Dose considerations for alcohol-based hand rubs

Journal of Hospital Infection 95 (2017) 175e182 Available online at www.sciencedirect.com

Journal of Hospital Infection journal homepage: www.elsevierhealth.com/journals/jhin

Dose considerations for alcohol-based hand rubs M.A.C. Wilkinson a, K. Ormandy b, C.R. Bradley a, *, A.P. Fraise a, J. Hines b a b

Hospital Infection Research Laboratory, Queen Elizabeth Hospital Birmingham, Edgbaston, Birmingham, UK Deb Group Ltd, Denby, Derbyshire, UK

A R T I C L E

I N F O

Article history: Received 3 March 2016 Accepted 20 December 2016 Available online 4 January 2017 Keywords: Alcohol-based hand rubs EN 1500 Hand hygiene User acceptability

S U M M A R Y

Background: Manufacturers’ recommended dosages for alcohol-based hand rubs are typically determined by measuring product efficacy using a model protocol such as EN 1500; however, anecdotal reports and informal observation suggests that in many cases users self-titrate to much lower doses in real-world application. Aim: To examine the interdependence of alcohol-based hand-rub volume on in-vivo efficacy using the EN 1500 standard test method, on drying time on users’ hands, and on their perceptions of acceptability. Methods: Three formulations were studied using EN 1500 and a modification of this method. The modification used volumes ranging from 0.5 to 3.0 mL and 30 s application. Drying times were recorded and user acceptability was established using a three-point scale (too long, OK, or too short). Dying times were analysed in relation to hand surface area. Findings: The drying time for all three products increased as a function of volume. The drying time displayed a positive association with volume and a negative association with hand surface area. The optimum volume for user acceptability was between 1.5 and 2 mL, yielding a drying time of between 20 and 30 s. Conclusion: Whereas EN 1500 is appropriate for establishing the efficacy of a hygienic hand-rub formulation compared to a benchmark, it does not reflect actual in-use conditions or the likely clinical effectiveness of the product. In particular, it fails to address the need to optimize the volume of application and user acceptability of the product. ª 2017 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved.

Introduction Alcohol-based hand rubs (ABHRs) have been used for the prevention of transmission of infections for many years and are widely endorsed, having been shown to be safe and highly effective.1e4 For effective infection prevention and control, a number of linked elements must be in place. First, hand hygiene must be performed by healthcare workers at the appropriate times * Corresponding author. Address: Hospital Infection Research Laboratory, Clinical Microbiology, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2WB, UK. Tel.: þ44 (0)121 627 2366; fax: þ44 (0)121 414 1682. E-mail address: [email protected] (C.R. Bradley).

where risk of transmission to the patient is greatest. Best practice guidelines such as ‘My 5 moments of hand hygiene’ have been developed and are widely endorsed.1,5 Second, an effective technique for hand sanitizing must be used to ensure that contamination is fully eliminated from frequently missed areas such as fingernails and intra-digital spaces. Again best practice methods are available and widely promoted.1 Third, a suitable ABHR formulation must be used, meeting acceptable standards for efficacy versus a broad spectrum of potential pathogens. The World Health Organization (WHO) has suggested two ‘free-to-use’ formulations, one based on ethyl alcohol and the other on isopropyl alcohol.6 More generally, various studies have demonstrated the efficacy of formulations comprising ethyl or propyl alcohol mixtures with water between 60% and 90% v/v.3 In order to validate the efficacy of

http://dx.doi.org/10.1016/j.jhin.2016.12.023 0195-6701/ª 2017 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved.

