Epidemiology and Infection Impact of a training program on the cambridge.org/hyg surveillance of Clostridioides difficile infection N. Sopena1,2, N. Freixas2,3, F. Bella2,4, J. Pérez2,5, A. Hornero2,6, E. Limon2,7, F. Gudiol2,8, M. Pujol2,8 and the VINCat Clostridioides difficile study group Original Paper 1 Infectious Diseases Department, Hospital Universitari Germans Trias i Pujol, Badalona, Spain; 2VINCat Program, Cite this article: Sopena N, Freixas N, Bella F, Catalonia, Spain; 3Infection Control Team, Hospital Universitari Mútua Terrassa, Terrassa, Spain; 4Infectious Pérez J, Hornero A, Limon E, Gudiol F, Pujol M, Diseases Unit, Hospital de Terrassa, Terrassa, Spain; 5Microbiology Laboratory, CATLAB, Viladecavalls, Spain; the VINCat Clostridioides difficile study group 6 Infection Control Team, Hospital Universitari de Bellvitge-IDIBELL, L’Hospitalet de Llobregat, Barcelona, Spain; (2019). Impact of a training program on the 7 surveillance of Clostridioides difficile infection. Nursing School of the Medicine and Health Sciences Faculty, Universitat de Barcelona and 8Infectious Diseases Epidemiology and Infection 147, e231, 1–7. Department, Hospital Universitari de Bellvitge-IDIBELL, L’Hospitalet de Llobregat, Barcelona, Spain https://doi.org/10.1017/S0950268819001080 Abstract Received: 12 March 2019 Revised: 14 May 2019 A high degree of vigilance and appropriate diagnostic methods are required to detect Accepted: 21 May 2019 Clostridioides difficile infection (CDI). We studied the effectiveness of a multimodal training program for improving CDI surveillance and prevention. Between 2011 and 2016, this pro- Key words: Clostridioides difficile; Clostridium difficile; gram was made available to healthcare staff of acute care hospitals in Catalonia. The program infection prevention; medical education; included an online course, two face-to-face workshops and dissemination of recommenda- surveillance tions on prevention and diagnosis. Adherence to the recommendations was evaluated through surveys administered to the infection control teams at the 38 participating hospitals. The Author for correspondence: Nieves Sopena, E-mail: nsopena. incidence of CDI increased from 2.20 cases/10 000 patient-days in 2011 to 3.41 in 2016 germanstrias@gencat.cat (P < 0.001). The number of hospitals that applied an optimal diagnostic algorithm rose from 32.0% to 71.1% (P = 0.002). Hospitals that applied an optimal diagnostic algorithm reported a higher overall incidence of CDI (3.62 vs. 1.92, P < 0.001), and hospitals that were more active in searching for cases reported higher rates of hospital-acquired CDI (1.76 vs. 0.84, P < 0.001). The results suggest that the application of a multimodal training strategy was associated with a significant rise in the reporting of CDI, as well as with an increase in the application of the optimal diagnostic algorithm. Introduction Clostridioides difficile (formerly Clostridium difficile) is the most frequent cause of healthcare-related infectious diarrhoea, which mainly affects elderly patients presenting risk factors such as the use of antibiotics [1]. Patients are exposed to spores of the organism through contact with the hands of the health personnel or the hospital environment [2]. The diagnosis of Clostridioides difficile infection (CDI) is based on clinical suspicion and on the application of adequate laboratory techniques. In recent years, microbiological diagno- sis has been enhanced by the incorporation of more sensitive tests in diagnostic algorithms such as molecular techniques [3]. The prevention of CDI in the healthcare setting depends on early diagnosis and the application of a set of measures including hand hygiene and the cleaning and disinfection of surfaces, and the proper use of antibiotics [1]. In developed countries, the incidence of CDI, including CDI of nosocomial origin, has risen in recent years [4–6]. However, according to a study conducted in Europe between 2011 and 2013, its incidence varies across countries from 0 to 36.3/10 000 patient-days [7], due in part to the low diagnostic suspicion and the application of suboptimal diagnostic methods; in fact, in that study, only 52% of hospitals used an optimal diagnostic algo- rithm [7]. The training of health professionals is fundamental in CDI prevention pro- grams in order to improve the detection of cases and to increase compliance with preventive measures [8]. The aim of this