Multi-Country Surveillance of Clostridioides difficile Demonstrates High Prevalence of Spores in Non-Healthcare Environmental Settings

Abstract

Background:

C. difficile spores are frequently isolated from hospital and non-healthcare settings but a worldwide analysis has not been done. The study objectives were to assess C. difficile spore contamination in the hospital and non-healthcare environments across a variety of countries.

Methods:

Field studies assessed hospital vs. non-healthcare C. difficile spore contamination in hospitals, non-healthcare buildings, outdoor environments, and shoes. Swabs were cultured anaerobically for C. difficile and typed using PCR-fluorescent ribotyping. C. difficile contamination by swabbing area and geographic locations were compared

Findings:

A total of 7,857 unique samples were collected primarily from the USA (89%) in addition to 9 other countries. The global prevalence of C difficile from environmental samples was 25.3% and did not differ between countries. In USA based studies, C. difficile contamination rates were similar for healthcare buildings (23.2%), non-healthcare buildings (23.4%), and outdoor spaces (24.7%). Floor samples had significantly higher (p<0.001) C. difficile contamination rate (46.5%) followed by non-floor samples (21.1%), and bathrooms (15.3%). In a comparison of USA to other country samples, C. difficile contamination rates were similar for USA samples (21.5%) compared to rest of world samples (22.3%; p=0.61). The most common ribotypes included F014–020 (15.7%), F106 (12.6%), F010 (8.9%), F027 (8.8%), and F002 (8.1%) and did not differ significantly between USA and non-USA samples. Finally, 546 of 1,218 (44.8%) shoe soles swabbed from the USA were contaminated with C. difficile spores.

Interpretation:

This large surveillance study of several countries demonstrated high prevalence of toxigenic C. difficile in non-healthcare environments with high contamination rates from floors and shoe soles.

Background

Clostridioides difficile is a spore-forming, Gram-positive enteric pathogen that causes a wide spectrum of disease ranging from mild diarrhea to pseudomembranous colitis and death.(1) C. difficile infection (CDI) is a common healthcare-associated infection that affects persons of advanced age, with recent antibiotic exposure, and/or with prior hospitalization.(2, 3) As a strict anaerobe, C. difficile is mainly spread is through the transmission of spores and thus infection control and prevention strategies (i.e., hand hygiene, contact precautions) coupled with antibiotic stewardship have been important measures to reduce the incidence of CDI.(4) Symptomatic CDI infection sheds C. difficile spores that are able to survive on environmental surfaces for months, and so isolation and terminal cleaning of CDI patient hospital rooms has become a standard practice.(5) However, there has been a recent shift in the global CDI epidemiology toward an increasing proportion of community-associated CDI (CA-CDI) cases.(6) In the United States (US), CA-CDI cases increased from 41% to more than 50% of cases between 2014 and 2018.(7, 8) Similarly, CA-CDI cases increased in Australia from 11.7% in 2010 to 26.8% in 2014.(9) This rise in community-associated cases led to the possibility that reservoirs of C. difficile spores exist outside the healthcare environment. This is supported by data from the United Kingdom that demonstrated over 65% of isolates within a geographic region were genetically distinct.(10)

Our research group previously identified C. difficile spore presence (contamination) in multiple non-healthcare settings in the Houston, Texas area, including private residences, public parks, and restaurants.(11, 12) As C. difficile is often isolated from soil or soil-related products(13), farm animals (14) and its spores are known to be easily spread by wind or human movement,(15, 16) these non-healthcare locations may serve as potential environmental reservoirs. Therefore, we hypothesize that the C. difficile contamination rates may be similar across the globe, but likely with regional differences in strain types. In addition, we demonstrated high rates of toxigenic C. difficile contamination on shoe soles,(11) consistent with other non-US, observational studies isolating C. difficile from shoes.(17, 18) This led to our hypothesis that shoe sampling may serve as an effective surrogate sampling method to identify regional differences in and the emergence of C. difficile strain types. We sought to address these hypotheses using a set of environmental C. difficile isolates obtained from three separate field studies across multiple countries. The objectives of this study were to 1) assess C. difficile contamination rates in the healthcare, non-healthcare (community), and outdoor environments in Texas, 2) compare C. difficile contamination rates in the US versus a variety of countries, and 3) assess the utility of shoe sole sampling as a method to identify regional differences in C. difficile strain types.

