short report Wien Klin Wochenschr DOI 10.1007/s00508-014-0549-x
Severe Clostridium difficile infection: incidence and risk factors at a tertiary care university hospital in Vienna, Austria Peter Starzengruber · Luigi Segagni Lusignani · Thomas Wrba · Dieter Mitteregger · Alexander Indra · Wolfgang Graninger · Elisabeth Presterl · Magda Diab-Elschahawi
Received: 15 January 2014 / Accepted: 13 April 2014 © Springer-Verlag Wien 2014
Summary Background Clostridium difficile infection (CDI) is the major cause of hospital-acquired bacterial diarrhoea. The incidence of CDI has been increasing in Canada, the US and Europe and severe cases are becoming more common. Methods A retrospective cohort study investigating all patients with an episode of CDI present at the Vienna University Hospital between 01 January 2012 and 31 December 2012 was conducted. All microbiologically confirmed C. difficile toxin positive cases were included, ribotyped and analysed regarding their clinical course. Results A total of 278 patients with CDI were recorded, with an overall CDI incidence of 5.23 per 10,000 patients-
E. Presterl, MD, MBA () · P. Starzengruber, MD · L. Segagni Lusignani, MD · M. Diab-Elschahawi, MD Department of Hospital Hygiene and Infection Control, Vienna General Hospital, Medical University Vienna, Waehringer Guertel 18–20, 1090 Vienna, Austria e-mail: [email protected] P. Starzengruber, MD · Prof. W. Graninger, MD Department of Medicine I, Division of Infectious Diseases and Tropical Medicine, Medical University of Vienna, Vienna, Austria T. Wrba, PhD Center for Medical Statistics, Informatics and Intelligent Systems, Medical University of Vienna, Vienna, Austria D. Mitteregger, MD Division of Clinical Microbiology, Department of Laboratory Medicine, Medical University Vienna, Vienna, Austria A. Indra, MD Austrian Agency for Health and Food Safety, Vienna, Austria
13
days. Around 84,5 % (235/278) of CDI cases would have been classified as severe CDI according to European Society of Clinical Microbiology and Infectious Diseases (ESCMID) if all criteria were used. According to Society for Healthcare Epidemiology of America and the Infectious Diseases Society of America (SHEA/IDSA) guidelines only 16.5 % (46/278) could be classified as severe; with a severe CDI incidence of 4.41 and 0.86 per 10,000 patient-days, respectively. Multivariate analysis showed only a co-morbidity index of ≥ 3 (p = 0.013) as independent risk factor for severe CDI. No link between ribotype 027 and severity or clustering was observed in our study population. Conclusions Special attention in terms of restrictive antibiotic prescription should be given to patients having a Charlson co-morbidity ≥ 3 at the time of hospital admission. SHEA/IDSA guidelines were more accurate than ESCMID criteria in predicting severe CDI in our collective, of mostly severely ill patients, in a tertiary care hospital setting. Keywords Clostridium difficile infection · Incidence · Risk factors · Severe case · Tertiary care hospital
Schwere Clostridium difficile Infektionen – Inzidenz und Risikofaktoren an der Universitätsklinik Wien Zusammenfassung Grundlagen Clostridium difficile ist zu einem der wichtigsten nosokomialen Erreger infektiöser Diarrhoe in Industriestaaten geworden. In Canada, den USA und Europa findet sich eine rasch steigende Inzidenz von C. difficile Infektionen (CDI) und mit dieser ein Anstieg von schweren Erkrankungsverläufen. Methodik Von 1. Jänner bis 31. Dezember 2012 wurden im Rahmen einer retrospektiven Kohortenstudie im
Severe Clostridium difficile infection
1
short report
Allgemeinen Krankenhaus der Stadt Wien – Medizinische Universitätsklinik alle mikrobiologisch gesicherten C. difficile Toxin positive Fälle eingeschlossen, ribotypisiert und ihr klinischer Verlauf analysiert. Ergebnisse Insgesamt wurden 278 PatientInnen mit CDI in die Studie aufgenommen. Es zeigte sich eine Gesamt-CDI Inzidenz von 5,23 pro 10.000 PatientInnentagen. Ein schwerer CDI Verlauf fand sich bei 84,5 % (235/278) der CDI Fällen, wenn die European Society of Clinical Microbiology and Infectious Diseases (ESCMID) Kriterien angewandt wurden. Zieht man die Kriterien der Society for Healthcare Epidemiology of America and the Infectious Diseases Society of America (SHEA/ IDSA) heran, so konnten nur 16,5 % (46/278) der