48 h.
For statistical evaluation, viable counts were processed as
described in the amended version of EN 12791.2 Pre- and post-
treatment colony counts per millilitre of sampling fluid were
expressed as decadic logarithms (log10). Pretreatment values of
study formulations and R were tested for significant differences by
Friedman analysis of variance per experimental run with an agreed
level of significance of P ¼ 0.05. From the intra-individual differ-
ences between log10 pretreatment minus log10 post-treatment
value, individual log10 reduction factors (log10 RFs) were calcu-
lated separately for both the immediate and the 3 h effects. The
differences between individual log10 RFs from study formulations
and the reference procedure were then calculated intra-
individually and tested for significance by a parameter-free
non-inferiority test according to HodgeseLehmann by means of
a self-developed computer program (M.K.).3 Non-inferiority was
supposed when the HodgeseLehmann upper one-sided 97.5%
confidence limit for the intra-individual differences in log10 RFs
between study formulation and R was smaller than the agreed
inferiority margin of 0.75 log10.
Results
No significant differences were found between the means of the
pretreatment bacterial counts in any of the three experimental
series (data not shown). Tables IeIII show the results of surgical
hand antisepsis with the WHO-recommended and with the
modified formulations. None of the WHO-recommended formula-
tions met the efficacy requirements of EN 12791 for surgical hand
preparation with 3 min application duration. Both formulations
Table II
Immediate and 3 h effects after a 5 min application of WHO-recommended
formulations WHO1 and WHO2 compared with the reference (R) of EN 12791
Formulation Application Mean (N ¼ 24) log10 reduction  SD
Immediate effect 3 h effect
WHO1 N  3 mL per 5 min 2.26  0.71a 0.95  0.61a
WHO2 N  3 mL per 5 min 1.46  0.51a 1.06  1.12a
R N  3 mL per 3 min 3.06  1.01 2.02  1.10
WHO1: ethanol 80% (v/v), hydrogen peroxide 0.125% (v/v), and glycerol 1.45% (v/v);
WHO2: 2-propanol 75% (v/v), hydrogen peroxide 0.125% (v/v), and glycerol 1.45%
(v/v); R: 1-propanol 60% (v/v).
a
Significantly inferior to the reference [HodgeseLehmann upper one-sided 97.5%
confidence limit of individual differences in log10 bacterial reductions between test
formulation and R  0.75 log10 (agreed inferiority margin)].
 M. Suchomel et al. / Journal of Hospital Infection 79 (2011) 115e118
Table III
Immediate and 3 h effects after a 5 min application of WHO1 and WHO2 modified
WHO-recommended formulations compared with the reference (R) of EN 12791
Formulation Application Mean (N ¼ 24) log10 reduction  SD
Immediate effect 3 h effect
WHO1-modified N  3 mL per 5 min 2.47  1.08 1.30  0.69a
WHO2-modified N  3 mL per 5 min 2.78  0.88 1.36  0.76a
R N  3 mL per 3 min 2.53  1.11 2.03  1.25
WHO1-modified: ethanol 80% (w/w), hydrogen peroxide 0.125% (v/v), and glycerol
1.45% (v/v); WHO2-modified: 2-propanol 75% (w/w), hydrogen peroxide 0.125% (v/v),
and glycerol 1.45% (v/v); R: 1-propanol 60% (v/v).
a
Significantly inferior to the reference [HodgeseLehmann upper one-sided 97.5%
confidence limit of individual differences in log10 bacterial reductions between test
formulation and R  0.75 log10 (agreed inferiority margin)].
were significantly less effective than R both immediately and 3 h
after hand antisepsis (Table I) and this remained the case even with
5 min duration of application (Table II).
However, increasing the volume percent concentrations (v/v) of
both alcohols by about 5% (80% and 75% w/w respectively) together
with a prolonged application time of 5 min rendered the immediate
effect of the two modified WHO formulations non-inferior to R
(Table III). Unfortunately, this was not observed with the 3 h effect,
which remained significantly inferior to R.
Discussion
Although both WHO-recommended formulations meet the
efficacy requirements of EN 1500 for testing hygienic hand disin-
fectants, the results of this study demonstrate that they do not meet
those of EN 12791, the European norm for testing products for pre-
surgical hand preparation.1 Moreover, even prolongation of the
application time from 3 min to 5 min, the longest exposure allowed
by the norm, did not result in a positive outcome.
Hence, the question arises as to the meaning of this finding for
clinical practice in the light of the WHO First Global Patient Safety
Challenge.4,5 Although the expert group was aware of the possible
deficiency in efficacy of the WHO formulations relative to the EN
12791, it was of the opinion ‘that the microbicidal activity of
surgical hand antisepsis is still an ongoing issue for research as due
to lack of epidemiological data, and there is no indication that the
efficacy of n-propanol (1-propanol) 60% v/v as a reference in EN
12791 finds a clinical correlate.’1 Indeed, the history of the devel-
opment of the European Norm for the validation of products for
pre-surgical hand antisepsis reveals that n-propanol 60% (v/v) was
chosen arbitrarily as a reference alcohol because it was shown to be
the most efficacious of three tested short-chain, aliphatic alcohols
(ethanol, 1- and 2-propanol) in in vivo laboratory comparisons on
the hands of volunteers.6 At that time, the intention was to choose,
for surgical hand preparation, the efficacy of the most active alcohol
as a yardstick for the performance of any surgical hand-rub product,
but at an acceptable and skin-tolerable concentration. In conse-
quence, the norm requires that the bactericidal activity of a surgical
hand rub shall not be inferior to that of the reference antiseptic
procedure employing 1-propanol 60% v/v. With ethanol and
2-propanol, this requirement can also be met, but only at higher
concentrations and/or longer duration of application. It is impor-
tant to recognize that the 1-propranol reference solution of the
norm does not contain any skin-protecting humectant and there-
fore should not be used in daily routine surgical hand preparation.