Methods

Field Studies

All samples were collected during field studies conducted between 2014–19 using procedures as detailed below. Sample swabbing was performed using pre-sterilized gauze lightly soaked with normal saline (0.85% NaCl). Field investigators swabbed an area approximately the size of their palm for each sampling location (approximately. 3 square inches) A negative control was included for each day of swabbing to monitor for sample cross-contamination during collection or storage. To assess C. difficile contamination rates between healthcare vs. non-healthcare settings, samples were collected from urban areas around Texas, USA with additional non-healthcare samples collected from across the US. To compare contamination rates between the US and other countries, environmental samples were collected from North America (Mexico), Central America (Peru, Guatemala), South America (Brazil), Europe (France, Germany, Italy), and Asia (Taiwan, India). For both objectives, samples were classified based on environmental location as ‘healthcare’ if collected from publicly accessible areas of healthcare systems (i.e., hospitals or emergency clinics), ‘non-healthcare’ if obtained from hotels, restaurants, chain stores, or public buildings (i.e., universities, museums, libraries), or ‘outdoor’ if collected from an outdoor public space (i.e., parks, metro stops). International samples were only collected from non-healthcare and outdoor environments. Samples were further subclassified based on swabbing area as ‘floor’, non-floor, or bathroom.

Shoe Sample Collection

Shoe bottoms from consenting participants in the community were swabbed as described above in 25 US states to assess for differences in contamination rates and RT distributions based on geographic location. All participants were community-based with no recent exposure to the healthcare setting. Additionally, paired sampling of shoe tops and soles was conducted in a subset of participants to investigate the same differences based on the shoe part sampled. The entire shoe top or shoe sole was sampled. No participant-identifiable data was collected for any of the shoe studies. These studies were deemed exempt from human subjects review by the Committee for the Protection of Research Subjects at the University of Houston.

Microbiologic Procedures

After collection, swabs were mailed back to the research labs at the University of Houston College of Pharmacy. Swabbed samples were enriched in brain heart infusion broth with 0.05% sodium taurocholate (Sigma Chemicals) and incubated anaerobically at 37°C for up to 5 days as previously described.(12) One milliliter of broth culture from each sample was centrifuged to concentrate the cells with the resulting pellet suspended in 100μL normal saline, plated onto selective cefoxitin-cycloserine-fructose agar (CCFA) (Anaerobic Systems, Morgan Hill, CA), and anaerobically incubated at 37°C for 48–72 hours (Forma Anaerobic System Model 1025/1029). Suspected C. difficile colonies were tested using latex agglutination reagent (Oxoid, Hampshire, England). Each batch of samples was processed with a positive and a negative control. The presence of toxin genes were determined using multiplex PCR.(12) All identified C. difficile strains underwent fluorescent ribotyping as previously described (11, 19). Reference strains included fluorescent (F) PCR ribotypes (RT) F027, F001, F053–163, F002, F014–020, F017, and F078, respectively. This technique does not distinguish between RTs F053 and F163, RTs F014 and F020, and RT F078 and F126. Therefore, these are reported as combined pairs (i.e., F053–163, F014–020, and F078–126). Ribotypes include F010, a non-toxigenic strain. F denotes linkage of our library with other known ribotypes, FP denotes a ribotype unique to our collection

Statistical Analysis

C. difficile contamination rates were calculated and compared first based on environmental location (i.e., healthcare versus non-healthcare); samples classified as non-healthcare or outdoor were combined for this analysis. Next, contamination rates were compared by swab area subclassification and both comparisons were analyzed using the Pearson’s chi-squared test. As samples were only collected from non-healthcare and outdoor settings internationally, healthcare samples were omitted from analysis when comparing contamination rates between the US and international environments. Contamination rates between shoe soles and shoe tops were analyzed similarly using the Pearson’s chi-squared test. A p value <0.05 was considered statistically significant. Descriptive statistics were used for RT distributions. All analyses were conducted using R software version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria) and SAS version 9.3 (SAS Institute, Cary, CN).

