High contamination rates of shoes of veterinarians, veterinary support staff and veterinary students with Clostridioides difficile spores

Abstract Clostridioides difficile is often found in animals and their environment. However, not much has been reported on veterinary clinics environment in terms of the spore load, prevalence and PCR ribotype diversity. The aim of this study was to assess the prevalence of C. difficile on shoe soles of veterinarians, veterinary support staff and veterinary students at the Veterinary Faculty campus. Altogether, 50 shoe sole swabs were collected, and the positivity rates ranged from 86.7% in swabs from veterinarians to 100% in swabs from support staff and students. Non‐toxigenic and toxigenic strains representing toxinotypes 0, IV and XIX were isolated and distributed into 17 different PCR ribotypes, most common being 010, 014/020, SLO002 and 009. PCR ribotype 010 was the most prevalent and isolated from shoe soles sampled in 6/7 areas. Students' shoes had highest ribotype diversity (15/17 PCR ribotypes) but showed a low overlap with ribotype isolated from vets and support staff shoes. Veterinary students are likely the main vectors of C. difficile spores transmissions among veterinary teaching clinics and the hospital.

the carcasses and intestinal contents (Candel-Pérez et al., 2019) but also in the environment of the slaughterhouse (Wu et al., 2017).
Fertilization with manure or compost contributes to contamination of the soil and water (Brown & Wilson, 2018). Dogs' paws were described as a possible source of C. difficile spores in households (Janezic et al., 2018). Moreover, dog's nasal discharge has been lately reported as a possible new source of C. difficile transmission (Rodriguez et al., 2019).
The veterinary clinics and hospitals are also likely to be contaminated, but only few studies report the presence of C. difficile spores in this environment (Rodriguez et al., 2016). The prevalence of C. difficile spores was estimated during the screening for different zoonotic pathogens in 101 veterinary hospitals in Canada (Murphy et al., 2010), in the Large Animal Clinic and Small Animal Clinic of the Ontario Veterinary College (Weese et al., 2000), and lately in Veterinary Hospital of Complutense University of Madrid (Villagómez-Estrada et al., 2019) showing highest contamination loads for isolation areas (29%) and on telephone, keyboards and taps (15%), dog walk entry (100%) and floor (8%), respectively. None of the reports has focused on the role of veterinary team on carriage the spores on their shoes, a sampling site that has recently gained much attention. First and subsequent reports from the United States and Bangladesh (Alam et al., 2014(Alam et al., , 2017Islam et al., 2019) showed varying loads of C. difficile on shoe swabs among all samples taken in the households and other urban areas. High level of shoe sole contamination was noted not only in households but also in hospitals in Slovenia (Janezic et al., 2018(Janezic et al., , 2020.
The aim of this study was to assess the C. difficile contamination rates on the shoe soles at the Faculty of Veterinary Medicine that conducts multidirectional professional activity in the field of veterinary services including didactic activities for the students.

| Sampling
The shoe sole swabs (n = 50) were collected from veterinarians (n = 15), support staff (n = 11) and veterinary students (n = 24) at one of the Faculties of Veterinary Medicine in Poland in February 2020.
The samples were collected at seven different locations ( Figure 1) from veterinarians and staff and in six out the seven locations F I G U R E 1 Schematic presentation of sampling locations. (1. Small Animal Clinic; 2. Food Hygiene Department; 3. Surgery Referral Clinic; 4. Internal Diseases Referral Clinic; 5. Reproduction Referral Clinic (including Ambulance crew); 6. Pathology Department; and 7. Infectious Diseases Referral Clinic) also from students attending the classes. The sponges (3M™) premoistened with NaCl (10 ml per sponge) were used for swabbing.
After the sampling, the sponges were kept at 4°C until testing. A pair of shoes was swabbed with one sponge.
Subsequently, 1.0 ml of enrichment culture was subjected to ethanol shock (each sample in duplicate) by adding 1.0 ml of absolute ethanol for 30 min. After centrifugation at 10,000 rpm for 10 min, pellet was resuspended and inoculated on selective medium CHROMID ® C. difficile (BioMerieux) and incubated anaerobically for 3 days. After incubation, up to 6 suspected colonies were plated on Columbia agar with 5% horse blood (BioMerieux). Putative colonies of C. difficile were first screened based on the colony morphology and identified by mass spectrometry (MALDI-TOF Biotyper System, Bruker).

