Reduction of the Microbiological Load on Hospital Surfaces Through Probiotic-Based Cleaning Procedures: A New Strategy to Control Nosocomial Infections

Alberta Vandini1, Alessia Frabetti1, Paola Antonioli2, Daniela Platano3, Alessio Branchini4, Maria Teresa Camerada1, Luca Lanzoni1, Piergiorgio Balboni1 and Sante Mazzacane1* 1CIAS Laboratory, Centre for the Study of Physical, Chemical and Microbiological Contamination of Highly Sterile Environments, Department of Architecture, University of Ferrara, Italy 2Infection Prevention Control and Risk Management Department, Sant’Anna University Hospital of Ferrara, Italy 3Department of Biomedical and Neuromotor Sciences, University of Bologna, Italy 4Department of Life Sciences and Biotechnology, University of Ferrara, Italy


Introduction
Healthcare-Associated Infections (HAIs) are the most frequent complications occurring in healthcare facilities. It has been estimated that about 5% to 15% of all hospitalized patients develop a HAI during hospitalization [1]. A three pointprevalence survey conducted in Italy showed a frequency of 6.7% HAI [2], with the most frequent HAI represented by lower respiratory tract infections, followed by urinary tract infections. In 2012, the European Centre for Disease Prevention and Control (ECDC) published European report on HAI frequency [3], stating that every year over 3 million patients contract a HAI in the European Union. The high incidence of this severe complication is responsible for a huge cost on the society and makes the abatement of HAIs a priority in healthcare industry. A great number of studies and surveys based on the screening of patients have been reported, but the actual sources of contamination have been poorly characterized. In this view, the contribution of environment and surfaces to HAI development is still a matter of discussion. Nevertheless, it is well known that surfaces act as reservoirs for microorganisms [4], and represent a potential risk of cross-contamination for patients, caused by direct or indirect contact of patients with such potentially contaminated surfaces. To reduce these risks, in healthcare facilities sanitation procedures are applied to each surface that directly or indirectly may come in contact with people and patients. Despite the existence of experimental evidence suggesting that a reasonable, strategic use of disinfectants is recommended, the routine use of these agents remains still a matter of debate [5][6][7]. Notwithstanding, a proper surface disinfection is highly recommended by all international guidelines as a primary procedure for preventing infections [8][9][10]. However, the widespread application of chemical disinfectants may account for several risks both for the environment and for the safety of users. Since microorganisms can adapt to a variety of environmental physical and chemical conditions, it is therefore not surprising that acquired resistance to extensively used antiseptics and disinfectants, together with the ever increasing antimicrobial resistance, has been reported in vitro [11][12][13][14]. For these reasons, there is increasing interest about the improvement of efficient, though sustainable, sanitation methodologies capable of containing or limiting the proliferation of pathogenic microorganisms. The importance of this issue is also outlined by recent research aimed to evaluate the antibacterial efficacy of natural detergents/biodetergents as a potential alternative to conventional disinfectants [15,16]. A very promising strategy, as previously suggested by Falagas and Makris [17], is represented by the use of non pathogenic microbiological products to colonize surfaces in order to counteract the proliferation of other bacterial species, in accordance with the competitive exclusion principle (Gause's law) [18]. This approach is based on the wellestablished adoption of probiotics as food supplements, referred as live microorganisms which, when administered in adequate amounts, confer a health benefit on the host. The problem of hospital hygiene is thus completely reversed, since the aim of the procedure is no more represented by a general disinfection, which minimizes the presence of any kind of microorganisms on hospital surfaces, but to counteract the development of potentially pathogenic strains, instead tolerating the presence of microorganisms that are not harmful to humans.
The broadening of the probiotic concept towards the environment has been designated as biocontrol when the application is antagonistic towards a certain pathogen [19]. This strategy has been already successfully applied for the abatement of Legionella in water systems [20], but there is no evidence about its effective application on surfaces by means of cleaning. Accordingly, such bio-stabilization through cleaning procedures may introduce a new conception of cleaning systems targeted to the establishment of a controlled and less harmful microbiota instead of a generic removal of microorganisms from the environment.
The aim of this study was to provide an experimental evaluation of the efficacy of a probiotic-based solution containing spore forms of Bacillus spp in comparison with traditional chlorine-based chemical disinfectants. As a study-model environment we chose the Sant'Anna University Hospital of Ferrara (Italy).
The microbiological screening was focused on the most common HAI-related microorganisms known to reside on surfaces, namely Staphylococcus aureus, Coliforms, Pseudomonas spp. and Candida spp.. The outcome should provide whether the strategy of bio-stabilization with a probiotic-based product could act as an effective and sustainable alternative to chemical disinfectants for treating inanimate surfaces, in particular those located in a nosocomial context.