study was to evaluate the effectiveness of a multimodal training program in © The Author(s) 2019. This is an Open Access improving the surveillance and prevention of CDI in hospitals participating in the nosocomial article, distributed under the terms of the Creative Commons Attribution licence (http:// infection surveillance program in Catalonia (VINCat) [9]. creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited. Methods This multicentre, prospective, quasi-experimental study designed by the VINCat program’s CDI study group assessed the multimodal training program by comparing the baseline phase (2011) with the post-training phase (2016). The items evaluated were the evolution of the number of hospitals participating in the CDI surveillance program, the evolution of 2 N. Sopena et al. the CDI rates reported, the percentage of centres using appropri- (6) Participants’ satisfaction survey regarding the training activ- ate CDI diagnostic methodology and the adequacy of the preven- ities, evaluating: online training, the diagnostic algorithm, tion measures applied at each centre. recommended precautions, information for the patient, the checklist of cleaning and disinfection of surfaces, recom- mended bibliography and feedback on the incidence rates Study setting reported by hospitals. A numerical scale from 1 to 5 was used (1: not very useful, 5: very useful), with scores 4 and 5 VINCat is a nosocomial infection surveillance and control pro- indicating a high level of satisfaction. gram which has been in operation at acute care hospitals in Catalonia since 2006 [9]. The present study was carried out in the hospitals in the VINCat program that reports cases of CDI. CDI surveillance All these hospitals have multidisciplinary infection control Cases of CDI were followed up continuously from 2011 to 2016, teams, and they are stratified into three groups according to using the methodology previously defined by the VINCat size: group 1 (more than 500 beds), group 2 (between 200 and program. 500 beds) and group 3 (fewer than 200 beds). Population under surveillance: adults over 18 years of age attended in any hospital sector (wards, emergency services, out- patient consultations, etc.) who met the definition of CDI. Intervention Asymptomatic patients were excluded, even if they were carriers Between 2011 and 2015, a multimodal training program was of a toxin-producing strain, as were patients with a history of implemented, which comprised the following elements: CDI and those at specific palliative care or convalescence units. Definition of CDI: a patient with acute diarrhoea (defined as (1) An online course for physicians, microbiologists and nursing three or more unformed bowel movements in 24 h), or toxic staff at Catalan hospitals. The course consisted of five parts: megacolon without another known cause, plus one of the follow- epidemiology and clinical manifestations of CDI; microbio- ing: (1) stool sample with a toxin A- or B-positive laboratory logical and non-microbiological diagnosis; transmission result for C. difficile, or detection of genes that encodes toxin by mechanisms, surveillance and prevention measures; treatment molecular testing; (2) endoscopic, surgical or histological examin- of diarrhoea associated with C. difficile; and CDI in special ation confirming the diagnosis of pseudomembranous colitis. populations. Each section included a text, summaries, algo- rithms, images and recommendations for further reading, Classification according to the site of acquisition of diarrhoea and ended with a test that participants had to complete before proceeding to the next section. Hospital-acquired CDI: infection identified >48 h after admission (2) Participation was voluntary. Healthcare professionals from all to the hospital and before discharge. hospitals in Catalonia could participate in the course, regard- Non-nosocomial healthcare-related CDI: infection starting in less of whether they reported or not cases of CDI to the the community or within 48 h of admission, in patients admitted VINCat program. The course was promoted by means of leaf- to a health centre (hospital, nursing home or community health lets, e-mails to the coordinators of the centres and members centre) in the 4 weeks prior to the onset of symptoms. of the Catalan Association of Infection Control Nurses, at Community-acquired CDI: infection starting in the commu- congresses and scientific workshops