Clostridium difficile infections in a university hospital in Greece are mainly associated with PCR ribotypes 017 and 126

RESEARCH ARTICLE Kachrimanidou et al., Journal of Medical Microbiology 2017;66:1774 1781 DOI 10.1099/jmm.0.000623 Clostridium difficile infections in a university hospital in Greece are mainly associated with PCR ribotypes 017 and 126 Melina Kachrimanidou, 1 Olga Tsachouridou, 2, * Ioannis A. Ziogas, 2 Eirini Christaki, 2 Efthymia Protonotariou, 3 Symeon Metallidis, 2 Lemonia Skoura 3 and Ed Kuijper 4 Abstract Purpose. Data regarding the incidence and molecular epidemiology of Clostridium difficile infections (CDIs) in Greece are limited. Methodology. A retrospective study of all laboratory-confirmed CDI cases in a university hospital during a 9-month period. Stool samples from inpatients with diarrhoea were tested with a combined glutamate dehydrogenase (GDH) and toxin enzyme immunoassay (EIA) test, as part of a two-step algorithm for CDI testing. All GDH-positive samples were cultured and isolates were further tested for the presence of toxin genes and characterized by PCR ribotyping. Results. The incidence of CDI in our hospital was 25 per 10 000 hospital admissions. Of 33 CDI cases, 72.7 % were hospitalacquired. Fourteen different PCR ribotypes were identified, of which 017 (21.2 %), 078/126 (15.1 %) and RT202 and RT106 (9 %) were the most prevalent. Most patients had a risk profile of recent antibiotic use, older age and comorbidities. Despite mild CDI clinical characteristics, six cases showed complications and led to 18.2 % mortality. Conclusion. The CDI incidence was comparable to that in other European countries. The hypervirulent PCR ribotype 027 was not found, whereas ribotypes 017 and 126 predominated. Most CDI cases were in patients who used antibiotics, emphasizing that antimicrobial stewardship should be considered as a cornerstone for the prevention of CDI. INTRODUCTION Clostridium difficile infections (CDIs) have emerged over the last few years as a major health problem, and are associated with prolonged hospitalization, morbidity and mortality [1, 2]. Despite this, clinical suspicion of CDI and therefore testing frequency remains low [3, 4]. CDI is defined as the acute onset of diarrhoea with a toxigenic C. difficile when other reasons for diarrhoea have been ruled out [5]. Exposure to antibiotics and contact with the spores of bacterium are the main risk factors for infection [6]. Comorbidities, use of proton-pump inhibitors (PPIs) and gastrointestinal tract operations are also considered to be predisposing factors [7, 8]. Data on CDI in Greece are limited [9]. Therefore, we designed a retrospective study to determine CDI incidence, combined with clinical and epidemiological data, and molecular characterization of C. difficile isolates. METHODS Patients and study design This is a retrospective study of all laboratory-confirmed CDI cases diagnosed at the Microbiology Department of AHEPA Hospital in Thessaloniki, Greece, a 600-bed tertiary hospital. The requirement for informed consent was waived, since the study was retrospective and patients records were anonymized and de-identified prior to analysis. During a 9- month period (June 2014 March 2015) 12 960 patient admissions occurred. Based on clinical criteria, physicians requested the microbiological laboratory to test 144 diarrhoeal stool samples for CDI. Diarrhoea was defined as at least three episodes of unformed stools in 24 h (Bristol stool chart: types 5 7) [10]. Clinical and laboratory data Detailed demographic, clinical (vital signs, physical examination and number of stools) and laboratory [white blood Received 7 August 2017; Accepted 13 October 2017 Author affiliations: 1 First Department of Microbiology, Aristotle University of Thessaloniki, Medical School, Thessaloniki, Greece; 2 First Internal Medicine Department, Infectious Diseases Division, AHEPA Hospital, Medical School, Aristotle University of Thessaloniki, Thessaloniki, Greece; 3 Department of Microbiology, AHEPA University Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece; 4 Department of Medical Microbiology, Center of Infectious Diseases, Leiden University Medical Center, Leiden, Netherlands. *Correspondence: Olga Tsachouridou, olgat_med@hotmail.com Keywords: Clostridium difficile; diarrhoea; risk factors; epidemiology; antibiotics; toxin genes. Abbreviations: CA, community-acquired; CBA, Columbia blood agar; CCEY, cycloserine cefoxitin egg yolk; CDI, Clostridium difficile infection; ESR, erythrocyte sedimentation rate; GDH, glutamate dehydrogenase; HA, hospital-acquired; PPI, proton-pump inhibitor. 000623 ã 2017 The Authors 1774