Identification of the pathogenic strains
The identification of pathogenic strains, initially identified through the Gram staining, was assessed by using the API 20 E (bioMérieux, Inc, Durham, NC, USA) or BBL Enterotube II (BD Diagnostic Systems) for Coliforms including Escherichia coli, the API Staph (20500 bioMérieux, Inc) for Staphylococcus aureus, the BBL Oxi/Ferm Tube II (BD Diagnostic Systems) for Pseudomonas spp., and the API AUX C for Candida spp..

Quantification of the microbial load
Counting was performed by preparing 10 -6 and 10 -7 dilutions of the test suspension in diluents buffer. A sample of 1 ml for each dilution was inoculated in duplicate using the pour plate technique [21]. Each 1 ml sample was transferred into separate Petri dishes and a volume of 15 ml of melted TSA, previously cooled to 45°C, was added. The number of Colony-forming Units (CFU) was determined by colony counting after incubation of TSA plates at 37°C for 20-24 h. A further incubation of plates for 20-24 hours was necessary. The numbers of CFU/ml in the bacterial stock suspension were finally calculated.

Sanitation procedures
Sanitation procedures were performed using microfiber mops, cleaned after each use according to the manufacturer's instructions and colour-coded according to the type of target surface. A dry dust mopping phase was followed by a wet cleaning phase fresh prepared just before each use, with aqueous solutions of either the probiotic-based or the chemicalbased solutions. They were the test and control preparations, respectively. The microfiber cloths were soaked into the solution and stored inside clean containers until use. Mopping phases were performed always by the same trained operator, in order to exclude or minimize the introduction of potential variables in the implementation of the procedures. Floors were treated with 700 ppm/m 2 active chlorine. Hand/body-touched surfaces such as doorknobs, bed frames, tables and chairs or sink, toilet and other bathroom fixtures, were treated with 140 ppm/m 2 active chlorine. These concentrations were used according to the manufacturer's specifications. Following the manufacturer's instructions (Chrisal), a preparation of 1.5 x 10 6 spores/m 2 of the probiotic-based solution was used for the treatment of the same surfaces describe above. During the testing, the environments were used in the usual way, either by personal care, and patients and visitors.

Contamination-controlled experiments (in-vitro tests)
In-vitro experiments were conducted by treating samples of the hospital's areas current materials (i.e. ceramic, PVC, rubber, vitreous-china) with the probiotic-based solution. A solution containing a known concentration (30 x 10 6 cells/ml, 15 ml/m 2 ) of Escherichia coli, Pseudomonas aeruginosa and Staphylococcus aureus strains was used for the contamination of the sampling surfaces. The bacterial load was measured by the determination of the colony count on RODAC plates (BD), containing TSA medium added with lecithin, histidine and Tween-20, in order to neutralize the action of disinfectants. The number of colonies was determined as total microbial count (TMC), or as specific colony count, by exploiting strain-specific medium. Control plates for sterility were used to guarantee sterility.