176

M.A.C. Wilkinson et al. / Journal of Hospital Infection 95 (2017) 175e182

such formulations, standard in-vivo test methods are available such as EN 1500, ASTM E-1174, and ASTM E-2755 which typically compare the performance of the test product under simulated contamination and handwashing conditions to that of a reference product or to an absolute reduction requirement, such as that recommended by the proposed rule published by the US Food and Drug Administration in 2015.7e10 Finally, for effective infection prevention, it is necessary to ensure that a product is used in an appropriate volume both in order to ensure complete coverage of the hands and to ensure that the hands remain wet for sufficient time for maximum effect. In their guideline for technique, WHO suggests 20e30 s as appropriate contact time for ABHR hand hygiene.11 This is supported by available studies on kill-time rates which indicate that alcohol is highly effective at times in excess of 15 s.1 Furthermore, the proposal of 20e30 s provides sufficient time for performance of all elements of the technique while remaining short enough to create a benefit in convenience for ABHR use compared to hand washing with soap and water, which WHO suggests should take 45e60 s.11 This element is not trivial; studies have demonstrated that convenience and product acceptability are key elements limiting compliance to hand hygiene best practice and recent studies have indicated that user expectation is for a rather short hygiene event, with wet time above 30 s poorly tolerated.12,13 However, it is widely regarded that the volume of ABHR recommended for use by manufacturers should be sufficient to pass the standard in-vivo efficacy test, typically EN 1500. Whereas this is an understandable position, the requirement of the test is determined by the volume and contact time of the reference product, in this case 23 mL 60% v/v isopropyl alcohol for 60 s. Generally, even highly efficacious formulations require a minimum of 3 mL to meet the test standard; it is not known whether 3 mL is appropriate to meet the WHO requirement of 20e30 s wet time. Recent work indicates that this volume of ABHR is likely to take longer than 30 s to dry on hands of various sizes, that full coverage may be achieved with smaller volumes, and that healthcare worker (HCW) acceptance is low.13,14 In this study, we attempt to clarify the interdependence of ABHR volume on in-vivo efficacy, drying time on users’ hands, and on user acceptability. We also consider how future standard test methods may be adapted to better provide a validation of ABHR efficacy under likely ‘real-world’ conditions and to guide realistic product dosage recommendations.

Methods

male), of whom two out of three were familiar with EN 1500, were recruited to test six volumes each of 60% IPA, WHOF1, and WHOF2. The volumes ranged from 0.5 to 3 mL, in 0.5 mL increments. The alcohol was applied to volunteers’ hands with a calibrated pipette, and rubbed in using the standard hand-rub procedure as described in EN 1500 (the Ayliffe technique). Volunteers self-reported when their hands were dry; the time from application was recorded. At the end of the test, the volunteers were asked to rate the time taken to dry on a threepoint scale: too short; OK; too long. The time taken to dry was withheld from volunteers in case it influenced their responses. The volunteers’ hands were checked to make sure they were visibly dry before each application; for all volunteers, applications were at least 30 min apart.

Antimicrobial efficacy and volume EN 1500 test: reduced volunteer To determine the antimicrobial efficacy of the three formulations, five of the volunteers (two female, three male) applied the same six volumes of 60% IPA, WHOF1, and WHOF2 in the manner of an EN1500 test. This involved washing the hands in a non-microbicidal standard soft soap as described in EN 1500, followed by immersion to the mid-metacarpal level in an overnight broth culture of Escherichia coli K12 (mean cfu/ mL ¼ 4.53108) for 5 s. The hands were then allowed to air-dry for 3 min. Subsequently, the hands were sampled for pre-counts by massaging the fingertips in sterile Petri dishes containing 10 mL tryptone soya broth (TSB) for 1 min. The requisite volume of alcohol was then added to the volunteers’ hands with a calibrated pipette, and rubbed in using the standard hand-rub procedure as described in EN 1500 for 30 s; the volunteers self-reported if and when their hands were dry. At the end of the rubbing procedure, the hands were sampled for post-counts by massaging the fingertips in sterile Petri dishes containing 10 mL TSB containing a validated neutralizer for 1 min. All samples were plated onto tryptone soya agar supplemented with 0.5 g/L sodium deoxycholate (tryptone soya selective agar, TSSA) and incubated at 37 C for 18e24 h followed by a further 24 h. Reduction factors were calculated by subtracting mean log10 post-values from mean log10 pre-values. The neutralizer comprised the following ingredients, per litre of distilled water: tryptone soy broth, 30 g; polysorbate 80, 30 mL; lecithin, 3 g; saponin, 30 g; sodium thiosulphate, 5 g; L-histidine, 1 g. This was shown to be non-toxic to the test organism and effective in neutralizing the reference and test products (data not shown).