CDI Fälle als schwer charakterisiert werden. Dies entspricht einer jeweiligen Inzidenz an schweren CDI Fälle von 4,41 und 0,86 pro 10.000 PatientInnentage nach ESCMID und SHEA/IDSA. Multivariate Analysen zeigten nur einen Co- Morbiditätsindex von ≥ 3 (p = 0,013) als unabhängiger Risikofaktor für eine schwere CDI. In dieser Studienpopulation konnte kein Zusammenhang zwischen dem „hypervirulenten“ Ribotyp 027 und Schwere oder Clustering beobachtet werden. Schlussfolgerungen Bei PatientInnen, die bereits bei der Aufnahme einen Charlson Co-Morbiditätsindex von ≥ 3 haben, sollte die Notwendigkeit jedes Antibiotika Einsatzes besonders kritisch abgewogen werden. SHEA/ IDSA Kriterien führten zu einer strikteren Einteilung schwerer CDI Fälle als die europäischen ESCMID Kriterien in unserem Patientenkollektiv an der Universitätsklinik Wien. Schlüsselwörter Clostridium difficile Infektion · Inzidenz · Risikofaktoren · Schwerer Verlauf · Universitätsklinik
Introduction The incidence of Clostridium difficile infection (CDI) has been increasing in Canada, the US and Europe and severe cases are becoming more common [1]. Severe cases of CDI have become reportable in Austria since 2010. Based on official hospital discharge data according to the International Classification of Diseases (10th revision), enterocolitis due to C. difficile given as primary diagnosis increased more than three-fold in Austria from 2003 until 2007, from 3.6 to 11.9 per 10,000 hospital discharges [2]. Accordingly, increasing numbers of C. difficile were found at the Vienna University Hospital (VUH) since 2004 (data not shown). The description of so-called hypervirulent strains of C. difficile in clinical settings [3] warrants the need for clinical course description. As no validated severity score indices for CDI are currently available, varying definitions are used for epidemiological purposes. Early identification of potentially severe CDI is important for the assessment of patient management options, both medical and surgical. The aim of our study was therefore to determine the inci-
2 Severe Clostridium difficile infection
dence of severe cases of C. difficile infection by comparing two different severity indices used in Europe and the US during a 1-year observation period at the VUH. Possible risk factors for a severe course of CDI and possible nosocomial transmissions during this time period were further evaluated.
Patients, materials and methods A retrospective cohort study investigating all patients with an episode of CDI present at the VUH between 01 January 2012 and 31 December 2012 was conducted. The VUH is a 2,119-bed tertiary care medical university teaching hospital in Vienna, Austria.
Inclusion criteria and data collection All patients with a positive stool sample for Clostridium difficile antigen glutamate dehydrogenase (GDH), detected by an enzyme-linked immunosorbent assay (C. DIFF CHEK—60TM, Alere) and tested positive by real-time PCR (Xpert® C. difficile, GeneXpert® System, Cepheid), were included and their medical records retrospectively reviewed for: (a) demographics, (b) length of hospital stay, (c) clinical course of CDI, (d) co-morbidity, (e) fatal outcome, (f ) admission ward (g) use of antibiotics before onset of CDI, (h) presence of ribotype 027 and (i) possible nosocomial transmission. Ribotyping was done as previously described [4]. An episode of CDI was defined according to the definitions of the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) [5]. Severe CDI was defined by Society for Healthcare Epidemiology of America and the Infectious Diseases Society of America (SHEA/IDSA) guidelines as an episode of CDI with either a marked leukocytosis (leukocyte count > 15 × 109/L) or a rise in serum creatinine (> 50 % above the baseline) and compared with ESCMID criteria for severe CDI [5, 6]. The Charlson co-morbidity index was used as a prognostic index for classifying co-morbid conditions [7]. All C. difficile isolates are routinely investigated for space-time clustering to evaluate patient to patient cross-infection. Therefore, all patients with CDI at the VUH are routinely entered in our in-house routine surveillance system (ArchimMed RDA-Plattform®). Possible secondary cases are also recorded and appropriate infection control measures are instituted.