cell count, erythrocyte sedimentation rate (ESR), creatinine, serum albumin and imaging tests] data were recorded. Patients gender, age, date of admission, time of CDI onset, type of ward, duration of hospitalization, concurrent medication and co-morbidities were obtained from their medical records. The severity of CDI infection was also assessed and classified based on current global recommendations [11, 12]. Characterization of hospital-acquired (HA-CDI) and community-acquired (CA-CDI) infection and cases of recurrent infection was also performed for all cases per guidelines [6]. Isolates and growth media Stool specimens were primarily tested for both glutamate dehydrogenase (GDH) and toxins A/B of C. difficile by an enzyme immunoassay (EIA; C. diff quick check complete). All GDH-positive samples underwent culture onto modified Brazier s cycloserine cefoxitin egg yolk (CCEY) agar (CCEY agar base containing cycloserine cefoxitin supplement and 5 % defibrinated horse blood; Bioprepare, Greece), and the plates were incubated anaerobically at 37 C for up to 5 days. A single colony was subcultured onto a Columbia blood agar (CBA) plate and incubated for 48 h, after which colonies giving the characteristic odour and fluorescence under UV illumination were obtained. For long-term storage, isolates were suspended in nutrient broth containing 10 % glycerol and stored at 80 C [13]. DNA extraction Genomic DNA was prepared from C. difficile grown on blood agar incubated anaerobically for 48 h. A few colonies were emulsified in Tris/EDTA (TE) buffer (Sigma Aldrich Co. Ltd., Gillingham, United Kingdom) and heated at 100 C for 10 min. Debris was removed by centrifugation at 13 500 r.p.m. for 2 min, and the supernatant was removed. DNA was stored at 20 C [13]. Detection of C. difficile toxin genes The presence of TcdA, tcdb, cdta and cdtb genes was investigated accordingly to previously described methods. The absence of the PaLoc was demonstrated using the primers lok1 and lok3, which amplify a 769-bp amplicon in strains without the PaLoc [13 16]. PCR ribotyping All PCR ribotyping of the cultured isolates described in the present study was performed at the Department of Medical Microbiology, National Reference Laboratory for Clostridium difficile, Leiden University Medical Center, Leiden, the Netherlands [17]. PCR ribotyping was performed by a standardized protocol for capillary gel-based PCR ribotyping and an international database with well-documented reference strains [18]. Statistical analysis The data were expressed as mean±standard deviation (SD) for continuous variables and as percentages for categorical data. The Kolmogorov Smirnov test was utilized for normality analysis of the parameters. Bivariate analyses were performed using the Student t-test, ANOVA model, the Chi-squared test and the Fisher exact test to analyse the relation between the outcome variables: CDI (healthcare vs community), severity (mild to moderate vs severe vs severe to complicated), and the quantitative and qualitative demographic and clinical variables, respectively. All of the tests were two-sided, and statistical significance was set at P<0.05. All analyses were carried out using the statistical package SPSS v 17.00 (Statistical Package for the Social Sciences, SPSS, Inc., Chicago, Il, USA). RESULTS Clinical characteristics Over 9 months, stool samples from 144 patients with diarrhoea were sent to the microbiological laboratory for CDI testing. Of the 144 samples tested by GDH/ToxAB, 76.4 % were negative, 15.3 % were only positive for GDH and 8.3 % were positive for both GDH and ToxAB. Out of 144 stool samples suspected to be CDI-positive, 33 (22.9 %) contained C. difficile with 14 distinct PCR-ribotypes. Two GDH-positive samples remained negative for C. difficile culture. Of the 33 isolates, 4 were non-toxigenic. The patients from which C. difficile was isolated were aged 1 91 years, with a mean age of 62 years. The incidence of CDI was 25 per 10 000 hospital admissions. Analysis of the patient demographics showed that most of them fitted the established risk profile, with almost two-thirds being older than 65 years, nearly all having significant comorbidities and all having received antibiotics in the 3 months prior to infection. Hospital-acquired vs community-acquired CDI distribution of C. difficile isolates in wards The incidence of HA-CDI (24 cases, 72.7 % of all CDI cases) in our hospital was higher than that for CA-CDI. With regard to other clinical data, half of the patients had been hospitalized recently, with a mean duration of current hospital stay of 26.5±28.7 days. The mean duration of hospitalization before CDI was diagnosed in patients with HA-CDI was 18.82±16.25 days. There was no difference between HA-CDI and CA-CDI in relation to the reason for hospitalization or past medical history (Table 1). Similarly, no significant difference was found between HA-CDI and CA-CDI concerning previous hospitalization in the last 6 months (62.5 vs 66.7 % respectively, P=1.000). Of the nine patients with CA-CDI, six patients (66.7 %) had experienced a previous hospitalization during the 3 months before the CDI was diagnosed. Variable rates of isolation from different wards were observed, with the highest rate of isolation being in internal medicine wards (18 cases, 54.5 %). The distribution of C. difficile isolates is shown in Fig. 1. 1775