Alcohol-based hand rubs EN 1500 test The formulations used were as follows:  WHOF1: ethanol, 80% v/v; glycerol, 1.45% v/v; hydrogen peroxide, 0.125% v/v.  WHOF2: isopropyl alcohol, 75% v/v; glycerol, 1.45% v/v; hydrogen peroxide, 0.125% v/v.  EN 1500 reference product/60% IPA: isopropyl alcohol, 60% v/v.

Drying time and volume In order to assess the relationship between volume of application and drying time, 15 volunteers (eight female, seven

Twenty volunteers (10 male, 10 female) performed a full EN 1500 test. The reference product was 23 mL of 60% IPA for 30 s each, giving a total of 6 mL over 60 s. The test products were 60% IPA, WHOF1, and WHOF2; each test product was applied as 3 mL for 30 s. This volume was chosen as a compromise between the need to dry within an acceptable/recommended time and the requirement of a sufficient level of antimicrobial efficacy. The order of product application for each volunteer was decided by a Latin-square design as described in EN 1500. The samples were processed in the same manner as for the reduced volunteer EN 1500. The mean cfu/mL of the E. coli broth was 2.60108. Each test product was tested for non-inferiority against the reference product. HodgeseLehmann upper

M.A.C. Wilkinson et al. / Journal of Hospital Infection 95 (2017) 175e182 one-sided 97.5% confidence limits for the difference in reduction factors produced by the reference product and each test product were calculated and compared to the agreed inferiority margin of 0.6. The full procedure is described in EN 1500 (2013).7

Measuring hand surface area To establish whether the surface area of volunteers’ hands had an effect on the drying time or antimicrobial efficacy of the products, the hand surface area (HSA) was calculated as in YaoWen and Chi-Yuang.15 In summary, the hand length (HL) was measured as the distance from the crease at the base of the hand to the tip of the middle finger; the hand breadth (HB) was measured as the distance between metacarpalephalangeal joints II and V. All measurements were recorded to the nearest half centimetre. The HSA in cm2 is then given by: HSA ¼ 2.48  HL  HB.

177

confidence intervals. The Spearman rank correlation coefficient was 0.807 for WHOF1, 0.860 for WHOF2, and 0.775 for 60% IPA (P < 0.001 for all three). When calculated across all volumes, the mean drying time was 24.22 s for WHOF1, 25.57 s for WHOF2 and 28.32 s for 60% IPA. A generalized linear regression model was fitted, which found that the drying time had a significant association with volume, product, gender, and HSA (P < 0.001, P ¼ 0.00514, P ¼ 0.00439, and P < 0.001, respectively). The drying time was positively associated with volume and male gender, and negatively associated with hand surface area. Whereas the model indicated significant variation between the products, for each volume, there was no significant difference between individual products (P > 0.05; Wilcoxon signed rank test, adjusted for multiple comparisons with the HolmeBonferroni method). Unsurprisingly, the data also suggest that the HSA of males was significantly larger than that of females (P < 0.001, ManneWhitney U-test).