and on the VINCat pro- nity or within 48 h of admission, with no admission to a health gram website. centre in the last 4 weeks. (3) Participants could access a student and tutor forum where Incidence rates were expressed in the number of cases per 10 they could ask and answer questions. Immediately after the 000 patient-days. study period, they were asked to answer a multi-choice quiz The coordinating centre provided feedback to each hospital on with 10–12 questions, each one with a correct answer and an annual basis, reporting their rates and comparing them with three distractors: one point was awarded for a correct answer the rest of the hospitals (means, medians and percentiles) and and zero points for an incorrect or no answer. The result was according to the size of the hospital. provided at the end of the test. The course was offered five times; in all, 890 health professionals participated. Survey of adherence to recommendations (4) Face-to-face training. Two workshops were conducted for the staff of the hospitals (physicians, microbiologists and nurses) To assess adherence to the recommended diagnostic methods and participating in the VINCat program: the first in 2011, preventive measures, a self-administered questionnaire was used attended by 129 participants, and the second in 2013, in 2011 (baseline measurement) and 2016 (post-training meas- attended by 149 participants. The workshops included feed- urement). A group of infection control experts validated the con- back on the incidence rates of CDI at the participating hospi- tent and structure of the questionnaire for comprehensibility and tals, measures to prevent transmission and discussion of consistency. The questionnaire had two sections, the first with practical cases with regard to surveillance criteria and micro- nine questions on diagnostic aspects and the second with seven biological diagnosis. questions on preventive measures. The 2016 questionnaire also (5) Preparation (and dissemination via the VINCat program incorporated the satisfaction survey on the different training activ- website) of recommended transmission prevention measures, ities and instruments provided. The questionnaire was distributed the microbiological diagnostic algorithm, information on CDI electronically by the coordinating centre, and the results were for patients and relatives, a checklist for cleaning and disinfec- entered into a database for further analysis. The diagnostic meth- tion of surfaces possibly contaminated by C. difficile, and odology was considered optimal when an algorithm including two recommended further reading. or more detection methods for toxigenic C. difficile was applied. Epidemiology and Infection 3 The most frequently used algorithm was the one that detected the 0.001 0.028 0.001 <0.001 value GDH enzyme by enzyme-linked immunosorbent assay as a P screening method, followed by the detection of the A/B toxin either by enzyme-linked immunosorbent assay or by molecular methods. Search intensity (i.e. the attention paid to identifying 2016 (47 hospitals) CDI cases) was considered adequate when the presence of toxi- 3.41 (3.21–3.61) 1.57 (1.43–1.70) 0.95 (0.84–1.06) 0.89 (0.79–0.99) MIa (95%CI) genic C. difficile was investigated in more than 80% of cases of nosocomial diarrhoea. The survey was answered by the person in charge of the infection control team at each hospital, after dis- cussing it with the team members (clinicians, nurses and microbiologists). 2015 (45 hospitals) Statistical analysis 3.62 (3.41–3.83) 1.45 (1.41–1.68) 1.09 (0.97–1.20) 0.99 (0.88–1.10) MIa (95%CI) CDI incidence rates are expressed as the number of CDI cases per 10 000 patient-days, with 95% confidence intervals (95% CI). Descriptive analysis was performed using counts and percentages for categorical and stratified continuous variables. To evaluate the impact of the training program, we compared the number of participating hospitals and the incidence density 2014 (43 hospitals) rate of CDI reported by the hospitals over the period between 3.34 (3.13–3.54) 1.49 (1.35–1.62) 1.07 (0.95–1.18) 0.78 (0.68–0.89) MIa (95%CI) 2011 and 2016. To assess the impact of the intervention on diag- nostic methods and prevention measures, we compared the com- Table 1. Evolution of rates of Clostridioides difficile infection and of the number of hospitals participating in the surveillance program pliance rates observed in the initial and the