Table 1. Demographics/clinical characteristics of patients at baseline and previous/current antibiotic use stratified by CDI type Community-acquired CDI, onset of CDI outside a healthcare unit or within 48 h following admission to a healthcare facility without residence in/discharge from a healthcare facility within the previous 12 weeks; hospital-acquired CDI, onset of CDI at least 48 h after admission to a healthcare unit; GI, gastrointestinal disease; CDI, Clostridium difficile infection; heart disease, heart failure/coronary disease; blood disease, myelodysplastic syndrome/leukaemia/ lymphomas. Hospital-acquired (n=24, 72.7 %) Community-acquired (n=9, 27.3 %) P-value Age at diagnosis (years) mean±sd 63.24±29.49 67.56±30.66 0.712 Sex n (%) male/female 12 (50.0 %)/12 (50.0 %) 3 (33.3 %)/6 (66.7 %) 0.697 Reason for hospitalization Fever n (%) 16 (66.6 %) 4 (44.4 %) 0.435 Blood disease n (%) 5 (20.8 %) 3 (33.3 %) 0.649 Pulmonary disease n (%) 11 (45.8 %) 1 (11.1 %) 0.113 Heart disease n (%) 3 (12.5 %) 0 (0.0 %) 0.549 GI disease n (%) 0 (0.0 %) 2 (22.2 %) 0.064 Renal disease n (%) 3 (12.5 %) 1 (11.1 %) 1.000 Past medical record Heart disease n (%) 8 (33.3 %) 5 (55.6 %) 0.254 Pulmonary disease n (%) 3 (12.5 %) 1 (11.1 %) 1.000 Renal disease n (%) 8 (33.3 %) 3 (33.3 %) 1.000 Immunosuppression n (%) 6 (25.0 %) 1 (11.1 %) 0.644 Blood disease n (%) 6 (25.0 %) 1 (11.1 %) 0.644 GI disease n (%) 2 (8.3 %) 0 (0.0 %) 1.000 Previous CDI n (%) 2 (8.3 %) 0 (0.0 %) 1.000 Previous hospitalization n (%) 15 (62.5 %) 6 (66.7 %) 1.000 PPI taken n (%) 15 (62.5 %) 6 (66.7 %) 1.000 Previous antibiotic use n (%) 15 (60.0 %) 7 (77.8 %) 0.439 Current antibiotic administration Penicillin n (%) 21 (87.5 %) 6 (66.7 %) 0.348 Cephalosporins n (%) 9 (37.5 %) 4 (44.4 %) 0.704 Penems n (%) 7 (29.1 %) 0 (0.0 %) 0.151 Fluoroquinolones n (%) 5 (20.8 %) 3 (33.3 %) 0.649 Previous antibiotic administration Cephalosporins n (%) 4 (16.0 %) 0 (0.0 %) 0.554 Penems n (%) 0 (0.0 %) 1 (11.1 %) 0.265 Fluoroquinolones n (%) 7 (28.0 %) 3 (33.3 %) 1.000 Antibiotic use and CDIs All patients with CDIs were receiving antibiotic treatment. Analysis of patients medical records showed that in HA- CDI cases, penicillins (87.5 %), cephalosporins (37.5 %), carabapenems (29.1 %) and fluoroquinolones (20.8 %) were the most commonly administered antibiotics prior to CDI (Table 1). The respective consumption rates in CA-CDI cases were penicillins (66.7 %), cephalosporins (44.4 %) and fluoroquinolones (33.3 %). Previous studies have shown that CDIs can occur long after antibiotic exposure, and therefore we looked for a history of antibiotic use in the 3 months prior to diagnosis. The most commonly prescribed class of antibiotic was fluoroquinolones in outpatients and penicillin in hospitalized patients. There was no significant difference between healthcare- and community-acquired infection with respect to the use of separate antibiotics (currently or previously administered) (Table 1). Regarding PPI administration, in most healthcare- and community-acquired CDI cases the patients had been receiving PPIs during the onset of CDI or in the previous 3 months. There was no difference concerning current or previous PPI intake (62.5 vs 66.7 % respectively, P=1.000) and CDI onset (Table 1). Severity of CDI and mortality Most CDI cases (52.8 %) had a mild-to-moderate course of infection, whereas 21.2 % (n=7) experienced a complicated course of infection (Table 2). Six patients with CDI (18.2 %) died during hospitalization and mortality was attributable to CDI. They were all older than 70 years and their infection was HA, except for one case. All patients had significant comorbidities (four with heart disease, two with chronic renal disease, two with blood disease and one with pulmonary disease), although they were immunocompetent and none of them had a confirmed previous CDI or presented 1776