Statistical analysis

Acceptability

Spearman rank correlation coefficients were calculated between the volumes and drying times for the three products (WHOF1, WHOF2, and IPA). The correlation between volume of application and log10 reduction factor (RF) produced by each product was assessed using Pearson product-moment correlation coefficients. Wilcoxon signed-rank tests with HolmeBonferroni adjustment for multiple comparisons were used both for pairwise comparisons between products of the drying time at different volumes of application, and for comparing log10 RF values between products in the full EN 1500 test. The ManneWhitney U-test was used to compare the hand surface areas of males and females. User acceptability and volume of application for the three products was compared using a Friedman rank sum test; this test was also used to compare the mean log10 RF values produced by each product, when calculated across all volumes. The KruskaleWallis test was used for assessing the difference in user acceptability between the products. A generalized linear regression model was fitted with drying time as the response variable, and with volume, product, gender, and HSA as the explanatory variables. This model employed a gammaresponse distribution with an inverse link function. A linear regression model was fitted using the data from the full EN 1500 test. The model had log10 RF as the response variable, and had product, gender, and HSA as the explanatory variables. Where parametric tests have been used, the data met any necessary assumptions. All statistical analyses were undertaken using: R version 3.2.3: ‘Wooden ChristmasTree’ (Copyright 2015, The R Foundation for Statistical Computing).

Figure 2 displays the user comments regarding drying time for the three products. The black dotted line represents the mean acceptable drying time, which was calculated across all three products as being equal to 25.33 s. Table I shows that there was a significant association between user acceptability and volume of application for all three products (P < 0.001; Friedman rank sum test), with user acceptability being greatest around the 1.5 mL and 2 mL applications. There was no evidence of difference between the products (P ¼ 0.4184; KruskaleWallis test).

Effect of volume upon bactericidal efficacy Figure 3 shows the log10 reduction factors produced by the different volumes of each product in the EN 1500-style test, along with linear regression lines and 95% confidence intervals. More than 12 years’ worth of EN 1500 tests at the Hospital Infection Research Laboratory have demonstrated that the reference product, 23 mL applications of 60% IPA over 60 s, produces a mean log10 RF of 5.16 (unpublished data). This value is represented by the thick dotted line on Figure 3. The thin dotted line represents the agreed inferiority margin of 0.6 log10 that is specified in EN 1500. The log10 RF values for all three products increased in a linear fashion as a function of volume, with a Pearson productmoment correlation coefficient of 0.733 for WHOF1, 0.614 for WHOF2, and 0.698 for 60% IPA (P < 0.001 for all three). There was no significant difference between the mean log10 RF values produced by each product, when calculated across all volumes (P ¼ 0.247; Friedman rank sum test). Table II shows a summary of the log10 RF values by product and volume.

Results EN 1500 results for WHOF1, WHOF2, and 60% IPA Effect of product, volume, gender, and hand surface area upon drying time The drying time for all three products increased as a function of volume. This increase was not linear and displayed evidence of heteroscedasticity. Figure 1 displays the spread of drying times per product, along with regression lines and 95%

Figure 4 show the results produced by 60% IPA, WHOF1, and WHOF2 (3 mL for 30 s) in a full EN 1500 test (20 volunteers), contrasted with the reference product (6 mL of 60% IPA for 60 s). The mean log10 RF produced by the reference product was 5.39, whereas those of IPA, WHOF1, and WHOF2 were 4.19, 4.56, and 4.80, respectively.

178

M.A.C. Wilkinson et al. / Journal of Hospital Infection 95 (2017) 175e182

80 60 WHOF1

40 20

60 WHOF2

Drying time (s)

80

40 20 80 60

IPA

40 20

1 Volume (mL)

2

3

Figure 1. Drying time by product and volume.

