What is the background to this?
Careful prescribing of antibiotics may help to delay the development and spread of antibiotic resistance. As pointed out in a MeReC bulletin on the management of common infections in primary care, it is important to reserve the use of antibiotics to those at higher risk of complications and in those where a bacterial cause is likely. Although antibiotics confer little average benefit to patients with sinusitis, there may be subgroups that may benefit. A recent meta-analysis attempted to identify whether common signs, symptoms, or specific characteristics of patients with acute rhinosinusitis can be used to identify a subgroup that would be benefit from antibiotic treatment.
What does this study claim?
The study found that, on average about 15 adult patients with rhinosinusitis-like complaints would need to be treated with antibiotics for one extra patient to be cured (i.e. being free of symptoms). However, taking into account the wide confidence intervals (CIs) around the mean estimate of their effect, the number of patients that would need to be treated to gain benefit could be as low as seven, but it is also possible that no patients benefit and one in every 190 would suffer harm.
Common clinical signs were unable to identify a subgroup of patients for whom treatment was justified. Only purulent discharge in the pharynx had some prognostic value, although eight patients with this sign still needed to be treated for one patient to benefit. The authors suggest that antibiotics are not warranted, even when symptoms are present for longer than seven to ten days, which is commonly recommended as a period of watchful waiting before use of antibiotics. They go on to suggest that treatment with antibiotics is only essential if symptoms are suggestive of serious complications (e.g. high fever, periorbital swelling, erythema or intense facial pain). The authors point out that their results do not apply to children or patients with suppressed immune systems.
How does this relate to other studies?
Previous reviews of the use of antibiotics have suggested that about 15% of patients may gain benefit from the use of antibiotics (e.g. Hickner et al 2001); a number needed to treat (NNT) of about seven. The present study suggest a higher NNT (fewer patients gaining benefit), but is likely to be more representative of the situation in primary care in the UK, as it excluded patients who were diagnosed on the basis of imaging, laboratory tests or bacterial culture.
A MeReC bulletin in 2006 reviewed the use of antibiotics in acute sinusitis and pointed out the need to restrict their use to patients with systemic illness, or several severe signs and symptoms which have lasted longer than seven to 10 days, or worsened after five to seven days. It advocated the use of watchful waiting or delayed prescription for most patients, recognising that two-thirds of patients experience resolution of symptoms without antibiotic treatment.
The present meta-analysis provides good-quality evidence confirming the limited benefits to be obtained from antibiotics for the treatment of adult rhinosinusitis in primary care. Routine use is clearly inappropriate, with, on average, about 93% of patients gaining no benefit. Even in patients with purulent discharge, antibiotics appear to have no effect in about 88% of cases. The study gives reassurance that a policy of symptomatic treatment and watchful waiting (or delayed prescription) of antibiotics is appropriate for most patients with acute rhinosinusitis. Only symptoms suggestive of serious complications would seem to justify immediate antibiotic treatment.
Action – Prescribers should continue to use antibiotics sparingly for the treatment of sinusitis. For most patients providing reassurance that the symptoms will resolve without antibiotic treatment and the use of watchful waiting will be all that is necessary. Only where symptoms are suggestive of serious complications (e.g. high fever, periorbital swelling, erythema or intense facial pain) should they be prescribed immediately. Prescribers should be aware a NICE guideline for respiratory tract infections is under development and is anticipated to be published in July this year.
Study details – Individual patients’ data from 2547 adults in nine double-blind randomised trials were checked and re-analysed for consistency. Analysis was by intention to treat. The outcome of interest was the proportion of patients cured at the time the primary outcome of the trial was assessed. ‘Cure’ was most often defined as being free of symptoms but this outcome did vary between studies. The overall effect of antibiotic treatment and the prognostic value of common signs and symptoms were assessed by calculating the NNT with antibiotics to cure one additional patient. In a meta-analysis of aggregated data, the estimated odds ratio (OR) for the overall treatment effect of antibiotics relative to placebo was 1.35 (95%CI 1.15 to 1.59). In an analysis of individual patients’ data, the estimated OR for the overall treatment effect was 1.37 (95% CI 1.13 to 1.66). The mean NNT for 10,000 simulated new patients was 15 (95%CI NNT[benefit] 7 to NNT[harm] 190), which implied that 15 patients had to be given antibiotics before one additional patient was cured. Patients who were older, reported symptoms for longer, or reported more severe symptoms also took longer to cure but were no more likely to benefit from antibiotics than other patients.