INTRODUCTION
Clostridium difficile (CD) is the leading cause of healthcare-associated diarrhea and has a high attributable mortality rate.Reference Davies, Longshaw and Davis 1 Characteristically, this infection occurs in elderly patients with comorbidities in whom the intestinal flora has been disrupted by previous antibiotic use.Reference Loo, Bourgault and Poirier 2 A recent national survey in the United States reported that CD was responsible for almost 500,000 infections annually and was associated with ~29,000 deaths in 2013.Reference Lessa, Mu and Bamberg 3 In a survey of 183 hospitals in 2010, CD was the most frequently reported healthcare-associated microorganism.Reference Magill, Edwards and Bamberg 4
Since 2001, the global emergence of the polymerase chain reaction (PCR) ribotype NAP1/027 C. difficile (CD-027) has been responsible for multiple outbreaks and increased disease severity in North America.Reference He, Miyajima and Roberts 5 This strain has spread to South America, Asia, Australia, and throughout Europe.Reference Valiente, Cairns and Wren 6
In France, the first large outbreak caused by this strain was described in the Nord-Pas- de-Calais region during 2006–2007, with a crude mortality rate close to 30%.Reference Birgand, Blanckaert and Carbonne 7 Since then, small sporadic clusters have been reported in southern Europe.Reference Di Bella, Paglia, Johnson and Petrosillo 8 , Reference Rodríguez-Pardo, Almirante and Bartolomé 9 In Marseille public hospital point-of-care laboratories, the first confirmed case of CD-027 was diagnosed in March 2013.Reference Lagier, Dubourg and Cassir 10 Herein, we describe the related large regional outbreak, and we analyze the role of long-term care facilities (LTCFs) in the spread of the disease.
METHODS
Study Population and Definition of CD Infection Cases
This study was conducted during a CD-027 outbreak in the Provence-Alpes-Côte-d’Azur region, France. The definition of CD infection (CDI) was based on the standard clinical and microbiological criteria given in the guidelines from the European Center for Disease Prevention and Control.Reference Crobach, Dekkers, Wilcox and Kuijper 11 Confirmed CD-027 cases were defined as CDI plus PCR-positive for CD-027. Probable but unconfirmed CD-027 cases were defined as CDI plus a positive immunoenzymatic assay (EIA) for Tox A/B plus possible contact with a confirmed CD-027 case. Only patients with probable or confirmed CD-027 infection were reported at the regional level. Clinical data were collected by the infection control team using a standardized questionnaire. Healthcare facilities (HCFs) included acute care hospitals (ACHs) and LTCFs. In this study, we referred to LTCFs as typical skilled nursing facilities for adults in which the average resident is elderly, requires long-term daily care for at least 6 months, and/or is likely to suffer from multiple chronic medical conditions or physical impairments. Residential care homes (RCHs) are referred to as nursing homes providing mostly non-medical custodial care for elderly adults who cannot live independently.
Characterization of CD Isolates
The Xpert C. difficile Epi PCR assay (Cepheid, La Serre, France) was used from April 2012 in the 2 point-of-care laboratories located in the Timone and North hospitals in Marseille.Reference Cohen-Bacrie, Ninove and Nougairède 12 This multiplex real-time PCR assay detects the genes encoding toxin B (tcdB) and the binary toxin (cdt) as well as the tcdC gene deletion at nt117, identifying CD-027.Reference Babady, Stiles, Ruggiero, Khosa, Huang and Shuptar 13
Spatio-temporal Analysis
The study area included the administrative region of Provence-Alpes-Côte-d’Azur (4.9 million inhabitants) with 351 HCFs and 614 RCHs. The study was conducted from January 2013 to December 2015. The outbreak occurred from January 2013 to October 2014, and we considered the endemic phase as the period from January 2015 to December 2015. Overall, the cases were reported by 25 HCFs and RCHs, all located in the same department (Bouches-du-Rhône).
Infection Control
All patients with diarrhea, whatever the age and context (community- or healthcare-acquired) were tested using the Xpert C. difficile Epi PCR assay (Cepheid). Stringent infection control measures were applied as a prophylactic precaution until the PCR results were obtained. Most of the patients who received a diagnosis of infection due to CD-027 were transferred to the infectious diseases unit in the North hospital, where the healthcare personnel are accustomed to applying stringent infection control measures in accordance with the CDC infection control guidelines.Reference Dubberke, Carling and Carrico 14 The compliance of the unit personnel was surveyed daily by the infection control team. When patients were transferred to other wards for complementary examinations, a member of the infection control team ensured that strict isolation precautions were followed and that appropriate environment cleaning was performed. Antibiotic stewardship was also reinforced during the outbreak period by re-evaluating all prescriptions in the dedicated infectious diseases unit. Transfers to other HCFs or RCHs were also restricted, and returning home was favored.