As a consequence, it is no great surprise that 60% IPA, WHOF1, and WHOF2 (3 mL for 30 s) all failed to fulfil the requirements of EN 1500 (2013). The 97.5% confidence limits for the difference in log10 RF values between the reference product and test product were 1.64 (IPA), 1.29 (WHOF1), and 1.01 (WHOF2). These were all in excess of the agreed inferiority margin of 0.6 log10. When comparing WHOF1 and WHOF2 with 3 mL of 60% IPA for 30 s, the 97.5% confidence limits for the difference in log10 RF values between the reference formulation and test product were 0.11 (WHOF1) and e0.09 (WHOF2). A linear regression model was fitted, with the results suggesting that the log10 RF was significantly associated with the reference product, with this association being positive (P < 0.001). Using a Wilcoxon signed rank test, adjusted for multiple comparisons with the HolmeBonferroni method, it is clear that the reference product produced a significantly higher log10 RF than IPA, WHOF1 and WHOF2 (P ¼ 0.0011, 0.0150 and 0.0304, respectively). The log10 RF did not appear to be affected by gender or HSA (P ¼ 0.5988 and 0.9782, respectively). Figure 5 displays the log10 RF values plotted against HSA for each product.

Discussion Although an EN standard, EN 1500 is designed to assess the efficacy of hygienic hand rubs; the test conditions do not

necessarily relate to practical use.7 The standard uses a set volume (23 mL of a reference alcohol) for a set time of 230 s. Also, from our experience the test and reference products are applied to hands that are often moist from the application of the culture. This does not relate to what actually happens in a clinical setting where the ABHR are applied to physically clean hands in smaller volumes for shorter rubbing times. ASTM E-1174 also has this drawback, which has been addressed in ASTM E-2755. Perhaps this should be taken into consideration in the development of future methods.16 Volunteers rated times longer than 30 s to be too long and 20e30 s to be acceptable. This is in line with WHO recommendations for time of application.11 A 3 mL application was considered to be too great a volume to dry in an acceptable time. Our results show that an acceptable volume was 1.5e2 mL. Other studies have also shown that 3 mL was perceived by the volunteers to be too long for drying in an acceptable time.13 The optimal volume and time of application is a balance between acceptability and efficacy. The log10 reductions obtained in the reduced volunteer EN 1500 showed a 2.8e3.8 log10 reduction when the smaller volume of 1.5e2 mL was used and >3.9 log10 when 3 mL was used. It is unclear whether the difference in log10 reductions obtained would have any clinical impact in practice. Kampf et al., when testing 15 volunteers, all non-healthcare workers, found that low volumes of ABHR

M.A.C. Wilkinson et al. / Journal of Hospital Infection 95 (2017) 175e182

179

80 60 WHOF1

40 20

60

Comment WHOF2

Drying time (s)

80

40 20

Too short Ok Too long

80 60 IPA

40 20

1

2

3

Volume (mL) Figure 2. Acceptability by volume and drying time.

dry more quickly, but lead to poor hand coverage.17 Macinga et al. found that volumes of ABHR capable of passing the methodology described in a draft revision of EN 1500 in 2009 all dried in >30 s; any difference in drying rates seemed to be largely due to alcohol concentration.18 Even 3 mL of ABHR, a volume that our data suggest takes longer than 30 s to dry, will often not pass EN 1500. Suchomel et al. found that 3 mL of WHOF1 and WHOF2, for 30 s, did not pass another draft revision of EN 1500 from 2012, but that 6 mL for 60 s was sufficient.19 This is in agreement with our findings. There is some evidence that increasing the alcohol content of ABHR may reduce the volume/time of application required to pass EN 1500.

Increasing the alcohol content of both WHO formulations led to them passing the draft revision of EN 1500 at 3 mL for 30 s.19 It must be noted, however, that this provisional version of the standard had an agreed margin of inferiority of 0.75 log10 units, as opposed to the 0.60 log10 units in the final version published in 2013; this means that only one of the formulations (WHOF2 modified to contain about 80% v/v IPA) actually met the requirements of EN 1500 (2013). Edmonds et al. found that ABHRs containing 70% v/v ethanol were capable of passing EN 1500 when applied as 3 mL for 30 s, but this was the 1997 version of the standard; the 2013 version is a more stringent test, containing as it does a requirement for a greater number of