Role of the funding source

This study was funded by the National Institute of Health. The NIH did not have any role in the study design, collection, analysis, interpretation of data, writing of the report, or in the decision to submit the paper for publication

Results

A total of 7,857 unique samples were collected from the US (n=6,966, 89%) and nine other countries (n=891, 11%). The overall prevalence of C. difficile contamination in the environment was 25% (n=1,988).

Healthcare vs. Non-healthcare Contamination Rates

A total of 4,668 samples were collected from 3,493 healthcare (75%), 458 non-healthcare (10%), and 717 outdoor (15%) environmental samples in the US. Most samples were collected from non-floor (n=3,992) locations, followed by floor (n=473) and bathroom (n=203) locations. Overall, 1,094 (23%) were positive for C. difficile and contamination rates were similar among the three environments (Figure 1a). Floor samples had a significantly higher (p<0.001) contamination rate (47%) than non-floor (21%) and bathroom (15%) samples (Figure 1b). Five hundred and ninety-one samples had ribotyping completed with notable differences between healthcare and non-healthcare environments, including emergence of a RT FP310 in the non-healthcare environments that was not matched to our reference library (Figure 3a).

Figure 1.

C. difficile contamination rates in healthcare and non-healthcare settings based on environmental location (1a) and swabbing area subclassification (1b).

Figure 3.

Ribotype distribution between healthcare and non-healthcare environments (3A) and in US and international environments (3B)

US vs. International Contamination Rates

To compare contamination rates, 3,932 environmental non-healthcare samples were collected from the US (n=3,041, 77%) and nine other countries (n=891, 23%). Most of these were collected from outdoor environments (n=2,807, 71%), while the remainder (n=1,125,29%) were collected from non-healthcare, public buildings. A total of 276 (7%) samples were swabbed from floors, while 3,434 (87%) and 222 (6%) were collected from non-floor surfaces and bathrooms, respectively. Overall, 854 (22%) of samples were positive for C. difficile (Figure 2). The C. difficile contamination rates were similar between the US and international environments (22% vs. 22%, respectively; p=0.61), but were higher in outdoor versus non-healthcare indoor environments (23% vs 18%, respectively; p<0.001). Floor samples had a significantly higher (p<0.001) contamination rate (36%) followed by non-floor (21%), and bathroom (16%) samples. There were no significant differences based on swabbing area noted between US and international samples. The most common RTs overall included F014–020 (16%), F106 (13%), F010 (9%), F027 (9%), and F002 (8%) and did not differ significantly between US and international samples (Figure 3b).

Figure 2.

Global differences in C. difficile environmental contamination rates based on environmental location (2a) and swabbing area subclassification (2b).

Shoe Sole Sampling

In total, 1,218 shoe soles were swabbed from 25 US states including most frequently Texas (n=776), Pennsylvania (n=150), Georgia (n=73), Maryland (n=59), Colorado (n=55), and Kansas (n=55). Overall, 546 (45%) shoes were contaminated with C. difficile. Rates of contamination ranged from 24% to 75% amongst states studied (p<0.001). Of these, 240 isolates were ribotyped: 43 (18%) were F106, 40 (17%) were F014–020, and 25 (10%) were F002. No sampling period showed a predominance of any single RT.

To compare the contamination rate based on shoe sampling location, 100 pairs of shoes were swabbed to yield 200 total shoe samples (n=100 shoe soles and n=100 shoe tops). Significantly more shoe soles (n=49, 49%) than shoe tops (n= 17, 17%) were positive for C. difficile (p<0.001). Of 39 isolates ribotyped, 34 (87%) were from shoe soles and 5 (13%) were from shoe tops. Although there was no statistically significant difference between RTs isolated from shoe soles versus tops (p=0.13), RT F014–020 was the most common RT isolated from shoe bottoms (33%) while RT F106 was the most commonly from shoe tops (20%).