In-situ trial
The same sanitation procedure described for the contamination-controlled tests was applied for the in situ trial in two different hospital settings in the Sant'Anna University Hospital of Ferrara. They are referred to as in-patients general medicine setting (GM) and out-patient setting (AU) and in which the probiotic-based and chemical-based procedures were used alternatively. The GM setting consisted of the S-Area and T-Area sections (these are the original names of the two areas used since the construction of the hospital), and the Ophthalmology/Cardiology and Orthopedics departments were part of the AU setting. The experimental design for cleaning procedures, sampling surfaces and areas are summarized in Table 1. Evaluation of the probiotic-based sanitation procedure was performed for four months (March -August 2011), in three different time periods and. At the end of the first month, the sanitation protocols were inverted within the area/division of each setting. In order to verify the effectiveness and stability of the probiotic-based protocol, at the end of the second month of sampling collection, the study was prolonged for two more months in the GM setting areas. This choice was due both to the results of the trial, which remained constant in the last period for the PCHS protocol, and to the not negligible fact that over 11,000 microbiological samples had already been carried out, with very high costs accordingly. During the trial, neither the cleaning procedure (PCHS or CBS) nor the products used were never changed. These all belonged to the same purchased batch. The packaging of the probiotic product and the chemical product did not require special measures for their conservation, as indicated by the manufacturers. During the first and second periods, samplings were collected once a week at 07:00 a.m. (time point t 1 ) and at 02:00 p.m. (time point t 2 ), 30 minutes and 7 hours after sanitation, respectively. During the third period, samples were collected twice a week at 02:00 p.m. (time point t 2 ), and at 06:00

Sampling Surfaces
Corridor floor (beginning and end) ceramic rubber ceramic rubber a.m. of the day following the sanitation procedure (time point t 3 ), immediately before the daily cleaning procedures. A total of 11,223 microbiological surface samples were processed following the microbiological environment monitoring methodology (MEM) [22]. Microbiological samplings were performed on two different floor locations in the same corridor (at the beginning and at the end of the corridor), randomly on one point on the toilet floor, and on the sink surface. In total, 33 different surfaces were sampled. Each measure was performed three times.

Calculation of the microbial load
The pathogen reduction obtainable with the PCHS protocol was evaluated in percentage terms compared to the values obtained with the CBS protocol. The average value of each pathogen (measured in CFU/100 cm 2 ) obtainable with the CBS protocol in periods 1 and 2 ( Table 1) was compared with that relating to the period 3, obtained with the probiotic product. For the first or second sampling periods, the calculation of the microbial load was performed by using the following formula: Where all the expressions are referred to: Calculation of the microbial load for data collected during the third sampling period was performed only for the GM setting by using the formula:

Ethics statement
The study protocol was reviewed and approved by the Ethics Committee, named Comitato Unico della Provincia di Ferrara (Unique Committee of the Ferrara Province), of the Azienda Ospedaliero-Universitaria of Ferrara (Ferrara, Italy). The Ethics Committee stated that a formal authorization was not necessary because the probiotic products would not be directly administered to patients but exploited for cleaning of hospital surfaces only. For this reason, the Committee waived the need for written informed consent from participants because of the observational nature of the study.

Statistical analysis
All statistical analyses were performed using the SPSS software (SPSS Inc., Chicago, IL, United States). Normality was tested by the Wilk-Shapiro test [23]. Differences among groups were tested by the non-parametric Mann-Whitney and Kruskall-Wallis tests. Results are indicated as mean ± standard deviation. P values ≤ 0.05 were considered as significant.

Evaluation of the probiotic-based solution in contamination-controlled experimental settings
To investigate the pathogenic strain reduction capacity exerted by the probiotic-based solution, the efficacy under contamination-controlled conditions was analyzed in vitro. Selected surfaces, such as ceramic, PVC and rubber, were contaminated with known amounts of Escherichia coli, Staphylococcus aureus and Pseudomonas aeruginosa and then the bacterial load was monitored over time. Before treatment with the probiotic-based solution similar average CFU/cm 2 values among the three strains on the contaminated surfaces were observed. Interestingly, a reduction in the amount of pathogens on test surfaces was clearly evident one hour after treatment, with a significant lowering in the number of colonies within 30 minutes for all the strains tested. A barely detectable bacterial load was maintained up to 5 days of measurement, while unvarying bacterial loads were recognized on the untreated control surfaces (Table 2).
These results indicate that the application of the probioticbased product results in a significant lowering of specific bacterial loads in a contaminated-controlled context.

