With the exception of non-medicated soaps, every new formulation for hand antisepsis should be tested for its antimicrobial efficacy to demonstrate that: (i) it has superior efficacy over normal soap; or (ii) it meets an agreed performance standard. The formulation with all its ingredients should be evaluated to ensure that humectants or rehydrating chemicals added to ensure better skin tolerance do not in any way compromise its antimicrobial action.
Many test methods are currently available for this purpose, but some are more useful and relevant than others. For example, determination of the minimum inhibitory concentration (MIC) of such formulations against bacteria has no direct bearing on the “killing effect” expected of such products in the field. Conditions in suspension and in vitro192 or ex vivo193 testing do not reflect those on human skin. Even simulated-use tests with subjects are considered by some as “too controlled”, prompting testing under in praxi or field conditions. Such field testing is difficult to control for extraneous influences. In addition, and importantly, the findings of field tests provide scant data on a given formulation’s ability to cause a measurable reduction in hand-transmitted nosocomial infections. While the ultimate approach in this context would be clinical trials, they are generally quite cumbersome and expensive. For instance, power analysis reveals that for demonstrating a reduction in hand-transmitted infections from 2% to 1% by changing to a presumably better hand antiseptic agent, almost 2500 subjects would be required in each of two experimental arms at the statistical pre-settings of α (unidirectional) = 0.05 and a power of 1-β = 0.9.194 For this reason, the number of such trials remains quite limited.195–197 To achieve a reduction from 7% to 5% would require 3100 subjects per arm. This reinforces the utility of well-controlled, economically affordable, in vivo laboratory-based tests to provide sufficient data to assess a given formulation’s potential benefits under field use.
10.1. Current methods
Direct comparisons of the results of in vivo efficacy testing of handwashing, antiseptic handwash, antiseptic handrub, and surgical hand antisepsis are not possible because of wide variations in test protocols. Such variations include: (i) whether hands are purposely contaminated with a test organism before use of the test agent; (ii) the method used to contaminate fingers or hands; (iii) the volume of hand hygiene product applied; (iv) the time the product is in contact with the skin; and (v) the method used to recover the organism from the skin after the test formulation has been used.
Despite the differences noted above, most testing falls into one of two major categories. One category is designed to evaluate handwash or handrub agents to eliminate transient pathogens from HCWs’ hands. In most such studies, the subjects’ hands are experimentally contaminated with the test organism before applying the test formulation. In the second category, which applies to pre-surgical scrubs, the objective is to evaluate the test formulation for its ability to reduce the release of naturally present resident flora from the hands. The basic experimental design of these methods is summarized below and the procedures are presented in detail in .
Basic experimental design of current methods to test the efficacy of hand hygiene and surgical hand preparation formulations.
In Europe, the most commonly used methods to test hand antiseptics are those of the European Committee for Standardization (CEN). In the USA and Canada, such formulations are regulated by the Food and Drug Administration (FDA)198 and Health Canada, respectively, which refer to the standards of ASTM International (formerly, the American Society for Testing and Materials).
It should be noted that the current group of experts recommends using the term “efficacy” to refer to the (possible) effect of the application of a hand hygiene formulation when tested in laboratory or in vivo situations. By contrast, it would recommend using the term “effectiveness” to refer to the clinical conditions under which hand hygiene products have been tested, such as field trials, where the impact of a hand hygiene formulation is monitored on the rates of cross-transmission of infection or resistance.199
10.1.1. Methods to test activity of hygienic handwash and handrub agents
The following in vivo methods use experimental contamination to test the capacity of a formulation to reduce the level of transient microflora on the hands without regard to the resident flora. The formulations to be tested are hand antiseptic agents intended for use by HCWs, except in the surgical area.
CEN standards: EN 1499 and EN 1500
In Europe, the most common methods for testing hygienic hand antiseptic agents are EN 1499200 and EN 1500.201 Briefly, the former standard requires 12–15 subjects, and the latter (in the forthcoming amendment) 18–22, and a culture of E. coli. Subjects are assigned randomly to two groups where one applies the test formulation and the other a standardized reference solution. In a consecutive run, the two groups reverse roles (cross-over design).
