Meta-analysis: antibiotics for acute sinusitis

Antibiotic treatment has a small therapeutic benefit over placebo in acute sinusitis with an increased risk of adverse effects, according to a meta-analysis.

The authors note that acute sinusitis is one of the more frequent reasons for prescription of an antibiotic in primary care, but that there is still controversy over whether they are effective. This review aimed to assess how safe and effective antibiotics were for acute sinusitis. The authors carried out a literature search primarily using Medline via PubMed, supplemented with Scopus. Eligible studies were randomised, double-blind placebo-controlled trials of any antibiotic for acute sinusitis. For each included study, data on its design and outcomes were extracted for meta-analysis; the primary outcome was cure or improvement in clinically assessable patients, between 7 and 15 days from the start of treatment.

Initial searches yielded 194 potentially relevant articles, of which 29 were selected on the basis of title and abstract; 16 of these, plus one further study identified through an updated search, fitted the eligibility criteria and were included in the meta-analysis. All 17 were high quality (Jadad score 4 or 5), and most were carried out in Europe or the US. Total number of participants was 3,291, including 376 children. The antibiotic most often used was amoxicillin (10 studies) followed by phenoxymethylpenicillin (4 studies) and co-amoxiclav (3 studies). Five studies had more than one antibiotic treatment group.

There were 16 studies (n=2,648) reporting the primary outcome of cure or improvement, and in these, more patients were cured or improved in the antibiotic group (OR 1.64; 95% CI, 1.35 to 2.00).

Twelve studies (n=1,813) specifically reported cure rates, and in these more patients were cured in the antibiotic groups compared to placebo (odds ratio 1.82; 95% CI, 1.34 to 2.46).

Twelve studies (n=1,963) reported adverse effects: in these, patients in the antibiotic group were more likely to report adverse effects (OR 1.87; 95% CI, 1.21 to 2.90). Study withdrawals because of adverse events did not differ between compared treatments, and neither did disease complications, or disease recurrence.

The authors conclude that antibiotic treatment has a small therapeutic benefit in acute sinusitis, but that this is matched with a corresponding increase in adverse events. There was no indication in the meta-analysis of any effect on disease complications or recurrence. Accordingly, they suggest that antibiotics should not be used routinely for these patients but should be reserved for those with a high probability of bacterial disease. To support this, research is needed to improve identification of this group.

Lancet Infect Dis 2008; 8: 543-52 (link to abstract)