post-training surveys. In 2016, at the end of the study, CDI incidence rates were com- pared in hospitals that applied an optimal diagnostic method- ology and in hospitals that applied a suboptimal methodology. Likewise, the rates of hospital-acquired CDI were compared 2013 (37 hospitals) 2.65 (2.45–2.85) 1.16 (1.02–1.29) 0.87 (0.75–0.98) 0.63 (0.53–0.72) between hospitals with an adequate search intensity and hospitals MIa (95%CI) with a lower intensity. For the comparison of proportions, the χ 2 test or Fisher’s exact test was used, as appropriate. For the comparison of the annual CDI incidence rates, the Cochran– Armitage trend test was used. A P value of 0.05 or less was considered statistically significant. For the statistical analysis, the EPIDAT V 3.1 program was used. 2012 (34 hospitals) 2.38 (2.19–2.57) 1.11 (0.97–1.24) 0.72 (0.61–0.82) 0.55 (0.46–0.65) MIa (95%CI) Results The number of hospitals reporting CDI rates to the VINCat pro- gram rose from 29 in 2011 to 47 in 2016. Of these, 25 centres completed the survey in 2011 (seven with >500 beds, 13 with 200–500 beds and five with <200 beds) and 38 centres in 2016 (five with >500 beds, 13 with 200–500 beds and 20 with <200 2011 (29 hospitals) 2.20 (2.00–2.40) 0.98 (0.85–1.11) 0.77 (0.65–0.89) 0.45 (0.36–0.54) beds). MIa (95%CI) During the 6 years of the study, 4931 cases of CDI were reported, of which 2196 (44.5%) were hospital-acquired, 1518 (30.8%) non-nosocomial healthcare-related and 1217 (24.7%) community-acquired. Table 1 shows the evolution of the reported incidence of CDI. There was a significant and constant increase in the overall incidence reported by the group of hospitals between Non-nosocomial healthcare-related CDI 2011 and 2016. This increase was also observed for hospital- Mean incidence: cases/10 000 patient-days. acquired CDI, non-nosocomial healthcare-related CDI and community-acquired CDI. This constant and significant upward CDI, Clostridioides difficile infection. trend was also observed when only the hospitals that reported Community-acquired CDI cases since 2011 were considered (Table 2). Hospital-acquired CDI Table 3 shows the results of the surveys on CDI diagnosis and prevention. With regard to the diagnostic methods, a significant increase in the application of the optimal diagnostic algorithm Overall CDI was observed. There was an increase in the search intensity for CDI in patients admitted from nursing homes or from commu- nity health centres and in patients with community-acquired diarrhoeas, as well as greater adherence to most of the preventive a 4 N. Sopena et al. measures recommended, although the differences did not reach P value 0.003 0.028 0.004 <0.001 statistical significance. Focusing solely on the seven hospitals that already used an optimal diagnostic algorithm in 2011 and presented an adequate search intensity, the overall incidence of CDI did not change sig- nificantly over the study period (from 3.51 cases/10 000 patient- 2016 MIa (95%CI) 3.27 (3.03–3.51) 1.47 (1.30–1.63) 0.98 (0.85–1.11) 0.82 (0.70–0.94) days (95% CI 3.09–3.92) in 2011 to 2.96 (95% CI 2.60–3.32) in 2016, P = 0.054); nor were there significant changes in the inci- dence of hospital-acquired CDI (from 1.74 (95% CI 1.45–2.03) in 2011 to 1.43 (95% CI 1.18–1-68) in 2016, P = 0.122). In 2016, the incidence of CDI reported by hospitals that used an optimal diagnostic methodology (3.62 cases/10 000 patient- days, 95% CI 3.34–3.90) was higher than that reported by hospi- 2015 MIa (95%CI) 3.54 (3.28–3.79) 1.58 (1.41–1.75) 1.09 (0.95–1.23) 0.86 (0.74–0.99) tals that used a suboptimal methodology (1.92 cases/10 000 patient-days, 95% CI 1.42–2.41, P < 0.001). Likewise, the inci- dence of hospital-acquired CDI reported by hospitals with an adequate search intensity (1.76 cases/10 000 patient-days, 95% CI 1.55–1.97) was higher than that reported by the hospitals with a lower search intensity (0.84 cases/10 000 patient-days, 95% CI 0.61–1.08, P < 0.001). Thirty-eight hospitals answered the survey on the degree of 2014 MIa (95%CI) 3.30 (3.06–3.55) 1.54 (1.37–1.71) 1.02 (0.88–1.15) 0.76 (0.64–0.87) Table 2. Evolution of rates of Clostridioides difficile infection in the 29 hospitals participating in the surveillance program from 2011 to 2016 satisfaction among health staff in relation to the training tools and documents provided. The rates of responses with a high degree of satisfaction (4 or 5 on a scale of 1 to 5) were 