Elsevier

The Lancet

Volume 349, Issue 9049, 8 February 1997, Pages 426-428
The Lancet

Hypothesis
Hospital-acquired Clostridium difficile diarrhoea and herd immunity

https://doi.org/10.1016/S0140-6736(97)80053-0Get rights and content

Summary

Clostridium difficile diarrhoea represents a significant health-service burden. We recently experienced an outbreak of C difficile diarrhoea associated with increased use of cefotaxime. The question we pose in this paper is how did the introduction and withdrawal of a single antibiotic so greatly affect rates of C difficile diarrhoea? Other antibiotics had nearly as high a risk of causing diarrhoea as cefotaxime, and the majority of patients never received cefotaxime. We believe that such outbreaks of C difficile diarrhoea are best understood in terms of a population model, and that taking antibiotics like cefotaxime should be thought of as a population rather than an individual risk factor. We postulate a herd-immunity model of C difficile diarrhoea, and examine the implications of this hypothesis.

Introduction

Although Clostridium difficile was first described in 1935, interest in this micro-organism increased greatly after 1977 when it was implicated in antibiotic-associated colitis. During the past 10 years, cases of hospital-acquired diarrhoea caused by C difficile have increased exponentially and it is now the commonest cause of hospital-acquired diarrhoea. C difficile diarrhoea represents a health-service burden, both in terms of direct costs and loss of potential admissions because of necessary ward closures. We reported1 an outbreak of C difficile diarrhoea in our geriatric wards that was associated with increased use of cefotaxime. Our increased prescribing of cefotaxime had been in response to published recommendations of the British Thoracic Society, which advocated the use of cefotaxime in community-acquired severe pneumonia.2 Our findings were consistent with the well-established association between antibiotic use and C difficile,1, 3 and previous findings that patients taking cefotaxime are at high risk of developing C difficile diarrhoea.4

The question we now pose is how did the introduction and withrdrawal of a single antibiotic so greatly affect rates of C difficile diarrhoea? Other antibiotics were associated with nearly as high a risk of C difficile diarrhoea as cefotaxime, and 58% of patients with this diarrhoea had never received cefotaxime. Why were there fewer non-cefotaxime cases of diarrhoea when cefotaxime was not in use? We knew that our experience was not unique; a similar pattern was observed for clindamycin in a hospital in Chicago,5 where antibiotic restriction reduced new case numbers far more than predicted by the number (75%) of patients for whom clindamycin was avoided. We argue that such outbreaks are best understood in terms of a population model, and that use of an antibiotic such as cefotaxime should be thought of as a risk factor for a population rather than for an individual. We postulate that a herd-immunity model of C difficile diarrhoea exists, and examine the implications of this hypothesis.

Section snippets

Herd immunity in the community

For an infection to spread within a population the number of secondary cases produced by a primary case must be greater than one. The average number of secondary cases produced from a primary case in a wholly susceptible population is known as the basic reproductive rate of the infection, R0.6 However, most populations contain a proportion of people who are immune to infection. In such populations the reproductive rate, R, depends both on R0 and on the proportion of the population who are

C difficile and immunity

There are few data to inform us whether C difficile comprises part of the normal bowel flora in community-dwelling elderly people. However, C difficile can be isolated from many symptom-free elderly patients in institutions and hospitals;8 whether these strains are pathogenic is unclear. Both C difficile colonisation and diarrhoea are thought to occur more readily when there is an alteration to the normal bowel flora.9 Such alterations are principally induced by antibiotics, especially if they

Herd-immunity model for C difficile

Any model of the nosocomial epidemiology of C difficile needs to distinguish between patients with no C difficile colonisation, symptom-free patients colonised with C difficile, and infected patients with toxin-induced diarrhoea caused by C difficile. Figure 1 summarises such a model with these variables which could form the basis for a solution for R0 in terms of a set of differential equations relating states of infection and colonisation. This model would be identical methodologically to

Implications of the herd-immunity hypothesis

As clinicians caring for the elderly, we are trained to look at patients as individuals with particular medical and non-medical problems. Our hypothesis suggests that if we wish to eliminate C difficile outbreaks in hospitals, we need to pay attention to the “herd” rather than just to individuals thought to be at risk. We can adopt a two-pronged approach to minimise R. First we can aim to reduce infectivity (Ro) by standard infection-control methods such as hand-washing, but these should be

References (16)

  • RM Anderson et al.

    Immunisation and herd immunity

    Lancet

    (1990)
  • M Impallomeni et al.

    Increased risk of Clostridium difficile diarrhoea in elderly patients receiving cefotaxime

    BMJ

    (1995)
  • Guidelines for the management of community-aquired pneumonia in adults admitted to hospital

    Br J Hosp Med

    (1993)
  • Clostridium difficile infection: prevention and management

    (1994)
  • H Funada et al.

    Cefotaxime associated diarrhoea and Clostridium difficile

    Jpn J Antibiot

    (1984)
  • SM Pear et al.

    Decrease in nosocomial Clostridium difficile-associated diarrhea by restricting clindamycin use

    Ann Intern Med

    (1994)
  • R Ross

    Some a priori pathometric equations

    BMJ

    (1915)
  • KJ Walker et al.

    Clostridium difficile colonization in residents of long-term care facilities: prevalence and risk factors

    J Am Geriatr Soc

    (1993)
There are more references available in the full text version of this article.

Cited by (53)

  • The projected effectiveness of Clostridium difficile vaccination as part of an integrated infection control strategy

    2016, Vaccine
    Citation Excerpt :

    In the case of C. difficile, these methods would allow the exploration of the impact of vaccination, taking into account the different modes of acquisition [25], as well as the indirect effect of prevention of onward transmission. Therefore, a dynamic transmission model [26–37] was developed to investigate the effectiveness of four vaccination strategies, in terms of preventing CDI in a hospital-ward setting with patients at greater risk of acquiring the infection, such as the Intensive Care Unit (ICU) [38]. Although ICUs can vary markedly in their case mix, their critically ill status often causes a state of immunosuppression [39], and requires high levels of antimicrobial prescribing [40].

  • Anaerobic bacteria 150 years after their discovery by Pasteur

    2015, Enfermedades Infecciosas y Microbiologia Clinica
  • Assessment of administrative data for evaluating the shifting acquisition of Clostridium difficile infection in England

    2012, Journal of Hospital Infection
    Citation Excerpt :

    This clinical presentation is the key to understanding the pathogenesis of infection since 5% of healthy adults and 66% of children in the population may be carriers of C. difficile without symptoms.15 CDI is associated with the widespread use of broad spectrum antibiotics, poor hospital hygiene, and is more common in certain high-risk patient groups.16–19 We found no studies describing the incidence of CDI among patients based on different socioeconomic classes, comorbidities, or admission methods admitted to English hospitals.

  • Nonobstructive Lung Disease and Thoracic Tumors

    2010, Brocklehurst's Textbook of Geriatric Medicine and Gerontology
View all citing articles on Scopus
View full text