Methods to assess antibacterial, antifungal and antiviral surfaces in relation to touch and droplet transfer: a review, gap-analysis and suggested approaches

Abstract To help assess whether a potentially antimicrobial material, surface, or coating provides antimicrobial efficacy, a number of standardised test methods have been developed internationally. Ideally, these methods should generate data that supports the materials efficacy when deployed in the intended end-use application. These methods can be categorised based on their methodological approach such as suspension tests, agar plate/zone diffusion tests, surface inoculation tests, surface growth tests or surface adhesion tests. To support those interested in antimicrobial coating efficacy, this review brings together an exhaustive list of methods (for porous and non-porous materials), exploring the methodological and environmental parameters used to quantify antibacterial, antifungal, or antiviral activity. This analysis demonstrates that antimicrobial efficacy methods that test either fungi or viruses are generally lacking, whilst methods that test bacteria, fungi and viruses are not designed to simulate end-use/lack realistic conditions. As such, a number of applications for antimicrobial activity across medical touch screens, medical textiles and gloves and transport seat textiles are explored as example applications, providing guidance on modifications to existing methods that may better simulate the intended end-use of antimicrobial materials.


INTRODUCTION
The use of antimicrobial surfaces and coatings to prevent the transfer (and potential subsequent infection [1]) of microorganisms is broad and traverses numerous applications across the built environment [1], hospitals [2], public transport [3] and high-touch devices such as mobile phones [4].To ensure that those involved in either the production, procurement, regulation or end-use of these materials can make appropriate, evidence-based decisions, it is essential that antimicrobial surfaces and coatings are assessed using methodology that provides robust, reproducible data that reflects the efficacy intended in use [5].
To assess whether a potential antimicrobial material, surface or coating provides antimicrobial efficacy, whether that be via biocide-release, contact activity, or reduced-adhesion [6], a number of standardised test methods (STMs) have been developed internationally.Ideally, these methods should generate data that supports the materials efficacy when deployed in the intended end-use application.At the very least, they should produce reproducible data that can enable efficacy performance comparisons when these data are generated in different laboratories (e.g.undertaking a ring trial).
There is a range of different methodological approaches for testing antimicrobial efficacy described in the literature, usually framed on specific materials (e.g.ceramics [7], plastics [8], carpet fibres [9]) or antimicrobial action (e.g.silver-ion release [10], copper [11], UV irradiation [12]).In almost every instance the focus has been on optimising the function of the antimicrobial effect under laboratory/controlled conditions, rather than simulating the environment and pattern of use that might prevail when that material is placed into service.Such an approach is useful during initial research, whether that be into the active substance/ mechanism or during studies on the compatibility with a final product or durability assessment.However, understanding the impact of end-use environmental conditions is important, and using a method that does not account for this may lead to inherent bias in data [13], as the antimicrobial activity demonstrated in the laboratory may well fail to be realised in practice [14].The analysis described in this review aims to highlight the gap between existing standard test methods for assessing antimicrobial efficacy of a material and model end-use environments and suggest potential areas for method development using a number of case study applications.

METHOD
A search was undertaken to collate all existing standards and established test methods relating to antimicrobial material efficacy testing by accessing standards repositories at BSOL (British Standards Online), AATCC (American Association for Textile Chemists and Colourists) and ASTM (American Society for Testing Materials).In select cases, other standards were added if they were deemed to be relevant (e.g.Japanese Industrial Standards [JIS] and ENV/JM/MONO(2007)17). Additionally, appropriate guidance documents from OECD were also included.A standard was excluded if it was assessing only the growth of microorganisms on the surface.
All relevant standards were sorted by whether they were assessing porous or non-porous surfaces (ceramics were deemed to be non-porous).They were then sorted by which category of STM was most relevant according to pre-determined definitions (Table 1).Various data (substrate type, temperature, relative humidity, incubation period and organisms used) were then extracted for each standard.
Three example end-use cases were formed based on some of the most likely scenarios for the implementation of antimicrobial materials.Whilst there are many other examples that could be described, those included in this document present a range of material types (porous and non-porous), a range of criticality (hospital wards through to mass-transport) and different contamination events (droplet transmission, direct touch transfer).In each case, an example of current practices based on the Organisation for Economic Co-operation and Development's (OECD [15]) guidance are compared to environmental conditions and methodological decisions that would be considered more realistic in an average setting for the end-use case.

