Potentially pathogenic bacteria isolated from neglected air and surfaces in hospitals

Hospital infections (HI) are a serious public health problem in many countries. Several studies have identified strains correlating to surgical site infections, many with multi-resistance. The goals of this study was to quantify, to identify and to verify the resistance profile of microorganisms collected at two hospitals settings, and to alert health professionals how environmental contamination can influence hospital infection rates. For air sampling in operating rooms, intensive care unit and materials sterilization center, the impaction method (Spin Air, IUL®) and passive sedimentation were used. For the isolation of bacteria on surfaces and uniforms contact plates (RODAC®) were used. Identification of the microorganisms was performed using Vitek® 2 Systems. The antibiograms were conducted according to the disk diffusion method recommended by CLSI. The surgical center of hospital B presented more than 500 CFU/m 3 in aerial microbial load. In the aerial microbiota of the sampled areas of both hospitals, M. luteus , S. haemolyticus and S. hominis spp hominis were the prevalent microorganisms, with a percentage greater than 30%. On the surfaces and uniforms there was a prevalence of M. luteus (40%) and S. hominis spp hominis (20%) among others, and some of the resistant strains were isolated from environments with microbial load within the recommended limits.


INTRODUCTION
To ensure a satisfactory internal air quality, the hospital environment must be constantly monitored in order to protect patients and workers from acquiring hospital infections and occupational diseases. (Birgand et al., 2015;Cabo Verde et al., 2015;Emuren, Ordinioha, 2016) The origin of microorganisms responsible for HI outbreaks may come from humans (Munoz-Price et al., 2012) (e.g.: hands and uniforms), system of climatization (Kumari et al., 1998), inefficient sterilization/cleaning procedure (Dancer, 2009), infiltration of humidity (Baughman, Arens, 1996) and external microbial load (Prussin, Marr, 2015), and these factors need to be associated with favorable conditions for microbial growth. (Park et al., 2013) Although some microorganisms found in the air, surfaces and uniforms of hospital environments are not considered pathogenic for healthy people, studies show that they are causing infections in people with compromised immune systems (Litvinov et al., 2015). In addition, hospital microorganisms are likely to be multidrug resistant (Cornejo-Juárez et al., 2015), which makes environmental monitoring an important tool to control and prevent nosocomial infection outbreaks, and reduce the rate of morbidity, mortality and costs. (Galvin et al., 2012;Neidell et al., 2012).
The present study aimed to evaluate and determine the resistance profile of microorganisms isolated from air, surfaces and uniforms, commonly neglected places, of different sectors of hospitals.

Study design
A prospective study on airbone, surface and uniforms microbiota was carried out in two small private hospitals (A and B) specialized in ophthalmologic, aesthetic, oncological and orthopedic surgeries, located in urban areas in the city of Goiânia/GO, and in three hospital sites: the operating room theater (ORT); the intensive care unit (ICU); the material and sterilization center (MSC). Three sampling campaigns were performed at each site at Hospital A and Hospital B, for six consecutive weeks.
In all ORTs evaluated there was a ventilation system equipped with 0.3 mm 99.97% Hight Efficiency Particulate Air (HEPA) filters. The ORT air pressure was ≥ 5 Pa higher than in adjacent rooms (only in Hospital A). In all ICUs and MSCs evaluated there was split type air conditioner and there was no high pressure compared to adjacent rooms.

Microbiological air sampling
Microbiological air counts were measured by impaction using air sampler (SpinAir ® , IUL S.A., USA) with a flow rate of 100 l/min (with a sampled volume of 1 m 3 ), and the equipment was approximately 1 m above the floor. All samples were collected during the normal work routine, and especially in the ORT, the collection was initiated after the first incision (one sampling per room for three weeks).
For passive sampling only TSA was used and these plates were exposed at predetermined locations for one hour. The incubation procedure was the same as the process that was used in the impaction method.

Surface and uniforms sampling
Surface and uniforms sampling were performed with the aid of the contact plates (Kasvi ® , Curitiba, Brazil) containing TSA with 0.1% lecithin (Scharlau ® , Barcelona, Spain) and 0.7% Tween 80 (Dinâmica ® , São Paulo, Brazil). The contact plates were gently pressed onto the surface or uniforms for 10 seconds (one per surface and uniform). After collecting, the plates were incubated under the same conditions and time as in the previous section.
In the ORT the collection was performed on a wall (closest to the area of action) and in the sterile surgical gown (chest and forearm). In the ICU, the collections were implemented on a wall, sink, bed and worker's uniform (chest). In the MSC, on a wall, autoclave cover and worker's uniform (chest), under aseptic conditions.

