A review of the epidemiology, risk factors and strain characteristics of Clostridium difficile among hospitalized patients: a pilot nested casecontrol study

Abstract number: O160

Cleary P., Quigley C., O'Brien S., Bolton E., Verlander N., Guleri A., Dodgson A., Fox A., Maxwell S., Braid K.

Introduction: The study aimed to determine the outcomes, mortality and morbidity from Clostridium difficile infections (CDI) in the North West of England, and to identify possible differences in risk factors between ribotypes.

Methods: Multicentre case control study at three large hospitals in NW England. Cases defined as hospitalized patients with positive Clostridium difficile toxin in diarrhoeal stool samples. Controls randomly selected from matched patients with negative stool samples, at a ratio of 1:1, frequency matched by 10 year age band. Data was abstracted at case note review. Stool samples from cases cultured and typed at regional reference laboratory.

Results: Between Sept 9, 2007 and Sept 30, 2008, data on 296 cases and 296 controls collected and analysed. Average age for controls and cases was 77 years. 40.2%(119/296) of controls and 48.3%(143/296) of cases were male. 48 (16%) controls and 97 (33%) cases died during the study period. Of the samples typed, 27.7% were 001, 26.6% were 001 and 20.9% were 027.

In the overall logistic regression model, length of stay, transfer from another NHS hospital, number of different antibiotics, number of days of quinolone antibiotics, degree of comorbidity and number of days of nitroimidazole antibiotics were identified as risk factors (see figure). Reduction in odds of CDI with H2-blocker therapy was an interesting finding.

On subgroup analysis the three main strains were all associated with length of hospital stay and transfer from another NHS hospital. All were associated with antibiotic use, especially quinolones. Only ribotype 001 was associated with cephalosporin use. PPI therapy was only a risk factor for ribotype 106.

The 014/020, 078 and 027 ribotypes had the highest observed rates of mortality and morbidity, but there were no significant difference in mortality or morbidity detected between different ribotype.

Discussion: This study did not find any difference in mortality or morbidity between different strains of Clostridium difficile, but did find some differences in risk factors between ribotypes. An unexpected association of H2-blocker therapy with a reduction in odds of CDAD was observed.

This study has some limitations. Diagnosis of Clostridium difficile associated disease was based on toxin detection, which may not be sensitive or specific. Details to be presented.

Figure. Multiple logistic regression model for all ribotypes.

Session Details

Date: 10/04/2010
Time: 00:00-00:00
Session name: Abstracts 20th European Congress of Clinical Microbiology and Infectious Diseases
Location: Vienna, Austria, 10 - 13 April 2010
Presentation type:
Back to top