Intended for healthcare professionals


Antibiotics for acute purulent rhinitis

BMJ 2002; 325 doi: (Published 07 December 2002) Cite this as: BMJ 2002;325:1311

Probably effective but not routinely recommended

  1. Bruce Arroll, associate professor (b.arroll{at},
  2. Timothy Kenealy, doctoral fellow
  1. Department of General Practice and Primary Health Care, University of Auckland, Private Bag 92019 Auckland, New Zealand

    Despite efforts to lower the prescribing of antibiotics for upper respiratory infections the prescription rates are still over 60%.1 Mucopurulent rhinitis is a component of the common cold, for which antibiotics are generally not effective.2 Guidelines specifically recommend against using antibiotics to treat rhinitis. 3 4 Nevertheless, the colour of the nasal discharge doubles the odds of being prescribed antibiotics.5 In one study no general practitioners said that they would give antibiotics for clear rhinitis but 72% would for purulent rhinitis.6 Moreover, a recent study of acute purulent rhinitis showed a significant reduction in the duration of purulent symptoms from 14 days for placebo to nine days for amoxicillin.7 These findings warrant a reconsideration of the evidence and guidelines for antibiotics in acute purulent rhinitis.

    One evidence based guideline in paediatrics (reviewed by the American Academy of Pediatrics) stated that mucopurulent rhinitis is not an indication for antimicrobial treatment unless it persists for more than 10 to 14 days.3 Although this may be sensible advice, it is referenced to only one study, by Todd et al, who found no benefit from cephalexin for mucopurulent nasopharyngitis at five to six days.8 In contrast to the text, the table in that guideline includes two other papers that actually found a benefit for antibiotics on rhinitis. 9 10

    In a study of children Taylor et al found a relative risk of 0.3 (95% confidence interval 0.11 to 0.82; number needed to treat (NNT)=10) for purulent rhinitis in patients taking amoxicillin or co-trimoxazole.9 In patients with clear rhinitis the relative risk was 0.6 (0.38 to 0.96; NNT=7). In a study of adults, Stott and West reported a benefit for rhinitis from treatment with doxycycline, although the study does not say if the nasal discharge was clear or purulent.10 Furthermore, it is unclear why the authors of the guideline chose 10-14 days as their decision time for giving antibiotics. The data in the study by Taylor et al were collected at day eight, those from Stott et al on day seven, and those from Todd et al at five to six days. The Cochrane review on purulent rhinitis that lasts more than 10 days was first published in 2000 and hence was not available to the authors of the guideline, yet no reference is made to the articles that were part of that review and published before 1998.11

    An evidence based guideline for adults (endorsed by the Centers for Disease Control and Prevention, the American Academy of Family Physicians, and the American College of Physicians-American Society of Internal Medicine Infectious Diseases Society of America) referred to one study as evidence for the lack of benefit from antibiotics for purulent rhinitis despite the fact that this study did not analyse purulent rhinitis as an outcome. 4 12 Moreover this study, by Kaiser et al, found a benefit in terms of a composite symptom score from amoxicillin with clavulanic acid in the 20% of patients whose nasopharyngeal aspirates tested positive for Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis.12 The guideline refers to three randomised controlled trials of antibiotics versus placebo for the common cold as evidence for no benefit from antibiotics in patients with the common cold, but only one of these analysed purulent rhinitis as an outcome and found no benefit from antibiotics.13

    In addition to the above five papers that address the use of antibiotics for acute rhinitis, we are aware of one other trial by Herne, published in French.14 This paper compares two antibiotics, xibornol and tetracycline, with placebo in treating infections of the upper respiratory tract. Herne reports rhinitis as a symptom outcome at day five but does not distinguish between clear and purulent rhinitis. When the discharge was considered to be clear, and pooled with the results of the study by Taylor et al, the resulting relative risk was 0.55 (0.35 to 0.84) (using Revman 4.1 meta-analysis, fixed effects).9 When the nasal discharge was considered to be purulent and added to the studies of Taylor et al,9 De Sutter et al7, and Todd et al,8 the relative risk was 0.69 (0.56 to 0.86). The studies by Howie and Stott did not report data that could be analysed. The study by Todd is the only one with a point estimate favouring placebo (relative risk 1.23, 0.85 to1.79). When this study is removed from the analysis the relative risk for the other three studies is 0.61 (0.47 to 0.79).

    Where does this leave us? The above studies provide some evidence that antibiotics are effective for acute rhinitis, and the Cochrane review on purulent rhinitis lasting 10 days or more found a benefit from antibiotics with a number needed to treat of six.11 We do not know which patients will benefit, although they may be a subset of patients from whom bacteria could be isolated such as those identified in the study by Kaiser et al.12 The above guidelines advise against antibiotics because of their ineffectiveness, which may not be true. Instead they might have cited a modest benefit of antibiotics for a condition that is rarely life threatening while pointing to the good reasons to reduce inappropriate use of antibiotics—to constrain antibiotic resistance, side effects, and cost, and to avoid encouraging patients to increase their use of health services.

    What can doctors do in the meantime? It seems reasonable to offer patients treatment for their symptoms. Not much research has been done on this, however, and medication for symptoms can have side effects. Practitioners also need to be aware that little work has been done in children. Perhaps doctors could consider delayed prescriptions in an attempt to meet demand from patients while maintaining evidence based integrity.15 Hopefully, increasing numbers of patients will accept that infections of the upper respiratory tract are typically self limiting or, as we often tell patients, “bodies are much cleverer than doctors.”

    To be truly evidence based the new answer would be to say to patients that the benefits from antibiotics in acute purulent rhinitis may range from no benefit to a one in 10 chance that they will work. If they are prepared to wait, their purulent rhinitis is likely to get better without them needing to be exposed to antibiotics, and that after 10 days they have a one in six chance of benefit. Future guidelines should reflect this. This advice differs little from the two guidelines above but is perhaps closer to the truth and evidence.


    • Competing interests None declared


    Log in

    Log in through your institution


    * For online subscription