Discussion

Although CDI is often considered a hospital-acquired infection, rates of CA-CDI are increasing.(6) The pathogenesis of CDI begins by ingestion of C. difficile spores and subsequent germination in a susceptible host..(20) C. difficile spores can survive in extreme conditions with varying temperatures and are likely to be highly prevalent in non-healthcare environments as well.(21, 22) We have previously demonstrated a high rate of C. difficile contamination in community settings surrounding Houston, Texas.(11, 12, 23) These prior studies provided the framework and expertise to expand our environmental surveillance to a variety of countries. Using multiple distinct field studies, we found similar rates of C. difficile contamination in non-healthcare and healthcare environmental surfaces. Contamination rates were similar between the US and other countries with comparable but differing distributions of RTs likely reflecting geographic distinctions. For example RT017, a common Asian C. difficile strain was observed more frequently in samples from Taiwan and India than other locations.(24) Floors were more likely to be contaminated than non-floor or bathroom surfaces. Finally, the rate of shoe sole contamination was similar to that from environmental floor samples and may be a useful sampling location surrogate to study the ecology of C. difficile contamination in a specific environment. To the best of our knowledge, this is the largest surveillance study to investigate C difficile environmental contamination and RT composition across various settings. Strengths of this study include a large sample size with pre-defined swabbing classifications and paired molecular typing to assess the global RT distribution of environmental C difficile.

There are several potential changes in surveillance and infection control practices that could be studied based on these findings. We have previously demonstrated that shoe soles were effective at transporting bacteria and, in a simulated environment, could contaminant hospital beds and other areas where hand-to-mouth transmission would be likely.(25) Other vectors such as wheelchairs have also been shown to increase risk of oral spore ingestion.(16) As floors were shown to be highly contaminated in our current study, it seems logical that an increased emphasis on floor cleaning in the areas surrounding patients at high risk for CDI could potentially decrease CDI risk. As C. difficile spores were highly prevalent in the community setting, antimicrobial stewardship and infection control efforts for high-risk patients in the community should also be studied. Additionally, studies investigating antibiotic exposure as a CDI risk have focused on high-risk antibiotics in hospitalized patients(26) and more studies are needed to identify high-risk patients in the community. Finally, swabbing shoe soles as a surveillance method is intriguing due to ease of collection and our observed high rates of positivity. However, multiple studies are needed prior to being able to use shoe sole sampling as an epidemiologic tool, including isolate comparisons of those identified on shoe soles and the environment they came from.

This study has certain limitations. We used a convenience sample to identify geographic locations to swab. As many of our field studies occur in Texas, we oversampled this area compared to the rest of the US or other countries. However, with almost 1,000 swabs collected at geographically distinct areas outside of the US, this still represents a large international sample size. Due to cost considerations, we only conduct ribotyping for one isolate per colony as part of our standard laboratory procedures. It is possible that more than one RT was present per sample, which may be especially important for future shoe studies as shoe soles may have multiple RTs present. We identified non-toxigenic strains (RT F010) as well as toxigenic strains in this study to better understand the complete ecology of C. difficile in these environments. Not all samples were successfully ribotyped which is a potential source of study bias. An analysis into the virulence of these environmental toxigenic strains compared to clinical strains was not conducted but is an area for future study. Finally, our sampling method only allows for a qualitative result of present or absence of C. difficile spores. The quantitative amount of spores in the environment will require further study

In conclusion, our large surveillance study demonstrated a high prevalence of toxigenic C. difficile in non-healthcare environments, driven by a high contamination rate from floors. C. difficile contamination rates and RT distributions were largely similar between US and other countries. Shoe soles are highly contaminated with C. difficile and may be a useful proxy to test for emerging RTs in the environment following future studies.

Importance.

We investigated community environmental contamination of Clostridium difficile spores based on reports of increasing incidence of community-acquired C. difficile infection and small case series that suggested a wider source of potential C. difficile environmental contamination. We conducted a multi-country project to study the community environmental contamination of C. difficile spores. Our study provides new evidence that non-healthcare public areas have high rates of toxigenic C. difficile contamination. C. difficile infection is often thought to result from recent acquisition of a C. difficile isolate in a health care center generally by direct or indirect horizontal transmission of the pathogen from patient to patient. These findings suggest that interventions beyond isolation of symptomatic patients should be targeted for prevention of C. difficile infection.

Highlights.

• This global project studied C. difficile community environmental contamination

• 7,857 unique samples were collected from the USA (89%) and 9 other countries.

• The global prevalence of C difficile from environmental samples was 25.3%.