After the regional alert, all healthcare facilities were informed and urged by the regional health agency to send their specimens to the central laboratory of Marseille public hospital point-of-care laboratories and/or to the national reference laboratory for anaerobic bacteria and C. difficile (Saint-Antoine Hospital, Paris, France). To the best of our knowledge, only these 2 laboratories ensured PCR-ribotype identification at the regional level, and no other laboratory reported any case of CD-027.
Statistical Analysis
Data analysis was performed using SPSS software (IBM, New York). Incidence rates were calculated as the ratio of the number of cases per 10,000 bed days. The χ2 test was used to compare the rates of unconfirmed CD-027 cases. All tests were considered significant at P<.05.
RESULTS
From January 2013 to October 2014, we identified a large CD-027 outbreak in the Bouches-du-Rhône department (southeastern France). Overall, 19 HCFs reported 144 CDI cases over a 22-month period, including 112 laboratory-confirmed CD-027 cases and 32 probable CD-027 cases. The incidence rate per 10,000 bed days was lower in LTCFs than in ACFs (0.05 vs 0.14; P<.001). After centralization of CD testing, the rate of confirmed CD-027 cases (compared with probable CD-027 cases) from LTCFs or RCHs increased significantly (69% vs 92%; P<.001). In total, 52 case patients (36.1%) had been hospitalized in the same long-term care facility (LTCF-1). Among them, 17 (30.9%) were transferred to other HCFs or RCHs, likely contributing to the spread of the epidemic strain. The epidemic curve (Figure 1) shows the timing of case patients and the prevention campaign. Overall, the outbreak developed over a 22-month period. The patient with the index case had no history of recent travel and was identified in January 2013 at LTCF-1. He was the first symptomatic patient in LTCF-1, positive for CD by EIA after a 3-month period without any post-antibiotic diarrhea. This patient’s diagnosis corresponded to the criteria of probable CD-027 case because he had been in contact with the first confirmed CDI-027 case hospitalized in the same LTCF-1 and diagnosed in February 2013 (ie, sent to the national reference laboratory for anaerobic bacteria and C. difficile). The epidemic curve exhibits 2 major peaks: the first from January 2013 to November 2013 with the highest number of cases in September 2013 (17 cases), and the second from January 2014 to October 2014 with the highest number of cases in March 2014 (14 cases). Spatial analysis of the reported cases highlighted the origin of the first part of the outbreak that corresponded to LTCF-1 (Online Supplementary Figure S1). Subsequently, the spread remained geographically limited to an urban area around LTCF-1. Overall, 10 clusters of >1 confirmed CD-027 case occurred in distinct HCFs or RCHs and formed part of the outbreak. Following a 2-month period with no cases, new 027 CDIs were identified. Overall, 22 confirmed CD-027 cases were reported from January 2015 to December 2015 and involved 6 additional HCFs. At the regional level, this time period corresponded to the endemic phase of the outbreak.
FIGURE 1 Confirmed (n=143) and probable (n=32) cases of PCR-ribotype NAP1/027 C. difficile (CD-027) infections in the Bouches-du-Rhône department (southeastern France) from January 2013 to December 2015.
Regarding confirmed CD-027 patients, the sex ratio and the median age were 0.53 and 84.2 years, respectively. The 30-day crude mortality rate was 31.2%. Most patients (95.7%) had received antibiotics within 3 months prior to the CD colitis diagnosis, including cephalosporins (30.5%), fluoroquinolones (29.2%), penicillin combinations (16.7%), carbapenems (8.3%), penicillins (10.4%), metronidazole (6.2%), and others (12.5%). In addition, 3 case patients (2.7%) fulfilled the criteria for community-acquired CDI, while 109 (97.3%) were classified as healthcare-associated belonging to a cluster of a minimum of 2 confirmed CD-027 cases.
During the study period, we tested 11,122 stool samples at the point-of-care laboratories of Marseille for the presence of C. difficile toxin(s) by PCR; among them, 203 (2%) were confirmed CD-027. While the rate of patients with CD-027 (compared with all patients tested) decreased significantly (P=.03), the rate of patients with non-027 CDI remained stable (Online Supplementary Figure S2).