In situ trial of the probiotic-based sanitation procedure
In order to test the efficacy of this new cleaning procedure on a real context, the experimental evaluation was performed in a hospital environment. Sampling data, obtained in the hospital settings, are shown in Tables 3-5. Table 3 shows the comparison of the microbial load between the two hospital settings (GM e AU, see Table 1) during the first and second experimental periods, in which a reduction in the microbial load of the monitored pathogens was still evident.
Mean values for the GM and AU hospital settings are not separately reported in Table 4 because no statistically significant differences in the microbiological load were found, despite the different type of health care activity. The PCHS and the CBS cleaning procedures have always been performed at 06:30 a.m..     As already mentioned, the microbiological load was monitored in three stages: half an hour after the application of the cleaning procedures (t 1 , 7 a.m.), about 6 ½ hours later (t 2 , 2 p.m.) and about 24 hours later (t 3 , 6 a.m. of the next day and before the new cleaning). The t 3 instant corresponds to the time when there was the maximum value of contamination of the surfaces. The mean TMC values observed at t 1 with the probiotic-based solution were significantly higher than those obtained with the chemicalbased product, due to the presence on treated surfaces of nonpathogenic microorganisms of the probiotic-based solution itself. Indeed, the probiotic microorganisms accounted for about 98% of TMC. The comparison of the two different protocols showed statistically significant differences in the microbial loads at time point t 1 . From Table 4 it can be observed that, for the CBS procedure, the S. aureus value was equal to 49.2 CFU/100 cm 2 at t 1 time (half an hour after cleaning), but it becomes 110.96 CFU/100 cm 2 at t 2 time (nearly 7 hours after cleaning). That is to say its value doubles after about 7 hours. A very similar trend is noted for Coliforms, Pseudomans spp and Candida spp.. On the contrary, with regard to the PCHS procedure, the value of S. aureus was equal to 58.86 and 81.65 CFU/100 cm 2 at t 1 time and at t 2 time, respectively. For other pathogens, the value even decreased as time goes by, probably because they are more sensitive to the action of the PCHS protocol's Bacillus spp., which continue to exert their competitive action for a long time even after their application (unlike chemical disinfectants, whose action expires quickly). This comparison is summarized in Table 5. Returning to Table 4, it is interesting to note the data obtained in the 3 rd period. When the PCHS procedure was applied for several weeks (as for the S-Area, ½ hours using the CBS protocol. It is therefore quite evident the effect of containment of pathogen levels in the interval of time between one cleaning procedure and the other, when using the product based on probiotics. The treatment with the probioticbased product reduced the potentially pathogenic microbial load, with the exception of the TMC and the Staphylococcus aureus and Candida spp. titers, while an increase was observed after the application of the chemical-based product. Conversely, as well as for other pathogens tested, a reduction in the microbial load of both Staphylococcus aureus and Candida spp. was observed at t 2 ( Table 5).

Reduction of the Microbiological Load on Hospital Surfaces Through Probiotic-Based Cleaning Procedures: A New Strategy to Control Nosocomial Infections
These results indicated a statistically significant general reduction of the potentially pathogenic species after the treatment with the probiotic-based product.

Microbiological stabilization effect of the probioticbased procedure
In order to verify the effectiveness and stability of the probiotic-based procedure, the study was prolonged for two more months within the GM setting areas (see Methods and Table 4).
Noticeably, when the bacterial load, collected at t 2 in the first and second period of the survey, was compared with the microbial load obtained in the third period, always at t 2 time, a significant reduction was obtained for all average bacterial loads, with the only exception of Coliforms (p=0.485) ( Table 6). These results suggested that the prolonged use of the probiotic-based protocol significantly improved the level of sanitation of treated surfaces, with a stabilization of the microbial load over time, with constant microbiological contamination values over time. This phenomenon was very evident in researches subsequently conducted; these results will be further described in other future works.
The percentage reduction of the bacterial loads for different surfaces, sampled at t 2 , confirmed that the probiotic-based procedure resulted in a general reduction (up to 80%) and stabilization of the pathogenic load (Table 7). In addition, data obtained in the third period, during which sampling was

7/10
performed 24 hours later the sanitation procedure with the probiotic-based protocol, showed that the average microbial loads were lower than those obtained at t 2 , by using the chemicalbased protocol.
Overall, these data indicate that, differently from observations obtained with the chemical-based product, the probiotic-based protocol resulted in an efficient stabilization of the microbial load, which was maintained at low levels within the 24 hours until the subsequent sanitation procedure.

