Abstract
Introduction: LRTI is the commonest acute presentation managed in primary care and still a major driver of antibiotic prescribing. Systematic reviews of placebo controlled studies are small (<1000).
Aim: To determine the effectiveness of amoxicillin for lower respiratory tract infection.
Methods: 2054 patients presenting with uncomplicated acute cough (<4 weeks) as the main symptom were randomised to amoxicillin 1g three times a day or placebo for 7 days. Patients completed validated symptom diaries for symptom severity (7 point scale) and duration. Notes were reviewed for repeat consultations
Results: 593 of trial population (28%) were aged 60+, and symptom severity documented and duration were documented in 87% of patients. There was no significant difference in symptoms severity in the first 4 days after seeing the doctor (placebo mean 1.69, antibiotic 1.62; difference -0.07 (-0.18 to 0.06)), and no significant difference in the proportion with moderately bad or worse symptoms at 7 days (47% vs 40% respectively, p=0.07 NNT 14). Among the subgroup of patients aged 60 or over there was no evidence of selective benefit. 5% more patients in the antibiotic group compared with the placebo group developed nausea, rash or diarrhoea (NNH 20).
Conclusion: Antibiotics are very unlikely to provide meaningful symptomatic benefit in LRTI for most patients, and any benefit is likely to be similar to the magnitude of harm.
- © 2011 ERS