with abdominal pain or bloody stools at admission. The strains causing severe infections with a fatal outcome belonged to PCR ribotypes 017 (two patients), 220, 106 and 050, with one unknown ribotype. The patients with a complicated infection were aged 66.17 (SD±26.92) and had chronic underlying diseases. Only one of these patients was immunocompromised. All of the patients had received antibiotics in the three months prior to admission, most commonly penicillin, carbapenems and fluoroquinolones. PCR ribotypes 017, 126, 202, 220, 106 and 050 were associated with the above complicated cases. There was a statistically significant correlation between severity status and fatal complication of CDI (death), as shown in Table 3. Patients with a complicated course had higher risk of in-hospital death, compared to patients with an uncomplicated course (mild moderate and severe 100 vs 0 %, 0 %; P<0001). No significant difference was found between HA and CA infection in relation to CDI symptoms, severity status and daily mean number of stools (Table 2). Remarkably, higher rates of severe CA-CDI compared to HA-CDI (33.3 vs 25.0 %) were recorded, and one fatal CA- CDI complicated case was observed. Interestingly, patients with mild or moderate CDI presented lower renal impairment during hospitalization monitoring, compared to those with severe and complicated infection (29.4 vs 77.8 %, 62.5 %; P=0.046). CDI treatment Most patients with CDIs (n=20, 60.6 %) received metronidazole 500 mg three times daily per os. Vancomycin 125 mg per os, four times daily, was administered to eight patients (24.2 %). Three patients received both regimens (9.1 %) and only two received fidaxomicin (6.06 %). The mean duration of therapy for all selected regimens was 9.4±5.0 days. Complicated or recurrent cases were equally distributed among patients who were treated with metronidazole or with vancomycin. Molecular characteristics Molecular detection of C. difficile toxin Twenty-nine out of the 33 C. difficile isolates were positive for both tcda and tcdb with PCR, although 7 isolates had a partial deletion in tcda (A-B+). The remaining four isolates were negative for both genes (A-B-, non-toxigenic C. difficile isolates). Despite the unclear clinical relevance of these non-toxigenic strains, we included them in the study as the stool specimens were A-/B-positive by enzyme immunoassay, and the patients responded well to C. difficile-specific therapy. Five of the A + B + C. difficile isolates produced a binary toxin (cdta/cdtb+), whilst the A - B + isolates were all binary toxinnegative. The five binary toxin isolates belonged to RT078/ 126, which were A+B+. In summary, four toxigenic profiles were revealed in this study: A+B+ CDT- (16/33, 48.5 %), A + B + CTD + (5/33, 15.2 %), A - B + CDT - (8/33, 24.2 %) and A-B-CDT- (4/33, 12.1 %). Typing PCR ribotyping was used for further characterization of the 33 clinical C. difficile isolates. Fourteen RTs were Fig. 1. Distribution of PCR ribotypes over time and by hospital wards, (a) monthly (b) according to hospital wards. CD, cardiology; HEM, haematology; ICU CD, coronary disease intensive care unit; ID, infectious disease unit; IM, internal medicine; ORL, otorhinolaryngology; PD, paediatric. 1777

Fig. 1. (cont.) recognized. The most prevalent RT in this study was 017 (21.2 %, 7/33). The next three most common RTs were 126 (12.1 %, 4/33), RT202 and RT106 (9 %, 3/33). The remaining RTs, 078, 012, 005, 024, 050, 070, 137 and 220, were detected sporadically and were represented by up to two isolates. The RT profiles of four isolates (4/33 12.1 %), one of which was toxin A/B-negative, differed from each other and did not match any of the reference RTs available in this study (provided by the Leiden Reference laboratory). Table 2. Symptoms and severity of infection, stratified by CDI type WBC, white blood cell count; mild-to-moderate disease, diarrhoea plus any additional signs or symptom not meeting severe or complicated criteria; severe disease, serum albumin <3 g dl 1 plus WBC15 000 cells mm 3 or abdominal tenderness; severe and complicated disease, any of the following, attributable to CDI: admission to intensive care unit for CDI or hypotension with or without use of vasopressors being required or fever 38.5 C or ileus or significant abdominal distension or mental status changes or WBC35 000 cells mm 3 or <2000 cells mm 3 or serum lactate levels >2.2 mmol l 1 or end organ failure (mechanical ventilation, renal failure, etc.). Symptoms Hospital (n=24, 72.7 %) Community (n=9, 27.3 %) P-value Fever n (%) 17 (70.8 %) 4 (50.0 %) 0.420 Abdominal pain n (%) 10 (41.6 %) 4 (44.4 %) 1.000 Bloody stools n (%) 4 (16.6 %) 2 (22.2 %) 0.644 Serum creatinine rise n (%) 12 (50.0 %) 5 (55.6 %) 1.000 WBC increase n (%) 15 (62.5 %) 5 (55.6 %) 1.000 Severity Mild moderate 12 (50.0 %) 5 (55.6 %) 0.578 Severe 6 (25.0 %) 3 (33.3 %) Severe complicated 6 (25.0 %) 1 (11.1 %) Number of daily stools Mean±SD 5.2±3.0 5.0±2.2 0.886 Median (IQR) 4.0 (3.0) 5.0 (3.0) 0.848 1778

