Commentary: Controversies in NICE guidance on antibiotic prescribing for self limiting respiratory tract infections in primary care

BMJ 2008; 337 doi: (Published 23 July 2008) Cite this as: BMJ 2008;337:a656
  1. Christopher C Butler, professor of primary care medicine, head of Department of Primary Care and Public Health1,
  2. Nicholas Francis, Medical Research Council fellow, Department of Primary Care and Public Health1
  1. 1Cardiff University, Cardiff CF14 4XN
  1. Correspondence: C C Butler butlercc{at}

    The key message in the new NICE guidance on antibiotic prescribing for respiratory tract infections is clear—the comprehensive reviews of the existing evidence base confirm that antibiotic treatment does not bring meaningful benefits for most patients.1 For most clinicians this will not be a new message, and indeed most will agree with the need for change and will already have made some changes to the way they manage these infections. But old habits die hard, and we in the United Kingdom still (unjustifiably) prescribe almost double the amount of antibiotics prescribed in, for example, Dutch primary care.2 Further progress is hampered by two main clinical challenges: how to identify accurately the few patients who are likely to develop serious illness or whose symptoms could be meaningfully ameliorated by prompt antibiotic treatment, and how to implement efficiently an evidence based prescribing decision while responding to patients’ concerns and expectations in the time constraints of real life general practice.3

    The guideline makes several evidence based recommendations about some groups that should be considered for immediate antibiotic treatment and/or further investigation, such as those who are systemically unwell. However, new, large studies are still needed to better identify specific features that alone or in combination indicate likely benefit from early antibiotic treatment and further testing. These will rely on national and international efforts to achieve sufficient power, as in the observational DESCARTE study focusing on sore throats ( and the observational, diagnostic, prognostic, and randomised studies in the GRACE Network of Excellence focusing on managing lower respiratory tract infections in the community ( Research should be ongoing as the complex interaction between social determinants of health and changing virulence of infecting organisms evolves.4

    Point of care rapid testing was beyond the guideline’s scope, which is a pity as these tests are widely used in primary care in some European countries to guide antibiotic treatment, and their potential role in practice is worth considering.

    These consultations are often a delicate dance around the use of antibiotics. The clinical departure point of the guidelines is eliciting patients’ expectations and concerns, and they urge clinicians to discuss the rationale for prescribing choices and and share evidence about the likely duration of symptoms, with and without antibiotic treatment. Such communication strategies have the potential to greatly improve the management of these infections, both by enhancing patient satisfaction and reducing antibiotic prescribing. However, the impact of advice to use these strategies without additional training in underpinning communication skills is unknown.

    The guideline identifies a “delayed prescription” strategy as one of the three main management options. This strategy, which has been shown to reduce antibiotic consumption and future consulting while maintaining satisfaction,5 will be a valuable tool for some clinicians. Delayed prescribing is appealing because it avoids having to say “no,” validates patients’ concerns, and seems to give patients control. However, its exact role in a context of enhanced prognostic and communication skills remains unclear. Studies are under way to examine whether a shared decision making approach (with general practitioners skilfully portraying risk of harms and benefits in understandable terms) also reduces antibiotic prescribing without reducing satisfaction or increasing time to recovery.6 7 Patients generally don’t want to take a drug when they recognise that it has a slim chance of helping them and may cause them harm, and clinicians should not view delayed prescribing as a substitute for high quality risk communication and shared decision making.

    Outcomes from delayed prescribing will need close review as the threshold for consulting goes up (those who are less sick are consulting less often).8 Applicability of delayed prescribing to contexts of high morbidity from infectious diseases and high social deprivation is also unknown.

    In the full guideline, the authors state that a delayed prescribing strategy may have advantages over a “no prescribing” strategy as it provides a “safety net.” A delayed prescribing strategy does not seem to be associated with any increase in complication rates.9 However, no evidence exists that delayed prescribing is any safer than not prescribing.10 The key to enhanced safety in either situation is empowering patients (again through enhanced communication skills and prognostic understanding) to make informed decisions about when to take further action (whether that means returning to consult their doctor or starting to use a delayed prescription).

    Most clinicians know what they should do but don’t always do it, largely because of nagging prognostic uncertainty and wanting to maintain relationships efficiently in time pressured consultations. These guidelines provide the most up to date, integrated review of what should be done and provide some useful advice on how to do it. They also serve to highlight the need for ongoing research that integrates biomedical, clinical, and communication sciences.


    Cite this as: BMJ 2008;337:a656


    • doi: 10.1136/bmj.a437
    • Contributors: Both authors contributed equally to the commentary. CCB is the guarantor.

    • Competing interests: None declared.


    View Abstract

    Sign in

    Log in through your institution