Distribution and correlation of methicillin-resistant coagulase-positive staphylococci (MRCoPS) between environmental surfaces, veterinary staff and dogs within a veterinary teaching hospital, Thailand

Background Coagulase-positive staphylococci (CoPS), a gram-positive cocci bacterium, is a group of bacteria causing dermatitis and septicemia in animal. Methicillin-resistant (MR) CoPS (MRCoPS) can nd on human, animal and environmental including medical equipment leading to nosocomial infection. This study aimed to determine the distribution and to analyse correlation of methicillin-resistant coagulase-positive staphylococci (MRCoPS) between environmental surfaces, veterinary staff, and dogs within a Veterinary Teaching Hospital (VTH) at Chulalongkorn University (CU) in Bangkok, Thailand. All isolates were characterized the antimicrobial susceptibilities, Staphylococcal cassette mec (SCCmec) typing and clone typing by Pulsed-Field Gel Electrophoresis (PFGE).


Results
In total, 88 CoPS isolates were obtained from 53 samples comprising 24 parts of oors (24/53, 45.3%), 14 items of medical instruments (14/53, 26.4%), 14 dogs (14/53, 26.4%) and a veterinarian (1/53, 1.9%). Sixty-two of 88 were MRCoPS locating in the ve hospital rooms. The highest MRCoPS species were MR Staphylococcus pseudointermedius (MRSP) isolates (57/62, 91.9%) retrieved from 26 samples and mainly discovered on the oors. Seven antibiogram patterns and two SCCmec types were resolved from 62 MRCoPS isolates. Most of the MRCoPS (93.6%) were resistant to at least three antibiotics. All MR Staphylococcus schleiferi subsp. coagulans (MRSSc) isolates displayed a triple antibiogram associated with the SCCmec type V, while the MRSP isolates showed various antibiograms and SCCmec types. The S. pseudintermedius in dogs and three rooms were related by their PFGE pattern A. The average percentage of MRCoPS-positive surfaces were typically high on the oors (20%). The dermatological room showed the highest of MRCoPS-positive surfaces on both the oor and medical instruments.

Conclusions
The distribution of CoPS and MRCoPS has become a relative common situation in veterinary hospitals, and so attention must be paid to limit cross-contamination between patients and the hospital environment. To general; please provide impact of this study digested as a take home message.

Background
Coagulase-positive Staphylococci (CoPS) are a group of Gram-positive cocci that cause dermatitis and septicaemia [1,2] The three signi cant CoPS members in animals are Staphylococcus (S.) aureus, S. pseudintermedius, and S. schleiferi subsp. coagulans [3,4]. Of importance is that canine CoPS can exchange the mobile genetic element "Staphylococcal cassette mec (SCCmec)" in both intra-and inter-species [5]. This mobile genetic element can carry many antibiotic resistance genes, especially the mecA gene (oxacillin resistant gene). Methicillin-resistant (MR) CoPS carry the mecA gene and are some of the most dangerous bacteria in human and veterinary hospitals [6,7]. This group of pathogens can be transmittable among human, animal, and environmental surfaces, including medical equipment. Typically, MR S. aureus (MRSA) is located at a low population level in veterinary hospitals and on small animal skin but can cause pathogen outbreaks, such as that at an equine hospital. Importantly, they cause infections and septicemia in humans [8,9].
The MR S. pseudintermedius (MRSP) commonly presents multidrug resistance to certain particular antibiotic groups, such as β-lactam, macrolide, lincosamide, uoroquinolone, and aminoglycoside [2]. Recently, the increasing distribution of MRSP in veterinary hospitals has become a high risk of surgical site infections and septicemia in animal patients and also a threat of organisation development and standardisation [10].
The distribution of MRSP in a veterinary hospital poses the risk of nosocomial infections [11][12][13]. Environmental surfaces, veterinary staff, and animal patients all play a role in both MRSP and (MRSSc) circulation in veterinary hospitals [11,12,14]. The surveillance of MRCoPS is essential in allowing the prediction of the distribution of these pathogens in hospitals [15]. MRSP has been discovered in veterinary hospitals on medical instruments, such as weight scales, examination tables, and stethoscopes [16]. The remaining MRSP in the environment is a potential cause for cross-contamination in veterinary hospitals [11]. Previously, an explosion caused by the MRSP ST71 isolate was of concern in a Finnish veterinary teaching hospital (VTH) [11]. In Thailand, veterinarians and pet owners can receive MRSP from pet and become a signi cant carrier sharing to medical devices. However, the factors affecting MRSP distribution were different in each hospital upon hygienic management [11,17].
A suitable hygienic strategy and hospital accreditation should be suggested to control bacterial distribution in veterinary hospitals [8,15], including hand hygiene, protective gear, and type of disinfectants [11]. However, there is no reported information on MRSP distribution in hospital environments related to the managements.
This study aimed to determine the distribution and correlation of MRCoPS among environmental surfaces, veterinary staff, and dogs within a Veterinary Teaching Hospital in Bangkok, Thailand and to characterise the clonal types and their antimicrobial susceptibilities and resistant gene cassette.