Statistical analysis Statistical analysis was performed using SPSS software, version 20.0. Beyond descriptive analysis, associations between the dependent variable (CDI severity according to SHEA/IDSA) and the independent variables were analysed using odds ratios (OR) and associated 95 % confidence intervals (CI). Proportions were compared using
13
short report
the Pearson’s chi-square test. Independent variables that showed p 65 years
26 (16.7 %)
130 (83.3 %)
156
≤ 7 days
20 (15.7 %)
107 (84.3 %)
127
> 7 days
26 (17.2 %)
125 (82.8 %)
151
P-value
Gender 0.74
0.95
Length of stay 0.742
Charlson index 3
2.54 (1.21–5.30)
0.013
Admission ward
0.84 (0.46–1.53)
0.576
Ribotype 027
0.99 (0.31–3.13)
0.997
P- value calculated by multiple logistic regression analysis CDI Clostridium difficile infection
Between 01 January 2012 and 31 December 2012, a total of 278 C. difficile positive patients were recorded with an overall CDI incidence of 5.23 per 10,000 patients-days. Out of these, 66.2 % (184/278) of patients received antimicrobial therapy within 30 days prior to the onset of CDI. Ribotype 027 was found in 9 % (25/278) of patients. Up to 84.5 % (235/278) of CDI cases would have been classified as severe CDI according to ESCMID if all criteria were used. If severity was defined according to SHEA/ IDSA guidelines based on either a marked leukocytosis (leukocyte count > 15 × 109/L) or creatinine ( > 50 % above the baseline), only 16.5 % (46/278) could be classified as severe; with a severe CDI incidence of 4.41 and 0.86 per 10,000 patients-day according to ESCMID and SHEA/ IDSA, respectively. About 59.3 % (156/278) of patients had moderate/severe co-morbidity indicated by a Charlson co-morbidity index of ≥ 3. A crude overall 30-day
Total: 278 Patients
Table 2 Results of the multivariate logistic regression analysis of clinical course of Clostridium difficile infection as dependent variable and investigated parameters as independent variables
0.463
mortality of 13.7 % (38/278) has been found and 15.1 % (42/278) of CDI patients died from all causes during their hospital stay; 50 % (21/42) of them were severe CDI cases according to SHEA/IDSA. Being a severe case according to SHEA/IDSA was significantly (OR = 8.44; CI (95 %) 4.05–17.07; p 0.20. The multivariate analysis revealed that severe CDI according to SHEA/IDSA was only significantly associated with a Charlson co-morbidity index of ≥ 3 (p
Severe Clostridium difficile infection: incidence and risk factors at a tertiary care university hospital in Vienna, Austria Peter Starzengruber · Luigi Segagni Lusignani · Thomas Wrba · Dieter Mitteregger · Alexander Indra · Wolfgang Graninger · Elisabeth Presterl · Magda Diab-Elschahawi
Received: 15 January 2014 / Accepted: 13 April 2014 © Springer-Verlag Wien 2014
Summary Background Clostridium difficile infection (CDI) is the major cause of hospital-acquired bacterial diarrhoea. The incidence of CDI has been increasing in Canada, the US and Europe and severe cases are becoming more common. Methods A retrospective cohort study investigating all patients with an episode of CDI present at the Vienna University Hospital between 01 January 2012 and 31 December 2012 was conducted. All microbiologically confirmed C. difficile toxin positive cases were included, ribotyped and analysed regarding their clinical course. Results A total of 278 patients with CDI were recorded, with an overall CDI incidence of 5.23 per 10,000 patients-
E. Presterl, MD, MBA () · P. Starzengruber, MD · L. Segagni Lusignani, MD · M. Diab-Elschahawi, MD Department of Hospital Hygiene and Infection Control, Vienna General Hospital, Medical University Vienna, Waehringer Guertel 18–20, 1090 Vienna, Austria e-mail: [email protected] P. Starzengruber, MD · Prof. W. Graninger, MD Department of Medicine I, Division of Infectious Diseases and Tropical Medicine, Medical University of Vienna, Vienna, Austria T. Wrba, PhD Center for Medical Statistics, Informatics and Intelligent Systems, Medical University of Vienna, Vienna, Austria D. Mitteregger, MD Division of Clinical Microbiology, Department of Laboratory Medicine, Medical University Vienna, Vienna, Austria A. Indra, MD Austrian Agency for Health and Food Safety, Vienna, Austria
13