Modern products do contain such humectants, as acceptability and
skin integrity are no less important than bactericidal efficacy;
efficacy may be grossly influenced by chemicals such as these, a fact
requiring every new product to be evaluated by consented,
117
standardized methods even if the product contains 1-propanol 60%
v/v as the active agent.
For several reasons, the choice of 1-propanol as a reference
antiseptic is now regarded by many experts as inappropriate.1 Thus,
the final statement of the WHO experts in this matter is: ‘Consid-
ering that other properties of the WHO recommended formula-
tions, such as their excellent tolerability, good acceptance by
health-care workers, and low cost are of key importance for a sus-
tained clinical effect, the above results are considered acceptable,
and it is the consensus opinion of the WHO expert group that the
two formulations can be used for surgical hand preparation.’1
For the above-mentioned reasons, we share unanimously this
view, but we are still of the opinion that it might be favourable if the
recommended formulations could be rendered non-inferior to the
European standard as, in the guideline itself, conformity with
a standardized test method is required and higher bactericidal
efficacy in surgical hand antisepsis would certainly not be of
disadvantage for the patient. Our efforts to make the formulations
compatible with the efficacy requirement of the norm by increasing
the concentrations of the alcohols through changing the recom-
mended volume concentrations into mass concentrations, together
with prolongation of the duration of application from 3 min to
5 min, rendered the immediate effects from inferior to non-inferior
to those of the reference EN 12791 antiseptic procedure. (With
ethanol, a mass concentration of 80% w/w equals a volume
concentration of about 85% v/v, whereas a mass concentration of 75%
w/w 2-propanol is equivalent to a volume concentration of about
80% v/v.) Thus, this modification of the recommended formulations
could be a feasible way to harmonize the standard for immediate
efficacy of the WHO formulations with the requirement of EN 12791.
However, there is still the observation of inferior 3 h efficacy
even of the modified antiseptic solutions. Until results of further
investigations are available, this phenomenon must remain unex-
plained, although it seems obvious that there is a connection with
the other components of the formulations, i.e. glycerol or hydrogen
peroxide, whereby the former is more suspected to have a negative
influence on the efficacy. Importantly, the WHO formulation based
on 80% v/v ethanol (WHO1 in the current study) was recently
compared in a double-blind, randomized, cross-over trial with soap
and water for surgical hand preparation with surgical site infection
as clinical outcome.7 There was no significant difference in terms of
outcome. One may speculate that higher alcohol concentrations
could have improved the result.
In conclusion, the WHO-recommended formulations do not
meet the European efficacy standard EN 12791 on products for pre-
surgical hand antisepsis, but the clinical significance of this finding
remains unknown until further clinical trials become available.
In practice the present reference solution is not used for surgical
hand preparation. For the sake of harmonisation with the efficacy
requirement of the EN 12791, an amendment of the WHO formu-
lations would easily be possible by increasing the presently rec-
ommended alcohol concentrations through changing their volume
(v/v) into mass (w/w) percent and to prolong the recommended
duration of application from about 3 min to 5 min. The negative
influence of the other components of the WHO formulations on the
3 h effect requires further investigation.
Acknowledgements
We would like to thank M. Weinlich for her excellent technical
assistance.
Conflict of interest statement
None declared. WHO takes no responsibility for the information
provided or for the views expressed in this paper.
118 M. Suchomel et al. / Journal of Hospital Infection 79 (2011) 115e118

Funding sources 4. Pittet D, Donaldson L. Clean care is safer care: a worldwide priority. Lancet
2005;366:1246e1247.
None.
5. Allegranzi B, Pittet D. Preventing infections acquired during health-care delivery.
Lancet 2008;372:1719e1720.
References 6. Rotter M, Koller W, Kundi M. Eignung dreier Alkohole für eine Standard-Desin-
fektionsmethode in der Wertbestimmung von Verfahren für die Hygienische
1. World Health Organization. WHO guidelines on hand hygiene in health care. Geneva: Händedesinfektion. [Usability of three alcohols for a standard disinfection
World Health Organization; 2009. method to be employed for the evaluation of procedures for the hygienic
2. European Norm (EN) 12791. Chemical disinfectants and antiseptics. Surgical hand disinfection of hands. In German]. Zentralbl Bakteriol Orig B 1977;164:428e438.
disinfection e test method and requirement (phase 2/step 2). Brussels: Comité 7. Nthumba PM, Poenaru E-S, Poenaru D, et al. Cluster-randomized, cross-over trial
Européen de Normalisation; 2005. of the efficacy of plain soap and water versus alcohol-based hand rub for
3. Lehmann EL. Nonparametrics: statistical methods based on ranks. San Francisco: surgical hand preparation in a rural hospital in Kenya. Br J Surg 2010;97:
Holden-Day; 1975. 1621e1628.