| PCR ribotyping
Crude DNA was isolated in a 5% (w/v) Chelex-100 Resin (BioRad) solution. The isolates were first screened by analysis of PCR ribotype patterns on Agarose BioReagent (Sigma-Aldrich). Only a single representative for clusters with identical profile and isolated from the same sponge was further PCR ribotyped on Certified Low Range Ultra Agarose (BioRad) after which the banding patterns of the ribotypes obtained were compared to those from in-house library consisting on almost 300 PCR ribotype reference strains with the use of BioNumerics software v7.6 (Applied Maths, Belgium).
A single representative per PCR ribotype from each sample was stored at −80°C.
F I G U R E 2 Representative strains of seventeen PCR ribotypes found among all Clostridioides difficile isolates obtained from shoe soles of veterinarians, supporting staff and students
Amplification of 115bp long insert with primer pair Lok1/Lok3 was performed to confirm non-toxigenic strains (Braun et al., 1996).

| Pulse-field gel electrophoresis (PFGE)
PFGE was performed after restriction with SacII as described by Janezic and Rupnik (2010). Standard protocol was modified, and the increased volume of proteinase K was used in the case of the strains that could not be typed (Fawley & Wilcox, 2002). The results were analysed with the use of BioNumerics software v7.6 (Applied Maths, Belgium).

| RE SULTS
Of 50 analysed shoe sole samples from different groups of volunteers (staff, veterinarians, students) from different departments at Veterinary Faculty, 48 (95%) were presumptively positive for C. difficile and 267 isolates were obtained. Ten of them represented other anaerobic spore-forming bacteria, that is Clostridium butyricum, Terrisporobacter glycolicus, Clostridium clostridioforme and Clostridium cadaveris. The remaining 257 isolates were identified as C. difficile and distributed into 17 different ribotypes ( Figure 2).

| Prevalence of Clostridioides difficile and ribotype distribution in shoe soles across different volunteer groups
Clostridioides difficile was detected in the samples from all groups. In veterinary students, veterinary nurses and technicians all taken samples were positive. Two negative samples were obtained from veterinarians (Table 1) yielding the 86.7% positivity rate in this group.
In 20 samples, more than one PCR C. difficile ribotype was found; two different PCR ribotypes were found in 19 samples (14 students, 3 supporting staff, 2 veterinarians), and three different ribotypes were found in one sample (student) ( Table 2).
PCR ribotypes with the highest number of isolates were 010, present on shoes from all analysed groups, and 014/020 present on shoes from vets and students (Table 1). The most prevalent PCR ribotypes found on the highest number of samples were 010, 014/020 and SLO002.
All groups shared only three of 17 PCR ribotypes. These were 009, 010 and SLO002. All PCR ribotypes isolated from veterinarian shoes were present on shoes from supporting staff and/or students. One PCR ribotype (039) was found only on the support staff shoes. The student shoes were the source of the majority of isolated ribotypes (88.9%) and consequently showed the highest diversity.

| Prevalence of Clostridioides difficile and ribotype distribution in different sampling areas
The samples were collected in seven different areas dealing with companion, farm and wild animals (Figure 1). There was a high PCR ribotype variability noted in terms of the sampling location (Table 3).
Given the high overall positivity rate, it is not surprising that C.
difficile was present in all tested locations. Two to eight PCR ribotypes were found at given location ( Table 3). The only two negative samples were collected at Food Hygiene Department and at Internal Diseases Referral Clinic.
The PCR ribotype diversity at a specific location is in congruence with the number of samples taken (Table 3). The most prevalent PCR ribotype 010 was isolated from all sampling locations except from Infectious Diseases Referral Clinic. The next most prevalent PCR ribotypes 009 and 014/020 were each isolated from the samples taken in 4 different areas. PCR ribotypes 012, SLO002 and SLO259 were present at 3 locations, while 001/072 and 046 were found at two locations (Table 3).
The only binary toxin-positive strain (PCR ribotype 023, toxinotype IV) was isolated from the student's shoes in the Surgery Referral Clinic.