If an antiseptic soap has been tested according to EN 1499,200 the mean log10 reduction by the formulation must be significantly higher than that obtained with the control (soft soap). For handrubs (EN 1500), the mean acceptable reduction with a test formulation shall not be significantly inferior to that with the reference alcohol-based handrub (isopropyl alcohol or isopropanol 60% volume).
ASTM standards
Currently, handwash or handrub agents are evaluated using this method in North America. The efficacy criteria of the FDA’s Tentative Final Monograph (TFM) are a 2-log
10 reduction of the indicator organism on each hand within 5 minutes after the first use, and a 3-log
10 reduction of the indicator organism on each hand within 5 minutes after the tenth use.
198The performance criteria in EN 1500 and in the TFM for alcohol-based handrubs are not the same.
48,198,201 Therefore, a formulation may pass the TFM criterion, but may not meet that of EN 1500 or vice versa.
203 It should be emphasized here that the level of reduction in microbial counts needed to produce a meaningful drop in the hand-borne spread of nosocomial pathogens remains unknown.
48,204ASTM E-1838 (fingerpad method for viruses)205 The fingerpad method can be applied with equal ease to handwash or handrub agents. When testing handwash agents, it can also measure reductions in the levels of viable virus after exposure to the test formulation alone, after post-treatment water rinsing and post-rinse drying of hands. This method also presents a lower risk to subjects because it entails contamination of smaller and well-defined areas on the skin in contrast to using whole hands (see below). The method can be applied to traditional as well as more recently discovered viruses such as caliciviruses.
206ASTM E-2276 (fingerpad method for bacteria)207 This method is for testing handwash or handrub against bacteria. It is similar in design and application to the method E-1838
205 described above for working with viruses.
ASTM E-2613 (fingerpad method for fungi)208 This method is for testing handwash or handrub against fungi. It is similar in design and application to the methods described above for working with viruses (E-1838)
205 and bacteria (E-2276).
207ASTM E-2011 (whole hand method for viruses)209 In this method, the entire surface of both hands is contaminated with the test virus, and the test handwash or handrub formulation is rubbed on them. The surface of both hands is eluted and the eluates assayed for viable virus.
10.1.2. Surgical hand preparation
In contrast to hygienic handwash or handrub, surgical hand preparation is directed against the resident hand flora. No experimental contamination of hands is used in any existing methods.
CEN standard: EN 12791 (surgical hand preparation)210
This European norm is comparable with that described in EN 1500, except that the bactericidal effect of a product is tested: (i) on clean, not experimentally contaminated hands; (ii) with 18–20 subjects; (iii) using the split-hands model by Michaud, McGrath & Goss211 to assess the immediate effect on one hand and a 3-hour effect (to detect a possible sustained effect) on the other, meanwhile gloved hand; (iv) in addition, a cross-over design is used but, contrary to hygienic hand antisepsis, the two experimental runs are separated by one week to enable regrowth of the resident flora; (v) the reference antisepsis procedure uses as many 3-ml portions of n-propanol 60% (v/v) as are necessary to keep hands wet for 3 minutes; thus, the total quantity used may vary according to the size and temperature of the hands and other factors; (vi) the product is used according to manufacturer’s instructions with a maximum allowed contact time of 5 minutes; (vii) the requirements are that the immediate and 3-hour effects of a product must not be significantly inferior to those of the reference hand antisepsis; and (viii) if there is a claim for sustained activity, the product must demonstrate a significantly lower bacterial count than the reference at 3 hours.
ASTM standard: ASTM E-1115 (surgical hand scrub)212
This test method is designed to measure the reduction in bacterial flora on the skin. It is intended for determining immediate and persistent microbial reductions, after single or repetitive treatments, or both. It may also be used to measure cumulative antimicrobial activity after repetitive treatments.