97.2% for the online training course, 94.1% for the annual feedback on the centre’s results, 94.1% for the recommendations for pre- ventive measures on the website, 93.9% for the diagnostic algo- rithm, 87.5% for the recommended further reading, 85.7% for 2013 MIa (95%CI) 2.72 (2.49–2.96) 1.24 (1.09–1.40) 0.91 (0.77–1.04) 0.57 (0.47–0.68) the information document for patients and relatives and 70.0% for the checklist of the cleaning and disinfection of surfaces in contact with patients with CDI. Discussion The training of health professionals is essential to the success of 2012 MIa (95%CI) programs for the prevention of nosocomial infections, including 2.49 (2.28–2.71) 1.14 (1.00–1.29) 0.79 (0.67–0.91) 0.56 (0.46–0.67) CDI [8]. Previous research has reported a lack of knowledge of CDI among healthcare professionals [10]. In the present study, training provided as part of a program of CDI surveillance and control was associated with significant increases in the use of the most sensitive diagnostic algorithm and in the rate of CDI cases reported at hospitals participating in the VINCat program in Catalonia, and with greater compliance 2011 MIa (95%CI) 2.26 (2.06–2.47) 1.01 (0.88–1.15) 0.78 (0.66–0.90) 0.47 (0.37–0.56) with diagnostic and preventive protocols. The training activity was also associated with a rise in the number of hospitals partici- pating in the CDI surveillance program, probably due to the par- ticipation of the hospital infection control teams in the training program. The mean incidence of CDI reported in Catalonia in 2016 (3.41 cases/10 000 patient-days) is similar to the rates in other Non-nosocomial healthcare-related CDI Mediterranean countries and lower than those reported in nor- Mean incidence: cases/10 000 patient-days. thern Europe [11–13]. In the present study, there was a significant increase in all three types of CDI from 2011 to 2016 – CDI, Clostridioides difficile infection. hospital-acquired, non-nosocomial healthcare-related and (above Community-acquired CDI all) in community-acquired infections. This increase in the inci- Hospital-acquired CDI dence of CDI has been reported in other studies and has been associated with the implementation of more sensitive diagnostic techniques [13]. In this regard, our analysis of the evolution of Overall CDI the incidence of CDI reported by hospitals which already used an optimal diagnostic algorithm in 2011 and presented an adequate search intensity showed no increase in incidence over the 6 years of the study, indicating that the increased incidence a Epidemiology and Infection 5 Table 3. Survey of diagnosis and preventive measures (2011 vs. 2016) 2011 (25 hospitals) 2016 (38 hospitals) N (%) N (%) P value CDI study in >80% of episodes of hospital-onset diarrhoea 16 (64.0) 23 (60.5) 0.78 CDI study in >50% of episodes of community-onset diarrhoea with history of antibiotic treatment 18 (72.0) 31 (81.2) 0.37 CDI study in >50% of episodes of diarrhoea in patients admitted from nursing homes or 9 (36.0) 22 (57.9) 0.08 community healthcare centres Only liquid or semiliquid faeces processed for CDI study 25 (100) 37 (97.4) 1.00 All the samples for CDI study sent to the laboratory in a bottle (not in a swab) 24 (96.0) 37 (97.4) 1.00 Optimal microbiological diagnostic methodologya 8 (32.0) 27 (71.1) 0.002 Availability of centre’s own protocol on diagnosis and preventive measures 15 (60.0) 25 (65.8) 0.64 Daily notification by the Microbiology Service of positive results for C. difficile 23 (92.0) 35 (92.1) 1.00 Introduction of contact precautions in all cases of CDI 25 (100) 38 (100) NS Recommendation to wash hands with soap and water instead of alcohol 23 (92.0) 38 (100) 0.15 Duration of precautions at least 48–72 h after the end of diarrhoea 21 (84.0) 34 (89.5) 0.70 Consideration of whether to suspend antibiotic treatment 25 (100) 38 (100) NS Consideration of whether to suspend antacid treatment 1 (4.0) 9 (23.7) 0.07 No post-treatment microbiological control performed in patients diagnosed with CDI 19 (76.0) 34 (89.5) 0.17 CDI, Clostridioides difficile infection. NS, not significant. a The diagnostic method was considered optimal when a diagnostic algorithm including two or more methods of detection of toxigenic C. difficile was applied. observed in the study as a whole is due to the improvements in the in search intensity in patients with previous antibiotic treatment. diagnostic methods recorded in most