Discussion
The present study investigated the effectiveness of a probiotic-based sanitation solution for hard surfaces both in a laboratory contamination-controlled setup and in a real setting, consisting of a study-model hospital. This study was based on the hypothesis that probiotic bacteria, which are defined as a preparation of viable microorganisms that bring a benefit to the host health [24], could colonize surfaces and counteract the proliferation of other bacterial species [17], including those recognized as potential pathogens for humans. Probiotics have been extensively used per os as food supplements for alterations, in both the qualitative and quantitative composition of human intestinal flora. Bacillus spp have been exploited as probiotics for a series of different applications, such as dietary supplementation in humans and animal feed, due to their capacity of stimulating the immune system [25]. In particular, Bacillus subtilis has been indicated as safe for human consumption [26,27], in live and spore forms [28] and there is increasing evidence and interest in the use of this non-pathogenic microorganism in the preparation of food [29].
The promising strategy, based on probiotics, could be shifted from the feed supplement purpose to a new application in which these bacteria could be involved in the colonization of inanimate surfaces in order to minimize and contrast the proliferation of other bacterial strains, potentially pathogenic to human subjects.
In this context, a probiotic-based sanitation procedure for hard surfaces can be hypothesized to have a role in the control of HAIs, which are the most frequent complications in healthcare facilities [1], and often develop from either direct or indirect contact with pathogen-contaminated surfaces.
The evidence for a proof-of-principle application of this strategy, stemmed from the results obtained in contaminationcontrolled studies, in which commercially available pathogenic strains such as Escherichia coli, Pseudomonas aeruginosa and Staphylococcus aureus were exploited. These experimental settings revealed that the probiotic-based solution resulted in a significant reduction in the pathogen populations used as reference, pointing towards the potential feasibility of this method on a real setting. To address the issue in a real context, the Sant'Anna University Hospital of Ferrara as a study-model setting was chosen. The observations of the probiotic-based solution indicate that its effectiveness in reducing and maintaining a low pathogen load was significantly more pronounced than that of a chemical-based disinfectant on all tested surfaces, over time. The reduction of pathogen load, observed using the probiotic-based treatment, could be likely due to bio-stabilization, according to the competitive exclusion principle (Gause's law) [18,30], and with other mechanisms related to natural antimicrobial activities, as demonstrated by several in-vitro studies [31][32][33] and by the recent isolation and description of a Bacillus strain, displaying a broad-spectrum biofilm inhibition activity [34].
We are aware of some limitations in this study, concerning the actual mechanisms and dynamics involved, which are beyond the purpose of the study itself. Noticeably, as mentioned above, one of the most well-known effects of probiotics is related to competitive exclusion [35], which confers an advantage to a certain bacterial strain on others, competing for the same resources. Nevertheless, several other mechanisms may contribute to this effect, such as quorum sensing/quenching [36], or the production of multifunctional molecules, known  as bio-surfactants, which have antagonistic anti-adhesive properties against microbial pathogens [37,38]. Among these, one of the most representative is surfactin, produced by Bacillus subtilis, which is a surface-active lipopeptide belonging to the extracellular polymeric substances (EPS) [39] displaying a potent antimicrobial activity [40]. In addition, other competition-related processes such as cannibalism and predation are involved in the formation of the biofilm and in the antimicrobial activity of Bacillus subtilis, respectively [41][42][43][44].
The proposed probiotic-based treatment, at variance with traditional chemical-based sanitation procedures, gradually reduced microbial contamination, with a consolidation of its effect in 8-12 weeks since the first application. Present data indicate that the probiotic-based procedure exerted a continuous activity, either by affecting the nutritional sources available for the development of microorganisms or, more likely, by forming anti-pathogen biofilms on inanimate surfaces. When a traditional chemical-based disinfectant was used, the potential pathogenic microbial load had a steep increase within 7 hours, resulting in a doubling or tripling of the number of CFU/cm 2 . Conversely, the increase of microbial load after the application of the probioticbased solution was modest within the next 24 hours. This observation suggests that the probiotic-based solution displayed a long-term stabilizing effect that strongly lowered the risk of peaks in pathogen loads between consecutive applications, which represents one of the most deleterious hurdles frequently witnessed for disinfections with chemical-based solutions.