Overview of existing standards -porous surfaces
Twenty two standards relating to porous surfaces were identified (Table 2), four of which were category one, five were category two and thirteen were category three.There were no methods that were in category four or five.The majority of the standards tested against bacteria (n=15), with some testing against fungi (n=7) and one method to test viruses.In one method both bacteria and fungi are tested, and all were using some variety of fabric or textile.Eleven standards used incubation temperature values between Table 1.Descriptions that can be used to categorise standardised methods based on the intended effect of an antimicrobial action or the end-use of the treated material

Category no. and name Description
Category I -suspension tests The material to be tested is immersed in a liquid containing the test species.The objective is to observe a reduction in the size of the population in the suspension.

Category II -agar plate / zone diffusion tests
The test material is placed into contact with a semi-solid growth medium that has been inoculated with the test species.The objective is to observe an effect on the growth of the organisms on the solid media (or on the test specimen).

Category III -surface inoculation tests
The test species is suspended in a liquid and then placed onto the test material.The objective is to observe a reduction in the size of the population recovered from the treated samples (often compared with no, or a smaller, reduction on the untreated ones).

Category IV -surface growth tests
The material to be tested is inoculated with a population of relevance to the material / application (either as single species or as a consortium).The inoculated samples are then incubated under conditions that encourage the growth of the organisms on the surface (either in growth chambers, flow cells or biofilm reactors).The objective is to observe the inhibition of growth on the treated sample when compared with the growth on untreated ones.

Category V -surface adhesion tests
The material to be tested is inoculated with a population of relevance to the material / application (either as single species or as a consortium).The samples are then incubated and processed to examine whether the treatment has an effect on the adhesion of the organisms to the surface (e.g. by direct microscopic examination, atomic force microscopy, etc.).Key to species and abbreviations used in Tables 1 and 2 Table

Overview of existing standards -non-porous surfaces
Eighteen relevant standards relating to non-porous surfaces were identified (Table 3), three of which were category one and fifteen were category three.There were no methods that were in category two, four, or five.The majority of the standards tested against bacteria (n=14), with some testing against viruses (n=3) and one method to test fungi.Seven standards specified a temperature of 35-37 °C, while ten standards specify 20-25 °C and the remaining standard being unspecified.The relative humidity is not stated in seven standards and is irrelevant in two standards as the materials are submerged.Of the remaining standards, six specify a high relative humidity of above 75 %, while three standards specify between 30-70 %.Seven standards specified an incubation period of 24-48 h, with the remaining eleven standards specifying less than 24 h.Ten of the bacterial standards included Escherichia coli, with other standards opting for S. aureus or Bacillus subtilis (although most standards accommodate multiple bacterial species).
A. niger was used for the fungal standard and influenza A, feline calicivirus or bacteriophage Q-beta was used for the viral standards.

Overview of existing standards -discussion
Whilst there is a relatively large number of standardised methods available for antimicrobial coating assessment, a number of methods can be seen as 'competitive' as they appear to address the same material / effect.In some cases, the methods can be simply   substituted for each other as they are either essentially identical (e.g.JIS Z 2801 [16] and ISO 22196 [17]) or they are capable of providing a similar amount of information with regards to the basic antimicrobial activity of a certain material.Conversely, some are very specific to a certain type of material or antimicrobial mechanism (e.g.ISO 27447 [18]).Whilst some element of global harmonisation towards method development exists, the development of STMs can operate on a regional/national basis (e.g.ASTM methods in the USA [19], NSF Norms in France [20]), the presence of certain trade organisations (e.g.AATCC in the USA [21], representing the textile and carpeting industries), or the need for pass criteria to satisfy certain brand-marks (e.g. the Kohkin brand mark associated with the 'pass' level in JIS Z 2801, or the criteria with [22,22] in Japan).In some cases, certain active substance producers have encouraged the development of standard methods that work well with their technology (presumably in the hope of gaining a competitive advantage in the market).A number of these regional methods have been converted into international norms (often stripped of their arbitrary pass / fail criteria) such as JIS Z 2801 ->ISO 22196, and [22] ->ISO 20743 [23] (combined with part of a French national standard).In some cases, the standards have been normalised further such as in the OECD tier one method for treated articles which presents a base method for both non-porous and porous materials through the harmonisation of the parameters used in ISO 22196 and the absorption method described in ISO 20743.
As described in Tables 2 and 3, there are numerous STMs available to examine the basic antimicrobial properties of treated materials, coatings, textiles, etc.Some can even be used to simulate actual exposure conditions (e.g. a flow cell biofilm method may be capable of accurately simulating the conditions present in pipelines or catheters).However, in most cases the methods only look at basic antimicrobial properties and do not simulate end-use scenarios sufficiently well to be capable of predicting performance in practice nor in supporting any claims made during product / material registration.Some approaches have been described that start to address this, but only in guideline form (OECD Guidance / ECHA Guidance / Nordic Council Guidance).
As such, taking an application or environment and considering in detail how an antimicrobial test method may be developed, and what the environmental parameters may be is an essential step forward for those interested in antimicrobial coatings.