In North America, this method is required to assess the activity of surgical scrubs.198 The TFM requires that formulations: (i) reduce the number of bacteria 1-log10 on each hand within 1 minute of product use and that the bacterial colony count on each hand does not subsequently exceed baseline within 6 hours on day 1; (ii) produce a 2-log10 reduction in bacterial counts on each hand within 1 minute of product use by the end of the second day of enumeration; and (iii) accomplish a 3-log10 reduction of bacterial counts on each hand within 1 minute of product use by the end of the fifth day when compared to the established baseline.198
10.2. Shortcomings of traditional test methods
10.2.1. Hygienic handwash and handrub; HCW handwash and handrub
A major obstacle for testing hand hygiene products to meet regulatory requirements is the cost, which can be prohibitive even for large multinational companies. Cases in point are the extensive and varied evaluations as specified in the TFM198; time-kill curves must also be established along with tests on the potential for development of antimicrobial resistance. In vivo, at least 54 subjects are necessary in each arm to test the product and a positive control, hence a minimum of 2 × 54 subjects. The immense expenditure would, however, be much smaller if the same subjects were used to test both formulations concurrently in two runs in a cross-over fashion as described in EN 1499 and EN 1500.200,201 The results could then be intra-individually compared, thus allowing a considerable reduction in sample size at the same statistical power.
Another shortcoming of existing test methods is the duration of hand treatments that require subjects to treat their hands with the hand hygiene product or a positive control for 30 seconds198 or 1 minute,200 despite the fact that the average duration of hand cleansing by HCWs has been observed to be less than 15 seconds in most studies.124,213–218 A few investigators have used 15-second handwashing or hygienic hand antisepsis protocols.151,219–222 Therefore, almost no data exist regarding the efficacy of antimicrobial soaps under conditions in which they are actually used. Similarly, some accepted methods for evaluating waterless antiseptic agents for use as antiseptic handrubs, such as the reference hand antisepsis in EN 1500,201 require that 3 ml of alcohol be rubbed into the hands for 30 seconds, followed by a repeat application of the same type. Again, this type of protocol does not reflect actual usage patterns among HCWs. However, it could be argued that non-inferiority in the efficacy of a test product as compared with the reference is easier to prove with longer skin contact. Or, inversely, to prove a difference between two treatments of very short duration, such as 15 seconds, under valid statistical settings is difficult and requires large sample sizes, i.e. large numbers of subjects. Therefore a reference treatment, which has usually been chosen for its comparatively high efficacy, may include longer skin contact than is usual in real practice. By this, the non-inferiority of a test product can be demonstrated with economically justifiable sample sizes.
The TFM,198 for instance, requires that a handwash to be used by HCWs demonstrates an in vivo reduction in the number of the indicator organisms on each hand by 2 log within 5 minutes after the first wash and by 3 log after the tenth wash. This requirement is inappropriate to the needs of working in a health-care setting for two reasons. First, to allow a preparation to reduce the bacterial release by only 2 log within a maximum time span of 5 minutes seems an unrealistically low requirement, as even with unmedicated soap and water a reduction of 3 log is achievable within 1 minute.48,223 Furthermore, 5 minutes is much too long to wait between two patients. Second, the necessity for residual action of a hand antisepsis formulation in the non-surgical area has been challenged.224–226 The current group of experts does not believe that for the aforementioned purpose a residual antimicrobial activity is necessary in the health-care setting. Rather, a fast and strong immediate effect against a broad spectrum of transient flora is required to render hands safe, not only in a very short time, but also already after the first application of the formulation. Therefore, the requirement that a product must demonstrate a stronger activity after the tenth wash than after the first seems difficult to justify.
An in-use test that is simple to use in the clinical setting to document microbial colonization is the fingerprint imprint method.72 This method entails taking imprints of the fingerpads and thumb on to a nutritive agar preferably containing neutralizers for the non-alcohol-based antiseptic agent in use. This is done by applying gentle pressure with the fingers and thumb individually on to the agar for 5 seconds. This method provides less accurate bacterial counts than the fingertip rinse method, but it has the advantage of ease of use in the field and provides good results when evaluating transient flora and their inactivation. The problem with such a qualitative method is that it often gives confounding results. Indeed, the bacterial count recovered after the use of the test formulation can be much higher than the one in controls because of the disaggregation of micro-colonies of resident bacteria.