Identification of microorganisms
Isolated bacteria were identified using the biochemical identification cards of the Vitek ® 2 Systems (BioMérieux, Inc., France).

Antimicrobial susceptibility
The antimicrobial susceptibility test was performed by disk diffusion for the microorganisms indicated by the Clinical & Laboratory Standards Institute guidelines. (CLSI, 2015) Staphylococcus aureus ATCC 25923, Pseudomonas aeruginosa ATCC 27853 and Escherichia coli ATCC 25922 were used as control microorganisms for the disk diffusion tests.

Data analysis
For the data analysis, the Analysis of Variance (ANOVA) of single factor with significance level p <0.01 using Microsoft Excel ® 2010 to determine differences between passive and impaction sampling was performed.

Microbiological air, surface and uniforms sampling
The results for the microbial load isolated in the air, surfaces and uniforms of the analyzed hospitals are present in Figures 1 and 2. In all sectors analyzed the microbial load was over 200 CFU/m 3 . In surfaces and uniforms microbial load indices were greater than 5 CFU/ cm 2 , reaching values of approximately 130 CFU/30 cm 2 in a sector of Hospital B ( Figure 2). As there was no ICU in Hospital B, the collection procedure was performed in the post-surgical rest room (PSRR).  In the sampling by impaction in cetrimide agar, a selective medium for P. aeruginosa, there was no growth in any evaluated environment.
In this study, a statistical difference (p <0.01) between the active and passive sampling methods was noticed in all sectors of both hospitals evaluated.

Identification of microorganisms
Regarding the bacterial isolates identified during the samplings, 79 and 96 microorganisms were isolated in hospital A and B, respectively, in which 24.7 and 36.5% were Staphylococcus subsp.
The microorganisms prevalent in the evaluated sectors of Hospital A and B are shown in Figures 3 to 8.
In the ORT of hospital A (Figure 3), there was also the isolation in air of a lower percentage of Staphylococcus. hominis subsp. hominis (6%) and Sphingomonas paucimobilis (6%) (Data not shown).
In the MSC of hospital B (Figure 8)

Antimicrobial susceptibility
In general, more than 90% of the bacteria isolated from both hospitals were resistant to at least one of the evaluated antibiotics. Simultaneous resistance to azithromycin and penicillin (all isolates, as described in Table 1) was quite common, especially in bacteria isolated from the air.
In the analyzed hospitals, the bacteria (Staphylococcus subsp. and others Gram-positive cocci) were resistant to oxacillin, with 31.25% from hospital A and 34.8% from hospital B, for which one and eight isolated were in the air of the surgical centers (Hospital A and B, respectively).
Extensively drug-resistant (XDR) bacteria (Table  I) were detected in several environments evaluated in air, surface and uniform samples. The largest number of XRD strains were isolated in the MSC and ORT of hospital B, both presenting five XRD strains (35.7%), followed by the ICU of hospital A with two XRD strains (14.3%). In some samples the isolated analyzed were resistant to azithromycin, oxacillin, penicillin, clindamycin and vancomycin (57,14% in all hospitals).