Example end-use scenario one: touch screen of a heart monitor in ICU
Most touch screens are constructed from materials that are essentially non-porous [24].In a clinical setting, they will be subject to low to moderate interaction and although human skin contact will occur, in many cases the operative will probably be wearing a disposable / surgical glove [25].As such, whilst some skin flora may be transferred to the touch screen, most of the microbial contamination delivered by touch is likely to be transferred from other surfaces via a glove [26].Deposition of microorganisms from the air can occur and it is possible that droplets (respiratory, etc.) could also be deposited on the touch screen [27].Most of the transfer / deposition will either be dry or be associated with very transient wetness.The environmental conditions within an ICU are likely to be constant [28][29][30] and although there may be some air movement from ventilation systems, they will be of low velocity [31].Temperature may range from 16 °C to 25 °C and relative humidity likely between 30 and 60 % [32].
As described in Table 3, there is no single STM that has exposure conditions that match those anticipated from the touch screen.ISO 22196 is intended for use with non-porous materials, but the exposure conditions require the full hydration of the surface of the material, and even if the temperature and contact time were aligned to those of a hospital ICU, it would still be a poor model due to the volume of liquid applied.In contrast, NSF S90-700 uses small droplets as an inoculum and so might be suited for simulating aerosol / droplet deposition but has some arbitrary time to dryness requirements and has no 'dry' contact component.
A method which simulates hand contact (Fig. 1) would likely present an extreme worst-case as a much larger number of organisms will be transferred than is likely to occur in an ICU [33].A further modification in which an intermediate, untreated surface is employed may provide a more representative model.This surface would be contaminated by either using a splash method (e.g.Fig. 2) or by applying a wet inoculum to the surface and allowing it to dry and using the resulting deposit as the inoculum to pick up using the transfer device.However, this would require significant additional work (parameters are described in Tables 4, 5 and 6).
This method could also be applied to determine the antiviral activity of a surface in a similar manner to bridge a further gap in current methodologies (Table 6).