Table I Mean drying time and acceptability for each product tested Product

WHOF1 WHOF2 60% IPA

Mean drying time (s) and acceptability (%) 0.5 mL

1 mL

1.5 mL

2 mL

2.5 mL

3 mL

10.07 (13.33%) 10.73 (13.33%) 11.40 (13.33%)

16.47 (53.33%) 16.33 (53.33%) 17.53 (33.33%)

22.13 (66.67%) 20.73 (60.00%) 27.07 (66.67%)

26.93 (53.33%) 31.00 (53.33%) 31.00 (73.33%)

33.13 (53.33%) 35.60 (26.67%) 37.20 (46.67%)

36.60 (20.00%) 39.00 (20.00%) 45.73 (26.67%)

WHOF1: ethanol, 80% v/v; glycerol, 1.45% v/v; hydrogen peroxide, 0.125% v/v. WHOF2: isopropyl alcohol, 75% v/v; glycerol, 1.45% v/v; hydrogen peroxide, 0.125% v/v. IPA: isopropyl alcohol, 60% v/v.

180

M.A.C. Wilkinson et al. / Journal of Hospital Infection 95 (2017) 175e182

6 5 WHOF1

4 3 2

5 WHOF2

Log10 reduction factor

6

4 3 2 6 5

IPA

4 3 2 2

1

3

Volume (mL) Figure 3. Reduction factor by product and volume.

volunteers and improved statistical analysis.20,21 Hence it is not entirely clear whether these ABHRs would have met the requirements of the new standard. Goroncy-Bermes et al. tested different volumes of ABHRs against EN 1500 (1997), in addition to measuring volunteer hand size.22 They found that volumes as low as 2.5 mL for 30 s could pass the standard, and that the bactericidal efficacy of ABHRs plateaued with increasing volume; this plateau occurred at a larger volume in male volunteers, possibly due to a tendency for men to have greater hand size. It must be noted, however, that the reference procedure

Table II Log10 reduction factor (RF) values by product and volume Product

WHOF1 WHOF2 60% IPA

Log10 RF 0.5 mL

1 mL

1.5 mL

2 mL

2.5 mL

3 mL

2.45 2.80 2.15

2.98 2.74 2.22

3.08 3.24 2.82

3.54 3.81 3.22

4.27 3.80 4.14

4.60 4.29 3.91

WHOF1: ethanol, 80% v/v; glycerol, 1.45% v/v; hydrogen peroxide, 0.125% v/v. WHOF2: isopropyl alcohol, 75% v/v; glycerol, 1.45% v/v; hydrogen peroxide, 0.125% v/v. IPA: isopropyl alcohol, 60% v/v.

deviated from that in the standard; instead of 23 mL of 60% IPA for a total of 60 s, 3 mL was added for 30 s, followed by up to 3 mL for a further 30 s, with the subsequent volume being the minimum necessary to keep the volunteers’ hands wet for the full 60 s. Thus the reference procedure comprised a range of volumes of 60% IPA, from 4 to 6 mL. Such a deviation in the reference procedure makes it difficult to ascertain whether such volumes of ABHR would be likely to pass a full EN 1500 (1997), let alone the updated version of EN 1500 (2013). Recent work by Bellissimo-Rodrigues et al. has suggested that small to medium volumes of ABHR (0.5e3.0 mL) produced significantly lower log10 RF values on volunteers with larger hands than on those with small or medium hands.14 Whereas our work did find a clear disparity in HSA between male and female volunteers, the full EN 1500 test we performed suggested that HSA was not an important factor in determining the log10 RF values obtained. This could be due to the fact that the range of hand sizes for our volunteers was smaller than for that studied by Bellissimo-Rodrigues et al.14 Additionally, we only measured the effect of HSA on the activity of 3 mL of ABHR; perhaps it would have more influence at lower volumes. This study has shown that volumes of ABHR that will dry in 20e30 s, the optimal time as perceived by the volunteers, are highly unlikely to fulfil the requirements of EN 1500. Although there was no significant difference between the reference

M.A.C. Wilkinson et al. / Journal of Hospital Infection 95 (2017) 175e182

7

Log10 reduction factor

6

5

4

3

Ref

WHOF1 WHOF2 Product

IPA

Figure 4. Reduction factors obtained by four products in EN 1500. Ref (23 mL, 60 s); WHOF1 (3 mL, 30 s); WHOF2 (3 mL, 30 s); IPA (3 mL, 30 s).