hospitals. The use of sub- However, a recent study found that only 36% of community- optimal diagnostic methods is a limiting factor in the diagnosis acquired CDIs had received antibiotics [17], and so other risk of CDI; in agreement with other studies [14], we found the factors must be considered [18]. mean incidence of CDI reported by hospitals that applied an opti- Early notification by the microbiologist, which was the norm mal diagnostic methodology (with a two-or-three step algorithm) at the vast majority of participating centres, has been associated to be significantly higher than at hospitals that did not. The with early initiation of antibiotic treatment and preventive mea- incorporation of PCR to the diagnostic algorithm increases the sures [19]. detection of cases, due to its greater sensitivity [12, 15], although Although only two-thirds of the hospitals had their own diag- it is important to restrict the study to liquid or semi-liquid stools nostic and/or preventive protocols, all had preventive measures (as our participating centres did) in order to avoid the diagnosis for the transmission of CDI. The high basal degree of adherence of CDI in asymptomatic patients carrying C. difficile. to the preventive measures made it difficult to obtain a statistically The lack of clinical suspicion is another fundamental factor in significant improvement in the course of the study. Nonetheless, CDI underdiagnoses, which may occur in as many as 25% of there was a non-significant increase in the implementation of cer- cases. In a multinational European study, when CDI was system- tain preventive measures, such as hand hygiene using soap and atically investigated in all faecal samples from patients with diar- water instead of alcohol (since the latter has a lower activity in rhoea referred to the Microbiology laboratory regardless of the presence of spores), and in the number of centres that consid- whether the physician had asked for the test, it was found that ered discontinuation of proton pump inhibitor (PPI) treatment in 23% of the cases of detected CDI, the study of C. difficile after CDI diagnosis, bearing in mind that the epidemiological had not been requested [7]. Probably, in these cases, clinicians association between PPI and CDI suggests that these agents had suspected other causes of diarrhoea. In our study, hospitals should be withdrawn unless they are essential [20, 21]. with adequate search intensity for nosocomial CDI reported a Regarding the degree of satisfaction with the various compo- significantly higher incidence than the rest. nents of the training program, the most valued was the online The incidence of CDI in patients from community health cen- training course, in which a large number of health professionals tres is little known and probably underdiagnosed, although these participated. This was followed by the recommendations on pre- patients may be a source of transmission of infection in acute care cautions for preventing CDI in health centres and the feedback hospitals [16]. The training carried out increased clinical suspi- with information on global and per hospital CDI rates. cion (though the difference was not statistically significant) in One possible limitation of the study is the use of a self- patients referred from nursing homes or from community health administered, non-anonymous survey instead of an external centres. audit. Although a certain bias in some of the responses cannot The incidence of community-acquired CDI practically doubled be ruled out, the respondents were informed that the results of during the period of the study, probably in relation to the increase their centre would not be published individually, thus minimizing 6 N. Sopena et al. the risk of such a bias. Another limitation of the study is the small NS, NF and FB drafted the manuscript. All authors revised critically the manu- number of hospitals that participated in the surveys on adherence script and approved the final version to be submitted. to diagnostic and control measures, which made it difficult to Financial support. This research did not receive any specific grant from obtain statistically significant differences between the two surveys. funding agencies in the public, commercial or not-for-profit sectors. In conclusion, the application of a multimodal training strat- egy was associated with an increase in the application of the opti- Conflict of