Example end-use scenario two: medical uniforms (scrubs)
Medical uniforms are commonly produced from polyester / cotton blends [34].They are normally porous and have a high moisture holding capacity.Over the course of a shift (8-12 h) it is likely that a medical professional would come into close contact with tens if not hundreds of patients, visitors, and colleagues where the potential for a contaminating event is high [35].Additionally, medical uniforms exist between two distinct environments (i) that of the wearer (body temperature, sweat/humidity) and (ii) the physical environment within which the wearer is placed (with different temperatures, humidity, etc.).All these factors are going to change how contaminating microorganisms interact with an antimicrobial textile, which should be considered when thinking about the efficacy assessment of antimicrobial activity [36].As the wearer is only likely to be wearing the uniform for one shift at a time, for example between 8-12 h, antimicrobial activity that takes longer than this may not be beneficial.A droplet deposition method would be well suited to simulate a contamination event rather than a fully wet test.If a gross contamination event occurred resulting in a wet uniform for a prolonged period, it is likely that the garment would be laundered [37].An antimicrobial effect may still be advantageous under these fully wet conditions to reduce the risk to the laundry staff who are involved in the laundering process.This activity could be determined by employing methods such as: IBRG TA22-004 [38], ISO 20743 [23], AATCC 100 [22,22,39] and the OECD method (ENV/JM/MONO (2014)18 [15]).However, in each case the method does not provide realistic environmental conditions, and as such, modifications may be required (Tables 7, 8 and 9).
Using a common method such as ISO 20743 would enable reproducible efficacy assessment, but as described above, the conditions the method requires would not be comparable to a uniform in a medical setting.In this scenario, various modifications can be suggested (Table 7).Target species can be selected to better simulate those considered important; nosocomial pathogens [40], suspended in more complex media (e.g.artificial blood [41], urine [42] etc.).Additionally, the temperature can be lowered to be more realistic of a 'warm day' which will accelerate drying onto the material [43].Assessment of the antimicrobial activity of  fabric following ageing [44] will also need to be determined to ensure activity throughout the lifetime of the garment is achieved.Current standards such as AATCC 61 [45], which is a method that employs accelerated laundering to determine the durability of a textile that is expected to undergo frequent laundering, could be employed to age the textile prior to efficacy studies.
Like the droplet deposition method described for touch screen applications in the medical setting, the OECD method can be used to consider the same droplet contaminating event on a medical uniform (Table 8).Like the modifications described in Table 7, selecting a temperature to accelerate drying of droplets and selecting a more complex suspending medium may be suitable.Additionally, reducing the volume of inoculum may better simulate aerosol deposition of droplets [46].Due to the intended use-time of a medical uniform, contact time can be reduced to a maximum of 8 h.
The nature of the environment means that those wearing medical uniforms will inevitably contact surfaces that are contaminated with pathogens.For example, this may occur when a medical professional is providing care to a patient in a bed, where the uniform may come in direct contact with both the bed linen (textile [47]) and bed rails (non-porous materials [48]).In either of these scenarios, a medical uniform that can actively kill contaminating microorganisms would be beneficial.However, despite both contamination events potentially co-occurring within a physically close space and time, the transfer of microorganisms from textile-to-textile and non-porous material-to-textile is different, and as such, efficacy assessment should consider these scenarios within the methodology (Table 9).The ISO 20743 standard defines three test methods, the absorption method is a fully wet method, the transfer method removes cells from an agar plate and transfers them to the test surface in a similar manner to the OECD hand contact simulation method, and a printing method where the bacterial cells are placed on a filter paper and then transferred onto the test fabric by printing using a defined weight.This method has the least moisture and nutrients transferred along with the inoculum.

Fig. 1 .
Fig. 1.A hand contact simulation protocol designed to assess the efficacy of non-porous antimicrobial surfaces.

Fig. 2 .
Fig. 2. A simulated splash protocol designed to assess the efficacy of non-porous antimicrobial surfaces.

Table 2 .
Overview of STMs relating to antimicrobial efficacy of porous surfaces and coatings Continued35-37 °C, six standards specify a temperature of 27-30 °C and the remaining five standards specify a temperature of 20-25 °C.The relative humidity was either not stated (five standards), was submerged (i.e.relative humidity was irrelevant, three standards), or was at a high relative humidity of above 90 % (humid chamber stated in eight standards, unspecified in three standards).Three standards stated a lower relative humidity of >70 %.Most standards stated an incubation period of 24-48 h (16 standards), while two standards specified greater than 48 h and four standards stated less than 24 h.Finally, Staphylococcus aureus was specified in all bacterial standards among others, and Aspergillus niger was equivalently specified in all fungal standards among other species.The viral standard specified influenza A or feline calicivirus.

Table 3 .
Overview of STMs relating to antimicrobial efficacy of non-porous surfaces and coatings Category relates to the five categories of the test method described in section '1.2 Standardised Testing Methods'. *

Table 4 .
Comparison of approaches regarding the use of antimicrobial coatings for touch-screens via droplet deposition

Table 5 .
Contamination of touch-screens via hand deposition of bacteria Environmental 20 °C and 50 % RH 20-24 °C and 50 % RH Use 24 °C as this will accelerate drying and so present the worst-case for a technology to function under and 50 % RH.

Table 6 .
Contamination of touch-screens via hand deposition of viruses