181

alcohol and the two WHO formulations when tested at 3 mL for a 30 s application, they did not pass the standard. EN 1500 is a useful standard for testing efficacy but is not useful for informing the recommended volume of application.13 A test using volumes and times that are acceptable to the user may comprise a useful addition/extension to the existing standard. This study demonstrates that it is not meaningful or relevant to link the performance of a given test product in EN 1500 to its recommended dosage or contact time. Nevertheless, it is clear that any such recommendations should be validated. Under the EU Biocides Regulation 528/2012, any bottle or cartridge with a pump is generally expected to provide a specific volume recommendation, expressed in number of pumps.23 To ensure compliance with the WHO guidelines, we suggest that product dose recommendations include a sufficient volume/number of applications to keep the hands wet for about 30 s.11 The use of such recommendations will help to prevent under-dosing when using ABHRs, the risk of which is very real, as demonstrated by Hines et al., who found that about 90% of healthcare worker hand hygiene events involved the application of <1 mL of ABHR.13 More research is required to determine the volume of hand rub and the duration of rubbing required to reduce the risk of transmission of infection. Hence, there is little clinical

7 6 Ref

5 4 3 7

WHOF1

6 4 3 7 6

WHOF2

Log10 reduction factor

5

5 4 3 7 6

IPA

5 4 3 300

350

400 Hand surface area (cm2)

450

Figure 5. Reduction factors obtained by product and hand surface area. Ref (23 mL, 60 s); WHOF1 (3 mL, 30 s); WHOF2 (3 mL, 30 s); IPA (3 mL, 30 s).

182

M.A.C. Wilkinson et al. / Journal of Hospital Infection 95 (2017) 175e182

evidence that can be used to determine dosage recommendations at present. In the meantime we propose that all products should be tested as described in EN 1500 to determine appropriate efficacy for the formulation under laboratory conditions, but this should not determine recommended dose or contact time. To establish this, products that fulfil the requirements of EN 1500 would then be subjected to a further test using the same methodology but with the manufacturer’s recommended volume and time of application (not less than 30 s), with the test product compared with the reference applied in the same manner. This additional test will establish that a given product that passes EN 1500 also performs at least as well as the reference product when compared at realistic contact times and volumes; surely an appropriate threshold for acceptability in the absence of any better? Other EN tests have additional test criteria and this could be one for EN 1500. Our study suggests that WHOF1 and WHOF2 would fulfil the requirements of this modified EN 1500, being non-inferior to the reference as assessed by the HodgeseLehmann test, when compared at the same volume and contact time (3 mL for 30 s). Conflict of interest statement K.O. and J.H. are employed by Deb Group Ltd. Funding sources Deb Group Ltd funded the materials to carry out this study.

References 1. World Health Organization. Guidelines on hand hygiene in health care. 2009. Available at: http://apps.who.int/iris/bitstream/ 10665/44102/1/9789241597906_eng.pdf [last accessed February 2016]. 2. Loveday HP, Wilson JA, Pratt RJ, et al. Epic 3: National evidencebased guidelines for preventing healthcare-associated infections in NHS hospitals in England. J Hosp Infect 2014;86 Suppl 1:S1eS70. 3. Rotter ML. Hygenic hand disinfection. Infect Control 1984;5:18e22. 4. Pittet D, Hugonnet S, Harbarth S, Mourouga P, et al. Effectiveness of a hospital-wide programme to improve compliance with hand hygiene. Infection Control Programme. Lancet 2000;356:1307e1312. 5. Sax H, Allegranzi B, Uc ¸kay I, Larson E, Boyce J, Pittet D. My five moments for hand hygiene: a user-centred design approach to understand, train, monitor and report hand hygiene. J Hosp Infect 2007;67:9e21. 6. World Health Organization. Guide to local production: WHO recommended hand rub formulations. 2010. http://www.who.int/gpsc/ 5may/Guide_to_Local_Production.pdf [last accessed February 2016]. 7. BS EN 1500: 2013. Chemical disinfectants and antiseptics. Hygienic handrub. Test method and requirements (phase 2/step 2).