Table 3. Relation of CDI and severity with complication of CDI type Percentage within death status. CDI All of the Greek isolates from RT 017 were only positive for tcdb with a partial deletion in tcda (A - B + CDT - ). Two patients >80 years old diagnosed with C. difficile healthcareassociated infection by RT 017 died. Both patients had severe comorbidity because of heart disease. All of the 126 RT isolates were found to be A + B + CTD +. Even though the presence of all three toxins (A + B + CTD + ) contributed to the more severe CDI course, only a 67-year-old patient in the cardiology ward presented a recurrent CDI. A single A + B + CTD + isolate typed as 078 was recovered from an 89-year-old haematology patient, who improved clinically after metronidazole treatment. Among the 33 cases, 4 were in respect of young children aged 2, 3 and 4 years old. The strains isolated from the three children indicated that their infections were acquired in the healthcare setting by ribotypes RT012 and RT024. DISCUSSION Death: no (n=27, 81.8 %) Death: yes (n=6, 18.2 %) P- value Healthcare n (%) 19 (79.1 %) 5 (83.3 %) 1.000 Community n (%) 8 (29.6 %) 1 (16.7 %) Severity Mild moderate n (%) 17 (62.9 %) 0 (0.0 %) <0.001 Severe n (%) 9 (33.3 %) 0 (0.0 %) Severe complicated n (%) 1 (3.7 %) 6 (100.0 %) Over a 9-month period, we found a CDI incidence of 25 per 10 000 hospital admissions in a Greek tertiary care hospital, which is similar to the rates reported in other European countries [19, 20]. However, a recently published pointprevalence survey of HA-CDI in Greek hospitals reported an overall incidence rate of CDI of 15.7 %, which is lower than the one found in our hospital (22.9 %) [9]. Of the patients diagnosed with CDI, 27.3 % of them acquired their infections in the community. The most frequently found PCR ribotypes were RT017 and RT126/078. Severe and complicated cases of CDI were observed in 48.5 % of patients and the overall mortality rate was 18.2 %. Remarkably, C. difficile toxin A-negative PCR ribotype 017 was the predominant type in our hospital. Historically, RT017 was initially reported in Asia, but it has now been reported worldwide, including in countries such as Poland, Korea, Japan, Bulgaria and Argentina [21], where it has been associated with severe disease [22 25]. Additionally, RT017 has been associated with a high level of resistance to clindamycin and erythromycin [21]. Studies of the outcome and clinical presentation of CDI caused by RT017 strains have discrepant results: some reported no difference from other ribotypes, while others reported more severe disease and worse outcome [26, 27]. The RT017, with its unique toxin profile and unusual global prevalence, has been overshadowed by the global outbreak of the ribotype 027 lineage. A recently published phylogeographic analysis of wholegenome sequencings from RT017 isolates of six continents showed a North American origin for ribotype 017, as has been found for the emerging epidemic RT027 lineage [28]. The authors also concluded that, despite having only one toxin, RT017 strains have evolved in parallel from at least two independent sources and can readily transmit between continents. The toxigenic and so-called hypervirulent PCR ribotype 078 is closely related to PCR ribotype 126 and differs by only one band by electropheris. C. difficile RT078 is one of the prevalent ribotypes in European hospitals, is generally characterized by the presence of all three toxins A + B + CTD + and is considered to be associated with complicated infection [26]. Similarly, ribotype 126 (RT126), which is phylogenetically close to RT078 [29], is also regarded as hypervirulent and has been implicated in cases of CDI. Epidemiological surveillance reports for C. difficile in Spain revealed that the RT078/126 complex is prevalent in human patients [30]. Interestingly, we found that RT126 was encountered more frequently in our hospital. According to Davies and colleagues, countries with low prevalence of ribotype 027 have higher overall ribotype diversity among C. difficile isolates, highlighting the diverse epidemiology of C. difficile across Europe. However, that study reported that ribotypes 078/126 and 017 were among the 10 most commonly isolated ribotypes from countries of southern Europe (Greece, Italy, Portugal and Spain) [31]. Previous studies from Europe have shown that the prevalence of cdta/cdtb in C. difficile human isolates varies from 1.6 