Diarrhoea associated with antibiotic use

BMJ 2007; 335 doi: https://doi.org/10.1136/bmj.39255.829120.47 (Published 12 July 2007) Cite this as: BMJ 2007;335:54
  1. Lynne V McFarland, affiliate associate professor
  1. Department of Health Services Research and Development, Puget Sound Veteran Administration Healthcare System, 1100 Olive Way #1400, Seattle, WA 98101, USA
  1. lvmcfarl{at}u.washington.edu

    Evidence support the use of probiotics, but effectiveness depends on the strain

    Diarrhoea is a common side effect of antibiotics; it may prolong hospital stay, increase the risk of other infections, develop into more serious forms of disease (colitis, toxic megacolon), and lead to premature discontinuation of the needed antibiotic. Diarrhoea associated with Clostridium difficile is a leading cause of iatrogenic outbreaks of diarrhoea, and considerably increases mortality and healthcare costs for inpatients.1 2 3 4 Antibiotic associated diarrhoea may develop in 5-30% of patients, with the rates increasing as the antibiotic spectrum gets broader.4

    Diarrhoea associated with antibiotic use may result from the disruption of the barrier of normally protective colonic microflora that are inadvertent targets of the inciting antibiotic. In 20-30% of these cases, an opportunistic pathogen, Clostridium difficile, takes advantage of this opening, colonises the intestine, and produces toxins, resulting in diarrhoea or colitis. A strategy to re-establish this microbial barrier is through the use of probiotics.5 In this week's BMJ Hickson and colleagues report a randomised, placebo-controlled trial of the effects of a mixture of three different strains of probiotics on prevention of diarrhoea associated with antibiotic use.6

    Probiotics are living, beneficial bacteria or yeasts that are taken orally to help restore the microbial balance in the intestinal tract.7 A multitude of probiotic products are available in the global marketplace, but only some are backed by evidence based clinical trials.8 9 10 11 12 The diversity of probiotic products has caused confusion among clinicians and patients as to which are effective and which are not.

    The trial by Hickson and colleagues randomised 135 older hospitalised patients receiving a new course of antibiotics to either a probiotic yoghurt drink (containing Lactobacillus casei DN 114 001, L bulgaricus, and Streptococcus thermophilus) or a placebo milkshake for the duration of the antibiotic plus one week. Patients were followed for an additional four weeks for the development of antibiotic associated diarrhoea or C difficile diarrhoea. Significantly fewer patients given the probiotic drink developed diarrhoea than did those given the placebo milkshake (odds ratio 0.25, 95% confidence interval 0.07 to 0.85). Although the rates of C difficile diseases were low, the probiotic drink also seemed to prevent this outcome (0% v 17% with placebo; adjusted rate ratio 17%, 7% to 27%). No adverse effects were reported.

    The trial has several strengths: the outcomes (diarrhoea associated with antibiotic use and C difficile diarrhoea) were defined, the doses (number of bacteria per day) of probiotics were provided, compliance was assessed, and treatments were blinded to patients and assessors.

    The trial is limited by its low generalisability due to the small proportion (8%) of people who were enrolled from the target pool of hospitalised patients receiving antibiotics. For a preventive treatment to be practical, it needs to be given widely to the population at risk. Another limitation is the lack of proper probiotic designations. Closely related bacterial strains have been shown to have differing abilities to act as efficacious probiotics,13 14 15 so the need to correctly identify the strain (not just the genus and species) of probiotic under investigation is paramount. The authors identified only one strain (L casei DN 114 001), and they (inappropriately) cited its brand name as part of its nomenclature. A description of the time of onset would have been helpful (while patient was taking antibiotics or delayed onset after antibiotics were stopped) so that clinicians know when they should expect diarrhoea to develop. The authors also say their analysis was intention to treat, but patients lost to follow-up were not included; although the loss to follow-up was moderate (22 patients; 16%) and the numbers were similar in the two groups.

    Despite these limitations, this study supports the findings from other randomised, double blinded, controlled trials that various probiotic preparations are effective for preventing diarrhoea associated with antibiotic use.8 9 10 11 12

    So how does this growing body of evidence translate into clinical practice? The study by Hickson and colleagues adds to the findings from recent meta-analyses that probiotics are effective for preventing diarrhoea associated with antibiotics.11 12 Some European hospitals routinely recommend probiotics as an adjunct to high risk antibiotics. Probiotics may be especially useful for patients with chronic infections (such as sinusitis or diabetic foot ulcers) that require repeated courses of antibiotics. Barriers to this practice include choosing an effective probiotic strain, the additional cost of the probiotic, and risks associated with use in immunocompromised patients. Probiotics can easily be given along with antibiotics as an adjunctive preventive treatment and seem to be well tolerated by both paediatric and elderly populations.12 14 16 A word of caution: clinicians need to consider which probiotic strain is supported by evidence from clinical trials, as not all probiotics have equal effectiveness. The evidence that probiotics are effective as a preventive measure for Clostridium difficile requires further trials, and more strains need to be tested for their probiotic potential.


    • Competing interests: LVM has been reimbursed by Laboratories Biocodex and Klaire Labs for attending several conferences and for speaker fees. The viewpoints in this article are solely those of the author and do not represent the viewpoints of the Veterans Administration.

    • Provenance and peer review: Commissioned, not externally peer reviewed.


    View Abstract

    Sign in

    Log in through your institution