8. ASTM International. E-1174-06. Standard test method for evaluation of the effectiveness of health care personnel or consumer handwash formulations. 2006. West Conshohocken, PA, ASTM International. 9. ASTM International. E-2755-15. Standard test method for determining the bacteria-eliminating effectiveness of healthcare personnel hand rub formulations using hands of adults. West Conshohocken, PA: ASTM International; 2015. 10. Federal Register/Vol. 80, No. 84/Friday, May 1st, 2015/Proposed Rules. 11. World Health Organization. Hand hygiene: why, how and when? 2009. Available at: http://www.who.int/gpsc/5may/Hand_Hygiene_Why_ How_and_When_Brochure.pdf [last accessed February 2016]. 12. Chassin MR, Mayer C, Nether K. Improving hand hygiene at eight hospitals in the United States by targeting specific causes of noncompliance. Jt Comm J Qual Patient Saf 2015;41:4e12. 13. Hines JD, Alper P, Eikelenboom-Boskamp A,Voss A, McGeer A. Product dose considerations for real-world hand sanitiser efficacy. Antimicrob Resist Infect Control 2013;2:P101. 14. Bellissimo-Rodrigues F, Soule H, Gayet-Ageron A, Martin Y, Pittet D. Should alcohol based handrub be customized to healthcare workers’ hand size? Infect Control Hosp Epidemiol 2016;37:219e221. 15. Hsu Y-W, Yu C-Y. Hand surface area estimation formula using 3D anthropometry. J Occup Environ Hyg 2010;7:633e639. 16. Arbogast JW, Okeke B, Brill F, et al. Advancement of the standard in vivo hand rub test methods: a critical comparison of the health care personnel handwash (ASTM E1174) and the hygiene handrub (EN1500). BMC Proc 2011;5(Suppl 6):P269. 17. Kampf G, Ruselack S, Eggerstedt S, Nowak N, Bashir M. Less and less e influence of volume on hand coverage and bactericidal efficacy in hand disinfection. BMC Infect Dis 2013;13:472. 18. Macinga DR, Shumaker DJ, Werner HP, et al. The relative influences of product volume, delivery format and alcohol concentration on dry-time and efficacy of alcohol-based hand rubs. BMC Infect Dis 2014;14:511. 19. Suchomel M, Kundi M, Pittet D, Weinlich M, Rotter ML. Testing of the World Health Organization recommended formulations in their application as hygienic hand rubs and proposals for increased efficacy. Am J Infect Control 2012;40:328e331. 20. Edmonds SL, Macinga DR, Mays-Suko P, et al. Comparative efficacy of commercially available alcohol-based hand rubs and World Health Organization-recommended hand rubs: formulation matters. Am J Infect Control 2012;40:521e525. 21. BS EN 1500: 1997. Chemical disinfectants and antiseptics. Hygienic handrub. Test method and requirements (phase 2/step 2). 22. Goroncy-Bermes P, Koburger T, Meyer B. Impact of the amount of hand rub applied in hygienic hand disinfection on the reduction of microbial counts on hands. J Hosp Infect 2010;74:212e218. 23. Health and Safety Executive. Packaging and labelling requirement for biocidal products. http://www.hse.gov.uk/biocides/eu-bpr/ packaging-labelling-requirements.htm [last accessed July 2016].