Distribution of CoPS and MRCoPS in the VTH-CU
A total of 253 samples were collected from 216 places of environment (8 units, 5 points of oor per room and twice collection per day), 23 veterinary staff, and 14 dog patients. In total, 88 CoPS isolates were found from 53 samples comprised of 24 parts of the oors (24/53, 45.3%), 14 medical instruments (14/53, 26.4%), 14 dogs (14/53, 26.4%) and one veterinarian (1/53, 1.9%). All rooms presented CoPS, especially the dermatological (27.3%) and gynaecology (22.7%) rooms. However, no CoPS were found in any disinfectant water in forceps jars, stethoscopes, branches and drug cabinets. Approximately 53% of the CoPS were found on oors. We found the three major canine CoPS in this study, with S. pseudintermedius having the highest population at 75 isolates, while S. schleiferi subsp. coagulans presented 12 isolates, and S. aureus was only obtained from a veterinarian and de ned as a methicillin-susceptible (MS) S. aureus (MSSA). This population of MSCoPS was about 2.5-fold higher than that of MRCoPS. The distribution of MSCoPS and MRCoPS on the oor and medical instruments is summarised in Antibiogram and SCCmec types of the MRCoPS A total of 62 MRCoPS were characterised by their antibiogram using six common antibiotics used in veterinary hospitals. Only the SCCmec type V was detected in these samples. The percentile of resistant antibiotic drugs is displayed in Figure 1, whereas the SCCmec type is included in Figures 2-4. Overall, seven antibiogram patterns and two SCCmec types were resolved from the 62 MRCoPS isolates. All MRCoPS were resistant to clindamycin (DA), and erythromycin (E), and most (93.6%) of the MRCoPS were resistant to at least three of the screened antibiotics with half of them having the same antibiogram patterns (CN-E-DA). All the MRSSc isolates were resistant to three antibiotics (CN-E-DA) associated with the SCCmec type V, while MRSP showed various antibiogram patterns and SCCmec types.
PFGE types All S. pseudintermedius and S. schleiferi subsp. coagulans were identi ed by PFGE patterns (Figures 2-4). After cfr9I cutting, S. schleiferi subsp. coagulans showed only two patterns (A and B), while S. pseudintermedius displayed 14 patterns (A-N). The relationship among S. pseudintermedius in the dogs and the post-surgery care, dermatological, and gynaecology rooms mainly presented the pattern A. Only 12 S. pseudintermedius isolates could be cut by ApaI only. From the ApaI restriction enzyme pattern, the DNA ngerprints were divided into three patterns that show a close relationship between the gynaecology, post-surgery care, and internal medical rooms.