days. Around 84,5 % (235/278) of CDI cases would have been classified as severe CDI according to European Society of Clinical Microbiology and Infectious Diseases (ESCMID) if all criteria were used. According to Society for Healthcare Epidemiology of America and the Infectious Diseases Society of America (SHEA/IDSA) guidelines only 16.5 % (46/278) could be classified as severe; with a severe CDI incidence of 4.41 and 0.86 per 10,000 patient-days, respectively. Multivariate analysis showed only a co-morbidity index of ≥ 3 (p = 0.013) as independent risk factor for severe CDI. No link between ribotype 027 and severity or clustering was observed in our study population. Conclusions Special attention in terms of restrictive antibiotic prescription should be given to patients having a Charlson co-morbidity ≥ 3 at the time of hospital admission. SHEA/IDSA guidelines were more accurate than ESCMID criteria in predicting severe CDI in our collective, of mostly severely ill patients, in a tertiary care hospital setting. Keywords Clostridium difficile infection · Incidence · Risk factors · Severe case · Tertiary care hospital
Schwere Clostridium difficile Infektionen – Inzidenz und Risikofaktoren an der Universitätsklinik Wien Zusammenfassung Grundlagen Clostridium difficile ist zu einem der wichtigsten nosokomialen Erreger infektiöser Diarrhoe in Industriestaaten geworden. In Canada, den USA und Europa findet sich eine rasch steigende Inzidenz von C. difficile Infektionen (CDI) und mit dieser ein Anstieg von schweren Erkrankungsverläufen. Methodik Von 1. Jänner bis 31. Dezember 2012 wurden im Rahmen einer retrospektiven Kohortenstudie im
Severe Clostridium difficile infection
1
short report
Allgemeinen Krankenhaus der Stadt Wien – Medizinische Universitätsklinik alle mikrobiologisch gesicherten C. difficile Toxin positive Fälle eingeschlossen, ribotypisiert und ihr klinischer Verlauf analysiert. Ergebnisse Insgesamt wurden 278 PatientInnen mit CDI in die Studie aufgenommen. Es zeigte sich eine Gesamt-CDI Inzidenz von 5,23 pro 10.000 PatientInnentagen. Ein schwerer CDI Verlauf fand sich bei 84,5 % (235/278) der CDI Fällen, wenn die European Society of Clinical Microbiology and Infectious Diseases (ESCMID) Kriterien angewandt wurden. Zieht man die Kriterien der Society for Healthcare Epidemiology of America and the Infectious Diseases Society of America (SHEA/ IDSA) heran, so konnten nur 16,5 % (46/278) der CDI Fälle als schwer charakterisiert werden. Dies entspricht einer jeweiligen Inzidenz an schweren CDI Fälle von 4,41 und 0,86 pro 10.000 PatientInnentage nach ESCMID und SHEA/IDSA. Multivariate Analysen zeigten nur einen Co- Morbiditätsindex von ≥ 3 (p = 0,013) als unabhängiger Risikofaktor für eine schwere CDI. In dieser Studienpopulation konnte kein Zusammenhang zwischen dem „hypervirulenten“ Ribotyp 027 und Schwere oder Clustering beobachtet werden. Schlussfolgerungen Bei PatientInnen, die bereits bei der Aufnahme einen Charlson Co-Morbiditätsindex von ≥ 3 haben, sollte die Notwendigkeit jedes Antibiotika Einsatzes besonders kritisch abgewogen werden. SHEA/ IDSA Kriterien führten zu einer strikteren Einteilung schwerer CDI Fälle als die europäischen ESCMID Kriterien in unserem Patientenkollektiv an der Universitätsklinik Wien. Schlüsselwörter Clostridium difficile Infektion · Inzidenz · Risikofaktoren · Schwerer Verlauf · Universitätsklinik
Introduction The incidence of Clostridium difficile infection (CDI) has been increasing in Canada, the US and Europe and severe cases are becoming more common [1]. Severe cases of CDI have become reportable in Austria since 2010. Based on official hospital discharge data according to the International Classification of Diseases (10th revision), enterocolitis due to C. difficile given as primary diagnosis increased more than three-fold in Austria from 2003 until 2007, from 3.6 to 11.9 per 10,000 hospital discharges [2]. Accordingly, increasing numbers of C. difficile were found at the Vienna University Hospital (VUH) since 2004 (data not shown). The description of so-called hypervirulent strains of C. difficile in clinical settings [3] warrants the need for clinical course description. As no validated severity score indices for CDI are currently available, varying definitions are used for epidemiological purposes. Early identification of potentially severe CDI is important for the assessment of patient management options, both medical and surgical. The aim of our study was therefore to determine the inci-
2 Severe Clostridium difficile infection
dence of severe cases of C. difficile infection by comparing two different severity indices used in Europe and the US during a 1-year observation period at the VUH. Possible risk factors for a severe course of CDI and possible nosocomial transmissions during this time period were further evaluated.
