Antibiotics for Adults with Clinically Diagnosed Acute Rhinosinusitis: A Meta-analysis of Individual Patient Data

This meta-analysis pooled nine randomized trials identified from the Cochrane, Medline and Embase databases to assess the prognostic value of common signs and symptoms of rhinosinusitis and the overall effect of antibiotic treatment of patients with this condition. Trials were selected from all those studies randomly assigning placebo vs. antibiotics in adult patients with rhinosinusitis complaints. The suitable trials used in the meta-analysis were selected by investigators based on the identification of patients for treatment based on clinical symptoms and physical findings of rhinosinusitis alone and not based on any diagnostic studies. Despite differences in the methods by which follow-up was obtained (e.g., patient diaries, telephone interview), the antibiotics used, and the clinical inclusion criteria, the selected trials maintained an intention-to-treat analysis and logical treatment of missing data. The intention-to-treat population from the nine trials consisted of 2640 patients where the primary outcome was clinical cure from the illness as defined by the authors of the original publications. In the aggregated data, the estimated odds ratio for overall treatment effect of antibiotics was 1.35 with a number needed to treat of 15 before one additional patient benefited from treatment with antibiotics. The treatment effect was almost independent of underlying risk, with an overall cure rate of 64% achieved at 14 days in patients without antibiotic treatment. Authors concluded that the majority of signs and symptoms have little prognostic value, with only purulent discharge in the pharynx and fever (odds ratios of 1.6 and 1.28, respectively) being the most likely to benefit from treatment with antibiotics. Authors also looked at length of symptoms before enrollment and found no difference in cure rates when patients presented 6 days or more from the onset of symptoms. Patients reporting a longer duration of symptoms took longer to achieve cure but received no benefit from the addition of antibiotics to their treatment regimen. The conclusions of this study were that antibiotics offer little benefit to patients with acute rhinosinusitis complaints and that particular signs and symptoms do not help distinguish patients whom are most likely to benefit from those who are not. Although many guidelines suggest such treatment after 7–10 days of symptoms, the authors suggest that antibiotics are not justified in this situation unless signs of a more serious complication are present.