to 20.8 %, which agrees with our data [32]. Infection with CDT-positive C. diffficile is associated with higher mortality and recurrence rates. This contrasts with our findings, as none of the CDT-positive patients had fatal outcomes, and only one patient experienced an episode of CDI recurrence [22]. In line with previous studies, most patients with CDI were of advanced age and had been exposed to antibiotics [9]. About three-quarters of cases were HA, in line with previous reports [19, 20, 27, 33]. However, since 2000 there have been several reports of increased incidence and severity of CDI with progressive increase of CA infection, a trend that was also evidenced in our hospital during the study period [34]. This is a finding that deserves more local and national attention. Of all the patients diagnosed with CDI, 27.3 % cases were acquired in the community. However, 63.6 % had experienced a hospitalization in the previous 3 months, suggesting that CA-CDIs could also be healthcare-associated. In the present study, all of the patients used antibiotics during the onset of CDI. The results are consistent with 1779

previous reports, since the prevalence of CDI is increasing globally because of high levels of antibiotic use, with thirdgeneration cephalosporins, fluoroquinolones and penicillin being the major predisposing factors for CDI [35]. Moreover, recent data from Greece confirmed the crucial role of current antibiotic treatment in the onset of CDI in patients with all-cause hospitalization [9]. The excessive use of fluoroquinolones in the Greek community is also illustrated in our data. The mortality rate (18.2 %) in our study was similar that in some other studies [36], while other authors have described even higher rates, related to the severity of CDI and the coexistence of comorbidities [37, 38]. Severe and complicated cases had a higher risk of death compared to mild and moderate cases [39, 40]. Recent reports indicate a remarkable increase in severe cases and deaths related to CDI [41 43]. Identifying patients who are at high risk for severe CDI early in the course of infection may improve outcomes. Our study has several limitations. Apart from its retrospective nature, another important limitation is the fact that the isolates originated solely from one hospital and this may not be representative of the national incidence rate or the general diversity of C. difficile. Furthermore, the relatively small sample size tested might have led to the under-representation of PCR ribotypes that can cause outbreaks in hospitals. To study the epidemiology of an emerging infection like CDI better, data on larger patient populations are needed. Nevertheless, this study is an important forward step towards improving the surveillance of CDI in Greece. Future prospective studies are needed to obtain more detailed information on CDI epidemiology in Greece. To conclude, the majority of C.difficile isolates in our hospital in Greece belonged to RT017, which is widely distributed in Europe but mainly found in Poland and Bulgaria. The emergence of community-acquired CDIs is alarming and probably due to excessive use of antibiotics, and it warrants continuous surveillance to prevent further spread of the toxigenic C. difficile isolates. Most CDI cases suffered from comorbidities and the mortality rates were high despite the absence of hypervirulent 027. This may have contributed to the low recurrence rate observed. However, to be able to draw firmer conclusions about the epidemiology of CDI in Greece, further large-scale multicentre sampling of C. difficile is necessary. Funding information The authors received no specific grant from any funding agency. Acknowledgements We would like to thank all staff members of the 1st Internal Medicine Department and Microbiology Department for their valuable cooperation. Conflicts of interest The authors declare that there are no conflicts of interest. Ethical statement No experimental work with humans or animals was performed. References 1. Kachrimanidou M, Malisiovas N. Clostridium difficile infection: a comprehensive review. Crit Rev Microbiol 2011;37:178 187. 2. Gabriel L, Beriot-Mathiot A. Hospitalization stay and costs attributable to Clostridium difficile infection: a critical review. J Hosp Infect 2014;88:12 21. 3. Chen Y, Glass K, Liu B, Riley TV, Korda R et al. A population-based longitudinal study of Clostridium difficile infection-related hospitalization in mid-age and older Australians. Epidemiol Infect 2017; 145:575 582. 4. Davies KA, Longshaw CM, Davis GL, Bouza E, Barbut F et al. Underdiagnosis of Clostridium difficile across Europe: the European, multicentre, prospective, biannual, point-prevalence study of Clostridium difficile infection in hospitalised patients with diarrhoea (EUCLID). Lancet Infect Dis 2014;14:1208 1219. 