Alcohol-based hand rubs are recommended and widely used for surgical hand disinfection in many countries 12. The main aim of the pre-operative treatment of hands is the reduction of resident skin bacteria to a minimum in order to reduce the risk of surgical site infections in case of perforated surgical gloves 3. Alcohols are considered to have better antimicrobial efficacy and dermal tolerance which are two clear advantages in comparison to antimicrobial soaps 45. The application time for alcohol-based surgical hand rubs has become shorter over the last decades, and there was increasing evidence that with well-formulated hand rubs an equivalent efficacy on the resident hand flora can be achieved at an application time of 1.5 min 67. At the same time evidence was cumulating to suggest that a 1 min hand wash which is recommended and often performed before the disinfection period may reduce the efficacy of the alcohol to some extent 8. But this effect has so far only been studied for a 3 min application time of the hand rubs. The influence of a preceding hand wash as well as of drying by evaporation of the hand rub on the efficacy of disinfection with an application time of 1.5 min is unknown. Therefore, these influences were studied by determination of the efficacy of a hand rub according to EN 12791 and compared with the 3 min reference treatment 9. The test design of EN 12791 ensures that the treated hands are sampled immediately after the specific application time even if they are still wet with the hand rub. In clinical practice, however, it is recommended for healthcare workers to put on surgical gloves only onto dry hands due to a better skin tolerance and a lower risk for impairment of glove integrity 210. Therefore, the test design of EN 12791 does not reflect the clinical practice of surgical hand disinfection in that particular point.

The 3 min reference disinfection reduced resident hand bacteria on average by 1.8 log₁₀ steps (immediate effect) and 1.4 log₁₀-steps (sustained effect; Table 1) An application of the hand rub after a 1 min hand wash was equally effective (immediate efficacy: 1.5 log₁₀-steps; sustained efficacy: 1.6 log₁₀-steps). An application of 1.5 min of the hand rub without a 1 min hand wash was slightly more effective (immediate efficacy: 2.3 log₁₀-steps; sustained efficacy: 1.7 log₁₀-steps). Application of the hand rub for 1.5 min followed by 1 min drying time revealed the best efficacy (immediate efficacy: 2.3 log₁₀-steps; sustained efficacy: 2.0 log₁₀-steps). A comparison of all three treatment variations and the reference treatment revealed a significant difference for the immediate efficacy (p = 0.026; Friedman test), but no significant difference for the sustained efficacy (p = 0.430). A post-hoc-test of pairwise differences for the immediate efficacy revealed only that method C was significantly more effective than method A (p < 0.05; Wilcoxon-Wilcox test). None of the pairwise comparisons of methods A to C with the reference reached statistical significance.

For the first time it was demonstrated that the efficacy requirements of EN 12791 can be fulfilled with a propanol-based hand rub applied for 1.5 minute time with three different application procedures. The lowest efficacy was observed with a preceding 1 min hand wash, the highest efficacy was found without a preceding hand wash and with an additional drying time of 1 min after the disinfection period.

It has been shown before that a 1 min hand wash followed by drying hands with a paper towel has a significant effect on skin hydration which significantly increases for up to 9 min 8. At the same time it was found that the efficacy of different types of alcohols is impaired by a hand wash to some extent 8. The value of a hand wash before the disinfection period has been questioned recently in a draft European recommendation on surgical hand disinfection for two reasons 12: First, a 1 min hand wash does not really reduce the number of skin bacteria to a relevant extent 413. Second, it has been suggested in some studies that the efficacy of the alcohol-based hand rub is even reduced if hands are washed before disinfection 1314. The difference in efficacy of the same treatment was even more remarkable with and without a 2 min hand wash 15. We can now add similar findings with an application time of 1.5 minutes.

Practical implications have to be adressed, too. The whole procedure of surgical hand disinfection is usually performed in the surgical theater block. If we assume that the surgeon enters the theater block with clean hands then there is no need to wash them before the disinfection period. If we assume that the surgeon enters the theater block with dirty hands then there is something wrong on another level. If we assume that hands have become dirty in the surgical theater block they should be washed immediately after soiling 1. In that respect there is no reason to include a 1 min hand wash into the standard procedure of pre-operative treatments of hands especially since recent evidence points out some clear disadvantages of the hand wash.

The best efficacy was found without a hand wash and with additional drying time after the application of the hand rub. The immediate effect was only marginally better (0.03 log₁₀-steps), the sustained effect improved somewhat more (0.3 log₁₀-steps). It is certainly correct that the efficacy of a hand rub which is applied for a specific time such as 1.5 min should be assessed exactly after 1.5 min. At the same time the application by the healthcare workers will be different in the hospital as they have to await the evaporation of any residual alcohol from their hands before gloving. But the effect of the additional drying time has to our knowledge never been investigated with an alcohol-based hand rub although it has a clear practical relevance. We observed a trend towards a better overall efficacy if hands are allowed to dry for 1 min after the disinfection period. This finding was expected and can be explained by the additional and variable contact time of the alcohol with the skin. That is why the healthcare worker can expect that the "real efficacy" of an alcohol-based hand rub for surgical hand disinfection is to some extent better than measured according to EN 12791.

An application of the propanol-based hand rub for 1.5 min after 1 min hand wash fulfills the efficacy requirements of EN 12791. The efficacy can be improved to some extent by omitting the preceding hand wash and by awaiting the evaporation of the alcohol which is clinical practice anyway. The preceding hand wash has the most negative effect on the immediate effect. Based on our data hands should not be routinely washed before the disinfection period unless there is a good reason for it such as visible soil.

The second author is paid employee of Bode Chemie GmbH & Co. KG, Hamburg, Germany.

GK and AK designed the study, NOH, GK, PK and TK analysed the data, NOH, GK and AK wrote the manuscript. All authors read and approved the final manuscript.