interest. None. mal diagnostic algorithm, a significant rise in the reporting of Ethical standards. Written informed consent was considered not necessary CDI, and an increase in the number of hospitals participating for the study, given that the study did not involve any patient interventions. in the VINCat program in Catalonia. The confidential information of the patients was protected according to Acknowledgements. The authors thank all the other members of the national normative and was anonymised for the purposes of this clinical VINCat Clostridioides difficile study group at Hospital Universitari Josep study. The participation of health professionals was voluntary. This manu- Trueta: Xavier Salgado, Marta Lora; Hospital Universitari de Bellvitge: script has been revised for its publication by the Clinical Research Ethics Purificación Martos, Jordi Niubó; Hospital Universitari Germans Trias i Committee of Bellvitge University Hospital (ref PR 435/18). Pujol: Gema Fernández, Laia Castellà; Hospital Clínic Barcelona: Silvia Valls, Gimena Santana; Hospital Universitari Mútua Terrassa: Maria López, Esther Calbo; Corporació Sanitària Parc Taulí: Lluís Falgueras, Marta Piriz; References Hospital Universitari del Mar/Hospital de l’Esperança: Juan Pablo 1. Mc Donald LC et al. (2018) Clinical practice guidelines for Clostridium Horcajada, Lluisa Sorlí; Hospital Universitari Santa Creu i Sant Pau: Joaquín difficile infection in adults and children: 2017 update by the Infectious López-Contreras, Maria Àngels Cotura; Hospital Universitari Arnau de Diseases Society of America (IDSA) and Society for Healthcare Vilanova: Alfredo Jover-Sáenz, María Ramírez-Hidalgo; Hospital Universitari Epidemiology of America (SHEA). Clinical Infectious Diseases 20, 1–48. Joan XXIII: Graciano García, Ester Picó; Hospital Verge de la Cinta de 2. Durovic A, Widmer AF and Tschdin-Stter S (2018) New insights into Tortosa: Mar Olga Pérez, Marie France Domenech; Centre Hospitalari transmission of Clostridium difficile infection-narrative review. Clinical Althaia: Dolors Mas, Rafel Pérez; Hospital General de l’Hospitalet: Ana Microbiology and Infection 24, 483–492. Coloma; Hospital de Terrassa: Laura Grau, Marta Andrés; Hospital 3. Crobach MJ et al. (2016) European Society of Clinical Microbiology and Universitari de Vic: Anna Vilamala, María Jesús Martínez; Hospital General Infectious Diseases update of the diagnostic guidance document for de Granollers: Jordi Cuquet, Rosa Vásquez; Hospital Universitari Sant Joan Clostridium difficile infection. Clinical Microbiology and Infection 22 de Reus: Antoni Castro, Simona Iftimie; Hospital Sant Rafael: Irene Sánchez, (suppl. 4), S63–S81. Mercè Clarós; Hospital Santa Caterina: Íngrid Vilaró, Marisa Jofre; Hospital 4. Bauer MP et al. (2011) Clostridium difficile infection in Europe: a Universitari Sagrat Cor: Glòria Garcia, Rosa Coll; Fundació Sanitària hospital-based survey. Lancet 377, 63–73. d’Igualada: Montserrat Brugués, Anna Marrón; Hospital de Mataró: Goretti 5. Freeman J et al. (2010) The changing epidemiology of Clostridium Sauca, Maria Pilar Barrufet; Hospital General de Catalunya: Mariló difficile infections. Clinical Microbiology Reviews 23, 529–549. Marimón; Centre Mèdic Teknon: Sònia Tortajada, Mònica Gallardo; 6. Asensio A et al. (2008) Increasing rates in Clostridium difficile infections Hospital de Barcelona: Montserrat Vaqué, Yolanda Meije; Hospital Sant (CDI) among hospitalised patients, Spain 1999–2007. European Surveillance Joan Despí Moisés Broggi: Ana Coloma; Hospital Delfos: Cristina Berbel, 13, pii: 18943. Irene Garcia; Clínica Girona: Josefina Serrat, Eva Palau; Hospital Comarcal 7. Davies KA et al. (2014) Underdiagnosis of Clostridium difficile across de Blanes: Àngels Garcia, Carme Gallés; Hospital de Campdevànol: Rosa Europe: the European, multicenter, prospective, biannual, point-prevalence Laborda, Ana Martínez; Hospital de Figueres: Maria Carmen Burgas, Pilar study of Clostridium difficile infection in hospitalized patients with diarrhea Girbal; Hospital Dos de Maig: Clara Sala, María José Moreno; Fundació (EUCLID). Lancet Infectious Diseases 14, 1208–1219. Sant Hospital de la Seu d’Urgell: Maria Teresa Ros, Joaquín Angas; Hospital 8. Fayerberg E, Bouchard J and Kellie SM (2013) Knowledge, attitudes and de l’Esperit Sant: Alex Smithson, Maria Teresa Bastida; Hospital Comarcal practice regarding Clostridium difficile: a survey of physicians in an aca- Móra d’Ebre: José Carlos de la Fuente, Montserrat Rovira; Hospital de demic medical center. American Journal of Infection Control 41, 266–269. Palamós: Anabel Martin-Urda, Teresa Aliu; HG Parc Sanitari Sant Joan de 9. Gudiol F et al. (2012) The development and successful implementation of Déu: Vicens Diaz-Brito, Encarna Moreno; Hospital de Sant Celoni: Carme the VINCat Program. Enfermedades Infecciosas Microbiología Clínica 30 Agusti, Inma Peña; Hospital Municipal de Badalona: Jordi Grau, Rosa María (suppl. 