5. van Dorp SM, Kinross P, Gastmeier P, Behnke M, Kola A et al. Standardised surveillance of Clostridium difficile infection in European acute care hospitals: a pilot study, 2013. Euro Surveill 2016; 21:29. 6. Cohen SH, Gerding DN, Johnson S, Kelly CP, Loo VG. et al. Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the society for healthcare epidemiology of America (SHEA) and the infectious diseases society of America (IDSA). Infect Control Hosp Epidemiol 2010;31:431 455. 7. Kwok CS, Arthur AK, Anibueze CI, Singh S, Cavallazzi R et al. Risk of Clostridium difficile infection with acid suppressing drugs and antibiotics: meta-analysis. Am J Gastroenterol 2012;107:1011 1019. 8. Arriola V, Tischendorf J, Musuuza J, Barker A, Rozelle JW et al. Assessing the risk of hospital-acquired Clostridium difficile infection with proton pump inhibitor use: a meta-analysis. Infect Control Hosp Epidemiol 2016;37:1408 1417. 9. Skoutelis A, Pefanis A, Tsiodras S, Sipsas NV, Lelekis M et al. Point-prevalence survey of healthcare facility-onset healthcareassociated Clostridium difficile infection in Greek hospitals outside the intensive care unit: The C. DEFINE study. PLoS One 2017;12: e0182799. 10. Saad RJ, Rao SS, Koch KL, Kuo B, Parkman HP et al. Do stool form and frequency correlate with whole-gut and colonic transit? Results from a multicenter study in constipated individuals and healthy controls. Am J Gastroenterol 2010;105:403 411. 11. Surawicz CM, Brandt LJ, Binion DG, Ananthakrishnan AN, Curry SR et al. Guidelines for diagnosis, treatment, and prevention of Clostridium difficile infections. Am J Gastroenterol 2013;108:478 498. 12. Debast SB, Bauer MP, Kuijper EJ, European Society of Clinical Microbiology and Infectious Diseases. European Society of clinical microbiology and infectious diseases: update of the treatment guidance document for Clostridium difficile infection. Clin Microbiol Infect 2014;20:1 26. 13. Griffiths D, Fawley W, Kachrimanidou M, Bowden R, Crook DW et al. Multilocus sequence typing of Clostridium difficile. J Clin Microbiol 2010;48:770 778. 14. Lemee L, Dhalluin A, Pestel-Caron M, Lemeland JF, Pons JL. Multilocus sequence typing analysis of human and animal Clostridium difficile isolates of various toxigenic types. J Clin Microbiol 2004;42: 2609 2617. 15. Braun V, Hundsberger T, Leukel P, Sauerborn M, von Eichel- Streiber C. Definition of the single integration site of the pathogenicity locus in Clostridium difficile. Gene 1996;181:29 38. 16. Stubbs S, Rupnik M, Gibert M, Brazier J, Duerden B et al. Production of actin-specific ADP-ribosyltransferase (binary toxin) by strains of Clostridium difficile. FEMS Microbiol Lett 2000;186:307 312. 17. Paltansing S, van den Berg RJ, Guseinova RA, Visser CE, van der Vorm ER et al. Characteristics and incidence of Clostridium difficile-associated disease in The Netherlands, 2005. Clin Microbiol Infect 2007;13:1058 1064. 1780

18. Fawley WN, Knetsch CW, MacCannell DR, Harmanus C, Du T et al. Development and validation of an internationally-standardized, high-resolution capillary gel-based electrophoresis PCR-ribotyping protocol for Clostridium difficile. PLoS One 2015;10:e0118150. 19. Rodríguez-Pardo D, Almirante B, Bartolome RM, Pomar V, Mirelis B et al. Epidemiology of Clostridium difficile infection and risk factors for unfavorable clinical outcomes: results of a hospital-based study in Barcelona, Spain. J Clin Microbiol 2013;51:1465 1473. 20. Taori SK, Wroe A, Hardie A, Gibb AP, Poxton IR. A prospective study of community-associated Clostridium difficile infections: the role of antibiotics and co-infections. J Infect 2014;69:134 144. 21. Goorhuis A, Legaria MC, van den Berg RJ, Harmanus C, Klaassen CH et al. Application of multiple-locus variable-number tandemrepeat analysis to determine clonal spread of toxin A-negative Clostridium difficile in a general hospital in Buenos Aires, Argentina. Clin Microbiol Infect 2009;15:1080 1086. 22. Barbut F, Mastrantonio P, Delmee M, Brazier J, Kuijper E et al. European study group on Clostridium difficile (ESGCD). Prospective study of Clostridium difficile infections in Europe with phenotypic and genotypic characterisation of the isolates. Clin Microbiol Infect 2007;13:1048 1057. 23. Pituch H, Brazier JS, Obuch-Woszczatynski P, Wultanska D, Meisel-Mikolajczyk F et al. Prevalence and association of PCR ribotypes of Clostridium difficile isolated from symptomatic patients from Warsaw with macrolide-lincosamide-streptogramin B (MLSB) type resistance. J Med Microbiol 2006;55:207 213. 