| Similarity between Clostridioides difficile strains across sampling sites and volunteer groups
Selected strains from PCR ribotypes found in several locations, and/or different volunteer groups were typed by PFGE to assess their identity. The results show that isolates within a given ribotype cluster together but can have identical or diverse PFGE profiles (Figure 3). PCR ribotype 014/020 grouped together with ribotype 106. PCR ribotype 012 was distributed between two main branches; while three strains formed coherent (but nonidentical) group, a single strain showed identical PFGE profile as RT 009 strain. This RT 012 strain also had tox-profile, same as ribotype 009. The ribotyping and toxinotyping were repeated for this strain with the same results.
Few clusters of potentially identical strains were detected ( Figure 3). Only one of them includes strains from two PCR ribotypes (009 and 012; cluster 2, described above). Each of the clusters 1, 5 and 6 includes strains isolated from different samples but in the same location and same volunteer group. All other clusters contain strains from two or more locations and two or more volunteer groups. The most numerous cluster of seven strains from PCR ribotype 010 (cluster 7) was spread across four locations and was found on student and support staff shoes.

| D ISCUSS I ON
Studies on C. difficile prevalence in the environment of veterinary settings are not numerous (Madewell et al., ,1995(Madewell et al., , , 1999Murphy et al., 2010;Rodriguez et al., 2016;Struble et al., 1994;Villagómez-Estrada et al., 2019;Weese & Armstrong, 2008), and to date, only one of them included C. difficile on the footwear of the personnel (Weese et al., 2000). Their reported positivity rate for the footwear of medical personnel in veterinary clinics was 12.5%. This is substantially lower compared with our results showing 86.7% of veterinarians and 100% supporting staff carrying C. difficile spores on their shoes. In non-hospital human reservoir, the reported rates are also lower and vary from 26.4% (Alam et al., 2017) through 39.7% (Alam et al. 2014) to 43% (Janezic et al., 2018). In contrast, in hospital environment the footwear positivity rate increases to 62%, but differed between two hospitals and between wards (Janezic et al., 2020). High positivity of shoe soles could imply the ongoing C. difficile outbreak, but the high PCR ribotype diversity did not confirm this. Similar as in hospitals also here in the veterinary setting, the overlap between ribotypes from different clinics/departments was low.
We have not sampled the animals upon arrival or the floors; therefore, we cannot speculate on the source of spore contamination. Floors were shown in other studies to be often contaminated by diverse C. difficile strains. Environmental C. difficile in veterinary teaching hospital was isolated from the floor surface by Weese et al. (2000). The contamination rates were 5.9% and 8% for Small and Large Animal Clinics, respectively. A report from Canada based on results obtained from 101 small animal clinics (Murphy et al., 2010)   Note: RT-PCR ribotype. a some PCR ribotypes included differed toxin gene profiles, and each of them is presented as a separate row. different areas of the veterinary teaching hospital, that is 078, 154, 014 and 039; two of which were found also in our study. However, our results show much higher diversity of PCR ribotypes disseminated on shoe soles from different clinics and areas in the Veterinary Faculty.
Our isolates belonged to ribotypes well known from human, animal F I G U R E 3 Pulse-field gel electrophoreses profiles of Clostridioides difficile isolates from different sampling locations and different volunteer groups. Ten clusters of strains with identical PFGE profile were found indicated with vertical blue line confirm the clonality of strains within the clusters. Although all three groups (veterinarians, veterinary staff and students) had comparable positivity rates, students are the ones who contributed the most to the diversity and several PCR ribotypes were found only on their shoes.
Because of this difference in the diversity of PCR ribotypes between students and employees, we concluded that students are more likely to collect the spores on their shoes outside the clinics/departments.
Thus, students could be considered as a significant source of contamination of the veterinary clinics and hospital from the external (outdoor) environment and vectors of the spore transmission between the departments and to other environments. However, the environmental samples within the clinics/departments or in the outdoor environment were not taken to experimentally support this assumption.
The shoes used in the environment of teaching veterinary clinics and hospital were highly contaminated with C. difficile. PCR ribotypes exhibit high variety and high overlap with common ribotypes found in humans, animals and environment. Several clusters of identical strains were found indicating transmissions within and between the areas. Changing shoes for the time spent in veterinary clinic/hospital practised by the employees seems to significantly reduce diversity of C. difficile ribotypes, probably suggesting also reduced transmission levels. The protocols for changing shoes while entering particular teaching units at Veterinary Faculties should be considered.

CO N FLI C T O F I NTE R E S T
None.

AUTH O R CO NTR I B UTI O N S
JW contributed to study design, experimental work, data curation and manuscript drafting. BW contributed to sample collection. AK contributed to experimental work. MR contributed to study design, supervision and manuscript preparation. All authors have read and approved the final version of the manuscript.

E TH I C A L A PPROVA L
Not applicable.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data available on request from the corresponding author.