3), 3–6. Benítez; Fundació Hospital Residència Sant Camil: David Blancas, Esther 10. Mavros MN et al. (2012) Underestimation of Clostridium difficile infec- Moreno; Hospital Comarcal Sant Bernabé: Sara Martínez; Hospital Sant tion among clinicians: an international survey. European Journal of Jaume de Calella: Roser Ferrer, Elisenda Capdevila; Hospital d’Olot i Clinical Microbiology and Infectious Disesases 31, 2439–2444. Comarcal de la Garrotxa: Ester Sanfeliu, Maria Montserrat Blasco; Hospital 11. Alicino C et al. (2016) Increasing incidence of Clostridium difficile infec- Sant Joan de Déu de Martorell: Helena Monzón, Silvia Sancliment; Hospital tions: results from a 5-year retrospective study in a large teaching hospital Sant Pau i Santa Tecla: Sebastián Hernández, David Castander; Hospital in the Italian region with the oldest population. Epidemiology and Santa Maria de Lleida: Irene Montardit; Pius Hospital de Valls: Marta Sanz, Infection 144, 2517–2526. Susana Sabaté; Clínica de Ponent: Teresita Gesé, Paula Judit Hernández; 12. Khanafer N et al. (2016) Clostridium difficile infection in a French univer- Hospital de Mollet: José María Tricas, Eva Redón; Hospital Comarcal sity hospital. Eight years of prospective surveillance study. Medicine d’Amposta: Manel Panisello, Rosa Maria Ferré; Hospital Comarcal de l’Alt (Baltimore) 95, e3874. Penedès: Maria Cuscó, Lourdes Gabarró; Clínica Quirón: Josep Farguell; 13. Rizzardi K et al. (2018) National surveillance for Clostridioides difficile Hospital del Vendrell: Elisabet Calaf, David Castander; Clínica del Vallès: infection, Sweden, 2009–2016. Emerging Infectious Diseases 24, 1617– Maria Carmen Fernández, Esther Oviedo; ICO l’Hospitalet de Llobregat: 1624. Carlota Gudiol, Adaia Albasanz-Puig; ICO Badalona: Montserrat Jiménez; 14. Bogaty C et al. (2017) Trends in the use of laboratory tests for the diag- Centre Coordinador VINCat: Gonçalo Rodrigues. The authors thank nosis of Clostridium difficile infection and association with incidence rates Angeles Jaen (Fundació Docència i Recerca Mútua Terrassa) for her assistance in Quebec, Canada, 2010–2014. American Journal of Infection Control 45, in the statistical analysis. 964–968. Author ORCIDs. N. Sopena, 0000-0001-6382-9500. 15. Longtin Y et al. (2013) Impact of the type of diagnostic assay on Clostridium difficile infection and complication rates in a mandatory Author contributions. NS, NF, FB, JP and FG were involved in the concep- reporting program. Clinical Infectious Diseases 56, 67–73. tion and design of the study. NF, EL, JP and AH were involved in data acqui- 16. Krishna A et al. (2017) Prevalence of Clostridium difficile infection in sition. FB, NF, NS and MP were involved in analysis and interpretation of data. acute care hospitals, long-term care facilities and outpatient clinics: is Epidemiology and Infection 7 Clostridium difficile infection underdiagnosed in long-term care facility 19. Polen CB et al. (2018) Impact of real-time notification of Clostridium patients? American Journal of Infection Control 45, 1157–1159. difficile test results and early initiation of effective antimicrobial therapy. 17. Chitnis AS et al. (2013) Epidemiology of community-associated American Journal of Infection Control 46, 538–541. Clostridium difficile infection, 2009 through 2011. JAMA Internal 20. Trifan A et al. (2017) Proton pump inhibitors therapy and risk of Medicine 173, 1359–1367. Clostridium difficile infection: systematic review and meta-analysis. 18. Na’amnih W et al. (2017) Incidence and risk factors for community and World Journal of Gastroenterology 23, 6500–6515. hospital acquisition of Clostridium difficile infection in the Tel Aviv 21. Riley TV and Kimura T (2018) The epidemiology of Clostridium difficile Sourasky Medical Center. Infection Control Hospital Epidemiology 38, infection in Japan: a systematic review. Infection Diseases Therapy 7, 912–920. 39–70.