10.2.2. Surgical handwash and handrub; surgical hand scrub; surgical hand preparation
As with hygienic hand antisepsis, a major shortcoming for testing surgical scrubs is the resource expenditure associated with the use of the TFM model. The required in vitro tests are the same as described under Part I, Section 10.2.1, above (see also ) No less than 130 subjects are necessary to test a product, together with an active control in the suggested parallel arm design. For some products, this number will even have to be multiplied for concomitant testing of the vehicle and perhaps of a placebo to demonstrate efficacy.198 As mentioned with the test model for HCW handwashes and described in EN 12791,210 this large number of subjects could be much reduced if the tests are not conducted with different populations of subjects for each arm but if the same individuals participate in each arm, being randomly allocated to the various components of a Latin square design, the experiments of which can be carried out at weekly intervals. The results are then treated as related samples with intra-individual comparison. Additionally, it is not clear why the vehicle or a placebo needs to be tested in parallel if a product is shown to be equivalent in its antimicrobial efficacy to an active control scrub. For the patient and for the surgeon, it is of no interest whether the product is sufficiently efficacious because of the active ingredient only or, perhaps, additionally by a synergistic or even antimicrobial effect of the vehicle.
In contrast to the requirement of EN 12791 where a sustained (or persistent) effect of the surgical scrub is optional, the TFM model requires a formulation to possess this feature (see above). However, the continued presence of a microbicidal chemical to produce a sustained effect may be unnecessary in view of the fact that volatile ingredients such as short-chain aliphatic alcohols (e.g. ethanol, iso-propanol, and n-propanol)48 appear fully capable of producing the same effect.227 With their strong antibacterial efficacy, the importance of a sustained effect is questionable, as regrowth of the skin flora takes several hours even without the explicitly sustained effect of the alcohols. Furthermore, whether a long-term effect (several days), such as recommended in the TFM model, is necessary or not remains a matter for discussion. It is, however, difficult to understand why the efficacy of a scrub is required to increase from the first to the fifth day of permanent use. Ethical considerations would suggest that the first patient on a Monday, when the required immediate bacterial reduction from baseline is only 1 log, should be treated under the same safety precautions as patients operated on the following Friday when, according to the TFM requirement, the log reduction has to be 3.0.
With regard to the statistical analysis of EN 12791, in which the efficacy of a product is compared with that of a reference (including a handrub with 60% n-propanol for 3 minutes), the currently suggested model of a comparative trial is no longer up to date. It should be exchanged for a non-inferiority trial. Furthermore, the latest CDC/HICPAC guideline for hand hygiene in health-care settings58 considers it as a shortcoming that in vivo laboratory test models use non-HCWs as surrogates for HCWs, as their hand flora may not reflect that on the hands of caregivers working in health-care settings. This argument is only valid for testing surgical scrubs, however, because protocols for evaluating hygienic handwash or rub preparations include experimental hand contamination. Besides, the antimicrobial spectrum of a product should be known from the results of preceding in vitro tests.
10.3. The need for better methods
Further studies will be needed to identify necessary amendments to the existing test methods and to evaluate amended protocols, to devise standardized protocols for obtaining more realistic views of microbial colonization, and to better estimate the risk of pathogen transfer and cross-transmission.72
To summarize, the following amendments to traditional test methods are needed.
The few existing protocols should be adapted so that they lead to comparable conclusions about the efficacy of hand hygiene products.
Protocols should be updated so that they can be performed with economically justifiable expenditure.
To be plausible, results of in vivo test models should show that they are realistic under practical conditions such as the duration of application, the choice of test organism, or the use of subjects.
Requirements for efficacy should not be formulated with a view to the efficacy of products available on the market, but in consideration of objectively identified needs.
In vivo studies in the laboratory on surgical hand preparation should be designed as clinical studies, i.e. to determine equivalence (non-inferiority) rather than comparative efficacy.
Protocols for controlled field trials should help to ensure that hand hygiene products are evaluated under more plausible, if not more realistic, conditions.
In addition, tests on the antimicrobial efficacy of hand hygiene products should be conducted in parallel with studies on the impact (effectiveness) of their use on cross-transmission of infection or resistance. Indeed, there is no doubt that results from well-controlled clinical studies are urgently needed to generate epidemiological data on the benefits of various groups of hand hygiene products on reducing the spread of HCAI, i.e. a more direct proof of clinical effectiveness.
