Effect of ribotype on all-cause mortality following Clostridium difficile infection

https://doi.org/10.1016/j.jhin.2013.04.008Get rights and content

Summary

Background

Clostridium difficile infection (CDI) is significantly associated with subsequent all-cause mortality. Although a number of studies have investigated mortality associated with CDI, few have compared all-cause mortality between ribotypes.

Aim

We aimed to estimate all-cause mortality following CDI and to investigate the relationship between mortality, ribotype and other available variables.

Methods

We undertook a retrospective cohort study. All patients with toxin-positive CDI in North East England between July 2009 and June 2011 were matched to death registration data. Differences in all-cause 30-day case fatality were explored using Poisson regression with robust error variances. For survival analysis, an accelerated failure time model with generalized gamma distribution was chosen.

Findings

In total, 1426 patients were included. All-cause case fatality was 10.2%, 16.4%, 25.7% and 38.1% at 7, 14, 30 and 90 days respectively. In multivariate analysis, ribotype 027 (risk ratio: 1.34; 95% confidence interval: 1.02–1.75) and ribotype 015 (0.46; 0.26–0.82) were significantly associated with higher and lower all-cause 30-day case fatality rates, respectively. In survival analysis, only ribotype 015 had significantly lower predicted mortality (P = 0.008). Patients whose infection was hospital-acquired had significantly higher predicted mortality (P < 0.001).

Conclusion

This is the first population-based study of comparative mortality between multiple ribotypes. Our study identified a high rate of all-cause mortality following CDI. We found evidence of variability in mortality between ribotypes in this cohort with mortality significantly higher for ribotype 027 at 30 days following diagnosis and significantly lower for ribotype 015.

Introduction

Clostridium difficile infection (CDI) is recognized as a widespread cause of healthcare-associated diarrhoea in Europe and North America.1, 2 Despite an ongoing decrease in the number of cases diagnosed in England, C. difficile remains a significant cause of mortality; 2704 deaths in England and Wales mentioned CDI on the death certificate in 2010.3, 4 The actual number of deaths caused by CDI is likely to be higher as using certified cause of death can underestimate the true number of attributable deaths.5

Death at 30 days is one of the most frequently used outcomes for investigating mortality caused by CDI.6 Published estimates of 30-day mortality vary widely, from 4.7% to 42%.7, 8 Estimates also vary by country, with some of the highest reported 30-day mortality from the UK; one weighted average 30-day mortality estimate based on studies in the UK was 30%.6 This is consistent with a recent study based in North East England which estimated 30-day all-cause mortality to be 32.5%.9 Many of these studies were based on a single hospital or subset of patients, and did not examine the effect of ribotype on mortality following CDI.

Ribotyping offers a way of understanding the molecular epidemiology of CDI. This involves fingerprinting the 16S–23S rRNA gene intergenic spacer region to assign a ribotype.10 At least 116 ribotypes have been recognized, but the total number of different ribotypes is likely to be higher.11

The epidemiology of C. difficile ribotypes in North East England during 2009/10 has previously been described.12 The five most numerous ribotypes were 001 (15.6%), 106 (11.7%), 027 (10.6%), 015 (7.7%) and 078 (5.6%). The ribotype distribution changed over time; from July 2009 to December 2010, the proportions of ribotypes 002 and 015 decreased and the proportions of ribotypes 016 and 023 were found to increase.12

Although several studies have investigated the mortality associated with CDI, relatively few studies have compared mortality between ribotypes. Some ribotypes such as 027 are considered to be ‘hypervirulent’, based on evidence of increased resistance to the fluoroquinolone class of antibiotics, greater sporulation efficiency, increased toxin production and production of binary toxin.13, 14, 15, 16 There is some evidence that infection with ribotype 027 is associated with higher mortality rates.17, 18, 19, 20 However, no association between severe CDI and ribotype 027 was observed in a number of smaller studies.21, 22, 23

We conducted a retrospective population-based cohort study to estimate mortality following CDI and investigate the relationship between mortality and ribotype of CDI.

Section snippets

Study design and participants

The North East Ribotyping project aimed to ribotype all toxin-positive C. difficile samples identified in the north east of England between April 2009 and March 2012 and has been previously described.12 This differs from the Clostridium difficile Ribotyping Network (CDRN) which requires samples to meet certain criteria before ribotyping.24 All samples from one hospital were excluded as toxin-negative samples could not be identified. C. difficile cultures were ribotyped using a previously

Results

In total, 2057 toxin-positive samples were submitted in the study period. A number of samples were excluded as C. difficile was not isolated (N = 192), no sample date was recorded (N = 2) or no valid NHS number could be obtained (N = 254). Also, where there was a later sample recorded for the same person, the earlier sample was excluded (N = 183). Therefore 1426 patients were included in this study.