Patients, materials and methods A retrospective cohort study investigating all patients with an episode of CDI present at the VUH between 01 January 2012 and 31 December 2012 was conducted. The VUH is a 2,119-bed tertiary care medical university teaching hospital in Vienna, Austria.
Inclusion criteria and data collection All patients with a positive stool sample for Clostridium difficile antigen glutamate dehydrogenase (GDH), detected by an enzyme-linked immunosorbent assay (C. DIFF CHEK—60TM, Alere) and tested positive by real-time PCR (Xpert® C. difficile, GeneXpert® System, Cepheid), were included and their medical records retrospectively reviewed for: (a) demographics, (b) length of hospital stay, (c) clinical course of CDI, (d) co-morbidity, (e) fatal outcome, (f ) admission ward (g) use of antibiotics before onset of CDI, (h) presence of ribotype 027 and (i) possible nosocomial transmission. Ribotyping was done as previously described [4]. An episode of CDI was defined according to the definitions of the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) [5]. Severe CDI was defined by Society for Healthcare Epidemiology of America and the Infectious Diseases Society of America (SHEA/IDSA) guidelines as an episode of CDI with either a marked leukocytosis (leukocyte count > 15 × 109/L) or a rise in serum creatinine (> 50 % above the baseline) and compared with ESCMID criteria for severe CDI [5, 6]. The Charlson co-morbidity index was used as a prognostic index for classifying co-morbid conditions [7]. All C. difficile isolates are routinely investigated for space-time clustering to evaluate patient to patient cross-infection. Therefore, all patients with CDI at the VUH are routinely entered in our in-house routine surveillance system (ArchimMed RDA-Plattform®). Possible secondary cases are also recorded and appropriate infection control measures are instituted.
Statistical analysis Statistical analysis was performed using SPSS software, version 20.0. Beyond descriptive analysis, associations between the dependent variable (CDI severity according to SHEA/IDSA) and the independent variables were analysed using odds ratios (OR) and associated 95 % confidence intervals (CI). Proportions were compared using
13
short report
the Pearson’s chi-square test. Independent variables that showed p 65 years
26 (16.7 %)
130 (83.3 %)
156
≤ 7 days
20 (15.7 %)
107 (84.3 %)
127
> 7 days
26 (17.2 %)
125 (82.8 %)
151
P-value
Gender 0.74
0.95
Length of stay 0.742
Charlson index 3
2.54 (1.21–5.30)
0.013
Admission ward
0.84 (0.46–1.53)
0.576
Ribotype 027
0.99 (0.31–3.13)
0.997
P- value calculated by multiple logistic regression analysis CDI Clostridium difficile infection
Between 01 January 2012 and 31 December 2012, a total of 278 C. difficile positive patients were recorded with an overall CDI incidence of 5.23 per 10,000 patients-days. Out of these, 66.2 % (184/278) of patients received antimicrobial therapy within 30 days prior to the onset of CDI. Ribotype 027 was found in 9 % (25/278) of patients. Up to 84.5 % (235/278) of CDI cases would have been classified as severe CDI according to ESCMID if all criteria were used. If severity was defined according to SHEA/ IDSA guidelines based on either a marked leukocytosis (leukocyte count > 15 × 109/L) or creatinine ( > 50 % above the baseline), only 16.5 % (46/278) could be classified as severe; with a severe CDI incidence of 4.41 and 0.86 per 10,000 patients-day according to ESCMID and SHEA/ IDSA, respectively. About 59.3 % (156/278) of patients had moderate/severe co-morbidity indicated by a Charlson co-morbidity index of ≥ 3. A crude overall 30-day
Total: 278 Patients
Table 2 Results of the multivariate logistic regression analysis of clinical course of Clostridium difficile infection as dependent variable and investigated parameters as independent variables
0.463
mortality of 13.7 % (38/278) has been found and 15.1 % (42/278) of CDI patients died from all causes during their hospital stay; 50 % (21/42) of them were severe CDI cases according to SHEA/IDSA. Being a severe case according to SHEA/IDSA was significantly (OR = 8.44; CI (95 %) 4.05–17.07; p 0.20. The multivariate analysis revealed that severe CDI according to SHEA/IDSA was only significantly associated with a Charlson co-morbidity index of ≥ 3 (p
