Intended for healthcare professionals


Delayed prescriptions

BMJ 2003; 327 doi: (Published 11 December 2003) Cite this as: BMJ 2003;327:1361
  1. Bruce Arroll, associate professor (b.arroll{at},
  2. Tim Kenealy, doctoral fellow,
  3. Felicity Goodyear-Smith, senior lecturer,
  4. Ngaire Kerse, senior lecturer
  1. Department of General Practice and Primary Health Care Faculty of Medical and Health Sciences University of Auckland PB 92019 Auckland New Zealand

    Can reduce antibiotic use in acute respiratory infections

    Although a reduction has occurred in the use of antibiotics for upper respiratory tract infections, international evidence indicates that they continue to be used for these conditions.1 This is in spite of Cochrane reviews indicating minimal or no benefit from antibiotics for sore throat, acute bronchitis, the common cold, and otitis media. This situation of potentially inappropriate prescribing prompted one commentator to suggest the use of delayed prescriptions (also known as “back-pocket,” “back-up,” or “as needed” prescriptions).2 These are prescriptions written with a proviso that they not be used immediately and only if symptoms do not improve.

    The first randomised trial of delayed prescriptions for respiratory symptoms was undertaken by Little et al (1997), who gave antibiotics, with the prescription to be filled immediately or after three days, or no antibiotics for acute sore throat.3 The immediate group filled 99% of the antibiotic prescriptions whereas the delayed group filled only 31% with no apparent serious harms. In the group not given any antibiotics 13% ended up filling an antibiotic prescription after a return visit to the doctor. Critics of delayed prescriptions say that the strategy increases the use of antibiotics by comparing the 13% with the 31%. This is not how we see the use of delayed prescriptions. Their use should be restricted to those patients who request antibiotics or whom their doctor thinks they want an antibiotic yet does not think one is immediately indicated. However, only one of the randomised trials has examined such a specific group.4

    Five controlled trials of delayed prescriptions have been published, conducted in patients with otitis media,5 6 sore throat,3 cough lasting seven days,7 and the common cold.4 In three trials the patients in the delayed prescription arm had more symptoms during the trial, which implies that patients are willing to tolerate some symptoms to avoid antibiotics.3 6 7 Ironically the study with the highest reduction in relative risk (75%) was the study in children with otitis media, in spite of the children having more symptoms.6 We speculate that parents may be more concerned about avoiding antibiotics in their children than in themselves—a view supported by a qualitative study of patients with sore throat.8

    The largest reductions in antibiotic consumption occurred in the three studies which required patients to return to the surgery to collect the prescriptions.3 6 7 Although most of the studies had pick-up suggestions of three days or less, the study on acute cough, which suggested waiting seven days, still produced a reduction in relative risk of 55%. An additional benefit of delayed prescriptions may be a reduction in repeat visits, at least for sore throat.9 The reduction in usage of antibiotics for infections of the upper respiratory tract through using delayed prescriptions is as effective as, and in many cases more effective than, educational projects.1011 However, no studies have directly compared delayed prescriptions with educational projects.

    Some interesting insights have been obtained from qualitative work in patients and doctors on delayed prescriptions.12 Not all general practitioners endorsed the use of delayed prescriptions—some had concerns that they may be missing or masking serious illness, with concomitant medicolegal issues. Some worried that their patients may consider them incompetent. General practitioners thought that the positive aspects of delayed prescriptions included avoiding side effects, reducing the drug bill, educating patients, and involving patients in decision making. Although reducing antibiotic resistance was a major issue for general practitioners it was not a concern for patients.12 More education for patients around this issue may be warranted, and we suggest that the delayed prescription be used as a tool to help improve patients' knowledge about infectious disease and awareness of the need for monitoring their own progress.

    Research is also needed on other methods of providing a barrier other than a patient's return to the practice if he or she is not getting better. Such barriers could be asking patients to wait seven days rather than three and post-dating prescriptions. If delayed prescriptions are to become routine then surgeries will need to have systems to hold the prescription at the front desk and to allow patients easy access for reassessment if concerned about their symptoms. They may also need to consider following up patients with delayed prescriptions to monitor adverse events.

    In the qualitative research on delayed prescriptions several general practitioners no longer used this strategy, once their patients had become “trained” not to expect antibiotics. As prescribing becomes more rational the need for delayed prescriptions for respiratory tract infections may in time become redundant.


    • Competing interests None declared


    View Abstract