DISCUSSION
The microbial load and the isolated microbiota in this study, the highest microbial load was found in the PSRR (566 CFU/m 3 , Hospital B) and the lowest in the ORT (124.5 CFU/m 3 , Hospital B), isolating predominantly Staphylococcus subsp. (26% S. haemolyticus and 20% S. epidermidis in general) and Micrococcus luteus (40%).
There are several studies that have displayed similar results to this work (Cabo Verde et al., 2015;Tang et al., 2013;Wan et al., 2011). This is exemplified by a Portuguese study (Cabo Verde et al., 2015) that found a microbial load ranging from 12 to 170 CFU/m 3 in ORT, 240 to 736 CFU/m 3 in ES (Emergency service) and 99 to 495 CFU/m 3 in SW (Surgical ward), isolating predominantly Staphylococcus subsp. (S. aureus, S. capitis, S. hominis, S. epidermidis and S. warneri) and Micrococcus (M. luteus and M. lylae).
The UK Department of Health Technical memorandum (HTM 03-01) (Department of Health/ Estates and Facilities Division, 2007) determines the ORT limits during activity of 180 CFU/m 3 . Based on this legislation only the ORT of hospital B presented high airborne microbial load, and for the other sites (ICU and MSC) there are no recommendations. One point that should be highlighted is that, although it is required according to hospital norms, the ORT of Hospital B does not present positive internal pressure higher than in adjacent rooms, a factor that contributes to the increase of aerial microbial load and, consequently, to the risk of infection.
The statistical difference (p <0.01) between active and passive sampling is probably due to active sampling being a quantitative method, its unit is expressed in m 3 of air, as opposed to passive sampling, which is not a quantitative method, since it is not possible to determine the amount of air coming into contact with the exposed plate.
The evaluation of air quality, surfaces and uniforms is an important tool for the investigation and prevention of hospital infection outbreaks. Compared to other studies, which only identified the strains isolated from air samples, surfaces and uniform, this study determined the resistance profile of the isolated microorganisms. Several multidrugresistant microorganisms were found, mainly in the MSC and in the ORT (Hospital B) and in the ICU (Hospital A).
These results are alarming, since, considering the exception of the MSC, where only healthcare workers have access, in the other areas the patient is exposed and comes in contact with these potential pathogens.
These findings show that in addition to the importance of assessing the microbial load of surfaces or environments, it is necessary to identify microorganisms and know their resistance profile (Table I -S. haemolyticus in the air at the ORT), where the aerial microbial load is within the recommended maximum limit of 180 CFU/m 3 (HTM 03-01), and it also showed that multidrug-resistant bacteria were found in the sampled microbiota.
The isolation of S. haemolyticus and S. epidermidis multi-drug resistant (Table I) is a big concern because, despite being a common bacterium of the human microbiota, studies have shown these bacteria as emerging pathogens in nosocomial infections mainly involving medical devices and the formation of biofilms (Pinheiro et al., 2016).
The same happens with S. hominis subsp. hominis that has been reported as a potential pathogen isolated in generalized infections (Voineagu et al., 2012).
The organism M. luteus, found in several sampled environments, has been described as the causative agent in endocarditis (Miltiadous, Elisaf, 2011) and central venous catheter infection (Oudiz et al., 2004).
Some microorganisms isolated in low percentages in this study are described in some cases of nosocomial infections, is the case of E. ludwigii reported as an agent causing an outbreak of bloodstream infection (Flores-Carrero et al., 2016); Bacillus cereus causing bacteremia, infection of skin, bones and joints (Veysseyre et al., 2015); S. lugdunensis causing bacteremia (Pereira et al., 2011); S. warneri causing endocarditis (Arslan et al., 2011) and S. cohnii subsp. urealyticus causing bacteremia (Soldera et al., 2013).
Multidrug-resistant bacteria are a worldwide public health problem, and investigating possible sources of these microorganisms is very important to prevent outbreaks. According to the results of this study, multi-resistant bacteria can be present in places that present risk to both the patients and healthcare workers (MSC, ORT and ICU, Table I).
Regarding patients, the risk is imminent because the hospitalized patient has a weakened immune system, which determines a predisposition to suffering an infection when coming into contact with a potential pathogen (Cabrera-Cancio, 2012).
In the case of healthcare professionals, when they come into contact with these microorganisms through the air or on work surfaces and even in uniforms, these workers can become reservoirs that assist in the dissemination of these pathogens (Chemaly et al., 2014).
The differential of this study is the investigation of the resistance profile of isolated microorganisms, a fact that may pose risks to health professionals but especially to patients exposed to these types of microorganisms. From this investigation, previously neglected sites should be monitored so as not to become a source of spread of resistant pathogens.

CONCLUSION
A large number of multi-d r ug resistant microorganisms (14 strains) have been isolated from environments, even within the recommended microbial load limits, that pose risks to patients such as the operating theater (ORT) and intensive care unit (ICU).
The results show that environmental monitoring, using an active sampling method, is an important tool, and should be adopted by hospital infection control committees to investigate, control and reduce the occurrence of nosocomial infections, mainly from sources such as the air, surfaces and uniforms.