1. Sinusitis is one of the top 10 diagnoses given by GPs, who on average see about 50 cases a year. Yet there’s widespread confusion on how to treat this condition. The new thinking can be summed up in two phrases: antibiotics are out, steroids are in.

2. Most cases of acute sinusitis are self-limiting. In large placebo-controlled trials, two-thirds of placebo patients saw symptoms vanish within 2 weeks.

3. In 2006, American doctors gave antibiotics to over 85% of sinusitus sufferers, at a cost of $2.4 billion. But recent research suggests antibiotics have little effect, even in the roughly 50% of cases where bacteria are clearly involved. Some ENT specialists, however, warn against overreacting to what remains a limited body of evidence.

4. The weight of evidence says this: the more serious the symptoms, the more likely antibiotics are to help. For milder cases, nasal topical steroids are often a better choice.

5. Purulence is key in identifying bacterial causes. The best clinical criteria were designed by Berg and Carenfelt, who identified 3 symptoms and 1 sign of bacterial infection: purulent rhinorrhea with unilateral predominance; local pain with unilateral predominance; bilateral purulent rhinorrhea; and pus in the nasal cavity. Spot 2 or more of these, and your diagnosis of acute bacterial sinusitis will rest on firm ground.

6. X-ray of the sinuses is obsolete and useless. Ultrasound is little better. Nasal endoscopy is preferable, but save it for refractory cases and red flags.

7. The main red flag for referral is unilateral nasal obstruction in an elderly patient. Refer these whether or not there’s pain or bloodstained discharge. Sinus tumours are quite rare — but they are also rarely benign. In the young with unilateral obstruction, first suspect a damaged septum.

8. Common pathogens behind bacterial sinusitis are S. pneumoniae, H. influenzae and M. catarrhalis. All have resistant strains, but penicillin is usually just as effective as newer antibiotics.

9. As for the best steroids, nasal formulations cause few adverse effects compared to systemic drugs. In children, it’s worth using the newer steroids, mometasone furoate or fluticasone, since budesonide has a demonstrably greater effect on growth.

10. Dig into the research and you’ll find little to back up the current trend towards steroids. Their stock has risen by default as that of antibiotics has fallen. But in trials comparing steroids and antibiotics, both struggled to outperform placebo, even in combination. Typically, about 67% of placebo patients would be symptom-free after 2 weeks. Steroids and antibiotics might push that to 71%. Like the common cold, it seems, sinusitis doesn’t let doctors decide its schedule.