Demographic and ribotyping details of study participants are shown in Table I. The majority of

Discussion

This is the first study of comparative mortality between multiple ribotypes, using a large population-based cohort to quantify mortality following CDI. This study provides the first ribotype-specific data on mortality for ribotypes that are common in England and Wales. These data improve our understanding of mortality caused by CDI, and will inform estimates of the burden of CDI infection.

There was variability in mortality among ribotypes; mortality was significantly higher for ribotype 027 at

Acknowledgements

The authors are grateful to the diagnostic microbiology laboratories in North East England for contributing samples for typing, J. Evans and B. Dowling for performing much of the ribotyping, A. Wilson for administrative assistance and S. Conti for statistical advice. The authors acknowledge the ONS as the provider of death registration data and declare that those who carried out the original collection and analysis of the data bear no responsibility for their further analysis or interpretation.

References (36)

  • M.R. Welfare et al.

    Co-morbidities as predictors of mortality in Clostridium difficile infection and derivation of the ARC predictive score

    J Hosp Infect

    (2011)
  • F. Barbut et al.

    Epidemiology and control of Clostridium difficile infections in healthcare settings

    Curr Opin Infect Dis

    (2011)
  • Office for National Statistics

    Deaths involving Clostridium difficile: England and Wales, 2006 to 2010

    (23 August 2011)
  • Healthcare Commission

    Investigation into outbreaks of Clostridium difficile at Maidstone and Tunbridge Wells NHS Trust

    (October 2007)
  • J. Pépin et al.

    Clostridium difficile-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity

    Can Med Assoc J

    (2004)
  • S.L. Stubbs et al.

    PCR targeted to the 16S–23S rRNA gene intergenic spacer region of Clostridium difficile and construction of a library consisting of 116 different PCR ribotypes

    J Clin Microbiol

    (1999)
  • J. Pépin et al.

    Emergence of fluoroquinolones as the predominant risk factor for Clostridium difficile-associated diarrhea: a cohort study during an epidemic in Quebec

    Clin Infect Dis

    (2005)
  • M. Merrigan et al.

    Human hypervirulent Clostridium difficile strains exhibit increased sporulation as well as robust toxin production

    J. Bacteriol

    (2010)
  • Cited by (23)

    • Japanese Clinical Practice Guidelines for Management of Clostridioides (Clostridium) difficile infection

      2022, Journal of Infection and Chemotherapy
      Citation Excerpt :

      Ribotype 027 was found to be a risk factor for death in a single-center retrospective study conducted in Canada by Labbé et al. (OR, 2.06; 95% CI, 1.00–4.22) [185] and in a retrospective study including 1,144 patients by Rao et al. (OR, 2.02; 95% CI, 1.19–3.43; P = 0.009) [203], as well as a risk factor for hospital mortality in a study of 292 patients by Bauer et al. (OR, 1.02; 95% CI, 0.53–1.96) [204]. A retrospective cohort study with 1,426 patients by Inns et al. showed the best prognosis for ribotype 015 with an RR of 0.46 (95% CI, 0.26–0.83), compared with an RR of 1.34 (95% CI, 1.02–1.7) for ribotype 027 [189]. A prospective case-control study including 139 patients by Wong et al. showed ribotype 002 as a risk factor (HR, 28; 95% CI, 1.1–7.0) [205].

    • Trends in mortality following Clostridium difficile infection in Scotland, 2010–2016: a retrospective cohort and case–control study

      2018, Journal of Hospital Infection
      Citation Excerpt :

      The prevalence of 027 in Scotland has changed greatly since 2008, from being one of the three most common PCR ribotypes found in Scotland to almost disappearing completely. The proportion of 027 in our study was 2.5% compared to 10.3% and 8.0% in the more recent English and Dutch studies, and, along with PCR ribotype 001, these were the only types in our study that had an OR >1, albeit not statistically significant [11,14]. We cannot discount the possibility that the lack of any association between 001 or 027 and higher mortality in our study might be due to the small numbers.

    View all citing articles on Scopus
    View full text