This study provides indications on the need for a careful use of the surface microbiological sampling through the TMC monitoring, for the evaluation of environmental contaminations, because of the wide range of the microbial load, in particular when chemical disinfectants are used. Surface sampling for the evaluation of the disinfectant efficacy should be performed within 30 minutes after application, but it does not represent a suitable method to describe the daily "average" contamination of surfaces. Differently, in the case of the probiotic-based treatment, the CFU/cm 2 count, targeted on a specific bacterial strain, provides a scientific value that is stronger than that yielded by the TMC, since the daily variation of the microbial load is modest.
A matter of debate could pertain the use of bacteria for sanitation procedures, notwithstanding the final goal of sanitation itself is not strictly the lowering of microorganisms in general, but rather the control or limitation of potentially pathogenic strain development, tolerating the presence of microorganisms harmless for humans. Indeed, regarding the risk related to the use of probiotic agents in indoor (or clinical) environments, the strict selection of strains must be of primary importance. It is worth noting that many probiotic strains are considered safe for use in humans [26,45,46], and that the Food and Drug Administration (FDA) classifies them as generally recognized as safe (GRAS) microorganisms. In addition, Bacillus subtilis has been listed on the Qualified Presumption of Safety (QPS) by the European Food Safety Authority since 2007 [47,48]. Nevertheless, we are aware that safety assessments, including antibiotic resistance and transferability, genetic stability, toxicity, remain important issues, still debated even for probiotics already designated for human consumption both in healthy and in sick individuals [49].
For these reasons, when working in healthcare areas, hosting immune-compromised patients, the need and availability of proper solid data about safety and toxicity of the strain used, should be the absolute priority. On the other hand, the major pitfall of current chemical disinfection procedures and agents is related to the re-growth/re-contamination leading to unstable bacterial loads, thus potentially exposing patients residing in hospital rooms, to several HAI-related complications. Noticeably, this picture prepares the ground for the notion that the replacement of such pathogenic strains with probiotics, displaying a safer reputation, should result in a lower risk for patients to develop HAIs.
The present study demonstrated that the strategy of probiotic-based sanitation results in an efficient and longterm lowering of pathogens both in contamination-controlled and hospital contexts. Our data indicated that i) a probioticbased sanitation procedure was significantly more effective (up to 80%) in reducing potentially pathogenic microbial loads than a traditional chlorine-based chemical disinfectant, and ii) the reduced microbial load was stably maintained at low levels throughout the 24 hours after application, despite of the presence of continuous and multiple sources of microbial recontaminations due to external natural contributors such as patients, visitors, hospital staff and moving materials.
However, further studies are needed in order to identify new standards for hygiene monitoring, and to better understand the mechanisms underlying this probiotic sanitation technique. Environmental probiotics alone may not be adequate to control all microbial species and, in addition, routine chemical cleaning procedures may interfere with probiotics [15]. Furthermore, it is an open challenge to determine whether, how and to what extent the strategy of probiotic-based sanitation of hard surfaces in hospital contexts will correlate with the number of HAIs.

Conclusion
In conclusion, this proposed strategy, based on the new PCHS system, provides evidence for the proof-of-principle of the strategy of bio-stabilization, and for probiotic-based products as a reliable alternative to traditional chemical disinfection of nosocomial surfaces, in particular of those directly or indirectly in contact with patients. However, this method is not proposed for environments that require a very low value of total microbial count (TMC), such as operating rooms, infectious disease wards, clean room, etc. In the future, the authors aspire to correlate the efficacy of this probiotic-based procedure with the reduction of nosocomial infections.
Finally, in order to evaluate the susceptibility or resistance of the Bacillus strains to antibiotics, the study was implemented by exploiting antibiogram tests on Bacillus spp. colonies, coming from the same type of the sample surfaces (e.g. floor) monitored in the in situ trial. The first results indicate that the isolated Bacillus spp. strains are susceptible to all the antibiotics tested, with the exception of penicillin, towards which the Bacillus sp is naturally resistant, as also indicated by antibiograms performed on the commercially-available ATCC strain. These preliminary data are not presented here but will be part of the future research work, where also new results, stemming from molecular assays (i.e. qPCR) for the evaluation of resistance factors, will be described.