Distribution of MRCoPS in the VTH-CU
All MRCoPS were separated into two origins ( oor and medical instruments) from eight VTH-CU places (Table 3).
Amoxicillin/clavulanic acid and E were the main antibiotics used in this hospital. The common routine for oor cleaning in the six rooms was to use a mop with a 2.5% (w/v) quaternary ammonium compound, twice daily. The dermatological room was the only room where the oor was cleaned with a broom, and this was once daily. The vaccination room was free from MRCoPS-positive places on both the oors and medical instruments. The average percentage of MRCoPS-positive surfaces were typically high on the oors (20%) compared to the medical devices (7.7%). In this study, the internal medicine room was chosen as the reference room due to its having the lowest MRCoPS-positive surfaces on both the oors and medical instruments. The dermatological room showed the highest of MRCoPS-positive surfaces for both the oor and medical devices.

Discussion
CoPS and MRCoPS were tentatively enumerated and isolated from medical instruments and oors in each hospital room. A high prevalence of MRCoPS, especially MRSP, was detected on the oors and medical devices in each room.
The possible source of MRSP could be identi ed, which can then lead to preventive and control measures in the VTH-CU. This cross-sectional study collected samples from environmental items, comprised of hand-touch sites, medical devices and materials, oor surfaces, veterinary staff, and dog patients in the VTH-CU, located in Bangkok's central area with a high customer ow rate (approximately 389 cases per day). The hospital has an increased risk of MRCoPS distribution and serves as a good model for study. Interestingly, the highest CoPS and MRCoPS levels were found in the dermatological room, especially on the oor and medical instruments.
One veterinarian carried MRSP with multidrug resistance in his nasal cavities. Human-contained MRSP in the nasal carriage is uncommon but closely associated with pet owners and veterinary staff [18]. Human carriers' prevalence has been reported to range from 1-5%, but the highest prevalence (8%) was reported in Thailand [18,19]. However, it can't be concluded that dogs and all high-touch sites in the VTH-CU were the potential sources of transmission, since the human strains were not linked to those of the dogs and environmental sources. The exposure time and host status may be signi cant factors in MRSP transmission [20,21]. However, many human-associated animal pair strains should be recruited and analysed in a further study.
In veterinary hospitals, antimicrobials are mostly administrated to the pets, leading to resistant commensal microbes on the skin [22]. This study showed that all dog patients contained CoPS on their nasal cavities, but only a few dogs harboured resistant pathogens. Nevertheless, the DNA ngerprint pattern (PFGE pattern A) showed the relationship between the MRSP clones and dogs' environment. It is possible that sharing of MRSP might occur in a veterinary hospital. A correlation between the PFGE patterns in the surgery room has been reported previously [10], but no correlation between the equipment and dog patients was found. However, this study showed the relationship between dogs and the environment in other hospital rooms. We assumed that MRSP contamination might be due to accidental sharing between dogs visiting the hospital. With prolonged use of antimicrobial administration to the dog patients, the clinical room had the highest risk of MRSP existence and maybe the potential source of the hospital Most of the strains were untypable for SCCmec using the previous recommendation. The untypable SCCmec was previously identi ed as ΨSCCmec57395, which showed multidrug resistance that was the majority in Thailand [22].
The frequency of speci c MRSP clones in an individual could be explained by selective pressure exerted by preexisting resistant strains during antimicrobial exposure. Besides, transmission of SCCmec between Staphylococcus spp. is of concern in many veterinary hospitals because this mobile genetic element can carry many resistance genes in its cassette. This study showed that at least two common veterinary drugs might not be useful for CoPS infections in the VTH-CU.
Moreover, we discovered Mupirocin (MUP)-resistant bacteria in one dog. Mupirocin is not a commonly used antibiotic in veterinary treatment, but it is often used in multidrug-resistant bacterial infection cases. This drug is recommended in human treatment more than in animal treatment. Thus, MUP-resistant bacteria are of great concern in human and animal hospitals [24,25].
The restriction enzymes crf9I, apaI, and sacII were previously recommended for cutting the DNA of MRSA ST 398 [26]. However, in this study, 12 S. pseudintermedius isolates could not be de ned by the Crf9I-based method. The apaI restriction enzyme, recently used for MRSA Sequence Type 398 [27], was found to cut these 12 isolates to obtain the DNA ngerprint patterns. Given that this enzyme can cut at the methylated restriction site that Cfr9I and SmaI cannot [27], then these 12 S. pseudintermedius isolates in this study might contain methylated restriction sites. However, next-generation sequencing in a further investigation could declare their whole genome in the future.
The oor of the VTH-CU was the most frequently found area for bacterial contamination. The use of a proper disinfectant and cleaning protocol could restrict the distribution of pathogens in hospitals and reduce the prevalence of nosocomial infections [28,29]. However, this study provides a shred of important evidence for revising the bacterial decontamination protocol in future hospital strategic programs. Mopping with a proper disinfectant [i.e. sodium hypochlorite, 2% (v/v) phenolic solution and 0.5% (v/v) chlorhexidine] has been recommended as a routine cleaning protocol in a veterinary hospital [28]. In the mopping method, the disinfectant has to contact the oor surface for at least 20 min, and then the mopping should be repeated. Additionally, we recommend changing the mop bucket twice daily: at the beginning of the day and the day's nal mopping.
After routine cleaning, MRCoPS was found at the highest level in ve units. In the VTH-CU, examination tables were cleaned immediately when the items were not in use. Even if the cleaning patterns were the same, residual MRSP were still found on the examination table of the gynaecology, post-surgery care, ICU, and dermatological rooms. Therefore, we suggest that the residual MRSP on some examination tables might result from cleaning practices and an insu cient exposure time to the disinfectant. Accordingly, we recommended revising the cleaning management to follow that of Portner and Johnson [2010]. Hand hygiene and appropriate disinfectant are recommended to control contamination on the examination tables. Concerning hand hygiene, the WHO recommended strict hand hygiene with alcohol and 0.5% (w/v) chlorhexidine [30], while for a low contact time (5 min), a peroxygen compound, is suitable to clean the examination table.