In March 2013, the Marseille public hospital point-of-care laboratories launched an alert that led to strengthened screening and reinforcement of hospital infection control measures. In October 2013, a regional prevention campaign was launched to help infection control and medical staffs to detect CDI cases early and to promptly implement barrier precautions, especially in the LTCFs. Enhanced control measures and specific disinfection procedures against CDI were recommended, including isolation precautions according to international standards, reinforcement of hand hygiene using alcohol-based hand rub solutions following hand washing with liquid soap, wearing gloves, using dedicating equipment, environmental cleaning with hypochlorite solutions (0.5%), and using a specific process for waste management. Because cases were still occurring after the first bundle of measures had been implemented, the campaign was reinforced in February 2014 with a focus on the implementation of regional cohorting to the reference infectious disease unit (whenever possible) with isolation in single rooms and dedicated healthcare personnel. Infection control audits were enhanced through surveillance of compliance with hand washing rules and use of gloves when handling body fluids, especially stools, in patients with documented CDI in the dedicated infectious disease unit.
DISCUSSION
We report a large outbreak of CDI caused by the PCR-ribotype NAP1/027 C. difficile strain in the Bouches-du-Rhône department of southeastern France. The prompt initial alert was facilitated by the accurate identification of the CD-027 epidemic clone in our laboratory and subsequent notification of case clusters or severe cases by HCFs.Reference Eckert, Coignard and Hebert 15 The rapid diagnosis of CDI is important to optimize patient management and infection control.Reference Crobach, Dekkers, Wilcox and Kuijper 11 Moreover, access to C. difficile typing is crucial to detect and control CD outbreaks according to the strains’ ribotype specificities.Reference Barbut, Mastrantonio and Delmée 16 In the present outbreak, the rate of probable but unconfirmed CD-027 cases decreased significantly after implementation of a regional campaign to centralize stool sample testing and to ensure PCR-ribotype identification. This outbreak highlighted the lack of molecular assays for C. difficile toxin detection in most of the peripheral laboratories.
In our report, the source of the outbreak was a single LTCF with secondary spread to other HCFs, RCHs, and the community. Notably, we found no relationship between the previously presumed index case (community-acquired CD-027) that was diagnosed in March 2013 in our LaboratoryReference Lagier, Dubourg and Cassir 10 and the highly probable index case from LTCF-1 diagnosed in January 2013. This finding emphasizes the fact that outbreaks are certainly better resolved with global analysis. Moreover, the spread to the community from this outbreak related to LTCF-1 is probably underestimated, primarily due to the risk of cross-transmission from asymptomatic CD-027 carriers.Reference Furuya-Kanamori and Marquess 17 LTCFs and RCHs are being increasingly recognized as important reservoirs of C. difficile. Reference Di Bella, Paglia, Johnson and Petrosillo 8 , Reference Riggs, Sethi, Zabarsky, Eckstein, Jump and Donskey 18 Henderson et alReference Henderson, Maddock and Andrews 19 identified significant issues regarding infection control practices in RCHs, including limited microbiological investigations, absence or delay of isolation, lack of protective equipment, inappropriate practice when cleaning the environment, and antibiotic misuse.
The cornerstone of the control of this outbreak was likely the cohorting of patients in a dedicated unit. This measure allows a specific care with a trained and educated staff and restricts transfers between HCFs and RCHs.Reference Debast, Vaessen, Choudry, Wiegers-Ligtvoet, van den Berg and Kuijper 20 , Reference Cherifi, Delmee, Van Broeck, Beyer, Byl and Mascart 21 In our infectious diseases ward, a specialized approach including fecal microbiota transplantation when indicated was implemented to optimize the management of CDI.Reference Lagier, Delord and Million 22 Although time-consuming, the daily presence of a member of the infection control team warranted better compliance of the prevention measures as well as regular training of healthcare professionals and dissemination of information to visitors.Reference Goldstein, Johnson and Maziade 23 During this outbreak, cohorting of patients in the infectious diseases ward was implemented only for patients with CD-027 infection, independently of the severity of the illness. Consequently, at the point-of-care laboratories of Marseille, while the rate of patients with CD-027 (compared to all patients tested) decreased significantly, the rate of patients with non-027 CDI remained stable (Online Supplementary Figure S2).
This study has some limitations. First, it is a retrospective study analyzing the clinical characteristics only from patients with confirmed or probable CD-027 infections. Second, we did not confirm clonality between strains. Third, this outbreak has been successfully controlled due to a multifaceted intervention. Therefore, we could not analyze the individual role of any of these interventions.
ACKNOWLEDGMENTS
The authors acknowledge the members of the infection control team,Valérie Roux, Georgette Grech and Yves Seccia, and the staff of the infectious diseases ward at Hôpital Nord, Marseille, France.
Financial support: No financial support was provided relevant to this article.
Potential conflicts of interest: All authors report no conflicts of interest relevant to this article.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit http://dx.doi.org/doi:10.1017/ice.2016.164.