24. Dobreva EG, Ivanov IN, Vathcheva-Dobrevska RS, Ivanova KI, Asseva GD et al. Advances in molecular surveillance of Clostridium difficile in Bulgaria. J Med Microbiol 2013;62:1428 1434. 25. Arvand M, Hauri AM, Zaiss NH, Witte W, Bettge-Weller G. Clostridium difficile ribotypes 001, 017, and 027 are associated with lethal C. difficile infection in Hesse, Germany. Euro Surveill 2009;14: pii=19403. 26. Goorhuis A, van der Kooi T, Vaessen N, Dekker FW, van den Berg R et al. Spread and epidemiology of Clostridium difficile polymerase chain reaction ribotype 027/toxinotype III in The Netherlands. Clin Infect Dis 2007;45:695 703. 27. Chitnis AS, Holzbauer SM, Belflower RM, Winston LG, Bamberg WM et al. Epidemiology of community-associated Clostridium difficile infection, 2009 through 2011. JAMA Intern Med 2013;173: 1359 1367. 28. Cairns MD, Preston MD, Hall CL, Gerding DN, Hawkey PM et al. Comparative genome analysis and global phylogeny of the toxin variant Clostridium difficile PCR ribotype 017 reveals the evolution of two independent sublineages. J Clin Microbiol 2017;55:865 876. 29. Schneeberg A, Neubauer H, Schmoock G, Baier S, Harlizius J et al. Clostridium difficile genotypes in piglet populations in Germany. J Clin Microbiol 2013;51:3796 3803. 30. Alcala L, Reigadas E, Marín M, Fernandez-Chico A, Catalan P et al. Comparison of genomera C. difficile and Xpert C. difficile as confirmatory tests in a multistep algorithm for diagnosis of Clostridium difficile infection. J Clin Microbiol 2015;53:332 335. 31. Davies KA, Ashwin H, Longshaw CM, Burns DA, Davis GL et al. Diversity of Clostridium difficile PCR ribotypes in Europe: results from the European, multicentre, prospective, biannual, pointprevalence study of Clostridium difficile infection in hospitalised patients with diarrhoea (EUCLID), 2012 and 2013. Euro Surveill 2016;21. 32. Søes LM, Brock I, Persson S, Simonsen J, Pribil Olsen KE et al. Clinical features of Clostridium difficile infection and molecular characterization of the isolated strains in a cohort of Danish hospitalized patients. Eur J Clin Microbiol Infect Dis 2012;31:185 192. 33. Silva Júnior M. Recent changes in Clostridium difficile infection. Einstein 2012;10:105 109. 34. Gupta A, Khanna S. Community-acquired Clostridium difficile infection: an increasing public health threat. Infect Drug Resist 2014;7: 63 72. 35. Deshpande A, Pasupuleti V, Thota P, Pant C, Rolston DD et al. Community-associated Clostridium difficile infection and antibiotics: a meta-analysis. J Antimicrob Chemother 2013;68:1951 1961. 36. Hota SS, Achonu C, Crowcroft NS, Harvey BJ, Lauwers A et al. Determining mortality rates attributable to Clostridium difficile infection. Emerg Infect Dis 2012;18:305 307. 37. Pepin J, Valiquette L, Cossette B. Mortality attributable to nosocomial Clostridium difficile-associated disease during an epidemic caused by a hypervirulent strain in Quebec. CMAJ 2005;173:1037 1042. 38. Mitchell BG, Gardner A. Mortality and Clostridium difficile infection: a review. Antimicrob Resist Infect Control 2012;1:20. 39. Abou Chakra CN, Pepin J, Sirard S, Valiquette L. Risk factors for recurrence, complications and mortality in Clostridium difficile infection: a systematic review. PLoS One 2014;9:e98400. 40. Wilson V, Cheek L, Satta G, Walker-Bone K, Cubbon M et al. Predictors of death after Clostridium difficile infection: a report on 128 strain-typed cases from a teaching hospital in the United Kingdom. Clin Infect Dis 2010;50:e77-81. 41. Dallal RM, Harbrecht BG, Boujoukas AJ, Sirio CA, Farkas LM et al. Fulminant Clostridium difficile: an underappreciated and increasing cause of death and complications. Ann Surg 2002;235:363 372. 42. Warny M, Pepin J, Fang A, Killgore G, Thompson A et al. Toxin production by an emerging strain of Clostridium difficile associated with outbreaks of severe disease in North America and Europe. Lancet 2005;366:1079 1084. 43. Arroyo LG, Kruth SA, Willey BM, Staempfli HR, Low DE et al. PCR ribotyping of Clostridium difficile isolates originating from human and animal sources. J Med Microbiol 2005;54:163 166. Five reasons to publish your next article with a Microbiology Society journal 1. The Microbiology Society is a not-for-profit organization. 2. We offer fast and rigorous peer review average time to first decision is 4 6 weeks. 3. Our journals have a global readership with subscriptions held in research institutions around the world. 4. 80% of our authors rate our submission process as excellent or very good. 5. Your article will be published on an interactive journal platform with advanced metrics. Find out more and submit your article at microbiologyresearch.org. 1781