Conclusions
In conclusion, distribution and correlation of CoPS and MRCoPS between the environmental surfaces, veterinary staff, and dogs within the VTH-CU in 2014-2016 were described. It is noteworthy that CoPS were discovered all around the VTH-CU, and some of them acted as antibiotic-resistant pathogens. The suitable sanitation, cleaning management, and annual monitoring must be practised to protect against this bacteria's distribution. In 2014, approximately 389 pet patients per day visited the VTH-CU, divided as 96, 30, 27, and eight at the general medicine, gynaecological, dermatological, and surgical rooms, respectively. The routine of cleaning management in the VTH-CU was comprised as follows. The oors were washed with 2.5% (w/v) quaternary ammonium compound (UMONIUM38 ® ; Laboratoire Huckert's International, Thailand) between 3.30-4.00 PM. The examination tables, stethoscopes, syringe plates, waiting branch, drug cabinet, keyboard, and knot door were cleaned with 0.5% (w/v) UMONIUM38 ® when the items were not in use. The cotton for wound dressing did not change rooms when empty.

Study design and ethic
The disinfectant water in forceps jars, which is 1% (w/v) povidone-iodine (Betadine ® solution; Pathumthani, Thailand), was changed every day in all rooms.

Populations and sample collections
Environmental samples A total of 216 samples were retrieved from the environmental surfaces in nine parts of the VTH-CU; the division of general medicine (M), vaccination room (V), gynaecology (G), dermatology (SW), surgery (S), post-surgery care (P), intensive care unit (ICU), and hallway (lower and upper; H). On the sample collection day, the medicine, vaccination, and dermatological rooms were located in the new building of the VTH-CU. The gynaecology, surgery, post-surgery care, and hallways were in parts of the old building. The criteria of sample collections are described in Table 1. The requirements for room selection were (i) having at least eight pet patients per day and (ii) the room was cleaned at least once per day. One environmental sample was collected two times within the same day; before clinics opened at 7.30-8.00 AM and after the clinics had been cleaned at 3.30-4.00 PM. In each room, samples were collected from ve parts of the oor [31] at the central examination room. In the case of medical instruments, if there was more than one item in that room, then the most used item was chosen for sampling [32]. All surfaces were sampled using sterile cotton swabs. The cotton was dipped into 2 mL of peptone saline dilution (PSD; 100 mg/mL peptone and 850 mg/mL sodium chloride in distilled water). The moistened swab was rolled on the surface, the cotton part cut off into the same tube and kept on ice until cultured.

Veterinary staff samples
Before this experiment started, the written informed consent was obtained from veterinary staff. Study in human was approved by the Ethical Review Committee for Research Involving Human Research Subjects, Health Science Group, CU (137/57), Research and innovation for society, Chulalongkorn University. Samples from veterinary staff were taken from one veterinary nurse per unit and 16 veterinarians who had worked more than 40 h/week at the VTH-CU for two years. Note that routinely, veterinarians and other staff wear a protective mask during working hours. They were asked to provide a sterile cotton swab wipe from their nasal cavities [18]. The sterile cotton swabs were dipped into 2 mL of peptone saline diluent (PSD) in a sterile test tube (No. 9820, Cole-Parmer ® , Thailand) before nasal sampling [18]. After sampling, the cotton part was cut into the same tube, contained on ice and cultured within 2 h.

Dog samples
Before this experiment started, the written informed consent was obtained from dog owners. Study in animal was approved by Institutional Animal Care and Use Committee (113/56), Faculty of veterinary science, Chulalongkorn university. A total of 14 samples were collected from the wound abscess of 14 dogs in surgery, post-surgery care, and the dermatological room on the same day as the staff and environmental samplings. Outpatient dogs with wound infection or dermatitis were chosen from each clinical room with the authorisation of the owner's veterinarian and permission. The moist, sterile cotton swabs were dipped in 2 mL of PSD into a sterile test tube [18], then rolled on the wound sites of dogs, and treated as above.

Identi cation of CoPS and MRCoPS
Samples were cultured within 2 h of the collection in an enrichment culture. A total of 0.1 mL of the suspension was spread on Baird-Parker agar (Difco TM , France) without and with 0.5 µg/mL oxacillin (screening plate) Staphylococci selection. After incubation at 35ºC for 48 h, black Staphylococci-like colonies were selected. Three presumptive Staphylococci colonies were puri ed on tryptic soy agar (TSA; Difco TM , France) and con rmed with the coagulase test. All CoPS were identi ed by primary and secondary biochemical tests, as previously reported [33]. Multiplex-PCR (M-PCR) and matrix-assisted laser desorption/ionisation-time of ight mass spectrometry (MALDI-TOF) were used for characterisation as previously reported [34,35], where S. aureus ATCC 25923, S. pseudintermedius CVMC 0108, S. schleiferi subsp. coagulans CVMC 0208 (canine origin), S. intermedius CVMP 0309, and S. delphini CVMP 0109 were used as control strains. All CoPS were screened using the oxacillin disk diffusion method. The oxacillin (1 µg) and cefoxitin (30 µg) breakpoints were used to con rm MRCoPS, as de ned by the CLSI recommendation [36]. All oxacillin-resistant isolates were characterised as MRCoPS and approved by the presence of the mecA gene by M-PCR as reported [37], and using S. aureus ATCC 25923 and MRSA N315 as the internal control strains.
The conserved fragments of the mec gene complex and ccr gene complex were detected by M-PCR to identify the SCCmec type [39]. To determine the clonal relationship between CoPS and MRCoPS, pulsed-eld gel electrophoresis (PFGE) was performed as recommended [18]. Brie y, bacterial DNA was plugged into Seakem ® agarose (Bio-Rad, USA) and cut with the Cfr9I enzyme. In the case of uncut samples, ApaI was used as the restriction enzyme [8]. The DNA fragments were then separated on a CHEFIII at 6 V/cm, 14 ºC, 120º in a 1% (w/v) pulsed-eld grade agarose gel with switching at 5-15 s for 18 h and 15-60 s for 5 h. Gels were stained with red safe TM Nucleic Acid Staining Solution (Scienti c, NSW, AUS), destained in water, and then digitally captured under UV light [40]. The Lambda Ladder PFG Marker (New England BioLabs, Beverly, Mass) was used as the size ladder. The Bionumeric program associated with the dice coe cient [1.5] was used for dendrogram construction (UPGMA with 1.0% position tolerance). Individual patterns were discriminated based on 80% similarity.   Figure 1